# PLANT-MICROBE-INSECT INTERACTIONS IN ECOSYSTEM MANAGEMENT AND AGRICULTURAL PRAXIS

EDITED BY : Gero Benckiser, Krishnamurthy Kumar, Anton Hartmann and Bernd Honermeier PUBLISHED IN : Frontiers in Plant Science and Frontiers in Microbiology

#### Frontiers Copyright Statement

© Copyright 2007-2019 Frontiers Media SA. All rights reserved. All content included on this site, such as text, graphics, logos, button icons, images, video/audio clips, downloads, data compilations and software, is the property of or is licensed to Frontiers Media SA ("Frontiers") or its licensees and/or subcontractors. The copyright in the text of individual articles is the property of their respective authors, subject to a license granted to Frontiers.

The compilation of articles constituting this e-book, wherever published, as well as the compilation of all other content on this site, is the exclusive property of Frontiers. For the conditions for downloading and copying of e-books from Frontiers' website, please see the Terms for Website Use. If purchasing Frontiers e-books from other websites or sources, the conditions of the website concerned apply.

Images and graphics not forming part of user-contributed materials may not be downloaded or copied without permission.

Individual articles may be downloaded and reproduced in accordance with the principles of the CC-BY licence subject to any copyright or other notices. They may not be re-sold as an e-book.

As author or other contributor you grant a CC-BY licence to others to reproduce your articles, including any graphics and third-party materials supplied by you, in accordance with the Conditions for Website Use and subject to any copyright notices which you include in connection with your articles and materials.

All copyright, and all rights therein, are protected by national and international copyright laws.

The above represents a summary only. For the full conditions see the Conditions for Authors and the Conditions for Website Use. ISSN 1664-8714 ISBN 978-2-88945-819-6 DOI 10.3389/978-2-88945-819-6

### About Frontiers

Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

### Frontiers Journal Series

The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

### Dedication to Quality

Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.

Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view. By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

### What are Frontiers Research Topics?

Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: researchtopics@frontiersin.org

# PLANT-MICROBE-INSECT INTERACTIONS IN ECOSYSTEM MANAGEMENT AND AGRICULTURAL PRAXIS

Topic Editors:

Gero Benckiser, University of Giessen, German Krishnamurthy Kumar, Tamil Nadu Agricultural University, India Anton Hartmann, Helmholtz Center Munich - German Research Center for Environmental Health, Germany Bernd Honermeier, University of Giessen, Germany

Plant residue with oribatide excrements is from Babel, U. (1975) Micromorphology of Soil organic matter In soil components. In J.E Gieseking (Ed.) *Soil Components: Volume 1: Organic components*. Berlin: Springer-Verlag.

Cover Image: Ant milking honeydew from an aphid. Image by Thomas Griessler.

Nature's high biomass productivity is based on biological N2 fixation (BNF) and biodiversity (Benckiser, 1997; Benckiser and Schnell, 2007). Although N2 makes up almost 80% of the atmosphere's volume living organisms need it in only small quantities, presumably due to the paucity of natural ways of transforming this recalcitrant dinitrogen into reactive compounds. N shortage is commonly the most important limiting factor in crop production. The synthesis of ammonium from nitrogen and hydrogen, the Haber–Bosch (H-B) process, invented more than 100 years ago, became the holy grail of synthetic inorganic chemistry and removed the most ubiquitous limit on crop yields. H-B opened the way for the development and adoption of high-yielding cultivars, for monoculturing by organic and precision farming. With N over fertilization and pesticide application monoculturing farmers could approach Nature's high biomass productivity by causing side effects the scientific world is investigating. This eBook presents the complexity the scientific world is facing in in understanding the soil-microbe-plant-animal cooperation, the millions of taxonomically, phylogenetically, and metabolically diverse above-below-ground species, involved in shaping the ever-changing biogeochemical process patterns being of great significance for food production networks and yield stability. Because ecosystem management and agricultural praxis are still largely conducted in isolation, the aim of this Frontiers' eBook is to gather and interconnect plant-microbe-insect interaction research of various disciplines, studied with a broad spectrum of modern physical-chemical, biochemical, and molecular biological, agronomical techniques. The goal of this Research Topic was to gain a better understanding of microbe-plantinsect compositions, functioning, interactions, health, fitness, and productivity. 42 Research Topic articles give insights into these topics, and are outlined and given a brief description below.

Kumar et al., Imam et al., Saxena et al.; Wei et al.: Modelling of biological networks and system biology approaches, Indian rice landrace diversity, and anthracnose management in chilli and a wide range of cruciferous crops.

Kamarudin et al.; Hashem et al.; Wei et al.; Guo et al.; Sharma et al.: *Henderosonia toruloidea* colonization of palm oil seedlings improves the thiamine biosynthesis, the cofactor vitamin B1, the growth as two, tripartite mutualistic symbiosis between mycorrhizal fungi, *Rhizobium radiobacter*, *Piriformospora* the cereal crop, or halotolerant, endophytic, plant growth promoting rhizobacteria the adaptation of peanuts on salt stress.

Han et al.; Egamberdieva et al.: The role of cooperation in rising fitness demonstrates also a favoured barley growth through the endophytic *N*-acyl homoserine lactone (AHL) producer *Acidovorax radicis* N35 and the medicinal plant *Hypericum perforatum* exhibits a remarkable activity against bacterial and fungal pathogens.

Wu et al.; Felestrino et al.: In monoculture *Pseudostellariae heterophylla* fields with significantly increasing *Fusarium oxysporum* wilt *Burkholderia* spp suppresses the fungus and specifically useable reservoir for plant growth promoting bacteria (PGPB) is the Brazilian Iron Quadrangle (IQ) soil and the from this soil enriched PGPB favour the reforestation of antropized soil.

Fatima and Anjum; Wu et al.: PGPB as the *Pseudomonas aeruginosa* strain PM12 induce systemic resistance in tomato against Fusarium and the soil rhizosphere microbial release of allelochemically acting phenolic acids mediates in rhizosphere soil of continuously monocultured *Radix pseudostellariae* medicinal plant fields a microflora shift between pathogenic microorganisms as *Talaromyces helicus* M. (KU355274) and *Kosakonia sacchari* W. (KU324465) and antagonistic bacterium, *Bacillus pumilus* that leads to increased replanting disease incidence.

Xia et al.; Wang et al.: In poplar growth, tissue development, and defence regulation plays the core motif TIF[F/Y]XG forming TIFY domain a role.

Anwar, Ali and Sajid; Manhas and Kaur; Kaur et al.; Nie et al.: Fungus-like growing actinomycetes as indole acetic acid (IAA) producing *Streptomyces nobilis, kunmingenesis, enissocaesilis*, and *hydrogenans*, of which *S*. *hydrogenans* forms a non-cytotoxic, non-mutagenic10-(2,2-dimethyl-cyclohexyl)-6,9-dihydroxy-4, 9-dimethyl-dec-2-enoic acid methyl ester showing significant inhibitory activity against fungal phytopathogens and a *Bacillus cereus* AR156 suppresses *Pseudomonas syringae* pv *tomato* DC3000 disease incidence in *Arabidopsis* plants.

Singh et al.; Alamgir et al.; Janahiraman et al.: Thyme oil (THY) in combination with conventional bactericides reduces the activity of common virulent bacterial rice plant pathogen as the biofilm forming *Xanthomonas oryzae* pv. *oryzae* by lowering bactericide dose and the histidine derived, sulphur-containing, non-proteinogenic amino acid ergothioneine, isolated from a moss, acts anti-oxidative by adding to the phyllospheric fitness, whereas methylotrophic, pink pigmented bacteria as *Delftia lacustris*, *Bacillus subtilis*, or *Bacillus cereus* exhibit direct antagonistic reactions against fungal pathogens.

Dutta et al.; Krithika and Balachandar; Shakeel et al.; Tamilselvi et al.: The list of plant production favouring rhizobacteria also includes isolates from Indian Assam tea estates as *Enterobacter lignolyticus*, *Bacillus pseudomycoides*, *Burkholderia* sp., or *Pseudomonas aeruginosa*, the calcite dissolving *Brevibacterium* sp. SOTI06, or the zinc (Zn) solubilizing bacteria *Bacillus* sp. and *Enterobacter cloacae*.

Islam et al.; Srivastava et al.: Such bacteria help translocating Zn toward grains and *Pseudomonas stutzeri*, *Bacillus subtilis*, *Stenotrophomonas maltophilia*, *Bacillus amyloliquefaciens* biocontrol asides economically important rice pathogens as *Pyricularia oryzae*, *Fusarium moniliforme* and *Rhizoctonia solani*.

Tadych et al.; Teshome et al.: More rot-resistant cranberry genotypes show early in fruit development higher levels of benzoic acid and more gradual decline in quinic acid levels and intercropping exhibits success in minimizing pea weevil larvae (*Bruchus pisorum* L.), entering developing seeds of *P. sativum*.

Harun-Or-Rashid et al.: Neoplastic genotypes as found in *P. sativum* seeds are also present in other crops such as sorghum and maize, suffering under insect-plantbacteria interactions, e.g. under a most destructive, plant sap sucking and plant viruses transmitting insect pest, under *Arabidopsis* leaves puncturing green peach aphid (GPA), *Myzus persicae*, against which *Bacillus velezensis* YC7010 can induce systemic resistance.

Zhang et al.: The nematode *Heterodera avenae*, another important field crop pest worldwide *Trichoderma longibrachiatum* controls in abundance by completely surrounding *H. avenae* eggs with dense mycelia and by lysing the egg content at late egg stage.

Prasad et al.; Lin et al.; Tan et al.; Trivedi et al.: Sap-sucking psyllids (*Diaphorina citri*) and citrus plants infecting *Candidatus* (*Ca*.) *Liberibacter asiaticus* modify plant volatile cues in a way that the predator, ladybird beetle *P. japonica*, cannot find the host plant to reduce the *D. citri* population and the tomato – whitefly – tomato yellow leaf curl virus (TYLCV) system unveils a linking of cellular levels with interdisciplinary scales.

Islam et al.; Li et al.: Bensulfuron-methyl, widely used in paddy soil for weed control, affects the metabolism of important phytohormones such as JA and SA and in sensitive plants (*Nicotiana tabacum*) the immune response against whitefly (*Bemisia tabaci*), aphids (*Myzus persicae*) and viruses (*Tobacco mosaic virus*) infections. A high level of nitrogen reduces the release of plant volatiles and more *Bemisia tabaci*  (Hemiptera: Aleyrodidae) are attracted.

Santamaria et al.: In this context is worthwhile to get interest in immunity gene (MATI) that is involved in mite attack and defence triggering by modulating the levels of photosynthetic pigments.

Qin et al.: Also of interest and that may be little studied: native grass of the Inner Mongolian steppe, *Achnatherum sibiricum*. The research suggesting that inherent endophyte infection significantly decreased the concentration of soluble sugar and amino acids while significantly increased the concentration of total phenolic content, and these metabolites may contribute to herbivore (*Locusta migratoria*) resistance of the host, while decreasing the tolerance of the host grass by mechanisms apart from endophyte-conferred alkaloid defence.

I hope this brief overview motivates to dive into this eBook on plant-microbe-insect interactions, which includes methods and suggestions for ecosystem management and agricultural praxis.

### References


Citation: Benckiser, G., Kumar, K., Hartmann, A., Honermeier, B., eds. (2019). Plant-Microbe-Insect Interactions in Ecosystem Management and Agricultural Praxis. Lausanne: Frontiers Media. doi: 10.3389/978-2-88945-819-6

# Table of Contents

*10 Editorial: Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent Research*

Gero Benckiser, Anton Hartmann, Krishnamurthy Kumar and Bernd Honermeier

*13 Recent Developments in Systems Biology and Metabolic Engineering of Plant–Microbe Interactions*

Vishal Kumar, Mehak Baweja, Puneet K. Singh and Pratyoosh Shukla

*25 Allele Mining and Selective Patterns of* Pi9 *Gene in a Set of Rice Landraces From India*

Jahangir Imam, Nimai P. Mandal, Mukund Variar and Pratyoosh Shukla


Wei Wei, Ying Xiong, Wenjun Zhu, Nancong Wang, Guogen Yang and Fang Peng


Huijuan Guo, Stefanie P. Glaeser, Ibrahim Alabid, Jafargholi Imani, Hossein Haghighi, Peter Kämpfer and Karl-Heinz Kogel

*110 Halotolerant Rhizobacteria Promote Growth and Enhance Salinity Tolerance in Peanut*

Sandeep Sharma, Jayant Kulkarni and Bhavanath Jha


Dilfuza Egamberdieva, Stephan Wirth, Undine Behrendt, Parvaiz Ahmad and Gabriele Berg

*146 Insights Into the Regulation of Rhizosphere Bacterial Communities by Application of Bio-organic Fertilizer in* Pseudostellaria heterophylla *Monoculture Regime*

Linkun Wu, Jun Chen, Hongmiao Wu, Xianjin Qin, Juanying Wang, Yanhong Wu, Muhammad U. Khan, Sheng Lin, Zhigang Xiao, Xiaomian Luo, Zhongyi Zhang and Wenxiong Lin

*160 Plant Growth Promoting Bacteria Associated With* Langsdorffia hypogaea*-Rhizosphere-Host Biological Interface: A Neglected Model of Bacterial Prospection*

Érica B. Felestrino, Iara F. Santiago, Luana da Silva Freitas, Luiz H. Rosa, Sérvio P. Ribeiro and Leandro M. Moreira

*175 Identification of a Potential ISR Determinant From* Pseudomonas aeruginosa *PM12 Against Fusarium Wilt in Tomato*

Sabin Fatima and Tehmina Anjum

*189 Insights Into the Mechanism of Proliferation on the Special Microbes Mediated by Phenolic Acids in the* Radix pseudostellariae *Rhizosphere Under Continuous Monoculture Regimes*

Hongmiao Wu, Junjian Xu, Juanying Wang, Xianjin Qin, Linkun Wu, Zhicheng Li, Sheng Lin, Weiwei Lin, Quan Zhu, Muhammad U. Khan and Wenxiong Lin

*204 Identification of TIFY Family Genes and Analysis of Their Expression Profiles in Response to Phytohormone Treatments and* Melampsora larici-populina *Infection in Poplar*

Wenxiu Xia, Hongyan Yu, Pei Cao, Jie Luo and Nian Wang

*215 Mixed Phenolic Acids Mediated Proliferation of Pathogens* Talaromyces helicus *and* Kosakonia sacchari *in Continuously Monocultured* Radix pseudostellariae *Rhizosphere Soil*

Hongmiao Wu, Linkun Wu, Juanying Wang, Quan Zhu, Sheng Lin, Jiahui Xu, Cailiang Zheng, Jun Chen, Xianjin Qin, Changxun Fang, Zhixing Zhang, Saadia Azeem and Wenxiong Lin


Rajesh K. Manhas and Talwinder Kaur


Akanksha Singh, Rupali Gupta, Sudeep Tandon and Rakesh Pandey


Abel Teshome, Tomas Bryngelsson, Esayas Mendesil, Salla Marttila and Mulatu Geleta


Samiksha Prasad, Jin Xu, Yunzeng Zhang and Nian Wang


Xiao L. Tan, Ju L. Chen, Giovanni Benelli, Nicolas Desneux, Xue Q. Yang, Tong X. Liu and Feng Ge

*468 Harnessing Host-Vector Microbiome for Sustainable Plant Disease Management of Phloem-Limited Bacteria*

Pankaj Trivedi, Chanda Trivedi, Jasmine Grinyer, Ian C. Anderson and Brajesh K. Singh


Shaikhul Islam, Abdul M. Akanda, Ananya Prova, Md. T. Islam and Md. M. Hossain

*502 MATI, a Novel Protein Involved in the Regulation of Herbivore-Associated Signaling Pathways*

M. Estrella Santamaría, Manuel Martinez, Ana Arnaiz, Félix Ortego, Vojislava Grbic and Isabel Diaz

*520 Effect of Endophyte Infection and Clipping Treatment on Resistance and Tolerance of* Achnatherum sibiricum

Junhua Qin, Yuan Gao, Hui Liu, Yong Zhou, Anzhi Ren and Yubao Gao

# Editorial: Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent Research

Gero Benckiser <sup>1</sup> \*, Anton Hartmann<sup>2</sup> , Krishnamurthy Kumar <sup>3</sup> and Bernd Honermeier <sup>4</sup>

<sup>1</sup> Applied Microbiology, Justus Liebig Universität Gießen, Giessen, Germany, <sup>2</sup> Deutsches Forschungszentrum für Gesundheit und Umwelt, Helmholtz Zentrum München, Munich, Germany, <sup>3</sup> Department of Agricultural Microbiology, Tamil Nadu Agricultural University, Coimbatore, India, <sup>4</sup> Department of Agronomy and Plant Breeding, Justus Liebig Universität Gießen, Giessen, Germany

Keywords: tripartite sebacinalean symbiosis, MATI regulation, virulence reduction, antimicrobial activity, nematode subpressing

### **Editorial on the Research Topic**

### **Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent Research**

The plethora of taxonomically, phylogenetically, and metabolically diverse soil- microbes, plants and animals is responsible for the dynamic observed in ecosystems. Especially the above- and below-ground species flexibility that drives biogeochemical processes is not well understood but is of great significance for food production networks and yield stability (Benckiser, 1997; Benckiser and Schnell, 2007)

Ecosystem management and agricultural praxis have until now been largely conducted in isolation but the recent interconnection of inextricably various disciplines has now given rise to a more holistic approach. This Frontiers' e-book on "Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent for Research" presents 42 contributions covering a broad spectrum of modern physico-chemical, biochemical, molecular biology and agronomy techniques with the aim to provide perspectives on how a better productivity, health and fitness could be achieved. The focus lies in the long term adapted composition and functioning of microbe-insect interactions, which should be brought in line with agricultural productivities.

In modeling of biological networks and system biology approaches Kumar et al. are introducing and giving insights into Indian rice landrace diversity Imam et al. as well as Saxena et al. and Wei et al. on anthracnose management in chilli and wide range of cruciferous crops. The cofactor vitamin B1 plays an important role in metabolic reactions of all living organisms and it is not surprising that thiamine biosynthesis of endophytic Henderosonia toruloidea colonizing palm oil seedlings improves their growth (Kamarudin et al.). In reports on two- or tripartite mutualistic symbioses Hashem et al. address the symbiosis between arbuscular mycorrhizal fungi (AMF) and "Formosa" azalea, Azalea gerrardii, Wei et al. describe how the mycorrhizal fungus Oidiodendron maius positively influences the growth of Rhododendron fortunei Lindl., Guo et al. describe fitness rise in a cooperation of Rhizobium radiobacter, Piriformospora and cereal crops, and Sharma et al. found salt stress adaptation of peanuts in the presence of halotolerant, endophytic, plant growth promoting rhizobacteria (PGPR). The endophyte Acidovorax radicis N35, a major producer of N-acyl homoserine lactone (AHL) type, favors growth of barley (Han et al.), the medicinal plant Hypericum perforatum shows a remarkable activity against bacterial and fungal pathogens (Egamberdieva et al.), and bacteria identified by pyrosequencing combined with qPCR effectively suppress Fusarium wilt in the Chinese medicinal plant Pseudostellariae heterophylla (Wu et al.).

Edited and reviewed by: Víctor Flors, Universitat Jaume I, Spain

\*Correspondence: Gero Benckiser gero.benckiser@umwelt.uni-giessen.de

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 05 April 2018 Accepted: 11 June 2018 Published: 28 June 2018

#### Citation:

Benckiser G, Hartmann A, Kumar K and Honermeier B (2018) Editorial: Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent Research. Front. Plant Sci. 9:931. doi: 10.3389/fpls.2018.00931

A selective pressure niche to plant growth promoting bacteria (PGPB) evolution is the Brazilian Iron Quadrangle (IQ) soil (Felestrino et al.). IQ soil is a specific reservoir for PGPB species useable in reforestation of antropized soil and one of those could be Pseudomonas aeruginosa PM12, which induces systemic resistance in tomato against Fusarium wilt (Fatima and Anjum). Phenolic, allelochemically acting acids apparently mediate shifts in soil rhizosphere microbial flora, which inter alia ameliorate tomato Fusarium wilt resistance in mono-cropping systems of the medicinal plant Radix pseudostellariae (Wu et al.).

The TIFY domain of poplar species contains approximately 36 conserved amino acids (Xia et al.). The forming core motif TIF[F/Y]XG plays a role in poplar growth, tissue development, and defense regulation. Fungus-like growing rhizosphere actinomycetes are known to promote growth and yield enhancement in wheat crops through the plant growth hormone indole acetic acid (IAA) and among rhizospheric IAA producers the most active and good candidates for developing biofertilizers are Streptomyces nobilis, Streptomyces kunmingenesis, and Streptomyces enissocaesilis (Anwar et al.). A candidate for agroactive compound development is also Streptomyces hydrogenans, (Manhas and Kaur). This non-cytotoxic, non-mutagenic, 10-(2,2-dimethyl-cyclohexyl)-6,9-dihydroxy-4,9-dimethyl-dec-2-enoic acid methyl ester producing actinomycete causes severe morphological alterations in Alternaria brassicicola (Kaur et al.).

Besides Streptomyces species, Bacillus cereus AR156 is also a suppressing agent (Nie et al.). AR156-treated Arabidopsis plants activate MAPK signaling and FRK1/WRKY53 gene expression, both of which are involved in pathogen associated molecular pattern (PAMP)-triggered immunity (PTI) and suppress Pseudomonas syringae pv. tomato DC3000 (Pst DC3000). The strength of induced systemic resistance (ISR) against Botrytis cinerea due to these bacteria remains unknown (Nie et al.). A common bacterial rice pathogen is the virulent, biofilm forming Xanthomonas oryzae pv. oryzae (Singh et al.). With a combined application of Thyme oil (THY) and conventional bactericides the motA, motB, flgE, rpfF gene activity can be reduced, extracellular endoglucanase, xylanase, cellobiosidase, and polygalacturonase genes are expressed, the bactericide dose can be lowered, and disease progression be diminished.

An antioxidative amino acid is the histidine-derived, sulfurcontaining, non-proteinogenic ergothioneine (EGT) that adds to phyllospheric fitness (Alamgir et al.). M. aquaticum strain 22A produces high amounts of EGT. It has been isolated from a moss and has been only poorly studied. This is also the case for the facultative methylotrophic, pink pigmented bacteria (PPFM), identified as Delftia lacustris, Bacillus subtilis, or B. cereus by 16S rRNA gene sequence analysis. They are directly antagonistic against fungal pathogens of tomato and regarded as biocontrol agents (Janahiraman et al.). Out of 217 rhizobacterial isolates obtained from six different Indian Assam tea estates, Dutta et al. identified Enterobacter lignolyticus, Bacillus pseudomycoides, Burkholderia sp., or P. aeruginosa by partial 16S rRNA gene sequence analysis and as bio-fertilizer candidates for tea crops. Tamilselvi et al. identified Brevibacterium sp. SOTI06 as calcite dissolving bacterium, Krithika and Balachandar identified Bacillus sp. or Enterobacter cloacae as zinc (Zn) solubilizing bacteria, and Shakeel et al. reports that E. cloacae releases under iron- sufficient and -deficient conditions Znregulated transporters and iron (Fe)-regulated transporterlike proteins (ZIP) into the rice rhizosphere. Islam et al. assessed by quantitative real-time reverse transcription PCR that the 234 isolates obtained from the roots of basmati-385 and basmati super rice varieties translocate Zn toward grains and suppress economically important rice pathogens as Pyricularia oryzae and Fusarium moniliforme. In Bangladesh, Pseudomonas stutzeri, B. subtilis, Stenotrophomonas maltophilia, Bacillus amyloliquefaciens, present in cucumber rhizosphere and identified by phylogenetic 16S rRNA sequences, characteristically change the morphology of Phytophthora capsici hyphae growing toward PGPR colonies and from B. amyloliquefaciens (SN13) it is known that it prolongs stress tolerance in rice and acts as biocontrol agent against Rhizoctonia solani (Srivastava et al.). The virulence of fruit rot fungi during cranberry fruit development influences the release of ROS suppressive compounds such as benzoic and quinic acids (Tadych et al.) and intercropping can minimize the success of pea weevil larvae (Bruchus pisorum L.; Teshome et al.). Some genotypes have the capacity to obstruct pea weevil larval entry into developing seeds [neoplastic (Np) genotypes] of P. sativum and intercropping Np genotypes with other crops such as sorghum and maize can facilitate neoplasm formation. Insect-plantbacteria interaction is also the subject of Harun-Or-Rashid et al., who studied the widespread yield loss of many crops caused by the green peach aphid, a most destructive, plant sap sucking and plant viruses transmitting insect pest. Harun-Or-Rashid et al. found that Bacillus velezensis YC7010 which induces systemic resistance against bacterial and fungal pathogens of rice, can also induce systemic resistance against the green peach aphid (GPA), Myzus persicae, puncturing Arabidopsis leaves. B. velezensis significantly reduces setting, feeding and reproduction of GPA on Arabidopsis leaves by strongly inducing the senescencepromoting gene PHYTOALEXIN DEFICIENT4 (PAD4) while suppressing BOTRYTIS-INDUCED KINASE1 (BIK1). Another important soil-borne pathogen that affects field crops worldwide is Heterodera avenae. Trichoderma longibrachiatum has the ability to control H. avenae by surrounding the eggs completely with dense mycelia, lysing the content at late egg stage, and inducing a defense response in wheat plants (Zhang et al.). Insect-plant-bacteria interactions are also observable in citrus plants (Prasad et al.; Lin et al.). The destructive citrus disease, Huanglongbing, breaks out because in sap-sucking psyllids (Diaphorina citri) and Candidatus (Ca.) Liberibacter asiaticus infected citrus plants, the volatile cues are modified in a way that the D. citri predator, ladybird beetle P. japonica cannot find the host plant and reduce the insect pathogen, D. citri, population.

Plant health and productivity is of complex nature and strongly influenced by the intimate interaction of plants with deleterious and beneficial organisms. Obligate parasites affecting plant phloem such as D. citri and Candidatus (Ca.) Liberibacter link cellular levels with interdisciplinary scales and in such systems also viruses play a key role as the tomato—whitefly— Tomato Yellow Leaf Curl Virus (TYLCV) system studied by Tan et al.. Viral influence is also asserted by Li et al., who report that soil bensulfuron-methyl residues exhibit significant effects on the infestation of Bemisia tabaci, Myzus persicae, and Tobacco mosaic virus in Nicotiana tabacum and Trivedi et al. expanded such insights by highlighting the molecular, ecological, and evolutionary aspects of interactions among insects, plants, and their associated microbial communities. Islam et al., having nitrogen in their research focus, detail that a high level of nitrogen makes tomato plants release less volatiles and attract more Bemisia tabaci (Hemiptera: Aleyrodidae) and Santamaria et al. highlight that MATI proteins unveil a potential in defense through photosynthetic pigment modulation.

The Achnatherum sibiricum article of Qin et al. adds an aspect to the "Frontiers research topic Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent Research", worthwhile to have a closer look. They report in their e-book contribution that A. sibiricum, a little-studied, native grass of the Inner Mongolian steppe, is highly infected by endophytes but is not producing detectable endophyte-related alkaloids as known from agronomically important grasses such as tall fescue and perennial ryegrass. Despite the herbivore resistance, the

REFERENCES

Benckiser, G. (ed.). (1997). Fauna in Soil Ecosystems – Recycling Processes, Nutrient Fluxes, and Agricultural Production. New York, NY: Marcel Dekker, Inc.

Benckiser, G., and Schnell, S. (eds.). (2007). Biodiversity in Agricultural Production Systems. Boca Raton, FL: Taylor and Francis.

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

resistance to hosting Locusta migratoria increases by mechanisms different from endophyte-conferred and alkaloid decline.

All the articles contributed to the research topic "Plant-Microbe-Insect Interaction: Source for Bio-fertilizers, Bio-medicines and Agent Research" reveal interesting interactive details, and show that our understanding of the complexity of agricultural production systems is only in its very beginning. Nevertheless, the given insights in soil properties-virus-microbeplant-insect interactions are hopefully a stimulus for continuing the discussion in this field.

### AUTHOR CONTRIBUTIONS

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

### ACKNOWLEDGMENTS

We profusely thank the Editors of Frontiers in Plant Science for their continued help and support in successfully bringing out a Special Issue on the above Research Topic in their Journal and subsequently this e-Book.

Copyright © 2018 Benckiser, Hartmann, Kumar and Honermeier. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Recent Developments in Systems Biology and Metabolic Engineering of Plant–Microbe Interactions

### Vishal Kumar, Mehak Baweja, Puneet K. Singh and Pratyoosh Shukla\*

Enzyme Technology and Protein Bioinformatics Laboratory, Department of Microbiology, Maharshi Dayanand University, Rohtak, India

Microorganisms play a crucial role in the sustainability of the various ecosystems. The characterization of various interactions between microorganisms and other biotic factors is a necessary footstep to understand the association and functions of microbial communities. Among the different microbial interactions in an ecosystem, plant– microbe interaction plays an important role to balance the ecosystem. The present review explores plant–microbe interactions using gene editing and system biology tools toward the comprehension in improvement of plant traits. Further, system biology tools like FBA (flux balance analysis), OptKnock, and constraint-based modeling helps in understanding such interactions as a whole. In addition, various gene editing tools have been summarized and a strategy has been hypothesized for the development of disease free plants. Furthermore, we have tried to summarize the predictions through data retrieved from various types of sources such as high throughput sequencing data (e.g., single nucleotide polymorphism detection, RNA-seq, proteomics) and metabolic models have been reconstructed from such sequences for species communities. It is well known fact that systems biology approaches and modeling of biological networks will enable us to learn the insight of such network and will also help further in understanding these interactions.

Keywords: plant–microbe interactions, signaling, systems biology, CRISPR-Cas, gene editing

## INTRODUCTION

Microbial interactions have a decisive role in the sustainability of the various ecosystems. The characterization of such interactions among microorganisms and other biotic factors is a necessary footstep to understand the association and functions of microbial communities. Among the different microbial interactions in an ecosystem, plant–microbe interaction plays an important role to balance the ecosystem. Plants produce a number of organic and inorganic compounds which results in a nutritionally enriched environment which is favorable for heavy colonization of diversity of microbes. Microorganisms may colonize the exteriorly (epiphytes) or interiorly (endophytes). Microbial communities can affect the plant physiology either positively or negatively in direct or indirect ways by various interactions mutualism, commensalism, amensalism, and pathogenic consequences. Endophytic bacteria is an example of plant–microbe interaction wherein bacteria live in a non-competitive environment of host plant tissue without any major damage to the host cell (James and Olivares, 1998). Endophytes are omnipresent in nearly all plants on earth. Endophytic microflora such as bacteria and fungi, are defined as microorganisms which

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Oswaldo Valdes-Lopez, National Autonomous University of Mexico, Mexico Sangeeta Negi, New Mexico Consortium, USA Joseph Davis Bagyaraj, Indian National Science Academy, Centre for Natural Biological Resources and Community Development, India

\*Correspondence:

Pratyoosh Shukla pratyoosh.shukla@gmail.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 21 July 2016 Accepted: 06 September 2016 Published: 26 September 2016

#### Citation:

Kumar V, Baweja M, Singh PK and Shukla P (2016) Recent Developments in Systems Biology and Metabolic Engineering of Plant–Microbe Interactions. Front. Plant Sci. 7:1421. doi: 10.3389/fpls.2016.01421

are present after surface sterilization of various plant parts such as root, shoot, seed, or nodules. It has been assumed that these endophytes originated from the rhizosphere, the seeds, and the aerial portion of plants (Seghers et al., 2004). The rhizospheric soil is a significant source of root endophytes (Gao et al., 2004; Castro-Sowinski et al., 2007; Imam et al., 2013a). These endophytic microbes are supposed to enter into the plant tissue by local fractures or cellulose degradation of the root system (Gough et al., 1997). Endophytes inside a plant tissue may either be restricted to the point of entry or extend throughout the plant. These bacteria generally colonize the intercellular spaces, and they have been isolated from all compartments including seeds. There are few studies on plant–microbe interactions on details about Avr protein, computational strategies for protein interactions, molecular diversity and interactions of virulence genes (Imam et al., 2013a,b,c, 2014, 2015a,b). Both types of bacteria either Gram-positive or Gram-negative have been isolated from different tissues of numerous types of plant species. A number of facultative endophytes have been reported from rice, maize, wheat, sorghum, cotton, potato, and Arabidopsis. Furthermore, several different bacterial species have been isolated from a single plant. Conventionally, to investigate the various plant–microbial interactions use of a number of laborious laboratory experiments such as growth assays and pot house experiments are required (Kato et al., 2005; Harcombe, 2010; Zeidan et al., 2010). However, these laborious experiments make them infeasible for large scale application. With the help of bioinformatics approaches these issues can be alleviated by predicting plant–microbe interactions for experimental validation (Freilich et al., 2011; Buffie et al., 2014; Lima-Mendez et al., 2015). These predictions are founded on different types of informational data, such as the measurement of species abundances from high throughput sequencing or reconstructed metabolic models for species communities. There are several reports in various related fields where use of gene editing, genome engineering, and advanced technologies are proving quite significantly addressed (Gupta and Shukla, 2015a,b, 2016). In addition, various other in silico methods could be relevant to analyze such interactions while understanding the large amount of published data (Pritchard and Birch, 2011; Xu et al., 2013; Dix et al., 2016). This review envisages the concept of systems biology and gene editing in plant–microbe interactions by deciphering these technologies in detail.

### PLANT–MICROBE INTERACTION AND ITS RELEVANCE

Microflora is an aggregation of several types of microbes to form heterogeneous communities which are necessary components in several ecological niches and composed of distinct proportions of various microorganisms. Microorganisms of microflora do not live isolated or independently, but in its place these populations actively interact with other biological members of the ecosystem within their ecological niche. These microbial interactions may take place with any of biological form such as animal–microbe interaction, microbe–microbe interaction, plant–microbe interaction, etc. Plants provide an excellent ecosystem for microbial interactions. The plant provides the variable environment to the microorganisms from aerial plant part to the stable root system for the interactions. On the basis of location of plant–microbe interaction, the microbes can be divided in two groups, phyllospheric microorganisms which interact with the aerial leaf surface of plants and rhizospheric which interact with roots of plants. Phyllospheric microorganisms are adapted to low humidity and high irradiation, helps to protect plants from airborne pathogens. Rhizosphere of plants is a nutritionally rich zone due to deposition of nutrition rich compounds such as amino acid, organic acid, vitamins, sugars, etc. secreted by the roots. There is a pictorial presentation of various microbiome in **Figure 1** showing both phyllospheric and rhizospheric microorganisms. The nutritional enriched environment around roots creates a favorable environment for the growth of soil microorganisms, which includes rhizosphere and the rhizoplane soil microbial communities. A number of microorganisms interact with different plant tissues or cells with various level of dependence. These interactions may be beneficial, harmful, or neutral for one or both the organisms on the basis of this attribute plant–microbe interactions are known as amensalism (neutral– negative), antagonism (negative–positive), commensalism (neutral–positive), competition (negative–negative), mutualism (positive–positive), and neutralism (neutral–neutral). The commensalism or mutualism are more frequent interactions found in plants, in which either one or both species gain benefit from the relationship respectively (Campbell, 1995). Mycorrhiza and genus Rhizobium symbionts are best example of mutualism interaction. There are a number of superb reviews reporting present research on plant–microbe interaction at the molecular level, plant responses to quorum-sensing signals from microbial communities, applications of plant–microbe interaction, microflora responses toward transgenic plants and other rhizospheric interactions (Bauer and Mathesius, 2004; Singh et al., 2004; Sørensen and Sessitsch, 2007; Fillion, 2008; Ryan et al., 2008). The examination and understanding of these plant–microbe interactions helps to figure out the insights of mechanism which may direct us to understand such concerns. These sustainable resources will be ecofriendly and helpful to clean up the pollution and gaseous effect on a global scale.

### SYSTEMS BIOLOGY APPROACHES IN PLANT–MICROBE INTERACTIONS

### Communication Systems

The life cycles of all the organisms from quorum sensing bacteria (Cornforth et al., 2014) to singing whales (Parks et al., 2014) are found on signaling pathways to convey information. Signaling system has played an important role in organismal evolution and the complexity of life (West et al., 2015). If both the donor as well as a receiver has a shared interest to propagate the reliable information then an effective signaling system can fetch a number of health benefits. The signaling pathway may be important from an evolutionary point of view because organisms can manipulate signals for their

profit (Mokkonen and Lindstedt, 2015). Now these days, there has been an escalating awareness in communication network between the plants and root microflora which have a symbiotic relationship (Miller and Oldroyd, 2012: Bakker et al., 2013; Andreo-Jimenez et al., 2015). The roots of plant are bordered by a massive amount of soil microorganisms consisting of tens of thousands species diversity (Bardgett and van der Putten, 2014). There should be an effective crosstalk between plant and surrounding microflora to establish a successful relationship. There should a better understanding of these molecular signaling pathways to access control over the microbial population. The researchers have made efforts from last decade to understand the molecular mechanism of communication in the rhizosphere (Guttman et al., 2014) but still we do not have sufficient knowledge to comprehend the evolutionary origins and stability of the rhizosphere communication system. Comprehension of major beneficial plant–microbe interactions such as arbuscular mycorrhizas and the plant growth promoting rhizobacteria (PGPR)–legume symbiosis have been changed over the past years. The PGPR–legume root symbiosis and arbuscular mycorrhizal (AM) symbioses are established by exchanging a number of signals as there is mutual identification of diffusible signal molecules generated from both plants and microbial partner. A common symbiotic pathway (CSP) is triggered by symbiotic signals produced by rhizospheric bacteria or fungi which are in form of lipo-chitooligosaccharides (LCOs). These LCOs are perceived via lysine-motif (LysM) receptors found on the plasma membrane of plant cell and activate the CSP which regulate the interactions between plant and rhizospheric microorganisms. LysM receptor families are found in both legume and nonlegume plants and receive signals from both rhizobia (Nod

factor signals) and AM fungi (Myc-LCO signals). A model of CSP triggered in plants has been described in **Figure 2** together with all the proteins and receptor molecules involved in signaling. Furthermore, in this review it has been tried to understand the signaling pathway among AM fungi and roots of their host plants, where organic food is exchanged for nutrients from soil. This symbiotic relationship is among the most prevalent and anticipated to have evolved roughly 450 Mya (Field et al., 2015). There are several evidences obtained that signaling pathways between AM fungi and roots of their plant hosts are so thriving that the components of this pathway have been recruited by plants to evolution of other symbiosis such as rhizobial N2-fixation (Geurts et al., 2012). Plants and microorganism use a signaling system to transmit information about their internal situation and their readiness for immigration or colonization, but how can these reach the desired recipients, and not others (Oldroyd, 2013). Theoretically, specific signaling is required at two levels a broader screening to identify or stimulating the mutualists and a finer screen, to distinguish high and low-quality strains within a mutualist microorganism (Werner and Kiers, 2015). Strigolactones are acting as a major plant signaling molecule in the symbiotic system of arbuscular mycorrhiza. Strigolactones are terpenoid lactones which are a byproduct of carotenoid metabolism (Bonfante and Genre, 2015). However, Strigolactones are plant hormones, which secondarily also act to attract AM fungi. Strigolactones act as a stimulus to initiate metabolic cycle of the AM fungus which promotes growth toward the roots (**Figure 3**; Gutjahr, 2014). The receptors for strigolactone in mycorrhizal fungi have not been yet discovered (Koltai, 2014) Different types of strigolactones have been emitted by different plants which

vary from host to attract specific fungal species or strains (Conn et al., 2015). The germinating AM fungal spores were activated by strigolactones derived from a root which execute a series of signal molecules such as chitooligosaccharides and lipochitooligosaccharides. These signal molecules activate a set of reactions in the plant root system and consequently the cytosolic concentration of calcium boosts which further induces gene expression of activated AM fungi which directs to the creation of the pre-penetration apparatus. The reacting root will secrete cut-in monomers, signaling the fungi to form a hypopodium and initiate arbuscular growth (Padje et al., 2016). The PGPR is known to synthesize the phytohormones, auxins. Auxin production can occur via multiple pathways by both plants as well as PGPRs. There are certain papers available which report that indole-3-acetic acid (IAA) is a natural auxin acting as signaling molecules in microorganisms. IAA affects gene expression in some of microorganisms, thus IAA act as a reciprocal signaling molecule in microbe–plant interactions (Spaepen and Vanderleyden, 2011). The bacterial gene expression is regulated under the control of IAA has been first described for ipdC gene of Azospirillum brasilense. It has been reported that IAA act as an inhibitory signal molecule for viral gene expression by Agrobacterium tumefaciens a phytopathogen (Liu and Nester, 2006). Furthermore, auxin level in plant–PGPR interactions affects different levels of nodule formation in plants such as auxin transport inhibition by the flavonoids which act as indicators of specification of founder cell and auxins accumulations initiate the nodule formation and differentiation (Mathesius, 2008).

### In silico Methods in Understanding Interactions

Systems biology is the study of genes, proteins and their interaction within a cell, tissue or whole organism. It also enables us to understand complex biological system and modeling it with the help of computational techniques. The interaction of host and pathogen in plants plays an important role in enhancing signaling cascade which brings change in the protein and eventually in the phenotypic expression. There are few notable studies on systems biology and molecular modeling tools to understand the microbial enzymes and similar proteins, but it lacks any further scope for studies of proteins involved in plant–microbe interaction (Singh and Shukla, 2011, 2015; Karthik and Shukla, 2012; Baweja et al., 2015, 2016; Singh et al., 2016). The study of in silico transcriptomes of both host and pathogen during the infection will contribute to the knowledge of changes occurring during the infection. There are different database which is dedicated to host–pathogen interaction. There is dynamic complexity in the plant–microbe interaction which occurs since edges represent processes in biological networks that may take time to occur and are dependent on the other factors in the network. Concentrations of metabolites in metabolic and signaling processes vary over time thus there could be several ways to model this time-dependent variation. Ordinary differential equations are employed for the analysis and calculation of biochemical process for metabolic kinetics studies. In such studies edges and node forms the complex,

edges are associated with some value of parameters such as binding coefficients. Edges comprise of values representing a quantity or concentration. There are variations in the value of nodes over the time as the substrate is utilized or byproduct is formed. Flux is the rate at which material flows, flux is associated with the edges and carry a certain value. Understanding flux and managing it helps in the regulating the biological process dynamics. The study of the dynamic behavior of interaction is complex to analyze even studying a small, dynamic behavior requires certain parameters and information which requires multiple dimension overview. The networks and their dynamic characteristics may be significant and these processes should be confirmed with valid experimental models. Topologies related to metabolomics of cell are dynamic between the compartments and they change over the time. It is obvious to mention here that concentrations (or counts) of active proteins, crucial metabolites within the interacting cell are more inconsistent than the topology of the metabolic model. This indicates a clear overview about that existence and these factors define the network topology. Furthermore, the amount of each active element in such system has varied significantly so such attributes are accessed by metabolomics, transcriptomics, and proteomics and these can be taken as significant markers to explicit the host– microbial interactions. There are examples in which microbes dominated the over the molecular control of the host and resulted in exceptional results including production of "zombie ants"

and mimicry of flowers by the fungi Ophiocordyceps unilateralis (Pontoppidan et al., 2009) and Puccinia monoica (Roy, 1994). Such examples exemplify the potential of microorganisms to control elegantly the physiological processes in host cells. In it quite important to mention here that such microbes have developed the capacity toward environment control and influence the surrounding factors. The systems biology approach helps to find out various ways toward the alteration of host plant cells. There are not many chances that all the symptoms that appear in the plant–microbe interaction come out as a disease, it is just the coincidental part that occurs. All pathogens are causing disease will not be the right thing to consider. The pathogens which attack the host first explore the most vulnerable element of the host network that could cause more disruption in the most economical way. By virtue of this, the host also develops its defense system and the pathogen attack may be detected only in those parts of the system which are structurally most responsive to these changes. Further, it is to mention here that host cell will be benefited because the reduction in the number of receptor and recognition proteins. Systems biology approach and mathematical modeling of the system could also lead us to develop novel strategies to control the disease. Apart from these, the metabolism of plant engineered in microbe will show the way to the production of different essential components which are commercially important such as fuel and pharmaceutical molecules. An overall depiction of the methods described above is given in **Figure 4**.

### Systems Biology Techniques for Deciphering Plant–Microbe Interaction

Metabolic engineering in microorganisms has been employed in different areas such as industrial microbiology, medical microbiology, and agricultural microbiology (Chotani et al., 2000; Nakamura and Whited, 2003). The targeted motive of metabolic engineering could be different, but the technology and platform remained unchanged. Recently, computational modeling emerged and changed the perspective to analyze metabolic engineering. Computational modeling anticipates the effect of genetic manipulations on metabolism, however, these methods need enzyme kinetic information that is still mostly unknown (Tepper and Shlomi, 2010). Constraint-based modeling (CBM), is an alternate which overcome these problems by examining the function of metabolic networks by relying on physical–chemical constraints (Price et al., 2003). There are certain genome-scale network models available for many microorganisms (Förster et al., 2003; Reed et al., 2003; Duarte et al., 2004). CBM has proved to be successful for large-scale microbial networks which involve metabolic engineering studies for different applications. A metabolic reconstruction is a wellstructured description of the network topology that enables derivation of genome-scale models (GEMs) that are used to mimic different metabolic states of an organism (Satish Kumar et al., 2007; Thiele and Palsson, 2010; Esvelt and Wang, 2013). Such technology has gained popularity for systems biology

studies as it enables the integration of omics and overall analysis to explore the interplay of metabolic networks (Saha et al., 2014). A few metabolic reconstructions have been developed for different plant species, including Arabidopsis (Poolman et al., 2009; de Oliveira Dal'Molin et al., 2010a), maize (de Oliveira Dal'Molin et al., 2010b; Saha et al., 2011), sugarcane, and sorghum (deOliveira Dal'Molin et al., 2010b). The effectors act outside the host cell and sometimes secrete small molecules that may affect the host and modifies its biochemistry, for example, coronatine. We understand systems biology perspectives can be well applied to study such effectors and their pathogenesis aspects. These studies are based on certain tools which help in analyzing large amount of genomic data, interactions, GEMs this is depicted in **Table 1**. OptKnock is a technique which searches for sets of gene knockouts that lead to the production of desired products (Burgard et al., 2003) and can be used for the same purpose which can resist the plant from harmful microbial compounds. On the other hand, OptStrain that not only allows gene knockouts, but also incorporate novel enzymecoding genes from different species to a given microbial genome (Pharkya et al., 2004). More recently, OptReg was developed, searching for manipulations in the form of up- and downregulation of metabolic enzymes in addition to gene knockouts to meet desired metabolite production (Pharkya and Maranas, 2006).

### Gene Editing: An Approach to Develop Customized Functions

The recombinant DNA technology has revolutionized the study of the genome to a next level to provide the opportunity for its application in various fields like agriculture, industries, etc. The techniques like gene editing are proving as potential techniques in improvement of crop characters such as enhancing yield, providing resistance from biotic and abiotic stress. This has been possible because of major gene editing tools like zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALEN), and clustered regularly interspaced short palindromic repeats (CRISPR-Cas) that introduce double strand break (DSB) in the target gene, which are repaired by the error-prone non-homologous end joining (NHEJ) pathway or homology-directed repair (HDR; Symington and Gautier, 2011).

ZFNs are artificial restriction enzymes that edit or cleave the specific target DNA by using zinc finger DNA-binding domain. The recognizing sequences viz. zinc finger domains can be artificially engineered to target specific sequences in the host. It consists of two DNA binding domains, the domain one is comprised of eukaryotic transcription factors and contain a zinc finger. The second domain includes the catalytic component, the nuclease FokI restriction enzyme that catalyzes the specific DNA sequences. ZFNs have successfully performed well in defining the functions of various genes from diverse organism, including proven highly valuable in defining the roles of numerous genes in cells from a variety of organisms, including fruit flies, humans, mice, and higher plants (Gaj et al., 2013). However, there are certain drawbacks of ZHN technology like difficulties in design, construction, cost, and uncertain success rates.

TALEN are restriction enzymes that cleave target DNA by utilizing TAL effector DNA binding domains. The specific targeting is aided by simple "code" that matches with the diamino acid sequence (repeat-variable di-residue) in ∼33–35 amino acid conserved target sequence. The progress in gene editing tools and development of various methods for easy synthesis and assembly of TALENs, allows the efficient editing at multiple sites. There have been various examples of the success of TALENs like knockout of the CCR5 gene for HIV resistance in human cells (Mussolino et al., 2011); destruction of the bacterial blight disease susceptibility gene in rice (Li et al., 2012); disruption of the LDL receptor in swine (Carlson et al., 2012); replacement of a tyrosine hydroxylase gene via TALEN-enhanced homologous recombination in zebrafish (Xiao et al., 2013; Zu et al., 2013).

#### TABLE 1 | Applications of tools related to systems biology.

fpls-07-01421 September 22, 2016 Time: 17:41 # 7


### CRISPR-Cas in Understanding Interactions

Gene editing has been highly appreciated for their ability to change the desired DNA fragment using engineered nucleases often called as molecular scissors. Since it edits the product according to fitment of the process it has various applications in a diversity of areas. The CRISPR-Cas system has been evidenced as most efficient, easy and simple (Kanchiswamy et al., 2016). CRISPR-Cas system, also known as third-generation programmable nuclease has a major role in crop protection. There are approximately 11 CRISPR-Cas systems have been reported. They can be distinguished into three types (Types I–III) which are further divided into 11 subtypes (Ma and Liu, 2016). Each type has its own specific Cas protein component which is named according to model organism.

Cas9 is a DNA endonuclease guided by RNA to target foreign DNA for inhibition (**Figure 5**) The guide RNAs (gRNAs) are derived from CRISPRs. CRISPRs consists of tandem arrays of a 30–40 bp short, direct repeat sequence which are separated by spacer sequences that matches the foreign sequence. Further transcription and processing of CRISPR produces mature CRISPR (cr)RNAs, the sequence flanked by signature CRISPR repeat tag at 5<sup>0</sup> and 3<sup>0</sup> end. The CRISPR (cr)RNAs form complex with Cas proteins to form a ribonucleoprotein (crRNP) that introduce cleavage in the DNA/RNA of the invader (Hale et al., 2012). One of the remarkable features of CRISPR is the specificity, that is aided by gRNA, that allows specific binding to target DNA and beauty of the system lies in the customized engineering of the gRNA. The specificity was enhanced by using double nickase and Cas9-nuclease fusion systems. Double nickase system allows binding of two gRNAs, both upstream as well as downstream preventing off target editing. This was further improved by using inactivated Cas9, i.e., without nuclease activity, fused with restriction enzymes. The nuclease activity of restriction enzyme only gets activated when both are in close proximity (Guilinger et al., 2014). The gene of interest can be inserted or deleted from the system with the help of CRISPR/Cas9 by introducing DSBs into a target site (Vanamee et al., 2001; Auer et al., 2014). Suitable expression construct is required for successful accomplishment of CRISPR-Cas sgRNA sequence(s), the codon-optimized variant of Cas9, strong promoters suitable to derive transcription of sgRNA and Cas9 (Raitskin and Patron, 2016). The importance of all these parameters was elucidated in a review by Schaeffer and Nakata (2015). With progress in computational techniques various computational tools like E-CRISP, CRISPR design tool, and CHOPCHOP have been developed that allow to identify the probable sequence of cleavage using input target sequences. Therefore, it helps to design gRNA (Hsu et al., 2013; Heigwer et al., 2014; Montague et al., 2014).

Once the target site is recognized by the gRNA, the nuclease Cas9 with the aid of its two domains RucV and HNH breaks the strand and generate blunt end DSB. Such DSB can be repaired by NHEJ that introduce mutation at the targeted site or by HDR, that may knock-in or replace the desired gene fragment at the target site using template DNA. There are various examples of gene editing utilized by different microbes (**Table 2**). Additionally, multiple editing in the same cell is possible using multiple gRNA that show various applications, like mutation in genes which are

TABLE 2 | Genome editing in different plant species by the CRISPR/Cas technology.


functionally related to control complex traits (Ma et al., 2015; Xie et al., 2015). In a study, expression of Cas9 and sgRNA genes in Arabidopsis and tobacco, caused a targeted cleavage of a nonfunctional GFP gene. Further mutation by NHEJ DNA repair led to the production of a strong green fluorescence in transforming leaf cells (Jiang et al., 2013, 2014).

To enhance the expression of Cas9 in plants, codon optimization is often used strategically (Fauser et al., 2014). For the expression of Cas9, constitutive promoters of ubiquitin genes of rice, Arabidopsis, and maize can attain the desired requirement of gene editing in monocot and dicot plants.

### Plant–Virus Interactions and Desired Trait Improvement

Earlier, the studies on trait improvement were based on plant breeding, somatic hybridization, and random mutagenesis, the process was tedious and time consuming. The trend of plant breeding was replaced by efficient and simple tools, i.e., CRISPR-Cas to introduce specific traits into the population. The effort was done to enhance the sensitivity toward the herbicide. The three oligonucleotides were targeted by CRISPR-Cas via A. tumefaciens. The transformation was done using single gRNA in a binary vector and successfully mutants were found to be sensitive to bentazon herbicide. A genome modification study was done for the first time in the maize utilizing TALENs and CRISPR-Cas and concluded that both the systems efficiently can be used for genome modification in maize (Liang et al., 2014). Similar studies were done in tobacco and it also suggested that CRISPR-Cas is an efficient genome modification tool (Gao et al., 2015). The studies were done to enhance the gene targeting and it was observed that virus mediated transformation showed a higher frequency

than the traditional A. tumefaciens T-DNA (Xu et al., 2014). Baltes et al. (2014) reported such finding in Nicotiana tabacum by using Gemini virus replicons to enhance the gene targeting and also revealed the DNA sequence editing using Gemini virus replicons. There have been a number of strategies for multiple gene targeting using multiple gRNA in a single plasmid vector described by Raitskin and Patron (2016). The Cas9 are now recently used to control the pests. In a study, the Cas9 was used to control the population of Drosophila melanogaster. Engineered endonuclease-based drive systems have been used to drive mutations into populations of pest species leading directly or indirectly to reduce population sizes (Reid and O'Brochta, 2016).

In near future, it is expected that CRISPR-Cas will prove as a remarkable tool to engineer plants to eradicate problems associated with crops like low yields, nutritional content, and resistance from biotic and abiotic factors. The technique can also be utilized to prevent the plant diseases by inhibiting the virus interaction with the plant system (**Figure 5**). The bacterial CRISPR-Cas could be used to inhibit the viral genetic material with the action of Cas9 as a nuclease thereby curtailing the establishment of viral infection in the plant (Ali et al., 2015; Baltes et al., 2015; Chaparro-Garcia et al., 2015; Ji et al., 2015). There are various examples where CRISPR-Cas system has proved to be successful in improving plant traits. In a rice plant, genetic modification was done in large chromosomal segments of sugar efflux transporter genes that resulted in 87–100% editing in T0 transgenic plants (Zhou et al., 2014). The gene function was first time revealed in the citrus fruit with the aid of CRISPR-Cas (Jia and Wang, 2014). CRISPR/Cas9 technology is most useful in woody plants that have long reproductive cycles, as they have the ability to acquire mutants in T0 generation (Fan et al., 2015; Tsai and Xue, 2015). Indeed, such results of gene editing empower the idea of the customized editing and desired expression in all living systems.

Certainly, successful development of the Cas9/sgRNA system for targeted gene modification and genome editing holds promise for boosting fundamental knowledge of plant biology as well as for designing crop plants with potential new agronomic, nutritional, and novel traits for the benefit of farmers and consumers.

### REFERENCES


### CONCLUSION

Microbes play a fundamental role in diverse ecosystems through microbial interactions with other biotic and abiotic components of the ecosystem. Plant–microbe interactions play an important role in plant health and ecological sustainability. So, comprehension of these interactions is very crucial to improve plant health and ecological sustainability. Recently, microbial interaction prediction using computational biology has become an extensively used approach to inspect the plant–microbial interactions. In this review, different computational methods developed by the computational data has been summarized to understand plant–microbe interactions. Several systems biology tools such as FBA (flux balance analysis), CBM, and OptKnock has been described to understand the metabolic pathways involved in plant–microbe interactions. Furthermore, gene editing tools such as TALENs and CRISPER-Cas have been described to control the pathogen interactions with plants to obtain customized plants. A snapshot of gene editing tools has been described to obtain disease free customized plants. There should be a better understanding of signaling pathways and metabolic networks to have an understanding of plant–microbial interactions. A combinatorial approach of computational biology and genomic tools has proven supportive to understand the communication pathway and metabolic pathway and provides an alternative to regulate these pathways to get a beneficial effect on plants with ecological sustainability.

### AUTHOR CONTRIBUTIONS

All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication.

### ACKNOWLEDGMENTS

VK is thankful to UGC New Delhi, India for awarding Junior Research Fellowship [F.17-63/2008 (SA-I)]. MB is grateful to Maharshi Dayanand University, Rohtak for URS.



verification by virulence analysis. Proc. Natl. Acad. Sci. U.S.A. 83, 499–504. doi: 10.1007/s40011-013-0186-6


new web-tool and a modified Cas9 enzyme. GM Crops Food 6, 243–252. doi: 10.1080/21645698.2015.1106063



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Kumar, Baweja, Singh and Shukla. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Allele Mining and Selective Patterns of Pi9 Gene in a Set of Rice Landraces from India

#### Jahangir Imam1,2† , Nimai P. Mandal<sup>1</sup> , Mukund Variar<sup>1</sup> and Pratyoosh Shukla<sup>2</sup> \*

<sup>1</sup> Biotechnology Laboratory, Central Rainfed Upland Rice Research Station, Hazaribagh, India, <sup>2</sup> Enzyme Technology and Protein Bioinformatics Laboratory, Department of Microbiology, Maharshi Dayanand University, Rohtak, India

Allelic variants of the broad-spectrum blast resistance gene, Pi9 (nucleotide binding site-leucine-rich repeat region) have been analyzed in Indian rice landraces. They were selected from the list of 338 rice landraces phenotyped in the rice blast nursery at central Rainfed Upland Rice Research Station, Hazaribag. Six of them were further selected on the basis of their resistance and susceptible pattern for virulence analysis and selective pattern study of Pi9 gene. The sequence analysis and phylogenetic study illustrated that such sequences are vastly homologous and clustered into two groups. All the blast resistance Pi9 alleles were grouped into one cluster, whereas Pi9 alleles of susceptible landraces formed another cluster even though these landraces have a low level of DNA polymorphisms. A total number of 136 polymorphic sites comprising of transitions, transversions, and insertion and deletions (InDels) were identified in the 2.9 kb sequence of Pi9 alleles. Lower variation in the form of mutations (77) (Transition + Transversion), and InDels (59) were observed in the Pi9 alleles isolated from rice landraces studied. The results showed that the Pi9 alleles of the selected rice landraces were less variable, suggesting that the rice landraces would have been exposed to less number of pathotypes across the country. The positive Tajima's D (0.33580), P > 0.10 (not significant) was observed among the seven rice landraces, which suggests the balancing selection of Pi9 alleles. The value of synonymous substitution (−0.43337) was less than the non-synonymous substitution (0.78808). The greater non-synonymous substitution than the synonymous means that the coding region, mainly the leucine-rich repeat domain was under diversified selection. In this study, the Pi9 gene has been subjected to balancing selection with low nucleotide diversity which is different from the earlier reports, this may be because of the closeness of the rice landraces, cultivated in the same region, and under low pathotype pressure.

Keywords: allele mining, rice landraces, polymorphism, blast resistance genes, selection pressure

### INTRODUCTION

Rice blast (Magnaporthe oryzae), the most serious diseases of rice causes significant yield loss globally and the complexity of pathogen, host, and microclimate have a profound effect on this (Valent, 1990; Teng et al., 1991; Kwon and Lee, 2002; Li et al., 2007). The blast fungus is both sexual and asexual in nature which resulted in the evolution of its different variants in field

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Tian-Qing Zheng, Chinese Academy of Agricultural Sciences, China Manish Kumar Pandey, International Crops Research Institute for the Semi-Arid Tropics, India

> \*Correspondence: Pratyoosh Shukla

pratyoosh.shukla@gmail.com

†Present address:

Jahangir Imam, State Forensic Science Laboratory, Directorate of Forensic Science, Department of Home and Disaster Management, Ranchi, India

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 17 September 2016 Accepted: 22 November 2016 Published: 15 December 2016

#### Citation:

Imam J, Mandal NP, Variar M and Shukla P (2016) Allele Mining and Selective Patterns of Pi9 Gene in a Set of Rice Landraces from India. Front. Plant Sci. 7:1846. doi: 10.3389/fpls.2016.01846

**25**

conditions (Xia et al., 1993). The high adaptation frequency and variations lead to the emergence of new races of the fungal population of M. oryzae leads to the breakdown of resistance in newly released rice cultivars in the fields (Kiyosawa et al., 1986; MacKill and Bonman, 1992; Valent and Chumley, 1994; Han et al., 2001). Under normal field condition, incomplete, or field resistance of blast disease is better options for the effective control of M. oryzae (Liu et al., 2005). The host–pathogen interaction can be better understood by the identification and characterization of both R and Avr genes and certain enzyme based studies on technological improvements through combined approaches (Imam et al., 2014a,d, 2015a, 2016; Baweja et al., 2016; Kumar et al., 2016). Till date, many blast resistance genes and QTLs have been recognized and cloned (Sharma et al., 2005; Imam et al., 2014b, 2015b). Most of the rice blast resistance genes cloned till date encode nucleotide binding site-leucine-rich repeat (NBS-LRR) proteins which suggest the common escape root involving a familiar resistance pathway to counter blast infections (Hammond-Kosack and Jones, 1997; Takken and Tameling, 2009; Liu et al., 2010; Imam et al., 2013b).

The transfer of valuable alleles found in the rice germplasm is generally employed by the plant breeders for the improvement of high-yielding varieties (Kumar et al., 2010). Natural mutation like transition, transvertion, point mutation, and insertion and deletions (InDels) is the main driving force for the generation and evolution of new alleles. With the availability of enormous database information, desired and superior alleles can be easily identified and retrieved (Kumar et al., 2010). The potential application of allele mining approach is in the identification of new haplotypes and evolution pattern study which helps in the rice improvement programs (Kumar et al., 2010). TILLING (Targeting Induced Local Lesions in Genomes) and sequencebased allele mining are the two main approach for sequence polymorphism study in the natural population of germplasm (Till et al., 2003; Kumar et al., 2010). Allele mining has emerged as an important approach for cloning and characterization of new and better forms of disease resistance genes. Isolation of orthologs provides insights into the evolutionary forces shaping the development that help identification of better alleles for future experiments (Ashkani et al., 2015). Because of its facile nature, this approach is being used extensively to identify alleles of agriculturally important traits. Wild as well as cultivated rice varieties has been studied for the blast resistance genes by allele mining approach (Geng et al., 2008; Huang et al., 2008; Yang et al., 2008). An extensive study of the Pi-ta locus was described from wild species of rice (Huang et al., 2008), cultivated (AA) and wild species and invasive weedy rice (Lee et al., 2009, 2011). Another gene, Pid3 studied from 36 rice lines of both cultivated and wild species indicated pseudogenization of Pid3 in japonica cultivars (Shang et al., 2009). Liu et al. (2011) reported divergent selection in Pi9 locus cloned from cultivated and wild species of rice. Most of the above mentioned blast genes were studied through sequence-based allele mining approach.

The Pi9, Pi2, and Piz-t, the three paralogs of the blast resistance gene family is now well characterized and resistance mechanism is known (Qu et al., 2006; Zhou et al., 2006). MacKill and Bonman (1992) discussed the origins of Piz-t and Pi2 genes from indica cultivars while Oryza minuta, a wild rice is the source of origin of Pi9 gene and high LRR positive selection (Amante-Bordeos et al., 1992; Zhou et al., 2007; Dai et al., 2010). The Pi9 locus contains at least six known resistance genes specific to the fungal pathogen M. oryzae and three R-genes from this locus (Pi9, Pi2, and Piz-t) have been cloned (Qu et al., 2006; Zhou et al., 2006). The resistance specificities of different broad spectrum rice blast resistance genes shown to be different from one another which mainly arise because of the different evolutionary changes in the NBS-LRR genes and its generation and also the helps in the adaptation to ever changing pathogen populations (Bergelson et al., 2001; Shen et al., 2006; Liu et al., 2007, 2013; Yang et al., 2008). For the preservation of the resistant germplasm, knowledge of the variation patterns of R-genes is important. Yang et al. (2006) worked on the genomewide allelic analysis of R-genes between two rice cultivars and categorize the variation patterns into four types, namely type I, type II, type II, and type IV from conserved to presence/absence genes. Basically type I and type II plays the main role in R-genes allelic polymorphism and resistance specificity and this give rise to rapid evolution in these blast resistance genes and enable them to adapt to the ever-changing pathogen population (Bergelson et al., 2001; Shen et al., 2006; Yang et al., 2008). The single-copy gene dominant group (type I), showed the lowest diversity (<0.005); the clustered-gene dominant group (type III), have a high level of diversity and the intermediate one (type II), and the presence or absence of R-genes in one genomes (P/A R-genes, type IV) (Yang et al., 2008). Different blast resistance genes showed different levels of polymorphism and diversity. The LRR region of Pi54 and Piz-t genes are more vulnerable for changes and leads to positive directional selection as compared to LRR regions encoded by Pi-km1 and Pi-km2 blast ressitance genes which are highly conserved (Ashikawa et al., 2010; Thakur et al., 2013). It is interesting to note that LRRs have direct interacting roles with effector proteins (Young and Innes, 2006). Higher levels of polymorphism were observed in the LRR region of Pi54 which helps in effector recognition and the evolutionary pressure by virulent M. oryzae races results in in variations in the LRR domain (Thakur et al., 2015).

Allelic polymorphisms in the R-genes are mainly driven by balancing selection and positive selection. The positive selection is the main evolutionary force which maintains the polymorphisms in the R-genes family plants (Bent and Mackey, 2007; Zhou et al., 2007). The genome structure at Pi9 locus is highly conserved but the LRR region showed high sequence variation giving rise to positive selection for Pi9 genes among the rice germplasm (Zhou et al., 2007; Dai et al., 2010; Liu et al., 2010). Previous data toward the prevalence of high pathotypic diversity of the M. oryzae population from Eastern India and very rare compatibility of Pi9 gene in field evaluations with isolates from this region provoked this study (Variar et al., 2009). Amongst the multi-genes near the waxy gene locus on chromosome 6, Pi9 was extra efficient than Pi2 (Piz-t) in the preceding investigation (Ballini et al., 2008; Imam et al., 2014b). Our recent studies on the rice and M. oryzae interaction

of isolates collected from India reveals the compatible and incompatible interactions between the R and the corresponding Avr genes (Imam et al., 2013b, 2014a, 2015a). Despite the fact that resistance mediated by single R-gene can be easily wrecked by emerging virulence, some cultivars with major resistance genes have stay resistant for a extended time without resistance loss (Khush and Jena, 2009). A likely rationale for the durability of Pi9-mediated resistance to blast is the fact that the gene presents broad-spectrum resistance to miscellaneous isolates. The germplams harboring the Pi9 gene identified in the earlier study originated from different Eastern Indian locations exhibited excellent resistance to several isolates from the region, which is appealing to hypothesize that the gene is effective and durable (Imam et al., 2014c). To better understand the genetic polymorphism and molecular evolution mechanism of the Pi9 alleles, we analyzed the 2.9 kb region of the Pi9 gene in six accessions of cultivated rice landraces. Therefore, the present investigation is taken up for the allele mining of NBS-LRR region of Pi9 gene from the rice landraces to better understand the sequence polymorphisms and its relevance in resistance and susceptibility pattern. The objectives of this study were (1) to isolate alleles of Pi9 blast resistance gene, (2) to understand the nucleotide diversity in NBS-LRR region of Pi9 gene, and (3) analysis of the molecular evolution and patterns of selection in this region.

### MATERIALS AND METHODS

### Plant Materials

A selected set of 338 rice landraces accessions which were reevaluated in the uniform blast nursery (UBN) were considered for allele mining of Pi9 gene. Out of 338, seven rice landraces accessions were taken for further analysis and allele mining (**Table 1**). The selection for allele mining was based on the result of the presence of Pi9 gene by STS marker and their disease score. A pair of dominant STS markers 195R-1 (5<sup>0</sup> -ATGGTCCTTTATCTTTATTG-3<sup>0</sup> ) and 195F-1 (5<sup>0</sup> - TTGCTCCATCTCCTCTGTT-3<sup>0</sup> ) derived from the Nbs2-Pi9 candidate gene was used to check the presence of Pi9 gene in the rice landraces in this study (Qu et al., 2006). Out of seven, one is the iso-genic line for Pi9 gene, three were resistant and rests three were susceptible to blast disease.

### Phenotype Evaluation of Landraces

A mixture of virulent isolates (Mo-ei-66, Mo-ei-79, Mo-ei-119, and Mo-ei-202) was used as inoculum for the phenotyping of selected rice landraces (Imam et al., 2013a). Oat Meal Agar (HiMedia, India) medium was used to maintain the fungal culture of each isolate. The Mathur's medium was used for the sporulation and multiplication of fungal spores. These cultures were preserved at 22◦C for 12–16 days under stable illumination from white fluorescent light (55 µF/Em/s) (Thakur et al., 2013). Conidia were split from the conidiophores which were used for the preparation of fungal spores and the inoculum were maintained to approximately 10<sup>5</sup> spores/ml. The leaf stage seedlings (2–3 in number) in replicated sets were spray-inoculated with 1 ml mixed spore suspension and then kept back in darkness at 27◦C and over 90% relative humidity for 24 h. In this experiment, positive control for Pi9 gene (IRBL9 w) and rice landraces was grown in plastic pots and maintained in phenotyping facility. After inoculation with mixed fungal cultures, the rice seedlings were maintained in the phenotyping chamber with desired temperature and humidity. Analysis of virulence was completed on the basis of reaction type using 0–5 standard evaluation scale. Resistance was scored based on no visible infection and no conidia produced from inoculated tissue (scores 0, 1, 2), while susceptibility was scored with a lesion >3 mm in length (score 3, 4, 5) and sporulating (Bonman et al., 1986).

### PCR and Sequencing

Overlapping oligos were designed using Primer 3 software<sup>1</sup> to amplify 2.9 kb NBS-LRR region of Pi9 gene (DQ285630) using primer walking technique (Thakur et al., 2015). A total of five primer pairs was designed to amplify the 2.9 kb region (**Table 2**). PCR was carried out from the isolated DNA of the isogenic line IRBL9-w and six rice landraces using Q5 high-fidelity DNA polymerase (New England Biolabs, Life Technologies, USA) to amplify full-length allele with high-fidelity with the following thermal cycling conditions: initial DNA denaturation at 95 ◦ C for 2 min followed by 30 cycles of 95◦ C for 30 s, 58 ◦ C for 30 s, 72◦ C for 1 min, final elongation at 72◦ C for 10 min and hold at 48◦ C. The amplified PCR products were then sequenced<sup>2</sup> and assembled. Phred/Phrap/Consed software (Ewing and Green, 1998) was used for the assembly of multiple reads of different fragments to form the full-length allele. For data analysis good quality (>Phred Phred 30) consensus sequence was used.

### Sequence Analysis

Alignment of assembled sequences and manual editing of blast resistance gene Pi9 was done by ClustalW (Thompson et al., 1994) and BioEdit Software version 7.0.9.0<sup>3</sup> . Pi9 gene sequence (DQ285630) was used as a reference for the prediction of gene coding regions by using Gene FGENESH<sup>4</sup> . The functional domain(s) which play an important role in mediating resistance were predicted using the online tools Pfam)<sup>5</sup> and SMART<sup>6</sup> . Phylogenetic analysis was performed with MEGA 4.0 (Tamura et al., 2007) using the Neighbor-Joining method (Saitou and Nei, 1987). All positions containing gaps and missing data were eliminated from the dataset (complete deletion option).

### Nucleotide Polymorphisms Analysis

Nucleotide polymorphism analysis of the aligned DNA sequences was done by DnaSP 5.10 program (Rozas et al., 2003).

<sup>1</sup>www.bioinfo.ut.ee

<sup>2</sup>www.chromous.com

<sup>3</sup>www.mbio.ncsu.edu

<sup>4</sup>www.softberry.com

<sup>5</sup>pfam.xfam.org

<sup>6</sup> smart.embl-heidelberg.de


TABLE 1 | Rice landraces sourced from NBPGR and their reaction to Magnaporthe oryzae at uniform blast nursery (UBN), Hazaribag selected for allele mining of Pi9 gene.

TABLE 2 | List of overlapping primers used for the amplificati1on of 2.9 kb nucleotide binding site-leucine-rich repeat (NBS-LRR) region of Pi9 gene using primer walking technique.


The Dna SP 5.10 program was used for the analysis of polymorphisms and Tajima's D test. The BioEdit software was used to calculate pairwise identity at DNA level. Synonymous and non-synonymous substitution (πsyn and πnon) were calculated to examine the selection at the NBS-LRR region of Pi9 gene.

### RESULTS

### Selection of Rice Landraces and Virulence Analysis

On the basis of pathotyping of 338 rice landraces at UBN, Hazaribag, six landraces comprising of three resistant and susceptible each, was selected for the allele mining of NBS-LRR region of Pi9 gene (**Table 1**). To further confirm their resistance and susceptibility, these rice landraces were along with the isogenic line for Pi9 gene IRBL9-w (control) were phenotyped with the mixture of virulent isolates discussed earlier (Imam et al., 2013a). The virulence analysis results showed that, out of six landraces, three were consistently resistant while three showed susceptibility to the mixture of virulent M. oryzae isolates (**Figure 1**). IRBL9-w, the isogenic lines for Pi9 gene was also given a resistant reaction.

### Sequence Characterization of the Pi9 Alleles

To determine the nucleotide diversity at the Pi9 allele, 2.9 kb long fragment were amplified from all the seven rice landraces by primer walking technique and sequenced (**Figure 1**). Only high-quality reads of the sequenced fragments were selected for analysis. About 99% (98%–100%) homology between the sequences was observed after pairwise alignment at the DNA level. Lower variation in the form of mutations (77) (Transition + Transversion), and InDels (59) was observed in the Pi9 alleles isolated from rice landraces selected. The phylogenetic tree was constructed based on the nucleotide sequences of seven rice landraces and one reference Pi9 (DQ285630) sequence (**Figures 2** and **3**). Phylogenetic analysis results in the formation of two groups, which clearly demonstrate the homology between the sequences. All the blast resistance Pi9 alleles were grouped into one cluster, whereas Pi9 alleles of susceptible landraces formed another cluster even though these landraces have a low level of DNA polymorphisms.

### Nucleotide Polymorphism of the Pi9 Alleles

A total number of 75 polymorphic sites were identified in the 2.9 kb sequence among all the Pi9 alleles using DnaSP program. Average pairwise nucleotide diversity (π) and silent Watterson's nucleotide diversity estimator (θw) over the Pi9 alleles were 0.01103 and 0.01011, respectively. The average number of nucleotide differences, k was 31.536 and θ (per site) from Eta was 0.01038. Low-diversified nucleotide diversity for Pi9 alleles was observed based on earlier published results. The results showed that the Pi9 alleles of the selected rice landraces were less variable, suggesting that these rice landraces would have been exposed to less number of pathotypes across the country. LRR region showed higher average nucleotide diversity than that of the NBS region and this clearly suggests the importance of LRR domain in the variation of the Pi9 alleles.

### Selection of Pi9 Alleles

We evaluated the neutral selection with the Tajima's D test to test the evolutionary selection dynamics of Pi9 alleles in the rice lanraces. Among the seven rice landraces, positive Tajima's D (0.33580) was observed, which signifies the balancing

selection among them, which is different from the earlier results (Tajima, 1989). The presence of less number of rare alleles may be the reason for the positive Tajima's D test. Average rates of non-synonymous and synonymous substitution (πsyn and πnon) were calculated to examine the selective patterns of Pi9 gene in the rice landraces. The synonymous (πsyn) and nonsynonymous (πnon) substitution in coding region as a whole were calculated in all the seven Pi9 alleles. In the coding region, the value of synonymous substitution (−0.43337) was less than the non-synonymous substitution (0.78808). The greater nonsynonymous substitution than the synonymous means that the coding region, mainly the LRR domain was under diversified selection. The Tajima's D ratio (Non-syn/Syn) was −1.81851 (<1), suggesting the low level of polymorphism in the coding regions of rice landraces. A haplotype distribution analysis was done for all the seven alleles to study mutations and polymorphisms. The study of sequence polymorphism leads to the identification of a total number of five (5) haplotypes (**Table 3**).

### DISCUSSION

The analysis of allelic variants of disease resistance gene imparts essential information regarding novel resistance gene generation and specificity. Earlier reports showed both higher as well as lower levels of sequence diversity at different R-gene locus (Yang et al., 2008). In this study, polymorphism of the Pi9 allele was investigated in seven rice landraces. Earlier study about the prevalence of high pathotypic diversity of the M. oryzae population of Eastern India and very rare compatibility of Pi9 gene in field evaluations with isolates from this region result in considering Pi9 gene further in rice landraces (Variar et al., 2009). Our earlier results of virulence analysis of 72 M. oryzae isolates against 26 differential variety revealed that matching virulence to all monogenic differentials carrying different resistant genes were present in the pathogen population, although the resistant check Tetep was resistant to all of them. The frequency of virulence on different monogenic lines ranged from 4.5 to 73%. Very low frequencies of isolates were virulent on Pi9 (4.5%) and Piz-5(Pi-2) (7%) followed by Pita-2 (16 and 18.2%) were reported (Alam et al., 2015). Therefore, complementary resistance spectra that exclude all the pathotypes of the pathogen are required for strategic resistant gene deployment. Pi9 and Pita-2 genes exhibited complementary resistance spectrum and excluded all the pathotypes of the pathogen. Therefore, Pi9 was taken into consideration for further study and analysis. The present result showed that the alleles of the rice landraces were mostly identical at the DNA sequence level, which further suggests the high level of conservation among Pi9 rice germplasms. A total number of 136 polymorphic sites comprising of transitions, transversions, and InDels were identified in the 2.9 kb sequence of Pi9 alleles. Simple InDels and Single nucleotide polymorphisms (SNPs) play a very important function in R-gene evolution (Shen et al., 2006). A single nucleotide difference in the regulatory region of Pi54 locus distinguishes resistant phenotype from the susceptible one (Sharma et al., 2005). The Pita gene when physically linked to a region called superlocus able to show resistance pattern (Jia and Martin, 2008; Lee et al., 2009). In cereal genomes, higher SNPs are detected in the in non-coding regions (one in 100–600 bp) (Gupta and Rustogi, 2004). Similarly, between O. sativa and O. rufipogon, the 26 kb region of DNA sequence showed higher variation (Rakshit et al., 2007). The results clearly showed 99% similarity and low polymorphism at the DNA level among all the seven rice landraces, however, presence of SNPs make it little variable at some regions. The low polymorphism in the DNA sequences of rice landraces reveals that these landraces are closely related were exposed to less number of pathotypes. Liu et al. (2011) demonstrated the intermediate level of polymorphism of the Pi9 alleles from 40 Oryza accessions of China are belonging to cultivate and wild species.

On the basis of a genome-wide analysis of allelic diversity in R-genes of the rice genome, four classes of diversification of R-genes are described (Yang et al., 2008). The present study

with seven rice landraces indicated that Pi9 allele belongs to type II category since it was neither highly conserved not highly diverse, even though it has low diversified alleles, similar to other blast resistance gene Pi54 (Thakur et al., 2015). Different studies showed that the rapid evolution of R-genes are driven by the high level of diversification (Type III and Type IV) and polymorphism (Shen et al., 2006; Yang et al., 2008; Liu et al., 2011). Pairwise allelic diversity, genomic organization, and the genealogical relationship among different genes have been the criteria to characterize the variation patterns which results in the categorization in four types of variation. Our studies also demonstrateed similar diversification of conserved (Type I; π < 0.5%), highly diversified (Type III;

FIGURE 3 | Phylogenetic tree of Pi9 allele based on nucleotide sequences of seven rice landraces along with one reference sequence.

TABLE 3 | List of identified haplotypes.


π > 0.5%), intermediated-diversified (Type II; π = 0.5−5%) and present/absent genes (Type IV) as previously published reports (Yang et al., 2008; Liu et al., 2011). Earlier study by Liu et al. (2011) suggest that both human and natural selection played a major role in evolutionary divergence of the Pi9 gene after the rice species differentiation. The allelic variation among the rice germplasm in the NBS-LRR region has increased our understanding of variation patterns. Earlier studies showed that variation level of R-gene was generally constant among the rice germplasms. This is now believed that the polymorphism content directly correlates to evolutionary changes (Shen et al., 2006; Yang et al., 2008). For the R-gene resistance specificity, LRR region is the major determinant which is largely responsible for the variation in the NBS-LRR genes (Collier and Moffett, 2009). It is also inetersting to note that among and within oryza species (wild and cultivated rice), LRR region showed more sequence variation than NBS region (Liu et al., 2011). Since Pi9 alleles showed Type II intermediate level of polymorphism, therefore, its evolution pattern is slow and intermediate during the course of time (Ding et al., 2007; Yang et al., 2008). The present study suggests the intermediate level of polymorphism in the Pi9 alleles which may be due to the mixed evolutionary pressure experienced by the gene during co-evolution of rice blast pathogen. In another study, among the cultivated rice, the Pita alleles showed the lowest rate of diversification as among other rice species (Lee et al., 2009). Low nucleotide variation was observed in the coding region (0.00067) of Pita alleles in US weedy rice as compared to non-coding regions (0.00161) (Lee et al., 2011). Interestingly, the phylogenetic analysis

showed that resistant and susceptible Pi9 alleles grouped into separate clusters. This is in line to Pi9 alleles wherein cultivated rice along with its ancestors clustered into one group and African cultivated rice along with its ancestors grouped into separate cluster, suggesting that same selection pressure has occurred in two groups during domestication and/or natural selection (Liu et al., 2011). Thakur et al. (2013) also demonstrated the grouping of resistant and susceptible Piz-t alleles in two sub-cluster. In R-gene evolution and development of resistance specificity, the LRR region plays the major role (Collier and Moffett, 2009). The present result also showed the high level of sequence variation in LRR region among the rice landraces.

### CONCLUSION

In R-gene evolution, balancing as well as positive selection has been observed and different test is used to calculate the selection pressure which drives the evolution of R-genes (Hudson et al., 1987; Tajima, 1989; McDonald and Kreitman, 1991). In this study, it appears to be balancing selection because of the minor positive Tajima's D test value, which is different from the earlier reports of Liu et al. (2011), which showed that the Pi9 gene is under positive selection. The reason for having positive Tajima's D test was the low nucleotide diversity within the rice germplasm.

### REFERENCES


This may be because of the closeness of the rice landraces, cultivated in the same region, and under low pathotype pressure. The Tajima's D ratio (Non-syn/Syn) is an indicative of selection pressure acting on the protein coding genes. Both balancing and purifying selections have been observed for the evolution of R-gene (Thakur et al., 2013).

### AUTHOR CONTRIBUTIONS

All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication.

### FUNDING

The authors acknowledge the National Agriculture Innovative Project- Component 4, Indian Council of Agricultural Research (ICAR), on Allele mining blast resistance genes for financial support for this research.

### ACKNOWLEDGMENT

This work was performed at the Central Rainfed Upland Rice Research Station, Hazaribag, Jharkhand, India.



confers race-specific resistance to the blast fungus. Genetics 176, 2541–2549. doi: 10.1534/genetics.107.075465


Till, B. J., Reynolds, S. H., Green, E. A., Codomo, C. A., Enns, L. C., and Johnson, J. E. (2003). Large-scale discovery of induced point mutations with high-throughput TILLING. Genome Res. 13, 524–530. doi: 10.1101/gr.977903

Valent, B. (1990). Rice blast as a model system for plant pathology. Phytopathology 80, 33–36. doi: 10.1094/Phyto-80-33


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Imam, Mandal, Variar and Shukla. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Chilli Anthracnose: The Epidemiology and Management

Amrita Saxena<sup>1</sup> , Richa Raghuwanshi <sup>2</sup> , Vijai Kumar Gupta<sup>3</sup> and Harikesh B. Singh<sup>4</sup> \*

*<sup>1</sup> Department of Botany, Banaras Hindu University, Varanasi, India, <sup>2</sup> Department of Botany, Mahila Mahavidyalaya, Banaras Hindu University, Varanasi, India, <sup>3</sup> Molecular Glycobiotechnology Group, Discipline of Biochemistry, National University of Ireland, Galway, Ireland, <sup>4</sup> Department of Mycology and Plant Pathology, Institute of Agricultural Sciences, Banaras Hindu University, Varanasi, India*

Indian cuisine is renowned and celebrated throughout the world for its spicy treat to the tongue. The flavor and aroma of the food generated due to the use of spices creates an indelible experience. Among the commonly utilized spices to stimulate the taste buds in Indian food, whole or powdered chilli constitutes an inevitable position. Besides being a vital ingredient of of Indian food, chilli occupy an important position as an economic commodity, a major share in Indian economy. Chilli also has uncountable benefits to human health. Fresh green chilli fruits contain more Vitamin C than found in citrus fruits, while red chilli fruits have more Vitamin A content than as found in carrots. The active component of the spice, Capsaicin possesses the antioxidant, anti-mutagenic, anti-carcinogenic and immunosuppressive activities having ability to inhibit bacterial growth and platelet aggregation. Though introduced by the Portuguese in the Seventeenth century, India has been one of the major producers and exporters of this crop. During 2010–2011, India was the leading exporter and producer of chilli in the world, but recently due to a decline in chilli production, it stands at third position in terms of its production. The decline in chilli production has been attributed to the diseases linked with crop like anthracnose or fruit rot causing the major share of crop loss. The disease causes severe damage to both mature fruits in the field as well as during their storage under favorable conditions, which amplifies the loss in yield and overall production of the crop. This review gives an account of the loss in production and yield procured in chili cultivation due to anthracnose disease in Indian sub-continent, with emphasis given to the sustainable management strategies against the conventionally recommended control for the disease. Also, the review highlights the various pathogenic species of *Colletotrichum* spp, the causal agent of the disease, associated with the host crop in the country. The information in the review will prove of immense importance for the groups targeting the problem, for giving a collective information on various aspects of the epidemiology and management of the disease.

Keywords: anthracnose, Capsicum spp., Colletotrichum capsici, epidemiology, disease management, biocontrol

### INTRODUCTION

Chilli (Capsicum annum L.) is one of the most important constituent of the cuisines of tropical and subtropical countries and the fourth major crop cultivated globally. Around 400 different varieties of chilies are cultivated throughout the globe. The hottest variety being "Carolina Reaper" developed by a grower Ed Currie of West Indies having the maximum pungency of about 2.2

Edited by:

*Gero Benckiser, University of Giessen, Germany*

#### Reviewed by:

*Susana Rodriguez-Couto, Ikerbasque, Spain Marcela Claudia Pagano, Universidade Federal de Minas Gerais, Brazil*

> \*Correspondence: *Harikesh B. Singh hbs1@rediffmail.com*

#### Specialty section:

*This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology*

Received: *23 June 2016* Accepted: *12 September 2016* Published: *30 September 2016*

#### Citation:

*Saxena A, Raghuwanshi R, Gupta VK and Singh HB (2016) Chilli Anthracnose: The Epidemiology and Management. Front. Microbiol. 7:1527. doi: 10.3389/fmicb.2016.01527* million SHU (Scoville Heat Units; PuckerButt Pepper Company, 2013). One of the hot chilli varieties of the world "Naga Jalokia," is the native of Tezpur in Assam, India. Numerous varieties of chilli are grown for vegetables, spices, condiments, sauces, and pickles occupying an indispensable position in Indian diet. Apart from the explicit importance of the crop in the diet, chilli is also used in other forms like medicines and beverages and also as an ornamental plant in the gardens. Nutrition wise these are enriched with high Vitamin A and C content; high iron, potassium, and magnesium content with the ability to boost the immune system and lower the cholesterol levels (Grubben and Mohamed El, 2004). India has been a leading producer, consumer and exporter of chilli especially in dried form. Various varieties of the crop are found in India and its quality varies among the states of the country.

### THE HOST CROP-CAPSICUM ANNUM L.

Known for over 9500 years, chilli is the native of Southern America and was first cultivated in Peru at around 7500 BC (MacNeish, 1964). It has been the first ever domesticated crop of America. During the course of evolution, three important species of Capsicum i.e., C. annuum, C. frutescens and C. chinense evolved from a common ancestor that grew wildly in the North of the Amazon basin (NW-Brazil, Columbia) spreading to the other parts of America. Initially, people cultivated them with other crops in order to protect their crop from the damage caused by birds. By the end of eighteenth century five species of chilli i.e., C. annuum L., C. baccatum L., C. chinense Jacq., C. frutescens L., and C. pubescens R. & P. were domesticated in different parts of the America (IBPGR, 1983).

Introduction of chilli to India is being credited to the voyage of Columbus, who brought the seeds from Spain, introducing it to Europe, which subsequently spread to Africa and Asia (Heiser, 1976). However, Columbus confused the pungent fruits of Capsicum with black pepper Piper nigrum L. calling them as red pepper owing to the red colored fruits. Capsicum is not related to the Piper genus. The popularity of the crop extended swiftly across Europe is moving to India, China and then to Japan. It was incorporated into all the Asian and European cuisines almost instantaneously unlike the other spices introduced from the Western Hemisphere. Since its introduction, it has been cultivated aggressively in all the tropical and sub-tropical countries.

In its course of evolution staring from Central America to the whole of Europe, Africa and Asia, copious terminologies have been used to name the pungent fruits. The phraseology in case of Capsicum is however very confusing. Names like pepper, chili, chile, chilli, aji, paprika are used to denote the pungent fruits of Capsicum. The term "Capsicum" is reserved for taxonomic discussion. The vernacular names include pepper, chile, and aji. The common term used to denote chilli in India is "Mirchi" in Hindi. Bell pepper generally refers to the blocky non pungent fruits of Capsicum known as "Shimla Mirch" in Hindi.

Belonging to the Solanaceae family, Capsicum genus consists of approximately 22 wild species and five domesticated species

FIGURE 1 | The botanical characteristics of Capsicum annum L., the chilli plant (A), the flower (B), immature green fruits (C) and the ripe red fruits (D).

(Bosland, 1994). Having chromosome number 2n = 24, Capsicum species may be herb or sub-shrub of height up to 2.5 m with extensively branched stem having hairy growth with purplish spots near the nodes. The tap root is strong with numerous lateral roots. Flowers are generally solitary, terminal, bisexual and pentamerous with campanulate to rotate corolla. Stamens are adnate at the base of the corolla tube with blue to purplish anthers. The ovary is superior having 2–4 chambers. Filiform style is found with capitate stigma. Chilli fruits are considered vegetable and are botanically berries (**Figure 1**).

Based on the fruit characteristics, i.e., color, pungency, shape, flavor, size, and their use, the types of chilli are classified (Bosland, 1992, 1996). It is a perennial crop and can be grown throughout the year. The major harvest season is between December and March with its supply reaching maximum during February-April (Bosland, 1996).

### USES AND IMPORTANCE

Chilli is used in all forms starting from fresh green fruits with ripe fruits along with its dried and powdered form. Fresh green pungent fruits are generally used in salads, stuffing, and as a flavoring agent in cooked meals. The non-pungent varieties are cooked as vegetables or processed with other food items for flavor (Welbaum, 2015). Very pungent varieties are consumed in small quantities generally considered as a condiment or spice for seasoning and for stimulating appetite. Hot peppers are also pickled in salt and vinegar used in ketchups as flavoring agents

(Grubben and Mohamed El, 2004). Apart from its extensive use in preparation of processed food, chilli varieties are also used as coloring agents in salad dressings, meat products, cosmetics, and even clothing.

Capsicum possesses various medicinal and nutritional values as well. It is interesting to quote that fresh green chilli fruits contain more Vitamin C than found in citrus fruits, while red chilli fruits have more Vitamin A content than as found in carrots (Osuna-García et al., 1998; Than et al., 2008). The active ingredient of the spice, capsaicin is a complex of capsaicinoid alkaloids found in variable concentration in different chilli varieties. It is found in abundance in the placental tissues and cross walls of the fruits. However, in very pungent fruits, it is distributed in all the fleshy parts of the fruit (Grubben and Mohamed El, 2004). The amount of capsaicin has been the measure of the pungency of the chilli variety and is generally expressed in Scoville Heat Units (SHU) (Scoville, 1912).

Capsaicin possesses the antioxidant, anti-mutagenic, anticarcinogenic, and immunosuppressive activities having the ability to inhibit bacterial growth and platelet aggregation. It is also used as anti-arthritic and anti-inflammatory agent. Regular consumption of chilli fruit is helpful against hemorrhoids, varicose veins, anorexia, and liver congestion. Pure and processed form of chilli extracts is used externally as rubefacient and analgesic in case of back pain, rheumatism, articular and muscular pains and swollen feet and even as antidote in case of poisoning. The two non-foods and non-pharmacological application of capsaicin are in the preparation of "pepper sprays" as weapons of self-defense and in preparation of "natural and organic pesticide" (Welbaum, 2015).

### GLOBAL PRODUCTION AND INDIAN SHARE

Being an important crop of tropical and sub-tropical countries, cultivated in large areas of Asia, Africa, South, and Central America and southern Europe, the total area under cultivation in the world is about 1.9 million hectares (Vanitha et al., 2013). About 42.2% (801,600 ha) cultivable land of the total area under chilli production at world level lies in India, producing approximately 21.4% (1.4 million t) of world's total chilli production (FAOSTAT, 2012). India has been the major producer as well as exporter of the dried chilli in the world (**Figure 2**). The other important producers of chilli include China, Mexico, Peru, Turkey, Thailand, Indonesia followed by countries in tropical Africa mainly Ghana and Ethiopia.

Chilli is a universal spice crop of India grown in almost all the states of the country. The quality of the chilli varies from state to state. For instance, chilli of Karnataka is known for its oil content, Gujarat quality is majorly known for its sharp color while that of Rajasthan is well known for making pickles. The chilli cultivated in Assam is famous for its pungency and that grown in Andhra Pradesh is mostly used as vegetables and liked by non-vegetarians. Major chilli producing states include Andhra Pradesh, Maharashtra, Karnataka, and Madhya Pradesh (Post harvest profile of chilli, 2009). Apart from being a large consumer and producer of chilli, India is also the largest exporter of the crop. Over 30% of the chilli produced in India is exported to countries of West Asia, East Asia, USA, Sri Lanka, and Bangladesh, most commonly in dried form (FAOSTAT, 2012).

### CONSTRAINTS IN CHILLI PRODUCTION

Though being an important spice crop grown worldwide, many constraints decrease production, causing significant reduction in yield and seed production. Plant diseases have been a major reason for the crop losses worldwide. Diseases caused by fungi, bacteria, viruses or nematodes have adversely affected chilli production in almost all parts of the world. A summary of the various diseases associated with the host crop chilli is outlined in **Table 1**.

Other than the losses due to the pests and pathogens, crop loss in post-harvest conditions further add in delimiting the yield and production of the crop (Prusky, 2011). Specifically, in developing countries, the post-harvest losses are more serious owing to poor transportation and storage facilities (Sharma et al., 2009). Moreover, due to the recent strengthening of the food security norms in the international markets, the trade of food

#### TABLE 1 | Diseases of chilli reported from different parts of the world.


#### *(Continued)*

#### TABLE 1 | Continued


contaminated with fungal toxins, mycotoxins has been declared unhealthy for human consumption (WHO, 2002). Species belonging to Aspergillus genus like A. flavus (Bankole et al., 2004) and A. parasiticus (Garcia-Villanova et al., 2004) have been majorly kept responsible for production of harmful aflatoxins B and G, whose contamination in food commodities has led to serious health problems in human, like cancer and aflatoxicosis (Jeffrey and Williams, 2005). Mycotoxin contamination in dried chillies has limited its export to major export destinations like the UK and USA. Post harvest loss due to aflatoxin contamination in chilli has been reported to be about 20% to about 100% of samples obtained from Turkey (Demircioglu and Filazi, 2010) and Malaysia (Reddy et al., 2011).

Among the large number of diseases affecting chilli cultivation, anthracnose disease caused by Colletotrichum species, bacterial wilt by Psuedomonas solanacearum and viral diseases like chilli veinal mottle virus (CVMV) infection and cucumber mosaic virus (CMV) infection have been most detrimental to chilli production (Than et al., 2008).

### ANTHRACNOSE DISEASE OF CHILLI

Anthracnose disease has been reported to be a major constraint in chilli production in tropical subtropical countries causing huge losses. India had been the largest producer and exporter of chilli, but since a few years the production has declined significantly and presently, India stands at the third number in terms of chilli production (FAOSTAT, 2012; **Figure 3**). An estimated annual loss of about 29.5%, amounting whopping figure of US\$ 491.67 million has been reported from India alone (Garg et al., 2014). In India, a calculated loss of 10–54% has been reported in yield of the crop due to the anthracnose disease (Lakshmesha et al., 2005; Ramachandran and Rathnamma, 2006). Significant losses have been reported from other parts of the world as well, like a significant amount of 20–80% loss has been accounted from Vietnam (Don et al., 2007) and about 10% from Korea (Byung, 2007). The loss is high owing to the post and pre harvest involvement of the pathogen causing a loss of 10–80% of the marketable yield of chilli fruits (Than et al., 2008).

The fruit lesion is the most economically important aspects of the disease as sometimes, even a small lesion on the fruit is enough to lower its market value thereby affecting the profitable yield of the crop (Manandhar et al., 1995). The disease is reported to affect almost all aerial parts of the plant. Chiefly, it causes fruit rot at both green and red stages primarily attacking ripe fruits, hence is also known by the name ripe fruit rot of chilli (Agrios, 2005). The disease is seed borne, soil borne, water borne and airborne and hence may lead to damage at the seedling stage or on the aerial parts of the plants. Many species of Colletotrichum have been associated with the pepper anthracnose in different countries. **Table 2** gives the brief outline of the different Colletotrichum species reported to be associated with the anthracnose of chilli in different parts of the world. However, in India, primarily three important species, namely, C. capsici, C. acutatum and C. gleosporoides have been reported to be linked with the disease, with C. capsici Syd. Butler and Bisby causing major damage at the ripe fruit stage of the plant (Ranathunge et al., 2012; Saxena et al., 2014).

### EPIDEMIOLOGY AND DISEASE SYMPTOMS

Environmental factors play an important role in deciding the severity and spread of any disease. The favorable host, pathogen and weather conditions lead to establishment of disease (Agrios, 2005). Thus, before proposing the management strategy of the disease, a thorough knowledge regarding the epidemiology of the disease should be studied. Anthracnose disease of chilli is generally most common among the tropical and sub-tropical countries. Hot and humid environmental conditions support the spread of the disease.

Other important environmental factors governing the severity of the disease include rainfall intensity and duration, humidity, leaf surface wetness and light. Amongst them

leaf surface wetness has been directly linked with the severity of the disease owing to the better establishment of the pathogen in respect of germination, attachment and penetration into host tissues (Than et al., 2008). The relationship between the environmental factors like rainfall intensity and duration and the prevailing temperature and humidity along with the crop geometry and inoculum spread leads to possible development of disease as well (Dodd et al., 1992). Temperature also affects the development of the disease and presence of surface wetness and competitive microbiota further favors the disease development (Royle and Butler, 1986). Temperature around 27◦C with relative humidity of 80% have reported to be the most optimum conditions for successful establishment of the disease in a given area (Roberts et al., 2001). The development of the disease also depends on the host cultivar, along with its resistance against the pathogen.

On attainment of favorable conditions, characteristic symptoms appear on the ripe chilli fruit, which appear as sunken circular or angular lesions (**Figure 4A**). Often multiple lesions coalesce to form severe fruit rot. Generally, the lesions are characterized by the presence of black colored spots in concentric rings at maturity. Initially, orange to pink conidial masses may be visible on the fruit surface. The dark spots when observed under a microscope are the acervuli structures containing setae hairs entrapping the conidia of the pathogen. Further, the pathogen forms micro sclerotia in plant debris or seed, soil, which is the mode of survival under unfavorable conditions. The pathogen infects all parts of the host plant, including stems and leaves (**Figures 4B,C**). Lesions on stems and leaves appear as small sunken grayish brown spots with dark margins, further on which development of acervuli in concentric rings could be easily seen.

### THE PATHOGEN-COLLETOTRICHUM CAPSICI

Colletotrichum spp. has been rated as among the ten most notorious pathogens in the world, causing heavy crop losses worldwide (Dean et al., 2012; **Figure 5**). Specifically, Colletotrichum is an asexual genus belonging to phylum Ascomycete and Coeleomycetes class of Fungi imperfectii (Dean et al., 2012). Despite significant developments in studies related to this plant-patho system, the taxonomic position of the pathogen remains unclear. The systematics of the fungal pathogens from this genus still exist in ambiguity with the number of species ranging from 29 to over 700 depending upon the criteria selected for separation (Sutton, 1992).

Being economically important pathogen, its host range varies from fruits, vegetables, ornamental plants to important staple food crops. The species of this genus are reported to cause anthracnose disease in more than 121 plant genera from 45 different plant families (Farr et al., 2016). It also causes blights of aerial plant parts and post-harvest rots. The damage may extend to severe economic loss in tropical and sub-tropical countries, causing infection to staple foods like bananas, sorghum, cassavas, legumes, and cereals (Bailey and Jeger, 1992). Particularly, this pathogen exhibit efficient infection in post-harvest conditions owing to its ability to cause latent infection, where the symptoms appear on the fruits only after the harvest or during storage or at the market shelf. Losses up to 100% have been recorded due to Colletotrichum spp. (Prusky, 1996). Another important parameter for its successful colonization and severe disease spread may be credited to its cosmopolitan nature. Many species of Colletotrichum may be found on a single host or single species may be able to infect different hosts (Sander and



Korsten, 2003). Broad, imprecise and often overlapping fungal plant relationships exist in Colletotrichum plant-patho system (Freeman and Shabi, 1996). With its ability to infect many hosts along with adapting to new environments, the pathogen poses serious threat to the different crop production system through cross infection problems (Photita et al., 2005). **Figure 6** gives the distribution of the sexual teleomorph of the Colletotrichum in different parts of the world.

Impact on global economic loss posed by the pathogen has triggered extensive studies on diverse aspects of the biology of the pathogen for better understanding of its infection process and host interaction mechanisms. In light of this, host specificity of the pathogen (Freeman, 2000; Correll et al., 2007) along with the biology involved in the infection mechanisms used by the pathogen (Perfect et al., 1999; O′Connell et al., 2000) and the various fungal-host interactions (Prusky et al., 2000) studies have been carried out. Studies related to genetic diversity and the epidemiology have also been reported (Freeman, 2000; Timmer and Brown, 2000). The genus has been used as a model for studying the genetic basis of symbiotic life styles (Rodriguez, 2000) leading to the development of infection and disease forecasting systems (Uddin et al., 2002). The use of molecular markers like DNA fragments analysis [e.g., Randomly

amplified polymorphic DNA (RAPD) and Arbitrarily primed (AP)-PCR] has improved the speed and accuracy in identification and characterization of Colletotrichum spp. (Lewis et al., 2004; Photita et al., 2005). Further the the nucleotide sequenceof the 5.8S gene and internal transcribed spacer (ITS) region has facilitated the construction of Colletotrichum species specific primers providing a rapid and accurate method for diagnostic purpose and phylogenetic analysis (Torres-Calzada et al., 2011). Many studies have been carried out to resolve the issue of species complex in Colletotrichum spp. in different areas of the world on various hosts like mango (Kamle et al., 2013), guava (Mohd Anuar et al., 2014), herbaceous plants (Photita et al., 2005), soursop (Álvarez et al., 2014), and also in medicinal plants as endophytes (Lima et al., 2012) using molecular marker analysis.

The anthracnose causing pathogen in chilli varieties have been reported to be Colletotrichum capsici (Sydow), Butler and Bisby (Than et al., 2008). Three pathotypes have been linked with the infection on ripe fruits, while two have been reported causing infection at mature green fruit stages (Mongkolporn et al., 2004; Montri et al., 2009). The convoluted relationships of the fungi with the host have left certain gaps in the knowledge of the infection process and the interactions of different species with chilli plant. Different species have been reported to be linked causing infection in different parts of the chilli plant, for instance, C. acutatum and C. gloeosporioides infect the fruits at

characteristic structures, i.e., conidia (B,C), setae (D), acervuli (E), and the acervuli as seen on surface of the chilli fruits (F).

all stages of development without causing much harm to leaves and stems of the plant, while C. coccodes and C. dematium mostly cause high infestation on leaves and aerial parts of the plant (Kim et al., 2004). Differential rates of infestation has also been reported by the species of the genus infecting different fruits of the plant. For instance, in Korea, Hong and Hwang (1998) reported C. gloeosporioides as the most prevalent species causing an infestation in chilli fruits at both ripe stages and unripe stage; while existence of C. capsici and C. acutatum as the prevalent species infectionsnfection on ripe and unripe chilli fruits respectively, in the North eastern region of India have been reported (Saxena et al., 2014).

### INFECTION STAGES AND DISEASE CYCLE

Colletotrichum employs different strategies for causing infection to the host plant which initiate from the intracellular hemibiotrophic mode to the intramural necrotrophic mode of nutrition (Bailey and Jeger, 1992). Liao et al. (2012) has reported an intermediate stage showing partial endophytic life style of the pathogen before adapting to the necrotrophic mode of nutrition in the host plant. Different species of this genus exhibit different mechanism of infection depending on the host infected. For instance, Peres et al. (2005) reported the epiphytic or endophytic mode of survival of C. acutatum in an orchard infected with the bitter rot of apple. Also, intramural necrotrophy by C. capsici was reported by Pring et al. (1995) while infecting cowpea leading to subsequent necrosis caused due to dissolution of cell wall structures. The biotrophic phase of infection by C. capsici is also well studied in the infection caused to broad bean or lentil characterized by the presence of large multilobed, multi septate, vesicular primary hyphae (Latunde-Dada and Lucas, 2007).

Initial infection starts with the attachment of the conidia to the host surface preceded with its germination and production of adhesive appresoria followed with its penetration into the host epidermis. This is further accompanied by the growth and colonization of plant tissue by the fungus, resulting in the formation of specific symptom structures that is acervuli containing the spores of the fungus for further spread (Prusky et al., 2000). The pathogen sometimes remains in quiescent state in the form of appresorial structures in tissues of unripe fruits and cause infection after the fruits ripe or mature (Than et al., 2008). A dendroid structure composed of multiple, thick-walled hyphal branches with swollen or sharp ends from the penetration pore of the appresorium has been recently reported to be an intermediate structure which penetrates the host cuticle layer and infect the epidermal cells during C. acutatum infestation in chilli (Liao et al., 2012).

Though the pre-penetration mechanisms exhibited by Colletotrichum species appear somewhat related to each other, the post penetration events such as spore adhesion, melanization and cutinization hold certain disparity. Based on the previous studies, four kind of infection strategies with varied hosts have been studied in C. acutatum plant patho system (Peres et al., 2005). First is the biotrophic growth of the pathogen, where

the formation of appresoria from the conidia is followed by the formation of secondary conidia which further infects and spreads the pathogen inside the host leaves (e.g., The biotrophic disease cycle in citrus leaves). The second is the subcuticular intramural necrotrophy with the development of wide and swollen hyphae in the anticlinal and periclinal walls of host epidermal cells (e.g., The necrotrophic disease cycle on strawberry). The third strategy is the hemibiotrophic mode of infection where the pathogenic hyphae interact with the infection vesicles within the host cells (e.g., The biotrophic disease cycle on blueberry fruits). The fourth type of interaction is the combination of hypertrophic and subcuticular intra and intracellular development of the pathogen generally observed during infestation of almond leaves and fruits.

As far as studies related to infection and colonization by Colletotrichum species i.e., C. gloeosporioides on susceptible chilli variety is considered, no biotrophic stage in the form of infection vesicle has been found during the infection (Kim et al., 2004). An increased number of small vacuole with the condensed cytoplasm in the epidermal cells followed with cell destruction extending to the sub epidermal cells of the plant due to the action of pathogen enzyme has been noticed during the early stages of infection in chilli plant. During the later stages, inter and intra cellular colonization of the pathogen occurs indicating the governance of necrotrophic mode of fungal growth. **Figures 7A,B** shows the diagrammatic representation and the microscopic representation of different stages of infection of Colletotrichum spp.

## DISEASE MANAGEMENT

Management of chilli anthracnose has been a burning issue for the agriculturists and the farmers as till date, no effective control measures has been proposed. The fall in the chilli production and the drop in fruit quality have further intensified the need for developing a sustainable approach for controlling the spread of the disease. No single management technique has been found to efficiently control the disease. Generally, using a combination of the different strategies like chemical control, biological control, physical control and intrinsic resistance has been recommended for managing the disease (Agrios, 2005). The management strategies for controlling Colletotrichum spp. from spreading and establishing a disease can be discussed under four broad categories: Use of cultural Practices, use of chemical control, use of resistant varieties and finally the use of biological control. **Table 3**, gives a summarized information on the strategies used

for controlling the anthracnose disease in chilli from different parts of the world.

stages of infection by the *Colletotrichum* spp. on chilli leaf as seen under microscope.

### Use of Cultural Practices

The pathogen being seed borne, wind borne and water borne apart from being soil borne, the practices to control its spread should target three main areas of disease free crop production in the field: proper drainage, crop rotation and removal of any infected plant parts of the field. Water splashes may easily spread the conidia of the pathogen from infected to uninfected plant parts. Also, relative humidity aids successful colonization of the pathogen. So, the field should have proper drainage and irrigation to prevent the outbreak of the disease. Also, proper distance between the plants should be maintained so as to reduce dense canopy, which gives way for creating moisture (Than et al., 2008). Another important practice is the use of transplants raised from disease free seeds of the chilli variety. The transplants should be kept weed free and away from other solanaceous crops. Ideally, the crop should be rotated after every 2–3 years with crops those are not the host of Colletotrichum (Roberts et al., 2001). The use of

#### TABLE 3 | Control measures for managing chilli anthracnose reported from different parts of the world.


*(Continued)*

#### TABLE 3 | Continued


rice straw and plastic mulches has also been reported for effective control of the disease (Vos et al., 1995).

### Use of Chemical Fungicides

In the absence of any accurate method of controlling the disease, chemical control has been sought as the most effective measure to control the spread of the disease. The longer time required for developing the resistant cultivar and the short span result of the use of fungicides further popularize this method of controlling the disease specifically for anthracnose disease (Wharton and Diéguez-Uribeondo, 2004). However, the remnant toxic residues of the chemicals in the fruits create hindrance to the expected export of the chilli products to other countries, in turn affecting the economy of the country. Also, relying on a single chemical component result in the development of resistance in the pathogenic isolates, which further augments the difficulty in the management of the disease (Staub, 1991; Than et al., 2008). Traditionally, recommended fungicide for the control of the disease is manganese ethylenebisdithiocarbamate (Maneb) (Smith, 2000) and carbendazim, though the use of both fungicides has been found ineffective under severe disease outbreak. The chemical fungicides generally recommended for controlling anthracnose disease are based on copper compounds, dithiocarbamates, benzimidazole and triazole compounds (Waller, 1992). Newer chemicals like strobilurins based fungicides (e.g., azoxystrobin, pyraclostrobin) have also been used for its management. However, only a few reports are available using this class of fungicide controlling chilli anthracnose under large field trials (Schilder et al., 2001; Lewis and Miller, 2003; Chen et al., 2009).

The effective control through the use of chemical fungicides is possible by the timely application during the critical period favorable for the onset of the disease. Generally, fungicides should be applied at young expanding tissues, including fruits, leaves and flowers to restrict the entry of the pathogen to the plant system (Wharton and Diéguez-Uribeondo, 2004). However, numerous reports on the destructive effects of the use of fungicides on farmers' health, economic status, and toxic contamination of the environment, particularly in developing countries cannot be ignored (Voorrips et al., 2004; Garg et al., 2014). Different classes of fungicides have specific mode of action along with their duration of effect on disease control. So, wise choice of fungicides by the farmers in a particular area, according to prevailing environmental conditions, should be taken into consideration. Rotation of two or more different classes of fungicides is highly recommended for increasing the chance of better protection against the disease in the fields (Förster et al., 2007).

### Use of Resistant Varieties

Developing resistance against the pathogen in the host is seeking to be the most important and sustainable approach for managing the disease. This strategy not only eliminates the losses caused due to the disease, but also remove the chemical and mechanical expense of the disease control (Agrios, 2005). The principle behind the use of resistant cultivars is to trigger the host defense response that in turn would inhibit or retard the growth of the pathogen involving the use of a single gene pair: a host resistance gene and the pathogen avirulence gene (Flor, 1971). In lieu of the existing biotechnological approach to manage diseases, certain successful resistant varieties of chilli against C. capsici have been reported from different parts of the world (Yoon, 2003; Voorrips et al., 2004; Garg et al., 2014). Though, not much success has been sought in developing resistant chilli varieties in the species Capsicum annum L., which is the only species grown worldwide (Park, 2007). The two major requirements before proceeding for developing the cultivar is the knowledge of the resistant varieties of Capsicum occurring wildly in the region and the different pathotypes of the pathogen found in that region. Many varieties resistant to Colletotrichum spp. and information regarding the pathotypes of the pathogen has been reported and is available AVRDC, 2003; Babu et al., 2011). However, the challenging task of resistant breeding is exceptionally difficult in Colletotrichumchilli pathosystem due to the association of more than one species of the pathogen with the disease (Sharma et al., 2005; Saxena et al., 2014) along with the differential ability of the pathogenic virulence (Montri et al., 2009).

Recently, a study carried by Garg et al. (2014) reported the existence of nine resistant varieties (BS-35, BS-20, BS-28, Punjab Lal, Bhut Jolokia, Taiwan-2, IC-383072, Pant C-1, and Lankamura Collection) of Capsicum spp. out of the 42 varieties existing in use in the area which could be employed for developing successful resistant cultivars through breeding programs. The information on the resistance varieties against Colletotrichum may also be utilized for studying the inheritance of the resistance from one generation to another (Kim et al., 2008) and also to locate and study the quantitative trait loci (QTLs) maps for resistance (Lee et al., 2010).

### Use of Botanicals and Biological Control Agents

Disease control through the use of botanicals and crude extracts of medicinal plants have been explored in recent years for their effective antifungal and antimicrobial properties. Their easy decomposition, non-residual activity and non phytotoxic properties further popularize their use for controlling phytopathogens. Several studies using crude plant extracts have also been conducted to access the control of Colletotrichum spp. on chilli (Ngullie et al., 2010; Johnny et al., 2011). They have shown different degree of effectiveness of extracts of sweetflag (Acorus calamus L.), palmrosa (Cymbopogon martini) oil, Ocimum sanctum leaf extract, neem (Azadirachta indica) oil, garlic, Piper betle L., Coleus aromaticus, plucao, and sabsua against the pathogen growth and spore germination.

Biocontrol strategy for disease management has stood up as a sustainable approach required for restoring the lost homeostasis of the environment. Though, for managing chilli anthracnose this particular strategy has not gained much momentum yet, the potential of using biocontrol agents (BCAs) for controlling the pathogen was elucidated way back by Lenné and Parbery (1976). The possibilities of using BCAs for controlling the post harvest loss of fruits has been illustrated by Jeger and Jeffries (1988) and Korsten and Jeffries (2000). Till date, the BCAs used for studying the antagonistic potential against Colletotrichum spp. affecting chilli crop include Psuedomonas fluorescens, Trichoderma spp., Bacillus subtilis, Candida oleophila, and Pichia guilliermondi (**Table 3**).

### Trichoderma As a Biocontrol Agent Against Colletotrichum spp.

Belonging to class Ascomycete, Trichoderma is a well-studied ubiquitous genus. Well known as a saprophytic fungus, it has high adaptive potential as evident from its ability to colonize wood, bark, agricultural wastes and other substrates apart from its omnipresence in a variety of soil types (Singh et al., 2012; Mukherjee et al., 2014). Its biocontrol potential has been well established against numerous important phytopathogens like Alternaria, Colletotrichum, Phytophthora, Pythium, Rhizoctonia, Sclerotinia, Verticillium etc. (Begum et al., 2008; Imtiaj and Lee, 2008; Jain et al., 2012; Singh et al., 2013). The mechanisms involved have been attributed to be mycoparasitim, antibiosis, competition for nutrients and space along with its ability to induce systemic resistance in the



plants against the pathogens (Harman, 2006; Shoresh et al., 2010; Hermosa et al., 2012). Also, the efficient enhancement in plant growth with significant increase in biomass has also been attributed to the application of Trichoderma species. (Yedidia et al., 2001; Jain et al., 2012). Recently, the ability of the fungus to induce biotic tolerance in plants by enhancing the mechanical strength of the plant system has been studied against phtytopathogenic infestation (Singh et al., 2013; Saxena et al., 2015).

Specifically, in the Colletotrichum plant pathosystem, its potential has been elucidated owing to its fast colonizing ability and mycoparasitic nature, which results in coiling and parallel growth of the pathogen (Begum et al., 2008; Živkovic et al., 2010). This property has been further attributed due to the secretion of extracellular enzymes, including glucanases, chitinases etc. that degrade the pathogenic mycelia thereby restricting its growth and further colonization in the host tissue (Harman, 2006; Vinale et al., 2008; Singh et al., 2012). The various studies reporting the application of Trichoderma species for the biological control of Colletotrichum in different host have been summarized in **Table 4**.

The mode of use of this fungus has been restricted to seed biopriming or root treatment of the plants. The effect of the foliar sprays of Trichoderma species to prevent spread of foliar disease has not been studied extensively. However, effective results have been reported by the foliar sprays of other antagonistic microbes like yeast, Psuedomonas, Bacillus etc. that managed to control the growth and colonization of the pathogenic fungus (Chanchaichaovivat et al., 2007; Anand et al., 2009; Sutarya et al., 2009). The reports have very well elucidated the enhanced efficiency of the microbes to combat the growth of pathogen at leaf and fruit surface acting as a first line of defense for the protection of the host plants. Similar approach for efficient control of foliar disease of Capsicum i.e., anthracnose could be proposed by using Trichoderma strains as well. There have been reports showing effective potential of Trichoderma species in controlling Colletotrichum infestation in other hosts like cowpea (Adebanjo and Bankole, 2004). Also report on effective colonization of Trichoderma species in the phylloplane of plants is well studied (Bae et al., 2011). In order to control chilli anthracnose very few attempts have been made to use Trichoderma isolates obtained from the phylloplane of healthy host plant. The need for an effective all around protection of the plant triggered our group to study the effectiveness of Trichoderma isolates dwelling at the phylloplane of the healthy plants to concur the pathogenic ingression. Recently, Trichoderma isolates from the phylloplane of healthy leaves were found equally successful in controlling disease incidence as evident from the elevated induction of defense related enzymes and reduced disease incidence on host plants (Saxena et al., 2016a).

### FUTURE PROSPECTS

Though the epidemic nature of the disease has been studied for ages, many areas are still unexplored in terms of host-pathogen interaction, its spread and effective management strategies. There lies an urgent need to develop an efficient integrated management strategy keeping in concern the different environmental factors and pathogenic resistance, driving the successful colonization of the pathogen in the host tissues. An insight into the pathogen's lifestyle would provide valuable information required to develop targets for developing resistant varieties of chilli against the pathogen. Also, modifications in conventionally recommended cultural practices suiting to a particular agroclimatic region will prove helpful in better management of the disease. More studies are required for acquiring in-depth information regarding various modes of infection by the pathogen and the pathogenic variability associated within a region with the post-harvest as well as pre-harvest loss in the crop production. The overall knowledge about the key aspects of a disease triangle will enable better management of the disease keeping track of the quality and quantity of the crop produced thereby contributing efficiently to the country's economy.

### REFERENCES


### AUTHOR CONTRIBUTIONS

AS has drafted the Manuscript; RR, HS, and VG have critically reviewed the draft for important intellectual content and provided substantial contribution for the concept and design of the Manuscript.

### ACKNOWLEDGMENTS

AS is grateful to Department of Science and Technology, Govt. Of India for providing INSPIRE Fellowship under the AORC Scheme.

using multiple mechanisms. Mol. Plant Microbe Interact. 24, 336–351. doi: 10.1094/MPMI-09-10-0221


FAOSTAT (2012). Production Data. Available online at: http://faostat.fao.org/


and HPLC analysis of aflatoxins and ochratoxin A in Spanish bee pollen. J. Agric. Food Chem. 52, 7235– 7239. doi: 10.1021/jf048882z


within the fruit cuticles of Capsicum spp.. Plant Pathol. 61, 262–270. doi: 10.1111/j.1365-3059.2011.02523.x


producing Bacillus vallismortis strain BS07. Biol. Control 65, 246–257. doi: 10.1016/j.biocontrol.2013.02.002


eds D. Prusky, S. Freeman, and M. B. Dickman (St. Paul, MN: The American Phytopathological Society), 232–244.


from grasses. Arch. Phytopathol. 46, 1641–1666. doi: 10.1080/03235408.2013. 771901


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Saxena, Raghuwanshi, Gupta and Singh. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# *Colletotrichum higginsianum* Mitogen-Activated Protein Kinase ChMK1: Role in Growth, Cell Wall Integrity, Colony Melanization, and Pathogenicity

Wei Wei <sup>1</sup> , Ying Xiong<sup>2</sup> , Wenjun Zhu<sup>3</sup> \*, Nancong Wang<sup>1</sup> , Guogen Yang<sup>4</sup> and Fang Peng<sup>3</sup>

*1 Institute for Interdisciplinary Research, Jianghan University, Wuhan, China, <sup>2</sup> Hefei Inzyme Information Technology Co., Ltd., Wuhan, China, <sup>3</sup> College of Biology and Pharmaceutical Engineering, Wuhan Polytechnic University, Wuhan, China, <sup>4</sup> The Provincial Key Lab of Plant Pathology of Hubei Province, College of Plant Science and Technology, Huazhong Agricultural University, Wuhan, China*

#### *Edited by:*

*Kumar Krishnamurthy, Tamil Nadu Agricultural University, India*

#### *Reviewed by:*

*Biswapriya Biswavas Misra, University of Florida, USA Pratyoosh Shukla, Maharshi Dayanand University, India*

> *\*Correspondence: Wenjun Zhu 82862108@qq.com*

#### *Specialty section:*

*This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology*

*Received: 08 May 2016 Accepted: 20 July 2016 Published: 03 August 2016*

#### *Citation:*

*Wei W, Xiong Y, Zhu W, Wang N, Yang G and Peng F (2016) Colletotrichum higginsianum Mitogen-Activated Protein Kinase ChMK1: Role in Growth, Cell Wall Integrity, Colony Melanization, and Pathogenicity. Front. Microbiol. 7:1212. doi: 10.3389/fmicb.2016.01212* *Colletotrichum higginsianum* is an economically important pathogen that causes anthracnose disease in a wide range of cruciferous crops. To facilitate the efficient control of anthracnose disease, it will be important to understand the mechanism by which the cruciferous crops and *C. higginsianum* interact. A key step in understanding this interaction is characterizing the mitogen-activated protein kinases (MAPK) signaling pathway of *C. higginsianum*. MAPK plays important roles in diverse physiological processes of multiple pathogens. In this study, a Fus3/Kss1-related MAPK gene, *ChMK1*, from *C. higginsianum* was analyzed. The results showed that the Fus3/Kss1-related MAPK ChMK1 plays a significant role in cell wall integrity. Targeted deletion of *ChMK1* resulted in a hypersensitivity to cell wall inhibitors, reduced conidiation and albinistic colonies. Further, the deletion mutant was also unable to form melanized appressorium, a specialized infection structure that is necessary for successful infection. Therefore, the deletion mutant loses pathogenicity on *A. thaliana* leaves, demonstrating that *ChMK1* plays an essential role in the early infection step. In addition, the *ChMK1* deletion mutant showed an attenuated growth rate that is different from that of its homolog in *Colletotrichum lagenarium*, indicating the diverse roles that Fus3/Kss1-related MAPKs plays in phytopathogenic fungi. Furthermore, the expression level of three melanin synthesis associated genes were clearly decreased in the albinistic *ChMK1* mutant compared to that of the wild type strain, suggesting that ChMK1 is also required for colony melanization in *C. higginsianum*.

Keywords: *Colletotrichum higginsianum*, MAPK, pathogenicity, cell wall integrity, growth rate

## INTRODUCTION

The hemibiotrophic fungal pathogen Colletotrichum higginsianum causes anthracnose disease. The disease typically manifests as small, dark, discrete spots or water-soaked, sunken lesions on the leaves, stems, or fruits of a wide range of cruciferous plants, including the model plant Arabidopsis thaliana (Narusaka et al., 2004; O'Connell et al., 2004; Birker et al., 2009; Ushimaru et al., 2010). C. higginsianum has emerged as an attractive model for studying both fungal pathogenicity and plant immune responses due to the availability of genomic and transcriptomic databases (O'Connell et al., 2012; Gan et al., 2013), and the ease of genetic manipulation of both the host plant and the pathogen (O'Connell et al., 2004; Zhang et al., 2006; Huser et al., 2009; Narusaka et al., 2009).

By recognizing host physical and chemical cues, C. higginsianum conidia differentiate melanized appressorium, an infection structure, at the tips of conidial germ tubes. Appressorium formation is required for successful infection since the fungus penetrates the cuticle and plant cell wall by utilization of enormous turgor pressure in melanized appressoria for further invasive growth (O'Connell et al., 2012). Thus, inhibition of melanized appressorium formation will facilitate the efficient control of anthracnose disease.

The infection-related morphogenesis and invasive growth of several appressorium-forming pathogens are regulated by many signal transduction pathways, especially the mitogenactivated protein kinase (MAPK) pathway (Xu and Hamer, 1996; Lev et al., 1999; Takano et al., 2000; Kojima et al., 2002; Bruno et al., 2004). To date, the MAPK cascades and MAPK signaling pathways are known to be involved in many major cell processes in fungi, such as stress responses, vegetative growth, pathogenicity, secondary metabolism, and cell wall integrity (Zhao et al., 2007; Turrà et al., 2014; Qi et al., 2015). CoMEKK1, a homolog of MAPKKK STE11, the up-stream regulator of Fus3/Kss1-related MAPK pathway, is essential for appressorium development and pathogenicity in C. orbiculare (Sakaguchi et al., 2010). Moreover, the CoIra1 in C. orbiculare is also involved in infection-related conidial germination and appressorium formation by proper regulation of Fus3/Kss1-related MAPK signaling pathways through CoRas2 on the up-stream (Harata and Kubo, 2014). In Magnaporthe oryzae, phosphodiesterase MoPdeH interacts with MoMck1, and functions upstream of the MoMck1–MoMkk1–MoMps1 MAPK pathway to regulate cell wall integrity (Yin et al., 2016). M. oryzae TRX2 interacts with Mst7, thus regulating the activation of Pmk1 MAPK via the Mst11-Mst7-Pmk1 MAPK pathway. Deletion of the TRX2 gene caused pleiotropic defects in conidiation, growth, responses to stresses, and plant infection progression (Zhang et al., 2016). By phosphorylation on MAPK Fmk1, Fusarium oxysporum Fbp1 regulates virulence, cell wall integrity, and invasive growth via the Fmk1 signal pathway (Miguel-Rojas and Hera, 2016). MAPKs related to the yeast Slt2, such as SLT2-type MAPK protein PsMPK1, and PsMPK7 from Phytophthora sojae are also important for hyphal growth, cell wall integrity, stress tolerance, ROS detoxification, and pathogenicity (Li et al., 2014; Gao et al., 2015). Mutation of three MAPK genes FoSlt2, FoMkk2, and FoBck1 respectively led to attenuated virulence and slower growth rate in F. oxysporum (Ding et al., 2015). The MAPK AaSLT2 in Alternaria alternate regulates conidiation, virulence, and melanin production (Yago et al., 2011). Therefore, these findings suggest significant roles for MAPK signaling pathways in multiple physiological processes of different microorganisms; inhibition on the MAPK signaling pathway of pathogens will disturb infection progresses and facilitate the efficient control of crop disease.

Although many studies have examined MAPK signaling pathways in other fungi, functional analysis in C. higginsianum is still required to understand the intricate roles in the A. thaliana– C. higginsianum interaction. To date, compared with other fungi, the specific roles of MAPK for infection-related morphogenesis remain largely unknown in C. higginsianum. Because the MAPK pathway contributes to multiple physiological processes of fungal pathogens (Zhang et al., 2016), characterization of C. higginsianum MAPK will help illuminate the mechanism of the cruciferous crops—C. higginsianum interaction and facilitate the efficient control of anthracnose disease. For characterizing the MAPK involved in appressorium formation and pathogenicity ability of C. higginsianum, we here investigated the functions of ChMK1 (CH063\_08490), an ortholog of PMK1 involved in appressorium differentiation and plant infection in M. oryzae (Bruno et al., 2004). Besides attenuated growth rate, reduced conidiation and albinistic colony, our results firstly indicated that targeted disruption of Fus3/Kss1-related MAPK gene ChMK1 leads to hypersensitivity to cell wall inhibitors. Moreover, the ChMK1 gene disruption mutant also failed to form appressorium and lost its pathogenicity. Here, we have reported the roles of ChMK1 in cell wall integrity, appressorium differentiation, and pathogenicity in C. higginsianum, and demonstrated that the MAPK signaling pathways play essential roles in this fungus.

### MATERIALS AND METHODS

### Fungal Strains, Plants, and Culture Condition

The wild-type C. higginsianum IMI349061 (CH-1) was cultured on potato dextrose agar (PDA) at 25◦C and stored in PDA slants at 4◦C for further use. A. thaliana wild-type Col-0 plants were used in this study. Plants were grown in growth chambers at day and night temperatures of 20 ± 2 ◦C, with 12 h of light and 12 h of darkness. Escherichia coli strain DH5α was used to propagate all plasmids and Agrobacterium tumefaciems strain EHA105 was used for fungal transformation.

### Bioinformatics Data and Programs Used in this Study

The publicly available genomic sequence database of C. higginsianum (http://genome.jgi.doe.gov/Colhi1/Colhi1. home.html) was used to characterize the CH063\_08490 gene. NCBI (http://www.ncbi.nlm.nih.gov/) and UniProt (http://www. uniprot.org/blast/) were used for Blastp analysis. The Clustal X program was used for amino acid alignments. MEGA program was used to produce the phylogenetic tree with unrooted neighbor-joining method. The secondary structure prediction was completed with Jnetpred program.

### *ChMK1* Gene Replacement and Complementary

For characterizing the ChMK1 gene, a ChMK1 replacement vector, pChMK1-3300, and a complementation vector, pChMK1- Com, were constructed. The replacement vector was constructed using the homologous recombination strategy. A 0.9-kb fragment upstream of the ChMK1 ORF in the C. higginsianum genome was amplified with primers Rep-up F and Rep-up R (**Table 1**) and cloned into the Hind III and Sal I sites on pMD18-hph, and the resulting construct was named pMD2-hph. Then a 1.0-kb fragment downstream of ChMK1 ORF was amplified with primers Rep-down F and Rep-down R (**Table 1**) and cloned between the Xba I and Kpn I sites in pMD2-hph, and the resulting construct was the ChMK1 gene replacement vector, pChMK1-3300 (**Figure S1**), which had the selective marker hph gene flanked by the ChMK1 ORF flanking sequences (**Figure 2A**).

The complementation vector was constructed that ChMK1 cDNA was amplified by RT-PCR with primers Com F and Com R (**Table 1**) and cloned into the same sites of pCIT vector, which contained the constitutive PtrpC promoter and terminator. Finally, the cDNA of ChMK1 was digested with Apa I and cloned into pNeoP3300, resulting in ChMK1 complementation vector, pChMK1-Com.

Agrobacterium-mediated transformation was performed as previously described (Liu et al., 2013) with some modification that plasmid-containing A. tumefaciems cultures (0.6 OD unit at 600 nm) were mixed 1:1 with C. higginsianum conidial suspension (10<sup>6</sup> spores/ml) in induction broth supplemented with 400µM acetosyringone and cultured for 6 h at 25◦C 200 rpm, and then cultured on a cellophane membrane laid on co-induction medium supplemented with 400µM acetosyringone at 22◦C for 48 h. The cellophane membrane were removed to new plates and overlaid with 15 ml PDA containing 500µg/ml of cephalosporin and 50µg/ml of hygromycin. After incubation at 25◦C for 72 h, transformants were transferred to

#### TABLE 1 | Primers used for vector construction and PCR.

Primers used for vector construction and PCR


PDA plates containing 50µg/ml of hygromycin for a second round of selection. Transformants were confirmed primarily by RT-PCR with primers RT F and RT R (**Table 1**) and further confirmed by Southern blot. The C. higginsianum β-tubulin gene (CH063\_04743) (**Table 1**) was used to normalize the RNA sample for each RT-PCR.

### DNA Manipulation and Southern Blot Analysis

The genomic DNA of C. higginsianum wild-type strain and other derivative mutants was extracted according to the procedure (Sambrook and Russell, 2001).

Southern blot analysis was performed following the method described (Liu et al., 2013). For each sample, 15–20µg genomic DNA was digested with Hind III, size-fractionated through a 0.8% agarose gel and mounted on positively charged nylon membrane. The nylon membrane was then hybridized with a probe amplified by primers RT-F and RT-R (**Table 1**) and labeled with digoxigenin (DIG)-dUTP using the PCR DIG Probe Synthesis Kit (Roche, Mannheim, Germany) in accordance with the manufacturer's instructions.

### RNA Manipulation and qRT-PCR

Total RNA of fungal strains was isolated with TriZOL reagent (Invitrogen, Carlsbad, USA) according to manufacturer's instructions and stored at −80◦C for further study. The total RNA samples were treated with DNase I (RNase Free) (Takara, Dalian, China) at 37◦C for 0.5 h and used to generate the first strand cDNA with RevertAidTM First Strand cDNA Synthesis Kit (Fermentas, St. Leon-Rot, Germany) according to manufacturer's instructions. The first strand cDNA was stored at −20◦C for further study.

Gene expression was analyzed by qRT-PCR using a Bio-Rad CFX96 (Bio-Rad, USA) and SYBR Green PCR mix (Bio-Rad, USA), according to the manufacturer's instructions. The C. higginsianum β-tubulin gene (CH063\_04743) was used to normalize the RNA sample for each qRT-PCR. For each gene, qRT-PCR assays were repeated at least twice, with each repetition having three replicates.

Primer pairs for qRT-PCR detections were designed across or flanking an intron and listed in **Table 1**.

### Characterization of ChMK1 Deletion Transformant and Wild-Type Strain

The growth rates were assayed that ChMK1 deletion transformant and the wild-type strain were firstly cultivated on PDA for 7 days, then the mycelial agar discs were taken from the active colony edge and inoculated on the center of the PDA petri dish at 25◦C before hyphal growth was examined. The colony morphology and conidiation of these strains were examined after being grown on PDA plate for 15 days at 25◦C.

The virulence was assayed that C. higginsianum strains were cultured on PDA at 25◦C for 5 days, and the conidia were collected and washed with sterile distilled water twice. The conidial suspension (10<sup>6</sup> conidia/ml) were spotted with 6µl droplets on the leaf surface of 4-weeks-old Arabidopsis. Inoculated leaves were incubated in darkness at 25◦C, and the symptoms were observed at 6 days post-inoculation (dpi).

Conidial germination and appressorial formation were observed that 10µl conidial suspension (10<sup>6</sup> conidia/ml) was spotted on plastic coverslips and incubated at 25◦C for 24 h. Then, the infection-related morphogenesis was examined by Nikon Eclipse 80i microscope (Nikon, Tokyo, Japan), under bright-field model using 40× fold magnification.

### Cell Wall Sensitivity Assay

The sensitivity of wild-type strain and the transformants to cell wall inhibitors were performed that mycelial plugs (4 mm in diameter) were respectively inoculated onto PDA containing 0.01% SDS, 300µg/ml Calcofluor White (CFW), and 300µg/ml Congo Red (CR) for 7 days as described (Zeng et al., 2012; Luo et al., 2014), and with PDA media as controls. Cell wall sensitivity to the compounds mentioned above was assayed by measuring the colony diameters as previously described (Fujioka et al., 2007; Valiante et al., 2009; Carbo and Perez-Martin, 2010).

### Statistical Analysis

The data assays were analyzed with Origin 7.5 (OriginLab Corporation, Massachusetts, USA) using ANOVA (one-way, P ≤ 0.05). Results of all graphs represent the mean value ± SD. Asterisks in the graphs indicate statistical differences, P ≤ 0.05.

### RESULTS

### CH063\_08490 is Similar to MAPK Protein

The SMART MODE (http://smart.embl-heidelberg.de/smart/ change\_mode.pl) analysis result indicates that CH063\_08490 contains a Serine/Threonine protein kinases catalytic (S\_TKc) domain (**Figure 1A**), which plays a key role in catalysis of protein phosphorylation. The BLAST searches of CH063\_08490 for homologous sequences resulted in significant similarity with sequences from M. oryzae PMK1 (XP\_003712175.1, Evalue: 0.0, 98.9% identity), Cochliobolus heterostrophus CHK1 (AF178977.1, E-value: 0.0, 94% identity), Botrytis cinerea BMP1 (AF205375.1, E-value: 0.0, 94.6% identity), F. oxysporum FMK1 (AAG01162.1, E-value: 0.0, 98.3% identity) and Colletotrichum lagenarium CMK1 (AAD50496.1, E-value: 0.0, 99.4% identity) that match to the MAPK proteins. Sequence multiple alignment analysis of these homologs revealed significant conservation in length and amino acid composition (**Figure 1B**). Phylogenetic analysis indicated that CH063\_08490 is closely related to CMK1 (**Figure 1C**), a MAPK in C. lagenarium involved in conidiation, appressorium formation, and pathogenicity (Takano et al., 2000).

In summary, the protein coded by CH063\_08490 resembles MAPK proteins, thus we named this gene as ChMK1 derived from C. higginsianum MAPK.

### Gene Disruption and Complementation of *ChMK1*

For function study of ChMK1 gene in C. higginsianum, we generated ChMK1 deletion mutants. Deletion transformants were screened by growing on PDA containing hygromycin and further confirmed by Southern blot and RT-PCR (**Figures 2C,D**). TABLE 2 | Comparison of growth rate and conidial production among *ChMK1* deletion mutants, complementary strain and wild-type strain.


*<sup>a</sup>Growth rate was detected by measuring the colony diameter of cultures incubated at 25*◦*C for 7 days. <sup>b</sup>Conidia produced by 15-days-old cultures and counted with haematocytometer. Different letters indicated statistically significant differences (P* = *0.05). Means and standard errors were calculated from three replicates.*

Two deletion mutants 1ChMK1-1 and 1ChMK1-8, which had the marker inserted into regions other than the ChMK1 gene, were selected for further analysis in this study. Furthermore, for complementation of the ChMK1 deletion mutant, the complementary vector pChMK1-Com was transformed into deletion mutant 1ChMK1-1, and the complemented transformants ChMK1-Com, confirmed by RT-PCR was selected for further analysis (**Figure 2D**).

### *ChMK1* Deletion Transformant Shows Abnormal Phenotype

The infection assay of ChMK1 deletion mutant was performed as described above. Conidia suspension of wild-type strain, ChMK1 deletion strain 1ChMK1-1 and ChMK1 complementary strain ChMK1-Com were inoculated on Arabidopsis leaves. At 6 dpi, the wild-type and ChMK1-Com strains could cause dark necrotic lesions (**Figure 3A**), whereas the ChMK1 deletion transformant failed to cause any lesion on the leaves of Arabidopsis after inoculating 1ChMK1-1 mutants for 6 days (**Figure 3A**), which is similar with previous study that the MAPK mutant of C. lagenarium did not develop any lesions, even inoculated on wounded plants (Takano et al., 2000). These results indicate that ChMK1 is essential for the pathogenicity of C. higginsianum on Arabidopsis.

The ChMK1 affects the melanin formation and growth rate of C. higginsianum. Compared with wild-type strain and ChMK1 complementary strain, the ChMK1 deletion transformant showed obvious albino colony (**Figure 2B**), reduction of conidiation (**Table 2**), and reduced growth rate on PDA (**Table 2**). Appressorium formation of the ChMK1 deletion mutants on the hydrophobic surface (plastic coverslips) was investigated by microscopic observation. Conidia of the wildtype strain and ChMK1 complementary strain could formed melanized appressoria on the plastic coverslips after incubation for 24 h, while ChMK1 deletion mutant failed to differentiate into appressoria (**Figure 3B**). It concluded from these results that ChMK1 regulate appressorium formation, thus to affect the pathogenicity of C. higginsianum.

### ChMK1 is Significant for the Maintenance of Cell-Wall Integrity

Previous studies indicated that MAPK signal pathway contributed to cell wall integrity in multiple fungi (Jeon et al., 2008; Zeng et al., 2012; Ding et al., 2015), in our study,

### FIGURE 1 | Continued

FMK1: AAG01162.1 (*F. oxysporum*); CMK1: AAD50496.1 (*C. lagenarium*). The conserved kinase activation residues TEY are labeled with asterisks at the top of the alignment. The secondary structure prediction was completed with *Jnetpred* program. The green arrows indicate β-strands structure and the red sticks indicate α-helix structure. (C) Phylogenetic analysis of ChMK1 of *C. higginsianum* and its homologs from other fungi. The amino acid sequences were analyzed by MEGA version 4 with Unrooted Neighbor-joining algorithm. Bootstrap values were calculated from 1000 bootstrap replicates. Only bootstrap support values >50% are shown. The black star indicates ChMK1.

the sensitivity of wild-type strain, ChMK1 complementary strain and ChMK1 deletion transformant to three cell wall inhibitors was also tested. The results indicated that ChMK1 deletion transformant was more sensitive to SDS, Calcofluor White (CFW) and Congo Red (CR) than wild-type strain and the complementary transformant (**Figure 4A**). Data showed that the inhibition of the growth rate of ChMK1 deletion transformant were higher than that of wild-type strain and

ChMK1 complementary strain when cultured on PDA amended with those cell wall inhibitors respectively (**Figure 4B**).

### ChMK1 Regulates the Expression of Melanin Biosynthesis-Associated Genes

Previous studies revealed a crucial role of MAPK signal pathway in melanin biosynthesis in several other fungi (Takano et al., 2000; Yago et al., 2011; Zeng et al., 2012). In our study, the C. higginsianum ChMK1 deletion mutant was also unable to form melanized colony as wild-type, we speculated that the biosynthesis of dihydroxynaphthalene (DHN) melanin was interrupted. In order to verify this assumption, the expression level of three major melanin biosynthetic-associated genes, βketoacyl synthase (PKS1, CH063\_03518), trihydroxynaphthalene reductase (THR1, CH063\_06688), and scytalone dehydratase (SCD1, CH063\_08047), were compared between ChMK1 deletion mutant, complementary strain and wild-type strain using qRT-PCR. The results indicated that the expression level of these three genes in ChMK1 deletion mutant mycelia was significantly reduced compared with that of the complementary and wild-type strains (**Figure 5**). Similar results were also obtained in other studies that the expression level of melanin biosyntheticassociated genes were decreased in MAPK mutants (Takano et al., 2000; Wei et al., 2016).

### DISCUSSION

In this study, we characterized the Fus3/Kss1-related MAPK ChMK1 to assess the roles of MAPK pathways in the fungal pathogenesis of C. higginsianum. Gene deletion and complementary analysis of ChMK1 demonstrated that this gene is essential for appressorium formation, pathogenicity,

conidiation production, growth rate, maintenance of cell-wall integrity, and melanin formation in C. higginsianum.

In F. oxysporum, C. lagenarium, M. oryzae, and Pyrenophora teres, deletion mutants of Fus3/Kss1-related MAPK genes showed normal growth rates (Xu and Hamer, 1996; Takano et al., 2000; Di Pietro et al., 2001; Ruiz-Roldan et al., 2001; Luque et al., 2016), whereas the opposite was observed in B. cinerea and Ustilago maydis (Mayorga and Gold, 1999; Zheng et al., 2000). In our study, the growth rate of ChMK1 deletion mutants was significantly decreased compared to wild type strain (**Table 2**),

supporting the diverse functions of Fus3/Kss1-related MAPKs in different plant fungal pathogens. Furthermore, the Fus3/Kss1 related MAPK signal pathway from several plant pathogen fungi plays significant roles in conidiation, appressorium formation, and fungal pathogenicity (Mayorga and Gold, 1999; Ruiz-Roldan et al., 2001; Zhao et al., 2007; Miguel-Rojas and Hera, 2016; Zhang et al., 2016). In this study, the albinistic ChMK1 deletion mutant also showed a significant reduction in conidiation production (**Table 2**) and defects in appressorium formation, thus losing pathogenicity (**Figure 3**). These results indicate the conserved roles of Fus3/Kss1-related MAPKs in phytopathogenic fungi.

SLT2-type MAPKs, another type of MAPKs, are involved in cell wall integrity and stress tolerance in multiple fungi and Phytophthora species (Gao et al., 2015). In Coniothyrium minitans, CmSlt2 was involved in conidiation, cell wall integrity, and mycoparasitism, targeted disruption of CmSlt2 led to hypersensitivity to cell wall-degrading enzymes and cell wall inhibitors caffeine, CFW and CR (Zeng et al., 2012). In F. oxysporum, disruption of the MAPK FoSlt2 gene resulted in increased sensitivity to H2O<sup>2</sup> and cell wall inhibitors CR and CFW (Ding et al., 2015). In M. oryzae, MoMck1–MoMkk1–MoMps1 MAPK pathway contributed to regulate cell wall integrity (Yin et al., 2016). Moreover, silencing of PsMPK1 in P. sojae caused increased hypersensitivity to cell wall-degrading enzymes cellulase and lysing enzyme (Li et al., 2014). Similar results were obtained in A. alternata, since the AaSLT2 mutants also displayed hypersensitivity to cell wall-degrading enzymes, CFW and CR (Yago et al., 2011). Nevertheless, until recently, there has been little experimental evidence for the contribution of Fus3/Kss1-related MAPKs to cell wall integrity in fungi. In this study, our results firstly demonstrated that the Fus3/Kss1-related MAPK ChMK1 from C. higginsianum also plays a significant role in cell wall integrity, and deletion of ChMK1 cause increased hypersensitivity to cell wall inhibitors (**Figure 4**).

Furthermore, the deletion mutants of ChMK1 showed obviously albino colony (**Figure 2B**). Since DHN melanin is required for melanization, which is essential for appressorial functions and virulence in Colletotrichum species (Rasmussen and Hanau, 1989; Lin et al., 2012), we speculated that the biosynthesis of DHN melanin was interrupted in the deletion mutant of ChMK1. The qRT-PCR results in this study demonstrated that the expression levels of PKS1, THR1, and SCD1, the major DHN melanin biosynthetic-associated genes (Takano et al., 1997), decreased significantly in ChMK1 deletion mutant when compared to wild-type strain (**Figure 5**), which was similar with other studies that the expression level of melanin biosynthetic-associated genes decreased in MAPK mutants (Takano et al., 2000; Wei et al., 2016), indicating that ChMK1 is involved in the biosynthesis of DHN melanin by regulating the expression of DHN melanin biosynthetic-associated genes.

Most fungal pathogens contain three MAPKs that are orthologs of the S. cerevisiae Fus3/Kss1, Slt2, and Hog1 MAPK. Recent study demonstrated that Fus3/Kss1, Slt2, and Hog1 MAPKs have distinct and complementary roles, and the positive and negative crosstalk between three MAPK pathways regulates stress adaptation, development and virulence in F. oxysporum (Luque et al., 2016). In M. oryzae, it was reported that the feedback between the cAMP and MAPK signaling pathways regulate appressorium morphogenesis and plant infection (Zhou et al., 2012). In C. lagenarium, Cmk1 MAPK cooperated with cAMP-PKA signaling pathway to regulate germination, appressorium formation and infectious growth, and mutation of these genes led to similar defects in germination, appressorium formation, and infectious growth (Takano et al., 2000; Yamauchi et al., 2004). In B. cinerea, Nox-, calcium-, and MAPKsignaling cascades incorporated with RasGAP scaffold protein BcIqg1 to regulate several developmental processes and virulence (Marschall and Tudzynski, 2016). Similar results were also obtained in mycoparasite C. minitans that MAPK cascade and Nox complex signal pathway are cross-linked and essential for pathogenicity, melanin synthesis and conidiation (Wei et al., 2016). In our study, deletion of ChMK1 resulted in decreased growth rate, defects in virulence, and appressorial formation, hypersensitivity to cell wall inhibitors, reduced conidiation, and albinistic colony. These phenotype of ChMK1 mutant are similar with that of the mutants of other signaling pathway in other fungi (Zhou et al., 2012; Luque et al., 2016; Marschall and Tudzynski, 2016; Wei et al., 2016). Thus, it was speculated that C. higginsianum MAPK cascade also cross-linked with other signal pathways possibly to regulate multiple physiological processes in C. higginsianum.

Briefly, we have analyzed the functions of ChMK1 in appressorium formation, pathogenicity, growth rate, conidiation and melanin production in C. higginsianum. Moreover, we also firstly reported that the Fus3/Kss1-related MAPK is involved in the maintenance of cell-wall integrity. The results described above will enhance our understanding of the mechanism underlying A. thaliana–C. higginsianum interaction and will facilitate the efficient control of cruciferous crops anthracnose disease.

### CONCLUSION

The Fus3/Kss1-related MAPK ChMK1 was experimentally confirmed to be essential to appressorium formation, pathogenicity, growth rate, conidiation production, and

### REFERENCES


melanin formation in C. higginsianum. Furthermore, our results also firstly showed the involvement of Fus3/Kss1-related MAPK in cell wall integrity, indicating that ChMK1 plays diverse and essential roles in this fungus.

### AUTHOR CONTRIBUTIONS

Conceived and designed the experiments: WW, WZ. Performed the experiments: WW, WZ. Analyzed the experiment data: WW, YX, WZ, NW, GY, FP. Contributed reagents/materials/analysis tools: WW, WZ. Wrote the paper: WW, WZ. All authors have read and approve the final manuscript.

### ACKNOWLEDGMENTS

The research was financially supported by the National Natural Science Foundation of China (31501587), Scientific Research Foundation Granted From Jianghan University, and Scientific Research Foundation Granted From Wuhan Polytechnic University. We thank reviewers for their kind suggestions.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01212

Figure S1 | The vector map of pChMK1-3300.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Wei, Xiong, Zhu, Wang, Yang and Peng. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Enhancement of Thiamine Biosynthesis in Oil Palm Seedlings by Colonization of Endophytic Fungus Hendersonia toruloidea

Amirah N. Kamarudin1,2, Kok S. Lai<sup>3</sup> , Dhilia U. Lamasudin<sup>3</sup> , Abu S. Idris<sup>2</sup> and Zetty N. Balia Yusof<sup>1</sup> \*

<sup>1</sup> Department of Biochemistry, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia, Serdang, Malaysia, <sup>2</sup> Ganoderma and Diseases Research Group, Biology Division, Malaysian Palm Oil Board, Kajang, Malaysia, <sup>3</sup> Department of Cell and Molecular Biology, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia, Serdang, Malaysia

Thiamine, or vitamin B1 plays an indispensable role as a cofactor in crucial metabolic reactions including glycolysis, pentose phosphate pathway and the tricarboxylic acid cycle in all living organisms. Thiamine has been shown to play a role in plant adaptation toward biotic and abiotic stresses. The modulation of thiamine biosynthetic genes in oil palm seedlings was evaluated in response to root colonization by endophytic Hendersonia toruloidea. Seven-month-old oil palm seedlings were inoculated with H. toruloidea and microscopic analyses were performed to visualize the localization of endophytic H. toruloidea in oil palm roots. Transmission electron microscopy confirmed that H. toruloidea colonized cortical cells. The expression of thiamine biosynthetic genes and accumulation of total thiamine in oil palm seedlings were also evaluated. Quantitative real-time PCR was performed to measure transcript abundances of four key thiamine biosynthesis genes (THI4, THIC, TH1, and TPK) on days 1, 7, 15, and 30 in response to H. toruloidea colonization. The results showed an increase of up to 12-fold in the expression of all gene transcripts on day 1 post-inoculation. On days 7, 15, and 30 post-inoculation, the relative expression levels of these genes were shown to be downregulated. Thiamine accumulation was observed on day 7 post-colonization and subsequently decreased until day 30. This work provides the first evidence for the enhancement of thiamine biosynthesis by endophytic colonization in oil palm seedlings.

Keywords: oil palm, endophytic fungi, thiamine biosynthesis, gene expression, endophytic colonization

## INTRODUCTION

Thiamine, also known as vitamin B1, is required for key metabolic processes in cellular organisms. The active form, thiamine pyrophosphate (TPP), is a cofactor in important metabolic reactions, notably glycolysis, tricarboxylic acid cycle, pentose phosphate pathway, and synthesis of branched amino acids (Goyer, 2010). The thiamine biosynthesis pathway in plants is similar to that in bacteria and yeast (Begley et al., 1999; Li et al., 2010). As shown in **Figure 1**, this pathway consists of two separate branches: the thiazole branch and pyrimidine branch. The pyrimidine moiety of thiamine, hydroxymethylpyrimidine phosphate (HMP), is produced from the precursor 5-aminoimidazole ribonucleotide (AIR) by the enzyme HMP synthase, encoded by the THIC gene.

#### Edited by:

Gero Benckiser, Justus Liebig University Giessen, Germany

#### Reviewed by:

Ömür Baysal, Mugla University, Turkey ˘ Michael Wink, Universität Heidelberg, Germany

> \*Correspondence: Zetty N. Balia Yusof zettynorhana@upm.edu.my

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 04 June 2017 Accepted: 03 October 2017 Published: 17 October 2017

#### Citation:

Kamarudin AN, Lai KS, Lamasudin DU, Idris AS and Balia Yusof ZN (2017) Enhancement of Thiamine Biosynthesis in Oil Palm Seedlings by Colonization of Endophytic Fungus Hendersonia toruloidea. Front. Plant Sci. 8:1799. doi: 10.3389/fpls.2017.01799

The thiazole moiety arises from NAD, glycine, and an S-donor, which form hydroxyethylthiazole phosphate; this is synthesized by the enzyme hydroxyethylphosphate synthase (THI4). The thiazole and pyrimidine moieties are joined together by the bifunctional enzyme (TH1) to form thiamine monophosphate (TMP) (Guan et al., 2014). TMP is dephosphorylated by a phosphatase known as thiamine monophosphate phosphatase (TH2) to form thiamine (Mimura et al., 2016). All of these steps occurs in the chloroplast (Pourcel et al., 2013). The last step is the phosphorylation of thiamine to its active form, TPP by the enzyme thiamine pyrophosphokinase (TPK), which occurs in the cytosol.

Thiamine metabolism is involved in adaptation to biotic and abiotic stresses in plants and microorganisms (Rapala-Kozik et al., 2012; Sylvander et al., 2013). For example, thiamine treatment enhances the resistance of soybean to charcoal rot disease (Monaim, 2011), of rice to sheath blight disease (Bahuguna et al., 2012), and of grapevine to Plasmapara viticola (Boubakri et al., 2012). The mechanism of disease suppression through application of thiamine is explained by the activation of a plethora of host defense responses. In Arabidopsis thaliana, thiamine treatment activates pathogenesisrelated protein (PR-1) and phenylalanine lyase (PAL). In addition, thiamine treatment in grapevine reduces downy mildew development in a dose-dependent manner by inducing hydrogen peroxide generation, callose disposition, and host resistance (HR) cell death. Similarly, thiamine treatment successfully controls charcoal rot disease in soybean plants by inducing defense-related enzymes including peroxidase (PO), polyphenol oxidase (PPO), PAL, and pathogenesis related (PR) chitinase (Monaim, 2011). Therefore, thiamine is thought to be involved in priming, which is important in mitigating crop-damaging diseases and stresses.

Oil palm is the most profitable oil-bearing crop in the world, yielding about 4–6 tons of oil per hectare (Murphy, 2014). However, the productivity of the oil palm is threatened by basal stem rot disease caused by Ganoderma boninense, which results in major economic losses and yield gaps (Barcelos et al., 2015). Therefore, studies on the proper management of the disease have been increasing (Hushiarian et al., 2013).

Endophytes, defined as fungi that are present in most plant tissues without causing any visible symptoms, have been utilized as biological control agents in preventive measures against the disease (Wilson, 1995). Several endophytic species have been thus utilized, namely Actinomycetes, Pseudomonas, Trichoderma, and Hendersonia (Sapak et al., 2008; Idris et al., 2010; Sundram et al., 2011). It was reported that the application of endophytes results in growth-promoting effects independent of the suppression of G. boninense. The endophytic fungus Hendersonia toruloidea, originally isolated from oil palm trunk and root tissues, has been used as a biocontrol agent. In vitro and nursery trial studies of endophytic application of H. toruloidea suppre**s**sed infection of the pathogenic fungus G. boninense (Idris et al., 2010).

In this study, we examined the responses of oil palm seedlings to colonization by H. toruloidea, specifically in terms of the expression of thiamine biosynthetic genes. The morphology and the colonization pattern of H. toruloidea were visualized with transmission electron microscopy (TEM). Furthermore, expression of thiamine biosynthetic genes, as well as accumulation of total thiamine and its intermediates, were compared.

## MATERIALS AND METHODS

### Fungal Strains, Growth Conditions, and Granular Bioformulation Preparation

A strain of the fungal endophyte H. toruloidea was previously isolated from healthy oil palms in disease-affected areas of MPOB Teluk Intan, Perak, Malaysia (Idris et al., 2010). Pure axenic cultures were sub-cultured on potato dextrose agar and incubated at 28◦C for 10 days. Conidial spores were scraped and poured into potato dextrose broth containing 9% jaggery. Fungal cultures were incubated at 28◦C for 4 days in a shaking incubator at 150 rpm. Fungal cultures were encapsulated in an alginate formulation containing kaolin, empty fruit bunch, and pectin (Idris et al., 2010).

### Plant Experimental Conditions

Seven-month-old seedlings of Dura × Pisifera variety oil palms were grown under nursery conditions at MPOB Nursery, Section 15, Bandar Baru Bangi, Selangor, Malaysia. Oil palms were inoculated with H. toruloidea by applying 50 g of the bioformulation (10<sup>7</sup> CFU/g) and drenching with tap water. Control plants were not treated. Spear leaves and roots were sampled (two replicates per treatment in three independent experiments) on days 0, 1, 7, 15, and 30 post-treatment, immediately frozen in liquid nitrogen, and stored at −80◦C until further analysis.

### TEM Analysis

Oil palm root sections were cut into 1-mm<sup>3</sup> slices. Root sections were put into separate vials and fixed in 4% glutaraldehyde for 2 days. Next, the root sections were washed with 0.1 M sodium cacodylate buffer three times for 30 min each. They were then post-fixed in 1% osmium tetroxide for 2 h at 4◦C before being washed again with 0.1 M sodium cacodylate buffer three times for 30 min each. A dehydration series of acetone (35, 50, 75, 95%) was used for 45 min each. The final dehydration with 100% acetone was performed three times for 1 h each. Each specimen was then embedded into a beam capsule, which was filled with resin mixture. This was polymerized in an oven at 60◦C for 48 h. Thick sectioning was performed by cutting the polymerized specimen into 1-µm thick sections using an ultramicrotome. The thick sections were stained with toluidine blue and dried on a hot plate. The stain was washed under running tap water. The area of interest was examined under a light microscope. For ultrathin sectioning, ultrathin sections were cut and selected for silver staining. The sections were selected with a grid and dried using filter paper. For the staining procedure, the sections were stained with uranyl acetate for 15 min and washed with double-distilled water. Next, they were stained with lead stain for 10 min and washed with double-distilled water. Lastly, the sections were viewed using TEM (Technai G2 Transmission Electron Microscope).

### RNA Isolation and Quantitative Real-time PCR Analysis

Total RNA was isolated from oil palm spear leaves using the CTAB method with modifications (Zeng and Yang, 2002). Genomic DNA was removed using DNase I (Promega, Madison, WI, United States) according to the manufacturer's instructions. Purified RNA samples (1 µg) were reverse-transcribed using GoScript cDNA synthesis kit (Promega). Specific primers (Supplementary Table S1) for quantitative real-time PCR were designed by Primer Premier 6.0 (Primer Biosoft, Palo Alto, CA, United States). Each 10-µl PCR reaction mixture contained 4.0 µl cDNA template, 5 µl 2× SYBR SensiFast Hi-Rox (Bioline, Taunton, MA, United States), and 0.4 µl 10 mM forward and reverse primers for each gene. Quantitative PCR was performed using a Bio-Rad CFX Connect 96 (Hercules, CA, United States). The cycling conditions were as follows: 2 min at 95◦C, followed by 40 cycles of 10 s at 95◦C and 30 s at 60◦C. Three biological replicates with three technical replicates each were assayed for each sample. Transcript levels of each gene were normalized to the reference genes tubulin and glyceraldehyde-3-phosphate using the method by Vandesompele et al. (2002). The 2−11CT method was used to analyze the relative changes in expression of THI4, THIC, TPK, and TH1 (Livak and Schmittgen, 2001).

### Analysis of Thiamine and Its Ester Phosphates by High Performance Liquid Chromatography (HPLC)

For HPLC, 5 g of leaf samples were ground in liquid nitrogen; 20 ml of 0.1 N hydrochloric acid (HCl) was then added and incubated at 37◦C for 16 h. Samples were then centrifuged at 4000 × g for 10 min. Samples were filtered using Whatman filter paper. Next, 2.5 ml of 5% trichloroacetic acid was added to the filtrate. The derivatization step was performed by adding 2.5 ml of freshly prepared 1% potassium ferricyanide in 15% NaOH, 250 µl

from the copyright holder.]

of phosphoric acid, and 750 µl of 0.1 N HCl. Samples were filtered using a 0.22-µm syringe filter before being added to amber vials. After sample derivatization into thiochrome and its esters, samples were analyzed by HPLC with fluorescence detection (Agilent 1290 Infinity UPLC, Palo Alto, CA, United States). The column used was a Kinetex 5 µm C18 100 Å, LC column, 100 mm × 4.6 mm (Phenomenex, Torrance, CA, United States). A gradient elution was used, where solvent A contained 10 mM sodium phosphate buffer, pH 7 and solvent B consisted of 100% methanol.

### RESULTS

### Colonization of Oil Palm Seedlings by Endophytic H. toruloidea

The colonization and localization of H. toruloidea inside the root was visualized with TEM. **Figure 2** shows a transverse section of oil palm root on days 1 and 30 post-colonization. H. toruloidea was found to reside primarily in the cells of the cortical tissues.

### The Effect of H. toruloidea Colonization on the Expression of Thiamine Biosynthetic Genes in Oil Palm

The expression of thiamine biosynthetic genes in oil palm seedlings during colonization by endophytic H. toruloidea was analyzed using quantitative real-time PCR at various time points: 1, 7, 15, and 30 days post-treatment (**Figure 3**). One day after the application of H. toruloidea, oil palm seedlings showed 12.9-, 3.65-, 1.65-, and 3.05-fold increased levels of expression of THI4, TPK, THIC, and TH1, respectively. Meanwhile on day 7, THI4, TPK, THIC, and TH1 were downregulated to levels that represented 2.85-, 0.92-, 0.44-, and 1.07-fold changes, respectively, when compared to levels in control seedlings. After 14 days, expression of THI4, THIC, TPK, and TH1 had continued to decrease to levels that were 0.23-, 0.21-, 0.73-, and 0.21-fold, respectively, those in control seedlings. Finally, at 30 days postcolonization, levels of expression of THI4, THIC, TPK, and TH1 were even lower, at 0.19-, 0.34-, 0.27-, and 0.57-fold, respectively, those in control seedlings.

### The Effect of H. toruloidea Colonization on Total Thiamine Accumulation in Oil Palm Seedlings

Well-separated peaks of thiamine (TF), thiamine monophosphate (TMP) and thiamine pyrophosphate (TPP) were detected at retention time (Rt) of 7.22, 3.35, and 3.04 min respectively (**Figure 4A**). **Figure 4B** showed that TMP could not be detected in oil palm leaves.

**Figure 5** summarizes the changes in the contents of total thiamine and its ester phosphates at each time point. The total content of thiamine is the sum of TPP and free thiamine (TF). TPP was present at higher levels than TF. There was no significant increase in total thiamine 1 day post-colonization. Interestingly, the total thiamine content in oil palm leaves was significantly enhanced by twofold on day 7 post-colonization. At subsequent time points (days 15 and 30), the total thiamine content had returned to control levels.

### DISCUSSION

This study was based on the hypothesis that thiamine biosynthesis would be upregulated upon endophytic colonization. Successful colonization by H. toruloidea was observed by TEM analysis, which revealed that the fungus resides inside cortical cells. Overall, THI4, TPK, THIC, and TH1 were upregulated 24 h after inoculation. A previous study had shown that thiamine and its intermediates are involved in systemic acquired resistance in various plant species, acting as signaling molecules (Rapala-Kozik et al., 2008; Tunc-Ozdemir et al., 2009). The observation that thiamine biosynthesis is upregulated as a result of colonization by an endophytic fungus is relatively novel. Because of the relatively long duration of the observation, the

downregulation of thiamine biosynthesis 7 days after inoculation can be seen as an adaptation to the establishment of H. toruloidea in oil palm seedlings.

We observed that the expression level of THI4, which is the first enzyme in the thiamine biosynthesis pathway, was remarkably high 1 day after endophytic colonization by H. toruloidea, with an increase of about 12-fold compared to levels of THIC, TPK, and TH1. It is implied that the upregulation of thiamine biosynthetic genes, specifically the increase in THI4 expression, was caused by the increased demand of TPP, which acts as a coenzyme in numerous physiological processes, including glycolysis, the pentose phosphate pathway, the synthesis of nucleic acids, and the synthesis of NADPH. The increased requirement for TPP-dependent enzymes was probably due to perturbations in metabolism as a result of stress and adaptive responses to H. toruloidea colonization in the oil palm seedlings (Rapala-Kozik et al., 2008). Aside from its role as a cofactor, the thiazole synthase family is also involved in other non-cofactor cellular functions. In the yeast Saccharomyces cerevisiae, THI4 mutants showed higher susceptibility to DNA damage (Machado et al., 1997). Interestingly, it was found that THI4, which codes for thiazole synthase in the fungal pathogen Verticillium dahlia, is required for stress tolerance against UV damage and vascular disease induction in tomato (Hoppenau et al., 2014). Thus, prominent increase in the expression of THI4 in response to H. toruloidea colonization can be explained by its dual role in thiamine biosynthesis and the stress response.

The application of endophytes is an excellent strategy for activating systemic acquired resistance in plants. Numerous studies have highlighted the potential role of fungal endophytes in plant protection through the upregulation of secondary metabolites, such as phytohormones, phenolic compounds, and defense enzymes (Gao et al., 2010; Kumara et al., 2014). Endophyte-mediated systemic acquired resistance could serve as a disease control strategy in the development of plants with enhanced disease resistance. The relationship between thiamine and systemic acquired resistance was established by Ahn et al. (2007), who showed that thiamine treatment elicited transient expression of PR genes and hypersensitive responses in rice, Arabidopsis, and cucumber. Similarly, thiamine treatment in rice induces priming, which results in higher hydrogen peroxide content, total phenolic accumulation, and phenylalanine lyase activity (Bahuguna et al., 2012). Moreover, when the pathogenic fungus Sclerotinia is inoculated on Arabidopsis, the plant experiences an upregulation in thiamine biosynthesis correlated with increased accumulation of thiamine, TMP, and TPP (Zhou et al., 2013). This implies that there is an increase in the de novo synthesis of endogenous thiamine upon Sclerotinia infection. Thiamine-treated plants exhibit resistance to the pathogenic fungus by producing reactive oxidative species (ROS) to promote defense signaling. Therefore, in relation to our study, it can be safely assumed that changes in thiamine biosynthesis, specifically the upregulation of THI4, are due to early adaptive defense responses in oil palm seedlings upon active colonization by H. toruloidea.

Ideally, it is minimally required that both THIC and THI4, the upstream enzymes for each of the two branches of thiamine biosynthesis (pyrimidine and thiazole), are upregulated in order to increase the total thiamine pool (Pourcel et al., 2013; Dong et al., 2016). We found that THI4 was significantly upregulated to 12.9-fold after 24 h of endophytic colonization. However, the magnitude of the change in expression of THIC upon

5 g of fresh oil palm leaf was used. Data are means ± SD of three replicates. Asterisks denote significant differences between control and inoculated seedlings (p < 0.05).

colonization with the endophytic fungus was not great, at about 1.65-fold. A small increase in the expression of THIC compared with that of THI4 was also demonstrated in oil palm seedlings infected with G. boninense (Balia Yusof et al., 2015). This can be explained by the fact that THIC is a highly complex energy-expensive enzyme and exhibits a bottleneck in overexpression. The synthesis of THIC requires a large energy investment, as structural studies have shown that THIC is made of an iron-sulfur cluster, S-adenosyl methionine, and ferredoxin– thioredoxin redox system (Colinas and Fitzpatrick, 2015). THIC may also be involved in other functions, such as oxygenic photosynthesis and circadian regulation (Colinas and Fitzpatrick, 2015). Apart from these, THIC, which is responsible for the synthesis of HMP-pyrophosphate (HMP-PP), is required for the synthesis of TMP by condensation of HMP-PP and HET-P. The imbalance in the magnitudes of the expression changes of THIC and THI4 could therefore be explained by the involvement of other sources of HMP-PP. In the yeast (S. cerevisiae) thiamine biosynthesis pathway, HMP-PP is obtained from the pyridoxal 5-phosphate (PLP) biosynthesis pathway (Li et al., 2010). This implies that THIC may not be significantly upregulated if the HMP-PP pool is already sufficient.

It is interesting to note that 7 and 15 days post-inoculation, the expression of thiamine biosynthetic genes was lower in inoculated seedlings than in control seedlings. This suggests a tight regulatory process, whereby the genes are switched off when the thiamine pool becomes sufficient. Moreover, it indicates that

the endophytic fungus H. toruloidea also synthesizes thiamine and therefore the thiamine biosynthetic machinery in the oil palm is repressed. Although plants generally synthesize thiamine, it may be more advantageous for them to obtain it from various external sources through biotic interactions (Helliwell et al., 2014). This was further supported by a study by Paerl et al. (2015), in which it was reported that during the co-culturing of the auxotrophic picoeukaryotic algae Ostreococcus lucimarinus and the bacterium Pseudoalteromonas sp., the algae was able to salvage thiamine from the bacterium.

In the present study, further evaluation of the effects of H. toruloidea colonization was performed through the quantification of total thiamine and its ester phosphates in oil palm leaves. Endophytes are known to be beneficial to the host through the production of secondary metabolites that can improve plant fitness. In this study, we assessed how colonization by the endophytic H. toruloidea affects total thiamine accumulation in oil palm. The significant twofold increase in the total thiamine content on day 7 post-colonization may lead to enhancement of the plant's metabolic fitness through the activation of TPP-dependent enzymes. However, we observed that the total thiamine content was restored to control levels after 14 days, which may signal adaptive processes. Constitutive thiamine accumulation is suggested to be detrimental to the plant itself. This is because increased activity of TPP-dependent enzymes will result in the over-influx of carbohydrate oxidation through the tricarboxylic acid cycle and pentose phosphate pathway. Previous observations have shown that elevated thiamine accumulation in Oryza sativa overexpressing NB-LRR genes, which function as intracellular immune receptors, resulted in growth retardation and chlorosis (Wang et al., 2016). Therefore, thiamine accumulation is believed to be beneficial but only to some extent. A certain physiological level of thiamine must be maintained for optimal growth and function in plants, and this has not yet been understood.

### REFERENCES


### CONCLUSION

The major findings presented here demonstrate that successful colonization of oil palm seedlings by H. toruloidea results in the upregulation of thiamine biosynthetic genes and increased accumulation of total thiamine. Subsequent attenuation of thiamine biosynthesis signals adaptation, which may be important in maintaining optimal growth and function in plants. Further molecular, biochemical, and physiological studies are needed to understand the role and function of thiamine in the oil palm stress response.

### AUTHOR CONTRIBUTIONS

AK performed the experiments, analyzed the data, and wrote the manuscript. ZBY designed the research, ZBY, AI, KL, and DL supervised the work and revised the manuscript.

### FUNDING

AK is grateful to the Malaysian Palm Oil Board Graduate Student Assistanceship Scheme (MPOB-GSAS) for providing the scholarship and Science Fund (5450799) from the Ministry of Science, Technology and Innovation Malaysia (MOSTI) for providing fund for the project.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2017.01799/ full#supplementary-material



salt and osmotic stress conditions is mediated by abscisic acid at the early stages of this stress response. BMC Plant Biol. 12:2. doi: 10.1186/1471-2229-12-2


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Kamarudin, Lai, Lamasudin, Idris and Balia Yusof. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# The Interaction between Arbuscular Mycorrhizal Fungi and Endophytic Bacteria Enhances Plant Growth of *Acacia gerrardii* under Salt Stress

Abeer Hashem1, 2, Elsayed F. Abd\_Allah<sup>3</sup> \*, Abdulaziz A. Alqarawi <sup>3</sup> , Asma A. Al-Huqail <sup>1</sup> , Stephan Wirth<sup>4</sup> and Dilfuza Egamberdieva<sup>4</sup>

 Department of Botany and Microbiology, Faculty of Science, King Saud University, Riyadh, Saudi Arabia, <sup>2</sup> Department of Mycology and Plant Disease Survey, Agriculture Research Center, Plant Pathology Research Institute, Giza, Egypt, Department of Plant Production, Faculty of Food and Agricultural Sciences, King Saud University, Riyadh, Saudi Arabia, Leibniz Centre for Agricultural Landscape Research, Institute of Landscape Biogeochemistry, Müncheberg, Germany

Microbes living symbiotically in plant tissues mutually cooperate with each other by

#### *Edited by:*

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### *Reviewed by:*

Biswapriya Biswavas Misra, University of Florida, USA Joseph Davis Bagyaraj, Indian National Science Academy, India

> *\*Correspondence:* Elsayed F. Abd\_Allah eabdallah@ksu.edu.sa

#### *Specialty section:*

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

*Received:* 17 February 2016 *Accepted:* 29 June 2016 *Published:* 19 July 2016

#### *Citation:*

Hashem A, Abd\_Allah EF, Alqarawi AA, Al-Huqail AA, Wirth S and Egamberdieva D (2016) The Interaction between Arbuscular Mycorrhizal Fungi and Endophytic Bacteria Enhances Plant Growth of Acacia gerrardii under Salt Stress. Front. Microbiol. 7:1089. doi: 10.3389/fmicb.2016.01089 providing nutrients for proliferation of the partner organism and have a beneficial effect on plant growth. However, few studies thus far have examined the interactive effect of endophytic bacteria and arbuscular mycorrhizal fungi (AMF) in hostile conditions and their potential to improve plant stress tolerance. In this study, we investigated how the synergistic interactions of endophytic bacteria and AMF affect plant growth, nodulation, nutrient acquisition and stress tolerance of Acacia gerrardii under salt stress. Plant growth varied between the treatments with both single inoculants and was higher in plants inoculated with the endophytic B. subtilis strain than with AMF. Co-inoculated A. gerrardii had a significantly greater shoot and root dry weight, nodule number, and leghemoglobin content than those inoculated with AMF or B. subtilis alone under salt stress. The endophytic B. subtilis could alleviate the adverse effect of salt on AMF colonization. The differences in nitrate and nitrite reductase and nitrogenase activities between uninoculated plants and those inoculated with AMF and B. subtilis together under stress were significant. Both inoculation treatments, either B. subtilis alone or combined with AMF, enhanced the N, P, K, Mg, and Ca contents and phosphatase activities in salt-stressed A. gerrardii tissues and reduced Na and Cl concentration, thereby protecting salt-stressed plants from ionic and osmotic stress-induced changes. In conclusion, our results indicate that endophytic bacteria and AMF contribute to a tripartite mutualistic symbiosis in A. gerrardii and are coordinately involved in the plant adaptation to salt stress tolerance.

Keywords: AMF, endophyte, *Acacia gerrardii*, salinity, nutrition

### INTRODUCTION

Salinity is a devastating environmental stress factor that severely affects plant growth and development (Barnawal et al., 2014). At the global level, particularly in arid and semiarid regions, salinity is considered an important constraint, and approximately 7% of global land has a high salt concentration, making this area unavailable for agriculture (Sheng et al., 2008; Ruiz-Lozano et al., 2012). Salinity reduces plant growth through osmotic as well as ionic constraints of major physiological and biochemical processes (Ahmad, 2010; Porcel et al., 2012; Abd\_Allah et al., 2015a). This may in turn alter the availability of nutrients for plant growth and affect the association with microbes living within the plant vicinity. Plants are colonized by microbes, including endophytes, nitrogen-fixing bacteria and mycorrhizal fungi, which closely cooperate with each other and can mediate important physiological processes, especially nutrient acquisition and plant tolerance to abiotic stresses (Egamberdieva, 2012; Egamberdieva et al., 2013, 2015; Berg et al., 2013; Ahanger et al., 2014a; Abd\_Allah et al., 2015a). Arbuscular mycorrhiza fungi (AMF) form beneficial symbiotic associations with most plants and play a vital role in plant growth under various conditions by modifying the root system and enhancing mobilization and the uptake of several essential elements. They have also been reported to stimulate plant stress tolerance by enhancing enzymatic as well as nonenzymatic antioxidant defense systems (Wu et al., 2014; Ahmad et al., 2015), lipid peroxidation (Abd\_Allah et al., 2015c), and phytohormone synthesis (Navarro et al., 2013). Endophytic bacteria colonize the internal tissues of their host plants and can promote growth, stress tolerance, and nutrient uptake and protect plants from soil-borne pathogens (Malfanova et al., 2011; Sessitsch et al., 2012). Microbes living symbiotically in plant tissues, such as mycorrhizal fungi, and nitrogen-fixing bacteria also mutually cooperate with each other by synthesizing biologically active compounds and providing nutrients for the survival and proliferation of their partner organism (Marschner et al., 2001; Egamberdieva et al., 2013). These synergies among endophytes are known to have beneficial effects on plants by improving the availability of nutrients to plants and inducing plant defense against various stresses, including drought and salinity.

The symbiotic relationship between legumes and nitrogenfixing rhizobia is susceptible to abiotic factors, such as nutrient deficiency, salinity, drought, acidity, and soil temperature, which induce failure of the infection and nodulation processes (Slattery et al., 2001; Bouhmouch et al., 2005; Egamberdieva et al., 2013). An improvement in legume-rhizobia symbiotic performance by AMF has been reported for faba bean (Vicia faba) (Yinsuo et al., 2004), lucerne (Medicago sativa) (Ardakani et al., 2009), lentil (Lens culinaris) (Xavier and Germida, 2002), and common bean (Phaseolus vulgaris) (Tajini et al., 2012). The positive effects on plant growth and stimulation of stress tolerance by synergistic interactions of root-colonizing, plant growth-promoting bacteria (PGPR) and AMF under hostile environments have been extensively reviewed by Nadeem et al. (2014). These microbes are believed to act as essential bio-ameliorators of stress by regulating the nutritional and hormonal balance (Abd\_Allah et al., 2015a,b; Egamberdieva et al., 2015, 2016) and inducing systemic tolerance to stress (Ruiz-Lozano et al., 2012).

Despite these beneficial associations of microbes, studies examining the interactions of endophytic bacteria and AMF in hostile environmental conditions are limited, especially where competition for nutrient and niches in the rhizosphere is high. This knowledge is important for our understanding of the relationship between AMF and endophytic bacteria and their potential effect on plant stress tolerance and for the development of crop management practices under hostile environmental conditions.

Acacia gerrardii Benth. (Talh tree) is a small leguminous shrub that is resistant to drought and salinity, forms nodules and can improve the fertility of salt-affected arid soils. This tree is widely used in Saudi Arabia for fuel, forage, medicine, food production and also for agroforestry. In this study, we hypothesize that the improved salt tolerance and growth of A. gerrardii are mediated by (i) the effect of endophytic bacteria on mycorrhizal development and colonization in the roots of A. gerrardii and (ii) nodulation, nutrient acquisition by a synergistic interaction between endophytic bacteria and AMF.

### MATERIALS AND METHODS

### Plants and Microorganisms Isolation of Endophytic Bacteria

Roots of Talh trees were collected from the top 20 cm of soil in a natural meadow at Khuraim in Riyadh, Saudi Arabia. The samples were wrapped in a plastic bag and brought to the laboratory, where they were incubated at 4◦C until further processing. One gram of roots was surface-sterilized by immersion in 70% ethanol, followed by 5% sodium hypochlorite for 5 min, and then rinsed in sterile distilled water four to six times to eliminate the chlorine. The sterilized roots were macerated, and the extracts were placed in a tube containing 9 ml of sterile phosphate-buffered saline and then serially diluted. A 100 µl aliquot from the appropriate dilutions was plated on tryptic soy agar (TSA, Difco Laboratories, Detroit, USA) supplemented with 4% NaCl. The plates were incubated at 28◦C for 3 days, and all colonies that displayed differentiable colony morphologies were selected from the plates and were re-streaked to purify the strains. To select strains with increased stress tolerance, the purified isolates were cultured in TSA medium supplemented with 3, 4, or 5% (w/v) NaCl.

### Identification of the Selected Strain

A highly salt-tolerant strain, which grew well in TSA medium containing 4% NaCl, was identified. DNA was isolated by a modified version of the Töpper et al. (2010) protocol. The filters were re-suspended in 250 µl of lysozyme solution (1 mg ml−<sup>1</sup> TE buffer, pH 7.4). Then, 250 µl of preheated (55◦C) lysis buffer (20 µg proteinase K ml−<sup>1</sup> 0.5% SDS) was added to the solution. After incubation for 30 min at 55◦C, 80 µl of 5 M NaCl and 100 µl of preheated (55◦C) CTAB [10% (w/v) hexadecyltrimethylammonium bromide in 0.7% NaCl] were added. The solution was incubated for a further 10 min at 65◦C followed by the addition of 500 µl chloroform: isoamylalcohol (24:1, v/v). The solution was centrifuged (16,000 × g, 5 min) to separate the DNA from the remaining cell debris. Then, the top phase was transferred to a fresh tube, and the DNA was precipitated with isopropanol and later resuspended in TE buffer (pH 7.4). Next, 16S rDNA was amplified by polymerase chain reaction (PCR) using universal forward 16SF (5′ -GAGTTTGATCCTGGCTCAG-3′ ) and reverse 16SR (5′ -GAAAGGAGGTGATCCAGCC-3′ ) primers (Mohanta et al., 2015). The PCR reactions were 25 µl and contained 5 µl of

5 × buffer (TaKaRa Bio Inc.), 0.5 µl of dNTP mixture (10 mM of each dNTP, TaKaRa Bio Inc.), 1 µl of 2% BSA (Promega), 0.5 µl of forward primer (10 µM), 0.5 µl of reverse primer (10 µM), 0.125 µl of One Taq DNA Polymerase (New England Biolabs), 15.375 MQ and 2 µl of template DNA. The PCR program (Bio-Rad DNA Engine) started with an initial denaturation step for 30 s at 94◦C followed by 30 cycles of 15 s at 94◦C, 30 s at 55◦C and 1.5 min at 68◦C. Before cooling to 4◦C, an extension period of 20 min at 68◦C was incorporated into the program. The PCR products were verified by gel electrophoresis on a 1.5% agarose gel stained with TAE (Tris-acetate-EDTA). Denaturing gradient gel electrophoresis (DGGE) was performed using the DCode system (Bio-Rad). Equal amounts of PCR products (6 µl) were loaded onto 8% acrylamide gels with a denaturing gradient of 30–55% [where 100% denaturing is defined as 7 M urea and 40% (v/v) formamide (Muyzer et al., 1995)] for optimal separation of the PCR products. DGGE gels were run for 19 h at 60 V and at 60◦C in 0.59 TAE buffer and stained for 30 min with SYBR Gold (Invitrogen) diluted 10,000-fold in 19 × TAE buffer. Gels were visualized and digitized using the Fujifilm Imaging System. The PCR product was purified, and nucleotide sequences were determined using automatic LI-COR DNA Sequencer 4000 L (Lincoln, USA). The sequences were identified using the basic local alignment search tool (BLAST) and comparisons with the GenBank nucleotide data bank from the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/).

### Plant Growth-Promoting Traits

The cellulose-degrading ability of bacterial isolates was analyzed by streaking inocula on cellulose Congo Red agar media, as described by Gupta et al. (2012). Zones of clearance around and beneath the colony were detected, indicating enzymatic degradation of cellulose.

The production of indole 3-acetic acid (IAA) was determined as described by Bano and Musarrat (2003). Briefly, bacterial strains were grown in TSB medium. After 3 days, 1 ml of each culture was pelleted by centrifugation, and the supernatant was discarded. Cell pellets were washed with 1 ml of PBS and re-suspended in PBS. One milliliter of cell suspension (corresponding to a cell density of 10<sup>7</sup> cells/ml) was added to 10 ml of TSB supplemented with tryptophan (100 µg/ml). After 3 days of cultivation, 2 ml aliquots of bacterial cultures were centrifuged at 13,000 × g for 10 min. One milliliter of supernatant was transferred to a fresh tube to which 100 µg/ml of 10 mM orthophosphoric acid and 2 ml of reagent (1 ml of 0.5 M FeCl<sup>3</sup> in 50 ml of 35% HClO4) were added. After 25 min, the absorbance of the sample was measured at 530 nm. The IAA concentration in cultures was calculated using a calibration curve of pure IAA as the standard.

The phosphate-solubilizing activity of the bacterial strains was determined on Pikovskaya agar (Pikovskaya, 1948) containing precipitated tricalcium phosphate. The bacterial culture grown in TSA medium for 2 days was streaked on the surface of Pikovskaya agar plates and incubated for 3 days. The presence of a clear zone around bacterial colonies was considered to be an indicator of positive P solubilization.

### Arbuscular Mycorrhizal Fungi

AMF were isolated from the soil surrounding the roots of A. gerrardii. AMF were extracted by wet sieving, decanting and sucrose density gradient centrifugation as described by Daniels and Skipper (1982) and modified by Utobo et al. (2011). Briefly, 100 g of soil was placed in a 10 L bucket, and 5 L of tap water was added to the soil and mixed well to produce a soil-water suspension. The suspension was left for 5 min to allow insoluble and heavy particles to settle, and the suspension was sequentially sieved through ASTM-500, ASTM-250, ASTM-150, and ASTM-50 sieves to extract the spores using a wet sieving and decanting method (Gerdemann and Nicolson, 1963). The sieved residues were filtered through Whatman filter paper No. 1. After water filtration, the filter paper was examined under a stereo-binocular microscope at 25 × magnification. Morphologically similar spores were selected for identification. AMF species were identified based on the description of subcellular structures (spore color, shape, surface ornamentation, spore contents, and wall structures) of asexual spores provided by the International Culture Collection of Vesicular and Arbuscular Mycorrhizal Fungi (INVAM 2012)<sup>1</sup> and other descriptive protocols (Bethlenfalvay and Yoder, 1981; Schüßler and Walker, 2010; Redecker et al., 2013).

### Propagation of AMF in Trap Cultures

The trap culture protocol described by Stutz and Morton (1996) was used in the current study to propagate the most efficient mycorrhizal isolates. Sterilized sand (121◦C for 3days) was inoculated with single spores from each mycorrhizal isolate. Surface-sterilized seeds [0.5% (v/v) NaOCl used] of Sorghum sudanense were sown (20 seeds/pot) at a depth of 2 cm in each pot (5 kg capacity) of trap cultures. The pots were incubated in a plant growth chamber at 25 ± 2 ◦C with 18 h photoperiod, 750 µmol m−<sup>2</sup> s <sup>−</sup><sup>1</sup> photosynthetic photon flux density, and 70–75% relative humidity for 3 months. Half-strength Hoagland's solution was used to irrigate the pots. The trap culture was used as the mycorrhizal inoculum and was added to the experimental soil as 25 g of trap soil culture (approx. 100 spores/g trap soil)/pot. Soil not inoculated with mycorrhiza served as the control.

### Determination of Arbuscular Mycorrhizal Colonization

At the end of the pot experiment (12 weeks), fine roots were collected from the lateral root system and fixed in formalin/acetic acid/alcohol (v/v/v) (FAA) solution until further processing. Roots were stained with trypan blue in lactophenol (Phillips and Hayman, 1970) and assessed for mycorrhizal infection. Roots that were pigmented after clearing were bleached in alkaline hydrogen peroxide (0.5% NH4OH and 0.5% H2O<sup>2</sup> v/v in water) to remove any phenolic compounds (Kormanik and McGraw, 1982) before acidification (0.05 M HCl). To assess mycorrhizal colonization, stained root segments (one cm in length) were mounted on glass slides with lactophenol and were observed under a digital computerized microscope (model DP-72, Olympus) at 20 ×

<sup>1</sup> INVAM - International Culture Collection of Arbuscular & Vesicular-Arbuscular Mycorrhizal Fungi. Available online at: www.invam.caf.wvu.edu/index.html (Accessed February 6, 2012).

magnification. A minimum of 50 segments for each replicate sample were observed to assess structural colonization of AMF associated with roots. Twenty or more segments were mounted on each slide and examined under the microscope. The presence of mycelia, vesicles and arbuscules was recorded and analyzed to assess structural colonization.

### Germination of Seeds

Acacia gerrardii seeds were surface-sterilized by immersion for 5 min in concentrated sulfuric acid followed by 3 min in 70% ethanol and were rinsed five times with sterile, distilled water.

Germination tests were carried out in Petri dishes (Ø 85 × 15 mm) containing 1% water agar. Salinity conditions were established by adding 50, 100, 150, 200, 250, and 300 mM NaCl. Twenty healthy and surface-sterilized seeds were placed on each Petri dish and were arranged in a randomized complete block design with three replications. Eventually, the Petri dishes were covered with a polyethylene sheet to avoid the loss of the moisture through evaporation and kept in the plant growth chamber at 28◦C. The seeds were observed daily, and the percent germination was recorded after 10 days of incubation. Seeds were considered to have germinated when the emerging radicles were greater than 0.5 cm long.

### Plant Growth Condition

This experiment was carried out in the growth chamber of the Plant Production Department, Faculty of Food & Agricultural Sciences, King Saud University, Riyadh, Saudi Arabia. Talh tree seeds were provided by a personal nursery in Alghat, Riyadh that produces tree seedlings. The soil used was loamy sand soil with the following properties (%): sand (87.6), clay (7.2), silt (5.2), organic carbon (0.12), total nitrogen (0.005), pH 7.5. The sand was washed with 1.0 N H2SO<sup>4</sup> for 1 h, followed by 1.0 N Ca carbonate, and was then washed using distilled water. The surface-sterilized seeds were sown in acid-washed sterile sand and kept in a plant growth chamber under the same conditions described for trap cultures for 1 month.

The bacterial strain was grown in TSB medium for 2 days, and 1 ml of culture suspension was pelleted by centrifugation. The supernatant was discarded, and the cell pellets were washed with 1 ml phosphate-buffered saline and diluted to a cell density of 10<sup>8</sup> cells/ml. Roots of 1-month-old seedlings were immersed in the bacterial suspension for 30 min and sown in pots (1 seedling per pot) filled with 2 kg of sandy loam soil mixed with mycorrhizal inoculum [i.e., 25 g of trap soil culture or approx. 100 spores/g trap soil (M = 80%)/pot]. The experiment was a completely randomized design with five replicates for each treatment: (i) Control without microbes, (ii) Bacillus subtilis (BS), (iii) AMF, and (iv) B. subtilis combined with AMF (BS+AMF). Plants were grown in a greenhouse for 12 weeks with average day/night temperatures of 25◦C/18◦C and were supplemented with tap water as required. Salinity was established by adding NaCl to the irrigation solution to obtain a constant concentration of 250 mM. At harvest, plants were harvested carefully, and shoots were separated from roots and dried to constant weight at 100◦C. Fresh nodules were used to estimate leghemoglobin (LB) levels as well as the activities of nitrite reductase, nitrate reductase, and nitrogenase. Fresh leaf samples were used to assess the content of photosynthetic pigments.

### Nodulation, Leghemoglobin, and Crude Protein Contents

The nodule fresh weight (FW) and the number of nodules per plant root were determined. The leghemoglobin concentration of root nodules was estimated using the method of Keilin and Wang (1945). Fresh nodules (2.0 g) were ground to a fine powder in liquid N<sup>2</sup> and transferred to 50 mM KPO<sup>4</sup> (pH 7.4) buffer containing 1 mM EDTA. The mixture was stirred until it thawed into a homogenate at a final temperature of 2◦C, transferred to centrifuge tubes and centrifuged at 4◦C and 10,000 × g for 10 min. The leghemoglobin-containing supernatant was collected and maintained at a known volume with 50 mM KPO<sup>4</sup> (pH 7.4) buffer as described above. The color intensity that developed was recorded spectrophotometrically at 710 nm against a blank, which contained 50 mM KPO<sup>4</sup> (pH 7.4) buffer with 1 mM EDTA. Leghemoglobin concentration was expressed as mg/g nodule (FW). The micro-Kjeldahl method described by Allen (1953) was used to estimate the total nitrogen content of oven-dried (110◦C for two successive weights) nodules. Crude protein concentration (%, w/w) was determined mathematically by multiplying total nitrogen content by the general factor for cereal protein determination, 6.25, as described in the AOAC Official Method 992.23 (AOAC Association of Official Analytical Chemists, 1995).

### Nitrate and Nitrite Reductase and Nitrogenase Activity

Nitrate and nitrite reductase activities in the nodules were assayed using the methods of Hageman and Hucklesby (1971) and Finka et al. (1977), respectively. Phosphate buffer (pH 7.5) containing 0.1 M potassium nitrate and 5% n-propanol was used to extract the reductases. The optical density of the samples was measured spectrophotometrically at 550 nm. A standard curve of potassium nitrite was used as a reference. The activity of nitrate and nitrite reductases was expressed as µmNO<sup>2</sup> released/h/g FW and µmNO<sup>2</sup> disappeared/h/g fresh wt, respectively. Nitrogenase (EC 1.7.99.2) activity (ARA) was determined by acetylene reduction with the known weight of the nodules collected from all nodulated root portions of the plants, following the method of Herdina and Silsbury (1990). Gas samples were analyzed for ethylene produced in the reaction using a Shimadzu GC-14B gas chromatograph equipped with a Porapak R column (Ligero et al., 1986).

### Determination of Photosynthetic Pigments

Photosynthetic pigments were extracted from leaf samples in 80% acetone as described by Arnon (1949). Fresh leaf samples (0.5 g) were extracted in 80% acetone (v/v). The extracted material was centrifuged at 10, 000 × g for 10 min. The optical density of the supernatants was recorded at 480, 645, and 663 nm using a UV-visible spectrophotometer (T80 UV/VIS spectrometer, PG Instruments Ltd., USA). A blank with 80% acetone served as the control.

### Estimation of Acid and Alkaline Phosphatase Activity

Fresh root samples were extracted with 0.1 M acetate buffer (pH 5.0) and 0.1 M Tris HCI buffer (pH 8.2) for acid phosphatase (AP) (EC 3.1.3.2) and alkaline phosphatase (ALP) (EC 3.1.3.1), respectively, as described by Gianinazzi-Pearson and Gianinazzi (1976). AP and ALP were assayed following the methods of Ikawa et al. (1964) and Torriani (1967), respectively. In these methods, the reaction mixture of AP contained 0.2 ml of enzyme extract and 1.0 ml of 5.5 mM p-nitrophenol phosphate in 5.5 mM citrate buffer (pH 4.8). For the ALP assays, 0.05 M Tris-citrate (pH 8.5) was used instead of citrate buffer (pH 4.8). The reaction mixtures were incubated at 37◦C for 30 min, and the reactions were stopped by adding 10 ml of 200 mM NaOH. Absorbance was recorded at 410 nm, and activity was expressed as µmol p-nitrophenol released min/mg protein.

### Determination of Mineral Contents

To determine the mineral contents, oven-dried leaves were powdered, and the powder was digested with 98% H2SO<sup>4</sup> and 30% H2O2. The total nitrogen (N) content in leaf tissues was determined following the semi-micro Kjedahl procedure using a nitrogen analyzer (Kjedahl 2300; FOSS, Hoganas, Sweden). The phosphorus (P) content was estimated using the vanadomolybdophosphoric colorimetric method (Jackson, 1962). A standard curve (10–100 µg/ml) of potassium dihydrogen phosphate (KH2PO4) was used as the reference. The contents of Na+, K+, Mg2+, and Ca2<sup>+</sup> in plant leaves were estimated as described by Wolf (1982) using a flame photometer (Jenway Flame Photometer, Bibby Scientific Ltd., Stone-Staffs, UK). The chloride (Cl−) concentration was directly estimated in the digested extracts using a chloride analyzer (Model 926, Sherwood Scientific Ltd., Cambridge, UK), as described by Abd\_Allah et al. (2015b).

### Statistical Analyses

Duncan's multiple range test was performed using one-way analysis of variance (ANOVA) for a completely randomized design by SPSS-21 software, and significant differences in means were determined by the least significant differences (LSD) (p = 0.05) test. In addition, the correlation coefficients among the parameters studied were calculated and are presented in **Tables 3, 5**, Table S3.

## RESULTS

## Identification of Microorganisms

Using nucleotide homology and phylogenetic analysis of the 16S rRNA gene sequences, we grouped the microbe that showed the highest salt tolerance in a cluster containing B. subtilis (GenBank Accession Number: JX188065.1) with 99% sequence similarity. The genome sequence of B. subtilis BERA71 was deposited in GenBank with the accession number KX090253. The strain produced IAA in nutrient broth containing 1.5% (260 mM) NaCl and cellulose and also solubilized mineral phosphorus from tricalcium phosphate used in solid medium.

Photomicrographs in (**Figures 1A–F**) showing spore morphology of AMF (Claroideoglomus etunicatum, Rhizophagus intraradices, Funneliformis mosseae) used in the current study. **Figures 1A,B** Crushed spore of C. etunicatum showing depressions on the surface and two wall layers: L1, an outer permanent rigid layer with some plasticity and an uneven outer surface. L2, A layer consisting of laminae that increase in thickness (∼8.0 µm thick) is rigid, exhibiting some swelling and spreading when broken. R. intraradices: **Figures 1C,D** Crushed yellow-brown spore, globose or subglobose in shape, consists of three layers. L1, Outer layer has a sloughing spore wall (SSW), hyaline, mucilaginous spores that stain pale pink. L2: A rigid hyaline layer attached firmly to the underlying laminae [L3] as a sub-laminae layer of the spore wall (SLW). L3: Inner layer as laminae of the spore wall (LSW), which is continuous with the innermost layer of the subtending hypha. The subtending hypha (SH) is cylindrical to slightly flared with three layers that are continuous with the three layers of the spore wall. The plug partitions the spore from the hyphal contents (septum, S). A septum (S) occludes the hyphal attachment of a thin-walled spore of the pale morph close to the spore base. F. mosseae: **Figures 1E,F** Crushed spore showing the wall structure as three layers. L1, Outer wall, hyaline, mucilaginous, forming a sloughing granular layer (SSW). L2, Hyaline, generally rigid, consisting of a thin adherent sublayer attached firmly to the underlying laminae (L3). L3, Inner layer consisting of laminae with minute depressions covering the surface and separating L1 and L2. The sporocarp (SC) is subglobose, light brown, and surrounded by a peridium (P). Developed intact spores (IS) were observed. The subtending hypha is funnel-shaped with a width ranging between 16 and 32 µm (**Figure 1F**). As shown in **Figure 1E**, the subtending hypha consists of two layers (L1, L3), with the sum of thickness ranging between 2.4 and 4.8 µm. A germ tube emerges from the lumen (funnel-shaped) of the subtending hypha, originating from the occlusion-recovered septum. Sporocarps dramatically produce numerous infective hyphae (**Figure 1E**). The spores in the current isolates from soil adhering to the roots of A. gerrardii were identified as F. mosseae (syn. Glomus mosseae), R. intraradices (syn. Glomus intraradices), and C. etunicatum (syn. Glomus etunicatum).

## Plant Growth Parameters

The germination of A. gerrardii seeds was tested at NaCl concentrations of 50–300 mM. The results showed that increases in salt concentration decreased the germination rate of the seeds compared to that of the control seeds (water only) (from 87 to 3%). Because germination of the seeds was strongly impaired by 300 mM NaCl, we used 250 mM NaCl in all further salt stress experiments.

The growth of A. gerrardii was strongly impaired by salinity when plants were not inoculated with AMF or endophytic bacteria. Salinity reduced shoot height and shoot dry weight by 61 and 62%, respectively, and root length and dry weight were reduced by 35 and 38%, respectively (**Table 1**), compared with non-stressed plants. Salt-stressed shoots were 35% shorter and root depth 61% less in plants treated with 250 mM NaCl than non-stressed plant shoots and roots.

FIGURE 1 | (A–F) (40X). AMF spores were isolated from soil samples of A. gerrardii. (A,B) The microscopic investigation indicated that the crushed spores of C. etunicatum were globose, sub-globose in shape, orange to red-brown in color and consisted of two layers (L1 and L2),. (C,D) Describes the spore morphology of R. intraradices as globose to subglobose in shape and whose crushed spores consisted of three layers (L1, L2, and L3) and prominent subtending hyphae. (E,F) The microscopic investigation revealed that the spores of F. mosseae are clustered together within a compact peridium. The shape of the spores was globose to subglobose, and the spore wall consisted of three layers (L1, L2, and L3).



SH, shoot height; SDW, shoot dry weight; RD, root depth; RDW, root dry weight; ±, standard deviation.

Plant growth depended strongly on the presence or absence of both AMF and the endophytic bacterium, BS. The inoculation of plants with B. subtilis alone or with AMF enhanced the root and shoot growth of non-stressed A. gerrardii, and the difference between the uninoculated and co-inoculated plants was significant (**Table 1**). Roots inoculated with the endophytic BS alone were significantly longer (22%) than uninoculated roots and were longer (14%) than roots inoculated with AMF alone (**Table 1**). In general, plant growth responded positively to the BS inoculation compared to AMF alone.

Salt-stressed A. gerrardii inoculated with endophytic B. subtilis alone grew better than salt-treated uninoculated plants. Root dry weight increased by 40% and shoot dry weight by 118% in the presence of 250 mM NaCl (**Table 1**). Shoot height and root length also responded positively to a single inoculation of BS. Root and shoot growth varied between treatments with the single inoculants and were higher in plants inoculated with the endophytic BS strain than with AMF. However, the interaction of AMF and rhizobia affected plant productivity positively compared to a single inoculation. Co-inoculated A. gerrardii had significantly higher shoot and root weight than plants inoculated with AMF or BS alone under NaCl stress (**Table 1**, Figure S1).

### Colonization of AMF in Plant Roots

The colonization of fine A. gerrardii roots by AMF is shown in **Figures 2A–F**. The roots were colonized with different AMF morphological structures, such as vesicles (**Figure 2A**), spores (**Figure 2C**), mycelium hyphae, intraradical hyphae (**Figure 2B**), subtending hyphae (**Figure 2C**), coiled hyphae (**Figure 2E**), and arbuscules (**Figure 2F**).

The percentage of 67.6, 55.8, and 13.8 of AMF that colonized the roots of A. gerrardii were in the form of mycelia, vesicles and arbuscules, respectively, with a total spore density of 707.8 spores/g of experimental soil (Table S1). NaCl stress decreased spore density, the presence of mycelia, vesicles and arbuscules by 24.8, 63.6, 20.7, and 60.4%, respectively, compared with the control treatment. Endophytic BS alleviated the adverse impacts of salt on spore density and mycorrhizal fungal colonization, and total spore count, mycelium, vesicles and arbuscules were increased by 27, 96, 14, and 23%, respectively, compared with those in the salt stress treatment group. In the absence of salt stress, the endophytic BS significantly increased both spore intensity and mycelium by 78.6 and 29.5%, respectively. However, both vesicles and arbuscules decreased by 48.38 and 44.92%, respectively, compared to those of the control treatment.

### Number of Nodules, Nodule Fresh Weight, and Leghemoglobin Content

The number of nodules, nodule FW and leghemoglobin content were reduced by 80.8, 80.04, and 80.6%, respectively, in saltstressed plants relative to control, unstressed plants (**Table 2**). A. gerrardii grown in AMF-infected soil showed a higher number of nodules, nodule FW and leghemoglobin content (25.9, 51.8, and 18.02%, respectively) than plants grown in control soil in both non-saline and saline conditions. The root nodulation of A. gerrardii depended strongly on the presence or absence of AMF and/or the endophytic bacterium, B. subtilis (**Table 2**). BS and AMF alone improved the symbiotic performance of saltstressed A. gerrardii. However, plants co-inoculated with AMF and B. subtilis produced three times more nodules, nodule FW, and leghemoglobin content in 250 mM NaCl than those in uninoculated plants.

### Nitrate Reductase, Nitrite Reductase, Nitrogenase, and Crude Protein

Both AMF and endophytic BS applied alone to non-stressed and stressed plants increased the nitrate (NR) and nitrite reductase (NIR) as well as nitrogenase activity. Inoculation of AMF increased nitrate reductase, nitrite reductase and nitrogenase activity by 15.4, 10.1, and 20.7%, respectively, while BS enhanced the activity of these enzymes by 31.5, 32.2, and 40.06%, respectively (**Table 2**).

The combined inoculation of AMF and BS increased nitrate reductase, nitrite reductase, and nitrogenase activity in the nodules of A. gerrardii by 42.5, 43.7, and 48.2%, respectively (**Table 2**), while salt-stressed plants showed a 38.5, 46.4, and 72.8% decline in nitrate reductase, nitrite reductase, and nitrogenase activity, respectively. The differences in enzyme activities between uninoculated plants and those co-inoculated with AMF and BS under salt stress were significant, and nitrate reductase activity increased by 56%, nitrite reductase activity by 53% and nitrogenase activity by 189% (**Table 2**). The crude protein content increased in response to all microbial inoculation treatments regardless of whether salt was present. The combination of BS and AMF produced even better results because co-inoculated salt-stressed plant nodules contained significantly more protein than uninoculated salt-stressed ones (**Table 2**).

**Table 3** show the correlations between salt, mycorrhiza, and bacteria with number of nodules, nodule FW, leghemoglobin nodule FW, nitrate reductase, nitrite reductase, nitrogenase, total nitrogen content, and crude protein. The results indicated a slight correlation between mycorrhiza and nitrite reductase (0.185). However, there was a strong correlation between the number of nodules and nitrate reductase (0.996).

### Photosynthetic Pigments

The photosynthetic pigments [chlorophyll a (ChlA), chlorophyll b (ChlB)], carotenoids and total chlorophyll (TChl) content in A. gerrardii were lower in plants without microbial inoculants or with only one inoculant. When plants were grown in the presence of AMF or BS, photosynthetic pigments increased, ChlA by 19.4 or 30.3%, ChlB by 12.6 or 28.3%, carotenoids by 41.8 or 62.5% and total chlorophyll content by 15.4 or 32.5%, respectively, compared to uninoculated plants (Table S2). Salinity reduced the content of ChlA, ChlB, carotenoids and TChl in A. gerrardii by 32.9, 40.5, 93.01, and 41.1%, respectively. Plants inoculated with endophytic BS and grown in soil infested with AMF showed an increase in the content of photosynthetic pigments. The content of ChlA, ChlB, carotenoids and TChl varied between the single inoculant treatments and was higher in plants inoculated with the endophytic BS strain than in those

FIGURE 2 | (A–F) Photomicrographs of structural colonization of AMF in the roots of A. gerrardii. (A) Vesicles (V); (B) intraradical hypha (IH), and vesicles (V). (C) crushed spore (CS) and subtending hypha (SH). (D) Intact spore (IS). (E) Coiled hyphae (CH). (F) Arbuscule (AR), trunk (T) and Intraradical hypha (IH).

TABLE 2 | Nodule number, fresh weight, leghemoglobin content, nitrate and nitrite reductase, and nitrogenase activity of *Acacia gerrardii* nodules under salt stress after inoculation with *B. subtilis* and AMF alone and in combination.


NN, number of nodules; LG, leghemoglobin (mg g−<sup>1</sup> nodule FW); NR, nitrate reductase (µm NO<sup>2</sup> released/h/g FW), NiR, nitrite reductase (µm NO<sup>2</sup> disappeared/h/g FW); NG, nitrogenase (µmole ethylene/mg nodule FW/h); CP, crude protein (total nitrogen content as mg/g dry weight × 6.25); ±, standard deviation.

inoculated with AMF (Table S2). A weak correlation was found between mycorrhiza and carotenoids (0.203), while the strongest correlation (0.998) was between chlorophyll a + chlorophyll b and total photosynthetic pigments followed by the correlation between chlorophyll a and total photosynthetic pigments (0.992; Table S3).


TABLE 3 | Correlations (r) between salt, mycorrhiza, and bacteria with number of nodules, nodule fresh weight, leghemoglobin nodule fresh weight, nitrate reductase, nitrite reductase, nitrogenase, total nitrogen content, and crude protein.

Sal, salt; M, mycorrhiza; B, Bacillus subtilis; NN, number of nodules; NFW, nodule fresh weight; LG, leghemoglobin (mg g−<sup>1</sup> nodule FW); NR, nitrate reductase (µm NO<sup>2</sup> released/h/g FW); NiR, nitrite reductase (µm NO<sup>2</sup> disappeared/h/g FW); NG, nitrogenase (µmole ethylene/mg nodule FW/h); TN, total nitrogen content (mg/ g dry weight); CP, crude protein (total nitrogen content as mg/g dry weight × 6.25).

### Nutrient Contents

The nutrient contents decreased following treatment with 250 mM NaCl, N content by 72% and P by 59%, K by 53%, Mg by 65%, and Ca by 60% (**Table 4**). NaCl stress increased Na<sup>+</sup> and Cl<sup>−</sup> content compared to plants under non-stressed conditions. Generally, both inoculation treatments, either BS alone or combined with AMF, enhanced Na+, P, K+, Mg2+, and Ca2<sup>+</sup> contents in non-stressed and salt-stressed A. gerrardii tissues (**Table 4**). However, the highest concentrations were detected in co-inoculated plant tissues grown in the presence of 250 mM NaCl. The correlations between salt, mycorrhiza, and bacteria with sodium, potassium, magnesium, calcium and chloride contents are shown in **Table 5**. The strongest correlation was between calcium and potassium (0.935), while the weakest correlation was between mycorrhiza and potassium (0.211).

### Acid and Alkaline Phosphatases

Acid phosphatase (AP) activity significantly increased as a consequence of salinity in uninoculated plant tissues (**Figure 3**). The acid and ALP increased by 78.3 and 116.9%, respectively, in response to 250 mM NaCl compared to non-stressed control plants. Inoculation with AMF had a stimulatory effect on the activity of both phosphatases in non-stressed and stressed plants (**Figures 3A,B**). Following combined inoculation of AMF and BS, AP activity increased by 142.7 and 147.3% in 250 mM NaCl and in unstressed control plants, respectively.

## DISCUSSION

The results from our experiments showed that a tripartite interaction of AMF, endophytic BS and the host plant A. gerrardii may have the potential to remediate degraded sites and can compensate for abiotic stresses due to climate change. Although, several reports have shown a positive effect of dual inoculation with AMF and PGPR on plant growth and stress tolerance, such as AMF with Enterobacter radicincitans on faba bean (Vicia faba) (Almethyeb et al., 2013), AMF with Pseudomonas mendocina on lettuce (Lactuca sativa L.) (Kohler et al., 2009), AMF with Pseudomonas fluorescens on common bean (Phaseolis vulgaris L.) (Neeraj and Singh, 2011), and AMF with Azospirillum on rice (Oryza sativa. L) (Ruíz-Sánchez et al., 2011), little is known about the interactions with endophytes as well as the underlying mechanisms. In this study, we observed a significant growth benefit of the synergistic association of A. gerrardii with AMF and endophytic B. subtilis under salt stress.

We found that salt stress reduced AMF colonization in A. gerrardii, consistent with observations made by Alqarawi et al. (2014a) in Ephedra aphylla and Hashem et al. (2015) in Vigna unguiculata. Our results showed that combined inoculation of plants with AMF and endophytic B. subtilis resulted in increased AMF colonization, which is an important indicator of plant nutrition. The synergistic interaction of B. subtilis and AMF altered A. gerrardii plant fitness under salt stress, significantly increasing plant biomass, nodulation, leghemoglobin, and crude protein content compared with untreated plants. An increase in plant growth and amelioration of salt stress by AMF was reported by Abd\_Allah et al. (2015a) for sunflower (Helianthus annuus L.), by Aroca et al. (2013) for lettuce, and by Gomez-Bellot et al. (2015) for laurustinus plants (Viburnum tinus L.). Bacterial endophytes have also been shown to increase plant growth and tolerance to abiotic stresses, e.g., P. fluorescens (Ali et al., 2014), Paenibacillus yonginensis (Sukweenadhi et al., 2015), and Bacillus sp. (Andreolli et al., 2016). When colonizing plant tissues, microbes contribute multiple benefits, such as improved nutrient acquisition, tolerance to biotic and abiotic stresses, and modulation of plant defenses (Bordiec et al., 2011). In addition, the increase in nodulation may be due to a synergistic effect of the two types of microbes, namely symbiotic and endophytic, including naturally occurring rhizobia. Huang et al. (2011) reported a mutualistic symbiotic relationship between B. subtilis and a leguminous plant, Robinia pseudoacacia L. In this study, B. subtilis colonized plant roots in a manner



Na, sodium; K, potassium; Mg, magnesium; Ca, calcium; N, nitrogen; P, phosphorous; ±, standard deviation.

TABLE 5 | Correlations (r) between salt, mycorrhiza, and bacteria with sodium, potassium, magnesium, calcium, and chloride.


Sal, salt; M, mycorrhiza; B, Bacillus subtilis; Na, sodium; K, potassium; Mg, magnesium; Ca, calcium; Cl, chloride.

similar to the infection of root hairs by rhizobia and formed bacteroids inside plant cortical cells. Inoculation of plants with cellulose-producing B. subtilis resulted in more nodules and higher nitrogenase activity than the uninoculated control and AMF-inoculated plants. Rhizobial symbionts penetrate deeper plant tissues by producing cellulase, which can completely erode the root-hair wall at the site of infection (Sindhu and Dadarwal, 2001). The synthesis of cell wall-degrading enzymes by B. subtilis could help explain the mechanism underlying rhizobial entry into target root hair cells to form nodules.

Increased nitrogenase activity following treatments with B. subtilis or B. subtilis combined with AMF resulted from their positive impact on the activity of enzymes such as nitrate reductase and nitrite reductase. Nitrate and nitrite reductase control the conversion of nitrate into ammonia and result in the formation of amino acids (Iqbal et al., 2015). In Trifolium alexandrinum L. and Trifolium resupinatum L. (Zarea et al., 2011), as well as in Vicia faba (Hashem et al., 2014), inoculation of AMF enhanced plant growth by improving nitrogenase activity and nodule formation.

High salt concentrations induce alterations in the synthesis of chlorophyll-related proteins and components of the oxygenevolving complex, resulting in reduced photosynthetic efficiency (Alqarawi et al., 2014a,b). Altered de novo synthesis of proteins and the associated pigment-related components due to salinity has negative effects on the synthesis of photoassimilates and hence reduces the growth rate of plants. The combined inoculation of AMF and B. subtilis increase photosynthetic pigments, which may be a collective result of many positive changes induced by AMF and PGPR. The enhanced chlorophyll content due to AMF inoculation under normal as well as NaClstressed conditions corroborates the reports of Aroca et al. (2013) in lettuce, Alqarawi et al. (2014a) in E. aphylla and Abd\_Allah et al. (2015a) in Sesbania sesban. Recently, in saltstressed Brassica juncea, Ahmad et al. (2015) demonstrated the positive impact of Trichoderma harzianum inoculation on growth via improved chlorophyll synthesis. In Ocimum basilicum grown under water stress, inoculation of PGPR (Pseudomonas sp. and Bacillus lentus) increased chlorophyll synthesis as well photosynthetic electron transport and also mitigated the negative impact of water stress (Heidari and Golpayegani, 2012).

Salt stress inhibits the uptake of essential mineral elements, such as K, Mg, Ca, N, and P, because of the antagonistic relationship of sodium. By reducing the uptake of magnesium, salt stress affects plant photosynthetic efficiency by altering the synthesis of chlorophyll molecules. Reduced uptake of nitrogen directly affects the nitrogen metabolic potential as well as amino acid synthesis in plants (Näsholm et al., 2009). Improved plant nutrient uptake under salt stress conditions by AMF was reported in many studies e.g., for common bean (Phaseolus vulgaris)

(Abd\_Allah et al., 2015b), olive (Olea europaea L.) (Porras-Soriano et al., 2009), and wheat (Triticum aestivum L.) (Talaat and Shawky, 2014). Endophytic bacterial strains can also increase plant tolerance to abiotic stresses and improve nutrient uptake under multiple adverse conditions (Malfanova et al., 2011; Berg et al., 2013). The combined inoculation of plants with AMF and PGPR reduced the Na and Cl concentrations in plant tissues, thereby protecting salt-stressed plants from ionic and osmotic stress-induced changes. Their synergistic interaction resulted in an increase of N, P, and K uptake by plants. Endophytic bacterial strains appear to have some plant growth-promoting activities, such as IAA production and solubilization of phosphate, which together or alone might explain the capacity of B. subtilis to improve plant growth and nutrient acquisition (Malfanova et al., 2011). Messele and Pant (2012) observed improved nodulation, yield and P uptake in chickpea (Cicer arietinum) by phosphatesolubilizing Pseudomonas. An increase in P availability to plants through the action of phosphate-solubilizing bacteria (PSB) has also been reported for green gram (Vigna radiata L. Wilczek) (Vikram and Hamzehzarghani, 2008) and wheat (T. aestivum L.) (Panhwar et al., 2011). In earlier studies, it was reported that salt and drought stresses inhibit the production of plant growth regulators in plant tissues (Debez et al., 2001). The additional supply of hormones by endophytes in plant tissue can stimulate the root system, thereby facilitating the absorption of more nutrients from the soil, especially under stress conditions (Malfanova et al., 2011; Berg et al., 2013). Studies by Kavino et al. (2010), Heidari et al. (2011), and Lavakush et al. (2014) also support the positive impact of PGPR on the mineral nutrient status of plants under normal and stressed conditions in determining the physiological strength of a plant. Improved Mg content affects chlorophyll, while P and N contribute to the energy budget of a cell, and Ca serves as an important cellular messenger for downstream signaling (Ahanger et al., 2014b). Improved K uptake is associated with reduced Na uptake in AMF- and PGPR-inoculated plants and results in an enhanced K/Na ratio, an important aspect for the maintenance of physiological cellular functioning (Abd\_Allah et al., 2015a,b; Ahanger et al., 2015).

Phosphatases are responsible for the hydrolysis of a range of organic P compounds and provide mineral phosphate to the plant (Tazisong et al., 2015). In our study, increased acid and ALP was observed in A. gerrardii grown under salt stress. Similarly, enhanced AP activity under salt stress was also observed in Medicago sativa (Ehsanpour and Amini, 2003) and may be due to the fact that salt stress suppresses plant growth, P uptake, transport and utilization (Dracup et al., 1984). An increase in the activities of phosphatases (alkaline and acidic phosphatase) following the combined inoculation of A. gerrardii plants with AMF and B. subtilis support the findings of Amaya-Carpio et al. (2009) in Ipomoea carnea sp. fistulosa and Kebrabadi et al. (2014) in Fraxinus rotundifolia. During P deficiency or impaired P uptake, plants release or activate acid phosphates and exude carboxylates and phosphatases to enhance P solubilization and uptake (Veneklaas et al., 2003). An increase in the activity of phosphatases in inoculated plants, either with AMF or B. subtilis, alone or in combination, can contribute to the release of bound P to maximize its

### REFERENCES


uptake and transport. Maize plants inoculated with B. subtilis showed an increase in phosphatase activity compared to uninoculated controls (Hussain et al., 2013). Salinity-stressed plants showed lower P accumulation in the above-ground plant parts, which may be due to the toxicity of high salt concentrations.

### CONCLUSION

Our observations in this study indicate that endophytic bacteria and AMF that live within the plant tissues of A. gerrardii are coordinately involved in the plant's adaptation to stress tolerance. Inoculation of plants with AMF and endophytic B. subtilis increased plant growth and nutrient acquisition and improved symbiotic performance of A. gerrardii. In addition, endophytic B. subtilis increased AMF germination and root colonization of A. gerrardii under salt stress.

### AUTHOR CONTRIBUTIONS

AAH, Provided the seeds and seedlings. AH, Provided AMF and mycological analysis. EA, All biochemical analysis, writing and graphing. AAA, Statistical analysis. SW, Revision and editing. DE, writing, revision and editing.

### ACKNOWLEDGMENTS

The authors would like to extend their sincere appreciation to the Deanship of Scientific Research at King Saud University for its funding of this research through the Research Group Project no RGP-271.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01089


Enterobacter radicincitans affect nutrient uptake of faba bean and soil biological characteristics. Agric. Fores. Res. 3, 229–234. doi: 10.3220/LBF\_2013\_229-234


banana (Musa spp.) under field conditions. Appl. Soil Ecol. 45, 71–77. doi: 10.1016/j.apsoil.2010.02.003


Zarea, M. J., Karimi, N., Goltapeh, E. M., and Ghalavand, A. (2011). Effect of cropping systems and arbuscular mycorrhizal fungi on soil microbial activity and root nodule nitrogenase. J. Saudi Soc. Agri. Sci. 10, 109–120. doi: 10.1016/j.jssas.2011.04.003

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Hashem, Abd\_Allah, Alqarawi, Al-Huqail, Wirth and Egamberdieva. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# A New Oidiodendron maius Strain Isolated from Rhododendron fortunei and its Effects on Nitrogen Uptake and Plant Growth

Xiangying Wei1,2, Jianjun Chen1,2 \*, Chunying Zhang<sup>3</sup> \* and Dongming Pan<sup>1</sup> \*

<sup>1</sup> College of Horticulture, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> Department of Environmental Horticulrture and Mid-Florida Research and Education Center, Institute of Food and Agricultural Sciences, University of Florida, Apopka, FL, USA, <sup>3</sup> Shanghai Academy of Landscape Architecture Science and Planning, Shanghai, China

#### Edited by:

Gero Benckiser, University of Giessen, Germany

#### Reviewed by:

Raffaella Balestrini, National Research Council, Italy Carolyn Frances Scagel, United States Department of Agriculture – Agricultural Research Service, USA

#### \*Correspondence:

Jianjun Chen jjchen@ufl.edu Dongming Pan pdm666@126.com Chunying Zhang mayzhang55@163.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 26 February 2016 Accepted: 11 August 2016 Published: 23 August 2016

#### Citation:

Wei X, Chen J, Zhang C and Pan D (2016) A New Oidiodendron maius Strain Isolated from Rhododendron fortunei and its Effects on Nitrogen Uptake and Plant Growth. Front. Microbiol. 7:1327. doi: 10.3389/fmicb.2016.01327 A new mycorrhizal fungal strain was isolated from hair roots of Rhododendron fortunei Lindl. grown in Huading Forest Park, Zhejiang Province, China. Morphological characterization and internal transcribed spacer rDNA analysis suggested that it belongs to Oidiodendron maius Barron, and we designated it as strain Om19. Methods for culturing Om19 were established, and the ability of Om19 to form mycorrhizae on R. fortunei was evaluated in a peat-based substrate. Microscopic observations showed hyaline hyphae on the surface of hair roots and crowded hyphal complexes (hyphal coils) inside root cortical cells of R. fortunei after inoculation, indicating that the roots were well colonized by Om19. In a second experiment, fresh and dry weight of R. fortunei 2 months after Om19 inoculation were greater than uninoculated plants, and the total nitrogen absorbed by plants inoculated with Om19 was greater than the uninoculated controls. qRT-PCR analysis of five genes related to N uptake and metabolism (two nitrate transporters, an ammonium transporter, glutamine synthetase, and glutamate synthase) showed that these genes were highly upregulated with twofold to ninefold greater expression in plants inoculated with Om19 compared to uninoculated plants. In the third experiment, Om19 was inoculated into the peat-based substrate for growing Formosa azalea (Rhododendron indica 'Formosa'). 'Formosa' azalea plants grown in the inoculated substrate had larger canopies and root systems compared to uninoculated plants. Our results show that Om19 could be an important microbial tool for improving production of Rhododendron plants.

Keywords: ammonium transporters, ericoid mycorrhiza, nitrate transporters, nitrogen, Oidiodendron maius, Rhododendron

### INTRODUCTION

About 6,000 recognized species of soil fungi form mutualistic symbioses with roots of more than 90% of land plants, which are collectively known as mycorrhizas (Smith and Read, 2008). The symbiosis enables plants to adapt to different soil conditions, improves roots in nutrient acquisition, enhances plant resistance to root pathogen infection, and increases plant growth (Bergero et al., 2000; Smith and Read, 2008). Mycorrhizal fungi are classified as

**86**

ectomycorrhiza and endomycorrhiza depending on whether fungal hyphae colonize root intercellular spaces or the interior of root cells (Bonfante and Genre, 2010). Endomycorrhiza includes arbuscular, ericoid, and orchid mycorrhiza (Smith and Read, 2008).

Ericoid mycorrhizal (ERM) fungi specifically form symbiotic associations with roots of plants in the family Ericaceae (Perotto et al., 2012). Such ERM fungi include Hymenoscyphus ericae (Read) Korf and Kernan, Oidiodendron spp., Cadophora finlandia (Wang and Wilcox) Harrington and McNew, and Scytalidium vaccinii Dalpe, Litten and Sigler (Read, 1996). Among these, Oidiodendron maius Barron is one of the most widely investigated ERM fungi. Its genome has recently been sequenced (Kohler et al., 2015). The symbiosis of O. maius with roots of ericaceous plants is known to facilitate the exchange of nutrients (Rice and Currah, 2006). O. maius also play a crucial role in the protection of host plants against heavy metal toxicity (Daghino et al., 2016). O. maius was first identified by Barron (1962) from collections of peat soil in Canada, and then isolated from roots of an ericaceous plant in Japan (Tokumasu, 1973). Subsequently, O. maius has been recorded as ERM endophytes of several taxa in the Ericaceae (Hambleton and Currah, 1997; Addy et al., 2005) and is especially common in the roots of Rhododendron species (Usuki et al., 2003; Bougoure and Cairney, 2005; Zhang et al., 2009; Tian et al., 2011).

A characteristic of ERM fungi is their ability to improve nitrogen (N) uptake in plants (Read, 1996; Bucking and Kafle, 2015). However, there is still controversy regarding how N, and particularly which form of N, is absorbed by ericaceous plants. Plants can absorb either nitrate (NO<sup>3</sup> <sup>−</sup>) or ammonium. Cranberry (Vaccinium macrocarpon Ait.), a member of the family Ericaceae, was reported to be unable to take up NO<sup>3</sup> <sup>−</sup> as a sole source of N in hydroponic culture (Rosen et al., 1990; Smith, 1993). An explanation is that cranberry has adapted to acidic soil conditions where pH ranges from 4 to 5 and nitrification is typically negligible at soil pH below 5.5 (Paul and Clark, 1989), therefore the adaptation has resulted in the loss of plant capacity to absorb NO<sup>3</sup> <sup>−</sup>. However, recent reports showed that inoculation with the fungus Rhizoscyphus ericae increased the capacity of cranberry to absorb NO<sup>3</sup> <sup>−</sup> (Kosola et al., 2007). Compared to uninoculated plants, inoculating certain blueberry (Vaccinium corymbosum L.) cultivars with ERM fungi can increase nutrient concentration, particularly N, and plant growth (Scagel, 2005). Yin et al. (2010) also found that ERM fungi significantly increased the ability of Rhododendron fortunei to absorb N, especially in the form of NO<sup>3</sup> −.

Nitrate uptake is carried out by two nitrate transporter (NRT) systems: a low-affinity transport system (LATS, active at NO<sup>3</sup> − concentrations higher than 0.2 mM) and a high-affinity transport system (HATS, operating at NO<sup>3</sup> <sup>−</sup> concentrations lower than 0.2 mM; Malagoli et al., 2004). Ammonium is absorbed through ammonium transporters (AMTs; von Wittgenstein et al., 2014). Increased ammonium uptake triggers plant glutamine synthetase (GS) and glutamate synthase (GOGAT) activities as glutamine and glutamate play crucial roles in N metabolism. In the symbiosis of arbuscular mycorrhizal (AM) fungi with host plants, N is absorbed by the extraradical mycelia and converted into the amino acid arginine for transport into the intraradical mycelia (Gomez et al., 2009; Tian et al., 2010). After internal migration is complete, the arginine is broken down through the urease cycle into ammonium for transport into the plant (Govindarajulu et al., 2005). Thus far, whether or not ERM fungi act like AM fungi in facilitating N uptake in ericaceous plants is largely unknown.

An important effort in our mycorrhizal research program has been pursuing a better understanding of the potential of ERM fungi as biofertilizers for improving plant growth of ericaceous species. We believe that valuable O. maius strains could be isolated from hair roots of understorey R. fortunei in Chinese forest parks, and that some isolated strains could improve Rhododendron N uptake and growth in commercial nursery production. This report is intended to document a new strain of ERM fungi isolated from hair roots of R. fortunei and its identification and characterization. The effects of this new strain on plant growth, N absorption, and N metabolism related gene expression were also determined. A better understanding of ERM fungal mediated N uptake in host plants may help improve production of some economically important ericaceous crops such as blueberry, cranberry, and rhododendron.

### MATERIALS AND METHODS

### Isolation of Mycorrhizal Fungi

Plants of R. fortunei with a height less than 40 cm that were grown singly without any plants in a radius of 2 m were identified in Huading Forest Park (29◦ 15<sup>0</sup> N 121◦ 06<sup>0</sup> E), Zhejiang Province, China where mixed pines (Pinus taiwanensis) were grown. Root samples of five identified plants were excavated from the upper 15 cm of the soil using a shovel. Most of the soil was removed from root samples by shaking, and roots were placed in plastic bags and stored at 4◦C until used for fungal isolation.

Fungi were isolated from hair roots by direct plating (Stoyke and Currah, 1991; Zhang et al., 2009). Briefly, collected root samples were soaked in cool tap water and washed gently to remove soil. Hair roots removed from the root samples were surface sterilized by immersion in a 72% ethanol for 30 s, followed by immersion in 10% sodium hypochlorite for 15 min, and then rinsed four times in sterile distilled water. The sterilized hair roots were cut into 0.3–0.5 cm segments, plated on modified Melin-Norkans agar medium (MMN; Xiao and Berch, 1992), and incubated in the dark at 25◦C for 3–5 weeks. A total of 100 root pieces, 20 per plant, were cultured and those producing rapidly sporulating fungi were removed. Slower growing fungi were subcultured on 2% malt extract agar (MEA; Thom and Church, 1926), a total of 84 were isolated and maintained in the dark at 25◦C for 2–4 weeks for morphological identification (described below), and subcultures were also stored at 4◦C until further use. Ten of the slower growing isolates were randomly selected for a preliminary evaluation of their ability to form mycorrhizae. Results from this preliminary evaluation indicated that one isolate showed promise in forming mycorrhizal association with R. fortunei and improving seedling growth. This isolate was selected for further molecular identification and for plant

experiments described below. This isolate was later designated as Om19.

### Morphological Identification of Om19

Morphological characteristics including colony diameter, color, thickness, texture, and pigments, and reverse side of colony color were examined after Om19 was cultured on MEA for 14 days in the dark at 25◦C. To induce conidia and conidiophores, a small volume (about 50 µl) of 20% potato dextrose broth was dropped to each of two holes of a glass slide placed in a 9 cm plastic Petri dish, hyphae of isolate were added, and the slides were incubated in the dark at 25◦C for 7–21 days. Conidia and conidiophores were observed under a light microscope, and fungi were putatively identified with taxonomic keys in Barron (1962) and Domsch et al. (1980).

### Molecular Identification of Om19

The isolate of Om19 was grown in potato dextrose broth on a rotating shaker (150 rpm) at 25◦C in the dark for 5 days. Mycelia were collected by filtrating through Whatman filter paper No. 1 and genomic DNA was isolated from the mycelia using the modified cetyltrimethylammonium bromide method (Gardes and Bruns, 1993). The internal transcribed spacer (ITS) region was amplified using the ITS1 (5<sup>0</sup> TCCGTAGGTGAACCTG CGG 3<sup>0</sup> ) and ITS4 (5<sup>0</sup> TCCTCCGCTTATTGATATGC 3<sup>0</sup> ) primers. The amplifications were performed in a 50 µl reaction volume containing 50 ng of genomic DNA, 50 pmol of each primer, 100 µM of each dATP, dGTP, dCTP, and dTTP, 1 U of Taq polymerase, and 5 µl of PCR buffer. The tubes were incubated at 95◦C for 2 min and then subjected to 35 cycles as follows: 94◦C for 40 s, 60◦C for 40 s, and 72◦C for 45 s; a final incubation was carried out for another 5 min at 72◦C. The PCR products were digested with the restriction endonucleases, HinfI, RsaI, MspI, BsuRI, and TaqI according to the manufacturer's instructions. The restriction fragments were separated by electrophoresis in 2.5% (w/v) agarose gels. The base pair lengths of individual fragments were determined by comparison with a 50-bp ladder. Fragments smaller than 50 bp were not scored.

The PCR products were cloned with the PMD18-T easy vector system and analyzed by an ABI 3730XI automatic DNA sequencer. The sequence was submitted to the NCBI database under the accession number KU382495. The sequences of mycorrhizal species that most closely matched the sequence of Om19 were obtained by Basic Local Alignment Search Tool (BLAST) from the GenBank database. The alignments were performed with ClustalX and then adjusted to optimize the aligned sites. The sequences and selected fungi from GenBank database were analyzed by neighbor joining using distances from Kimura's two-parameter model with the MEGA5.0 software system. To assess support for nodes, 1,000 bootstrap replications were performed.

### Mycorrhizae Synthesis of Om19

The Om19 was examined for its ability to colonize hair roots of R. fortunei seedlings. Seeds of R. fortunei were rinsed in running tap water for 2 h and surface sterilized four times, 5 min each, using 25% (v v−<sup>1</sup> ) solution of commercial bleach (8.25% NaOCl) followed by rinsing in sterile distilled water three times. The sterilized seeds were germinated on a half-strength Economou and Read medium (Economou and Read, 1984). The medium was supplemented with 1.5% (w v−<sup>1</sup> ) sucrose and 0.7% (w v−<sup>1</sup> ) agar with pH adjusted to 5.2, autoclaved at 121◦C for 30 min, and 30 ml was transferred into each culture vessel (150 ml). Seeds were germinated in a culture room under a 16 h photoperiod provided by cool-white fluorescent lamps at a photon flux density of 50 µmol m−<sup>2</sup> s −1 and a temperature of 25◦C.

A peat based substrate was formulated by mixing dry Klasmann peat (Geeste, Germany) with dry sand (initially washed two times with tap water, then five times with deionized water and dried) at 2:1 ratio based on volume. The organic matter and total N content of the Klasmann peat were 965.3 g/kg and 0.89%, respectively. A MMN nutrient solution devoid of malt extract and glucose was prepared where the N source was replaced by Ca(NO3)<sup>2</sup> with a final N concentration at 3.79 mM and its pH was adjusted to 5.2. The substrate was moistened with the modified MMN solution at 3:2 ratio by volume, and its pH was tested to be 5.2 by press extraction method (Scoggins et al., 2002). Twenty cylindrical vessels (400 mL) were filled with 100 mL of substrate, covered with caps, and autoclaved at 121◦C for 30 min.

Two months after seed germination, seedlings were transferred under aseptic conditions to the culture vessels containing the sterilized peat-based substrate with five seedlings per vessel. Mycelium of Om19, cultured on MEA for 2 weeks, was collected using a sterile 5-mm cork borer. After extra medium beneath the top 1 mm layer of MEA with hyphae was removed, the 5-mm diameter disks were cut into half and inoculated into the peat-based substrate next to each of the five R. fortunei seedlings. Ten culture vessels were inoculated with Om19, and the other 10 were used as control without inoculation. The experiment was arranged as a randomized complete block design with 10 blocks (replications). Plants were grown in a culture room under the conditions described above. After 2 months, shoots were harvested from each vessel and roots were removed from substrate by rinsing with sterile deionized water. At harvest, the number of leaves and roots, and the length and fresh weight of shoots (both stem and leaves) and roots (entire roots) of each seedling were recorded, and mean of five seedlings per vessel were calculated.

Roots from one randomly selected seedling per vessel were immediately fixed in formaldehyde–acetic acid–ethanol (FAA) for 24 h and cleared at 90◦C for 1 h in 10% KOH. The roots were then acidified with 1% HCl and stained in a lactophenoltrypan blue (0.05% trypan blue in lactophenol) for 5 min at 90◦C (Phillips and Hayman, 1970). Stained roots were cleared with fresh lactophenol, cut to 5 mm segments, and examined under a light microscope for ERM structures. Root colonization was quantitatively assessed using the method described by Biermann and Linderman (1981) as the percentage of root length with internal hyphal coils. A minimum eight segments from the control plant and 25 segments from the Om19-inoculated plant randomly selected per vessel were examined under the light microscope. Additionally, roots of randomly selected seedlings from the two treatments were observed under scanning electronic

microscope (SEM) using the method described by Bonfante-Fasolo and Gianinazzi-Pearson (1979). All specimens were coated with gold and platinum and examined using a SEM at 20 kV.

### Effects of Om19 on Plant Growth and N Uptake

Two experiments were performed to determine the effect of Om19 on Rhododendron growth and N uptake. In the first experiment, 60 culture vessels (400 ml) were filled with 100 ml of a peat-based substrate prepared as described above. Twomonth old seedlings of R. fortunei were transplanted into culture vessels (five per vessel), and 30 vessels were inoculated with Om19 as described above. The remaining 30 were uninoculated as the control treatment. The experiment was arranged as a randomized complete block design with five blocks, and each treatment had six vessels per block. Plants were grown in a culture room under the conditions described above without supplying any additional nutrients. Two months after inoculation, seedlings were collected by gently washing away of substrate from roots. A root with a length of 3 cm was sampled from seedlings of five randomly selected vessels per treatment and used for examining Om19 colonization as described above. Fresh weight of seedlings from each block (30 seedlings) was recorded after blotting with paper towel, and dry weights were also determined after oven-drying at 80◦C for 48 h. Substrate pH at the end of the experiment was also recorded using the press extraction method. The seedlings (30 seedlings together) from each block were analyzed for N content using CNS Auto-Analyzer (VarioMAX, Elementar Americas, Mt. Laurel, NJ, USA), and total N per 30 seedlings was calculated.

In a second experiment, Om19 was evaluated in a greenhouse with commercial production practices. Rooted cuttings of a popular Formosa azalea (Rhododendron indica 'Formosa') were potted in a peat-based substrate (peat and sand at 2:1 ratio by volume) in 20 10-cm diameter containers (0.5 L). The Om19 inoculum was prepared by transferring 6 plugs (5-mm-diameter) of MEA culture of Om19 into 1 L of liquid MEA and incubating in a rotary shaker (1,500 rpm) at 25◦C for 10 days in the dark. The culture was fragmented with a blender, diluted with sterile deionized water in a 1:1 ratio, and inoculated by pouring the inoculum mix onto the substrate of 10 containers, 10 ml each. Plants in the other 10 containers were used as controls by receiving 10 ml of 50% of diluted MEA without the inoculum. The experiment was set as a randomized complete block design with 10 blocks. A week after transplanting, all plants were fertigated with Peters Professional 20–20–20 General Purpose Fertilizer (Scotts Co., Marysville, OH, USA) with N at 100 mg L −1 and every other week thereafter. Plants were grown in a shaded and evaporative pad cooled greenhouse under a maximum photosynthetically active radiation of 285 µmol m−<sup>2</sup> s −1 . Temperatures ranged from 18.3 to 32.2◦C and relative humidity from 50 to 80%. After 4 months, the number of leaves and canopy height of each plant were recorded, and substrate was removed from roots by washing with tap water. Plants were photographed while roots were immersed in water.

### qRT-PCR Analysis of Genes Related to N Metabolism

Expression of some key genes related to N uptake and N metabolism in Om19 inoculated and uninoculated plants was investigated in an experiment similar to the above culture room experiment with R. fortunei. Seedlings of R. fortunei were grown in 400 ml vessels containing the sterilized peat-based substrate described above. A total of 120 vessels were prepared in order to collect enough root tissue for RNA extraction. Sixty vessels were inoculated with Om19 and the rest were uninoculated as controls. The experiment was arranged as a randomized complete block design with three blocks and each treatment was replicated 20 times per block. Plants were grown in a culture room as described above. A root with a length of 3 cm was sampled from seedlings of five randomly selected vessels per treatment and assessed for root colonization as describe above. All roots were harvested 2 months after inoculation, and total RNAs were extracted with TRIzol reagent (Invitrogen, USA) and treated with RNase-free DNase I (Takara Biotechnology, China). qRT-PCR was carried out to analyze the expression level of RfNRT1-1 and RfNRT1-2 (NRTs), RfAMT (ammonium transporter), RfGS (GS), and RfGOGAT (glutamate synthase). These genes were isolated from roots of R. fortunei, and their sequences were submitted to the GenBank with accession numbers KX506094, KX506095, KX506096, KX506097, and KX506098 for RfAMT, RfNRT1-1, RfNRT1-2, RfGS, and RfGOGAT, respectively. The sequence homology of these genes to those in other plant species is presented in Supplementary Table S1. Primers specifically for R. fortunei were designed according to the cDNAs with Primer Premier software (version 5.0; Supplementary Table S2). EF1α was used as an internal control gene. The first strand cDNA was synthesized using the PrimeScript first cDNA Synthesis Kit (Takara, Dalian, China). qRT-PCR was performed in a 20 µL reaction mixture containing 2x SYBR Master Premix Ex Taq II 12.5 µL (Takara, Dalian, China), 1 µL of cDNA template (1:5 dilution), and 1 µL of each corresponding primer for the genes of interest and EF1α. qRT-PCR of three biological replicates (20 seedlings from each block as a replicate) for each sample was performed for 5 s at 95◦C, 10 s at 56◦C, and 20 s at 72◦C using a LightCycler 480 II System (Roche Applied Science, Indianapolis, IN, USA). The relative expression levels were normalized and calibrated according to the 2−11CT method (Schmittgen and Livak, 2008) where EF1α was an internal reference and seedlings before Om19 inoculation was control. For a given gene, the relative expression level was expressed as mean ± SE with three replicates.

### Data Analysis

Plant growth data from mycorrhizae synthesis and fresh and dry weight as well as N content data from plant-growth experiments were analyzed separately by experiments using a mixed model where treatment effects were considered fixed and block effects considered random. After checking for normality and homogeneity of variances using SPSS Statistics (SPSS Inc., Chicago, IL, USA), all data were subjected to analysis of variance using SPSS 19.00 statistical software, and the significance probability for treatment effects was evaluated at P ≤ 0.05 or 0.01 level.

### RESULTS

### Morphological Characteristics

fmicb-07-01327 August 20, 2016 Time: 14:30 # 5

The Om19 formed a thin gray-white colony after incubation in the dark at 25◦C for 14 days on MEA medium. Diameter of the colony ranged from 20 to 25 mm (**Figure 1A**). Hyphae of the strain were smooth and transparent, diameter of the hyphae ranged from 1.2 to 2.0 µM. When cultured on 20% potato dextrose broth, tall, and erect conidiophores bearing a head of divergent, branched undulating chains of conidia were viewed under microscope (**Figure 1B**). Conidia had thin-walled, hyaline, subglobose to elongated spores (**Figure 1C**) that appeared to have a single nucleus per spore (**Figure 1D**).

### Molecular Characterization

Molecular analysis of Om19 resulted in a 457 bp ITS rDNA sequence (**Figure 2**). The ITS rDNA sequence was compared to the available sequences obtained by BLAST from the GenBank database. The sequence has 99.8% identity to the sequence of O. maius var. citrinum UAMH 1525 and O. maius UAMH 1540. Thus, the strain was placed in the cluster with O. maius var. citrinum UAMH 1525 and O. maius UAMH 1540, which is supported by a bootstrap of 100% (**Figure 3**). Both O. maius var. citrinum UAMH 1525 and O. maius UAMH 1540 were identified by Barron (1962). The former was isolated from a cedar bog, Canada and the latter was isolated from peat soil in Canada. The Om19 has 91.6 and 92.9% similarity to the other Oidiodendron species and also to Myxotrichum cancellatum UAMH1991.

### Mycorrhizae Synthesis

Microscopic examination of inoculated R. fortunei seedlings showed that roots were infected by hyphae (**Figure 4A**); the intracellular hyphal growth was observed in epidermal cells (**Figure 4B**). Some root epidermal and cortical cells completely filled with mycelium after inoculation (**Figure 4C**). The mean percent root length colonized by Om19 ranged from 65 to 72%. There were no fungal structures in root cortical cells of control seedlings (**Figure 4D**). SEM observation also showed that mycelia heavily surrounded roots of seedlings inoculated with Om19 (**Figure 4E**), but the surfaces of control roots were clear (**Figure 4F**). Transverse section of a hair root from a plant inoculated with Om19 showed dense hyphal growth in epidermal cells and a few cortical cells. Some of the epidermal cells in roots from plants inoculated with Om19 appeared to be deformed (**Figure 4G**). Transverse section analysis of the control root showed no hyphal growth inside root cells (**Figure 4H**).

Seedlings inoculated with Om19 were almost two times larger compared to the controls (**Table 1**). Inoculated plants had more roots, greater root length, longer shoots and greater shoot and root fresh weight (**Table 1**).

FIGURE 1 | Morphological characteristics of the mycorrhizal fungus Oidiodendron maius var. maius strain Om19 isolated from hair roots of Rhododendron fortunei. (A) Colony of Om19 cultured on malt extract agar (MEA) medium for 14 days. (B) Dematiaceous conidiophores of Om19. (C) Cylindrical arthroconidia of Om19. (D) Cylindrical arthroconidia with one nucleus.

FIGURE 3 | Neighbor-joining phylogenetic tree based on ITS rDNA sequence data of mycorrhizal fungus O. maius var. maius strain Om19 isolated from hair roots of R. fortunei, along with known ericoid endophytes and selected fungal species from GenBank with high sequence similarity. Numerical values above the branches indicate bootstrap percentiles from 1000 replicates. Bootstrap numbers over 50% are

indicated. Horizontal branch lengths are proportional to the scale of substitutions.

### Plant Growth and N Uptake

When R. fortunei seedlings were grown in a peat-based substrate for 2 months, substrate pH ranged from 4.9 to 5.2 across all treatments. Similar to the mycorrhizae synthesis experiment, roots inoculated with Om19 were colonized, and the percent root length colonization ranged from 65 to 72%. The Om19-colonized seedlings grew substantially larger than controls (**Figure 5**). After 2 months of growth, seedlings inoculated with Om19 measured 105% greater fresh weight and 84% greater dry weight than the control seedlings (**Table 2**). Inoculation with Om19 had no significant influence on N concentrations in seedlings, however, total N of inoculated plants was 61% greater than the controls.

Results from the greenhouse trial using commercial production practices showed that Om19 inoculation promoted the growth of rooted cuttings of Formosa azalea (**Figure 6**). Om19-inoculated Formosa azalea plants had more leaves (23) and were taller (16 cm) compared to control plants (14 leaves and 10 cm tall). Furthermore, plants inoculated with Om19 appeared to have more abundant roots compared to the controls (**Figure 6**).

### qRT-PCR Analysis

The Om19 colonized the roots of R. fortunei seedlings with the percent root length colonization similar to the mycorrhizae synthesis experiment. The qRT-PCR analysis of five genes related to N uptake and metabolism showed that all were increasingly expressed in roots of R. fortunei seedlings colonized by Om19 compared to those uninoculated controls (**Figure 7**). The highest expression was RfNRT1-1, almost a ninefold increase compared to its counterpart in the control seedlings. The expression of RfNRT1-2 was 3.5, RfAMT was upregulated more than threefold. RfGS expression was 2.8-fold higher, and the expressions of RfGOGAT was fivefold.

### DISCUSSION

The present study isolated a new O. maius strain called Om19 from hair roots of R. fortunei grown under mixed pine forests in east China. The Om19 was able to colonize roots of R. fortunei, and the colonization improved the growth of both R. fortunei and Formosa azalea. Investigation of the expression of genes related to N uptake (RfAMT, RfNRT1-1, and RfNRT1-2) and N metabolism (RfGS and RfGOGAT) showed that they were highly upregulated in Om19-colonized roots of R. fortunei. Total N of inoculated plants was significantly greater and dry weights were much higher than in the non-colonized control seedlings. The Om19 is promising and could potentially be used as a microbial agent for improving Rhododendron and possibly blueberry production.

### Isolate Identification

Rhododendron fortunei is native to mountainous areas in east China (Liberty Hyde Bailey Hortorum, 1976). The Huading Forest Park is one of the centers of R. fortunei origin. We were able to identify 84 isolates from five plant roots. As indicated by Zhang et al. (2010) and Tian et al. (2011), R. fortunei roots harbor a rich microbial ecosystem in their endogenous habitat.

The isolated Om19 reproduced asexually by forming conidia with a single haploid nucleus (**Figure 1**) and was morphologically similar to O. maius as described by Barron (1962) and Rice and Currah (2005). ITS rDNA analysis showed a high degree of sequence identity (99.8%) to O. maius as reported by Sigler and Gibas (2005) and Zhang et al. (2009) and placed it in the O. maius cluster in the neighbor-joining phylogenetic tree (**Figure 3**). It thus was named as O. maius Om19.

The Om19 can form mycorrhizae with roots of R. fortunei. Root epidermal and cortical cells were filled with mycelium, and transverse section of a hair root showed hyphal growth in epidermal and cortical cells (**Figure 4**). Seedlings colonized by Om19 were significantly larger with more roots (**Table 1**; **Figures 5** and **6**). Our results concurs with those of Jansa and Vosatka (2000), who documented that ERM fungi promoted

FIGURE 4 | Microscopic observation of R. fortunei roots inoculated or not with mycorrhizal fungus O. maius var. maius strain Om19. (A) A root was infected by Om19 hyphae after inoculation. (B) Hyphal growth inside epidermal cells of roots from inoculated plants. (C) Hyphae proliferation through all epidermal cells of a root. (D) Roots of control seedlings (without inoculation) with no mycorrhizal colonization. (E) SEM showing mycelia surrounding roots of inoculated seedlings. (F) No mycelia were visible on the surface of control roots. (G) Transverse section of a hair root from an inoculated seedling showing hyphal growth in the cortical cells and some distorted cells. (H) Transverse section of a hair root from a control seedling showing no hyphal growth.

growth of Rhododendron microcutting in peat-based substrates. Yin et al. (2010) also reported that R. fortunei plants inoculated with an EMR fungus grew better than the uninoculated controls. These results support our hypothesis that valuable O. maius strains could be isolated from hair roots of understorey R. fortunei in Chinese forest parks.



∗∗ indicate significant difference between seedlings inoculated and uninoculated with Om19 at P < 0.01 level (n = 10).

FIGURE 5 | Seedling of R. fortunei inoculated (A) or not (B) with mycorrhizal fungus O. maius var. maius strain Om19 and grown in a peat-based substrate for 2 months.

TABLE 2 | Fresh and dry weights, tissue nitrogen (N) concentration, and total N of 30 R. fortunei seedlings grown for 2 months in a peat-based substrate inoculated or not with mycorrhizal fungus Om19.


<sup>∗</sup> and ∗∗ indicate significant difference between seedlings inoculated and uninoculated with Om19 at P < 0.05 and P < 0.01 levels (n = 5).

### Plant N Uptake and Growth

This study evaluated N uptake and seedling growth of R. fortunei inoculated with Om19 under well controlled environmental conditions. Total N absorbed by Om19-colonized seedlings was 61% greater than the control seedlings. Considering the fact that the readily available N in the substrate was NO<sup>3</sup> <sup>−</sup> and that the pH

FIGURE 6 | Growth of Formosa azaleas (Rhododendron indica 'Formosa') inoculated or not with mycorrhizal fungus O. maius var. maius strain Om19, and grown in a shaded greenhouse for 4 months. Containers with uninoculated (A) and inoculated plants (B). Uninoculated (C) and inoculated (D) plants removed from substrate and roots submerged in water.

FIGURE 7 | The expression level of five genes (RfNRT1-1, RfNRT1-2, RfAMT, RfGS, and RfGOGAT) related to N-uptake and metabolism in roots of R. fortunei seedlings 2 months after being inoculated with a mycorrhizal fungus O. maius var. maius strain Om19. The expression levels were normalized based on the expression of the internal control gene and corresponding gene expressed in the control seedlings (seedlings before Om19 inoculation). The bars represent standard errors of three replicates (n = 3).

of the substrate ranged from 4.9 to 5.2, the higher total N value in the Om19-colonized plants may suggest that the mycelium of Om19 could act in a similar function as AM fungi (Gomez

et al., 2009; Tian et al., 2010) by absorbing NO<sup>3</sup> <sup>−</sup>, converting it into arginine, and releasing ammonium to plants. Ammonium release would increase AMT expression and also trigger plant GS and GOGAT activities. As shown in **Figure 7**, the expression of RfAMT increased more than threefold in Om19-colonized roots. Genes in the family of AMT varied from 6 to 14, depending plant species; AMT1-1 and AMT1-3 were reported to contribute to 30–35% of ammonium transport in plants (Masclaux-Daubresse et al., 2010). Whether or not the RfAMT identified in this study belongs to AMT1-1, AMT1-3, or other AMT requires further investigation. Concomitantly, the expression of RfGS increased 2.7 and RfGOGAT 5.1-fold in Om19-inoculated roots (**Figure 7**). GS is a key metabolic enzyme that synthesizes glutamine from glutamate, leading to the entrance of organic N in cellular metabolic pathways such as the biosynthesis of amino acids, nucleic acids, and complex polysaccharides. As a result, irrespective of low pH in the substrates, Om19 colonized roots were able to take up more applied N and improve plant growth.

Besides Om19-mediated uptake of NO<sup>3</sup> <sup>−</sup>, Rhododendron seedlings per se could directly absorb NO<sup>3</sup> <sup>−</sup> from the substrate. This is because NO<sup>3</sup> <sup>−</sup> was the readily available N in the peat-based substrate and both RfNRT1-1 and RfNRT1-2 were highly upregulated in Om19-colonized seedlings (**Figure 7**). As mentioned previously, plants have LATS and HATS for NO<sup>3</sup> − uptake. In general, the LATS consists of the NRT1 family and the HATS comprises the NRT2 family (Sun and Zheng, 2015). There are 53 NRT1 genes and 7 NRT2 genes in Arabidopsis (Tsay et al., 2007). Recent studies showed that NRT1.1 is actually a dual-affinity transporter regulating NO<sup>3</sup> <sup>−</sup> uptake by changing its affinity for NO<sup>3</sup> <sup>−</sup> depending on the availability of NO<sup>3</sup> <sup>−</sup> in the soil (Tsay, 2014; Sun and Zheng, 2015). Since the readily available NO<sup>3</sup> <sup>−</sup> was limited in the peat-based substrate, NRT1.1 might play an important role in NO<sup>3</sup> <sup>−</sup> absorption by changing its affinity for NO<sup>3</sup> <sup>−</sup>. At the present, we are not completely certain if either RfNRT1-1 or RfNRT1-2 in R. fortunei plays the same roles as NRT1.1 in Arabidopsis. The expression of RfNRT1-1 and RfNRT1-2 does suggest that NRTs were active in Om19-colonized roots.

The next question is why the upregulation of RfNRT1-1 and RfNRT1-2 was greater in Om19-colonized roots than in control roots. One explanation could be that Om19 mediated N uptake enhanced GS/GOGAT cycle (Lea and Forde, 1994), increased Rubisco activity (Masclaux-Daubresse et al., 2010), and elevated NRT gene expression, thus NO<sup>3</sup> − uptake. Wirth et al. (2007) showed that upregulation of NRTs (AtNRT2.1 and AtNRT 1-1) was related to the concentration of glucose 6-phosphate. The direct coupling of NO<sup>3</sup> <sup>−</sup> assimilation and photosynthesis in chloroplasts is energy efficient and is known as nitrate photoassimilation (Searles and Bloom, 2003). Additionally, several NRT genes were reported to play dual nutrient transport/signaling roles, called transceptors, sensing N availability in soil and regulating N uptake and allocation in plants (Gojon et al., 2011; Krapp et al., 2014; Zhang et al., 2014). In a study of NRT gene expression in tomato plants, the expression of NRT2;3 was higher in AM-colonized tomato roots than in controls; this was explained as AM-colonization positively affecting nitrate uptake from soil and nitrate allocation to the plant partner (Hildebrandt et al., 2002). The increased uptake of NO<sup>3</sup> <sup>−</sup> may also improve plant absorption of other ions. For example, a sevenfold increase in N uptake by rhododendron (Rhododendron 'H-1 P.J.M.') was associated with a threefold– fourfold increase in the uptake rate of phosphorus, potassium, and sulfur, and ∼twofold increase in the uptake rate of magnesium and calcium (Scagel et al., 2008). Similar results were observed in blueberry cultivars inoculated with ERM fungi (Scagel, 2005).

The higher level of total N absorbed by Om19-colonized plants (**Table 2**) may also suggest that Om19 might enzymatically degrade organic N from substrate. A recent study showed that O. maius symbionent expressed a full complement of plant cell wall-degrading enzymes in symbiosis, suggesting its saprotrophic ability in sphagnum peat (Kohler et al., 2015). ERM fungi are able to gain access to polymeric sources of N in the form of peptides, pure proteins, or protein-polyphenol complexes (Smith and Read, 2008). The Klasmann peat contains 0.89% N, the decomposition may release organic N for plant absorption. At this point, whether Om19 could enzymatically degrade organic N from the peat is unknown. Further studies for determining its saprotrophic ability and its practical application to be a microbial fertilizer will be fully explored.

Nevertheless, this study suggests that the Om19 can effectively colonize R. fortunei, and the colonization improves plant absorption of NO<sup>3</sup> <sup>−</sup> under acidic growing conditions. Our results thus disagree with the notion that NO<sup>3</sup> <sup>−</sup> availability is always negligible in acid soils (Paul and Clark, 1989) but support the claims of Scagel (2005) and Kosola et al. (2007) that NO<sup>3</sup> <sup>−</sup> is an important N source for plants in the family Ericaceae. ERM colonization under acidic soil conditions probably plays a critical role for plants in the heather family to capture NO<sup>3</sup> <sup>−</sup>-N.

### Practical Application

Plants in the family Ericaceae have some unique characteristics: (1) cortical cells never form root hairs, instead their roots are known as hair roots, (2) they have adapted to soils with low pH and low nutrient status, and (3) roots are commonly associated mycorrhizal fungi, mainly ERM fungi. Increasing evidence suggests that the ability to form symbiotic relationships with ERM fungi is critically important to the success of ericaceous plants in these stressful environments (Read, 1996; Cairney and Meharg, 2003; Perotto et al., 2012). Some of these plants, such as cranberry, bilberry, blueberry, and rhododendrons are economically important horticultural crops. However, production of these crops in a sustainable manner has been challenged.

Commercially these crops are produced in acidic soils or acidic substrates, and chemical fertilizers must constantly be applied. Applied N often is leached or is less available to plants because of acidic growing conditions. N leaching has been an environmental concern in commercial production of horticultural crops (Goulding, 2006). Different strategies have been proposed to reduce N leaching (Chen et al., 2001), but little attention has been given to the microbial fertilizers for improving N use efficiency and reducing nutrient leaching (Scagel, 2005). Berruti et al. (2016) recently documented that AM

fungi as natural biofertilizers can positively affect plant growth in both controlled and open-field conditions. The present study shows that isolate Om19 is able to colonize R. fortunei in an acidic substrate and effectively use applied NO<sup>3</sup> <sup>−</sup>, resulting in increased plant growth. It is worth noting that Om19-inoculated Formaosa azaleas grown in a greenhouse fertilized with a Peters Professional 20–20–20 had visually larger root systems than uninoculated plants. These results suggest that Om19 may be able to colonize different Rhododendron species when plants are grown with a common commercial fertilizer containing different forms of N. Taken together, our results demonstrate that Om19 is a potentially important ERM fungus and can be used as a biofertilizer for improving production of Rhododendron and possibly other ericaceous plants.

### AUTHOR CONTRIBUTIONS

CZ, DP, and JC conceived and designed the experiments. XW conducted the experiments, analyzed the data, and drafted the manuscript. JC participated in data analysis, wrote and revised the manuscript. The final version was approved by all authors.

### REFERENCES


### FUNDING

The authors would like to thank National Natural Science Foundation of China (No. 30972409) and the Scientific Research Foundation of Graduate School at the Fujian Agriculture and Forestry University (324-1122YB026) for supporting this study.

### ACKNOWLEDGMENT

Dr. Hao Sun at the Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang, China for assistance in submission of the ITS sequence, and Mrs. Barb Henny for critical review of the manuscript.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01327


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Wei, Chen, Zhang and Pan. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fmicb-07-01327 August 20, 2016 Time: 14:30 # 11

# The Abundance of Endofungal Bacterium Rhizobium radiobacter (syn. Agrobacterium tumefaciens) Increases in Its Fungal Host Piriformospora indica during the Tripartite Sebacinalean Symbiosis with Higher Plants

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Marc Tad Nishimura, Colorado State University, USA Philipp Franken, Leibniz Institute of Vegetable and Ornamental Crops, Germany

#### \*Correspondence:

Stefanie P. Glaeser stefanie.glaeser@umwelt.unigiessen.de Karl-Heinz Kogel karl-heinz.kogel@agrar.uni-giessen.de

> †These authors shared first authorship.

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Microbiology

Received: 07 November 2016 Accepted: 28 March 2017 Published: 13 April 2017

#### Citation:

Guo H, Glaeser SP, Alabid I, Imani J, Haghighi H, Kämpfer P and Kogel K-H (2017) The Abundance of Endofungal Bacterium Rhizobium radiobacter (syn. Agrobacterium tumefaciens) Increases in Its Fungal Host Piriformospora indica during the Tripartite Sebacinalean Symbiosis with Higher Plants. Front. Microbiol. 8:629. doi: 10.3389/fmicb.2017.00629 Huijuan Guo<sup>1</sup>† , Stefanie P. Glaeser<sup>2</sup> \* † , Ibrahim Alabid<sup>1</sup> , Jafargholi Imani<sup>1</sup> , Hossein Haghighi<sup>2</sup> , Peter Kämpfer<sup>2</sup> and Karl-Heinz Kogel<sup>1</sup> \*

1 Institute of Phytopathology, Research Centre for BioSystems, Land Use and Nutrition, Justus-Liebig-University Giessen, Giessen, Germany, <sup>2</sup> Institute of Applied Microbiology, Research Centre for BioSystems, Land Use and Nutrition, Justus-Liebig-University Giessen, Giessen, Germany

Rhizobium radiobacter (syn. Agrobacterium tumefaciens, syn. "Agrobacterium fabrum") is an endofungal bacterium of the fungal mutualist Piriformospora (syn. Serendipita) indica (Basidiomycota), which together form a tripartite Sebacinalean symbiosis with a broad range of plants. R. radiobacter strain F4 (RrF4), isolated from P. indica DSM 11827, induces growth promotion and systemic resistance in cereal crops, including barley and wheat, suggesting that R. radiobacter contributes to a successful symbiosis. Here, we studied the impact of endobacteria on the morphology and the beneficial activity of P. indica during interactions with plants. Low numbers of endobacteria were detected in the axenically grown P. indica (long term lab-cultured, lcPiri) whereas mycelia colonizing the plant root contained increased numbers of bacteria. Higher numbers of endobacteria were also found in axenic cultures of P. indica that was freshly reisolated (riPiri) from plant roots, though numbers dropped during repeated axenic re-cultivation. Prolonged treatments of P. indica cultures with various antibiotics could not completely eliminate the bacterium, though the number of detectable endobacteria decreased significantly, resulting in partial-cured P. indica (pcPiri). pcPiri showed reduced growth in axenic cultures and poor sporulation. Consistent with this, pcPiri also showed reduced plant growth promotion and reduced systemic resistance against powdery mildew infection as compared with riPiri and lcPiri. These results are consistent with the assumption that the endobacterium R. radiobacter improves P. indica's fitness and thus contributes to the success of the tripartite Sebacinalean symbiosis.

Keywords: P. Indica, endofungal bacteria, tripartite symbiosis, endobacteria, plant growth promotion bacteria, endophytes, root colonization

### INTRODUCTION

fmicb-08-00629 April 11, 2017 Time: 16:11 # 2

Plant-microbe interactions have promoted the development and evolution of land plants. In mutualistic relationships, the plant host benefits from associated microbes by increased mineral nutrient supply, higher tolerance against abiotic stress and increased resistance to pathogens and pests, while the beneficial microbes profit from photosynthates and find protection in the plant's sphere (Kim et al., 1997; German et al., 2000; Van Wees et al., 2008; Bonfante and Anca, 2009; Pieterse et al., 2009; Breuillin-Sessoms et al., 2015). Rhizobacteria and mycorrhizal fungi are typical examples of mutualistic microbes. For instance, the biological nitrogen fixing Rhizobium spp. converts stable nitrogen gas into the biologically useful form ammonia as nitrogen source for the leguminous family (Long, 1989; Parniske, 2000; Oldroyd et al., 2011). Next to their nutritional benefits, arbuscular mycorrhizae fungi (AMF) also induce resistance in plants against necrotrophic pathogens and insects through jasmonic acid and ethylene (JA-ET) dependent signaling pathways (Liu et al., 2007; Pozo and Azcón-Aguilar, 2007). In general, beneficial effects on plants have a high relevance in natural and agricultural ecosystems because of the reduced need of industrial fertilizer in agricultural soils (Yang et al., 2009; Weyens et al., 2009).

Knowledge on the interaction of beneficial fungi and terrestrial plants (bipartite relationships) became even more complex once endofungal bacteria were discovered that form tripartite symbioses with fungi and plants. Endofungal bacteria as symbionts residing in mycelia and spores were first described as Bacteria-Like Organisms (BLOs) in 1970s (Mosse, 1970). Further research showed that such bacteria are either vertically transmitted through vegetative spores, or horizontally transmitted when they are released by the fungal host and subsequently infect newly developed mycelium (Partida-Martinez et al., 2007; Lackner et al., 2009). Endofungal bacteria have been discovered in three clades of beneficial fungi including AMF (Bonfante and Anca, 2009; Naumann et al., 2010; Salvioli et al., 2016), ectomycorrhizal basidiomycetes (Bertaux et al., 2005) and Sebacinalean endophytes (Sharma et al., 2008; Sharma and Kogel, 2009). Bacteria associated with fungi of the genera Piriformospora and Sebacina belong to two genera of Gram-negative (Rhizobium and Acinetobacter) and two genera of Gram-positive (Paenibacillus and Rhodococcus) bacteria (Sharma et al., 2008). Moreover, substantial understanding about the complex role of such bacteria came from the discovery of endobacteria in the rice pathogenic fungus Rhizopus microsporus (Partida-Martinez and Hertweck, 2005; Moebius et al., 2014).

Rhizobium radiobacter was first discovered in the cytoplasm of the root-colonizing, endophytic Sebacinalean fungus Piriformospora indica. The strain R. radiobacter F4 (RrF4) was isolated from its fungal host and could be grown in axenic and liquid cultures (Sharma et al., 2008). Using denaturing gradient gel electrophoresis (DGGE) of 16S rRNA gene fragments amplified with universal bacterial 16S rRNA gene targeting primers, one single DNA band was detected in fungal DNA extracts which had the same motility in the DGGE gel as the 16S rRNA gene product amplified from the pure culture of RrF4. These data confirmed that P. indica contains a single bacterial species. RrF4 was described as a rod-shaped, Gram-negative bacterium and identified by 16S rRNA gene sequencing and genome comparisons as the Alphaproteobacterium R. radiobacter (syn. Agrobacterium tumefaciens, syn. "Agrobacterium fabrum"; Sharma et al., 2008; Glaeser et al., 2016). Fluorescence in situ hybridization (FISH) and PCR amplification detected low numbers of R. radiobacter in P. indica (Sharma et al., 2008) similar to the ectomycorrhizal Laccaria bicolor that contains 1-20 bacteria per fungal cell (Bertaux et al., 2003, 2005). Interestingly, the isolated strain RrF4 and its fungal host P. indica showed similar colonization pattern in plant roots as they colonized mainly the maturation zone, and entered into the rhizodermal and cortical layers (Deshmukh et al., 2006; Schäfer et al., 2009; Jacobs et al., 2011; Qiang et al., 2012; Glaeser et al., 2016). In contrast to other endobacteria, the genome size of RrF4 cells is not reduced (Lackner et al., 2011; Fujimura et al., 2014; Naito et al., 2015; Glaeser et al., 2016). The full genome sequencing shows high similarity to the plant pathogenic A. tumefaciens (syn. "Agrobacterium fabrum") C58 except vibrant differences in two plasmids, especially the tumor-inducing plasmid (pTi) without T-DNA on it (Goodner et al., 2001; Wood et al., 2001; Lassalle et al., 2011; Slater et al., 2013; Glaeser et al., 2016).

Plant inoculated with RrF4 showed increased shoot and root biomass and pathogen resistance against the powdery mildew fungus Blumeria graminis f. sp. hordei (Bgh) in barley, Pseudomonas syringae pv. tomato DC3000 (Pst) in Arabidopsis thaliana, and Xanthomonas translucens pv. translucens (Xtt) in wheat, resembling the beneficial activity induced by P. indica (Waller et al., 2005; Sharma et al., 2008; Varma et al., 2012; Oberwinkler et al., 2014; Ye et al., 2014; Glaeser et al., 2016). Moreover, systemic resistance mediated by RrF4, alike P. indica, requires a functional jasmonate-based ISR pathway (Stein et al., 2008; Jacobs et al., 2011; Glaeser et al., 2016).

The beneficial effects elicited by RrF4 in inoculated plants suggest that R. radiobacter contributes to a successful beneficial symbiosis. Unambiguous elucidation of such a role has been hampered by the fact that all attempts to cure P. indica from endobacteria have failed (Glaeser et al., 2016). In the present study, we assessed factors affecting the amount of endofungal bacteria under different growth conditions of fungal cultures. We show that reduced numbers of P. indica-associated bacteria lead to changes in the fungal morphology, affects vegetative fungal reproduction, and impedes the biological activity of the fungus. Furthermore, the number of the bacteria especially increases at early stages of root colonization by P. indica, suggesting that R. radiobacter supports the establishment of the tripartite Sebacinalean symbiosis.

### MATERIALS AND METHODS

### Plant, Fungal, and Bacterial Materials

Barley (Hordeum vulgare) cultivar Golden Promise and Arabidopsis thaliana ecotype Columbia-0 (Col-0, N1092) were used. Seeds were surface sterilized and germinated on sterilized

filter paper in jars (barley) or germinated on solid <sup>1</sup>/<sup>2</sup> Murashige-Skoog (MS) medium supplied with sucrose and solidified with 0.4% gelrite (Arabidopsis).

Piriformospora indica DSM 11827 was obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), Braunschweig, Germany. This isolate origins from a sample collected in the Indian Thar desert in 1997 (Verma et al., 1998). The fungus was propagated on modified complete medium (CM, 20 g/L glucose, 2 g/L peptone, 1 g/L yeast extract, 1g/L casamino acids, 1 ml/L 1000× microelements, and 50 ml/L 20 × salt solution) at room temperature (Pham et al., 2004).

R. radiobacter F4 is a subculture of strain PABac-DSM isolated from P. indica DSM 11827 (Sharma et al., 2008), cultured on YEP medium (5 g/L beef extract, 1 g/L yeast extract, 5 g/L casein hydrolysate, 5 g/L sucrose, and 0.49 g/L MgSO4·7H2O) at room temperature.

### Plant Inoculation

Chlamydospores were collected from 3-week-old P. indica cultures, washed and suspended in 0.002% tween-20 with 500,000 chlamydospores per mL. Three-day-old barley seedlings or 7-day-old Arabidopsis seedlings were dip-inoculated in the chlamydospore solution for 1.5 h. Seedlings dipped into tween-20 were used as control. Inoculated barley seedlings were cultured on <sup>1</sup>/<sup>2</sup> MS or in pots containing 3:1 mixture of expanded clay (Seramis <sup>R</sup> , Masterfoods) and Oil Dri <sup>R</sup> (Damolin), fertilized with an aqueous solution of Wuxal Super 8/8/6 (1:1000 v:v; Haug, Düsseldorf, Germany) every week, in a climate chamber under a 16 h photoperiod and 22/18◦C day/night (60% rel. humidity, and a photon flux density of 240 µmol m−<sup>2</sup> s −1 ). Inoculated Arabidopsis seedlings were grown vertically on <sup>1</sup>/<sup>2</sup> MS in squared (100 mm) petri dish in climate chamber under short day conditions (8/16 h light/dark, 22/18◦C, 60% rel. humidity, and a photon flux density of 183 µmol m−<sup>2</sup> s −1 ). Roots were harvested at 7 and 14 dpi.

### Re-isolation of P. indica

Three-day-old barley seedlings were dip-inoculated with chlamydospores of P. indica and cultured on <sup>1</sup>/<sup>2</sup> MS medium under sterile condition. Roots were collected after 2 weeks, washed in 70% (v/v) ethanol for 1 min, followed by sodium hypochlorite (3% active chlorine) for 1 min, and cultured on CM agar medium.

### Co-cultivation of P. indica

Chlamydospores were collected from 3-week-old axenic P. indica cultures and cultured for 3 days in liquid CM medium. For the cocultivation with GFP-RrF4 (Glaeser et al., 2016), overnight grown GFP-tagged RrF4 was collected, re-suspended and added to the P. indica culture with final bacterial concentrations of OD<sup>600</sup> 0.1, 0.01, and 0.001. For the co-cultivation with plant tissue, P. indica cultures were supplemented with root extract and root pieces, respectively, from 7-day-old barley seedlings. Briefly, root extract was collected by grinding 500 mg roots with mortar and pestle in liquid CM medium, followed by passing through a 0.2 µm filter. Root pieces (0.5 cm) were cut from roots with a scissors. The mycelium from each culture was harvested after 7 days to assess the amount of R. radiobacter bacteria.

### Production of Protoplasts from P. indica

Chlamydospores were collected from 3-week-old P. indica, cultured in liquid CM medium with 130 rpm at 28◦C for 7 days, harvested by filtering with miracloth and finally washed with 0.9% NaCl solution. The mycelium was crushed with a blender and cultured again for re-generation. After 3 days, the mycelium was filtered, washed, and dissolved in 2% Trichoderma harzianum lysing enzyme (Sigma–Aldrich, USA) in SMC buffer (1.33 M sorbitol, 50 mM CaCl2, 20 mM MES buffer pH 5.8). After incubating at 37◦C for 1 h, 10 mL ice cold STC buffer (1.33 M sorbitol, 50 mM CaCl2, 10 mM TrisHCl pH 7.5) was added to stop the enzyme reaction. Protoplasts were collected by centrifugation at 4,000 rpm at 4◦C for 10 min, dissolved in 100 µL ice cold STC buffer after washing three times, and then plated on CM agar containing 300 µg/mL spectinomycin and 300 µg/mL ciprofloxacin. After 5 days, single colonies from single protoplasts were picked and transferred to fresh CM agar medium plates containing both antibiotics. After culturing for 3 weeks, DNA was extracted from each fungal sample and used for detection of endobacteria with real-time PCR (qPCR) using specific ITS primers. Chlamydospores were collected from these plates, and used for the second round of propagation. P. indica cultures showing no endobacteria in the third round were treated as partially cured P. indica cultures (pcPiri).

### Quantification of Endofungal Bacteria by Real-time PCR (qPCR) and DGGE

DNA was extracted with the NucleoSpin <sup>R</sup> Soil DNA extraction Kit (Macherey-Nagel, Germany) with lysis buffer SL1 according to manufactures' instructions. DNA concentrations were measured using a NanoDrop ND-1000 (Peqlab Biotechnology, Erlangen, Germany) and adjusted to 25 ng/µL for qPCR analysis. QPCR reactions were performed as described in detail by Glaeser et al. (2016) using a Sybr Green I based detection method. The relative amount of endobacteria was thereby quantified with RrF4 specific 16S rRNA-23S rRNA gene internal transcribed spacer (ITS) targeting primers (ITS\_Rhf and ITS\_Rhr; Sharma et al., 2008) (**Table 1**). The relative amount of P. indica was quantified with P. indica translation elongation factor EF-1α Tef targeting


primers (Bütehorn et al., 2000) (**Table 1**). Denaturing gradient gel electrophoresis (DGGE) of 16S rRNA gene fragments amplified with universal bacterial 16S rRNA gene targeting primers was performed as described by Glaeser et al. (2010).

### Fluorescence In Situ Hybridization (FISH) and Staining of Mycelia

Fluorescence In situ Hybridization (FISH) was used to detect the endobacterium in fungal materials. Three-week-old P. indica and overnight cultured RrF4 were fixed in 50% ethanol for 3–4 h at 4 ◦C, pelleted by centrifugation 5 min with 6,000 rpm at 4◦C, and suspended in 1:1 mixture of 1× PBS and 99.9% ethanol after three times washing with PBS. Samples were pipetted on the six recesses-microscope slide (coated with 0.1% gelatin and 0.01% chromium potassium sulfate), dried at 46◦C and dehydrated in an increasing ethanol series 50, 80, and 96% (v/v) three min for each. Nine µL hybridization buffer (5 M NaCl, 1 M Tris-HCl, 10% SDS) mixed with 1 µL probe (50 ng/µL) were added to each sample and incubated at 46◦C for 1.5 h for hybridization. The hybridization buffer and excess probe were washed with washing buffer (0.5 M EDTA, 1 M Tris-HCl, 10% SDS, 5 M NaCl). Subsequently, the slide was incubated in washing buffer at 48◦C for 15 min, and dried at room temperature for microscopic analysis. For staining of fungal DNA, samples were incubated with 10 µL 4',6-diamidine-2' phenylindole dihydrochloride (DAPI, Sigma) for 10 min and removed by rinsing with distilled water. The air-dried samples were mounted in AF1 anti-fading reagent (Citifluor Ltd., London, UK), and observed with epi-fluorescence microscope (Leica DM 5000B, Germany). The oligonucleotide probes used in this research were EUB-338-mix, including EUB-338, EUB-338-II, and EUB-338-III for bacteria (Daims et al., 1999), and Rh-1247 for Rhizobium (Ludwig et al., 1998) (**Table 1**). The probes were labeled with fluorescein isothiocyanate (FITC, Sigma). The excitation and emission wavelengths were 488 and 530 nm for FITC-labeled EUB-338, and 358 and 461 nm for DAPI staining.

Fluorescent wheat germ agglutinin (WGA) staining was used to detect fungi in barley roots. Root samples were washed in distilled water three times and fixed in fixation solution (chloroform:ethanol:trichloroacetic acid [20%:80%:0.15%]) for 24 h. Subsequently, the materials were treated with 10% KOH for 30 s and washed in 1× PBS buffer (pH 7.4) three times for 5 min. Thereafter, root material was stained in 10 µg/mL WGA solution (Alexa Fluor 488 [WGAAF488], Molecular ProbesTM, ThermoFisher Scientific, Germany) containing 0.02% surfactant Silwet L-77. Vacuum infiltration was applied during the root staining with 25 mm Hg for 1 min. After washing with 1× PBS buffer, root samples were mounted on glass slides for epifluorescence microscopy. WGAAF<sup>488</sup> was excited with 488-nm and detected at 505–540 nm.

### Biological Activity of P. indica

Barley seedlings were dip-inoculated with chlamydospores of P. indica and cultured in pots containing 3:1 expanded clay and Oil Dri <sup>R</sup> (fertilized with Wuxal 8/8/6) for the growth promotion assay. Shoot and root fresh weight (FW) was measured after 3 weeks. Seedlings cultured in pots containing soil were used for the pathogen assay. Detached leaf-segment assays were performed on third leaves of 3-week-old plants with powdery mildew. Leaf segments were plated on water agar (1.5% agar) containing 5% benzimidazole, and inoculated with 15 conidia mm−<sup>2</sup> of Blumeria graminis f. sp. hordei (Bgh race A6) for 10 min. Pustules were counted with a binocular microscope after 6 days.

## RESULTS

### Detection of Endofungal R. radiobacter

Piriformospora indica-associated endofungal R. radiobacter was originally discovered in mycelia by FISH (Sharma et al., 2008). Corroborating this prior result, we detected low numbers of endobacteria in both mycelia and chlamydospores of steadily grown axenic P. indica cultures (**Figures 1A,B**). To exclude the possibility that the hyphal wall constituted a barrier for FISH probes and thus conditioned the low number of fluorescent signals, the FISH method was also applied on crushed hyphae (**Figure 1C**) and fungal protoplasts, which confirmed the low abundance of endobacteria. Analysis with a Rhizobium specific probe confirmed the results (Supplementary Figure S1). Interestingly, while the endobacterium detected in the mycelium was coccoid-shaped and much smaller in size, isolated RrF4 cells detected by FISH showed rod-shaped cells with a mean size of 1.2–2.0 µm in length and 0.7–0.9 µm in width (**Figure 1D**). Also, consistent with the previous study (Sharma et al., 2008), denaturing gradient gel electrophoresis (DGGE) of 16S rRNA gene fragments amplified with universal bacterial 16S rRNA gene targeting primers detected only one single DNA band in fungal DNA extracts, which had the same motility in the DGGE gel as the 16S rRNA gene product amplified from the pure culture of RrF4 (Supplementary Figure S2).

### The Amount of Endofungal Bacteria Varies with the Type of Fungal Culture

We addressed the question whether the number of endofungal bacteria varies in P. indica when grown under different conditions. Therefore, the amount of R. radiobacter was quantified in the fungus growing in liquid culture vs. fungus colonizing plant roots. The absolute number of bacteria was calculated with the standard curve for amplification of RrF4's ITS, while the amount of P. indica was calculated using a standard curve based on the amplification of the fungal Tef gene (Basiewicz et al., 2012). As indicated by melt curve analysis, only primer dimers could be detected for a standard containing 100 ITS targets per PCR reaction (Supplementary Figure S3). Based on the information that RrF4 has three rrn operons (Glaeser et al., 2016) the detection limit of R. radiobacter by quantitative RT-PCR is approximately 33 bacteria cells per PCR reaction. Thus, in order to detect endobacterium efficiently, approximately 500 mg mycelium was necessary for DNA extraction. The relative number of bacteria based on the genome ratio between endobacteria and P. indica is shown in **Figure 2**. We found increased numbers of R. radiobacter in fungal mycelia colonizing

barley roots as compared with P. indica from liquid cultures (lcPiri). Moreover, the number of bacteria detected in root samples was higher at 7 dpi than at 14 dpi as shown in three independent biological replicates (**Figure 2A**).

Next, we addressed the question whether the number of R. radiobacter in mycelia grown in liquid culture was affected by plant derived compounds. For this purpose, a liquid culture started with chlamydospores was supplemented with root segments or extracts of fresh barley roots, respectively, and harvested for R. radiobacter quantification after 1 week. qPCR analysis revealed increased relative amounts of bacteria in the supplemented P. indica cultures compared with the non-supplemented fungal culture (**Figure 2B**).

### Quantification of Endobacteria in P. indica Isolated from Plant Roots

Since the relative amount of bacterial cells increased in P. indica during colonization of barley roots, we addressed the question whether the number of bacteria remained high in axenic cultures of P. indica when the fungal inoculum was re-isolated from roots. To this end, P. indica-colonized barley roots were surface sterilized and subsequently incubated on solid or liquid CM medium. Once fungal hyphae that resided inside the root tissue broke through the root surface, they formed colonies around the root pieces within a week (**Figures 3A,B**). Fungus from this culture was recorded as re-isolated P. indica (riPiri-1); it was further transferred and sub-cultured as riPiri-2. As shown in **Figure 3C**, the relative amount of endobacteria increased significantly in riPiri compared to the lcPiri, and the amount in riPiri-1 also was significantly higher than in riPiri-2. These results together show that the amount of R. radiobacter increases in the tripartite Sebacinalean symbiosis, whereas it decreases during liquid or axenic culturing of the fungal host in the absence of a plant root. A release of bacterial cells from the P. indica cultures was not observed, neither on agar plates nor in liquid medium.

### Antibiotics Reduce the Number of Bacteria in P. indica

While biological activities of both, P. indica and the isolated strain RrF4, have been demonstrated many times, it is not known whether R. radiobacter is required for a successful Sebacinalean symbiosis (Sharma et al., 2008; Glaeser et al., 2016). To further address this question, we used antibiotics to cure P. indica of endofungal bacteria. To this end, fungal protoplasts were treated with antibiotics. After three rounds of protoplastation in the presence of a mixture of 300 µg/ml spectinomycin and 300 µg/ml ciprofloxacin, growth of P. indica cultures was delayed compared with the culture from protoplasts that were not treated with antibiotics (**Figures 4A,B**). The diameter of the

colonies that derived from single antibiotic-treated protoplasts had a size of 3 cm, while the diameter of control colonies was approximately 6 cm (**Figures 4C,D**). No apparent alterations in morphology of fungal hyphae were observed, while formation of vegetative chlamydospores was clearly reduced in P. indica cultures under antibiotics selection (**Figures 4E,F**). Moreover, DAPI staining revealed differences in the shape of fungal nuclei: hyphae of antibiotics-treated P. indica culture had rodshaped nuclei (**Figure 4G**), while spherical-shaped nuclei were observed in non-treated mycelia (**Figure 4H**). Significantly, no endobacteria could be detected by qPCR in antibiotics-treated P. indica cultures. However, prolonged fungal cultivation in the absence of antibiotics resulted in the recovery of a low number of endobacterial cells.

### Re-introduction of RrF4 Cells into P. indica Failed

GFP-tagged RrF4 cells were co-cultured with P. indica to address the question whether isolated RrF4 can invade fungal hyphae. GFP-tagged RrF4 cells were added to a three-day-old liquid culture of P. indica started from chlamydospores. As shown in the **Figure 5**, many GFP-tagged RrF4 cells stuck around the hyphae (**Figure 5A**) and the surface of chlamydospores (**Figure 5B**), while they were not detected inside these structures. As the fungal wall may constitute a barrier for bacteria, protoplasts of P. indica were also incubated with GFP-tagged RrF4, but this strategy could neither improve the uptake of bacteria (data not shown).

### Biological Activity Exhibited by riPiri, pcPiri, and lcPiri

To answer the question whether P. indica cultures that contain different amount of endobacteria exhibit comparable biological activities, biomass and systemic resistance of barley seedlings were measured upon inoculation with riPiri, pcPiri, and lcPiri, respectively. By 7 dpi, roots were checked for fungal colonization

(C) Relative amount of endofungal bacteria in cultures of riPiri-1 and sub-culture riPiri-2 was quantified with ITS targets of RrF4 related to the Tef gene of P. indica. Mean values and standard errors based on three independent biological replicates. Different letters indicate statistically significant differences tested by one-way analysis of variance performed with the Tukey test (P < 0.05).

FIGURE 4 | Piriformospora indica cultures grown from single protoplasts in the presence of antibiotics. (A) Fungal colonies regenerated from single protoplasts of lcPiri cultures grown for 5 days on CM medium containing 300 µg/mL spectinomycin and 300 µg/mL ciprofloxacin. (B) Colonies grown on CM medium without antibiotics. (C) Single colony picked up from A and propagated on medium with the same antibiotics combination for 10 days. (D) Single colony picked up from B and propagated without antibiotics for 10 days. (E): Microscopy of P. indica mycelium from image C (cultured in the presence of antibiotics). (F) Microscopy of the P. indica culture from image D (cultured in the absence of antibiotics). Chlamydospores are clearly visible in this culture. (G,H) Nuclei were stained with DAPI and show blue fluorescence under the fluorescence microscope. Mycelium from an antibiotics-treated (G) and a control culture (H).

FIGURE 5 | GFP-tagged RrF4 sticking around P. indica. Chlamydospores of P. indica were germinated in liquid CM medium for 3 days, before a GFP-RrF4 suspension was added to the culture. (A) GFP-tagged RrF4 stacking around the hyphae. The square box showed GFP-tagged RrF4 around a hyphal tip. (B) GFP-tagged RrF4 at the surface of chlamydospores. Bars indicated 10 µm.

and the amount of endobacteria. The Tef gene of P. indica was used for the quantification of the fungus, while the specific ITS primer system was used for the quantification of R. radiobacter bacteria. We found that the relative amount of endobacteria in root samples colonized by riPiri was significantly higher compared with roots colonized by lcPiri and pcPiri (**Figure 6A**). Yet, that bacteria were detectable in root samples colonized by pcPiri shows that single protoplast cultivation and antibiotics treatment did not completely cure P. indica from R. radiobacter. riPiri also showed the highest colonization density on barley roots as compared with pcPiri and lcPiri (**Figure 6B**). Consistent with this finding, high amounts of pear-shaped chlamydospores and mycelium were observed in root inoculated with riPiri when roots were stained with chitin-specific WGA-Alexa Fluor 488 to visualize mycelia (**Figure 6C**), while less colonization was seen in roots inoculated with pcPiri (**Figure 6D**). Importantly, less chlamydospores were seen in mycelia of lcPiri and most significantly in pcPiri, suggesting that vegetative reproduction is negative affected when the number of bacteria is limited.

Next, we assessed fungal growth promotion activity. Threeweek-old barley plants inoculated with riPiri, lcPiri, and pcPiri were harvested for biomass analysis. Compared with non-inoculated plants, plants treated with either fungal culture showed some growth promotion (**Figure 7A**). Accordingly, the shoot FW of colonized plants was always higher compared to non-colonized control plants, though it was increased significantly only in seedlings treated with riPiri (17.7%) and lcPiri (15.9%) (**Figure 7B**). Moreover, root FW slightly increased in all colonized plants though the effect was not significant (**Figure 7C**). We also assessed the resistance-inducing activity of P. indica cultures against Bgh in barley. Barley seedlings were dip-inoculated with crushed P. indica cultures. Three weeks later, third leaves were harvested and inoculated with Bgh conidia in a detached leaf assay to assess powdery mildew resistance. The pustules on leaves from plants inoculated with riPiri and lcPiri were equally reduced by 21% compared to control plants, while there was no significant difference for plants infected with pcPiri (2.3%) compared with untreated controls (**Figure 7D**).

FIGURE 6 | Colonization of barley roots by riPiri, lcPiri, and pcPiri. Three-day-old seedlings were dip-inoculated with respective mycelia. Seedlings were grown in soil in a growth chamber and harvested after 1 week for quantification and WGA-staining. (A) Relative amount of R. radiobacter bacteria based on the genome ratio of RrF4 vs. P. indica on barley roots. (B) Quantification of the amount of P. indica mycelium on barley roots. (C) Pear-shaped chlamydospores and mycelium in a root inoculated with riPiri. (D) Mycelium in a root inoculated with pcPiri. Mean values and standard errors of three independent biological replicates are given. Different letters on the top of the bars indicate statistically significant differences tested by one-way analysis of variance performed with the Tukey test (p < 0.05). Bars indicated 10 µm.

We also assessed the biological activity of the P. indica cultures in Arabidopsis (**Figure 8**). Consistent with the above data, treatment of Arabidopsis seedlings with riPiri resulted in a statistically significant increase in shoot and root FWs. Moreover,

riPiri colonized Arabidopsis roots much stronger than lcPiri and pcPiri, which is consistent with the hypothesis that bacteria support the fitness of the endophytic fungus.

### DISCUSSION

### R. radiobacter in Piriformospora indica

In the present work, we show a positive correlation between the number of endofungal R. radiobacter that are associated with the fungal mutualist P. indica and the biological activity the mutualist exerts on plants. Furthermore, there is an indication for P. indica's requirement for endobacteria for full vegetative propagation via chlamydospores. Endobacteria were detected by independent methods including FISH, qPCR, and sub-culturing of the isolated strain RrF4. In continuous fungal cultures the endobacteria reach very low numbers. This low amount and limited detection have been the bottlenecks for the analysis of the role endobacterium in the tripartite Sebacinalean symbiosis. The finding that the amount of endofungal bacteria increased when P. indica colonized roots opened new possibility to study endofungal bacteria's role in the tripartite symbiosis. The higher amount of bacteria in riPiri that was isolated directly from barley roots, and the steady decrease during axenic subculture suggest that a plant host promotes the propagation of the endofungal bacterium in its host fungus. Similar observation is known from the fungus Gigaspora margarita, in which the expression of the ftsZ gene of the endocellular bacterium Candidatus Glomeribacter gigasporarum was up-regulated when the fungal host colonized on plant (Anca et al., 2009).

### Endobacteria Influence the Fitness of P. indica

In order to generate P. indica mycelia with substantially lowered numbers of bacterial cells, we combined strategies recently published to remove endobacteria from its fungal hosts. A bacteria-free fungus Rhizopus microspores was obtained through antibiotics treatment (Partida-Martinez and Hertweck, 2005), while repeated passages through single-spore inoculation of plants was used to dilute the initial Candidatus Glomeribacter gigasporarum (CaGg) population in the spores of the AM fungus Gigaspora margarita (Lumini et al., 2007). We show here that a combination of single fungal protoplast propagation and antibiotics treatment at least transiently reduced the

number of active bacterial cells in P. indica. P. indica depleted in R. radiobacter showed delayed germination from chlamydospores, reduced growth in axenic cultures and less sporulation. Although we cannot exclude that the effect comes from antibiotics treatment, those morphological changes could be induced by the reduced number of endobacteria. Because there was no morphology change at the beginning of antibiotics treatment, those changes only happen after several rounds of selection. That endofungal bacteria support the fitness of its fungal host has been demonstrated with pathogenic and mutualistic fungi. Cured Rhizopus microspores show no sporulation (Partida-Martinez and Hertweck, 2005; Partida-Martinez et al., 2007), cured fungus Gigaspora margarita shows limited changes in spore morphology, while the symbiosis with this endobacterium CaGg increased the environmental fitness and bioenergetics potential of the AM fungal host (Lumini et al., 2007; Salvioli et al., 2016). P. indica cultures treated with antibiotics (pcPiri) initially seemed to be completely cured from bacteria. However, when inoculated to barley seedlings, bacteria were again detected in plant root samples. However, the relative amount of bacteria was significantly reduced compared with root samples colonized by steadily grown P. indica. We concluded that antibiotics-treated P. indica culture was only partially cured from endobacteria, and the fitness and asexual reproduction of the fungal host is compromised with reduced endobacteria.

### Endobacteria Exhibit Beneficial Activity on Plant Hosts

The colonization and induced biological activity were compared among three types of P. indica cultures. The determination of the genome ratio of R. radiobacter vs. P. indica in plant roots inoculated with pcPiri, lcPiri, and riPiri confirmed the differences among these three P. indica cultures with respect to the presence of endobacteria. The riPiri culture contained the highest number of endobacteria, while pcPiri contained the lowest number. The riPiri culture showed most efficient colonization of barley and Arabidopsis roots with high amounts of mycelium and chlamydospores. These results support the hypothesis that the amount of endobacteria had an effect on both fungal colonization and sporulation. Compared with control plants, plants either treated with riPiri or lcPiri, respectively, had significantly increased shoot weights, which is consistent with earlier findings that P. indica promotes plant growth (Waller et al., 2005; Sharma et al., 2008; Qiang et al., 2012). Although the increase in bacterial cell numbers in riPiri was significant, numbers were still low as shown by qPCR and confirmed by low

numbers of cells detected by FISH analysis (data not shown). Even a small but significant increase in R. radiobacter cell numbers has an effect on the biological activity of P. indica. The mechanism behind the activation of the fungal activity is not resolved. Genome analysis of the endobacteria Candidatus Glomeribacter gigasporarum in Gigaspora margarita indicated that the endobacterium could provide vitamin B12 to the fungal host and plant host (Ghignone et al., 2012). The same genes for the synthesis pathway of vitamin B12 were also present in the genome of RrF4 (Glaeser et al., 2016). This indicates that RrF4 may also support P. indica with vitamine B12. Salvioli et al. (2016) showed that the endofungal bacteria can increased the fitness of root colonizing fungi and thereby may affects their colonization efficiency and biological activity. Taken together, these data are consistent with the hypothesis that the endobacterium mediates growth promotion of plants and thus contributes to the biological activity induced by its fungal host P. indica. Whether the effects of the endobacteria on the plant growth are direct or indirect (by activating P. indica) still needs further investigations.

### Cultured RrF4 Cells Do Not Invade the Fungus

The co-culture of GFP-tagged RrF4 and P. indica showed GFP-RrF4 cells sticking around the hyphae and chlamydospores instead of entry into hyphae. Type 2 secretion system (T2SS) in endosymbiont Burkholderia was previously shown to be central for the active invasion into living fungus Rhizopus, because T2SS releases chitinase, chitosanase and chitin-binding proteins to soften the cell wall of fungus and allow the bacterial entry into fungal hyphae (Moebius et al., 2014). Based on this information, we checked the genome sequence of RrF4. Not surprisingly, neither the whole gene cluster that is responsible for the encoding of T2SS components nor the genes synthesizing chitinase and chitosanase is existing in the genome of RrF4 (Glaeser et al., 2016). Instead RrF4 has a type IV secretion system, which is a remarkable characteristic in Agrobacterium and mediates the transfer of plasmid DNA fragment into plant genome (Sharma et al., 2008; Fronzes et al., 2009; Glaeser et al., 2016). The missing responsible genes and T2SS in RrF4 could be the reason for the unsuccessful invasion of RrF4 into P. indica. However, protoplastation of hyphae and incubation with the RrF4 cells also did not result in bacterial uptake indicating that additional factors may be required for bacterial transfer. Our co-culture experiment showed that RrF4 sticking around the fungus including the tip of hyphae and spores, which are the active part from P. indica. Those hyphae tip and germinating spores can be potential entry sites for bacteria.

### Endobacteria Rhizobium in a VBNC State

Some Gram-negative bacteria have extraordinary ability to survive under harsh environment. They enter into a specific growth stage of low metabolic activity, the viable but nonculturable (VBNC) state, which is similar to a stationary growth phase in a lab culture. Bacteria, including Agrobacterium and Rhizobium species (Alexander et al., 1999), enter the VBNC state if the environment is not sufficient enough to keep steady growth (Llorens et al., 2010). In the VBNC state, bacterial cells perform morphological and physiological adaptations to the changed conditions. The bacterial cells become smaller as the result of reductive division and dwarfing, and more resistant against different kinds of harm because of the formation of cell envelopes (Nyström, 2004; Llorens et al., 2010). Given that bacteria decreased in abundance and changed to the small spherical shape (compared to the rod-shaped cells in pure culture) as shown by FISH and Sybr Green I straining, we assume that R. radiobacter switched into the VBNC state in axenic longterm P. indica lab-cultures. In this VBNC state, bacterial cells are more resistant to antibiotics, which could be one reason for the insufficient antibiotics treatment to cure endobacteria from their fungal host. Because the VBNC cells are less metabolically active, the concentration of ribosomes decreased. This may have resulted in a reduced detection efficiency by FISH. Our data are consistent with the hypothesis that bacteria resuscitate from the VBNC to become metabolically active and growing cells in the presence of the plant or plant-derived extracts which may trigger the activity of the bacterium by, e.g., supplying it with enough nutrition and better propagation conditions. This could explain the increased amount of R. radiobacter when P. indica colonized plant roots.

## CONCLUSION

The data presented here support the hypothesis that the endobacterium of P. indica regulates the fitness and sporulation of P. indica, influences the colonization on plant roots, and further contributes to the biological activity induced by P. indica. There is robust association between endobacteria and P. indica, since P. indica cannot be fully cured from the endobacterium, while free RrF4 did not re-enter mycelium. Our studies suggest R. radiobacter contributes to the tripartite Sebacinalean symbiosis but further studies are needed to fully elucidate the mechanism exerted by the endofungal bacteria during the mutualistic tripartite Sebacinalean interactions.

## AUTHOR CONTRIBUTIONS

HG, SG, JI, and K-HK designed the research. HG, IA, and HH performed the experiments. HG, SG, JI, and K-HK analyzed data. K-HK and PK got the funding. HG, SG, and K-HK wrote the manuscript.

## FUNDING

This work was supported by the German Science Foundation (DFG) to K-HK (KO 1208/24-1) and PK (DFG KA 875/8-1).

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb.2017. 00629/full#supplementary-material

### REFERENCES

fmicb-08-00629 April 11, 2017 Time: 16:11 # 12



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Guo, Glaeser, Alabid, Imani, Haghighi, Kämpfer and Kogel. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Halotolerant Rhizobacteria Promote Growth and Enhance Salinity Tolerance in Peanut

Sandeep Sharma1, 2 \*, Jayant Kulkarni 1, 2 and Bhavanath Jha1, 2 \*

<sup>1</sup> Central Salt and Marine Chemicals Research Institute, CSIR, Bhavnagar, India, <sup>2</sup> Academy of Scientific and Innovative Research, CSIR, New Delhi, India

Use of Plant growth promoting rhizobacteria (PGPR) is a promising strategy to improve the crop production under optimal or sub-optimal conditions. In the present study, five diazotrophic salt tolerant bacteria were isolated from the roots of a halophyte, Arthrocnemum indicum. The isolates were partially characterized in vitro for plant growth promoting traits and evaluated for their potential to promote growth and enhanced salt tolerance in peanut. The 16S rRNA gene sequence homology indicated that these bacterial isolates belong to the genera, Klebsiella, Pseudomonas, Agrobacterium, and Ochrobactrum. All isolates were nifH positive and able to produce indole -3-acetic acid (ranging from 11.5 to 19.1µg ml−<sup>1</sup> ). The isolates showed phosphate solubilisation activity (ranging from 1.4 to 55.6µg phosphate /mg dry weight), 1-aminocyclopropane-1-carboxylate deaminase activity (0.1 to 0.31 µmol α-kB/µg protein/h) and were capable of reducing acetylene in acetylene reduction assay (ranging from 0.95 to 1.8µmol C2H<sup>4</sup> mg protein/h). These isolates successfully colonized the peanut roots and were capable of promoting the growth under non-stress condition. A significant increase in total nitrogen (N) content (up to 76%) was observed over the non-inoculated control. All isolates showed tolerance to NaCl ranging from 4 to 8% in nutrient broth medium. Under salt stress, inoculated peanut seedlings maintained ion homeostasis, accumulated less reactive oxygen species (ROS) and showed enhanced growth compared to non-inoculated seedlings. Overall, the present study has characterized several potential bacterial strains that showed an enhanced growth promotion effect on peanut under control as well as saline conditions. The results show the possibility to reduce chemical fertilizer inputs and may promote the use of bio-inoculants.

Keywords: halotolerance, plant growth promoting rhizobacteria (PGPR), salinity stress, ion homeostasis, reactive oxygen species, Arthrocnemum, IAA production, acetylene reduction

### INTRODUCTION

Peanut (Arachis hypogaea L) is an important cash crop of the leguminous family, grown in most of the arid, and semi-arid regions. Worldwide, peanut is cultivated on 26.4 million ha with a total annual production of 39.46 million metric tons (FAO, 2012; Sarkar et al., 2014). The production of peanut has shown to be repeatedly hampered by various abiotic stresses, such as high salt and drought, which ultimately, lead to severe loss in yield. Peanut being moderately salt tolerant,

#### Edited by:

Anton Hartmann, Helmholtz Zentrum München, Germany

#### Reviewed by:

Raffaella Balestrini, National Research Council, Italy Stijn Spaepen, Max Planck Institute for Plant Breeding Research, Germany

#### \*Correspondence:

Sandeep Sharma sksbhu@gmail.com Bhavanath Jha bjha@csmcri.org

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 19 April 2016 Accepted: 26 September 2016 Published: 13 October 2016

#### Citation:

Sharma S, Kulkarni J and Jha B (2016) Halotolerant Rhizobacteria Promote Growth and Enhance Salinity Tolerance in Peanut. Front. Microbiol. 7:1600. doi: 10.3389/fmicb.2016.01600 suffers massively by salinity stress due to its growth habitat in arid or semi-arid regions (Tanji and Kielen, 2002; Sun et al., 2013; Tiwari et al., 2015). Salinity stress has detrimental effects on almost every aspect of peanut growth and development including seed germination, early seedling establishment, photosynthesis, pod formation, total biomass, and finally, on yield production (Salwa et al., 2010; Qin et al., 2011; El-Akhal et al., 2013). It is, therefore, necessary to improve the salinity tolerance of peanut to minimize the yield loss.

Several approaches, including traditional breeding, and genetic engineering have been used to improve the salinity tolerance of peanut. However, such interventions have low success rate, mainly due to complexity of salinity tolerance and narrow genetic variability among germplasm accessions (Krishna et al., 2015). Use of bacterial inoculation, in particular, plant growth promoting rhizobacteria (PGPR), is effective and ecofriendly to improve plant stress tolerance. Several reports have shown that PGPR effectively improve growth of a wide range of agricultural crops under environmental stress conditions (Bacilio et al., 2004; Mayak et al., 2004b; Yuwono et al., 2005; Jha et al., 2009; Nabti et al., 2010; Ji et al., 2014; Islam S. et al., 2015; Majeed et al., 2015; Rolli et al., 2015; Timmusk et al., 2015; Zahid et al., 2015). In addition, the ability of PGPR to serve as bio-fertilizer or phyto-stimulator helps in maintaining the soil fertility, thereby providing a promising alternative to chemical fertilizers and pesticides for the sustainable agriculture (Majeed et al., 2015).

PGPR are free-living soil microbes that colonize roots and stimulate plant growth (Baldani et al., 1997; Schmid et al., 2009). A number of mechanisms are involved in plant growth promotion by PGPR. These include acquisition of nutrients, fixation of atmospheric nitrogen (N), phosphorus (P) solubilization, siderophore production, hydrocyanic (HCN) production, modulation of plant hormone and antagonistic action against biotic pathogens. The N-fixation, P- availability and the hormonal response have direct involvement in plant growth promotion; however, other mechanisms indirectly support the plant growth (Gontia et al., 2011; Bhattacharyya and Jha, 2012; Glick, 2012; Estrada et al., 2013; Vacheron et al., 2013; Abd El Daim et al., 2014).

The PGPR use several mechanisms to protect the plant growth under various abiotic stresses. Rhizobacteria activate plant antioxidant defense machinery by upregulating the activity of key enzymes, such as superoxide dismutase (SOD), peroxidase and catalase that scavenges overproducing reactive oxygen species (ROS) and protect the plants from salt toxicity (Jha and Subramanian, 2014; Islam F. et al., 2015). Under salinity stress, PGPR-inoculated plants gain increased efficiency to uptake selective ions to maintain a higher K+/Na<sup>+</sup> ratio than noninoculated plants (Shukla et al., 2012; Islam F. et al., 2015). PGPR producing 1-aminocyclopropane-1-carboxylate (ACC) deaminase, an enzyme that converts plant ethylene precursor ACC to ammonia and α-ketobutyrate. This metabolic event decreases plant ethylene level which in turn resumes plant growth under abiotic stresses (Glick, 2014; Singh et al., 2015). PGPRinoculated plants have also been shown to have a change in root architecture. This may be due to increased indole -3-acetic acid (IAA) level that enables plant to uptake more nutrients under salinity stress condition (Vacheron et al., 2013; Goswami et al., 2014). A number of rhizobacteria emit stress-related volatile compounds that enhance plant biomass and survival under severe drought stress (Timmusk et al., 2014). Recently, Rolli et al. (2015) studied eight osmotolerant bacterial isolates that showed stress-dependent plant growth promoting activities and were capable in improving grapevine growth under drought stress. A mutant strain of Paenibacillus polymyxa, A26△sfp, that had inactivation of A26 Sfp-type 4′ -phosphopantetheinyl transferase enzyme (Sfp-type PPTase) showed greatly enhanced biofilm activity and produced two times higher plant survival and three times increased wheat biomass under drought stress (Timmusk et al., 2015). This mutant strain is also an efficient antagonistic agent against Fusarium spp. causing Fusarium head blight disease in cereals (Abd El Daim et al., 2014).

Halophytes are extremely salt tolerant plants which usually grow near a coastal area where no cultivation occurs. The rhizosphere of halophytic plants represent ideal source for isolation of various groups of salt tolerant rhizobacteria that could enhance the growth of different crops under salinity stress (Shukla et al., 2012; Bharti et al., 2013; Ramadoss et al., 2013; Goswami et al., 2014; Jha et al., 2015). Efforts have been made to isolate few halotolerant bacteria that confers salt tolerance to agricultural crops. Examples of such bacteria are Brachybacterium sp. (Jha et al., 2012; Shukla et al., 2012), B. licheniformis (Goswami et al., 2014), Exiguobacterium oxidotolerans (Bharti et al., 2013), Pseudomonas sp. (Rosenberg, 1983; Egamberdieva et al., 2015), and Hallobacillussp. (Ramadoss et al., 2013). However, scant information is available for potential halotolerant PGPR that promotes salinity tolerance to peanut.

Arthrocnemum indicum (Chenopodiaceae family) is a stem succulent perennial halophytic shrub, generally found in tropical salt marshes that are frequently inundated with seawater. It has high antioxidant and anti-radical activity (Boulaaba et al., 2013). Moreover, it is a potential source of anti-cancer molecules, used in the treatment of snakebites and scorpion stings thus, has medicinal significance (Boulaaba et al., 2013). The present study was undertaken to isolate bacterial strains from the roots of A. indicum, their characterization and efficacy to test their ability to promote growth and tolerate salinity of peanut. Our results indicate that the A. indicum roots have several bacteria that enhances growth and confer salt tolerance to peanut seedlings.

### MATERIALS AND METHODS

### Isolation of Bacterial Isolates

Bacterial isolation from the roots of Arthrocnemum indicum was performed according to Jha et al. (2012). Briefly, roots (0.5 g fresh weight) were washed thoroughly, homogenized with 0.5X PBS (9.5 ml), serially diluted and grown on nitrogen-free semisolid NFb medium containing up to 4% NaCl and malate as a sole carbon source at 30◦C. Bacteria were subcultured two more times. Finally, bacteria were streaked onto solid NFb medium with 20 mg l−<sup>1</sup> of yeast extract. Single, distinct colonies were analyzed for pellicle formation and purified further.

### Physiological and Biochemical Characterization

Biochemical tests for utilization of different carbon sources, such as sugars, organic acids, and amino acids were performed using BIOLOG identification system (BIOLOG MicrostationTM, Biolog Inc., Hayward, CA). GEN III MicroPlateTM analyses 94 phenotypic tests: 71 carbon source and 23 chemical sensitivity assays were inoculated according to the BIOLOG manufacturer's directions. Plates were covered and incubated at 30◦C for 24 h. Measurement of MicroPlate were taken with a BIOLOG microplate reader. The enzyme activities (Amylase, Gelatinase, pectinase, lipase, Catalase) were measured as described previously by Jha et al. (2012). Motility tests for PGPR were performed using Motility Test Medium (Himedia, Mumbai) (Camper et al., 1993). For ammonia production, isolates were cultured in peptone broth and incubated at 30◦C for 48–72 h. Following incubation, Nessler's reagent (Sigma, USA) was used to detect brown to yellow color formation as described previously by Jha et al. (2012).

### Molecular Characterization of Bacterial Isolates

Genomic DNA was isolated from bacterial strains using standard protocol (Sambrook et al., 1989). The universal primers 27F (5′ - AGAGTTTGATCMTGGCTCAG-3′ ) and 1492R (5′ - TACGGYTACCTTGTTACGACT-3′ ) were used to amplify 16S gene sequences by PCR (Lane, 1991). Amplified gene sequences were gel purified using QIAquick gel extraction kit (Qiagen, Germany) and sequenced. Sequence comparison was then performed with obtained 16S rRNA gene sequences against sequences in NCBI and Silva rRNA database (https://www.arbsilva.de/). For nifH screening, a 360-bp fragment was amplified using PolF and PolR primers (Poly et al., 2001) and the amplified DNA fragment was purified and sequenced. Phylogenetic analysis of 16S rRNA and nifH amino acid sequences were performed using MEGA version 6. The Neighbour- Joining method (Saitou and Nei, 1987) and bootstrap analysis (Felsenstein, 1985) were performed using 1000 bootstrap replications and evolutionary distance were computed with the Maximum Likelihood method (Tamura et al., 2004).

### Bioassays for Plant Growth Promoting Traits

### Biological Nitrogen Fixation (BNF)

BNF was measured by acetylene reduction/ethylene production assay (ARA) as described previously by Majeed et al. (2015). Briefly, pure bacterial cultures were inoculated in 20 ml airtight vials containing 5 ml semisolid nitrogen free NFb medium and grown for 48 h at 28◦C. Following pellicle formation, 10% (v/v) acetylene gas was injected into the vials, which were incubated for 16 h at 30◦C. Samples were then analyzed by gas chromatography (GC) (Shimadzu, Japan) using RTX5 column (Restek, USA) and a H2-flame ionization detector (FID). The peaks of acetylene and ethylene were also confirmed by GC-MS (Shimadzu, Japan). All the experiments were performed in triplicates.

### Indole-3-Acetic Acid (IAA)

Indole-3-Acetic Acid (IAA) production of bacteria was determined by colorimetric method as described by Patten and Glick (2002). 50µl of overnight grown cultures were transferred into 10 ml of nitrogen-free NFb medium supplemented the presence and absence of 0.05% L-tryptophan. Cultures were then incubated at 28◦C with continue shaking at 180 rpm for 48 h. The density of each culture was measured spectrophotometrically at 600 nm, and then the culture was centrifuged at 13000 rpm for 5 min. One ml aliquot of resulting supernatant was then mixed with 2 ml Salkowski's reagent and incubated in the dark at room temperature for 20 min. IAA production was confirmed by observing pink color formation, and the absorbance was measured at 530 nm. A known IAA standard was used to determine IAA concentrations. A. brasilense strain Sp7 was used as a positive control in the present study.

### Phosphate (P) Solubilisation

The bacterial cultures were inoculated into 10 ml Pikovskaya broth (Himedia, Mumbai) containing tri-calcium phosphates and incubated at 28◦C for 72 h. Following incubation, 1 ml of culture was withdrawn and the cell-free supernatant was mixed with 4 ml of Chen reagent and incubated at 37◦C for 1 h 30 min. The reaction product, phosphomolybdate, was then determined by measuring OD at 619 nm after adjustment to a final volume of 8 ml (Goldstein, 1986).

### Assays for Other Growth Promoting Traits

ACC deaminase activity was determined as described previously by Penrose and Glick (2003). The formation of orange-yellowish halos surrounding bacterial colonies on CAS agar plates after 48 h of incubation at 30◦C indicated siderophore production (Schwyn and Neilands, 1987). The zinc solubilisation assay was performed according to Fasim et al. (2002) and HCN production was carried out according to the method of Lorck (1948).

### Bacterial Growth Promotion Effects on Peanut

Peanut cv. GG 20 seeds were surface sterilized with 2% sodium hypochlorite and rinsed 4–5 times with sterilized double-distilled water. Seeds of uniform size were placed in sterile cotton (soaked in ½MS medium) in a tissue culture bottle and kept in the dark for 2 days before transfer to a growth chamber. The growth conditions were set at 26◦C and 16/8 h light/dark cycle (350µmol m−<sup>2</sup> s-1 light intensity). After a week, seedlings were transferred to hydroponics containing half MS medium (without sucrose and nitrogen source) with and without bacterial inoculum and growth parameters were measured after 21 days. For inoculation, single bacterial colony was inoculated in 5 ml of half DYGS medium (Kirchhof et al., 2001) and kept at 30◦C on shaker incubator (180 rpm) for 24 h. The bacterial culture was adjusted to OD 0.6 that correspond to approximately 10<sup>8</sup> cells/ml. Five ml of this bacterial culture was added to 250 ml of ½MS medium. The hydroponic medium was changed weekly. The shoot and root samples were dried separated and used for dry weight measurement.

In soil pot experiment, peanut seeds were pre-inoculated with bacterial inoculum (MBE02, MBE03, and positive strain) and kept in dark for 2 d and the seedlings were transferred to the pots (3 kg soil capacity, three seedlings/pot) filled with garden soil. Soil samples were autoclaved twice before the experiment. One ml of bacterial culture (OD 0.6) was placed at the base of the seedlings after 3 weeks of germination. Seedlings without inoculation served as control. The plants were harvested after 60 d for root and shoot biomass measurement. For field trial, seedlings were grown in 2 × 4 m plots with the spacing of 30 cm between rows and 15 cm between plants. The bacterial treatments were given in a similar way as described for greenhouse experiment. The treatments, including control, were replicated thrice in randomized complete block design. N and P were applied at a half rate of recommended dose. Standard agronomical practices were followed to maintain the plots. The plant biomass was measured after 40 days of germination.

For salt treatment, 12 d old peanut seedlings, grown in half MS medium were treated with 100 mM NaCl with or without bacterial inoculum and growth were measured after 18 d.

### Root Colonization

Root colonization assay was performed as described previously by Majeed et al. (2015). Root samples were collected from bacteria inoculated peanut seedlings at 10th and 20th day postinoculation. Root surface was thoroughly washed with running tap water to remove weakly bound cells. After washing, the plant roots were blotted dry, weighed and 1 g was homogenized with 10 ml autoclaved distilled water using sterile mortar and pestle. Serial dilutions (up to 10−<sup>7</sup> ) were then plated on ½DYGS medium plates and incubated at 30◦C for 24–48 h. Colony forming unit (CFU) per gram root was determined as described previously by Islam S. et al. (2015). Three replications were used for each treatment and the experiment was repeated three times.

### Ion Analysis and Nitrogen Measurement

For ion analysis, 0.5 gram of dried shoot and root was digested with 5 ml of perchloric acid and nitric acid solution (3:1). The solution was heated on a hot plate and diluted to 25 ml with autoclaved deionized water and filtered through a 0.22µm filter. Na+, K+, and Ca2<sup>+</sup> content was measured using an inductively coupled plasma optical emission spectroscopy (ICP). For nitrogen measurement, dried shoot and root sample were crushed into fine powder and analyzed using an Elemental analyzer (Elementar, Vario Micro Cube, Germany).

### In vivo Localization of Peroxide and Superoxide Radicals

The upper most leaves of 3–4 different plants were combined and immersed into 0.5 mg/ml nitroblue tetrazolium (NBT, Sigma) solution in 10 mM phosphate buffer (pH 7.8) for 2 h in the dark for superoxide detection. Whereas, for H2O<sup>2</sup> visualization, leaves were stained with 3,3'-diaminobenzidine tetrahydrochloride (DAB) solution (in 10 mM phosphate buffer, pH 3.8) for 6 h in the dark. Thereafter, for both the assays, samples were exposed to light and treated with destaining solution (ethanol: acetic acid: glycerol; 3:1:1 v/v) for 15 min at 95◦C, and then rehydrated in 40% glycerol.

### Salt Sensitivity Test of PGPR

Salt tolerance of the PGPR in the presence of a nitrogen source was observed in nutrient broth (NB) medium supplemented with 1–20% NaCl. Fresh bacterial cultures were inoculated in 5 ml NB medium and incubated at 30◦C with constant shaking at 180 rpm for 24 h. Bacterial growth was then determined by measuring the OD at 600 nm. Additionally, salt tolerance of PGPR in nitrogenfree semisolid NFb was also tested. Bacteria were enriched in semisolid NFb containing up to 4% NaCl (w/v) for 5–7 days at 30◦C (Jha et al., 2012).

### RNA Isolation, cDNA Synthesis and Real-Time PCR

Total RNA was isolated from peanut seedlings using Tri reagent (Sigma, USA) according to the manufacturer's instructions. RNA was quantified using a ND-1000 spectrophotometer (NanoDrop), and 0.5 to 1µg of RNA was reverse transcribed by using ImProm-IITM reverse transcription kit (Promega, USA). For quantitative PCR analysis, 5–10 fold diluted cDNA was used in a reaction mixture with QuantiFast SYBR Green PCR reaction kit (Qiagen, Germany) and real time quantification was performed using Real-Time iQ5 Cycler (Bio-Rad, USA). Two to three biological samples for each treatment were processed in triplicates. Ah-actin was used as an internal control and relative fold change was determined by 2−11Ct method (Livak and Schmittgen, 2001). The real time primers used in the present study were previously reported by Yang et al. (2013) and Tiwari et al. (2015) and are given in Supplementary Table 4.

### Statistical Analysis

Data analysis was performed by using IBM SPSS statistics 19. Means of different treatments were compared by one way ANOVA using Student-Newman-Keuls test (SNK test) at 5% probability level (P < 0.05). The real time data was analyzed by one way ANOVA using Dunnett's test. Wherever needed, means among different treatments were compared by t-test (P < 0.05).

### RESULTS

### Enrichment of Bacterial Isolates and nifH Sequence Analysis

Five different bacterial isolates (MBE01, MBE02, MBE03, MBE04, and MBE05) obtained from roots of A. indicum were capable to grow on nitrogen free semi-solid NFb medium containing up to 4% NaCl and malate as sole carbon source. The growth of bacteria under these conditions indicated the ability of bacterial isolates to fix atmospheric N, which was further confirmed by amplification of nifH gene and the acetylene reduction assay (see below). The sequencing of the amplified products showed similarity with nifH gene. The phylogenetic tree based on NifH amino acid sequences was constructed (Supplementary Figure 1) and the sequences were submitted to Genbank under accession numbers KX215161, KX215162, KX215163, KX215164, and KX215165 for isolates MBE01, MBE02, MBE03, MBE04, and MBE05, respectively.

### Biochemical and Molecular Characterization of Bacterial Isolates

Biochemical analysis revealed all the isolates to be gram-negative, with an ability to exhibit catalase activity and produced ammonia. Except MBE02, rest of the isolates were positive in motility test (**Table 1** and Supplementary Table 1). Other biochemical and physiological parameters were analyzed and are summarized in Supplementary Table 1.

Sequence analysis of 16S rRNA gene revealed that isolates MBE01, MBE02, MBE03, MBE04, MBE05 to share sequence identity with A. tumefaciens (98%), Klebsiella sp. (100%), Ochrobactrum anthropi (99%), P. stutzeri (99%), and Pseudomonas sp. (99%), respectively. A phylogenetic tree was constructed based on the 16S rRNA sequences shows the taxonomic positions of the isolates is shown in (Supplementary Figure 2). The sequences were submitted to Genbank under accession numbers KX083679, KX083680, KX083681, KX083682, and KX083683 for the isolates MBE01 to MBE05, respectively.

In a comparison between nifH and 16S rRNA tree, it was observed that NifH protein of MBE01 was highly related to Azospirillum species; however, it had similarity with A. tumefaciens in 16S tree. Similarly, MBE03 fall in the group of Pseudomonas with MBE04 and MBE05 but it belong to Ochrobactrum in 16S tree (Supplementary Figures 1, 2). The incongruence of these dendrograms as regards the position of MBE01 and MBE03 suggests the possibility of lateral transfer of nifH (Haukka et al., 1998).

### Estimation of Plant Growth Promoting Traits

Plant growth promoting traits, such as IAA production, Psolubilization, acetylene reduction activity, siderophore, and HCN production, were analyzed to evaluate the putative plant growth promoting activities of isolates. Results obtained indicated all of the five isolates were able to synthesize IAA. MBE03 and MBE04 had the lowest ability to produce IAA; however, MBE02 had highest IAA concentration followed by MBE01 (**Table 1**). None of the bacterial isolates were able to synthesize IAA in the absence of external tryptophan (data not shown). A. brasilense Sp7, used as a positive control, showed higher IAA level (∼1.7 to 2.9-fold) than other isolates. This is consistent with previously published reports where Azospirillum strains were shown to produce high IAA content (Akbari et al., 2007).

MBE04 and MBE05 showed lowest; however, MBE02 had highest P-solubilization activity followed by MBE01. All bacteria isolates were positive for nitrogenase activity where MBE02 and MBE01 showed maximum ARA. The highest ACC deaminase activity was observed for MBE04 followed by MBE05 isolate (**Table 1**). The positive control values obtained from these assays are given in **Table 1**. Other PGPR traits, such as siderophore and HCN production and zinc solubilization were, also, measured and shown in (Supplementary Table 2).

### PGPR Promotes Peanut Growth under Non-stress Condition

The efficacy of bacterial isolates as PGPR in growth promotion of peanut seedlings was evaluated. The results indicated that all isolates including positive control significantly increased the growth of peanut compared to control (SNK test P < 0.05; **Table 2**). The relative increase in shoot and root length varied between 14–70% and 12.9–36%, respectively. Improvement in shoot and root biomass ranged between 21–44% and 36–64% over the non-inoculated control, respectively. MBE02 exhibited highest desirable trait for plant biomass (**Table 2**).

N contents were measured in the shoot and the root of bacteria inoculated and non-inoculated peanut seedlings and shown as total plant N in **Table 2**. A significant increase in the N content was observed for inoculated seedlings, values varied between 27.4 ± 1.96 and 37.1 ± 3.19 mg g−<sup>1</sup> as compared to 21.0 ± 1.8 mg g −1 for control. The isolate MBE01 caused highest N content.

The growth promotion ability of two selected isolates (MBE02, MBE03) and the positive control were tested in a pot experiment under greenhouse conditions using sterile soil. All inoculated plants had increased biomass as compared to non-inoculated control (**Figures 1A–C**). Similar results were obtained for peanut cultivated in field plots (**Figures 1D,E**).

Plate count for root colonization efficiency of bacterial isolates revealed that isolates were able to colonize the roots, among which, highest colonization rate was observed for MBE05 at 10th and 20th day after inoculation, followed by MBE04. (Supplementary Figure 3).


"+" corresponds to a positive response. Data are means ± SE (n = 3–5). Experiments were repeated twice observing the same trend. A. brasilense strain Sp7 was used as positive control.

<sup>a</sup>µg phosphate/mg dry weight. \*µmol α-ketobutyrate µg protein-1 h -1 .

)


TABLE 2 | Effect of PGPR treatment on various growth parameters of peanut under non-stress condition.

Data are means ± SE and combined from 2 to 3 independent experiment (n = 10–20). Different letter indicates significant difference between data of the same column and calculated by one way ANOVA SNK test (P < 0.05). A. brasilense strain Sp7 served as positive control.

represented as a dash line. A. brasilense Sp7 was used as positive control. (C) A representative pot of inoculated and non-inoculated peanut seedlings of 60 days after germination. (D) Pre-inoculated peanut seeds were grown in field plots and measurement was taken after 40 days of germination. Data are means ± SE (n = 14–15). Significant differences are shown by Asterisks (\*) calculated by t-test (P < 0.05). (E) Data from (D) is shown as % increase as compared to the non-inoculated control represented by dash line.

## Halotolerant PGPR Enhances Salt Tolerance in Peanut

Since bacterial isolates originate from halophytic plant, salt tolerance capacity of the isolates were evaluated by growing them in nutrient broth (NB) medium supplemented with different concentrations of NaCl. Maximum NaCl tolerance was shown by MBE02 (up to 8%); the rest failed to exhibit the same trait (Supplementary Table 3). To determine if bacterial isolates could enhance salt tolerance in peanut seedlings, 12 d old seedlings in the presence of 100 mM NaCl and PGPR were grown and growth measurement was monitored 18 days. All isolates were found to significantly improve the seedling growth over the control (**Figure 2**). Relative increase in shoot and root biomass for PGPR treated seedlings ranged between 19–31% and 45– 64%, respectively. Other parameters, such as plant fresh weight and length were determined and are shown in Supplementary Figure 4.

Results obtained with estimation of intracellular ions with shoot and root of bacteria treated and non-treated peanut seedlings under salinity stress are summarized in **Table 3**. Seedlings inoculated with MBE01, MBE02, and MBE05 had significantly lower shoot Na+/K<sup>+</sup> ratios in comparison to noninoculated seedlings (SNK test P < 0.05). Rest of the isolates failed to exhibit similar effect. The shoot Ca<sup>+</sup> content was, also, significantly higher in seedlings inoculated with all isolates but, not with the MBE01 (**Table 3**).

Treatment of seedlings with MBE01 significantly lowered the root Na+/K<sup>+</sup> ratio as compared to non-treated samples. However, other bacterial isolates failed to exhibit similar effect (**Table 3**). The root Ca<sup>+</sup> was found to be

altered when the seedlings were treated with MBE02 and MBE03.

We next determined the accumulation of superoxide and hydrogen peroxide by using nitroblue tetrazolium (NBT) and 3,3′ -diaminobenzidine (DAB) staining. Salt treated seedlings in the absence of bacterial inoculum accumulated more reactive oxygen species (ROS) than the seedlings inoculated with PGPR (**Figure 3A**). This shows PGPR treatment might have led to differential regulation on the expression of antioxidant genes, such as ascorbate peroxidase (APX), catalase (CAT), and superoxide dismutase (SOD) (**Figure 3B**). Catalase levels were significantly induced by MBE01 and MBE04 but not by others. Expression of APX was significantly induced by MBE03 but decreased by MBE02. Isolates MBE04 and MBE05 induced expression of SOD whereas, the same was significantly reduced when peanut seedlings were treated with MBE03. The other isolates had no significant effect on SOD level (**Figure 3B**).

### DISCUSSION

Soil salinity is a major obstacle for the production of agriculture crops growing in arid and semi-arid regions. Use of halophilic or salt tolerant PGPR is an effective approach that has been employed successfully in various crops to improve their growth and tolerance under salt stress condition. The halotolerant bacteria are able to withstand high salt concentration because of their capability to accumulate compatible osmolyte to maintain intracellular osmotic balance (Nabti et al., 2015). These bacteria are positive for multiple stress-related traits that may facilitate plants to survive under growth inhibitory levels of salt (Rohban et al., 2009; Siddikee et al., 2010; Bharti et al., 2013). In the present study, salt tolerant rhizobacteria were isolated from the roots of A. indicum and growth potential of these bacteria were determined under control and salinity stress. The isolates were tolerant to 4–8% NaCl in NB medium and 3–4% NaCl concentration in nitrogen free Nfb medium. They were gram negative, nifH positive and belong to the genera Klebsiella, Pseudomonas, Agrobacterium, and Ochrobactrum. Pseudomonas is the most commonly reported genera in PGPR and the isolates belonging to this genera have shown to be involved in conferring



Data shown are means ± SE and combined from two independent experiments (n = 8-10). Different letter indicates significant difference calculated by one way ANOVA SNK test (P < 0.05).

salinity tolerance in various crop species (Jha et al., 2012). Similar reports are available for salt tolerant Agrobacterium and Klebsiella (Shukla et al., 2012; Liu et al., 2016). However, Ochrobactrum has mostly been used in phytoremediation (Al-Mailem et al., 2010) and its role in enhancing salinity tolerance of crop plants has not been studied extensively. The present study has demonstrated that halotolerant Ochrobactrum improves peanut growth under salinity stress condition.

The present isolates showed response for ACC deaminase and catalase activity. PGPR with ACC deaminase activity have been known to protect plants against environmental stresses by reducing the ethylene levels (Mayak et al., 2004b; Singh et al., 2015). Similarly, catalase, present in most of the aerobic bacteria, helps in maintaining plant ROS levels during stress (Cowell et al., 1994). Our observations on the above traits are in consistent with previous reports where diverse halotolerant bacteria were found to exhibit both of the enzymatic activities to promote plant growth under environmental stress conditions (Siddikee et al., 2010; Jha et al., 2012).

All of the isolates significantly increased the growth of inoculated peanut seedlings under non-stress conditions. The synthesis of phytohormone IAA is a frequently used mechanism of PGPR to enhance plant growth (Dimkpa et al., 2009; Glick, 2012). IAA regulates several aspects of growth and development by controlling critical biological process, such as lateral root initiation, cell enlargement, cell division and increase root surface area that helps in an uptake of soil nutrients (Zhao, 2010). All bacterial isolates, studied here, produce a significant quantity of IAA and their levels are similar or, even, higher than other PGPR that promote growth in various crop species (Karnwal, 2009; Majeed et al., 2015; Zahid et al., 2015).

Nitrogen fixation also contributes to plant growth promotion. For example, a recent study has provided direct evidence through radiolabelling experiment that bacteria (Herbaspirillum seropedicae and A. brasilense) inoculated plants incorporate biologically fixed N in major metabolism processes to promote the plant growth (Pankievicz et al., 2015). All bacterial isolates of the present study reduced acetylene, indicating that they are capable of fixing N. In accordance, a significant increase in the N content in inoculated plants grown in N-free medium was observed. This indicates active biological N fixation is well achieved in the presence of our isolates. Similar observations have also been reported in several plant species previously (Malik et al., 1997; Requena et al., 1997; Figueiredo et al., 2008; Majeed et al., 2015; Pankievicz et al., 2015). These observations imply that the increase in N content and modulation of IAA contents may contribute to peanut growth promotion by PGPR.

The present study also revealed that PGPR isolates enhance salt tolerance of peanut seedlings. Among several mechanisms used by PGPR, maintenance of low Na+/K<sup>+</sup> is considered as a predominant mechanism that favors plant growth under high salinity (Munns and Tester, 2008). In the present study, peanut seedlings treated with three isolates (MBE01, MBE02, and MBE05) showed lower shoot Na+/K<sup>+</sup> ratio than non-inoculated seedlings under salinity stress. These results were in accordance with previous studies where bacterial inoculum has been shown to maintain low Na+/K<sup>+</sup> ratio in various crop plants to reduce salt toxicity(Ozawa et al., 2007; Bano and Fatima, 2009; Shukla et al., 2012; Ramadoss et al., 2013).

Two bacterial isolates (MBE03 and MBE04) enhanced peanut growth under salinity stress; however, no change in shoot Na+/K<sup>+</sup> ratio was observed. Similar observation was previously reported for inoculated tomato (Mayak et al., 2004a). This observation suggested that the strains use alternative mechanisms to favor the plant growth under salinity. In the present study, four isolates including MBE03, and MBE04 had increased shoot Ca<sup>+</sup> accumulation than non-inoculated control. Ca<sup>+</sup> is an important secondary molecule that plays a vital role in salt signaling (Kader and Lindberg, 2010). Maintenance of high Ca<sup>+</sup> levels is a potential mechanism to reduce the damage caused by salt stress (Yang et al., 2016). The PGPR may stimulate plants to selectively take up Ca<sup>+</sup> to maintain a high Ca+/Na<sup>+</sup> ratio.

Salinity stress leads to excess ROS production and cellular toxicity in plants (Munns and Tester, 2008). The counteract severe effect of oxidative stress plants activate their antioxidant defense machinery that scavenges the excess ROS and maintain redox homeostasis (Munns and Tester, 2008; Tiwari et al., 2013). In the present study, PGPR inoculated seedlings accumulated less reactive oxygen species and affected transcript levels of antioxidant genes. The observation implies that the differential expression of antioxidant genes might be involved in the

### REFERENCES


regulation of ROS level in PGPR treated seedlings under salinity stress. This is in agreement with previous studies where PGPR treatment activated antioxidant defense response, thereby leading to low level accumulation of ROS in plants under salinity stress(Heidari and Golpayegani, 2012; Upadhyay et al., 2012; Gururani et al., 2013). However, we cannot rule out the possibility of other mechanisms rhizobacteria may use to control excess ROS accumulation in inoculated peanut seedlings under salinity stress. In summary, our present results have shown that selective ion uptake and redox homeostasis is an important protective mechanism that PGPR may use to confer salinity stress tolerance in peanut.

Overall, the present study characterized potential halotolerant PGPR that are attributed with several traits related to plant growth promotion, efficiently colonize the roots, increase total plant N and promotes peanut growth under controlled condition. Peanut inoculated with bacterial isolates maintain ion homeostasis and ROS levels under salt stress condition. Future studies focusing on characterization of these strains under multilocation field trials are in progress to commercialize them as biofertilizers.

### AUTHOR CONTRIBUTIONS

SS performed research, design the experiments, analyze data and wrote the paper; JK performed research and help in data analysis; BJ conceived the research and wrote the paper.

### ACKNOWLEDGMENTS

CSIR-CSMCRI communication number: 058/2016. The authors acknowledge financial support from the Council of Scientific and Industrial Research (CSIR) and the Government of India, New Delhi [Plant-microbe and Soil Interaction (BSC0117)]. SS acknowledge CSIR, New Delhi for the financial support received in the form of CSIR-Quick Hire Scheme.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01600


from leguminous trees growing in Africa and Latin America. Appl. Environ. Microbiol. 64, 419–426.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Sharma, Kulkarni and Jha. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Systemic Responses of Barley to the 3-hydroxy-decanoyl-homoserine Lactone Producing Plant Beneficial Endophyte *Acidovorax radicis* N35

Shengcai Han<sup>1</sup> , Dan Li <sup>1</sup> , Eva Trost <sup>2</sup> , Klaus F. Mayer <sup>2</sup> , A. Corina Vlot <sup>3</sup> , Werner Heller <sup>3</sup> , Michael Schmid<sup>1</sup> , Anton Hartmann<sup>1</sup> and Michael Rothballer <sup>1</sup> \*

<sup>1</sup> Research Unit Microbe-Plant Interactions, Department Environmental Sciences, Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Neuherberg, Germany, <sup>2</sup> Research Unit Plant Genome and Systems Biology, Department Environmental Sciences, Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Neuherberg, Germany, <sup>3</sup> Department Environmental Sciences, Institute of Biochemical Plant Pathology, Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Neuherberg, Germany

#### *Edited by:*

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### *Reviewed by:*

Ulrike Mathesius, Australian National University, Australia Sylvia Schnell, Justus-Liebig University, Germany

*\*Correspondence:* Michael Rothballer rothballer@helmholtz-muenchen.de

#### *Specialty section:*

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

*Received:* 31 August 2016 *Accepted:* 25 November 2016 *Published:* 12 December 2016

#### *Citation:*

Han S, Li D, Trost E, Mayer KF, Vlot AC, Heller W, Schmid M, Hartmann A and Rothballer M (2016) Systemic Responses of Barley to the 3-hydroxy-decanoyl-homoserine Lactone Producing Plant Beneficial Endophyte Acidovorax radicis N35. Front. Plant Sci. 7:1868. doi: 10.3389/fpls.2016.01868 Quorum sensing auto-inducers of the N-acyl homoserine lactone (AHL) type produced by Gram-negative bacteria have different effects on plants including stimulation on root growth and/or priming or acquirement of systemic resistance in plants. In this communication the influence of AHL production of the plant growth promoting endophytic rhizosphere bacterium Acidovorax radicis N35 on barley seedlings was investigated. A. radicis N35 produces 3-hydroxy-C10-homoserine lactone (3-OH-C10-HSL) as the major AHL compound. To study the influence of this QS autoinducer on the interaction with barley, the araI-biosynthesis gene was deleted. The comparison of inoculation effects of the A. radicis N35 wild type and the araI mutant resulted in remarkable differences. While the N35 wild type colonized plant roots effectively in microcolonies, the araI mutant occurred at the root surface as single cells. Furthermore, in a mixed inoculum the wild type was much more prevalent in colonization than the araI mutant documenting that the araI mutation affected root colonization. Nevertheless, a significant plant growth promoting effect could be shown after inoculation of barley with the wild type and the araI mutant in soil after 2 months cultivation. While A. radicis N35 wild type showed only a very weak induction of early defense responses in plant RNA expression analysis, the araI mutant caused increased expression of flavonoid biosynthesis genes. This was corroborated by the accumulation of several flavonoid compounds such as saponarin and lutonarin in leaves of root inoculated barley seedlings. Thus, although the exact role of the flavonoids in this plant response is not clear yet, it can be concluded, that the synthesis of AHLs by A. radicis has implications on the perception by the host plant barley and thereby contributes to the establishment and function of the bacteria-plant interaction.

Keywords: *Acidovorax radicis*, 3-OH-C10-homoserine lactone, plant growth promoting bacteria (PGPB), systemic plant responses, flavonoids, endophytes

### INTRODUCTION

In the rhizosphere, microbes are selectively enriched as compared to the surrounding bulk soil due to the availability of plant root exudates. Plant growth promoting bacteria (PGPB) are part of this microbial community exhibiting beneficial effects on plants, like biocontrol activity toward plant pathogenic organisms and promotion of plant growth due to enhanced supply of limiting nutrients like phosphate, nitrogen and essential trace elements like ferric iron. Induced systemic resistance (ISR) caused by root associated bacteria enhances the defense even in foliar tissues for later pathogen attack (Lugtenberg and Kamilova, 2009). Many molecules, so-called MAMPS (microbial associated molecular patterns) including lipopolysaccharide (LPS), exopolysaccharide (EPS), and microbial flagella elicit ISRresponses (Berendsen et al., 2012). In addition, small secondary metabolites such as the siderophore pyoverdin, the antibiotics 2,4-DAPG and lipopeptides, pseudobactins, pyocyanin, and certain biosurfactants belong to the complex spectrum of elicitors of plant responses upon contact with microbes (De Vleesschauwer et al., 2008; De Vleesschauwer and Höfte, 2009; Guillaume et al., 2012; Chowdhury et al., 2015). Also volatile organic compounds, for instance 2R, 3R-butanediol, were shown to induce plant resistance (Cortes-Barco et al., 2010). PGPB cause ISR because they initiate a priming of specific initial plant responses, upon surface or endophytic colonization. The priming status includes no upregulation of pathogen related (PR) genes, which is required in systemic acquired resistance (SAR), known to be induced by plant pathogens. Upon additional specific stress situations, PR proteins were potentially activated (Pieterse et al., 1996; Ahn et al., 2007). Quorum sensing (QS) compounds of Gram-negative bacteria, like N-acyl homoserine lactones (AHL), were also found to cause systemic ISR-like responses in different plants (De Vleesschauwer et al., 2008; De Vleesschauwer and Höfte, 2009; Cortes-Barco et al., 2010). Compared to the already advanced knowledge on the responses of plants to a large number of systemic defense elicitors, details about the perception of plants regarding these QS signal molecules are still scarce, despite the importance these bacterial messenger molecules must have considering their presence throughout the entire plant evolution.

In many Gram-negative bacteria, luxI-luxR type quorum sensing (QS) systems use N-acyl-homoserine lactones (AHLs) as auto-inducing signals (Fuqua and Greenberg, 2002). The length of the acyl-residues of AHLs produced by the I-gene, varies from 4 to 18 carbon atoms and hydroxyl- or carbonylgroup substitutions are found at the C3-position. Most of these AHL signal compounds are able to diffuse through bacterial membranes freely, while specific transporters were found for AHLs with long chain fatty acid residues (Krol and Becker, 2014). Specific luxR-type receptors or transcription factors bind AHL signal molecules at elevated intracellular concentrations leading to increased expression of the luxI-gene. Subsequently, specific gene expression is activated or suppressed by binding and releasing the AHL-LuxR transcription factor from specific gene promoter regions. AHLs dependent QS circuits are global regulons; they control a wide range of biological functions including swarming motility, bioluminescence, plasmid conjugative transfer, biofilm formation, antibiotic biosynthesis, and the production of virulence factors in plant and animal pathogens (Eberl, 1999; Waters and Bassler, 2005). Since these AHL autoinducers also convey information about the surrounding and habitat quality of the cells, AHLs play a central role in optimizing the expression of their genetic repertoire and thus have an important efficiency optimizing function (Hense et al., 2007). It turned out that AHLs not only allow bacterial populations to interact with each other but are also recognized as signals by their eukaryotic hosts. C12- and C16- side chain AHL molecules are able to induce a specific and extensive proteome response in Medicago truncatula (Mathesius et al., 2003). Using in situ bioreporter bacteria for AHLs, a production of AHLs by Serratia liquefaciens MG1 and Pseudomonas putida IsoF colonizing the rhizoplane of tomato roots were demonstrated (Gantner et al., 2006). These strains exert beneficial effects on tomato plants when inoculated to roots, since it could be shown, that the ISR-like response toward the leaf attacking fungus Alternaria alternata was dependent on the production of C6-and C8-side chain AHLs by S. liquefaciens MG1 (Schuhegger et al., 2006). In contrast, in Arabidopsis thaliana, short side chain AHLs induced phytohormonal changes in the plants and an enhancement of root growth, but no priming of pathogen response (von Rad et al., 2008). In recent years, a series of specific perception responses in different plants were reported toward the addition of long-side chain AHLs and AHL producing bacteria to roots, as summarized by Schikora et al. (2016). While most of the effects of AHLs on plants were documented when AHLs were applied as pure compounds to the medium at the roots, much less is known about how AHL production by PGPB located on or inside the root contributes to the plant's perception of these bacteria. This is not only because root colonizers produce many other substances besides AHLs the plant will respond to, but also because due to the variable bacterial colonization pattern on the root surface, which ranges from microcolonies to dense biofilms. Therefore the AHL concentration will vary quite a lot locally, which is not well-reflected by the application of an average AHL concentration to the plant growth medium.

In plant response toward bacteria flavonoids play an important role. A high diversity of flavonoids are known in different plants (Hassan and Mathesius, 2012). In barley, the most abundant flavonoids are saponarin and lutonarin (Kamiyama and Shibamoto, 2012). Flavonoids contribute to biotic or abiotic stress resistance toward oxidative damage. They are known for their antioxidant, fungicide, bactericide, and anti-pest properties (Treutter, 2005; Cushnie and Lamb, 2011; Hassan and Mathesius, 2012). Flavonoid biosynthesis genes are expressed in a tightly regulated manner and include early flavonoid biosynthesis genes (EBG) like chalcone synthase (CHS; Hassett et al., 1999), chalcone-flavonone isomerase (CFI), 4-coumarate-CoA ligase (4- CL), and UDP-glucuronosyltransferase (UGT; Besseau et al., 2007). The AHLs 3-oxo-C12-HSL and 3-oxo-C14-HSL were found to induce several of these flavonoid synthesis genes (Mathesius et al., 2003; Schenk et al., 2014).

The model organism used in this study was the type strain N35 of Acidovorax radicis, an endophytic Gram-negative bacterium originally isolated from wheat roots (Li et al., 2011). In the genus Acidovorax, pathogenic as well as saprophytic or beneficial species are known. The majority of the Acidovorax spp. are phytopathogenic for diverse plants, but there are also ubiquitously distributed saprophytic environmental Acidovorax spp. in rhizosphere and water habitats, like A. delafieldii, A. defluvii, A. temperans, and A. soli which are more closely related to A. radicis. A. radicis N35 can colonize the surface and endosphere of barley roots and shows the ability to promote barley growth in soil under certain conditions. In its genome, a homologous luxI-luxR type gene pair was identified (Li et al., 2011). The N-acyl-homoserine lactone produced by A. radicis N35 was identified as N-(3-OH-C10)-homoserine lactone using high performance liquid chromatography and FT-ICR-mass spectrometry (Fekete et al., 2007).

The objective of this study was to investigate the effect of AHL production of A. radicis on root colonization and the perception by barley plants. Therefore, we compared the wild type strain N35 and an AHL negative mutant with disrupted araI gene in their influence on barley seedlings using RNA-sequencing of leaves of inoculated barley plants and q-PCR. The analysis was focused on the flavonoid biosynthesis as part of the defense response. The results indicated that the AHLs produced by A. radicis N35 reduced systemic defense responses like flavonoid accumulation in response to the colonization by this endophytic bacterium.

### MATERIALS AND METHODS

### Strains, Culture Media, and Growth Conditions

All strains and plasmids used in this study are listed in **Table 1**. A. radicis N35 was isolated from surface sterilized wheat roots (Li et al., 2011). It was grown in NB complex medium at 30◦C at 180 rpm. Kanamycin (Km, 50µg/ml) was supplemented to growth media of YFP-labeled A. radicis N35. The A. radicis N35 araI mutant was grown in NB medium containing 20µg/ml tetracycline (Tc); for the GFP-labeled A. radicis N35 araI mutant, Km 50µg/ml and Tc 20µg/ml was added. Agrobacterium tumefaciens A136 (with plasmids pCF218 and pCF372) was cultured in NB medium with Tc 5µg/ml.

For the inoculation of barley plants, 50 ml overnight culture of A. radicis N35 wild type and araI mutant strains were harvested using 4000 g by centrifugation (Eppendorf 5417R, Eppendorf, Hamburg, Germany) for 10 min at room temperature, and the supernatant was discarded. The cells were washed twice with 50 ml of 1x PBS and thereafter the cell concentration was adjusted to an optical density (OD435nm) of 1.5 (equal to 10<sup>8</sup> cfu/ml) in 20 ml 1x PBS solution measured using a spectral photometer (CE3021, Cecil, Cambridge, England).

### Characterization of AHL Production Using AHL Biosensor Strain

AHL production of A. radicis N35 and its AHL deficient araI mutant were examined via a traI-lacZ fusion sensor plasmids in A. tumefaciens A136, which lacks the Ti plasmid and harbors the two plasmids pCF218 and pCF372. These two plasmids encode the traR and traI-lacZ fusion genes, respectively. These bio-reporter constructs allow highly efficient detection of AHLs (Stickler et al., 1998). The sensor strain was streaked to the center of an LB or NB agar plate containing 40µg/ml X-gal, and the test bacterial strains were cross-streaked close to the biosensor. The culture plates were incubated at 30◦C in the dark for 24–48 h. AHL production was detected via the activation of the reporter fusion traI-lacZ. In the presence of AHLs, beta-galactosidase activity was induced at the contact area of test and sensor strain. The metabolization of X-gal to the insoluble blue colored 5 bromo-4-chloro-3-hydroxyindole dimer indicates the presence of AHL molecules.

### Construction of an *araI* Mutant Strain

For knock-out mutagenesis in A. radicis N35, the sacB based gene replacement vector pEX18Gm described by Hoang et al. (1998) was used. First, a DNA cassette was constructed, which carried the araI target gene (amplified with primer pair AHLsyn-s2 GCCAGCTTGTCATAGGACTC and AHLsyn-as2 ATGCACCTCCAGAAAACG) disrupted by a Tc antibiotic marker (tet gene amplified with primer pair TcR-s AAAGTCTACTCAGGTCGAGG and TcR-as3 AAAGTAGACGACGAAAGGC). This cassette was cloned into the gene replacement vector pEX18Gm. Subsequently this constructed gene replacement plasmid was transferred into electrocompetent A. radicis N35 cells by electroporation. In the target cell a homologous recombination event occurred after pairing of the constructed DNA cassette with the homologous region in the genome of A. radicis N35, which led to an insertion


of the whole constructed pEX18 plasmid into the genome of N35. The cells with integrated plasmid were selected on NB medium containing antibiotics. These merodiploids were resolved by plating on NB medium containing 5% sucrose, which led to cell death if the sacB gene was expressed. Only cells where the sacB gene together with the gentamycin selective marker was eliminated from the genome by a second homologous recombination could survive on sucrose containing medium. The resulting insertion mutants A. radicis N35 araI::tet carried a disrupted dysfunctional araI gene (**Figure S1**). The success of the knock-out mutagenesis was verified with PCR using the araI specific primers and by sequencing of the PCR products (ABI Prism, Applied Biosystems, Carlsbad, CA, USA). AHL production was then visualized by the A. tumefaciens A136 AHL biosensor as described above.

### Construction of Fluorescence Labeled *A. radicis* N35 and *araI* Mutant Strains

For YFP-labeling of A. radicis N35 araI::tet, plasmid pBBR1MCS-2-YFP, a YFP expressing broad-host range vector, and for GFPlabeling of the N35 wild type, plasmid pBBR1MCS-2-GFP were used. After isolation using a NucleoSpin plasmid kit (Macherey & Nagel, Düren, Germany) the plasmid was transferred to electro-competent cells of A. radicis N35 as described by Dower et al. (1988). Electroporation was performed with a Gene Pulser instrument (Bio-Rad, Munich, Germany) using a voltage of 2.5 kV for 4.5–5.5 ms. The resulting transformants were selected on Km containing NB plates and examined for specific fluorescence with an epifluorescence microscope at an excitation wavelength of 488 nm.

### Inoculation and Growth of Barley Seedlings in Axenic System

Before germination, seeds of barley (Hordeum vulgare L.) cultivar Barke were surface sterilized to eliminate microbial contaminations using the method described by Rothballer et al. (2008). The method was slightly modified by using antibiotics (streptomycin 250µg/ml and penicillin 600µg/ml) for 20 min before testing the seeds on NB plates for 2 days at 30◦C in the dark for contaminations. After washing at least three times in sterilized water, 2 days old axenic barley seedlings with comparable root lengths of about 2 cm were selected for inoculation according to Li et al. (2012). Seeds were immersed in suspensions of A. radicis N35 and its derivative strains for 1 h before planting. For the inoculation with single bacterial strains or a bacterial mixture (v/v 1:1), the 2 days old barley seedlings were incubated in the bacterial suspension for 1 h at room temperature. Axenic cultivation of barley was performed in sealed and autoclaved glass tubes (3 cm width, 50 cm length, AG, Mainz, Germany) filled with 50 g autoclaved glass beads and 10 ml of sterile Murashige and Skoog mineral salt medium (Duchefa Biochemie, Haarlem, The Netherlands). The barley seedlings were grown at a 12 h photoperiod (metal halide lamps of 400 W) under a 23◦C / 18◦C day/night cycle for 10 days maximum until three-leaf stage in a growth chamber. The roots were harvested by taking the whole plant out of the glass tube, and rinsing off the adhering material with sterile 1x PBS.

### Visualization of Fluorescence Protein Labeled Bacteria

To visualize the GFP or YFP tagged A. radicis N35 colonizing barley roots, freshly harvested roots of barley were embedded in Citifluor and placed on glass slides. For each inoculation 6 root pieces of about 1 cm were observed. The fluorescence was detected using a confocal laser scanning microscope, CLSM 510 Meta (Zeiss, Oberkochen, Germany). The excitation wavelength at 488 nm was produced by an argon ion laser, the others at 543 and 633 nm by helium/neon lasers. Barley roots show autofluorescence which allows the visualization of the root structure. In the CLSM-images, roots were shown in magenta, GFP- labeled bacteria in green, and YFP-labeled bacteria in red color. CLSM lambda mode was used to discriminate between the very similar emission wavelengths of 510 nm for GFP and 530 nm for YFP (excitation for both 488 nm).

### Visualization of Bacteria Using Fluorescence *In Situ* Hybridization

The FISH-method as described in Manz et al. (1992) and Amann et al. (1991) was applied modified for root samples as described in Rothballer et al. (2015). For A. radicis N35 the specific probe ACISP 145 (TTTCGCTCCGTTATCCCC), combined with an equimolar mixture of the universal bacterial probes EUB 338 I (GCTGCCTCCCGTAGGA), EUB 338 II (GCAGCCACCCGTAGGTGT), and EUB 338 III (GCTGCCACCCGTAGGTGT) were used. The specific fluorescence label was visualized by a CLSM using appropriate excitation wavelengths.

### Soil Cultivation of Barley and Sample Preparation

For the cultivation of barley in soil, commercial "Graberde" (nutrient limited substrate, Alpenflor, Weilheim, Germany) was mixed with sand (v/v 1:1). Each pot (10 cm height, 8 cm diameter) was filled with the same volume of soil substrate. One liter of tap water was added to initially water the pots. Barley seeds were germinated on paper towel by incubation at room temperature for 3 days (non-sterile conditions). Seedlings without inoculation were used as control. For bacterial inoculation, seeds were treated with cell suspensions of A. radicis N35 or the A. radicis N35 araI mutant (10<sup>8</sup> cells ml−<sup>1</sup> per seedling) for 1 h, as described above. In the plant growth promotion experiment for each treatment 15 pots with only one plant per pot were cultivated for 2 weeks or 2 months. The plants were watered twice a week. Throughout the experiment, the plants were fertilized once each week with Hoagland solution (10 ml 50x stock, diluted in 1 l water). Barley plants were grown under greenhouse conditions at temperatures of 15–25◦C during the day and 10–15◦C during the night.

For RNA-sequencing, q-PCR and HPLC analysis for each treatment four leaves of 2 weeks old barley seedlings, inoculated 10 days prior to harvest were pooled. At this time point under these cultivation conditions plants were in the three leave developmental stage and their height was <20 cm excluding roots. Always the second leaves were harvested. The leaves were shock frozen in liquid nitrogen and ground in a mortar resulting in about 150 mg of sample material. Fifty milligrams of this homogenate was transferred to a 2 ml Eppendorf tube and used for RNA or flavonoid extraction.

### RNA-Sequencing

Total RNA was isolated from the prepared barley leave samples using RNeasy plant mini kit (QIAGEN, Hilden, Germany) according to the manufacturer's instruction. For each sample cDNA was generated using high capacity cDNA reverse transcription kit (Applied Biosystems, Carlsbad, CA, USA). cDNA-libraries were sequenced by the KFB Regensburg (Regensburg, Germany) using a HiSeq 2500 (Illumina, San Diego, CA, USA) in single-read mode running 100 cycles. Bioinformatic analysis was performed as described by Dugar et al. (2013). To ensure high sequence quality, Illumina reads in FASTQ format were trimmed with a cutoff phred score of 20 by the program fastq\_quality\_trimmer from FASTX-Toolkit version 0.0.13. The alignment of reads, coverage calculation, genewise read quantification, and differential gene expression were performed with READ emption which was relying on segemehl and DEseq version X (Hoffmann et al., 2009). Visual inspection of the coverages was performed using the integrated genome browser (IGB, http://bioviz.org/igb/index.html).

### q-PCR

Isolation of total RNA and generation of cDNA were performed as described above. Quantitative PCR (q-PCR) was performed using the primers as listed in **Table 2** and the SYBR green kit (PeQSTAR) of the real-time PCR system (PeQSTAR SEQ, VWR, Darmstadt, Germany). All primer pairs were verified by melting curves showing only one peak and a slope value close to −3.33. Transcript accumulation was analyzed using relative quantification with the software sigma plot. The q-PCR results are the average of three technical

### repetitions per sample and five independent plant inoculation experiments.

### HPLC Quantification of Flavone Glycosides in Barley Leaves

Ten microliters of methanol (HPLC-grade) was added for every mg of frozen sample material prepared as described above. The samples were then vortexed for 1 h on a lab shaker at 700 rpm in the dark. Samples were then centrifuged for 10 min at 11,000 rpm. The supernatants were transferred to a new cap and stored until HPLC-measurement at −80◦C. For HPLC analysis, a reversed-phase HPLC system was applied. A linear gradient over 45 min was applied with 100% solution A (2% formic acid containing 0.1% ammonium formate) to 100% solution B (0.1% ammonium formate in 88% methanol) and maintained for another 5 min. Finally, the absorbance of the eluent was measured at 280 nm (Yin et al., 2012).

### RESULTS

### Construction of AHL Synthase Mutant and *araI* Complemented Strain

To investigate the QS function in A. radicis N35, an AHL deficient mutant was constructed by disrupting the AI synthase gene araI with a tetracycline resistance gene (1.5 Kb). A complementary strain was produced by cloning the A. radicis wild type araI gene into the broad host range vector pBBR1MCS-2 and transferring this construct into the araI knock-out strain. PCR amplification with araI specific primers using as a template a complete DNA extract (containing genomic and plasmid DNA) from the araI mutant and complemented mutant resulted in the expected band sizes (**Figure 1A**). Subsequent sequencing verified the identity of the PCR amplificates and correct construction of the knock-out cassette and complementary plasmid. To test the AHL production, the AHL biosensor A.


FIGURE 1 | PCR verification (A) and a galactosidase biosensor plate assay detecting AHL production (B) of the A. radicis N35 wildtype, araI knock-out mutant (araI::tet), and complemented araI knock-out mutant (araI::tet C).

tumefaciens A136 (carrying pCF218 or pCF372) was applied. This strain can detect various types of AHLs, especially C10- HSL including the hydroxyl- or oxy-derivative at position C3 (Stickler et al., 1998). **Figure 1B** shows AHL production indicated by the blue color only for the wild type and the complemented araI mutant, and not for the uncomplemented mutant, which proves the successful knock-out of the araI gene.

### Colonization of Roots Using Differentially Fluorescence Labeled Wild Type and *araI* Mutant Strains

In order to analyze if AHL production of A. radicis N35 had an influence on the colonization ability on barley roots, differentially GFP/YFP-labeled wild type and araI mutant strains were applied. The differentially fluorescence labeled bacteria were applied to barley roots separately as well as in equal mixtures, and the barley seedlings were cultivated under axenic conditions for 1 week. After harvesting and washing the roots, the colonization behavior of wild type and araI mutant on the roots was examined using a CLSM in lambda mode, which allows to distinguish the fluorescence of GFP and YFP based on their specific emission spectrum. Both wild type and araI mutant colonized barley roots well when applied separately, although the araI mutant showed more of a single cell colonization pattern and less microcolony formation than the wild type. In general, most bacteria were found to colonize the basal parts of roots especially in the root hairs and the branching sites of side roots. When the A. radicis N35 wild type and araI mutant were applied in a 1:1 mixed inoculum, the wild type clearly dominated in colonization over the AHL negative strain (**Figure 2**). This indicates that AHL production by A. radicis N35 is important for its competitive colonization ability on barley roots.

### Plant Growth Promotion Effect

To assess whether a growth promoting effect of A. radicis was possibly dependent on AHL production, seedlings were inoculated with N35 wild type and the araI mutant strains or not inoculated as control and grown under axenic conditions in the growth chamber or in non-sterile soil in the greenhouse. After 2 weeks or 2 months in the soil system and 2 weeks in the axenic system, barley plants were harvested and total fresh weight of the plants was measured (**Figure 3**). In the soil system, a significant growth promotion effect in response to inoculation with A. radicis N35 and the araI mutant on total plant fresh weight was found after 2 months (**Figure 3A**). In the axenic system no significant stimulation of growth was detectable after 2 weeks upon inoculation (**Figure 3C**). When the colonization of roots was analyzed using FISH, only a few cells of A. radicis N35 could be detected after 2 weeks, and it was not detectable anymore after 2 months in the soil system (not shown). In the axenic system, the colonization by A. radicis N35 was very well detectable with FISH (not shown) and GFP labeled cells during the whole growth period of 2 weeks (**Figure 2** and **Figure S2**).

### Barley Transcriptome Analysis

To investigate which plant genes were differentially regulated in barley leaves after inoculation in response to colonization by AHL producing and non-producing A. radicis N35 compared to the un-inoculated control plants, the plant transcriptome was analyzed via next generation sequencing alongside with a series of specific qPCR assays for verification of the sequencing results. In the barley leaf transcriptome sequencing analysis a number of gene transcripts were found to be significantly enhanced or suppressed by A. radicis N35 and/or the araI mutant at 10 days post inoculation (dpi) compared to the uninoculated control plant (**Figure 4**). These plant transcripts can be classified into two groups: (1) AHL independent transcripts,

correlated to the presence or absence of A. radicis N35, regardless if wild type or mutant were inoculated, and (2) AHL dependent transcripts, correlated to whether or not inoculated A. radicis N35 was able to produce AHL. Interestingly, transcript sequencing results from leaves after 10 dpi indicated that the transcription of several flavonoid synthesis genes (Besseau et al., 2007) was upregulated when plants were inoculated with the araI mutant (transcript group 2), including UDPglycosyltransferase-like protein (UGT), CFI, chalcone synthase (CS), 4-coumarate-CoA ligase (4-CL) and chaperone protein (DnaJ). Thus, these five transcripts were selected for verifying the results by qPCR. Additionally, two genes of transcript group 1 were also selected, namely F-box family-3 gene (fb-3, 1) and the E3 ubiquitin-protein ligase PRT1 gene (PRT1), which were downregulated in response to inoculation with A. radicis N35 compared to the uninoculated control. Primers were designed for each of these seven genes (**Table 2**) and tested based on the standard and melting curve (**Figure S4**). The q-PCR assays (**Figure 5**) confirmed the transcriptomic sequencing data for the five group 2 transcripts related to the flavonoid pathway. The expression of these genes was between two- and fourfold higher with the A. radicis N35 araI mutant than with the AHL producing wild type. The downregulation of the two group 1 genes by both mutant and wild type could also be confirmed.

### Content of Saponarin and Lutonarin in Barley Leaves

The HPLC-analysis revealed that in the leaves of the tested barley cultivar Barke the concentration of saponarin was generally higher than of lutonarin. In plants inoculated with the A. radicis N35 araI mutant, the contents of saponarin and lutonarin were ∼two-fold compared to plants inoculated with the wild type or non-inoculated controls (**Figure 6** and **Figure S3**). In

and 2 weeks, respectively. (C) Total fresh weight of barley seedlings 2 weeks after inoculation with A. radicis N35 wild type or the araI mutant, respectively, under axenic growth condition. Different letters indicate significant differences among treatment groups (p < 0.05).

addition the amount of lutonarin methylether in araI mutant inoculated plants reached almost twice the level detectable in wild type inoculated and non-inoculated control plants. These results corroborate the data from the gene transcriptome profiling by sequencing and q-PCR, which leads toward the conclusion that AHL signaling by barley root colonizing A. radicis N35 impacts flavonoid production in barley leaves.

### DISCUSSION

### Influence of AHLs on the Root Colonization of *A. radicis* N35

AHL production by A. radicis N35 has a positive effect on the colonization of roots, since an AHL defective mutant was less successful in root colonization (**Figure 3**). The QS-deficient araI mutant strain showed colonization mostly by single cells spread randomly over the root surface, while the N35 wild type cells grew more aggregated in microcolonies. This result corroborates the observation by Li (2010), who showed in a 1:1-mixture of GFP-labeled N35 wild type cells and the SYTO orange labeled araI mutant that only a few mutants colonized the roots, while wild type cells showed dense colonization. Even though the mutants were colonized together with the wild type in these experiments, the AHL deficient cells could obviously not profit from the produced AHL by the wild type in their vicinity. However, it has been shown previously (Hense et al., 2007), that quorum sensing on the rhizoplane must be considered as a strictly localized phenomenon, mainly taking place in microcolonies forming in small niches on the root surface confined by diffusion limitation. Here, the signaling substances can locally reach very high concentrations far above the average values which are measurable by taking e.g., liquid samples from the rhizosphere. Thus, although some mutant cells might profit from AHL producing neighboring wild type cells, the overall AHL concentration these mutants were exposed to, was probably below the threshold levels required for the autoinduction process. Consequently, QS regulated genes were not activated which might have led to the observed low competitive colonization phenotype.

In A. radicis N35, also phenotypic variants showed reduced root colonization. However, in contrast to araI mutants these variants had much reduced ability of plant growth promotion (Li et al., 2012). The reduced colonization of the araI mutant could be caused by a reduced tolerance toward reactive oxygen species (ROS) released by barley roots upon first contact with microbes as has been found in the case of the endophyte Gluconacetobacter diazotrophicus during colonization of rice roots (Alquéres et al., 2013). In this case, ROS-quenching enzymes catalase and superoxide dismutase of the endophyte have a major role in the degradation of ROS released by the host plants during early host defense. In P. aeruginosa, QS was found to be involved in the stress tolerance, and luxI-type QS-deficient mutants (lasI, rhlI, and lasI rhlI) have defective expression of catalase (CAT) and superoxide dismutase activities (SOD). These mutants were more sensitive to oxidative stress than the parental strain (Hassett et al., 1999). Another study showed that the QS based stress tolerance can make it more difficult to quench quorum sensing activities and help to prevent social cheating (García-Contreras et al., 2015). In the co-inoculation experiment, the quorum sensing araI mutant may behave even as such a quorum sensing cheater, because it does not produce any AHL. The positive influence of QS on biofilm formation was shown in several studies. For instance, in the Pseudomonas fluorescence 2p24, the pcoI coded AHL synthase mutant resulted in seriously decreased biofilm formation, leading to less root

FIGURE 4 | Transcriptome analysis by RNA sequencing from 4 pooled leaves of barley plants grown in parallel. Total RNA was isolated from barley leaves 10 days after root inoculation with A. radicis N35 wild type and the araI mutant, respectively. Red (upregulated) and green (downregulated) colors represent an at least three-fold difference in the amount of detected gene transcripts for the respective gene between the analyzed samples. Black color means no change between the expression levels of the found transcripts.

(B) Fb-3, F-Box family-3 and E3 ubiquitin-protein ligase (PRT1).

colonization ability (Wei and Zhang, 2006). In P. aeruginosa, a quorum sensing lasI mutant formed flat undifferentiated biofilms which are more sensitive to the biocide sodium dodecyl sulfate than the wild type. These flat biofilm types of AHL defective mutants could be restored by exogenous addition of AHLs (Davies et al., 1998). Also in Sinorhizobium fredii SMH12, micro-colony biofilm formation was found regulated by QS and in Rhizobium spp., the biofilm formation is dependent on the production of AHLs (Davies et al., 1998; Rinaudi and Giordano, 2010; Pérez-Montaño et al., 2014). It could be shown, that exogenous addition of AHLs could promote the biofilm formation by Acidovorax sp. strain MR-S7 (Kusada et al., 2014). The importance of QS for the biofilm formation could be due to secretion of important compounds like extracellular DNA, the biosurfactant rhamnolipid and the secretion of the BapA-protein as shown in P. aeruginosa (Tolker-Nielsen, 2015). Furthermore, QS-compounds play an important role in P. fluorescence 2p24 for its colonization on wheat roots and development of biocontrol ability toward the take-all disease fungus (Wei and Zhang, 2006). In Burkholderia phytofirmans PsJN QS was also found to be important for its competitive biofilm formation and efficient colonization of roots and beneficial interaction with A. thaliana as plant host (Zúñiga et al., 2013). Thus, there is an increasing knowledge about the important role of AHLs in plant beneficial rhizosphere bacteria and endophytes in different plant systems and their involvement in different mechanisms of plant growth promotion. It could be even shown, that the exopolysaccharide production in S. fredii NGR234 was modified by AHL production (Krysciak et al., 2014).

### Influence of AHL Producing *A. radicis* on the Biosynthesis and Content of Flavonoids in Barley Leaves

Compared with A. radicis N35 wild type bacteria, the colonization of roots by the AHL deficient araI mutant caused an accumulation of saponarin and lutonarin in barley leaves (**Figure 6**). This indicates that AHLs themselves or bacterial components induced by AHLs are involved in the regulation / induction of flavonoid biosynthesis in the host plant. A direct stimulatory effect of AHLs on the induction of flavonoid biosynthesis was first found in M. truncatula. In this case, 3-oxo-C12-HSL was shown to activate the transcription of chalcone synthase genes in white clover roots (Mathesius et al., 2003). In the A. radicis N35—barley interaction, a different AHL (3-OH-C10-HSL) is operating, which may have caused an inhibition of flavonoid biosynthesis. In A. thaliana, the influence of the length of the acyl chain and the substitution at the AHL C3 position were shown to cause different systemic responses (Schikora et al., 2011). The contrasting response of barley to 3-OH-C10 HSL may also be due to the fact that the monocotyledonous barley may respond differently to AHLs than the dicotyledonous white clover. On the other hand, since QS autoinducers are able to regulate bacterial surface exopolysaccharide production (Krysciak et al., 2014), the lack of AHLs in the A. radicis araI mutant could also have resulted in considerable changes in the surface exopolysaccharide structure and this may have caused a different plant response.

Flavonoids are known to help plants to acquire resistance toward various biotic and abiotic stresses (Treutter, 2005). For example, flavonoids were found in relation to drought stress in winter wheat (Ma et al., 2014) and could exhibit antifungal activities, e.g., in the carnation-Fusarium pathosystem (Curir et al., 2004). Furthermore, lutonarin and saponarin isolated from barley sprouts have been shown to be effective against bacterial pathogenesis enzymes (Park et al., 2014). Finally both lutonarin and saponarin are frequently discussed as antioxidative and antimicrobial agents in alternative medicine, as e.g., recently reviewed in Lahouar et al. (2015). The enhanced accumulation of the two flavone glycosides saponarin and lutonarin in barley leaves caused by the colonization of the roots by the A. radicis N35 araI mutant is an example for this kind of defense response. The expression of several flavonoid biosynthesis genes was upregulated due to inoculation with

the A. radicis N35 araI mutant. This clearly indicated that the AHL deficient mutant strains activated a defense response. Three closely related R2R3-MYBs transcription factors (MYB11, MYB12, and MYB111) redundantly activate the transcription of early flavonoid biosynthesis genes (EBGs). The UDPglycosyltransferases UGT91A1 and UGT84A1 together with CHS, CHI, and F3H, FLS1 are controlled by this R2R3MYB factors in Arabidopsis (Stracke et al., 2007). However, no data were obtained for the regulation at the transcriptional level. The flavonoid accumulation is not only regulated at the transcriptional, but also at the post-transcriptional level through PAL-degradation mediated by Kelch domain-containing F-box (KFB) proteins and degradation of E3 ubiquitin ligase (PRT1) complexes leading to the suppression of the phenylpropanoid pathway (Feder et al., 2015). In this study, it could be demonstrated by the transcriptomic sequencing results and confirmed by q-PCR, that the expression of F-box protein and E3 ubiquitin-protein ligase were downregulated (**Figure 5**). F-box family proteins are components of the SCF-protein complex, which is involved in the proteome degradation pathway. This process is for example important for plant development and immunity response to various stress condition (Thines et al., 2007). In addition, also an upregulation of dnaJ expression after inoculation with the araI mutant was shown mediated by transcription factor SG7 MYB (**Figure 5**). Its expression was found to correlate with flavonoid related genes and to be under the control of MYB transcription factors (Stracke et al., 2007). The upregulation of dnaJ expression also correlated with the upregulation of the flavonoid biosynthesis genes and flavonoid accumulation. DnaJ is also involved in salt stress resistance and known to interact with Hsp70 in the heat shock resistance process (Zhu et al., 1993). Since dnaJ expression was also found to be involved in regulation of saline tolerance, it is reasonable to test whether the higher expression of dnaJ will also result in increased salt stress tolerance by the plants.

### Integrated Role of AHLs by *A. radicis* in Plant Perception

Due to common evolutionary history, plants, and microbes have developed an elaborate system of mutual detection, cooperation, or deterrence. In the first recognition step, the most important function is the plants' innate immune system recognizing MAMPS and diverse microbial elicitors. On the microbial side the response to plant surface structures and exudates have a central role in recognition. The quorum sensing communication system of rhizobacteria based on AHL compounds may be considered as an integrated part in the perception of bacteria by plants. In the plant growth beneficial endophyte A. radicis N35, 3-OH-C10-HSL is the dominant AHL (Fekete et al., 2007).

However, many Gram-negative plant pathogenic rhizobacteria also synthesize AHLs, although with different chain lengths and other functional groups. Since the onset of virulence is regulated by these auto-inducers, the plant is expected to learn about the presence of AHLs in their vicinity as soon as possible. In the case of pathogens, the network of multiple interactions concludes to initiate full expression of the defense cascade, while in the case of beneficial endophytes the response is dampened or completely suppressed allowing a cooperative interaction. There are several examples, that AHL compounds applied to rooting solutions of plants can exert diverse beneficial effects on plants, which include growth promotion as well as priming or induction of pathogen resistance in the host. This was shown in different plant species, such as M. truncatula, tomato, A. thaliana, and barley (Mathesius et al., 2003; Schuhegger et al., 2006; von Rad et al., 2008; Schenk et al., 2014). However, it is much less clear, what role AHLs of a beneficial root colonizing bacterium play in the concert of interaction with all its other compounds in the plant's recognition and perception system. In the current study, it could be shown, that the production of 3-OH-C10-HSL during the colonization process by the plant growth promoting, endophytic A. radicis N35 is able to efficiently influence the plant response and reduce the onset of a defense cascade. Whether this is caused by a direct interaction with AHLs or by an indirect effect through the induction of e.g., a different surface structure of the bacteria in the presence of AHLs, which is not recognized as a pathogenic signal, is not known yet. Nevertheless, the response of barley plants to A. radicis N35 wild type is characterized by the absence of expression of several genes involved in flavonoid biosynthesis in the plant leading to an attenuated defense response. Apparently, the 3-OH-C10-HSL-production is playing a major role in this process. Future detailed studies need to

### REFERENCES


focus on the role of the quorum sensing compound 3-OH-C10-HSL on the regulation of expression of enzymes involved in the modification of the fine structure of the cell surface lipo- and exopolysaccharides or of type III secretion systems or other transport systems potentially involved in bacteria-host interactions.

### AUTHOR CONTRIBUTIONS

SH and DL performed the experimental work and wrote the paper; ET and KM did bioinformatic data analysis; AV and WH did plant metabolite analysis and discussion; MS, AH, and MR conceived the research and wrote the paper.

### FUNDING

This study was funded by the China Scholarship Council (CSC), file No. 2011911941 to SH.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016. 01868/full#supplementary-material

Figure S1 | Construction of *araI* mutant in *A. radicis* N35.

Figure S2 | Colonization of barley roots by *A. radicis* N35 GFP in a monoxenic system at different time points after inoculation (dpi as indicated in the pictures).

Figure S3 | Upper part: HPLC-separation of compounds with absorbance at 280 nm (see MM). Lower part: UV-spectra of the different eluted compounds.

Figure S4 | qPCR primer melting curves.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Han, Li, Trost, Mayer, Vlot, Heller, Schmid, Hartmann and Rothballer. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Antimicrobial Activity of Medicinal Plants Correlates with the Proportion of Antagonistic Endophytes

Dilfuza Egamberdieva<sup>1</sup> \*, Stephan Wirth<sup>1</sup> , Undine Behrendt<sup>1</sup> , Parvaiz Ahmad2,3 and Gabriele Berg<sup>4</sup>

1 Institute of Landscape Biogeochemistry, Leibniz Centre for Agricultural Landscape Research, Müncheberg, Germany, <sup>2</sup> Department of Botany and Microbiology, Faculty of Science, King Saud University, Riyadh, Saudi Arabia, <sup>3</sup> Department of Botany, Sri Pratap College, Srinagar, India, <sup>4</sup> Institute of Environmental Biotechnology, Graz University of Technology, Graz, Austria

Medicinal plants are known to harbor potential endophytic microbes, due to their bioactive compounds. In a first study of ongoing research, endophytic bacteria were isolated from two medicinal plants, Hypericum perforatum and Ziziphora capitata with contrasting antimicrobial activities from the Chatkal Biosphere Reserve of Uzbekistan, and their plant-specific traits involved in biocontrol and plant growth promotion were evaluated. Plant extracts of H. perforatum exhibited a remarkable activity against bacterial and fungal pathogens, whereas extracts of Z. capitata did not exhibit any potential antimicrobial activity. Matrix-assisted laser desorption ionization (MALDI) timeof-flight (TOF) mass spectrometry (MS) was used to identify plant associated culturable endophytic bacteria. The isolated culturable endophytes associated with H. perforatum belong to eight genera (Arthrobacter, Achromobacter, Bacillus, Enterobacter, Erwinia, Pseudomonas, Pantoea, Serratia, and Stenotrophomonas). The endophytic isolates from Z. capitata also contain those genera except Arthrobacter, Serratia, and Stenotrophomonas. H. perforatum with antibacterial activity supported more bacteria with antagonistic activity, as compared to Z. capitata. The antagonistic isolates were able to control tomato root rot caused by Fusarium oxysporum and stimulated plant growth under greenhouse conditions and could thus be a cost-effective source for agro-based biological control agents.

Keywords: Hypericum perforatum, Ziziphora capitata, endophytic bacteria, plant growth traits, antimicrobial activity, antagonism

### INTRODUCTION

Medicinal plants are traditionally used worldwide as remedies for the treatment of various diseases, including asthma, gastrointestinal symptoms, skin disorders, respiratory and urinary problems, and hepatic and cardiovascular disease (Van Wyk and Wink, 2004; Tian et al., 2014). These plants synthesize a diverse array of biologically active compounds (Bajguz, 2007; Cushnie et al., 2014) that are important for them to survive and flourish in the natural environment, including protective functions with respect to abiotic stresses derived from temperature, water status, mineral nutrient supply and to insect pests (Simmonds, 2003; Treutter, 2006; Vardhini and Anjum, 2015). The composition of biologically active compounds of medicinal plants varies widely depending on the plant species, soil type and on their association with microbes (Zhao et al., 2011; Morsy, 2014).

#### Edited by:

Gero Benckiser, Justus-Liebig-University Giessen, Germany

#### Reviewed by:

Roberta Fulthorpe, University of Toronto Scarborough, Canada Stefanie P. Glaeser, Justus-Liebig-University Giessen, Germany

> \*Correspondence: Dilfuza Egamberdieva

egamberdieva@yahoo.com

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Microbiology

Received: 27 August 2016 Accepted: 27 January 2017 Published: 09 February 2017

#### Citation:

Egamberdieva D, Wirth S, Behrendt U, Ahmad P and Berg G (2017) Antimicrobial Activity of Medicinal Plants Correlates with the Proportion of Antagonistic Endophytes. Front. Microbiol. 8:199. doi: 10.3389/fmicb.2017.00199

These bioactive secondary metabolites synthesized by medicinal plants can also strongly affect plant-associated microbial communities and their physiological functions (Qi et al., 2012; Philippot et al., 2013; Chaparro et al., 2014; reviewed in Köberl et al., 2013). Moreover, plants rely on their microbiome for specific traits and activities, including growth promotion, nutrient acquisition, induced systemic resistance and tolerance to abiotic stress factors (Egamberdieva et al., 2010, 2011; Malfanova et al., 2011; Sessitsch et al., 2013; Berg et al., 2014). Although a vast number of medicinal plants have been well-studied with respect to their phytochemical constitutes and pharmacological properties, their microbiome and the physiological interactions between host and microbes remain poorly understood (Köberl et al., 2014).

The plant-associated microbiome consists of distinct microbial communities living in the roots, shoots and endosphere (Beneduzi et al., 2012; Berg et al., 2014). The rhizosphere of many plants is well-studied and known to be a potential source for selecting beneficial microbes that can positively affect plant health (Weller et al., 2002; Berendsen et al., 2012; Philippot et al., 2013). Hence, understanding the response of microbial communities to alterations in the physiochemical environment of the rhizosphere may provide valuable insights into the microbial ecology of plant-associated bacteria. Köberl et al. (2013) observed a high abundance of antagonistic bacteria in the rhizosphere of the medicinal plants Matricaria chamomilla, Calendula officinalis, and Solanum distichum. The root-associated bacteria of Ajuga bracteosa exhibited a wide range of plant growth promoting activities by producing siderophores and indole acetic acid and exhibiting antioxidant activity (Kumar et al., 2012). Recently, endophytic microorganisms have been under increased investigation due to their intimate interaction with the host (Hardoim et al., 2015); it is believed that the phytochemical constitutes of plants are related either directly or indirectly to endophytic microbes and their interactions with host plants (Chandra, 2012; Qi et al., 2012). Despite first studies of endophytes in medicinal plants (Bharti et al., 2012; López-Fuentes et al., 2012; Miller et al., 2012; El-Deeb et al., 2013; Egamberdieva and Teixeira da Silva, 2015), the potential of medicinal plants is far from exhausted.

Therefore, the current exploratory study was designed to evaluate whether medicinal plants with contrasting antimicrobial activities have an impact on plant-specific traits involved in biocontrol and plant growth promotion of root-associated culturable endophytic bacteria. In first experiments of ongoing research, we studied Ziziphora capitata L. (Field basil) and Hypericum perforatum L. (St John's wort) from the Chatkal Biosphere Reserve of Uzbekistan, an isolated protected area in Western Tien Shan province, which significantly surpasses other areas with respect to the absolute number of endemic species (Kogure et al., 2004). Z. capitata L. is a medicinal and aromatic plant of the Lamiaceae family, which is traditionally used for the treatment of various ailments, such as heart disease, inflammation, depression, diarrhea, fever, skin disorders, hepatic diseases, and edema (Sonboli et al., 2006). The Ziziphora species are rich in essential oils, flavanoids and sterols (Zhaparkulova et al., 2015). The major component of essential oil found in several species of Ziziphora is pulegone, which has strong antibacterial and antifungal activity (Sonboli et al., 2006), but Z. capitata does not contain pulegone (Ebrahimi et al., 2009). H. perforatum is a species in the family Hypericaceae and is known for analgesic, sedative, antihelmintic, antiinflammatory, and antibacterial properties (Dall'Agnol et al., 2003). H. perforatum contains a wide range of biological active compounds, such as essential oils, tannins, flavonoids, xanthones, and hyperforin as an antibiotic substance (Jurgenliemk and Nahrstedt, 2002). The crude extracts of H. perforatum exhibited higher antibacterial activity against Gram-positive than Gramnegative bacteria (Sarkisian et al., 2012). The aim of this study was to isolate and characterize endophytic bacteria from two medicinal plants, H. perforatum and Z. capitata, with contrasting antimicrobial activities and evaluate their plant-specific traits involved in biocontrol and plant growth promotion.

### MATERIALS AND METHODS

### Collection of Plant Samples

Hypericum perforatum (Hypericaceae) and Ziziphora capitata (Lamiaceae) plants were collected during the summer (June 2013, the plant's flowering stage) from Chatkal Biosphere Reserve of Uzbekistan, western part of Tien Shan mountain (41◦ 08<sup>0</sup> N; 69◦ 59<sup>0</sup> E). This biosphere reserve is situated in the Tashkent Region within the Chatkal mountain range (1.110–4.000 m above sea level) of the West Tien-Shan Mountains and is unique for its significant role in biodiversity conservation and ethnobotany. The climate is characterized by average annual temperatures ranging from 20 to 25◦C with increased annual precipitation from plains to mountains, reaching 700–800 mm.

### Preparation of Plant Extracts

The aerial parts of H. perforatum and Z. capitata were dried in the laboratory excluding direct sun light at room temperature for 6–7 days and ground into a fine powder by mortar and pestle. Approximately, 10 g of plant powder was extracted with 50 ml of methanol for 24 h in a dark room temperature. Subsequently, the solvent was evaporated in a rotary vacuum evaporator at 40◦C and re-suspended in dimethyl sulfoxide (DMSO). The homogenate was filtered through Whatman No. 1 filter paper, centrifuged at 5000 g for 15 min and sterilized by filtration through 0.22-µm sterile filters (Millipore, Bedford, MA, USA). The filtrates were stored at −4 ◦C and used for in vitro screening of antimicrobial activity.

### Antimicrobial Activity of Plant Extracts

The extracts were individually tested against the following pathogenic microorganisms: Klebsiella oxytoca 6653, K. pneumoniae 40602, K. aerogenes NCTC 8172, Citrobacter freundii 82073, Staphylococcus aureus MRSA 16, Enterococcus faecalis NCTC 775, Providencia rettgeri NCIMB 9570, Pseudomonas aeruginosa NCTC 6749, Escherichia coli NCTC 9001 and Fusarium solani, Fusarium oxysporum, and Alternaria alternata. Reference strains and clinical isolates were obtained from the Department of Microbiology, Manchester Metropolitan

University, UK, and the National Culture Type Collection (NCTC), UK. The fungal strains were obtained from the Department of Microbiology and Biotechnology, National University of Uzbekistan. Each plant extract was dissolved in dimethyl sulfoxide (DMSO), sterilized by filtration using a sintered glass filter, and stored at 4◦C. The antimicrobial activity of the extracts was tested using the agar well-diffusion method. Microorganisms were grown overnight at 30◦C in Mueller-Hinton Broth (Oxoid, Basingstoke, UK) supplemented with 5% horse blood, and 100 µl of suspension containing 10<sup>6</sup> CFU ml−<sup>1</sup> of bacteria was spread on the surface of Mueller-Hinton agar plates. Wells with 6-mm diameters were cut off and filled with 50 µL of each extract (10 mg ml−<sup>1</sup> ). Ampicillin (Sigma-Aldrich, Steinheim, Germany) (0.5 mg ml−<sup>1</sup> ), nystatin (Sigma-Aldrich, Steinheim, Germany) (1 mg ml−<sup>1</sup> ) and DMSO were used as controls. Fungal strains were grown on potato dextrose agar plates (PDA; Difco Laboratories, Detroit, MI, USA) at 28◦C for 5 days. Small piece of fungal culture were placed in the middle of Petri plates. Each antimicrobial assay was performed in triplicate. The plates were incubated at an appropriate growth temperature for 2 days for bacterial strains (37◦C) and 4 days for fungal strains (30◦C). The assessment of antimicrobial activity was based on the measurement of inhibition zones on the agar surface around the well.

### Isolation of Endophytic Bacteria

Three plants from each species of H. perforatum and Z. capitata including roots (20–30 cm depth) were randomly collected about 1 m apart from each other from an area of 100 m<sup>2</sup> in the Chatkal Biosphere Reserve. The whole plants, along with root systems, were wrapped in plastic bags, and brought to the laboratory on same day and immediately stored at 4◦C. The isolation of bacterial strains was carried out on the next day to minimize storage effects.

The root systems of the collected plants were separated from the shoots, soil adhering to the roots was removed and roots were carefully washed under running water, taking care to minimize root injury. Three plants of each species were used to determine the number of bacterial colonies cultured from the root tissue. For the bacterial isolation, root tissues were pooled from each of three replicate plants. The roots were surface sterilized by immersion in 70% (v/v) ethanol, following by shaking in 5% (w/v) sodium hypochlorite solution for 5 min. Subsequently, the roots were rinsed in sterile distilled water six times. To test the efficiency of sterilization, the sterile roots were incubated in TSA medium for 2 days at 28◦C, and no infestation was observed.

Sterilized roots were weighed aseptically (1 g) and macerated in a mortar employing phosphate buffered saline (PBS) (20 mM sodium phosphate, 150 mM NaCl, pH 7.4) in a laminar air flow cabinet. The extracts were placed in a tube containing 9 ml sterile PBS and shaken with a vortex for 1 min. The supernatant was collected and serially diluted (101–10<sup>5</sup> ) in PBS, and 100 µl from appropriate dilutions were spread on Tryptic Soy Agar (TSA, Difco Laboratories, Detroit, MI, USA) plates in triplicate. The plates were incubated at 28◦C, and colony forming units (cfu) g−<sup>1</sup> root tissue were determined on the third day. A representative number of colonies that exhibited differentiable colony morphologies were picked from plates and were re-streaked for the purification of the isolates. The pure bacterial cultures were preserved on plates at 4◦C for the further analyses. In addition all bacterial isolates were stored in Tryptic Soy broth (TSB) (Difco) with 30% glycerol at −80◦C.

### Identification of Endophytic Bacterial Strains

The identification of bacterial isolates was performed by matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) as described previously (Egamberdieva et al., 2016). The sample preparation was performed according to the ethanol/formic acid extraction protocol recommended by Bruker Daltonics (Bremen, Germany) and was described in Egamberdieva et al. (2016). Briefly, the isolates were cultured on TSA medium (Difco Laboratories, Detroit, MI, USA) for 24 h, and approximately 10 mg of cell mass was suspended in 300 µL water (LC–MS CHOMASOLV <sup>R</sup> ; Honeywell) and vortexed to generate a homogenous suspension. The suspension was mixed with 900 µL ethanol (≥99.8% GC; Sigma-Aldrich) and centrifuged. The pellet was resuspended in 50 µL 70% formic acid (v/v) and subsequently carefully mixed with 50 µL acetonitrile. After centrifugation, aliquots of 1 µL supernatant were placed immediately on spots of a MALDI target. Each spot was allowed to dry and subsequently overlaid with 1 µL of matrix (α-ciano-4-hydroxycinnamic acid in 50% aqueous acetonitrile containing 2.5% trifluoroacetic acid). Mass spectra were acquired using a MALDI-TOF MS spectrometer in a linear positive mode (MicroflexTMLT, Bruker Daltonics, Bremen, Germany) in a mass range of 2–20 kDa. A bacterial test standard (BTS, Bruker Daltonics, Bremen, Germany) was used for instrument calibration. The raw spectra were imported into the MALDI BiotyperTM software and then processed and analyzed using standard pattern matching against the reference spectra in the MALDI BiotyperTM reference database (version 3.0, Bruker Daltonics, Bremen, Germany). A calculated matching score (score value) provided a measure of the probability of a correct classification.

### In vitro Screening for Plant Beneficial Traits

The production of IAA (indole 3-acetic acid) was determined as described by Bano and Musarrat (2003). The IAA concentration in culture was calculated using a calibration curve of pure IAA as a standard. The cellulose-degrading ability of bacterial isolates was analyzed by streaking inocula on cellulose (Sigma-Aldrich, St. Louis, MO, USA) Congo-Red agar media as described by Pratima et al. (2012). Furthermore, β-1,3 glucanase activity was tested using the substrate lichenan (Sigma-Aldrich, St. Louis, MO, USA) in top agar plates (Walsh et al., 1995), and protease activity was determined by using 5% skimmed milk agar plates (Brown and Foster, 1970). The production of HCN by bacterial isolates was measured using the protocol described by Castric (1975).

The bacterial isolates were tested in vitro for their antagonistic activities against the following pathogenic fungi: Fusarium

oxysporum f.sp. radicis-lycopersici (Forl), F. solani, F. culmorum, Gaeumannomyces graminis pv. tritici (Ggt), Alternaria alternata, and Botrytis cinerea and the oomycete Pythium ultimum. The bacterial isolates were grown in TSB broth for 3 days, and 50 µl bacterial cultures were dropped into a hole of PDA plates (4 mm in diameter). Fungal strains for inoculation were grown in peptone dextrose agar (PDA) plates at 28◦C for 5 days. Disks of fresh cultures of the fungus (5 mm diameter) were cut out and placed 2 cm away from the hole filled with bacterial filtrate. The plates were sealed with Parafilm <sup>R</sup> M and incubated at 28◦C in darkness until the fungi had grown over the control plates without bacteria. Antifungal activity was recorded as the width of the zone of growth inhibition between the fungi and the bacteria tested.

### Biological Control of Tomato Root Rot

Bacterial isolates with antagonistic activity against the majority of tested fungal pathogens, were tested for their ability to control tomato root rot caused by F. oxysporum f.sp. radicis-lycopersici. For the inoculation of soil, F. oxysporum was grown in PDA plates for 5 days. Small pieces of agar from the growing edge of the colony were homogenized and used to inoculate 300 ml of Chapek-Dox medium, which was kept under aeration (110 rpm) at 28◦C. After 3 days, the spore suspension was filtrated with sterile glass wool to remove the mycelium. The concentration of spores in the inoculum was adjusted to 10<sup>7</sup> spores ml−<sup>1</sup> by microscopic enumeration with a cell-counting haemocytometer and mixed thoroughly with potting soil to obtain a concentration of approximately 10<sup>7</sup> spores kg−<sup>1</sup> soil. The tomato seeds of the cultivar Fuji Pink (Sakata, Japan) were sterilized by stirring with 70% ethanol for 5 min and in household bleach (adjusted to approximately 5% sodium hypochlorite) for 3 min. Subsequently, the seeds were washed several times with sterile distilled water. After germination in sterile Petri plates, the seeds were placed in a bacterial suspension of 1 × 10<sup>8</sup> CFU ml−<sup>1</sup> prepared as described above and shaken gently for 10 min. The inoculated seeds were sown in plastic pots, and each treatment contained four groups of 24 plants. The plants were grown in a growth chamber under controlled conditions (16 h light, 8 h dark), at temperature light 28◦C, dark 20◦C and relative humidity 60%. After 3 weeks, the plants were removed from the soil, washed and examined for foot and root rot symptoms as indicated by browning and lesions. Roots without any disease symptoms were classified as healthy.

### Plant Growth Stimulation

To test whether bacterial isolates were capable of stimulating plant growth, a pot experiment was conducted in the controlled plant growth chamber. Tomato seeds (Solanum lycopersicum. cv. Fuji Pink, Sakata, Japan) were surface-sterilized as described above. Surface-sterilized seeds were transferred to plastic Petri dishes and germinated for 4 days in a dark room at 25◦C. The bacterial isolates were grown overnight in TSB, and 1 ml of each culture was pelleted by centrifugation (10.000 × g for 10 min). Cell pellets were washed with 1 ml PBS, re-suspended in PBS and cell suspensions were adjusted to OD<sup>620</sup> nm = 0.1 (0.2 for Bacillus and Arthrobacter) that correspond to a cell density of about 107–10<sup>8</sup> cells ml−<sup>1</sup> . Germinated tomato seeds were placed in the bacterial suspension with a sterile forceps and shaken gently. After 10 min, the inoculated seeds were aseptically planted into a plastic pot filled with potting soil (N 250 mg l−<sup>1</sup> , P 120 mg l −1 , K 700 mg l−<sup>1</sup> , pH 6.0, Floragard GmbH, Germany) to a depth of approximately 1.5 cm. Non-inoculated plants were used as negative controls. Each experiment included six plants per treatment with three replications (total 18 plants) and pots were set-up in a randomized design. Plants were grown in a growth chamber under the conditions described above.

### Statistical Analyses

The data were subjected to one-way analysis of variance (ANOVA) in the software package SPSS-22 statistical software (SPSS, Inc., Chicago, IL, USA). Mean comparisons were conducted by the least significant difference (LSD) (P = 0.05) test.

### RESULTS

### The Antimicrobial Activity of Plant Extracts

The inhibitory effect of extracts from Z. capitata and H. perforatum, which were tested against diverse enteric pathogens (A. baumanii 60649, K. oxytoca 6653, K. pneumoniae 40602, K. aerogenes NCTC 8172, C. freundii 82073, S. aureus MRSA 16, E. faecalis NCTC 775, Proteus rettgeri NCIMB 9570, P. aeruginosa NCTC 6749, and E. coli NCTC 9001) at a concentration of 10 mg ml−<sup>1</sup> , resulted in different extents of inhibition (**Table 1**). The strains A. baumanii 60649, E. coli NCTC 9001, E. faecalis NCTC 775, K. oxytoca 6653, K. pneumoniae 40602, P. aeruginosa NCTC 6749, and S. aureus MRSA 16 were inhibited by the extract of H. perforatum. However, extract of Z. capitata did not exhibit any potential antibacterial activity against the 10 tested pathogens. Extracts of H. perforatum exhibited potential antifungal activity against F. oxysporum and A. alternata, whereas the extract of Z. capitata did not exhibit any inhibitory activity against the tested fungal strains.

### Enumeration, Isolation, and Identification of Endophytic Bacteria

The total number of endophytic bacterial isolates in the root tissue of Z. capitata was significantly higher (4.5 ± 0.8 × 10<sup>3</sup> CFU g−<sup>1</sup> of fresh root tissue) than in H. perforatum roots (2.6 ± 0.71 × 10<sup>3</sup> CFU g−<sup>1</sup> of fresh root tissue). Isolates were chosen randomly from the dilution plates exhibiting different colonial morphology, forms, texture, and color from each plate. A total of 18 bacterial isolates were derived from H. perforatum and 15 isolates from Z. capitata. Taxonomic investigation by MALDI-TOF MS revealed that the majority of strains were identified with secure genus identification and probable species identification (**Table 2**). The endophytes from the root of H. perforatum were affiliated with nine genera, whereas 14 isolates were identified at the species level. Achromobacter was the predominant genus, which was followed by the genus Pseudomonas. Furthermore, isolates affiliated with the genera Arthrobacter, Bacillus, Erwinia, Pantoea, Serratia, and


Stenotrophomonas were found. The most abundant species was Achromobacter piechaudii (S22a, S7a, S7) (**Table 2**). A total of five bacterial genera were isolated from the root of Z. capitata (**Table 2**). The most abundant isolates of Z. capitata were also identified as A. piechaudii (M11, M6, M31, M24, M41). Members of the genera Serratia, Stenotrophomonas, and Erwinia were not identified among the endophytes from Z. capitata .

### Beneficial Plant Traits of Endophytic Bacteria

All endophytes isolated from H. perforatum and Z. capitata were screened for multiple plant growth-promoting traits. Most of the bacterial isolates exhibited one or more plant growthpromoting activities. The production of phytohormone IAA by bacterial isolates is presented in **Table 2**. The highest level of IAA production was observed for Arthrobacter crystallopoietes S1 (19.8 µg ml − 1 ), A. piechaudii S7 (17.5 µg ml − 1 ), Achromobacter sp. S14 (15.2 µg ml − 1 ), Pantoea agglomerans S22 (12.3 µg ml − 1 ), and Bacillus cereus S40 (12.2 µg ml − 1 ), which were isolated from H. perforatum. Two isolates, Enterobacter cloacae M20 and P. agglomerans M13, isolated from Z. capitata also exhibited high IAA production in culture media (12.2 and 16.1 µg ml − 1 , respectively). The presence of tryptophan did not stimulate auxin production in the majority of the isolates, whereas only four isolates from Z. capitata revealed an increase in IAA synthesis: Bacillus sp. S2 (15.5 µg ml − 1 ), P. agglomerans S22 ( µg ml − 1 ) , Serratia liquefaciens S26 (15.0 µg ml − 1 ), Stenotrophomonas sp. S9 (16.3 µg ml − 1 ) and two isolates, B. altitudinis M9a (49.9 µg ml − 1 ) and P. agglomerans M13 (52.5 µg ml − 1 ) from H. perforatum (**Table 2**). All isolates isolated from H. perforatum , except Achromobacter spanius S23, Bacillus sp. S2 and isolate S9, were able to produce one or more cell wall-degrading enzymes. In contrast, only four isolates from Z. capitata (A. piechaudii M41, Achromobacter sp. M19, E. cloacae M20 and isolate M8) were able to produce proteases, and only one isolate, A. spanius M18, produced cellulase and β-1,3-glucanase. HCN was not produced by any isolate from Z. capitata, whereas seven isolates isolated from H. perforatum were able to produce HCN (**Table 2**).

Antagonistic activity was recorded for endophytes against plant pathogenic fungi F. oxysporum f. sp. radicis-lycopersici , F. solani , F. culmorum, G. graminis pv. tritici , A. alternata, and B. cinerea and the oomycete P. ultimum. As presented in **Table 3** , all isolates from H. perforatum exhibited antagonistic behavior to one or more of the tested plant pathogenic fungi. The isolates A. crystallopoietes S1, A. piechaudii S7, Pseudomonas koreensis S25, Pseudomonas pseudoalcaligenes S24, and Stenotrophomonas sp. S9 were highly effective against six fungal pathogens and exhibited the highest inhibition of mycelial growth (**Figure 1**). Among the isolates from Z. capitata, only P. agglomerans M13 exhibited antagonistic activity against five fungal pathogens, but the in vitro inhibition of the mycelium was lower than that of the other isolates. In general, H. perforatum, which exhibited a broad spectrum of antimicrobial activity, supported a higher proportion of antagonistic endophytes compared with Z. capitata .

TABLE 1 |

Antimicrobial

 activity of extracts obtained from

Hypericum

 perforatum

and

Ziziphora capitataa.


TABLE 2 | Matrix-assisted laser desorption ionization (MALDI) biotyper-based identification of culturable endophytic bacteria isolated from the root of Hypericum perforatum and Ziziphora capitata, and traits related to biocontrol and/or plant growth-promoting activity of bacterial strains.

<sup>∗</sup>++, score value 2.300–3.000, highly probable species identification; +, score value 2.000–2.299, secure genus identification and probable species identification; +, score value 1.700–1.999, probable genus identification; #Bacillus anthracis, B. cereus, B. mycoides, B. pseudomycoides, B. thuringiensis and B. weihenstephanensis are closely related and members of the Bacillus cereus group, which cannot be differentiated by MALDI−TOF MS.

### Biological Control and Plant Growth Promotion

The bacterial isolates that exhibited antagonistic activity against a wide range of fungal pathogens in vitro were selected to evaluate their ability to suppress tomato foot and root rot caused by F. oxysporum f. sp. radicis-lycopersici in a pot experiment. In non-infested soil, the portion of diseased plants was 2%, whereas in the presence of the pathogen, the portion of plants that exhibited disease symptoms increased to 38% (**Figure 2**). The selected antagonistic bacterial isolates A. crystallopoietes S1, Bacillus sp. S2, B. cereus S40, P. koreensis S25, S. liquefaciens S26, and Stenotrophomonas sp. S9, exhibited a statistically significant (P < 0.05) disease reduction (up to 9%) compared with Fusarium-infected control plants (**Figure 2**). Several isolates, namely A. piechaudii S7, Pseudomonas putida S19, Pseudomonas thivervalensis S5, P. pseudoalcaligenes S24, and P. agglomerans M13 reduced disease incident, but the effects were not significant.

The antagonistic endophytic bacterial isolates were also effective on the growth of tomato plants under controlled conditions (**Figure 3**). Statistical analysis showed that growth stimulatory effects of the isolates A. crystallopoietes S1, A. spanius S23, Bacillus sp. S2, P. putida S19, and Stenotrophomonas sp. S9 increased plant biomass significantly (P < 0.05) between 30 and 41%. However, four strains namely A. piechaudii S7, B. cereus S40, P. koreensis S25, and P. pseudoalcaligenes S24 reduced plant


TABLE 3 | Antagonistic activity of culturable endophytic bacterial isolates associated with Hypericum perforatum and Ziziphora capitata against soil-borne fungal pathogens.

+ Presence of antagonism; − absence of antagonism.

growth response, leading to a decrease in plant dry biomass between 5.3 and 11.4% (**Figure 3**).

### DISCUSSION

In our study, we analyzed the antimicrobial activity of plant extracts of H. perforatum and Z. capitata, and characterized plant beneficial traits of their associated culturable endophytic bacteria. Both parameters exhibited a relationship – Hypericum plant extracts exhibited greater antimicrobial activity and harbored a higher abundance of endophytes with antagonistic activity than Ziziphora, which lacks antimicrobial activity. In detail, H. perforatum was proved to possess potential antimicrobial activity against a wide range of pathogenic bacteria (A. baumanii, E. coli, E. faecalis, K. oxytoca, K. pneumoniae, P. aeruginosa, S. aureus) as well as fungi (F. oxysporum, A. alternata), whereas the extract of Z. capitata did not exhibit any inhibitory activity against the tested microbes. A similar observation for H. perforatum was reported by Maleš et al. (2006), who found that methanol extracts exhibited strong antibacterial activity against S. aureus, S. epidermidis, E. faecalis, and Bacillus subtilis.

In our study, we observed a lower number of endophytes in H. perforatum compared to Z. capitata that exhibited antibacterial activity. This is consistent with the report of Ahmed et al. (2014), who also reported a smaller microbial population in the rhizosphere of M. chamomilla, which possesses abundant antibacterial activity against pathogenic bacteria (Munir et al.,

2014). Our findings suggest that host plants differing in their antibacterial activity exhibited selective effects on physiological properties of endophytes. The understanding of interactions of endophytic bacteria with host plants includes the production of phytohormones, siderophores, and antifungal compounds, which have been well-documented previously by various authors (Berg et al., 2013, 2014; Cho et al., 2015; Egamberdieva et al., 2015a,b). Endophytic bacteria can also improve plant growth by protecting plants against soil-borne diseases or various environmental stresses (Berg et al., 2014; Cao et al., 2014). We have observed that endophytic bacteria associated with both investigated plants exhibited multiple plant beneficial activities, such as the production of IAA, HCN and cell-wall-degrading enzymes. Moreover, the endophytic bacteria associated with H. perforatum demonstrated higher antagonistic activity as compared with endophytes of Z. capitata. This observation is consistent with Gorluk et al. (2009), who also reported a higher proportion of antagonistic endophytes associated with

Chelidonium majus L., which is known for its antimicrobial potential (Zuo et al., 2008; Baker and Satish, 2013) against fungal pathogens. Furthermore, it is has been documented that endophytic microbes associated with medicinal plants may produce the same metabolites as their hosts and have been considered a potential source of biologically active metabolites (Mehanni and Safwat, 2010). For example, endophytic species (e.g., Pseudomonas, Bacillus) associated with Aloe vera exhibit antibacterial activity against human pathogenic bacteria, such as S. aureus, Streptococcus pyogenes, P. aeruginosa, and E. coli (Nejatzadeh-Barandoz, 2013), and produce bioactive compounds with antimicrobial activities (Akinsanya et al., 2015).

In our study, endophytic isolates which exhibited antagonistic activity against a wide range of fungal pathogens were evaluated for their capability to suppress tomato foot and root rot caused by F. oxysporum. All selected bacterial isolates of A. crystallopoietes S1, Bacillus sp. S2, B. cereus S40, P. koreensis S25, S. liquefaciens S26, and Stenotrophomonas sp. S9, exhibited statistically significant disease reduction compared with the Fusarium-infected control plants. These observations demonstrate the capability of endophytes to protect plants from soil-borne diseases. In accordance with these results, there is a report of the biological control of Verticillium wilt disease of cotton by endophytic bacteria B. subtilis KDRE 01 and B. megaterium KDRE 25, isolated from the medical plant Sophora alopecuroides (Lin et al., 2013). It has been also reported that Stenotrophomonas maltophilia which is an antagonist against Ralstonia solanacearum significantly suppressed potato brown rot in Egyptian clay soil (Messiha et al., 2007). Moreover, five isolates namely A. crystallopoietes S1, A. spanius S23, Bacillus sp. S2, P. putida S19, and Stenotrophomonas sp. S9 with antifungal activity exhibited enhancement of tomato growth. This finding is consistent with Wei et al. (2014), who also observed an enhanced growth of tomato plants by B. subtilis isolated from the rhizosphere of the traditional Chinese medicinal herb Trichosanthes kirilowii. In another study, endophytic bacteria isolated from a common weed Cassia occidentalis used in several traditional medicines, were able to produce IAA and stimulated growth of mung bean in pot experiments (Arun et al., 2012).

### CONCLUSION

The results from our pilot study of ongoing research provide insights about plant beneficial traits of culturable endophytic bacteria associated with the medicinal plants H. perforatum and Z. capitata with contrasting antimicrobial activities. We observed that H. perforatum with antibacterial activity supported more bacteria with antagonistic activity, as compared to Z. capitata. The antagonistic isolates were able to control tomato root rot caused by F. oxysporum under greenhouse conditions and could be a cost effective source for agro-based biological control agents. However, these findings indicate that further research is necessary to resolve the impact of medicinal plant species with contrasting antimicrobial activity on the endophytic microbial community in more detail, and to identify biological active compounds produced by the hosts and their endophytes.

### AUTHOR CONTRIBUTIONS

DE, SW, and GB did experimental design work. DE and UB conducted experiments. PA analyzed the data. DE, SW, UB, and GB wrote the manuscript. All authors read and approved the Manuscript.

### ACKNOWLEDGMENTS

The authors extend their appreciation to the Deanship of Scientific Research, College of Sciences Research Centre, King Saud University, Riyadh, Saudi Arabia for supporting the project. A Fellowship was provided to DE by the Alexander von Humboldt Foundation.

### REFERENCES

fmicb-08-00199 February 7, 2017 Time: 14:14 # 10



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer SPG and handling Editor declared their shared affiliation, and the reviewer SPG declared a past co-authorship with one of the authors UB to the handling Editor, who ensured that the process met the standards of a fair and objective review.

Copyright © 2017 Egamberdieva, Wirth, Behrendt, Ahmad and Berg. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Insights into the Regulation of Rhizosphere Bacterial Communities by Application of Bio-organic Fertilizer in Pseudostellaria heterophylla Monoculture Regime

Linkun Wu1,2† , Jun Chen1,2† , Hongmiao Wu1,2, Xianjin Qin2,3, Juanying Wang1,2 , Yanhong Wu1,2, Muhammad U. Khan1,2, Sheng Lin1,2, Zhigang Xiao1,2, Xiaomian Luo3,4 , Zhongyi Zhang3,4 and Wenxiong Lin1,4 \*

<sup>1</sup> College of Life Sciences, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> Key Laboratory of Crop Ecology and Molecular Physiology, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>3</sup> College of Crop Science, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>4</sup> Fujian Provincial Key Laboratory of Agroecological Processing and Safety Monitoring, Fujian Agriculture and Forestry University, Fuzhou, China

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Erik Limpens, Wageningen University and Research Centre, Netherlands Lianghui Ji, Temasek Life Sciences Laboratory, Singapore

\*Correspondence:

Wenxiong Lin lwx@fafu.edu.cn †These authors have contributed

### Specialty section:

equally to this work.

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 06 July 2016 Accepted: 25 October 2016 Published: 16 November 2016

#### Citation:

Wu L, Chen J, Wu H, Qin X, Wang J, Wu Y, Khan MU, Lin S, Xiao Z, Luo X, Zhang Z and Lin W (2016) Insights into the Regulation of Rhizosphere Bacterial Communities by Application of Bio-organic Fertilizer in Pseudostellaria heterophylla Monoculture Regime. Front. Microbiol. 7:1788. doi: 10.3389/fmicb.2016.01788 The biomass and quality of Pseudostellariae heterophylla suffers a significant decline under monoculture. Since rhizosphere miobiome plays crucial roles in soil health, deep pyrosequencing combined with qPCR was applied to characterize the composition and structure of soil bacterial community under monoculture and different amendments. The results showed compared with the 1st-year planted (FP), 2nd-year monoculture of P. heterophylla (SP) led to a significant decline in yield and resulted in a significant increase in Fusarium oxysporum but a decline in Burkholderia spp. Bio-organic fertilizer (MT) formulated by combining antagonistic bacteria with organic matter could significantly promote the yield by regulating rhizosphere bacterial community. However, organic fertilizer (MO) without antagonistic bacteria could not suppress Fusarium wilt. Multivariate statistics analysis showed a distinct separation between the healthy samples (FP and MT) and the unhealthy samples (SP and MO), suggesting a strong relationship between soil microbial community and plant performance. Furthermore, we found the application of bio-organic fertilizer MT could significantly increase the bacterial community diversity and restructure microbial community with relatively fewer pathogenic F. oxysporum and more beneficial Burkholderia spp. In conclusion, the application of novel bio-organic fertilizer could effectively suppress Fusarium wilt by enriching the antagonistic bacteria and enhancing the bacterial diversity.

Keywords: Pseudostellaria heterophylla, replant disease, bio-organic fertilizer, microbial community, deep pyrosequencing

### INTRODUCTION

Pseudostellaria heterophylla, belonging to the family Caryophyllaceae, is highly valued in traditional Chinese medicine. It provides cures for ailments including anorexia, spleen deficiency, and palpitations because of its various active components including saponins, polysaccharides, and cyclopeptides (Zhao W.O. et al., 2015). It is mainly produced in Ningde City, Fujian Province,

southeast China, which is known as a geo-authentic production zone with the most suitable soil and climate conditions for P. heterophylla. However, consecutive monoculture of this plant has led to a serious decline of biomass and quality of its underground tubers. Fields used for P. heterophylla cultivation can only be replanted once every 4 years (Wu H. et al., 2016; Wu L. et al., 2016). In addition, farmers generally apply a copious amount of pesticides and fertilizers to maintain production levels under this monoculture regime, but such application raises production costs, causes excessive pesticide residues, and leads to environmental pollution. In recent years, the market demand for P. heterophylla has forced farmers to plant the crop in fields outside of the geo-authentic production areas. However, P. heterophylla produced in these areas cannot be assured of quality because of the unsuitable environmental conditions (Wu et al., 2013). Therefore, it has become a priority to explore the mechanisms of consecutive monoculture problems, also known as replant disease, and strategies to effectively control these problems in P. heterophylla.

A growing body of evidence suggests that plant-microbe interactions play crucial roles in soil quality and crop health (Nunan et al., 2005; Lakshmanan et al., 2014; Macdonald and Singh, 2014). Xiong et al. (2015a) found that long-term consecutive monoculture of black pepper (Piper nigrum L.) led to a significant decline in soil bacterial abundance, especially in the Pseudomonas spp. Li et al. (2014) attributed the problems associated with consecutive monoculture in peanut to the changes in the structure of the soil microbial community induced by root exudates rather than to direct allelopathy. Wu L. et al. (2016) found that P. heterophylla monoculture can significantly increase the amount of Fusarium oxysporum in the rhizosphere, and root exudates could promote the growth of this soil-borne pathogen. Besides, previous studies demonstrated that a decrease in soil microbial diversity or the population size of antagonistic bacteria could result in the occurrence of soil-borne diseases (Latz et al., 2012; Shen et al., 2014). Mendes et al. (2011) used a PhyloChip-based metagenomic approach to compare the rhizospheric microbiomes of disease-suppressive and diseaseconducive soils and indicated that the abundance of particular bacterial taxa including Burkholderiaceae, Pseudomonadaceae, and Xanthomonadales was more abundant in suppressive soil than in conducive soil and was more abundant in transplantation soil (conducive soil + 10% suppressive soil) than in the conducive soil. Therefore, increasing attention has been paid to the roles of soil microbial ecology in causing and controlling of replant disease (Mazzola and Manici, 2012; Köberl et al., 2013; Philippot et al., 2013; Cha et al., 2016). Unfortunately, however, few studies have been carried out to understand the relationship between the soil bacterial community and replant disease of P. heterophylla, as well as the methods to overcome replant disease of this plant.

Manipulation of the rhizosphere microflora to favor beneficial microorganisms was considered to be an effective approach in suppressing the soil-borne pathogens and recovering the microbial populations damaged by pathogens (Babalola, 2010; Qiu et al., 2012). Nadhrah et al. (2015) found that formulated bio-organic fertilizer containing Burkholderia GanoEB2 could effectively suppress basal stem rot in oil palm caused by the fungal pathogen Ganoderma boninense. Pal et al. (2001) found that plant growth promoting rhizobacteria (Bacillus spp. and Pseudomonas sp. EM85) were strongly antagonistic to Fusarium moniliforme, Fusarium graminearum and Macrophomina phaseolina, causal agents of maize root diseases, and Bacillus spp. MRF was also found to produce IAA, solubilized tri-calcium phosphate and fixed nitrogen from the atmosphere. Qiu et al. (2012) indicated that a bio-organic fertilizer which was a combination of manure compost with antagonistic microorganisms (Bacillus sp., Paenibacillus sp., etc.) was an effective approach to suppress Fusarium wilt of cucumber plants by regulating the microbial community of rhizosphere soil. Previous studies demonstrated that beneficial microorganisms could better colonize plant roots and protect plants from soil-borne diseases when they were applied to the soil with organic matter (OM) or manure (El-Hassan and Gowen, 2006; Yang et al., 2011; Yuan et al., 2014). The application of OM or manure can enhance the suppression of soil-borne diseases possibly by providing nutrients for the growth of beneficial microorganisms, by regulating the physicochemical property in the soil, or by inducing changes in the structure of the soil microbial community (Gorissen et al., 2004; Shen et al., 2014; Wen et al., 2015). Therefore, in this study, several beneficial bacteria (Burkholderia spp., Bacillus spp.) that have shown strong antagonistic activities against F. oxysporum were used as inocula to fortify organic fertilizer for the purpose of suppressing P. heterophylla Fusarium wilt (Supplementary Figure S1). In addition, our previous studies found that phenolic acids in the root exudates of medicinal plants mediated the increase of soil-borne pathogens (Wu H. et al., 2016; Wu L. et al., 2016). Some Bacillus spp. with a capacity of phenolic acid degradation (Supplementary Figure S2) were selected as the experimental strains for bio-organic fertilizer fermentation in order to degrade the phenolic acids released by P. heterophylla.

In the current study, we examined the shifts in bacterial communities under different soil amendments by using deep pyrosequencing combined with DNA barcoding. The objective of this investigation is to consider in detail the changes in diversity and composition of soil bacterial communities under P. heterophylla monoculture and explore the efficiency and underlying mechanisms of a novel bio-organic fertilizer in controlling replant disease of this plant.

### MATERIALS AND METHODS

### Fertilizer Preparation and Application

The microbial fertilizer NO.2 used in this study was a mixture of the OM, the effective microorganisms and antagonistic bacteria (Burkholderia spp. including Burkholderia sp. 4, Burkholderia sp. 8; Bacillus spp. including Bacillus pumilus, Bacillus cereus, etc.), which were previously isolated and stored in our laboratory. In particular, Burkholderia sp. 4 and Burkholderia sp. 8 possess the prnD gene (KX885433, KX894563), responsible for the production of the antifungal compound pyrrolnitrin (PRN), and show very strong antagonistic activities

against F. oxysporum, an agent known to cause wilt and rot disease of P. heterophylla (Supplementary Figure S1). Bacillus pumilus was found to promote plant growth (unpublished data). Besides, some Bacillus spp. including Bacillus cereus, Bacillus subtilis with a capacity of phenolic acids degradation (Supplementary Figure S2) were used as inocula to ferment OM for preparing the novel bio-organic fertilizer. The effective microorganisms and antagonistic bacteria were incubated separately in the liquid-state fermentation and then mixed together for solid-state fermentation. The OM consists of soybean meal and fish meal with 2:1 weight ratio. After solid-state fermentation using the OM as a substrate (30% moisture content) at 37◦C for 48 h and then at room temperature for 4 weeks, the microbial fertilizer contained 1.0 × 10<sup>9</sup> ∼ 5.0 × 10<sup>9</sup> CFU g−<sup>1</sup> of antagonistic bacteria. The fertilizer NO.1 was just the OM with the same fermentation process as microbial fertilizer NO.2 except the beneficial microbial inocula. When treating the field plots, the solidstate fermented fertilizer was completely mixed with soils, flooded with water and then mulched with black film for 1 month.

### Experiment Design and Soil Sampling

The P. heterophylla 'Zheshen 2,' a cultivar widely planted in the geo-authentic production zones, was selected as the experimental material. The experiment was conducted at Xiapu County, Ningde City, Fujian Province, China (27◦ 08<sup>0</sup> N, 119◦ 88<sup>0</sup> E), with an annual mean precipitation of 1550 mm and annual mean temperature of 16◦C. A field previously cultivated with rice (Oryza sativa) was used for experiments. The experimental design and sampling time are shown in **Figure 1**. More specifically, CK, FP and SP represent the control with no P. heterophylla cultivation, the 1st-year planted and 2nd-year monocultured plots, respectively. AMO, AMT represent the 1-year cultivated plots that treated with equal amounts of fertilizers NO.1, NO.2 for 1 month. NMF represent non-microbial fertilizer, namely the 1-year cultivated plots without microbial fertilizer treatment and was sampled at the same time as AMO and AMT. MO and MT represent the plots treated with microbial fertilizers NO.1 (Microbial fertilizer NO. ONE, MO) and NO.2 (Microbial fertilizer NO. Two, MT) for 7 months. Each treatment had three replicate plots and the study plots were completely randomized. To ensure the accuracy and repeatability, all treatments were organized within a single field site with the same climatic conditions and subjected to the same fertilization protocol and field management during the whole experimental period.

After 1 month of microbial fertilizer application (November 14, 2014), soil physical and chemical properties (temperature, moisture, electrical conductivity, pH, and redox potential) were detected in situ by using HH2 Moisture Meter (Delta-T Devices, Ltd, England), pH meter (Spectrum Technologies, Inc., USA) and ORP Depolarization Automatic Analyzer (Nanjing Chuan-Di Instrument & Equipment Co., Ltd, China). And then the soils were sampled (NMF, AMO, and AMT) from five random locations for each treatment. At the expansion stages of tuberous roots (May 15, 2015), rhizosphere soils of P. heterophylla (FP, SP, MO, and MT) as well as control soils (fallow, CK) were sampled because of the pronounced difference in growth status between different treatments on this date (**Figure 2**). The rhizosphere soil tightly attached to tuberous roots of P. heterophylla was brushed off and collected. The collected soils were sieved through 2 mm mesh and immediately used for total soil DNA extraction. The rest of the soil samples were air-dried and used to determine soil nutrients (Wu et al., 2013). Briefly, available nitrogen (AN) was determined by alkaline hydrolysis method and total N (TN) was measured by Kjeldahl method. Available phosphorus (AP) was extracted using 0.5 mol/L NaHCO<sup>3</sup> and then determined by Mo–Sb colorimetry method. Available potassium (AK) was extracted using CH3COONH<sup>4</sup> and then measured by flame atomic absorption spectrometry. The total P and K (TP

and TK) was calculated by first digesting the soil using the H2SO4–HClO<sup>4</sup> and then measuring the level as described for AP and AK.

### DNA Extraction and Deep Pyrosequencing

For each soil sample, the extraction of total soil DNA was carried out in three replicates using a BioFast soil Genomic DNA Extraction kit (BioFlux, Hangzhou, China) following the manufacturer's instructions. The DNA quality was monitored on 1% agarose gels. The DNA concentration was determined using a Nanodrop 2000C Spectrophotometer (Thermo Scientific, USA). Variable regions 3 to 4 (V3– V4) of bacterial 16S rRNA gene were amplified with the specific primers 341F (5<sup>0</sup> -CCTAYGGGRBGCASCAG-3<sup>0</sup> ) and 806R (5<sup>0</sup> -GGACTACNNGGGT ATCTAAT-3<sup>0</sup> ). All polymerase chain reaction (PCR) reactions were carried out using Phusion <sup>R</sup> High-Fidelity PCR Master Mix (New England Biolabs).

The PCR products of each sample were pooled in equimolar concentrations and purified using a Qiagen Gel Extraction Kit (Qiagen, Germany). Sequencing libraries were generated with TruSeq <sup>R</sup> DNA PCR-Free Sample Preparation Kit (Illumina, USA) following the manufacturer's instructions, and then sequenced on an Illumina HiSeq2500 platform.

### Operational Taxonomic Unit (OTU)-Based Sequence Analysis

After pyrosequencing, raw sequences were assigned to the individual sample according to the unique barcode. The lowquality tags were excluded according to the QIIME (V1.7.0) quality-controlled process (Caporaso et al., 2010). Next, the chimera sequences were removed via the reference database (Gold database<sup>1</sup> ) based on the comparison using the UCHIME algorithm (Edgar et al., 2011). Finally, the obtained effective tags were clustered into operational taxonomic units (OTUs) with a cutoff of 97% similarity. The singleton OTUs that contain only one sequence in all 12 samples were removed before further analysis. The remaining sequences were taxonomically classified using Ribosomal Database Project (RDP) classifier (Version 2.2) (Wang et al., 2007) via the GreenGene Database<sup>2</sup> .

### Statistical Analyses of Pyrosequencing Data

After data normalization, alpha and beta diversities were calculated to analyze the complexity of species diversity within a sample and differences between samples, respectively (Xiong et al., 2015b). An OTU-based analysis was carried out to calculate the following five indices (alpha diversity): observed-species (Sobs), Chao1, ACE (Abundance-based Coverage Estimator), Shannon and Simpson diversity indices (Sun et al., 2014). Chao1 and ACE were calculated to estimate the richness of each sample (Chao, 1984; Shen et al., 2014). Shannon and Simpson indices were calculated to estimate the diversity within each individual sample (Lemos et al., 2011; Chen et al., 2012). A weighted UniFrac distance based on the abundance of lineages was calculated to analyze the differences in overall bacterial community composition and structure (beta diversity). Principal coordinate analysis (PCoA) and the unweighted pair-group method with arithmetic means (UPGMA) clustering was performed based on the weighted UniFrac distance.

One-way analysis of variance (ANOVA) followed by the Tukey's test (P < 0.05, n = 3) was carried out for multiple comparisons. Spearman correlation coefficients between dominant bacterial taxa and yield of tuberous roots were calculated using SPSS v20.0 (SPSS, Inc., USA). Analysis of similarity (ANOSIM) was used to examine the statistical significance between samples. Similarity percentage (SIMPER) analysis was conducted to assess the relative contribution (%) of each taxon to the dissimilarity between samples. Both ANOSIM and SIMPER analyses were carried out in three replicates by using the PRIMER V5 software package (PRIMER-E, Ltd, UK) (Rees et al., 2004).

### Quantitative PCR for Genus Burkholderia, F. oxysporum and prnD Gene

Quantitative PCR (qPCR) was carried out to quantify the Burkholderia, F. oxysporum and prnD gene in different soil samples by using the taxon-specific primers (Supplementary Table S1). The reaction mixture (15 µl) for qPCR consists of 7.5 µl 2× SYBR green I SuperReal Premix (TIANGEN, Beijing, China), 0.5 µl of each primer (10 µM) and template DNA (20∼40 ng of total soil DNA or a serial dilution of plasmid DNA

<sup>1</sup>http://drive5.com/uchime/uchime\_download.html

<sup>2</sup>http://greengenes.lbl.gov/cgi-bin/nph-index.cgi

for standard curves). Three independent quantitative PCR assays were performed for each treatment.

### RESULTS

### Effects of Consecutive Monoculture and Bio-organic Fertilizer Application on the Growth of P. heterophylla

The yields of P. heterophylla tuberous roots were detected under consecutive monoculture and bio-organic fertilizer application. Consecutive monoculture of P. heterophylla (SP) led to a serious decline in the above- and below-ground biomass, and an increase in pest and disease problems. When the 1-year cultivated plots were treated with the microbial fertilizer NO.2 (MT), the growth status of P. heterophylla greatly improved as compared with the 2nd-year monocultured plots (SP). However, the fertilizer NO.1 (MT) could not increase the yield of P. heterophylla tuberous roots. The fresh weights of tuberous roots in the 1st-year planted plots (FP), MT, SP, and MO were 656.3 kg/666.7 m<sup>2</sup> , 661.1 kg/666.7 m<sup>2</sup> , 364.3 kg/666.7 m<sup>2</sup> , and 400.7 kg/666.7 m<sup>2</sup> , respectively (**Figure 2**).

### Soil Physical and Chemical Properties of Different Treatments

One month after an application of microbial fertilizers, soil physical and chemical properties were detected in situ. The results showed that soils amended with microbial fertilizers for 1 month (AMO and AMT) had a higher level of moisture, pH, and electrical conductivity, especially for AMT. NMF had the lowest level of pH. Moreover, the fallow plots had the highest redox potential (ORP), while the soils amended with microbial fertilizer NO.2 had the lowest ORP value (**Figure 3**). The results suggest that application of microbial fertilizer produced by antagonistic bacteria could change the soil physical and chemical properties in a monoculture regime.

Furthermore, analysis of soil nutrients at the expansion stage of tuberous roots showed that the contents of available nitrogen (AN), available phosphorus (AP), and available potassium (AK) were significantly higher in SP than in FP and MT. Total nitrogen (TN) was significantly higher in FP and MT than in SP, while

total phosphorus (TP) was significantly higher in SP than in FP and MT. Except for the AK and total potassium (TK), the rest of the soil nutrients were significantly higher in MO than in MT. TN, AN, and TK were significantly higher in FP than in MT while there was no significant difference in TP, AP, and AK between FP and MT (**Table 1**).

### OTU Cluster and Species Annotation

Deep 16S rRNA pyrosequencing was applied to assess the responses of the soil bacterial community to bio-organic fertilizer application in a P. heterophylla monoculture regime. Across all soil samples, a total of 1,233,132 effective tags with a species annotation were obtained, with an average of 54,502 effective tags per sample. In total, 74,900 singletons, accounting for 5.7% of total tags, were removed from the dataset before further analysis (Supplementary Figure S3). Rarefaction curves demonstrated that the number of observed OTUs tended to reach a plateau at 40,000 sequences (**Figure 4**). At a 97% sequence similarity cut-off, we obtained a sum of 62,464 OTUs across the 24 soil samples. The OTU numbers in CK, NMF, AMO, AMT, FP, SP, MO, and MT plots were 3,452, 3,482, 2,119, 2,269, 2,607, 2,281, 2,200, and 2,410, respectively (Supplementary Figure S3). On average, we were able to classify about 99.7% of effective sequences at the phylum level and 75.3% at the family level, but only 34.2% at the genus level (Supplementary Figure S4).

### Alpha Diversity Indices

Alpha diversity was calculated to analyze the complexity of species diversity within a sample. Among of them, the Chao1 estimator and the abundance-based coverage estimator (ACE) were calculated to estimate the richness of each sample. Both the Shannon and Simpson indices were calculated to estimate the diversity within each individual sample. In this study, the alpha diversity indices of soil bacterial community were calculated with a cutoff of 44,990 sequences. The bacterial community showed a significantly higher Shannon's diversity index in FP than in SP (P < 0.05), and higher richness and Simpson's indices (not significant, P > 0.05). Both the Shannon and Simpson diversity indices were significantly higher in MT than in MO (P < 0.05). The richness indices including the observed species, Chao1 and ACE indices, were higher in MT than in MO, but the difference was not significant (P > 0.05). No significant difference was observed in the Shannon and Simpson diversity indices between AMO and AMT (**Table 2**).

### Beta Diversity Indices

Beta diversity was calculated to analyze the differences in overall bacterial community composition and structure between samples. In this study, the weighted Unifrac distance between FP and MT was a minimum of 0.101 (the purple outline), suggesting a high similarity of the soil bacterial community between these two healthy samples. However, a relatively great dissimilarity was observed between AMO and other treatments and between AMT and other treatments (the orange outline). The similarity among

FIGURE 4 | Rarefaction curves of bacterial communities based on observed OTUs at 97% sequence similarity for individual samples. CK, FP, and SP represent the control with no P. heterophylla cultivation, the 1st-year planted and 2nd-year monocultured plots, respectively. AMO, AMT represent the 1-year cultivated plots that treated with equal amounts of microbial fertilizers NO.1, NO.2 for 1 month, respectively. NMF represents the 1-year cultivated plots without microbial fertilizer treatment and was sampled at the same time as AMO and AMT. MO and MT represent the plots treated with microbial fertilizers NO.1 and NO.2 for 7 months, respectively. The numbers followed by the treatments represent the three replicates.


CK, FP, and SP represent the control with no P. heterophylla cultivation, the 1st-year planted and 2nd-year monocultured plots, respectively. MO and MT represent the plots treated with microbial fertilizers NO.1 and NO.2 for 7 months, respectively. TN, AN, TP, AP, TK, and AK represent total nitrogen, available nitrogen, total phosphorus, available phosphorus, total potassium, available potassium, respectively. Different letters in columns show significant differences determined by Tukey's test (P ≤ 0.05, n = 3).



CK, FP, and SP represent the control with no P. heterophylla cultivation, the 1st-year planted and 2nd-year monocultured plots, respectively. AMO, AMT represent the 1-year cultivated plots that treated with equal amounts of microbial fertilizers NO.1, NO.2 for 1 month, respectively. NMF represents the 1-year cultivated plots without microbial fertilizer treatment and was sampled at the same time as AMO and AMT. MO and MT represent the plots treated with microbial fertilizers NO.1 and NO.2 for 7 months, respectively. Different letters in columns show significant differences determined by Tukey's test (P ≤ 0.05, n = 3).

SP, NMF, and MO was high, as indicated by the low weighted Unifrac distances among each other (the red outline) (**Figure 5**).

### PCoA and UPGMA Clustering

To compare bacterial community structures across all samples, PCoA and UPGMA clustering were performed based on the weighted UniFrac distance. The PCoA analysis based on the weighted Unifrac distance revealed distinct differences in soil bacterial community structure between different treatments. The first two components (PC1 and PC2) of PCoA explained 61.77 and 15.20% of the total bacterial community variations, respectively. The bacterial communities of the MO-treated (MO) soil and the 2nd-year monoculture soil (SP) (both unhealthy samples) formed a separate group. However, the bacterial

communities of the MT-treated soil (MT) and the 1st-year cultivated soil (FP) (both healthy samples) formed a separate group, which were distinctly different from those in the unhealthy samples (**Figure 6A**).

Furthermore, UPGMA clustering showed similar bacterial community structure for the same treatment in triplicate and obvious differences between different treatments, as observed in the three highly supported clusters (clusters I, II, and III). Bacterial community structure from FP, MT, and CK were clustered together (cluster I) and were separated from the soil samples from SP, MO, and NMF, which were grouped together (cluster II). The bacterial community structure from soil samples that were amended with microbial fertilizers NO.1 and NO.2 for 1 month clustered together (cluster III), which was distinguished from cluster I and cluster II (**Figure 6B**). The results from PCoA and UPGMA cluster analyses indicated that microbial fertilizer NO.2 could repair the imbalance in the bacterial community in replanted soils, but the microbial fertilizer NO.1 could not create a healthy bacterial community structure, although, it altered the bacterial community after 1 month of treatment.

### Venn Diagram Analysis

Venn diagram analysis was carried out to detect the exclusive and shared OTUs between the healthy (FP and MT) and unhealthy samples (SP and MO). The results showed that the number of OTUs exclusively found in FP was 483 (13.3%). The number of OTUs exclusively found in SP was 443 (12.9%). The number of OTUs only shared in FP and MT was 269 (6.0%), and that only shared in SP and MO was 137 (3.1%). Moreover, high similarity was observed between the pie charts A and B and between the pie charts C and D. The percentage of OTUs shared in FP, SP, MO, and MT was 32.6% (1,767 species), and these were mainly assigned to the class Proteobacteria (52.1%), Acidobacteria (11.5%), and Actinobacteria (7.4%) (**Figure 7**). The abundance of these 1,767 OTUs shared in the healthy and unhealthy samples accounted for up to 93.9, 96.4, 95.1, and 96.1% of the total detected taxa in FP, MT, SP, and MO plots, respectively.

### Shifts in Soil Bacterial Community Structure under Consecutive Monoculture

The taxonomy of OTUs in each sample was assigned using the Ribosomal Database Project (RDP) classifier via the GreenGene Database. The results showed that the bacterial OTUs were comprised mainly of 10 phyla, Proteobacteria, Acidobacteria, Bacteroidetes, Firmicutes, Actinobacteria, Chloroflexi, Crenarchaeota, Gemmatimonadetes, Nitrospirae, and Verrucomicrobia. Proteobacteria was the dominant microbial taxa, and accounted for approximately 50% of the total population in each sample. Compared with NMF, those bacteria belonging to Firmicutes and Bacteroidetes significantly increased after bio-organic fertilizer application (MO and MT), while Acidobacteria, Chloroflexi, and Gemmatimonadetes significantly decreased. When the treated soils were planted with P. heterophylla under aerobic conditions, those bacteria belonging to Acidobacteria and Chloroflexi significantly increased, but Bacteroidetes and Firmicutes significantly decreased (Supplementary Figure S5).

Analysis of similarity of the pyrosequencing data consisting of the relative abundance of genera (logarithmic transformation and standardization) showed that the bacterial communities differed significantly between eight different treatments (ANOSIM Global

FIGURE 6 | Principal coordinate analysis (PCoA) (A) and hierarchical clustering (B) of bacterial communities based on weighted Unifrac algorithm for different soil samples. CK, FP, and SP represent the control with no P. heterophylla cultivation, the 1st-year planted and 2nd-year monocultured plots, respectively. AMO, AMT represent the 1-year cultivated plots that treated with equal amounts of microbial fertilizers NO.1, NO.2 for 1 month, respectively. NMF represents the 1-year cultivated plots without microbial fertilizer treatment and was sampled at the same time as AMO and AMT. MO and MT represent the plots treated with microbial fertilizers NO.1 and NO.2 for 7 months, respectively. The numbers followed by the treatments represent the three replicates.

R = 0.977, P = 0.001). ANOSIM analysis based on the presence/absence of genera also showed a significant difference among eight treatments in bacterial communities (ANOSIM Global R = 0.901, P = 0.001). Furthermore, similarity percentage analysis (SIMPER) analysis using the relative abundance of genera (logarithmic transformation and standardization) showed that the pairwise dissimilarity in bacterial communities between MO and MT was 26.57%, and was 28.34% between FP and SP. The top shared genera that contributed approximately 60% to the observed differences between MO and MT and between FP and SP included Hylemonella, Candidatus koribacter, Dokdonella, Dyella, Rhodoplanes, Burkholderia, Kaistobacter, and Clostridium (**Table 3**). Moreover, the relative abundances of Burkholderia, Kaistobacter, and Clostridium were significantly (P < 0.01) positively correlated with the yield of P. heterophylla tuberous roots (**Table 3**). Compared with the 2nd-year monocultured (SP) and microbial fertilizer NO.1 (MO) treatments, the relative abundance of genus Burkholderia in the microbial fertilizer NO.2 (MT) treatment increased by 453.7 and 116.0%, respectively.

### Quantification of F. oxysporum, Genus Burkholderia, and the prnD Gene

The amounts of genus Burkholderia, potential biocontrol agents, and F. oxysporum, the main agent causing rot disease of this plant were detected by quantitative PCR. The results showed that the relative abundance of F. oxysporum was significantly greater in the SP and MO (unhealthy soils) than in the FP and MT (healthy soils) while the abundance of Burkholderia was significantly lower in the SP and MO (unhealthy soils) than in the FP and MT (healthy soils) (**Figure 8**). The results were consistent with the deep pyrosequencing analysis (**Table 3**). However, application of microbial fertilizer NO.2 produced


TABLE 3 | Top shared genera with 60% cumulative contribution to the dissimilarity between MO and MT, and between FP and SP.

Similarity percentage (SIMPER) analysis was conducted to assess the relative contribution (%) of each taxon to the dissimilarity between samples (MO and MT, FP and SP) based on the relative abundance of each genus (logarithmic transformation, standardization). FP and SP represent the 1st-year planted and 2nd-year monocultured plots, respectively. MO and MT represent the plots treated with microbial fertilizer NO.1 and NO.2 for 7 months. § Different letters within a line show significant differences determined by Tukey's test (P ≤ 0.05, n = 3). #Spearman correlation coefficients between the abundance of bacterial taxa and the yield of P. heterophylla tuberous roots. ∗∗ mean P < 0.01, respectively.

by antagonistic bacteria increased the amounts of beneficial Burkholderia spp. in the rhizosphere. Besides, quantitative PCR of prnD, a gene responsible for the production of the antibiotic pyrrolnitrin, showed that the population harboring prnD gene was significantly greater in the healthy soils (FP and MT) than in the unhealthy soils (SP and MO) (**Figure 8**).

### DISCUSSION

The consecutive monoculture problem, also known as replant disease, commonly occurs in the cultivation of a range of crops in intensive agriculture, especially in the production of medicinal plants. Approximately 70% of medicinal plant species using tuberous roots have various degrees of consecutive monoculture problems (Zhang and Lin, 2009), which substantially limits the development of traditional Chinese medicines. Our field experiment showed that consecutive monoculture of P. heterophylla resulted in a significant decline in the yield of tuberous roots (**Figure 2**). Many factors have been thought to be responsible for the consecutive monoculture problem of P. heterophylla, including soil physical and chemical properties, accumulation of allelochemicals released by roots, and shifts in the soil microbial community (Wu L. et al., 2016). In this study, we found that the content of most of the soil nutrients were significantly higher in the unhealthy samples (SP and MO) than in the healthy samples (NP and MT) (**Table 1**). Moreover, our study showed the utilization of OM without antagonistic bacteria (MO) in monocultured plots could improve soil nutrients but not effectively eliminate replant disease of this plant. However, the application of microbial fertilizers could significantly increase the pH value and soil electrical conductivity. The monocultured plots without microbial fertilizer treatment (NMF) were found to have the lowest pH (**Figure 3**). Previous studies have shown that consecutive monoculture of many plants could lead to a significant decrease in soil pH (Meriles et al., 2009; Wu et al., 2015). Many pathogenic fungi including Fusarium spp. and Pythium spp. are adapted to a slightly low pH (Husson, 2013). Both soil pH and electrical conductivity were recognized as important factors affecting microbial processes and ecology in soil (Gelsomino and Cacco, 2006; Hogberg et al., 2007). Additionally, application of microbial fertilizer NO.2 could significantly increase the soil electrical conductivity but decrease the soil ORP (**Figure 3**). OM was reported to play a dominant role in affecting soil ORP, leading to a lowering of soil ORP when soil OM increased (Husson, 2013). Many researchers have indicated that the development of several plant pathogens (Fusarium spp. and Rhizoctonia solani) were closely related to a high soil ORP (Takehara et al., 2004; Husson, 2013). Some agricultural practices used to reduce soil ORP including the application of decomposable organic material and flooding or covering the soil with plastic could effectively control soil-borne pathogens (Blok et al., 2000; Mowlick et al., 2012; Wen et al., 2015). Wen et al. (2015) found that biological soil disinfestations (BSD) conducted by incorporating organic amendments into the soil under flooding conditions could rapidly decrease the soil ORP, which would contribute to pathogen inactivation. The results mentioned above suggested that the soil physical and chemical properties may indirectly influence P. heterophylla performance by modulating the soil microbial community structure.

Since researchers indicated that allelochemicals released by roots were not sufficient to directly inhibit neighboring plants or the host plant, more attention has been paid to the belowground microbial community structure and function diversity mediated by rhizodeposition (Berendsen et al., 2012; Chaparro et al., 2014; Haney and Ausubel, 2015; Xiong et al., 2016). In this study, deep pyrosequencing combined with the qPCR technique was used to assess the changes in the microbial community under monoculture and different amendments. The results demonstrated that consecutive monoculture of P. heterophylla led to a significant variation in the bacterial community structure in the rhizosphere, but the OM fermented

by antagonistic bacteria (MT) could recover the bacterial community composition (**Figure 5**). It was found that the application of bio-organic fertilizer could significantly increase those bacteria belonging to Firmicutes and Bacteroidetes but significantly decrease Acidobacteria (Supplementary Figure S5), which was in accordance with previous reports (Mowlick et al., 2012; Huang et al., 2015). Furthermore, both PCoA analysis and UPGMA clustering showed a distinct separation between the healthy samples (FP and MT) and the unhealthy samples (SP and MO), suggesting the strong relationship between soil microbial community and plant performance (**Figure 6**). In addition, our findings showed that the healthy samples (FP and MT) had significantly higher bacterial community diversity than the unhealthy samples (SP and MO) (**Table 2**). A high microbial diversity was considered to be responsible for the development and maintenance of disease suppressive soils (Van Bruggen and Semenov, 2000; Qiu et al., 2012; Yuan et al., 2014). Many researchers showed that organic amendments could induce general suppression in the soil by increasing soil microbial activity, enhancing competition for nutrients and ecological niches available for soil-borne pathogens (Kotsou et al., 2004; Shen et al., 2014).

Moreover, we found that consecutive monoculture of P. heterophylla led to a significant increase in the relative abundance of F. oxysporum but a decrease in the potential beneficial bacteria including Burkholderia and Clostridium (Mowlick et al., 2012; Ho et al., 2015; Wen et al., 2016). Our previous studies by culture-dependent and culture-independent analysis demonstrated that F. oxysporum and other Fusarium spp. are the main causative agents of wilt and rot disease of this plant (Zhao Y. et al., 2015; Wu L. et al., 2016). Our previous studies have demonstrated that the phenolic acids identified in the root exudates of P. heterophylla could significantly inhibit the growth of beneficial Burkholderia sp. but promote the growth of soil-borne F. oxysporum (Lin et al., 2015; Wu L. et al., 2016). However, bio-organic fertilizer, in particular, MT containing antagonistic bacteria could rebalance the beneficial and detrimental microbial residents in soil, leading to relatively more beneficial microorganisms and relatively fewer pathogenic microorganisms (**Figure 8**; **Table 3**). Compared with the 2ndyear monocultured (SP) and microbial fertilizer NO.1 (MO) treatments, the relative abundance of genus Burkholderia in the microbial fertilizer NO.2 (MT) treatment increased by 453.7 and 116.0%, respectively. The increase might be mainly due to the organic amendments and the addition of Bacillus spp. with a capacity of degradation of phenolic acids into bioorganic fertilizer. The application of OM can provide nutrients for the growth of beneficial microorganisms and regulate the physicochemical property in the soil, which may be favorable for the growth of beneficial microorganism and contribute to pathogen inactivation (Husson, 2013; Shen et al., 2014; Wen et al., 2015). In this study, we also found that the application of OM could significantly increase soil pH but decrease soil redox potential (ORP) (**Figure 3**). However, it should be noted that we were only able to classify approximately 4.9% of the effective sequences at the species level based on the pyrosequencing data of hypervariable regions 3–4 (V3–V4, 420 bp) of the bacterial 16S rRNA gene (Supplementary Figure S4). The Burkholderia spp. added in the bio-fertilizer in this study could not been identified at the species level. Therefore, more studies on species-level classification and analysis are needed.

Biological control is considered to be the most promising technique for plant disease prevention by favoring beneficial microorganisms that are directly antagonistic to root pathogens (Qiu et al., 2012; Shen et al., 2014). A growing number of Burkholderia species have been reported to be beneficial for plants, including biological control of fungal diseases, promotion of plant growth and induction of systemic resistance (Compant et al., 2008; Ho et al., 2015). The Burkholderia strains used for solid fermentation in this study proved to have a strong antagonistic activity against F. oxysporum (Supplementary Figure S1), the main agent causing replant disease of P. heterophylla (Wu L. et al., 2016). Therefore, the mechanism underlying the reduction of replant disease by the application of Burkholderiafortified organic fertilizer might be attributed not only to general suppression but also to specific suppression referring to an increase in the specific microorganisms with antagonistic activity against the specific pathogen (Kotsou et al., 2004). Pyrrolnitrin (PRN), a chlorinated phenylpyrrole antibiotic catalyzed by the

prnD gene product, is produced by many strains of Burkholderia spp. (de Souza and Raaijmakers, 2003). In this study, quantitative PCR analysis of the prnD gene showed that consecutive monoculture of P. heterophylla led to a significant decline in the population harboring prnD gene. In contrast, the application of bio-organic fertilizer could significantly increase the amount of prnD gene in the monocultured soils (**Figure 8**). Latz et al. (2012) found that soil suppressiveness against pathogens by fostering beneficial bacterial communities was at maximum when the abundance of PRN and 2,4-diacetylphloroglucinol (DAPG) producer was high in soil.

### CONCLUSION

Consecutive monoculture of P. heterophylla can alter the microbial community in the rhizosphere, including enrichment of host-specific pathogens at the expense of beneficial microorganisms. However, application of bio-organic fertilizer could restructure the microbial community in the rhizosphere and effectively suppress Fusarium wilt by enriching the antagonistic bacteria and enhancing the bacterial diversity. Our results give a promising strategy for replant disease control by manipulating the rhizosphere microbiome to keep soil healthy and enhance P. heterophylla production and sustainability.

### AUTHOR CONTRIBUTIONS

WL and LW conceived the study; LW, WL, JC, and MK wrote the paper. LW, JC, HW, XQ, JW, XL, and ZX performed experiments;

### REFERENCES


LW, JC, and SL performed the statistical analyses; ZZ, YW, and JC are involved in field management and soil sampling. All authors discussed the results and commented on the manuscript.

### FUNDING

This work was supported by grants from the National Natural Science Foundation of China (Grant no. 81303170, U1205021, 81573530, and 31401950), the Outstanding Youth Scientific Fund of Fujian Agriculture and Forestry University (Grant no. XJQ201501) and the Health and Family Planning Program of Fujian Province (Grant no.WKJ-FJ-34).

### ACKNOWLEDGMENTS

We thank National Natural Science Foundation of China and the Health and Family Planning Program of Fujian Province for providing the funds used in this work. This work was supported by Fujian-Taiwan Joint Innovative Center for Germplasm Resources and Cultivation of Crop (FJ 2011 Program, NO. 2015-75).

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01788/full#supplementary-material


soil disinfestation with incorporation of various organic matters. Eur. J. Plant Pathol. 143, 223–235. doi: 10.1007/s10658-015-0676-x


plants by regulating microbial community of rhizosphere soil. Biol. Fert. Soils 48, 807–816. doi: 10.1007/s00374-012-0675-4


fmicb-07-01788 November 14, 2016 Time: 12:50 # 13


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Wu, Chen, Wu, Qin, Wang, Wu, Khan, Lin, Xiao, Luo, Zhang and Lin. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Plant Growth Promoting Bacteria Associated with Langsdorffia hypogaea-Rhizosphere-Host Biological Interface: A Neglected Model of Bacterial Prospection

Érica B. Felestrino1,2, Iara F. Santiago<sup>3</sup> , Luana da Silva Freitas<sup>4</sup> , Luiz H. Rosa<sup>3</sup> , Sérvio P. Ribeiro<sup>4</sup> and Leandro M. Moreira1,2 \*

<sup>1</sup> Núcleo de Pesquisas em Ciências Biológicas, Universidade Federal de Ouro Preto, Ouro Preto, Brazil, <sup>2</sup> Laboratório de Genômica e Interação Microrganismos-Ambiente, Departamento de Ciências Biológicas, Instituto de Ciências Exatas e Biológicas, Universidade Federal de Ouro Preto, Campus Morro do Cruzeiro, Ouro Preto, Brazil, <sup>3</sup> Laboratório de Ecologia e Biotecnologia de Leveduras, Departamento de Microbiologia, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil, <sup>4</sup> Programa de Pós-Graduação em Biomas Tropicais, Departamento de Biodiversidade, Evolução e Meio Ambiente, Instituto de Ciências Exatas e Biológicas, Universidade Federal de Ouro Preto, Ouro Preto, Brazil

Edited by: Gero Benckiser, University of Giessen, Germany

### Reviewed by:

Daolong Dou, Nanjing Agricultural University, China Dananjeyan Balachandar, Tamil Nadu Agricultural University, India

\*Correspondence:

Leandro M. Moreira lmmorei@gmail.com

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Microbiology

Received: 29 July 2016 Accepted: 24 January 2017 Published: 10 February 2017

#### Citation:

Felestrino ÉB, Santiago IF, Freitas LS, Rosa LH, Ribeiro SP and Moreira LM (2017) Plant Growth Promoting Bacteria Associated with Langsdorffia hypogaea-Rhizosphere-Host Biological Interface: A Neglected Model of Bacterial Prospection. Front. Microbiol. 8:172. doi: 10.3389/fmicb.2017.00172 Soil is a habitat where plant roots and microorganisms interact. In the region of the Brazilian Iron Quadrangle (IQ), studies involving the interaction between microbiota and plants have been neglected. Even more neglected are the studies involving the holoparasite plant Langsdorffia hypogaea Mart. (Balanophoraceae). The geomorphological peculiarities of IQ soil, rich in iron ore, as well as the model of interaction between L. hypogaea, its hosts and the soil provide a unique niche that acts as selective pressure to the evolution of plant growth-promoting bacteria (PGPB). The aim of this study was to prospect the bacterial microbiota of holoparasitic plant L. hypogaea, its plant host and corresponding rhizosphere of IQ soil, and to analyze the potential of these isolates as PGPB. We obtained samples of 11 individuals of L. hypogaea containing fragments of host and rhizosphere remnants, resulting in 81 isolates associated with Firmicutes and Proteobacteria phyla. The ability to produce siderophores, hydrocyanic acid (HCN), indole-3-acetic acid (IAA), nitrogen (N2) fixation, hydrolytic enzymes secretion and inhibition of enteropathogens, and phytopathogens were evaluated. Of the total isolates, 62, 86, and 93% produced, respectively, siderophores, IAA, and were able to fix N2. In addition, 27 and 20% of isolates inhibited the growth of enteropathogens and phytopathogens, respectively, and 58% were able to produce at least one hydrolytic activity investigated. The high number of isolates that produce siderophores and indole-3-acetic acid suggests that this microbiota may be important for adaptation of plants to IQ. The results demonstrate for the first time the biological importance of Brazilian IQ species as reservoirs of specific microbiotas that might be used as PGPB on agricultural land or antropized soils that needs to be reforested.

Keywords: Langsdorffia hypongaea, bioprospecting, biotechnological potential, plant growth-promoting bacteria, Brazilian Iron Quadrangle, IAA and siderophores

## INTRODUCTION

fmicb-08-00172 February 8, 2017 Time: 14:51 # 2

Throughout evolution, plants have developed adaptive mechanisms related to interactions with microorganisms (Zilber-Rosenberg and Rosenberg, 2008). Accordingly, plants comprise a complex host system, made up of different microhabitats that can be simultaneously colonized by a great diversity of endophytic and epiphytic microorganisms (Lodewyckx et al., 2002). This microbial community is essential for the development of plants since it facilitates the absortion of nutrients and at the same time provides protection against phytopathogens (Fungi, oomycetes, bacteria, viroses, protozoa, and nematodes) and herbivores (Lynch and Whipps, 1990).

The rhizosphere or portion of soil that has close contact with the plant roots represents a highly dynamic environment that enables the interaction of roots with beneficial and pathogenic microorganisms, invertebrates and even root systems of other plants (Bais et al., 2006; Raaijmakers et al., 2009). The communication between the plant roots and organisms present in the rhizosphere is based on the production and secretion of chemicals that can cause different responses depending on the sensitivity or responsiveness of microorganisms present in this highly dynamic environment (Jones et al., 1994; Bertin et al., 2003; Badri et al., 2009).

Some microorganisms present in this plant rhizosphereinterface have the ability to solubilize mineral phosphates, among other soil nutrients (Rodriguez and Fraga, 1999). Many of them synthesize, provide or increase the production of plant hormones such as indole-3-acetic acid (IAA), gibberellic acid, cytokines and ethylene (Costacurta and Vanderleyden, 1995); promote associative nitrogen fixation (Richardson et al., 2009); and produce siderophores (Kloepper et al., 1980), hydrolytic enzymes such as glucanases, chitinases, proteases, cellulases, and amylases (Bashan and de-Bashan, 2005), hydrocyanic acid (HCN) (Voisard et al., 1989), and even antimicrobial agents (Compant et al., 2005). All these features allow classify them as plant growth-promoting bacteria (PGPB) (Bashan and Holguin, 1998). Accordingly, plant growth is favored by the influence of the direct or indirect action of these microorganisms, which features them as important biotool of agronomic and environmental interest (Mirza et al., 2001; Ramamoorthy et al., 2001; Vessey, 2003). Besides this potential (Moore et al., 2003), these microorganisms are commercially important when capable of producing enzymes with different applicabilities in specific sectors. In the same way, secondary metabolites produced by these microorganisms have been used in medicine, when they have antibiotic, antitumor, antifungal, or antiparasitic activity (Bertin et al., 2003; Glick, 2010).

Therefore, understanding distinct interactions of the microbiota with soil and plants allow not only a better understanding of the biological models studied, but also prospecting potential uses of this microbiota or even its biomolecules in a biotechnological perspective. In this context, the search for new microorganisms and natural processes in environments with unique characteristics and that are neglected in biological studies are fundamental, and this is the case of the Brazilian Iron Quadrangle. The geomorphological peculiarities of this soil rich in iron ore as well as the model of interaction with plants provide a unique niche that acts as a selective pressure to the evolution of PGPR. Furthermore, these peculiarities make this environment a potential hostspot of microbial diversity. Belonging to a geologically very old craton that covers about 7200 km<sup>2</sup> , the IQ extends between southeast of Ouro Preto and northeast of Belo Horizonte, continuing to the south of Serra do Espinhaço. In this region, there are rocky outcrops that have a naturally high contamination of soil with heavy metals, which makes the environment very adverse for many plant species. Despite this adverse condition, the IQ presents a great floristic diversity with high levels of endemism (Jacobi and do Carmo, 2008). Due to its association with an extensive deposit of iron ore, and since it is one of the least studied ecosystems in Brazil, IQ has a seriously threatened biodiversity due to the intense mining activity associated with its iron outcrops.

Among the plant species threatened by this anthropic activity in IQ, there have been few studies particularly on holoparasitic plants, as in the case of Langsdorffia hypogaea MART, the model of this study, being Asteraceae (Guatteria genus), Fabaceae (Dalbergia genus), Melastomataceae (Miconia genus), and Myrsinaceae (Myrsine genus) the most representative families of potential host plants from L. hypogaea (Vale, 2013). There are approximately 4200 species of parasitic plants distributed in 18 families and 274 genera (Nickrent, 2002). Langsdorffia hypogaea is one of the 44 species of plants described belonging to the family Balanophoraceae, which includes herbaceous angiosperms, achlorophyllous plants and holoparasites of roots of trees, shrubs, and even herbaceous plants (Hsiao et al., 1995). In Brazil, this holoparasite is found in the Amazon, Caatinga, Cerrado, and Atlantic Forest (Cardoso, 2014), and although it is not threatened by extinction, due to its wide distribution, it is considered at risk because of substantial habitat loss, due to global warming (Miles et al., 2004) and human use for obtaining wax (Pott et al., 2004). In some places, however, it is considered "Rare," and it is on the list of threatened flora of the state of Paraná, Brazil (Sema/Gtz, 1995). In fact, a series of local compromising extinctions can be disrupting the gene flow of this species so vagile in its biology of dispersion, and historical events may no longer correspond to the effective state of extinction threat. Morphologically, L. hypogaea has two regions that are well distinguishable anatomically, i.e., a basal vegetative body and an apical reproductive region. The vegetative body or rhizome is irregularly cylindrical, elongate and epigeal, with tomentose and fleshy appearance (Hsiao et al., 1995). The reproductive region is represented by dioecious, fleshy, unisexual inflorescences that erupt from ascending vegetative branches, encircle at the base by a sheath of bracts where the fruits are drupaceous and small (Hansen, 1980; Cardoso et al., 2011) (**Figures 1A–D**). Haustoria extend from the vegetative body and attach to the roots of host plants (Nickrent, 2002). Although there is direct connection with its host, part of vegetative body stays surrounded by soil, allowing intimate contact with organisms that live in the rhizosphere (**Figure 1E**).

Taking into account this anatomic-physiological perspective and composition of fixation substrate of the plant described above, some questions involving this model of interaction are

yet to be answered to better comprehend the biology of the species: understand if the microbiota living in association with L. hypogaea is shared to the host plant as well as with the rhizosphere, and verify if the potential of these bacterial isolates contribute to the adaptive mechanism of these plants in such a hostile environment as the soil from ferruginous fields.

In an attempt to answer these and other related questions, we carried out bacterial prospecting. The main objective was to identify which bacterial species were associated with L. hypogaea, the host and specific rhizosphere of the soil of the semidual seasonal forests of Brigida hill, basically composed by sandy-clay textures, low concentration of P and K, high concentration of Ca, Mg, Fe, As, and Sb (Vale, 2013), and pH values ranging from 3.9 to 6.2 in the first 20 cm deep (Filho et al., 2010). In parallel, biochemical assays were used to investigate the biotechnological potential of these isolates, which could be eventually used for various purposes.

### MATERIALS AND METHODS

### Location and Sampling of Plants and Rhizosphere Soil

The collections were made in Serra da Brigida (central point: 20◦ 210 3500S, 43◦ 300 1100W), which is part of Parque Natural Municipal das Andorinhas and is in southern part of Environmental Protection Area Cachoeira das Andorinhas, within IQ (Ferreira, 2011), municipality of Ouro Preto, Minas Gerais – Brazil (**Supplementary Figure S1**). The topography of the region is sustained by itabirites and quartzites. Itabirites are iron formations, metamorphic and strongly oxidized, showing discontinuous bodies with high ore content (>64% Fe) (Rosière and Chemale, 2000). Sandy and flooded soils are absent and they have large amounts of humic substances (Jacobi and do Carmo, 2008). In this forest fragment, we randomly collected 11 individuals of L. hypongaea, containing fragments of parasitized roots and remnants of corresponding rhizosphere. The samples were stored in sterile plastic bags and processed on the same day of collection.

### Isolation of Bacteria, Media and Culture Conditions

The samples of L. hypogaea were externally disinfected using chlorine solution 2.5% by 2 min. We used for each sample five inner fragments of L. hypogaea root (∼1.0 × 0.5 cm) inoculated in Luria-Bertani (LB) medium (Maniatis et al., 1982) containing 0.03 mg/L thiophanate, with pH adjusted to 6.0. The host root was washed following a standardized sequence of solutions for surface disinfection (9 g/L NaCl – 2 min, 70% alcohol – 2 min, 2,5% sodium hypochlorite – 2 min and 9 g/l NaCl – 2 min), and similarly, five fragments of each root of plant host were inoculated in LB medium. For the isolation of microorganisms present in the rhizosphere and in fragments of plant host root containing traces of soil (approximately 2 g), the samples were placed in a 10 mL of saline solution (0,5 NaCl g/L) for 10 min. Next, an aliquot of 100 µL of this solution was inoculated in selective LB medium. All plates were incubated at 25–28◦C for a period of up to 10 days, and the microorganisms grown were isolated in new 60 mm diameter Petri dishes containing the same culture medium. The colonies isolated were photographed (front and back, data not shown) and grouped according to their origin. All isolates were cataloged and preserved in 30% glycerol and stored at −80◦C.

### DNA Extraction, Amplification, and Sequencing

DNA of the isolates was extracted using the CTAB/NaCl protocol (Doyle and Doyle, 1987). The primers 27f and 1492r were used for amplification of bacterial 16S rRNA gene (Lane et al., 1985). The 50-µL PCR mixture contained 20–50 ng of DNA, 250 pmol of each primer, 5 µL 10x PCR buffer, 2.5 U rTaq DNA polymeraseTM (Invitrogen), and 100 µM deoxynucleoside triphosphate mixture. The PCR program consisted of initial denaturation at 95◦C for 5 min, followed by 35 cycles of 1 min of denaturation at 95◦C, 45 s of annealing at 47◦C and 2 min of extension at 72◦C and a final extension for 10 min at 72◦C, utilizing a 2720 ThermalCyclerTM (Applied Biosystems). The amplicons generated by PCR were verified in 1% agarose gels and purified using 20% PEG-8000 in 2.5 M NaCl (Arbeli and Fuentes, 2007). The product obtained was quantified by spectrophotometry using a NanoDrop ND 1000TM (NanoDrop Technologies). Sequencing was carried with the DYEnamicTMTM kit (Amersham Biosciences, USA) in combination with the MegaBACE 1000TM automated sequencing system (Amersham Biosciences, USA). The sequencing reactions were performed with 100–150 ng purified DNA and the reagents in the DYEnamicTMTM kit (Amersham Biosciences, USA), using the manufacturer's recommendations. The program consisted of 36 cycles with an initial denaturation at 95◦C for 25 min, followed by 15 s of annealing at 50◦C and 3 min of extension at 60◦C. After cycling, the reaction product was transferred to a 96-well sequencing plate to be precipitated.

For precipitation, 1 µL of 7.5 M ammonium acetate was added to each well. Next, 28 µL of absolute ethanol (Merck, USA) were added. The plate was vortexed and incubated for 20 min at room temperature, protected from light. Afterward, the plate was centrifuged for 45 min at 3200 × g and the supernatant was discarded. Next, 150 µL of 70% ethanol were added. The plate was centrifuged again for 15 min at 3200 × g and the supernatant was then discarded. The plate was allowed to stand for 20 min, protected from light, para evaporation of ethanol. Precipitated DNA in each well of the plate was then resuspended in 10 µL of loading buffer (present in sequencing kit). The plate was vortexed for 2 min, centrifuged for 1 s at 800 × g and stored at 4◦C, protected from light, until injection of samples in a MegaBACE 1000TM sequencer (Amersham Biosciences, USA).

### Determination of Sequences and Phylogenetic Analysis

The contigs were assembled using the forward and reverse sequences of each 16S rRNA gene amplicon using Phred (Ewing and Green, 1998; Ewing et al., 1998). The DNA sequences were analyzed utilizing the BLASTn program (Altschul et al., 1997). The sequences were aligned using the program Muscle (Edgar, 2004) and then curated by the program G-block (Castresana, 2000). A neighbor-joining phylogenetic tree was then determined using PhyML 3.0 (Anisimova and Gascuel, 2006) followed by TreeDyn (Chevenet et al., 2006), and the statistical robustness of the analysis was estimated by bootstrapping with 1,000 replicates.

### Nucleotide Sequence Accession Numbers

All sequences obtained in this study were deposited in GenBank, according to the accession numbers given in **Table 2**.

### Production of IAA

The production of IAA was determined by the method of Bric et al. (1991), with modifications. Bacterial producers of IAA were identified by the change in the color of the nitrocelulose disk from yellow (negative result) to red (positive result). The assays were made in triplicates and only those that achieved a positive result in at least two of them were accounted in the analysis.

### Production of Ammonium Ions

The production of ammonium ions was determined by the indophenol method (Verdouw et al., 1978), using Proteus sp. as positive control (Vince et al., 1973). The samples were read at 690 nm, and the absorbance values were compared with a control condition. The assays were made in triplicates and only those that the absorbance values were greater than or equal to the Proteus values in at least two of the assays were accounted in the analysis.

### Production of Siderophores

The production of siderophores was based on the work of Schwyn and Neilands (1987), with modifications. To remove traces of iron in medium it was made a pretreatment with hidroxiquinolone followed by separation using dropping funnel (Pierre et al., 2003). An orange shade of the culture medium around the regions where bacteria grew was indicative of production of siderophores. The assays were made in triplicates and only those that achieved a positive result in at least two of them were accounted in the analysis.

### Nitrogen Fixation

Nitrogen fixation capacity was investigated using two serial methods: Nitrogen-free combined carbon (NFCC) semi-solid medium, free of N<sup>2</sup> (pH 5.7), supplemented with 5 g/L mannitol and 5 g/L sucrose (Dobereiner et al., 1976) and PCR using universal nifH primers (Burgmann et al., 2004). The Change in color of medium from yellow to green was indicative of the capacity to fix N2. This assay was made in triplicates and only those that achieved a positive result in at least two of them were accounted in the analysis, which was confirmed by amplification of nifH fragment evaluated in 1.2% agarose gel. Bradyrhizobium elkanii BR96 was used as positive control.

### Production of Hydrocyanic Acid (HCN)

The production of HCN was determined by the method of Bakker and Schippers (1987), with modifications. A change in yellow color of filter paper (negative result) to brown (positive result) indicated production of HCN. The assays were made in triplicates and only those that achieved a positive result in at least two of themwere accounted in the analysis.


TABLE 1 | Distribution of bacterial representativeness of taxa found in Langsdorffia hypogaea, host and rhizosphere.

### Production of Amylase, Cellulase, and Protease

Amylase activity was determined in 90-mm Petri dishes containing Yeast nitrogen base (YNB) medium (2%) containing 2 g/L soluble amide, 5 g/L peptone, and 1 g/L yeast extract, with pH adjusted to 6.0 (Strauss et al., 2001). Cellulase activity was determined in 90-mm Petri containing YNB medium (2%) supplemented with 0.5 g/L cellobiose and 1 g/L carboxymethyl cellulose (Teather and Wood, 1982). Protease activity was determined in 90-mm Petri dishes containing culture medium composed of 20 g/L casein, 5 g/L peptone, 3 g/L yeast extract, 10 g/L glucose, and 20 g/L agar, with pH adjusted to 5.0 (Strauss et al., 2001). After 3 days of growth at 28◦C, the bacterial isolates in all assays producing a transparent halos determined the amylase, cellulose, or protease activity in the respective isolates and assays.

### Inhibition of Entero- and Phytopathogens

To investigate a possible antimicrobial activity of the isolates, antimicrobial assays were performed against the following targets: Staphylococcus aureus ATCC 29213 (opportunistic human pathogen), Bacillus cereus ATCC 11778 (associated with food poisoning), Klebsiella pneumoniae ATTC 4352 (causal agent of pneumonia), Shigella flexneri (associated with dysentery) and Xanthomonas citri subsp. citri 306 (causal agent of citrus canker) using solid LB in a direct inhibition test. The assays were made in triplicates and only those that achieved a positive result in at least two of them were accounted in the analysis.

To investigate the capacity of isolates to inhibit the growth of the phytopathogenic fungus Fusarium oxysporum f.sp. lini (causal agent of various diseases in plants), the isolates were grown in LB and then transferred to 60-mm Petri dish containing potato agar. The bacterial isolates were inoculated to make a square of approximately 2 cm on the culture medium. After 2 days of growth of these isolates at 28◦C, a fraction of 4 mm<sup>2</sup> of culture medium containing F. oxysporum was inoculated exactly at the center of the plate (center of square), starting at a pre-growth of 5 days at 28◦C. The bacteria capable of producing some anti-Fusarium substance hindered the growth of the fungus, compared to the growth profile of Fusarium under control conditions in the absence of bacterial isolate. The values to determine per cent inhibition were obtained by the formula (%) = (C × E)/C × 100, where C is the diameter of the Fusarium culture in control, and E is the diameter in the presence of bacterial isolate (Zhao et al., 2014). The assays were made in triplicates and the average of per cent inhibition was ploted in radar graph.

### RESULTS

### Characterization of Bacterial Population

Of the 11 samples collected in four different points of Serra da Brigida-MG, we selected and preserved 81 isolates (**Table 1**).

#### TABLE 2 | List of isolates representing the 75 OTUs from L. hypogaea, host and rhizosphere.


(Continued)

#### TABLE 2 | Continued

fmicb-08-00172 February 8, 2017 Time: 14:51 # 7


However, isolate L5 was non-viable after storage in glycerol. The growth rate of isolates L13, L15, L21, H57, and H79 were unsatisfactory, resulting in the extraction of insufficient DNA and sequencing of low quality. Although sample L4, isolated from L. hypogaea, showed good quality of sequencing and satisfactory assembly of contigs using the assembly parameters (see Materials and Methods), it still showed low percentage of similarity (76.8%) and identity (77%) compared with sequences deposited in GenBank of NCBI, featuring it as a potential new organism to be investigated rigorously.

In the three niches analyzed, we found isolates of the phyla Firmicutes (34), belonging to the genera Bacillus (17), Lysinibacillus (13), Paenibacillus (3), and Viridibacillus (1), and Proteobacteria (40), belonging to the genera Serratia (26), Klebsiella (3), Rahnella (2), Citrobacter (1), Enterobacter (5), Shewanella (1), Raoultella (1), and Pseudomonas (1) (**Table 2**). Four of these genera (Serratia, Bacillus, Lysinibacillus, and Enterobacter) were found in all three niches analyzed, where the first genera was the most representative. Isolates of the genera Citrobacter, Paenibacillus, and Shewanella were found only in the plant host. Isolates of the genera Viridibacillus, Pseudomonas, and Raoultella were found only in the rhizosphere (**Table 2**). Isolates of the genera Klebsiella and Rahnella were found exclusively shared between L. hypogaea and the rhizosphere.

Phylogenetic analysis of the isolates confirmed the sequencing results and allowed the grouping of the bacterial isolates from different niches to the taxa represented, Firmicutes and Proteobacteria. The isolates from L. hypogaea (L3, L16, L22, and L8), the isolates from the plant host (H55, H60, H64, H67, H71, H75, H76, and H80), and the isolates from the rhizosphere (R38 and R51) were grouped in a different clade (**Figure 2**).

For the biochemical tests, 66 isolates were analyzed since nine of these isolates showed reduced growth rate, prevented an accurate analysis of the results in the proposed biochemical assays. Of these, 63 isolates showed growth in nitrogen-free medium, representing 14 (93.4%) of the 15 isolates of L. hypogaea, 23 (95.8%) of the 24 isolates from plant host and 26 (96.3%) of the 27 isolates from rhizosphere (**Figure 3A**). From these isolates, the presence of nifH was confirmed in 23 genomes by PCR analysis (**Supplementary Figure S2**). With regard to siderophore production, more than 62% of the isolates were capable of producing these compounds. All isolates were able to produce ammonium ions, at different concentrations. Meanwhile, only 15 isolates (22.72%) were capable of producing hydrocyanic acid. More than 86% of the isolates were capable of producing IAA, of which 13 were from L. hypogaea (**Figure 3A**). In attempt to understand the potential of each of these isolates with regard to the capacity to fix atmospheric N2, to produce siderophores, IAA, and HCN, Venn diagrams were prepared for each of the three niches analyzed (**Figure 3B**). Especially the isolates L11, H55, H71, and H80 yielded positive results for all analyses. Of the isolates obtained from L. hypogaea, four were capable of to fix N<sup>2</sup> and to produce IAA (L13, L16, L18, and L28), and seven were able to produce IAA and siderophores, and to fix N<sup>2</sup> (L1,

L2, L3, L4, L8, L10, and L12). Of the isolates obtained from the rhizosphere, six were capable of producing IAA and fixing N<sup>2</sup> (R23, R26, R33, R36, R39, and R40). Eleven were capable of producing IAA and siderophores, and even fixing N<sup>2</sup> (R30, R31, R35, R37, R38, R43, R44, R45, R46, R51, and R52), while another six were capable of producing IAA and HCN, and fixing N<sup>2</sup> (R25, R27, R41, R42, R50, and R54). Of the isolates obtained from the host plant, six were capable of producing IAA and fixing N<sup>2</sup> (H64, H73, H74, H75, H77, and H78). Two were capable of producing siderophores and fixing N<sup>2</sup> (H58 and H60), whereas ten were able to produce IAA and siderophores, and to fix N<sup>2</sup> (H59, H61, H62, H65, H66, H67, H68, H72, H79, and H80).

Regarding hydrolytic enzymes assay, 11 isolates were able to produce amylase. None of these isolates was obtained from

(B) Venn diagrams indicating the potential of each isolate with regard to the biochemical performed (IAA, HCN, N2, and Side) in each niche evaluated (Langs, Rizhos, and Host). (C) Venn diagrams showing the potents of each isolate with regard to the enzymatic assays performed (Cel, Pro, and Amy) in each niche evaluated (Langs, Rhizos, and Host). <sup>∗</sup>Denotes the isolates whose nifH were amplified by PCR.

L. hypogaea, the isolates obtained from the plant host were the most representative in this analysis. Thirty isolates were capable of producing cellulase, with greatest representativeness among the isolates from the rhizosphere. Thirty isolates were able to produce proteases, with egual representativeness among the isolates from the plant host and rhizosphere. Similarly, a Venn diagram pointing out the isolates and their origin with respect to the production of hydrolytic enzymes was constructed (**Figure 3C**). Notably, isolate H67 was capable of producing all three types of hydrolytic enzymes investigated.

Twenty isolates were able to inhibit the growth of enteropathogens. They include 11 inhibitors of Staphylococus aureus (L1, R51, R52, R53, R54, H62, H63, H65, H69, H70, and H72), 11 inhibitors of Klebisiella pneumoniae (L9, L13, L14, R29, R32, R34, R46, R49, and H69), and 10 inhibitors of Shiguella flexneri (L1, R51, R52, R53, R54, H62, H63, H65, H69, and H70) (**Figure 4A**). No isolate was capable of inhibiting the growth of Bacillus cereus, and only one isolate inhibited the growth of Xanthomonas citri strain 306 (H65). Nineteen isolates inhibited the growth of Fusarium (**Figure 4B**), 5 of themwere obtained from L. hypogaea (L1, L2, L3, L20, and L28), eight from the rhizosphere (R26, R30, R33, R39, R50, R51, R52, and R55) and six from the plant host (H58, H70, H77, H78, and H80) (**Figure 4D**). The potential of some of these isolates is indicated in **Figure 4C**, ranging from 28% in isolates from L. hypogaea to above 95% in the isolates R50, R51, R52, H55, and H58 from rizhosphere and host, respectively.

FIGURE 4 | Analysis of inhibition of entero- and phytopathogens. (A) Profile of growth inhibition of isolates regarding the target S. aureus. (B) Profile of growth inhibition of Fusarium induced by three of the 19 isolates positive for this analysis. (C) Radar demonstrating the representativeness and percentage of growth inhibition of Fusarium of the 19 isolates positive for this analysis. Note that the majority of isolates from the rhizosphere inhibited almost 90% of growth of the fungus. (D) Venn diagrams showing the potentials of each isolate with regard to inhibition of enteropathogens (Sa, S. auereus; Sf, S. flexneri; and Kp, K. pneumoniae) and of Fusarium oxysporum (Fo) in each niche evaluated (Langs, Rizhos, and Host).

## DISCUSSION

### Identification of Microbiota Associated with L. hypogaea and Its Interactions

The importance and complexity of the rhizosphere in the interaction with plants and other organisms in which they live have been reported in many studies (Bais et al., 2006; Vacheron et al., 2013; Zhang et al., 2014). Similarly, the modifications in the physiological profile of plants as a result of alterations in the chemical composition of the rhizosphere or the microbiota contained therein have also been characterized, especially with regard to the interaction between plants of agricultural interest (Saleem et al., 2007; Bhattacharyya and Jha, 2012; Dodd and Ruiz-Lozano, 2012; Nadeem et al., 2014). Contrary to these advances, studies seeking to understand the identification of microbiota associated with parasitic plants, especially holoparasitic ones are limited or lacking. This relation of strict dependence on its host makes it a very interesting target for understanding the degree of importance of microbiota associated with the adaptation and resultant survival of species in the most diverse environments where they occur. Accordingly, this is the first work aimed at understanding the composition and importance of microbiota in the biological interface holoparasitic plant-host-rhizosphere, using L. hypogaea as a model. Not only is the biological model distinctive for studies bioprospecting, the sampling site, the IQ, is characterized as an environment also neglected in studies involving its microbiota. Because it is an environment enriched in iron compounds and other heavy metals, whose soils are extremely compacted, adaptation of this microbiota can be directly related not only to the soil features, but also related to the adaptation of plants in this environment, and may uncover

new organisms and the biotechnological potential of this specific microbiota.

Although there are no known genera that are unique for L. hypongaea, isolate L4 did not show significant identity (over 80%) with any other isolate with sequence deposited in databanks and therefore deserves attention. This isolate was capable of producing ammonium ions in high concentration, IAA and siderophores and fix nitrogen. This activity could generate better or complementary nutritional conditions compared to those offered by the plant host such as for a tree growing on soils lacking nutrients. A key adaptive aspect of this would be the decrease in risk of mortality of the plant host. Regardless of nutrients that it provides, physical sustentation and supply of water and carbohydrates to the parasite is totally dependent on host plants, and their death is a great adaptive disadvantage (Lopez Pascua et al., 2014). Thus, the co-association with a microbiota capable of providing the necessary nutrients saves the parasitized tree from irreversible stress. However, more studies are necessary to confirm these preliminary results.

Similarly, isolates of the genera Klebisiella and Rahnella were found only in L. hypogaea and the rhizosphere. This allows us to infer that there may be a direct association of soil bacteria with the holoparasitic plant, and that a possible relation of complementarity makes the parasitic plant not only dependent on the plant host for survival but also dependent on specific soil bacteria. This allows us to raise the prospect that perhaps the concept of botanical holoparasitism needs to be rethought, taking the scientific community to undertake further research in this area of knowledge. Although need the host plant for aquisition of carbohydrates, since it does not have the capacity of producing them itself (by the chlorophyll absence), L. hypogaea could also depend closely on PGPB to assist in the development of its roots, aquisition of ions from the rhizosphere or even in the interaction with its host plant through their haustoria. Finally, isolates of the genera Serratia, Bacillus, Lysinibacillus, and Enterobacter were found in the three niches, reinforcing this exchange of microbiota, now a more complex perspective. These results reinforce the prospect that in a community, the ecological interaction of plants and microorganisms is directly related to the rigor of the habitat and the ability of colonization. In other words, the mutualistic interactions with microorganisms would be the basis of evolution of adaptations to inhospitable environments. The theoretical concept of species "A" of adversity strategists [sensu (Greenslade, 1983)], in contrast to the artificial r-k continuum of Pianka (1970; Taylor et al., 1990), or the understanding of "template habitats" (Southwood, 1977; Korfiatis and Stamou, 1999) and specialization in habitats (consequently, the whole concept and use of bio-indicators) can make sense only in the light of these interactions. Therefore, a specific habitat provides characteristic conditions that promote the growth of certain microorganisms, which in turn facilitate the development of other species associated with them.

Thus, much less time adjustment to selective pressures imposed by oligotrophic and contaminated soils would be necessary to ensure the invasion of environment, always done by species that evolve from less hostile niches. This possibility of having one microbiological micro-habitat that minimizes the natural hostility of the environment can change the ecologicalevolutionary perception of biodiversity evolution, and even the basic theoretical models that guide our understanding of these processes.

### PGP Activities of the Isolates

About 95% of isolates showed growth in medium combined nitrogen, thereby demonstrating diazotrophic activity. This may have a direct relation with an environment whose soil is highly leached, oligotrophic and contaminated, like these montane ecosystems or any other soil of Brazilian savannas (Goodland, 1971; Goodland and Pollard, 1973; Batmanian and Haridasan, 1985; Haridasan, 2008). The understanding of the evolutionary costs of colonizing these environments are well studied (Ribeiro and Brown, 2006) and consistent with the existing theoretical propositions (Herms and Mattson, 1992; Fine et al., 2004). This is the first time that the role of mutualistic microbiota was taken into consideraction as a fundamental adaptive mechanism.

Iron can be available in the soil as Fe2<sup>+</sup> or Fe3+, where the latter is less soluble but more abundant. Siderophores are molecules secreted by some bacterial species that chelate Fe3<sup>+</sup> converting it to Fe2+, which as a consequence is internalized by specific cellular receptors for these ions (Neilands, 1995). From a competition point of view, microorganisms that are able to utilize siderophores as a mechanism of acquisition of Fe3<sup>+</sup> make it available for them consequently decreasing the availability of iron for possible competing microorganisms in the same niche (Hibbing et al., 2010). Often when this competition occurs because of a phytopathogenic organism, this resource becomes used as an indirect mechanism of plant growth since it controls the growth of these phytopathogens (Perez-Montano et al., 2014). In the same perspective, many plants only have receptors for siderophores, depending intimately on the their production by microorganisms that live in symbiosis or cooperation with these plants for acquiring iron from the environment, essential for their growth (Crowley et al., 1991). More than half of the bacterial isolates obtained here were producers of siderophores (**Figure 3**). Although this number was high, it was expected since the environment favors the adaptation of bacteria capable of surviving in such high concentrations of iron. Therefore, further study of the structural composition and regulation of the synthesis of these substances in these microorganisms is necessary and may lead to the discovery of new biomolecules with ion-chelating activity. With regard to the region where the plants were isolated, this perspective becomes even more interesting, because it is an environment classically reported as rich in arsenic, and it is possible that these microorganisms make use of these molecules as an adaptive alternative to the presence of this element (Gonçalves and Lena, 2013).

Similarly, about 86% of isolates were capable of producing IAA. Since this hormone helps to increase root growth, this could explain the adaptation of plants of these niches to extremely hard soil, which provides surface branching of roots, favoring not only anchoring but also procurement of water and minerals. These bacteria may have great importance in the study model proposed since it is a holoparasitic plant whose capacity to produce this hormone can be rather low in relation to an autotrophic

plant (Magnus et al., 1982), which could explain its intimate dependence on IAA-producing bacteria.

Far less representative, but of similar importance were the isolates in which we identified hydrolytic enzyme activities. These enzymes are of great industrial interest, since they can optimize the manufacture of products of economic interest such as in the case of glucose for fermentation processes obtained through cellulolytic or amylolytic activity, or of amino acids and peptides widely used in the food, pharmaceutical and chemical industry, obtained from proteolytic activity (Dalmaso et al., 2015). In an environmental microbiological context, all these enzymes can be of fundamental importance in the process of adaptation to a specific niche.

Individually, some genera deserve attention because of the previous results described in the literature. R54, for example, was isolated from the rhizosphere and showed similarity to Pseudomonas fluorescens. A recent study involving the strains PA4C2 and PA3G8 of P. fluorescens, also isolated from the rhizosphere, were found to be able to inhibit the growth of the phytopathogen Dickeya (Cigna et al., 2015), which causes diseases in herbaceous plants. In out study, R54 was also found to be a potential inhibitor of enteropathogens since it blocked the growth of Staphylococcus aureus and Shigella flexneri. Specific strains of Pseudomonas have also been described as inducing systemic resistance in cloves, cucumbers, radishes, tobacco, and Arabidopsis, which raisethe possibility of these bacteria being a potential growth inhibitors of phytopathogens in wild plants. Similarly to Pseudomonas, induced systemic resistance has also been described for different strains of Bacillus spp., including the specific species B. amyloliquifaciens, B. subtilis, B. pasteurii, B. cereus, B. pumilus, B. mycoides, and B. sphaericus (Choudhary et al., 2007). When inoculated or present in specific organisms, they induce a significant reduction in the incidence or severity of diseases in various hosts (Choudhary et al., 2007). In this study, various isolates showed similarity with bacteria of the genus Bacillus, including the strains B. cereus and B. mycoides. Some of these isolates showed positive results for all the biochemical assays performed (H80) or inhibited the growth of three of the four enteropathogens investigated (H69).

Bacteria of the genus Enterobacter has been associated with numerous biological models. Enterobacter sp. strain EJ01 isolated from Dianthus japonicus thunb (China Sea rose) was described as a bacterium capable of aiding vegetative growth, besides alleviating salt stress in tomato and Arabidopsis (Kim et al., 2014). Of the five isolates similar to Enterobacter identified in this study, all produced siderophores and fixed N2, and only two were capable of producing IAA. In another study, Serratia marcescens isolated from the rhizosphere of the coconut tree was found to fix nitrogen and to produce IAA and siderophores, among other compounds investigated (George et al., 2013), highlighting the importance of these genera in the support and growth of these plants. Of the 26 isolates that showed similarity to Serratia, 25 isolates were capable of fixing N<sup>2</sup> and producing IAA, and 15 were capable of producing siderophores.

Paenibacillus yonginensis DCY84 was evaluated in growth with Arabidopsis thaliana subjected to salt, drought and heavy metal stress, and the study showed that plants treated with this bacterial isolate were more resistant than the untreated control plants (Sukweenadhi et al., 2015). Our isolates H59, H66, and H70 showed similarity to this genus and were able to produce siderophores and IAA and to fix N2, besides inhibiting enteropathogens. Another recent work isolated bacteria from the rhizospheric soil of Populus euphratica and identified ten strains that induced a significant increase in dry weight of buds and roots of wheat (Wang et al., 2014). These isolates were identified as being from the genera Pseudomonas, Bacillus, Stenotrophomonas, and Serratia. Among these strains, Serratia sp. 1–9 and Pseudomonas sp. 23/05 were the most effective strains. Both produced auxin, and significantly increased production when grown under simulated dry conditions, leading to a direct effect on promoting plant growth under drought stress (Wang et al., 2014). Similarly, a work identified 12 endophytic bacteria characterized as diazotrophic, two species belonging to the genus Paenibacillus, three to the genus Mycobacterium, three to the genus Bacillus, and four to the genus Klebsiella (Ji et al., 2014). Rice seeds treated with these bacteria showed improved growth, increase in height and dry weight and antagonistic effects against pathogenic fungi (Ji et al., 2014). Our isolates that showed high identity to Paenibacillus (H59 – 99% and H66 – 98%), were also capable of fixing N<sup>2</sup> and were thus diazotrophic. The isolate H70, although showing similarity to Paenibacillus, was not able to fix N2, but did inhibit the growth of enteropathogens and Fusarium. All isolates that showed similarity to Klebisiella (L16, L28, and R39) were also capable of fixing N2, besides producing IAA.

### Perspectives of Use of Isolates

In an agroecological context, there is currently an emerging demand for the development of sustainable agriculture, to decrease our dependence on agrochemical farming and its harmful consequences to the environment (Bhardwaj et al., 2014). The utilization of PGPB to increase farm production has become an important alternative. Similarly, alternative methods for pest control attracted attention, and biological control has been considered a viable solution for various diseases that are difficult to control (Cespedes et al., 2015). This practice aims to maintain a balance in the agroecosystem, so that the host, in the presence of a pathogen or pest, does not suffer significant damage due tothe controlling action exerted by non-pathogenic organisms (Meldau et al., 2012). Thus, understanding microbial relations in soils and plants can lead to the discovery of microorganisms with great agricultural potential and other applications as well.

Besides agroecological importance, all potential presented by microbiota isolated from these neglected biological niches drives the search for new products and processes with potential pharmacological and for environmental bioremediation. This was evident by the ability of some isolates to inhibit three out of four investigated enteropathogenic species, besides Xanthomonas and Fusarium strains. Analysis of genomic composition and metabolites produced by these isolates could uncover new metabolic pathways associated with the synthesis of new biomolecules of pharmaceutical interest. In the same way, the isolates associated with production of IAA and siderophores may be used as bacterials consortia to induce native plant growth in areas degradeted by human action (de-Bashan et al., 2012).

### CONCLUSION

fmicb-08-00172 February 8, 2017 Time: 14:51 # 13

The integration of biological data found in this study suggests a hypothetical complex network of interaction and mutual dependence between the niches analyzed and the isolated bacteria. Classically, holoparasitic plants draw all necessary nutrients from their host plant. However, the results of this study show that the microbiota in L. hypogaea can also be of benefit by supplying nutrients essential for the survival of the plant. This hypothesis needs further and in-depth studies to become valid. From an ecological perspective, this is the first report of the potential of bacteria isolated from the IQ region in producing these siderophores and IAA, which allowed us to infer that part of the adaptive process of these plants in ferroginous fields can be a result of the ability of a large percentage of isolates to produce these compounds. Siderophores would be key for the chelation of iron in the Fe3<sup>+</sup> state, existing in high concentrations in these ferroginous fields, and the production of IAA by a large number of isolates demonstrated how essential this compound would be for the induction of extensive root system of plants that survive in this environment. The characteristics of the soil from this area are extremely adverse, with low supply of water, requiring the plants to obtain nutrients from more superficial regions. Thus, it is possible that the survival of plants in this environment have some relationship with the presence and interaction with these microorganisms, which could also justify the large endemic plant found in this environment. From a biotechnological aspect, the perspectives and results obtained with this work point to the potential of developing a bacterial consortium that could be used as an indispensable tool in the recovery of areas degraded by anthropic actions, especially in this region where mining activities are eliminating important species of the biome of ferroginous fields. Therefore, the results presented in this work emphasize the importance of studying biological models neglected and differentiated such as the holoparasite plants and ferruginous soils from IQ since they are propitious sources for finding new compounds with biotechnological potential.

### REFERENCES


### AUTHOR CONTRIBUTIONS

ÉF and LF collected samples. ÉF and LM conceived and designed the experiments. ÉF and LM performed all experiments and analysis. ÉF, LM, IS, and LR contributed with reagents, materials and analysis tools. ÉF and LM prepared the figures and tables. ÉF, LM, and SR wrote the paper.

### ACKNOWLEDGMENTS

We thank Prof. Jesus A. Ferro, Prof. Alessandro Varani and Agda Facincani (FCAV-UNESP) for their help with sequencing and assembly of contigs. We are grateful to Prof. Renata Guerra de Sá Cota (DECBI-UFOP), Silvana de Queiróz Silva (DECBI-UFOP), Cinthia Lopes de Brito Magalhães (DECBI-UFOP), Maria Catarina Megumi Kasuya (UFV), and Cornélio de Freitas Carvalho (DEQUI-UFOP), who all contributed to this study in some way. This work was supported by the National Council for Scientific and Technological Development – CNPq (Project 481226/2013-3), Fundação de Amparo à Pesquisa do Estado de Minas Gerais – FAPEMIG (Project CBB APQ-02387-14) and UFOP scientific grants granted to LM. SR is granted researcher from CNPq.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2017.00172/full#supplementary-material

FIGURE S1 | Geographical location of Serra da Brigida, collection site of L. hypogaea. Serra da Brigida is located around the city of Ouro Preto, state of Minas Gerais – Brazil. The stars indicate the sampling points for Langsdorffia hypogaea. Adaptated from Atlas Digital GeoAmbiental (http://institutopristino. org.br/atlas/).

FIGURE S2 | Presence of nifH confirmed by PCR analysis.

spp-mediated plant growth-stimulation. Soil Biol. Biochem. 19, 451–457. doi: 10.1016/0038-0717(87)90037-X



ethylene-insensitive plant genotype in nature. Front. Plant Sci. 3:112. doi: 10. 3389/fpls.2012.00112


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Felestrino, Santiago, Freitas, Rosa, Ribeiro and Moreira. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Identification of a Potential ISR Determinant from Pseudomonas aeruginosa PM12 against Fusarium Wilt in Tomato

Sabin Fatima and Tehmina Anjum\*

Institute of Agricultural Sciences, University of the Punjab, Lahore, Pakistan

Biocontrol of plant diseases through induction of systemic resistance is an environmental friendly substitute to chemicals in crop protection measures. Different biotic and abiotic elicitors can trigger the plant for induced resistance. Present study was designed to explore the potential of Pseudomonas aeruginosa PM12 in inducing systemic resistance in tomato against Fusarium wilt. Initially the bioactive compound, responsible for ISR, was separated and identified from extracellular filtrate of P. aeruginosa PM12. After that purification and characterization of the bacterial crude extracts was carried out through a series of organic solvents. The fractions exhibiting ISR activity were further divided into sub-fractions through column chromatography. Sub fraction showing maximum ISR activity was subjected to Gas chromatography/mass spectrometry for the identification of compounds. Analytical result showed three compounds in the ISR active sub-fraction viz: 3-hydroxy-5-methoxy benzene methanol (HMB), eugenol and tyrosine. Subsequent bioassays proved that HMB is the potential ISR determinant that significantly ameliorated Fusarium wilt of tomato when applied as soil drench method at the rate of 10 mM. In the next step of this study, GC-MS analysis was performed to detect changes induced in primary and secondary metabolites of tomato plants by the ISR determinant. Plants were treated with HMB and Fusarium oxysporum in different combinations showing intensive re- modulations in defense related pathways. This work concludes that HMB is the potential elicitor involved in dynamic reprogramming of plant pathways which functionally contributes in defense responses. Furthermore the use of biocontrol agents as natural enemies of soil borne pathogens besides enhancing production potential of crop can provide a complementary tactic for sustainable integrated pest management.

Keywords: Fusarium wilt, tomato, induced systemic resistance, Pseudomonas, 3-hydroxy-5-methoxy benzene methanol

## INTRODUCTION

Tomato (Solanum lycopersicum L.) is a member of family Solanaceae, cultivated worldwide, ranked first among the processing crops and second as a vegetable crop. In Pakistan it is cultivated on about 58.196 thousand hectares with an annual production of 574.052 thousand tons (FAO, 2013). It contains valuable nutrients like vitamin A and C, calcium, phosphorus, potassium and magnesium

#### Edited by:

Gero Benckiser, Justus Liebig Universität Gießen, Germany

#### Reviewed by:

Javier Plasencia, National Autonomous University of Mexico, Mexico Michael Rothballer, Helmholtz Center Munich, Germany

> \*Correspondence: Tehmina Anjum tehminaanjum@yahoo.com

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 19 January 2017 Accepted: 08 May 2017 Published: 31 May 2017

#### Citation:

Fatima S and Anjum T (2017) Identification of a Potential ISR Determinant from Pseudomonas aeruginosa PM12 against Fusarium Wilt in Tomato. Front. Plant Sci. 8:848. doi: 10.3389/fpls.2017.00848

(United States Department of Agriculture [USDA], 2009). It is also a source of an antioxidant compound named lycopene that has been found helpful against cancer (Miller et al., 2002). Tomato wilt caused by the fungus Fusarium oxysporum Schlecht. f. sp. lycopersici (Sacc.) W.C. Snyder et H.N. Hansen (FOL) is an alarming disease, causing yield losses up to 25% (Fravel et al., 2005; Park et al., 2013). Disease control strategies include use of resistant tomato varieties with cultural, chemical and biological control (Agrios, 2005; Pottorf, 2006). Cultural control lost its effectiveness as pathogen has a wide host range. Use of resistant varieties is futile due to chances of mutation in Fusarium spp. Chemical control is now losing its ground due to adverse effects of chemicals on environment and soil microbiota that calls for alternative inputs with lower dependency on chemicals for sustainable agriculture (Lucas, 2011). Recently used biological control employed induced systemic resistance (ISR) mechanism using rhizospheric plant growth-promoting bacteria against fungal pathogens (Pieterse et al., 2014).

Plants are devoid of an immune system which is the chief characteristic of mammals. To deal with different types of pathogens, plants have weakly inducible or constitutive defense systems including plant cell walls, cuticles, phytoanticipins (Underwood, 2012; Newman et al., 2013) and inducible defenses in which plants activate their immune system by the stimulus of signal molecules known as elicitors (Henry et al., 2012; Maffei et al., 2012; Newman et al., 2013). Elicitors can be derived from natural/living organisms like plants and microbes or generated synthetically (Walters et al., 2013).

Plant pathogens stimulate the host plant to activate defense responses against the invaders but this weak defense reaction will not limit the spread of the pathogen into the host plant (Thordal-Christensen, 2003). However, the defense responses can be enhanced by triggering the plant before pathogen attack. Interaction of some rhizobacteria with the plant roots has proven to increase plant resistance against some pathogenic bacteria, fungi and viruses. This phenomenon is termed as ISR (Lugtenberg and Kamilova, 2009). Application of microorganisms to control diseases, which is a form of biological control, is an environment-friendly approach (Lugtenberg and Kamilova, 2009). The direct mechanism of PGPR in biocontrol involves antagonism to soil borne pathogens (Supplementary Data Sheet 1) and indirect mechanism depends on induction of systemic resistance (Choudhary, 2011; Glick, 2012). Generally competition for nutrients, niche exclusion, ISR and antifungal metabolites production are some of the chief modes of biocontrol activity in PGPR (Lugtenberg and Kamilova, 2009). Many rhizobacteria have been reported to produce antifungal metabolites like hydrogen cyanide (HCN), phenazines, pyrrolnitrin, 2, 4-diacetylphloroglucinol, pyoluteorin, viscosinamide, and tensin (Bhattacharyya and Jha, 2012). In 2009, De Vleesschauwer and Höfte (2009) coined the term ISR for resistance induced by PGPR which was found to be independent of the pathway involved.

In case of microbially induced resistance the best option is the use of plant growth promoting rhizobacteria (PGPR) as potential elicitors of plant defense mechanisms (Liu et al., 1995). On the onset of plant colonization by the rhizobacteria, metabolic changes may occur in the host, i.e., production of phytoalexins (Van Peer et al., 1991), accumulation of pathogenesis related (PR) Proteins (Zdor and Anderson, 1992) or deposition of structural barriers, etc. (Benhamou et al., 1996a,b).

Origin of elicitor compounds may be biological (plant or microbe derived) or synthetic like beta-amino-butyric acid (BABA), cis-jasmone and acibenzolar-S-methyl (ASM) (Walters et al., 2013). Plants generally recognize three types of chemical elicitors viz; microbe-associated molecular patterns (MAMPs) derived from beneficial microbes, pathogen-associated molecular patterns (PAMPs) released by pathogenic microbes and damageassociated molecular patterns (DAMPs) produced by plants on injury by insects or herbivores or even during microbial degradation (Henry et al., 2012; Newman et al., 2013). These aforementioned molecules are called "patterns that elicit immunity" (PEIs) which are recognized by plants through transmembrane pattern recognition receptors (PRRs) (Jones and Dangl, 2006; Maffei et al., 2012; Newman et al., 2013). After recognition of MAMPs or DAMPs elicitors pattern-triggered immunity (PTI) is activated in plants. This stimulation of defense reaction restricts the pathogen making plant resistant to additional pathogen attack through the mechanism of induction of systemic resistance (Henry et al., 2012).

Bacteria belonging to the genus Pseudomonas involve pathogen inhibition via competition and/or antagonism (Haas and Défago, 2005) and by developing direct interactions with the host plants through ISR (Bakker et al., 2007). Pseudomonas aeruginosa PM12, used in this study was isolated from healthy tomato roots from vegetable garden of University of the Punjab, Lahore, Pakistan (Fatima and Anjum, 2016). It was characterized on molecular grounds through sequencing of 16S rRNA (900bp) and alignment at GeneBank (NCBI, MaryLand), allotted accession number KT966743 (Fatima and Anjum, 2017).

Induced systemic resistance is involved in synthesis of enhanced levels of various secondary metabolites engaged in plant defense mechanisms. These include phytoalexins that increase plant resistance through their toxic action against pathogens. Phytoalexins, i.e., sesquiterpenoids are synthesized by the members of Solanaceae and isoflavonoids are produced in individuals of the Papilionaceae. They inhibit germination of fungal spores and retard fungal growth. Plant tissues exhibiting ISR show increased activity of phenylalanine ammonia lyase (PAL) that leads to resistance response in plants (Verhagen et al., 2010).

Metabolomics is one of the most rapidly growing areas of contemporary science. This technology is now being used to route out reprograming and metabolic fluctuations in plant pathways (De Vos et al., 2007; Zhao et al., 2015). Current investigation revealed that compatible host pathogen interactions are characterized by a lower level of certain defense-related mechanisms compared with host pathogen interactions in the presence of bacterial ISR elicitor that leads to a more dynamic metabolic response over the course of colonization. Furthermore, our results demonstrate that elicitor (HMB) from P. aeruginosa (PM12) induces production of secondary metabolites involved in defense pathways of tested plant.

### MATERIALS AND METHODS

### Fungal and Bacterial Strains

Virulent strain of F. oxysporum isolated from diseased tomato plants growing in vegetable garden of University of the Punjab, Lahore was cultured on potato dextrose agar (PDA difco) for 10 days. Conidia were harvested by gentle scraping in sterile water and pathogen inoculum was prepared by adjusting the concentration to 10<sup>5</sup> conidia/ml using haemocytometer. P. aeruginosa (PM12) was grown on LB broth medium (100 ml) for 24 h at 35◦C. For collecting extracellular metabolites culture was pelleted by centrifugation at 4000 g for 15 min, and the supernatant obtained was processed for ISR assay. Intracellular metabolites were extracted using sonication. Bacterial cell lysis was performed six times through sonication at resonance amplitude for 15 s at 4◦C.

### Preliminary Screening of ISR Determinants from P. aeruginosa PM12

This study was performed to identify the involvement of intracellular metabolites and cell-free culture filtrates (CFCF) of P. aeruginosa for inducing systemic resistance against fungal wilt in tomato. Two-weeks old seedlings of Fusarium wilt susceptible tomato variety "Rio-Grande" seeded in sterilized pot media. In total, 50 ml of both extracts (intracellular and extracellular) were supplied to the allocated plastic pots containing 0.5 kg sterilized soil and after a period of 3 days these pots were inoculated with 50 ml of pathogen @ 10<sup>5</sup> conidia/ml. Positive control consisted of P. aeruginosa PM12 suspension made in water by adjusting the concentration to 10<sup>4</sup> cfu/ml. Fifty milli liter of sterile distilled water was provided to the pots designated as untreated control. All the pots were incubated for 14 days under greenhouse conditions. There were ten replicates against each treatment and experiment was conducted twice. To determine the disease index (DI), wilting was scored based on the criteria developed by Epp (1987) (0 = no wilt symptoms; 1 = less than 25% of the plant turned yellow; 2 = yellowing and browning covered nearly 50% of plant; 3 = whole plant turned brown and died). The equation described by Cachinero et al., 2002 was used to calculate the DI.

$$\text{DI} = \{ (\text{ $\Sigma$  ni} \times \text{si})/(\text{N} \times \text{S}) \} \times 100$$

where, ni = the number of tomato plants with wilt symptoms, si = value of the symptom score, N = the total number of tested plants, and S = the highest value of the symptom score.

### Isolation of Bioactive Compound(s) from Extracellular Metabolites of P. aeruginosa (PM12)

Methodology proposed by Sumayo et al. (2013) was used to isolate bioactive compound (s) from extracellular metabolites of P. aeruginosa (PM12). Extracellular metabolites of P. aeruginosa PM12 were obtained as described previously and extracted twice with double volumes of organic solvents such as ethyl acetate, chloroform, n-hexane and n butanol. All the organic extracts were dried in a rotary evaporator at 50◦C, mixed in 10% DMSO and subjected to ISR experimentation.

### Primary Screening of Bacterial Metabolites Extracted with Various Solvents for ISR Activity

Seeds of the variety "Rio-Grande" were sown in sterilized media and after 2 weeks of emergence seedlings were transplanted in plastic pots (4 inches diameter) containing sterilized sandy loam soil. Pots were subjected to pathogen inoculum and bacterial metabolites extracted with various organic solvents for each solvent separately. After 1 week of incubation data regarding the disease development was recorded and the solvent treatment found most suitable for the disease reduction was further screened for ISR determinant/s.

### Column Chromatography of the Active Metabolites and Selection of Sub Fraction Involved in ISR

Bacterial metabolite exhibiting highest ISR activity was further fractioned through silica gel column chromatography. Column was washed with double volumes of methanol and ethyl acetate. The extract was fractioned using stepwise elution method with increased concentration of methanol in ethyl acetate. Obtained eluates were evaporated and mixed in 10% DMSO and checked for ISR activity in a test tube assay. Negative control received pathogen inoculum with 10% DMSO. Seeds of the variety "Rio-Grande" were surface sterilized with 1% NaOCl and sown in culture tubes containing Murashige and Skoog (MS) medium. Tubes were incubated in a growth chamber at 25◦C for seedling development. Seedlings were then provided with different bioactive fractions for ISR activity separately. After 2 days of seedlings treatment, 10 µl of pathogen inoculum was provided to each tube @ 10<sup>5</sup> condia/ml. After 1 week of incubation data was recorded regarding DI. Experiment was repeated twice and there were ten replicates for each treatment.

### Identification of ISR Determinant(s) by GC-MS Analysis

Induced systemic resistance active sub-fraction obtained from silica gel chromatography was further subjected to instrumental analysis for identification of potential ISR determinant/s. This study was executed in Agilent GC/MS apparatus including capillary column (0.25 ID × 30 M × 0.25 µM film thicknesses) in which electron ionization was used as ion source. Helium with flow rate of 1.0 ml/min was supplied as carrier gas. Column temperature of apparatus was maintained at 30◦C for 3 min which was increased at 50◦C/min to 180◦C and by 40◦C/min to 200◦C.

### ISR Bioassay with Pure Compounds

Another independent experiment was performed to screen ISR active biochemicals from P. aeruginosa. Biochemicals found in ISR active sub-fraction were purchased from the market. Three concentrations, i.e., 0.1, 1.0 and 10 mM of each of the pure biochemicals were applied as soil drench method, application of inoculum (bacterial/pathogen origin) to the soil surrounding

the roots of transplanted seedlings of tomatoes, at the rate of 50 ml/pot (Algam et al., 2005). Data regarding DI was taken as described earlier after 15 days of treatment applications.

### Analysis of Metabolite Profile of Tomato Plants in the Presence of ISR Elicitor from P. aeruginosa PM12

It was found in the previous experiment that HMB acted as the most active compound involved in elicitation of ISR in tested tomato plants therefore another independent experiment was carried out to study metabolic transitions brought about in the tomato plants under the influence of HMB elicitor.

### Plant Growth and Inoculation

Seedlings of the tomato variety 'Rio-Grande' were raised in sterilized sandy loam soil media. After 3 weeks of emergence seedlings were transplanted @ 1 seedling/ pot in plastic pots (4 inches diameter) containing sterilized sandy loam soil. Four treatments were made in this experiment viz; T1 = Plants receiving ISR elicitor (HMB) + FOL, T2 = Plant receiving ISR elicitor (HMB) alone, T3 = Plant receiving FOL to serve as pathogen control, T4 = Plants receiving distilled sterilized water to serve as untreated control.

After 3 days of seedling transplantation, 50 ml of 10 mM ISR elicitor (Pure compound) was supplied to the tomato plants. After 2 days of treatment, FOL inoculum was provided @ 50 ml of conidial suspension (see Fungal and Bacterial Strains). Plants were incubated under greenhouse conditions. Experiment was conducted twice having 15 plants in each replicate.

### Extraction of Plant Metabolites

After 1 week of pathogen challenge to the seedlings plant metabolites were extracted. Leaf samples (1 g) from young shoots were taken from treated and untreated plants separately. Samples were crushed in liquid nitrogen and the resulting powder was dissolved in 10 ml of extraction solution comprising of methanol, chloroform and water mixed in a ratio of 80:10:10, respectively. Dissolved material was kept overnight at room temperature. Afterward metabolite extraction was done using micro filters.

### **Derivatization**

Derivatization transforms a chemical compound into its derivative having similar chemical structure suitable for GC-MS analysis. Plant extract (0.3 ml) was mixed with 0.1 ml of ribitol (internal standard) inside a glass container. It was dried using N<sup>2</sup> gas. After drying added 25 µl of methoxyamine hydrochloride (MOX), mixed it and left overnight at room temperature. Afterward, 80 µl of N-methyl-N trimethylsilyltrifluoroacetamide (MSTFA) was supplied to the solution and left for 2 h at room temperature (Warth et al., 2014).

### **GCMS Analysis of Derivatized Plant Metabolites**

Gas chromatography-mass spectrometry set up consisted of capillary column (0.25 ID × 30 m × 0.25 µm film thicknesses) and electron ionization was used as an ion source. Helium was used as a carrier gas at the flow rate of 1.0 ml/min. Temperature of the capillary column was adjusted at 30◦C for 3 min which was raised to 50◦C /min to 180◦C and by 40◦C/min to 200◦C / min. Derivatized plant extract was injected in GC-MS machine @ 1 µl.

### **Plant Metabolites Analysis**

Metabolite analysis of tomato was performed using method devised by Lisec et al. (2006). Metabolite identification was carried out by comparing spectrum with NIST library. Levels of different metabolites were determined through Mzmine software package<sup>1</sup> ; values obtained were log2-transformed (Steinfath et al., 2008) and normalized to show identical medium peak sizes per sample group. ClustVis<sup>2</sup> online tool was used to make heatmap.

### Statistical Analysis

The data was analyzed by performing ANOVA, and the significance of difference between the treatments was determined by DNMRT at P < 0.05 using software DSASTAT (Onofri, 2007).

## RESULTS

### Initial Screening of Metabolites Involved in ISR from Pseudomonas aeruginosa PM12

Extracellular metabolites of P. aeruginosa (PM12) were found to be actively involved in suppressing Fusarium wilt of tomato under greenhouse conditions (**Table 1**). Significant reductions in the DI up to 73% were recorded in case of the extracellular metabolites that was almost equal to the efficacy of live cells, i.e., 75.86%. Intracellular metabolites were not found capable in suppressing the disease to a significant level. Resultantly, it was concluded that extracellular metabolites are involved in disease suppression and contains ISR determinant/s (**Figure 1**).

### Searching of ISR Determinants from Extracellular Metabolites of Bacterial Strains

In this experiment extracellular metabolites/CFCF of P. aeruginosa PM12 were examined for the presence of ISR determinant/s. Extraction of extracellular metabolites was done by using different organic solvents and then applied to the plants

<sup>1</sup>http://mzmine.sourceforge.net/

<sup>2</sup>http://biit.cs.ut.ee/clustvis/



Capital letters represents levels of significance as governed by ANOVA and DNMRT at (P > 0.05). Values with ± signs represent standard error between different replicates of same treatment.

as soil drench method and was challenged with the pathogen afterward. Compound/s present in ethyl acetate fraction of extracellular metabolites significantly reduced wilt disease in tomato plants when observed on visual grounds (**Figure 2**).

### Identification of ISR Determinant by GC/MS Analysis

Ethyl acetate fraction of extracellular metabolites exhibiting maximum ISR activity was further processed via column chromatography using step wise elution method and ISR experiment was performed in a test tube assay. Sub-fraction showing maximum disease suppression was then analyzed through GC/MS. Chromatogram obtained from GC/MS analyses showed three peaks that were identified as 3-hydroxy-5-methoxy benzene methanol (HMB), eugenol and tyrosine, respectively (**Figure 3**).

Biochemicals obtained were again tested for ISR activity in their pure form. The ISR bioassay showed that only HMB significantly (P ≤ 0.05) reduced disease severity to 76.25%

at 10 mM concentration. Whereas eugenol and tyrosine were not found efficient in reducing disease severity at the same concentration (**Figure 4**).

### Metablomic Analysis of Tomato Plants

This study explored metabolic transitions observed in the tomato variety "Rio-Grande" under the stimulus of ISR elicitor (HMB) using GC/MS analysis. Whole metabolome of plants inoculated with HMB and pathogen in either combination were studied to elucidate ISR mechanism. In four different treatments, 94 metabolites were detected some of which were up regulated and some were down regulated as shown by different color schemes in heat map (**Figure 5**). Metabolite values obtained using Mzmine software were log2-transformed (Steinfath et al., 2008). Heat map was generated using online ClustVis tool with row wise scaling and correlation-based clustering. Most pronounced changes were depicted in metabolite levels of the plants receiving elicitor + pathogen making plants more resistant against pathogen attack by timely accumulation of defense related chemicals (**Figure 5**). Changes brought about in central metabolites

FIGURE 4 | Influence of soil drench application of pure biochemicals present in ISR active sub-fraction on the disease development on tomato plants after inoculation with Fusarium wilt pathogen. HMB = 3-hydroxy-5-methoxy benzene methanol. Vertical bars represents standard errors. Asterisks indicate statistically significant reduction in disease index as compared to pathogen control as governed by ANOVA at (P < 0.05).

under the stimulus of different treatments were normalized to respective control and were expressed in fold change values. Most metabolic fluctuations were documented in primary and secondary metabolism, signaling and defense pathways. Treatment receiving elicitor and pathogen showed variation in primary metabolism with different intensities. Upregulation of various metabolites involved in phenylpropanoid pathway was documented in case of the elicitor, e.g., 4-hydroxybenzene,

cinnamate and tryptophan were increased to 3.19-, 2.31-, and 1.24-folds, respectively (**Figure 5**). Similarly plants treated with elicitor and pathogen significantly enhanced levels of most of the metabolites taking part in amino acid metabolism and TCA pathway (**Figure 5**). These findings illustrate that synergistic effect of the elicitor and the pathogen in inducing systemic resistance was more pronounced due to the active participation of the pathogen in some cases or elicitor in other.

Plants treated with the elicitor (HMB) showed enhanced levels of fructose, glucose and sucrose to 3.4-, 3.1-, and 1.2-folds, respectively. Whereas concentration of trehalose was found to be 0.9-fold less as compared to the untreated control (**Figure 5**).

In case of phosphorylated metabolites there was an increasing trend in the concentration of fructose-6- phosphate, glucose-6-phosphate and myo-inositol-phosphate to 3.9-, 3.3-, and 1.9-folds, respectively. Whereas concentration of glycerol-3 phosphate (0.9-fold) was decreased to 0.9-fold in the presence of the elicitor in comparison to the untreated control.

Plants treated with the elicitor showed increased levels of organic acids including α ketoglutarate (2.7-fold), malate (2.8 fold), oxaloacetate (2.3-fold), citrate (2.1-fold) and succinate (1.4 fold). Whereas low levels of fumarate (0.4-fold) and cis- aconitate (0.3-fold) were recorded in comparison to the untreated control.

Elicitor treated plants also showed raised quantities of polyamines like picolinic acid (7.7-fold), pipecolic acid (4.7-fold), putrescine (2.8-fold), and spermidine (1.4-fold).

Amino acids like arginine, glutamine and proline levels were significantly increased to 3.04, 2.7 and 2.32-folds in case of the elicitor whereas leucine, tyrosine and threonine were decreased in comparison to the untreated control (**Figure 5**). Methionine acts as a precursor for ethylene biosynthesis through S-adenosyl methionine (SAM) (Bleecker and Kende, 2000). Methionine levels were significantly raised in plants under the influence of elicitor and elicitor + pathogen treatments.

Levels of salicylic acid were significantly upregulated in all the treatments in comparison to the untreated control. Plants treated with the elicitor and the pathogen showed increased levels of phenylalanine to 2.8-folds whereas application of the elicitor resulted in an increase of 1.2-folds as compared to the non-treated control. Metabolic pathways were drawn to show transitions in various metabolites in the presence of elicitor HMB (**Figures 6–8**). Metabolites belonging to the phenylpropanoid, glycolysis, photosynthesis and signal transduction pathways were significantly increased by the elicitor (HMB) in treated tomato plants.

Alterations in the metabolites of the plants receiving elicitor and elicitor + pathogen revealed that HMB elicitor from P. aeruginosa is responsible for inducing systemic resistance against Fusarium wilt in tomato.

### DISCUSSION

### Isolating ISR Determinant(s) from Pseudomonas aeruginosa PM12

Recent attempts in exploiting non-pathogenic rhizospheric microbes in biocontrol programs have shown their involvement as antagonists and plant defense stimulators. However, the exact nature of biochemicals imparting systemic resistance in plants against diseases remains poorly understood. This research was performed to isolate potential ISR determinant/s from P. aeruginosa (PM12) against F. oxysporum f. sp. lycopersici in tomato plant.

Pseudomonas aeruginosa PM12 is a non-pathogenic strain that was isolated from tomato rhizosphere. To avoid the use of whole bacterium as it is categorized as opportunistic pathogen we focused on the signal compound/s involved in eliciting defense mechanisms of the plant and identified the ISR determinant from extracellular metabolites of P. aeruginosa PM12. Moreover opportunistic pathogens are microbes that are capable of causing disease only in people who are immunocompromised or are otherwise especially susceptible. A critical issue concerns the pathogenic strains proposed for registration as biopesticides, since these strains are typically isolated from the environment, for example agricultural fields, rather than as clinical specimens. As such, these strains have no history of actually causing disease and may not be able to do so (GPO, 1999).

Extracellular metabolites and live cells of P. aeruginosa PM12 showed significant (P ≤ 0.05) protection against Fusarium wilt whereas the intracellular metabolites were unable to provide any considerable control (**Table 1**). The ethyl acetate fraction of extracellular metabolites of the bacterium exhibited maximum ISR activity and was sub fractioned using column chromatography. The active sub fraction was then analyzed through GC/MS. The chromatogram obtained identified three compounds, i.e., HMB, tyrosine and eugenol. Reconfirmation of active biochemical was achieved using pure compounds which revealed HMB as the most capable component in mitigating Fusarium wilt in tomato.

The elicitor compound, HMB, isolated from P. aeruginosa (PM12), is not yet reported for its biological activity against Fusarium wilt so this is the first finding describing its involvement in systemic resistance in tomato. Colonization of plant roots with P. fluorescens enhanced plant defenses against pathogen by sensitizing the plant through signal compounds (Verhagen et al., 2004). Literature showed the use of benzene in making pesticides, detergents, drugs, dyes, rubbers, explosives and lubricants (Ceresana, 2014). Furthermore antifungal and antimicrobial activity of poly substituted benzene derivatives has been reported. The substituent attached to the benzene ring and its position affects the activity of the compound against pathogen (Uyanik et al., 2009). Similarly elicitor benzothiazole isolated from Pseudomonas sp. has shown inhibition of mycelial growth and germination from ascospores and sclerotia of Sclerotinia sclerotiorum (Fernando et al., 2005). Aldehydes, ketones and benzenes from Bacillus amyloliquefaciens have been known to restrict mycelial growth and spore germination of F. oxysporum (Yuan et al., 2012). Bacterial volatile compounds help in maintaining plant health through ISR and production of antifungal compounds (Weisskopf, 2013). As benzene is involved in protection of plants against different pests so it is concluded that benzene is the actual molecule in elicitor compound HMB for imparting resistance against FOL in tomato.

A considerable diversity of the elicitors involved in ISR has been observed in case of Pseudomonads. They include the building blocks of bacterial cells as well as extracellular compounds synthesized by the microbes (Jankiewicz and Kol-tonowicz, 2012). In current investigation HMB was discovered as an ISR determinant from P. aeruginosa PM12. Nowadays awareness regarding concept of immunity development in plants by the use of beneficial microbes is gaining momentum. In conventional agriculture system the identification of MAMPs from beneficial microbes and their

use in formulation stimulates resistance in plants against pest attack.

Use of biocontrol agents in elicitation of resistance in plants is gaining importance these days. Some strains of Pseudomonas are clearly potent inducers of systemic resistance in plants (Van Wees et al., 1999; Meena et al., 2000).

### Mechanism behind Systemic Resistance Induced by Pseudomonas aeruginosa PM12 in Tomato Plants

Biological inducers have been known to elicit metabolites involved in different pathways like pentose pathway, signaling transduction and tricarboxylic acid cyclic, etc. (Verhagen et al., 2004; Shoresh and Harman, 2008; Damodaran et al., 2010; Brotman et al., 2012). In second phase of this study heat map was developed to show variations in metabolites under the influence of HMB and accentuated in physiological pathways. In this investigation upregulation of metabolites involved in phenylpropanoid pathway may lead to resistance development in tomato plants treated with the elicitor HMB.

A significant decrease in fumaric acid was recorded in the elicitor and elicitor + pathogen treated plants. These results are in connection with the findings of Doornbos et al. (2009) who inoculated Arabidopsis plants with the PGP and ISR-inducing bacterium WCS358r and observed decrease in intermediates of Krebs cycle that showed an increased demand for carbon respiration.

Brotman et al. (2012) reported that putrescine induction by Trichoderma spp. was related to growth promotion in Arabidopsis thaliana. Polyamines were also involved in various defense responses (Tonon et al., 2004) similarly in this study upregulation of putrescine was noted.

In plant microbe interactions, sugars serve as signaling molecules for resistance induction against the pathogens and this resistance is termed as "sweet immunity" (Bolouri Moghaddam and Van den Ende, 2013). Like sucrose is involved in activating plant immune responses against pathogens (Tauzin and Giardina, 2014). Sucrose is known to trigger the production of isoflavonoids against F. oxysporum in lupine (Morkunas et al., 2011). Trehalose stimulates the activity of phenyl ammonia lyase and peroxidase against powdery mildew in wheat (Reignault et al., 2001). Fructose, sucrose and glucose elicited PR-protein transcripts in tobacco in a SAindependent pathway (Herbers et al., 1996). Moreover Kawano and Shimamoto (2013) and Yamaguchi et al. (2013) found that chitin is responsible for stimulation of Mitogen-Activated Protein Kinase (MAPK) cascade, lignin and phytoalexins production in rice. In this study significant increase in quantities of some carbohydrates like fructose, glucose, mannose, etc., were also

documented that are reportedly involve in plant growth and development.

Salicylic acid is involved in activation of defense responses in plants after challenging with a pathogen (Klessig et al., 2000). In a study enhanced levels of salicylic acid were documented when plants were challenged with pathogen or with P. fluorescens (Kumar and Sharma, 2013). Our results also showed upregulation of salicylic acid in the presence of the bacterial elicitor HMB.

Glutamate levels were significantly raised in treatments receiving the elicitor or elicitor and pathogen in combination. These results are in line with findings of Brotman et al. (2012). Glutamate acts as signaling compound in nitrogen signaling pathway (Forde and Lea, 2007).

Biochemicals like polyamines are involved in different physiological processes of the plants (Tonon et al., 2004). In this study plants primed with the elicitor (HMB) showed significant increase in polyamine levels. Likewise an increasing trend was

fpls-08-00848 May 30, 2017 Time: 16:58 # 11

also observed in organic acids in the HMB + pathogen treated plants. Shikimate pathway is stimulated by the synthesis of aromatic compounds which is responsible for defense response of plants (Tzin and Galili, 2010). Moreover shikimic acid concentration was also found to be enhanced under the effect of the bacterial elicitor HMB. Shikimic acid and malonic acid pathways are involved in phenolics synthesis leading to plant defense against pathogens. Phenolics include a variety of defense related compounds like anthocyanins, phytoalexins, flavonoids, furanocoumarins tannins and lignin (Freeman and Beattie, 2008). Present study illustrated the significance of metabolic dynamics engaged in defense mechanisms of the tomato plants infected with F. oxysporum ensured triggering of ISR by the bacterial elicitor HMB.

Phenylpropanoids like lignin, phytoalexins, flavonoids, tannins, etc., are the secondary metabolites of the plants produced under stress conditions (Vogt, 2010). They provide resistance in plants against phytopathogens through accumulation of lignin that acts as a barrier for penetration in plant tissues. Enzymes involved in phenylpropanoid pathways viz; PPO, PO and PAL impart rigidity and mechanical strength to the plant cell walls through production of lignin metabolites that act as barrier for invading fungal pathogens (Mohammadi and Kazemi, 2002; Anjum et al., 2012). Metabolic dynamics engaged in defense mechanisms of the tomato plants infected with F. oxysporum ensured elicitation of ISR by the bacterial elicitor HMB. Enhanced metabolite upregulation under influence of the elicitor + pathogen treatment may be due to the synergistic effect of ISR and SAR that may lead to better cope with the stress condition of the plant by activation of defense machinery.

### REFERENCES


### CONCLUSION

Findings of our study revealed that priming of plant roots with P. aeruginosa helps in suppressing Fusarium wilt by releasing an ISR elicitor HMB. Furthermore this elicitor compound is engaged in elicitation of defense mechanisms in tomato plants through metabolic transitions that leads to resistance induction against Fusarium wilt pathogen. Research regarding targeting of genes controlling the regulation of this elicitor compound will undoubtedly help in pest management practices in a sustainable manner.

### AUTHOR CONTRIBUTIONS

SF and TA designed the study. TA revised the manuscript critically. All authors read and approved the final version of the manuscript.

### ACKNOWLEDGMENT

We are thankful to Forman Christian College, Lahore, Pakistan for providing us assistance in GC/MS analysis.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.00848/ full#supplementary-material




Combating them: Science, Technology and Education, ed. A. M. éndez-Vilas (Badajoz: Formatex Research Center), 1352–1363.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Fatima and Anjum. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Insights into the Mechanism of Proliferation on the Special Microbes Mediated by Phenolic Acids in the Radix pseudostellariae Rhizosphere under Continuous Monoculture Regimes

Hongmiao Wu1,2† , Junjian Xu1,2† , Juanying Wang1,2, Xianjin Qin2,3, Linkun Wu1,2 , Zhicheng Li1,2, Sheng Lin1,2, Weiwei Lin1,2, Quan Zhu1,2, Muhammad U. Khan1,2 and Wenxiong Lin1,2,3 \*

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Birinchi Kumar Sarma, Banaras Hindu University, India Pratyoosh Shukla, Maharshi Dayanand University, India

#### \*Correspondence:

Wenxiong Lin wenxiong181@163.com †These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 20 October 2016 Accepted: 11 April 2017 Published: 02 May 2017

#### Citation:

Wu H, Xu J, Wang J, Qin X, Wu L, Li Z, Lin S, Lin W, Zhu Q, Khan MU and Lin W (2017) Insights into the Mechanism of Proliferation on the Special Microbes Mediated by Phenolic Acids in the Radix pseudostellariae Rhizosphere under Continuous Monoculture Regimes. Front. Plant Sci. 8:659. doi: 10.3389/fpls.2017.00659 <sup>1</sup> Fujian Provincial Key Laboratory of Agroecological Processing and Safety Monitoring, College of Life Sciences, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> Key Laboratory of Crop Ecology and Molecular Physiology, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>3</sup> Key Laboratory of Ministry of Education for Genetics, Breeding and Multiple Utilization of Crops/College of Crop Science, Fujian Agriculture and Forestry University, Fuzhou, China

As potent allelochemicals, phenolic acids are believed to be associated with replanting disease and cause microflora shift and structural disorder in the rhizosphere soil of continuously monocultured Radix pseudostellariae. The transcriptome sequencing was used to reveal the mechanisms underlying the differential response of pathogenic bacterium Kosakonia sacchari and beneficial bacterium Bacillus pumilus on their interactions with phenolic acids, the main allelochemicals in root exudates of R. pseudostellariae in the monoculture system. The microbes were inoculated in the pots containing soil and the medicinal plant in this study. The results showed that the addition of beneficial B. pumilus to the 2-year planted soil significantly decreased the activity of soil urease, catalase, sucrase, and cellulase and increased the activity of chitinase compared with those in the 2nd-year monocropping rhizosphere soil without any treatment. However, opposite results were obtained when K. sacchari was added. Transcriptome analysis showed that vanillin enhanced glycolysis/gluconeogenesis, fatty acid biosynthesis, pentose phosphate, bacterial chemotaxis, flagellar assembly, and phosphotransferase system pathway in K. sacchari. However, protocatechuic acid, a metabolite produced by K. sacchari from vanillin, had negative effects on the citrate cycle and biosynthesis of novobiocin, phenylalanine, tyrosine, and tryptophan in B. pumilus. Concurrently, the protocatechuic acid decreased the biofilm formation of B. pumilus. These results unveiled the mechanisms how phenolic acids differentially mediate the shifts of microbial flora in rhizosphere soil, leading to the proliferation of pathogenic bacteria (i.e., K. sacchari) and the attenuation of beneficial bacteria (i.e., B. pumilus) under the monocropping system of R. pseudostellariae.

Keywords: bacteria, microbe, monocropping, phenolic acid, rhizosphere

### INTRODUCTION

fpls-08-00659 April 29, 2017 Time: 12:26 # 2

Radix pseudostellariae L. belongs to the Caryophyllaceae family. It is a common and popular medicine in China. It is mainly produced in Fujian Province in the southeast of China. R. pseudostellariae contains ginseng saponins, polysaccharides, amino acids, flavonoids, and cyclic peptides, which can be used as a cure for spleen deficiency, anorexia, and palpitations (Zhao W.O. et al., 2015). The typical average annual yield of R. pseudostellariae is about 5000 tons, accounting for more than 22 million dollars per year. However, this medicinal plant is also affected by the replanting disease, in which the consecutive monoculture of R. pseudostellariae leads to a serious decline in the biomass and quality of underground tubers. Consecutive monoculture of R. pseudostellariae caused the yield reduction by 33.3%, the polysaccharide content and ginseng saponins Rb1 of the tuberous root were reduced by 88.08 and 44.33%, respectively (Zeng et al., 2012). Zhang and Lin (2009) and Wu et al. (2016) reported that more than 70% of medicinal plants were attacked by the replanting diseases, leading to serious and negative influences on both the soil and the plant health. Moreover, farmers have increased the amount of pesticides and fertilizers to find the solution to these problems, at the cost of excessive pesticide residues, more capital investment, soil degradation, and environmental pollution. Therefore, it has become a top priority to elucidate the underlying mechanisms of replanting disease, especially in the case of medicinal plant production.

Root exudates are perceived as chemical signals of communication between roots and microorganisms (Hirsch et al., 2003; Neal et al., 2012; Zhang et al., 2013; Yuan et al., 2015; Rugova et al., 2017; Vergani et al., 2017). A growing body of evidence suggests that the imbalanced microbial populations mediated by the root exudates play crucial roles in replanting disease (Wu et al., 2015). As one of the important root exudate, phenolic acids are involved in allelopathy and cause replanting disease (Bais et al., 2004, 2006; Kulmatiski et al., 2008; Liu et al., 2017). The phenolic acids are prone to change the soil microbial community and indirectly cause autotoxicity to the monocultured plants (Zhou and Wu, 2011; Zhou et al., 2012; Li et al., 2014; Liu et al., 2017). Previous results have also shown that consecutive monoculture of R. pseudostellariae can significantly increase the number of pathogenic microorganisms such as Fusarium oxysporum, Talaromyces helicus, Kosakonia sacchari, and so forth in the rhizosphere. Phenolic acids could promote the growth of these soil-borne pathogens (Zhao Y. et al., 2015; Wu et al., 2016). However, these phenolic acids did not show any autotoxicity toward R. pseudostellariae. Besides, previous studies demonstrated that K. sacchari produced protocatechuic acid when consuming phenolic acids and its metabolite, and protocatechuic acid negatively affected the growth of beneficial bacterium Bacillus pumilus (Wu et al., 2016). The pathogenic bacterium K. sacchari and the beneficial bacterium B. pumilus were isolated from the rhizosphere soil of the monocultured R. pseudostellariae. They are thought to be involved in the replanting disease of R. pseudostellariae (Wu et al., 2016). The studies also depicted that B. pumilus was not pathogenic to R. pseudostellariae but suppressed the mycelial growth of pathogens such as F. oxysporum and T. helicus. However, the intrinsic mechanism of the relationship between these special microbes and the root exudates is still not clarified.

Sequencing and analysis of expressed sequence tags have been widely used by most researchers for molecular marker discovery and gene expression profiling (Venturini et al., 2013; Mudalkar et al., 2014). The comparative transcriptome sequences were used in this study to identify the differentially expressed genes (DEG) in the bacteria under the treatment of phenolic acids in root exudates from the monocultured R. pseudostellariae. The influence of special microbes (K. sacchari and B. pumilus) on the physiological characteristics of R. pseudostellariae and its monoculture rhizosphere soil was also analyzed. This study focused on the underlying mechanisms related to the imbalance of microbial populations mediated by phenolic acids. It aimed to provide useful information and insights into the molecular ecological mechanism of destabilization of microbial populations mediated by phenolic acids present in secreted root exudates.

### MATERIALS AND METHODS

### R. pseudostellariae Plant Preparation

The tubers of R. pseudostellariae were transplanted into pots (13-cm bottom diameter, 22-cm top diameter, and 20 cm height) containing different soils, which were sampled from the upper soil layer (0–15 cm) of the fields in the experimental station of Fujian Agriculture and Forestry University, Fuzhou, China. One of the soil samples was called as newly planted soil, which was never planted with R. pseudostellariae before, and the others were sampled from the field monocultured with R. pseudostellariae plants for 1 year. Each pot had four R. pseudostellariae plants. The young plants were grown in pots and incubated in the greenhouse.

The beneficial bacterium B. pumilus and the pathogenic bacterium K. sacchari of R. pseudostellariae were grown in Luria–Bertani (LB) medium at 37◦C and 200 rpm. The specific bacterial cells were collected and washed three times with doubledistilled water until an OD600 of 1 was reached, and then resuspended in the same double-distilled water (OD600 = 1). The bacterial cells were added to the two kinds of soils when R. pseudostellariae plants reached the five-leaf stage. In the sequential pot experiment, specific bacterial treatments were given to soils four times, which were called as newly planted soil and continuously monocultured soil, as mentioned earlier and in **Table 1**. The inoculum at the time of each treatment contained 60 mL of the bacterial cells in total. The pot experiment was conducted in three replicates. The details of the treatments are shown in **Table 1**.

### Physicochemical Properties of R. pseudostellariae and Rhizosphere Soil under the Treatments

Plant Chlorophyll Content and Soil Enzyme Activities The soil urease activity was determined using the phenol–chloroform method and expressed as mg NH3·N


TABLE 1 | Different treatments of Radix pseudostellariae.

CK, unplanted soil; FY, first cropping year; SY, second cropping year; BP, Bacillus pumilus in the second cropping year; KS, Kosakonia sacchari in the first cropping year.

g −1 soil 24 h−<sup>1</sup> . The cellulase activity was detected by the colorimetric anthracenone method and expressed as mg glucose g −1 soil 24 h−<sup>1</sup> . The sucrase activity was determined by the 3,5-dinitrosalicylic acid anthracenone method and expressed as mg glucose g−<sup>1</sup> soil 24 h−<sup>1</sup> . The soil dehydrogenase activity was analyzed using the 2,3,5-triphenyltetrazolium chloride method on an oven-dried soil and expressed as mg TF g−<sup>1</sup> soil 24 h−<sup>1</sup> . The chlorophyll content of R. pseudostellariae and the soil acid protease, chitinase, acid phosphatase, and catalase activities were determined using a plant and soil kit (Comin, Suzhou, China) according to the manufacturer's protocol.

### Phenolic Extraction and Determination

Soil phenolics were extracted as previously described (Zhou and Wu, 2011). Briefly, 5 g of soil was shaken in 25 mL of 1M NaOH for 24 h (200 rpm, 30◦C) and then spun in a vortex generator for 30 min at maximum speed. After centrifugation at 10,000 rpm for 15 min, the supernatant was acidified to 2.5 with 9M HCl and extracted with ethyl acetate. The residue was dissolved in 5 mL of methanol using ultrasound for 5 min and maintained in the dark at 4◦C.

The methanol solution of soil extracts was analyzed with the Waters HPLC system (C18 column: Inertsil ODS-SP, 4.6 × 250 mm, 5 µm). The mobile phase was a mixture of methanol and 2% acetic acid. Detection was performed at 280 nm. Protocatechuic acid and vanillin were identified and quantified by comparing the retention time and area with the pure standard.

### Transcriptome Analysis of Bacteria Bacterial Cell and RNA Extraction

The LB liquid culture medium was diluted six times and sterilized for 20 min. The culture medium was cooled, an appropriate amount of each protocatechuic acid and vanillin, passed through a 0.22-µm ultrafiltration membrane, was added to make the concentration up to 120 µmol/L, which was the closest to actual conditions detected in the rhizosphere soil. The K. sacchari culture (50 µL), which had already been activated, was added to the vanillin tube and placed on a thermostatic shaker at 200 rpm for 6.5 h at 37◦C. B. pumilus was added to protocatechuic acid tube and cultured for 7.5 h at 37◦C. The control received only sterile water to substitute the effects of phenolics. Then, all the bacterial cells were centrifuged, washed twice with RNase-free water, and maintained at −80◦C.

The total RNA was isolated from each tissue sample using the RNAiso Plus kit (TaKaRa, Dalian, China) according to the manufacturer's protocol. The purity and content of each RNA were measured using the Qubit RNA Assay Kit in Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, CA, USA) and confirmed using 1.2% agarose gels.

### cDNA Library Construction and Sequencing

The probes were used to eliminate the rRNA sequence of prokaryotes. Then, the mRNA was concentrated using oligo (dT) magnetic adsorption and then broken into fragments, which were used as templates to synthesize first-and second-strand cDNA. The double-stranded cDNA was further purified using the QIAQuick Polymerase Chain Reaction (PCR) extraction kit (Qiagen, Hilden, Germany), resolved for final reparation and poly (A) addition, and then connected with different sequencing adaptors. The library was sequenced by Biomarker Technologies Co., Ltd. (Beijing, China) using the Illumina HiSeq 2500 system.

### Transcriptome Assembly, Gene Annotation and Expression, and Gene Ontology and Kyoto Encyclopedia of Genes and Genomes Annotation

The raw sequencing data reads were first cleaned by removing adaptor sequences, poly-N and low-quality reads. All the downstream analyses were based on high-quality data. For de novo assembly, the clean reads were mapped back to the contigs by Trinity (Grabherr et al., 2011) with the parameters set at a similarity of 90%. Subsequently, the contigs were assembled to construct transcripts with pair-end information and clustered to obtain unigenes.

A sequence similarity search was performed against several databases to investigate the putative functions of the unigenes based on sequence or domain alignment. All unigenes were compared with genes in the non-redundant (NR) databases (National Center for Biotechnology Information non-redundant protein<sup>1</sup> ), Swiss-Prot<sup>2</sup> , Kyoto Encyclopedia of Genes and Genomes (KEGG<sup>3</sup> ), Clusters of Orthologous Groups (COG<sup>4</sup> ),

<sup>1</sup>http://www.ncbi.nlm.nih.gov

<sup>2</sup>http://www.expasy.ch/sprot

<sup>3</sup>http://www.genome.jp/kegg

<sup>4</sup>http://www.ncbi.nlm.nih.gov/COG

and Gene Ontology (GO<sup>5</sup> ) (Ashburner et al., 2000; Tatusov et al., 2000; Kanehisa et al., 2004). Homology search against the NR database was performed to identify top-hit species by BLASTX with a cut-off E-value < 10−<sup>5</sup> . The Blast2GO program (Conesa et al., 2005) was used to obtain the functional classification, and the WEGO software (Ye et al., 2006) was employed to perform the distribution of GO functional classification.

### Quantitative Real-Time PCR Analysis

The expression analysis of the selected genes was assessed using quantitative real-time PCR (qRT-PCR). The reaction conditions were as follows: 95◦C for 15 min, followed by 40 cycles of 95◦C for 50 s, 55◦C for 40 s, and 72◦C for 40 s. At least four biologically independent replicates were used for each sample. All the primer sequences involved in the experiment used for qRT-PCR are listed in Supplementary Tables S1, S2.

### Biofilm Formation Assay

An amendment assay was performed to determine the effects of the phenolic acids on bacterial biofilm formation, as described

<sup>5</sup>http://www.geneontology.org/

by previous reports in 96-well microtiter plates (Hamon and Lazazzera, 2001). In short, protocatechuic acid and vanillin, with concentration up to 120 µmol/L, were added to the culture medium and incubated at 37◦C for 36 h. Then, the biomass of the biofilms was harvested and stained with 250 µL of 0.1% crystal violet. The bound crystal violet was further solubilized with 250 µL of 4:1 (v:v) ethanol and acetone. The multifunctional plate reader SpectraMax i3 analysis system (Multi-Mode Detection Platform, USA) was used to measure the OD570 of the solution in each well.

### Statistical Analysis

The GO annotation was analyzed using the Blast2GO software<sup>6</sup> . Functional classification of the unigenes was performed using the WEGO software. Differences among the treatments were calculated and statistically analyzed using the analysis of variance and the least significant difference multiple range test (P < 0.05). The Statistical Package for the GraphPad Prism version 5.1 and the Data Processing System version 7.05 were used for statistical analysis.

<sup>6</sup>https://www.blast2go.com/

Columns with different letters are statistically different (LSD-test, p < 0.05).

#### TABLE 2 | Summary of data generated in the transcriptome sequence of bacteria.


TABLE 3 | Statistics of differentially expressed genes under the bacterial treatments.


COG, clusters of orthologous groups; GO, gene ontology; KEGG, kyoto encyclopedia of genes and genomes; NR, non-redundant.

### RESULTS

### Physicochemical Properties of R. pseudostellariae and Rhizosphere Soil under the Treatments

**Figure 1** shows that B. pumilus had a better effect on the growth of R. pseudostellariae in the soil in the second cropping year. B. pumilus significantly promoted the chlorophyll content and plant biomass of R. pseudostellariae, while the opposite was true in the case of pathogenic bacterium K. sacchari, as shown in Supplementary Figure S1.

The activity of soil chitinase and acid protease decreased, and the activity of soil urease, cellulase, and sucrase significantly increased in the soil as it was continuously monocultured for 2 years. The added B. pumilus significantly increased the activity of soil acid phosphatase, dehydrogenase, and acid protease compared with enzyme activities in the soil monocultured for 2 years without the beneficial bacterial treatment. However, the added pathogenic K. sacchari significantly decreased the activity of soil chitinase, acid phosphatase, cellulase, dehydrogenase, and acid protease (**Figure 2**) compared with the enzyme activities in the newly planted soil without the bacterial addition.

B. pumilus.

in K. sacchari.

fpls-08-00659 April 29, 2017 Time: 12:26 # 7

HPLC analysis showed that the protocatechuic acid levels tended to increase in the rhizosphere soil as the number of continuous monoculture years increased. It was also found that the vanillin tended to decrease, but protocatechuic acid increased in the newly planted soil treated with the pathogenic K. sacchari, compared with the phenolic acids in the newly planted soils without the pathogen treatment (Supplementary Figures S2, S3).

### De Novo Assembly of Transcriptome

The transcriptome of bacteria was characterized, and a number of genes associated with functional regulation and metabolism were obtained. Sequencing of the bacterial transcriptome yielded about 21,797,093 clean reads with 93.62% Q30 bases and 5.48 Gb with 43.16% GC ratio in B. pumilus. The 24,161,020 clean reads with 93.55% Q30 bases and 6.08 Gb with 54.06% GC ratio were obtained in K. sacchari (**Table 2**). Transcripts were assembled into unigenes, yielding 131 DEG with downregulation of 77 in B. pumilus. The number of DEG detected in K. sacchari was up to 331, including 235 upregulated unigenes (**Table 3**).

### COG Annotation

To assess the validity and integrity of the transcriptome sequence in B. pumilus, 131 DEG annotated in the NR database were assigned to the COG database to classify potential functions. In total, 125 DEG were aligned to the 16 COG classifications. Among them, assignments to (P) inorganic ion transport and metabolism made up the majority (28, 22.4%), followed by (E) amino acid transport and metabolism (22, 17.6%), (K) transcription (11, 8.8%), and (R) general function prediction only (10, 8%) (**Figure 3**).

A total of 299 DEG in K. sacchari were aligned to the 20 COG classifications. The majority part was (G) carbohydrate transport and metabolism (43, 14.38%), followed by (E) amino acid transport and metabolism (39, 13.04%), (S) function unknown (39, 13.04%) and (P) inorganic ion transport and metabolism (31, 10.37%), (R) general function prediction only (31, 10.37%), and (C) energy production and conversion (22, 7.36%) (**Figure 4**).

### GO Annotation of Gardenia Transcriptome

The Blast2GO software was used to assemble the DEG in B. pumilus. These DEG were assigned to 25 classifications. In the category of cellular component, most genes were associated with cell, cell part, membrane, membrane part, macromolecular complex, and organelle. In the category of molecular function, genes belonged to catalytic activity, binding, transporter activity, protein binding transcription factor activity, and nutrient reservoir activity. In the category of biological process, the dominant subcategories were genes associated with the metabolic process, single-organism process, cellular process, localization, biological regulation, stimulatory response, cellular component organization or biogenesis, developmental process, biological adhesion, and signaling (**Figure 5**).

The DEG in K. sacchari were assigned to 28 classifications. Among the category of cellular component, the dominant subcategories were genes associated with cell, cell part, membrane, membrane part, and organelle. The category of molecular function included most of the genes associated with catalytic activity, binding, transporter activity, nucleic binding transcription factor activity, structural, molecule activity, electron carrier activity, antioxidant activity, and molecular transducer activity. The metabolic process, single-organism process, cellular process, localization, biological regulation, stimulatory response, locomotion, and cellular component organization or biogenesis were under the category of biological process (**Figure 6**).

### KEGG Pathway Annotation

The KEGG database was used to understand the interaction of genes and metabolic biological functions in B. pumilus. The 94 DEG were assigned to 33 KEGG pathways. The three KEGG categories were based on pathway data, including metabolism (61, 64.89%), environmental information processing (26, 27.66%), and genetic information processing (7, 7.45%). Among the pathway subgroups, the pathway of ABC transporters had the most unigenes (25), followed by phenylalanine, tyrosine, and tryptophan biosynthesis (8), and glycine, serine, and threonine metabolism (5). The metabolism pathways of citrate cycle (3), fatty acid biosynthesis (1), arginine and proline metabolism (3), histidine metabolism (2), and novobiocin biosynthesis (2) also had several DEG (**Figure 7**).

A total of 201 DEG were assigned to 55 KEGG pathways, including metabolism (145, 72.14%), environmental information processing (45, 22.39%), cellular processing (8, 3.98%), and genetic information processing (3, 1.49%) in K. sacchari. The pathways of ABC transporters (26), fructose and mannose metabolism (13), pyruvate metabolism (11), phosphotransferase system (PTS; 10), and glycolysis/gluconeogenesis (9) were the dominant unigenes. The ratio of the four classifications showed that some DEG were focused on the flagellar assembly (6), fatty acid biosynthesis (2), bacterial chemotaxis (2), and phenylalanine, tyrosine, and tryptophan biosynthesis (1) (**Figure 8**).

### Validation of Expression of Selected DEG Using qRT-PCR

Based on the KEGG pathway enrichment scatter diagram (Supplementary Figure S4), the expression of downregulated DEG involved in the important metabolic pathways (citrate cycle; fatty acid metabolism; phenylalanine, tyrosine, and tryptophan biosynthesis; and novobiocin biosynthesis) were determined by real-time PCR in B. pumilus. The results showed that the trend of these unigenes was similar to the RPKM values. Based on the DEG and metabolic pathway, it was concluded that the protocatechuic acid had a negative effect on the growth and biocontrol of B. pumilus (**Figure 9**).

For K. sacchari, the KEGG pathway enrichment of the upregulated DEG in glycolysis/gluconeogenesis, fatty acid biosynthesis, pentose phosphate, bacterial chemotaxis, flagellar assembly, and PTS pathway was also determined by qRT-PCR (Supplementary Figure S5). Vanillin enhanced these pathways, having a positive effect on the growth of the pathogenic K. sacchari (**Figure 10**).

### Effects of Phenolic Acids on the Biofilm Formation of Specific Bacteria

The results showed that vanillin significantly stimulated the biofilm formation of the pathogenic bacterium K. sacchari at a dose of 120 µmol/L, while protocatechuic acid showed inhibitory effects on the beneficial bacterium B. pumilus (Supplementary Figure S6).

### DISCUSSION

Replanting disease is a common example of typical negative plant–soil feedback, which accounts for the serious decline in the biomass and quality of crops when the same crop or its related species are monocultured consecutively in the same soil (Huang et al., 2013). Replanting disease has become a prevalent problem recently in the production of many annual crops, which are being subjected to intensive consecutive monoculture, such as Rehmannia glutinosa, cucumber, peanut, and tobacco (Zhou et al., 2012; Li et al., 2014; Santhanam et al., 2015; Wu et al.,

2015). Many factors have been thought to be responsible for the replanting disease, including soil nutrient imbalance, soil physical and chemical properties, accumulation of autotoxins generated by roots, and change in the soil microbial community structure (Wu et al., 2011; Huang et al., 2013; Zhao et al., 2016). With more research focused on this field, many studies have revealed that root exudates produced by plants have a propensity to shape the rhizosphere microbiome directly or indirectly and also have some influences on the growth of plants (Li et al., 2014; Venturelli et al., 2015; Wu et al., 2016). A previous study found that the root exudates of R. glutinosa could promote the growth and toxin production of pathogenic F. oxysporum while inhibiting the growth of the beneficial bacteria Pseudomonas sp (Wu et al., 2015). Liu et al. (2017) have elucidated the benzoic acid significantly increased the abundances of bacterial and fungal, and reduced bacteria-tofungi ratio in the rhizosphere soil of peanut. Moreover, the continuous monoculture of R. pseudostellariae was found to differentially mediate the shifts in rhizosphere microbes, showing that the number of the pathogenic bacterium K. sacchari in the rhizosphere soil of R. pseudostellariae significantly increased with the increase in monocropping years, and the opposite trend was observed in the case of B. pumilus (Wu et al., 2016). This study demonstrated the positive effect of B. pumilus on the growth of R. pseudostellariae, suggesting that the addition of beneficial bacterium B. pumilus to the monocultured rhizosphere soil of the medicinal plants significantly decreased the activity of soil urease, catalase, sucrase, and cellulase and increased the activity of chitinase compared with those in the 2nd year monocultured soil without the treatment. These changes could contribute to the better growth of the plants and inhibit the proliferation of pathogenic fungi (Mauch et al., 1988; Zeng et al., 2007). However, the added K. sacchari shifted most soil enzyme activities, creating a disease-conducive environment to inhibit the growth of R. pseudostellariae. The results further confirmed the imbalance of microorganism population structure, which in turn changed the properties of rhizosphere biology including the amounts and activities of soil enzymes by mediating the phenolic acid exudates under the cropping system.

The microbes in the rhizosphere, referred to as the second genome of the plants, and the plants are able to shape their rhizosphere microbiome by hosting specific microbial communities (Berendsen et al., 2012). Zhou and Wu (2011) and Zhou et al. (2012) reported that the root exudates of cucumber could change soil microbial communities and promote the growth of F. oxysporum f.sp. cucumerinum (the hostspecific soil-borne pathogen of cucumber) in the continuously monocropped system. For R. pseudostellariae, K. sacchari and B. pumilus have been identified as the specific microbial flora involved in the replanting disease (Wu et al., 2016). However, the rhizosphere is a complex system. The catabolism of the root exudates and their stimulatory effects on the microbial community structure has been explored by studying the effects of single low-molecular-weight organic molecules in simple systems (Landi et al., 2006; Renella et al., 2007; Li et al., 2014). This study found vanillin as one of the main allelochemicals in the root exudates of the medicinal plants, which enhanced the glycolysis/gluconeogenesis, and pentose phosphate, and PTS pathway, leading to high metabolism and a good balance between carbon and nitrogen in the treated pathogenic K. sacchari. Moreover, the genes associated with the fatty acid biosynthesis, flagellar assembly, and bacterial chemotaxis were upregulated in the pathogenic bacteria under treatment (**Figure 11**). Chemotaxis and colonization are the two primary elements of plant– microbe interactions. Moreover, bacterial biofilm formation on plant roots is a visualized performance of effective colonization (Sood, 2003; Timmusk et al., 2005; Li et al., 2013). It means that vanillin promotes colonization and growth of the treated K. sacchari and leads to an increased incidence of the disease. Moreover, the metabolite (protocatechuic acid) of K. sacchari had a negative effect on the citrate cycle, novobiocin biosynthesis, and phenylalanine, tyrosine, and tryptophan biosynthesis of the beneficial B. pumilus (**Figure 12**). It also promoted the fatty acid metabolism and inhibited the biofilm formation of B. pumilus under treatment. The results suggested that the metabolite (protocatechuic acid) of K. sacchari in the rhizosphere soil was not conducive to the colonization of B. pumilus and significantly reduced its antagonistic ability against the specific pathogens in the monoculture system.

The complex plant-associated microbial communities are crucial for plant health. However, the microbial activity in

soil is greatly influenced by plant root exudation, resulting in an increased or decreased microbial biomass and activity around the roots (Bais et al., 2006; Doornbos et al., 2012). For most plant species, consecutive monoculture creates a new environment for the build-up of specialized plant pathogens (Bennett et al., 2012). In this process, the root exudates of plants play important roles. Phenolic acids act as signals to attract K. sacchari by chemotaxis response. K. sacchari utilized vanillin for its proliferation. The stimulation of vanillin caused high metabolism and maintained a good balance between carbon and nitrogen in K. sacchari. The colonization of K. sacchari could be promoted by the biofilm formation. K. sacchari altered the soil enzyme activities and caused the disease in R. pseudostellariae. Meanwhile, the metabolite of K. sacchari had a negative effect on the growth of beneficial bacteria. Wang et al. (2013) reported that Pseudomonas utilized and catalyzed the conversion of (−)-catechin into protocatechuic acid, which then became an active agent in the allelopathy of understory species in the rhizosphere of Rhododendron formosanum. The results showed that the pathogenic bacterium K. sacchari was able to convert vanillin, one of the main allelochemicals in root exudates of

FIGURE 12 | Schematic presentation of some biological pathways affected in B. pumilus under protocatechuic acid stress. Most differentially expressed genes were integrated and are indicated in red (upregulated) or green (downregulated), respectively.

R. pseudostellariae, into protocatechuic acid, and reduced the number of beneficial bacteria via chemical transformation in the monoculture cropping of R. pseudostellariae. The function of the root exudates differs in plant species and their rhizosphere microbes. The present study unraveled the molecular mechanism underlying differential changes in specific hosting microflora mediated by the root exudates of R. pseudostellariae in the monoculture system. The results gave a promising strategy for the control of replanting disease by rhizosphere management, which includes root exudate remediation, organic amendment application, and rhizosphere microbiome manipulation. In short, rhizosphere management is bound to lead to a new round of the Green Revolution, which will have a huge influence on the sustainability development of agriculture.

### AUTHOR CONTRIBUTIONS

WXL and HW conceived the study. HMW, WXL, and MK wrote the paper. HMW, XQ, QZ, JW, and ZL performed experiments. HMW, LW, and SL performed the statistical analyses. HMW, SL, and WWL are involved in field management and soil sampling. MK, JW, and JX have revised the manuscript. All authors discussed the results and commented on the manuscript.

### REFERENCES


### FUNDING

This work was supported by the National Science Foundation of China for providing the funds (81573530, 31401950), Scientific Research Foundation of Graduate School of Fujian Agriculture and Forestry University (324-1122YB031), Major agricultural extension services of Fujian Province (KNJ-153015), Fujian-Taiwan Joint Innovative Center for Germplasm Resources and Cultivation of Crop (Fujian 2011 Program, No. 2015-75, China) and Natural Science Foundation of Fujian Province (2017J01803).

### ACKNOWLEDGMENTS

We thank Scientific Research Foundation of Graduate School of Fujian Agriculture and Forestry University for providing the funds used in this work. We also thank Weng Hongwu Academic Innovation Research Fund of Peking University and Original Research Fund of Peking University.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.00659/ full#supplementary-material


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Wu, Xu, Wang, Qin, Wu, Li, Lin, Lin, Zhu, Khan and Lin. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpls-08-00659 April 29, 2017 Time: 12:26 # 15

# Identification of TIFY Family Genes and Analysis of Their Expression Profiles in Response to Phytohormone Treatments and Melampsora larici-populina Infection in Poplar

### Wenxiu Xia, Hongyan Yu, Pei Cao, Jie Luo and Nian Wang\*

College of Horticulture and Forestry Sciences, Huazhong Agricultural University, Wuhan, China

#### Edited by:

Gero Benckiser, University of Giessen, Germany

#### Reviewed by:

Daolong Dou, Nanjing Agricultural University, China Mario Serrano, Center for Genomic Sciences (UNAM), Mexico

> \*Correspondence: Nian Wang wangn@mail.hzau.edu.cn

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 06 January 2017 Accepted: 21 March 2017 Published: 05 April 2017

#### Citation:

Xia W, Yu H, Cao P, Luo J and Wang N (2017) Identification of TIFY Family Genes and Analysis of Their Expression Profiles in Response to Phytohormone Treatments and Melampsora larici-populina Infection in Poplar. Front. Plant Sci. 8:493. doi: 10.3389/fpls.2017.00493 The TIFY domain contains approximately 36 conserved amino acids that form the core motif TIF[F/Y]XG, and they were reported to play important roles in plant growth, tissue development and defense regulation. Moreover, more and more evidence has shown that some members of the TIFY gene family perform their functions by modulating plant hormone signaling pathways. Poplar trees are found worldwide, and they comprise approximately 30 species. Benefit from the importance of poplar and its advanced platform, this tree is considered to be the model perennial plant. Here, we conducted a genome-wide identification of TIFY genes in poplar, and 24 TIFY genes were found. These 24 TIFY genes were assigned to different subfamilies according to the presence or absence of domains and motifs that they harbored. Careful analyses of their locations, structures, evolution and duplication patterns revealed an overview of this gene family in poplar. The expression profiles of these 24 TIFY genes were then analyzed in different tissues using publicly available expression data; their expression profiles following different JA/SA treatments and infection with leaf rust pathogen were also carefully examined by qRT-PCR assays. Based on their expression profiles, the functions of a number of TIFY genes could be predicted. By performing this study, we have provided valuable information for further functional characterisation of TIFY genes in poplar and candidate genes for the improvement of poplar disease resistance.

Keywords: TIFY domain gene family, poplar, jasmonic acid, salicylic acid, gene expression, Melampsora larici-populina

### INTRODUCTION

The TIFY domain contains approximately 36 conserved amino acids (AA) that form the core motif TIF[F/Y]XG (Vanholme et al., 2007; Bai et al., 2011). This conserved domain characterizes a plantspecific family of transcription factor (TF) genes called the TIFY gene family. TIFY genes were first characterized in Arabidopsis, and the gene AT4G24470 was reported to be a putative TF involved in inflorescence and flower development (Nishii et al., 2000). However, this gene was annotated

as a ZIM gene in this study (Nishii et al., 2000). With a genomewide survey of the Arabidopsis genome and because of the confusing use of the ZIM nomenclature, proteins containing TIFY domains were renamed as the TIFY gene family (Vanholme et al., 2007). The TIFY gene family can be classified into four subfamilies, TIFY, JAZ, ZML, and PPD, depending on whether they contain additional domains/motifs (Vanholme et al., 2007; Bai et al., 2011). Proteins with only the TIFY (PF06200) domain are classified as the TIFY subfamily; proteins with both the TIFY and jasmonate ZIM domains (JAZ, PF09425) are classified as the JAZ subfamily (Staswick, 2008); and proteins containing the TIFY domain and the CCT (PF06203) and/or ZML (PF00320) motif are classified as the ZML subfamily. In the PPD subfamily, the proteins contain the TIFY and PPD domains; proteins in this subfamily also sometimes contain a truncated JAZ domain (White, 2006).

To date, the functions of several TIFY genes have been fully investigated, and some of them have been found to play important roles in different biological processes. Of the four TIFY subfamilies, the functions of JAZ proteins are the most clear, and they have been found to be involved in the jasmonate (JA) signaling pathway. JAZ proteins act as JA repressors to inhibit TFs that regulate early JA-responsive genes (Pauwels and Goossens, 2011); in contrast, JA can also induce degradation of JAZ proteins, thereby allowing the expression of its response genes (Chung and Howe, 2009; Niu et al., 2011). Since JA is a key phytohormone in plant development, JAZ proteins were also found to play critical roles in the regulation of numerous aspects of plant development. In Astragalus sinicus, AsJAZ1 was found to interact with AsB2510 and participated in nodule development and nitrogen fixation (Li et al., 2015). In Arabidopsis, deletion of the two PPD genes (at the same locus) increased leaf lamina size and resulted in dome-shaped rather than flat leaves. Siliques were also altered in shape because of the additional lamina growth (White, 2006). JAZ and DELLA proteins were also found to bind to the WDrepeat/bHLH/MYB complex to modulate the synergistic effects of gibberellin and JA signaling; thus, these two types of proteins can integrate different hormonal signals to synergistically regulate plant development (Qi et al., 2014). Moreover, JAZ proteins were also found to play important roles in plant defense. In wild soybean, transcription of GsJAZ2 increased following exposure to different abiotic stresses including salt, alkali, cold and drought. Overexpression of GsJAZ2 in an Arabidopsis line resulted in enhanced tolerance to salt and alkali stresses (Zhu et al., 2012). In rice, OsJAZ8 was reported to regulate host immunity by modulation of JA-responsive volatile compounds (Taniguchi et al., 2014). In addition to the JAZ genes, other genes of the TIFY subfamilies were found to play roles in plant development and defense. In wild soybean, GsTIFY10, which could be induced by bicarbonate, salinity stress and JA, was isolated and overexpressed in Arabidopsis. The transgenic plants showed enhanced tolerance to bicarbonate stress during seed germination and during the early- and adult-seedling developmental stages (Zhu et al., 2011). Based on the preceding information, we conclude that the plantspecific TIFY proteins are very important in the regulation of plant development and defense.

Due to the importance of the TIFY gene family and benefitting from the accelerated release of genome data, a number of studies have focused on the genome-wide investigation and characterisation of TIFY genes in different plant species. In the respective dicot and monocot model plants, Arabidopsis and rice, 18 and 20 TIFY genes were reported, respectively (Ye et al., 2009; Bai et al., 2011). In addition, 18 TIFY family proteins were found in pigeonpea [Cajanus cajan (L.) Millsp.] (Sirhindi et al., 2016); 30 TIFY genes were found in apple (Malus × domestica Borkh.) (Li et al., 2014); 19 TIFY genes were found in grape (Vitis vinifera) (Zhang et al., 2012); 28 TIFY genes were found in the Gossypium raimondii genome (He et al., 2015); 21 TIFY genes were found in Brachypodium distachyon (Zhang et al., 2015); and 34 TIFY genes in found in wild soybean (Glycine soja) (Zhu et al., 2013). In these studies, the expression profiles of TIFY genes were investigated and their functions were predicted based on their preferentially transcriptional abundance in a given tissue or under a given stress condition. This information is very valuable for guiding further characterisation of the functions of TIFY genes. In poplar, 25 TIFY proteins were identified using the Populus trichocarpa genome version 1.0 annotations when global comparisons of the gene family were performed in a number of plant species (Bai et al., 2011). Recently, the P. trichocarpa genome annotation was updated to version 3.0, and the quality of the reference genomic sequences was also greatly improved. Therefore, it is necessary to conduct a genome-wide survey of TIFY genes in poplar using the P. trichocarpa genome annotation version 3.0, in addition to analyzing their expression profiles under different conditions.

Poplar (Populus spp.) is a tree that is found worldwide, and it comprises approximately 30 species. Due to its characteristics of being fast-growing, widely used and strong ability to adapt, poplar is grown in most of the Northern hemisphere. Most of the species in the genus can be used for wood, pulp, paper, and fuel. In some areas, poplar is also an important landscape tree. Moreover, poplar species usually have small genome sizes, and some species can easily be created transgenic lines. In 2006, the draft genomic sequence of P. trichocarpa was released (Tuskan et al., 2006). Subsequently, the genomic sequence and annotations have been updated several times and high quality genomic information has recently become available online<sup>1</sup> (Wullschleger et al., 2013). Moreover, additional genomic information of value for the species is publicly available, including the expression database and the whole genomic sequence of P. euphratica (Ma et al., 2013; Sundell et al., 2015). According to the above information, poplar is considered to be a model plant in perennial tree species.

To our knowledge, no comprehensive genome-wide survey and characterisation of TIFY genes in poplar has been carried out. In this study, we mainly focused on the responses of TIFY genes to phytohormone treatments and biotic stresses, and we aimed to identify TIFY genes involved in poplar biotic defenses. To this end, we first conducted a genome-wide identification of TIFY genes using the P. trichocarpa genome annotation version 3.0; 24 TIFY genes were found. Based on the presence of certain domains/motifs within these 24 TIFY genes, they were assigned to different subfamilies. The duplication types of each of the TIFY

<sup>1</sup>http://popgenie.org/

genes were simulated and their phylogeny, gene structures and locations within the genome were also investigated. Based on the publicly available expression data, the expression profiles of the 24 TIFY genes in different tissues were subsequently analyzed; their expression profiles after different JA/SA treatments and M. larici-populina infection were also carefully examined by qRT-PCR assays. Based on their expression profiles, the functions of a number of TIFY genes could be predicted. By performing this study, we have provided valuable information for further functional characterisation of TIFY genes in poplar and candidate genes for the improvement of poplar disease resistance.

### MATERIALS AND METHODS

### Identification of TIFY Genes in the Poplar Genome

To identify members of the TIFY gene family in poplar, the Hidden Markov Model (HMM) profile of the TIFY domain (PF06200) was downloaded from the pfam database (Pfam<sup>2</sup> ). The reference genome P. trichocarpa was downloaded from the Phytozome database<sup>3</sup> . Genome annotation version 3.0 of P. trichocarpa was used in this study. To search the 73,013 predicted proteins of the P. trichocarpa genome annotation version 3.0, 759 seed sequences in the PF06200 HMM profile were used. Two softwares, blastp and HMMER, were used to perform a search for proteins harboring the TIFY domain. Proteins of P. trichocarpa that showed E-values above 1e-6 in the search results of blastp or HMMER were considered to be candidate TIFY domain genes. Of the proteins from the same gene model, only the longest ones (the "0.1" gene model of TIFY genes in the P. trichocarpa genome annotation version 3.0) were kept for further study. To confirm the existence of the TIFY domain and the Jas, CCT and ZML motifs, candidate TIFY domain proteins were used to search the Pfam database. The Pfam accession numbers for the three motifs are PF09425, PF06203 and PF00320, respectively. We also checked for the existence of the PPD domain/motif in each of the candidate TIFY genes. We noted that there is no sequence information for the PPD domain/motif in the Pfam database. The previously annotated PPD genes in Arabidopsis and grape were used to build the conserved PPD sequence. Briefly, the previously annotated PPD sequences were first aligned by ClustX2.1, and an msf result was produced and imported into hmmbuild implemented in HMMER. The resulting PPD HMM profile was used to search all of the candidate TIFY genes.

### Analyses of Phylogeny, Genomic Structures, Chromosomal Locations and Gene Duplications

The software MEGA5 was used to analyze the phylogeny of all of the TIFY genes (Tamura et al., 2011). The protein sequences

<sup>2</sup>http://pfam.xfam.org/

of all of the identified poplar TIFY genes were imported into ClustalX2.1 to perform a complete alignment, and the resultant multiple alignment file was imported into MEGA5. Unrooted phylogenetic trees were constructed using the Neighbor-Joining (NJ) method, and the bootstrap test was carried out with 1,000 iterations. For constructing phylogenetic trees among the four species, the protein sequences of the three other species were obtained from Phytozome database<sup>4</sup> . The deduced TIFY protein information was obtained from Arabidopsis thaliana (Bai et al., 2011), Oryza sativa (Bai et al., 2011) and Vitis vinifera (Zhang et al., 2012).

Gene structures and their chromosomal locations were obtained from the P. trichocarpa genome annotation version 3.0. The gene structures were displayed using a custom R script. The duplication pattern for all of the poplar genes was analyzed using MCScanX software. Briefly, the 41,335 poplar gene models (the "0.1" gene model in the P. trichocarpa genome annotation version 3.0 represents all genes models if a gene has more than one alternative transcript) were extracted from the poplar genome. An all-vs.-all local blast for the 41,335 poplar gene models using the local blast software with E-values under 1e-4 was carried out. The blast output was imported into MCScanX software following the procedure. All of the 41,335 genes were classified into four types using default criteria, including segmental, tandem, proximal, and dispersed duplications. The duplication pattern for each poplar TIFY gene was then determined according to the results.

### Microarray-Based Expression Analysis of Poplar TIFY Genes Using Publicly Available Data

The PopGenExpress database was employed to investigate the expression levels of all of the poplar TIFY genes in different tissues<sup>5</sup> (Wilkins et al., 2009). Briefly, the correspondence between microarray probes and poplar genes was obtained from the annotation files deposited for the microarray data. Because the annotation released for the poplar microarray used in the PopGenExpress database was based on P. trichocarpa genome annotation version 2.0, the relationship between the probes and poplar genes annotated in genome version 3.0 was transformed based on the correspondence between versions 2.0 and 3.0 of the P. trichocarpa genome annotation. For genes that had no corresponding probes in genome version 3.0, an additional reciprocal blast with released sequences for all of the probes was employed. If a given gene and probe pair were each other's top hits in the reciprocal blast results, then the probe was considered to represent the gene. The relative expression levels of all of the TIFY genes in xylem, roots, mature and young leaves, male and female catkins were obtained from the PopGenExpress database. A heat map for the expression levels of all of the TIFY genes in the selected tissues was prepared with Cluster 3.0 software (de Hoon et al., 2004); the dendrogram was visualized using JavaTreeview<sup>6</sup> .

<sup>3</sup>https://phytozome.jgi.doe.gov/pz/portal.html

<sup>4</sup>https://phytozome.jgi.doe.gov/pz/portal.html

<sup>5</sup>http://www.bar.utoronto.ca/

<sup>6</sup>https://sourceforge.net/projects/jtreeview/

### Plant Material and Stress Treatments

The hybrid poplar variety "NL895" (P. euramericana) was used in this study. The female parent of NL895 is P. deltoids cv. Lux" (I-69) and the male parent is also a hybrid variety "I-45" (P. × euramericana). Tissue culture seedlings of "NL895" were grown in 3-L pots containing a sand-peat (50:50, v/v) mixture, and the pots were placed in a greenhouse under a 16/8 h photoperiod. The light intensity in the greenhouse was set at 12,000 Lx. The temperatures were set at 28 and 25◦C for day and night, respectively. The seedlings were watered twice per month with Hoagland's complete nutrient solution. Additionally, the seedlings were watered with distilled water 2–3 times a week. Seedlings with 6–10 fully expanded leaves were used for the treatments.

For the plant hormone treatments, "NL895" seedlings were treated with 0.2 mM jasmonic acid (JA) and 0.5 mM salicylic acid (SA). Leaves with LPIs (leaf plastochron indices) ranging from 4 to 10 were sprayed with JA or SA solution. To ensure that the hormone treatments were uniform, leaves were sprayed on both sides just until the leaf surfaces began to form small drops. Leaf samples were collected at 0, 2, 6, 12, and 24 h after treatment. Sprayed leaves from each individual poplar plant were pooled into one biological sample. Three replicates were carried out for each treatment.

For the leaf rust pathogen infection treatments, a virulent M. larici-populina strain used as fungal material in this study was obtained from P. simonii Carr., which showed typical leaf rust symptoms (yellow bubbles on the lower surface of leaf) in the summer of 2015. This strain was also used as the fungal material in another study (Wang et al., 2017). The urediniospores of M. larici-populina isolates were allowed to multiply on 1-year-old potted P. simonii Carr. cuttings and diluted into an urediniospore suspension in agar-water (0.1% agar in distilled water). The concentrations of the urediniospore suspensions were 1–2 mg/ml. Fully expanded leaves from "NL895" with leaf plastochron indices (LPIs) of 4–10 from each branch were spray-inoculated on their abaxial surfaces with the prepared urediniospore suspension. Leaves from one separate cutting were considered to be one sample. Leaf samples were collected after 0, 2, 4, and 8 days treatments. The samples were immediately frozen in liquid nitrogen and stored at −80◦C until required for further analysis.

### Quantitative Real-Time RT-PCR Analysis

Total RNA was isolated using an RNAprep Pure Kit (for Plants) according to the manufacturer's protocol [TIANGEN Biotech (Beijing) Co., Ltd., Beijing, China]. The quality of all of the RNA samples was examined by performing agarose gel electrophoresis. The first strand cDNA was synthesized using the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix Kit (TransBionovo Co., Ltd., Beijing, China). The primers used for the qRT-PCR assay of TIFY genes are listed in Supplementary Table S2. PCR products from cDNA or genomic DNA templates using these primer pairs were initially sent for sequencing to confirm that the correct targets had been amplified (data not shown). The correct primer pairs were used for subsequent qRT-PCR assays that were performed on the LightCycler 96 (Roche) platform by using the FastStart Essential DNA Green Master Mix (Roche). For each sample, two reference genes (ACTIN and 18S) were used to standardize the mRNA abundance, and three replications were performed. The 2−11Cq method was used to calculate the relative gene expression based on the qRT-PCR data (Livak and Schmittgen, 2001).

### RESULTS

### Characteristics of TIFY Family Genes in Poplar

A genome-wide search of TIFY family genes in poplar yielded a total of 24 non-redundant genes that were found to harbor the TIFY domain. The 24 non-redundant genes represent a total of 99 transcripts in the P. trichocarpa genome annotation version 3.0 (**Table 1**) and all the 99 transcripts have the TIFY domain. To avoid complexity in the subsequent study, only the "0.1" gene model for the 24 non-redundant genes were kept in further analyses. These genes were considered to be TIFY family genes (**Table 1**). Their phylogeny and gene structures are shown in **Figures 1A,B**, respectively. Of these 24 TIFY domain-containing genes, 12 of the predicted proteins contain both the TIFY domain and the Jas motif; these proteins were designated as PtJAZ1 through 12. Seven proteins have both a TIFY domain and CCT and ZML motifs, and one protein has both a TIFY domain and a CCT motif; thus, these eight proteins were designated as PtZML1 through 8. Two proteins contain the TIFY domain and the PPD motif, and they were designated as PtPPD1 and PtPPD2. Note that PtPPD1 also has a Jas motif in its protein structure. The remaining two proteins contain only the TIFY domain, and they were designated as PtTIFY1 and PtTIFY2. All of the domains or motifs located in the 24 TIFY genes are illustrated in **Figure 1C**. The nomenclatures of these genes were designated based on both their subfamilies and their locations on the poplar chromosomes (**Figure 2**). The correspondence of our nomenclatures for the identified TIFY family genes and their original gene IDs released by the P. trichocarpa genome annotation version 3.0 are shown in **Table 1**. In previous report, 25 TIFY proteins were identified using the P. trichocarpa genome version 1.0 annotations when global comparisons of the gene family were performed in a number of plant species (Bai et al., 2011). When comparing our results with theirs, we found PtZML4 and 5 were corresponding to 3 ZML proteins in Bai's results; while the other 22 TIFY genes are consistent in the two identifications.

According to **Figure 1B**, all of the TIFY family genes have more than one exon. PtJAZ9 has the smallest number of exons at 2, whereas PtZML5 has the largest number of exons at 11. The genomic sequences of most of these genes are less than 6 kb in length, but PtZML4 has the longest genomic sequence of approximately 12 kb. The gene PtJAZ9 has the shortest open reading frame (ORF) of 149 AA, whereas PtPPD2 has the longest ORF of 454 AA. Insights into the phylogeny and gene structures of genes has indicated that proteins that are closely related phylogenetically tend to have similar gene structures.

#### TABLE 1 | TIFY family of genes in poplar.

fpls-08-00493 April 3, 2017 Time: 14:46 # 5


For example, PtJAZ1 and PtJAZ4 have the same number of exons in the same arrangement and tend to cluster together in the phylogenetic tree; this phenomenon was also observed for PtTIFY1 and PtTIFY2 and PtZML1 and PtZML3. In order to further insight into the phylogeny of TIFY proteins within plant, TIFY family proteins of three other representative species,

including Arabidopsis, rice and grapes, were also employed to construct a phytogenic tree. Clearly, TIFY proteins in the four plants can be grouped into four clades (**Supplementary Figure S1**). According to this figure, proteins in the clade 1 and 2 tend to only have TIFY domain and have ZML motif, respectively; while proteins in the clade 3 and 4 have no clear difference when comparing with their domains or motifs.

The chromosomal locations of these 24 TIFY family genes is shown in **Figure 2** and **Table 1**, and they are found on 13 of the 19 chromosomes. Chromosomes 4, 9, 13, 14, 16, and 19 contain no TIFY genes. Because duplication usually contributes to the expansion of gene families, we investigated the duplication patterns of each of the genes. In total, 19 genes were produced by whole genome duplication (WGD), and two genes were produced by tandem duplication. The other three genes did not show duplication patterns. The two tandemly duplicated genes, PtZML2 and Pt ZML6, were found to be tandemly copied from PtZML1 and PtZML5 (**Figure 2**), respectively, whereas both PtZML1 and PtZML5 were produced by WGD. This result suggested that PtZML1 and PtZML5 were the ancestors of PtZML2 and PtZML6, respectively. The gene duplication patterns for all 24 of the TIFY genes showed that WGD played large roles in this gene family in poplar.

### In silico Analysis of Gene Expression Profiles in Different Tissues

Since the transcriptional abundance of a gene in different tissues is usually indicative of its function, we investigated the expression profiles of the 24 TIFY family genes in poplar. By analyzing of the publicly available expression data from the PopGenExpress database, we were able to obtain the relative expression levels of all 24 TIFY family genes in six tissues: xylem, roots, mature and young leaves and male and female catkins (Supplementary Table S1 and **Figure 3**). All 24 genes showed very high gene expression levels in female (7.11 ± 4.54) and male catkins (5.84 ± 1.63), whereas the lowest expression levels were found in roots (Supplementary Table S1). This result may suggest that the TIFY family of genes plays important roles in poplar catkin development.

Genes that are expressed abundantly in one or several specific tissues usually indicate their function related to the formation of these organs or the corresponding plant development. Further insights into the expression levels for each individual gene indicated that TIFY genes showed extremely high expression levels in some tissues. We filtered these types of genes when their expression was greater than two-fold the average expression level of all TIFY genes in the tissue. In total, seven TIFY genes were preferentially expressed in a given tissue, and some tissues had more than one preferentially expressed gene. PtJAZ6 and PtJAZ9 were preferentially expressed in xylem; PtJAZ2 and PtJAZ9 were preferentially expressed in roots and mature leaves; PtJAZ1 and PtJAZ2 were preferentially expressed in both male and female catkins, and PtJAZ9 was preferentially expressed in female catkins; PtJAZ7 and PtJAZ11 were preferentially expressed in young leaves (Supplementary Table S1 and **Figure 3**). These data suggest that PtJAZ1, PtJAZ2, PtJAZ6, PtJAZ7, and PtJAZ9 could be responsible for the formation of these tissues or contribute to the functions of these tissues. By summarizing these results, we were able to find that a large proportion of genes in the JAZ subfamily were filtered out by our method,

although some genes in other subfamilies also showed relatively high abundance in the seven selected tissues. This suggests that most of the JAZ subfamily in the TIFY gene family plays important roles in poplar development. In addition, the expression values for the two reference genes, Actin and 18S, across the five developmental tissues were also obtained and listed in Supplementary Table S1. The data revealed these two genes were expressed similarly in different tissues and it suggested that these two genes were suitable for the subsequent qRT-PCR assay.

### Expression Profiles of Poplar TIFY Genes in Response to JA and SA Treatments

JA and SA are hormones that play important roles in signal transduction when plants are challenged with biotic and abiotic stresses. Treating plants with JA, SA or their derivatives MeJA and MeSA can modulate the symptoms of pathogenic or herbivory damage on plants (Yamada et al., 2012; Zhang et al., 2016). To investigate which genes in TIFY gene family can respond to these two plant hormones, we examined the expression profiles of TIFY genes in poplar leaves after treatment with MeJA and SA solutions. The relative expression levels were assayed by qRT-PCR. In the description to convenience, CK indicated untreated samples; JA-2 h, JA-6 h, JA-12 h, JA-24 h indicated samples collected from leaves after 2, 6, 12 and 24 h of MeJA treatment, respectively; and SA-2 h, SA-6 h, SA-12 h, and SA-24 h indicated samples collected from leaves after 2, 6, 12, and 24 h of SA treatment, respectively. In total, we were able to examine the expression profiles for 20 of the 24 TIFY genes; the expression profiles of PtJAZ8, PtPPD2, PtZML4 and PtZML5 were not examined due to unsuccessful primer design. The relative expression profiles of all 20 genes for all nine samples are shown in **Figure 4**.

Fold-changes in expression larger than 2.0 or less than 0.5 and also with a p-value less than 0.05 between CK and time points after hormone treatment was considered that gene expression was influenced by MeJA or SA treatment. According to the data presented in **Figure 4** and this criterion, gene expression patterns could be classified in four groups. The first group showed that both MeJA and SA increased gene expression at some time points but decreased it at other time points. This group included PtJAZ1, PtJAZ4 and PtJAZ7. The second group showed that both MeJA and SA decreased gene expression at all of the time points (except SA-24 h). This group included PtJAZ2, PtJAZ9, PtZML3, PtZML6, PtZML8, PtTIFY1 and PtTIFY2. Expression of most of the genes in this group was significantly decreased at JA-2 h, JA-6 h, JA-12 h, JA-24 h, SA-2 h, SA-6 h, and SA-12 h. In the third group, MeJA decreased gene expression at most time points, whereas SA increased it at some time points and decreased it at others. This group included PtJAZ3, PtJAZ5, PtJAZ10, PtPPD1, PtZML1, PtZML2 and PtZML7. In the fourth group, MeJA increased gene expression at some time points and decreased gene expression at the other time points, whereas SA decreased gene expression at all of the time points. This group included PtJAZ6, PtJAZ11 and PtJAZ12.

### Expression Profiles of Poplar TIFY Genes in Response to the Poplar Leaf Rust Pathogen M. larici-populina

Previous reports suggested JAZ protein might act as TFs in response to rust fungi (Duplessis et al., 2009). To identify TIFY genes that respond to poplar leaf rust disease, we collected

leaf samples from the poplar hybrid variety "NL895" that had been infected with M. larici-populina (see Materials and Methods); these samples were designated as CK, PI2, PI4, and PI8. The CK represented samples collected immediately after leaf rust infection, while PI2, PI4, and PI8 represented samples collected from leaves treated by urediniospore suspension of M. larici-populina after 2, 4, and 8 days, respectively. The symptom of leaf rust on leaves of "NL895" at these four selected time points were different. No visible symptom was observed at CK and 2 dpi; while very small white milk-white and plenty of uredinia could be observed at 4 and 8 dpi, respectively (data not shown). Previously, 48 hpi (equal to 2 dpi) was reported to be a key time point for the rust development (Rinaldi et al., 2007). Therefore, our plant material could be able to represent all key developmental stages of M. larici-populina on poplar leaves.

The relative expression of genes for PI2, PI4, and PI8 versus CK for all of the 24 TIFY genes was determined by qRT-PCR assays. We were able to test the expression of 18 of the 24 TIFY genes (**Figure 5**). According to the expression profiles of these 18 TIFY genes and using the criterion of fold changes less than 0.5 or larger than 2 and also with a p-value less than 0.05 when comparing CK with the other three treatments, we were able to classify them into four groups. The first group showed that gene expression was reduced by M. larici-populina infection at all the three stages of PI2, PI4, and PI8 when comparing with CK. This group included PtJAZ5, PtJAZ7, PtZML3, PtZML5 and PtZML6. The second group showed similar gene expression level of CK and PI8, while gene expression was largely reduced at PI2 and PI4 stages. This group had five genes and it included PtJAZ1, PtJAZ11, PtPPD1, PtTIFY1 and PtZML2. The third group showed that gene expression was similar at CK, PI2 and PI4 and highly increased at PI8. This group had six genes and it included PtJAZ2, PtJAZ6, PtJAZ9, PtJAZ12, PtTIFY2 and PtZML1. The fourth group showed that gene expression was reduced at PI2 and PI4 and highly increased at PI8. This group had two genes and it included PtJAZ3 and PtZML8. Based on these data, we could conclude that all tested TIFY genes respond to M. larici-populina infection and they also showed different responsive patterns.

FIGURE 5 | qRT-PCR assays of the expression profiles of the poplar TIFY gene family in response to M. larici-populina. The hybrid poplar variety "NL895" was used as plant material. CK indicated the control samples; 2, 4, and 8 dpi indicated samples collected from "NL895" leaves after 2, 4, and 8 days treatment by M. larici-populina inoculation. The virulent M. larici-populina strain used as fungal material in this qRT-PCR assay (see Materials and Methods). Y-axis indicates relative expression folds when comparison with CK.

### DISCUSSION

In this study, we were able to identify 24 TIFY genes in the poplar genome using the P. trichocarpa genome annotation version 3.0. However, the similar total number of TIFY genes between the two identifications could suggest the accuracy of our results. In other plants (e.g., Arabidopsis, rice, apple, soybean, and grape), the number of TIFY genes ranges from 18 to 34 (Ye et al., 2009; Bai et al., 2011; Zhang et al., 2012; Li et al., 2014; Sirhindi et al., 2016). This result suggested that the number of TIFY genes in poplar is within the normal range. In a number of previous studies, the expansion of a given gene family usually indicates preferential roles that they played during the plant's development. A largely expanded gene family in a plant genome sometimes results in unique characteristics for the plant. For example, when the genomes of P. trichocarpa and P. euphratica were compared, several gene families likely to be involved in tolerance to salt stress were found to be present in significantly greater gene copy numbers within the P. euphratica lineage (Ma et al., 2013). The expansion of TIFY genes in most plants did not show clear differences according to the information we obtained; this phenomenon might indicate that the function of TIFY genes in plants is essential and that they are involved in the basic plant processes.

The duplication pattern of a gene usually reveals how the gene was generated, how its function evolved and what roles it may played in plant growth and development (Wang et al., 2013). WGD is predicted to produce new genomic regions based on an ancestral genome, and this event occurs once or several times over many (sometimes hundreds of) millions of years (Wang et al., 2011). After WGD, the functions of the newly produced genes are usually modified to reduce genomic redundancy and could increase the plant's ability to adapt to various growth environments (Flagel and Wendel, 2009; Magadum et al., 2013). In other words, gene families that were chiefly produced by WGD usually existed in the "ancestral genome" before duplication. This indicates that the functions of members of a gene family are indispensable for plant growth and development in a given species. In addition, genes in families that were mainly produced by WGD usually evolve slowly (Wang et al., 2013; Soltis and Soltis, 2016). In this study, we found that 19 of the 24 TIFY genes were produced by WGD. This revealed that WGD made a large contribution to the generation of the TIFY gene family in poplar and this result is consistent with reports from the poplar genome sequencing project (Tuskan et al., 2006). It also suggested that the TIFY genes are indispensable for poplar growth and development; this prediction is consistent with the prediction from gene numbers in TIFY families of most of the plants investigated.

Since JA and SA are important hormones for signal transduction when plants encounter pathogens, we chose to investigate whether there was commonality between the TIFY gene expression profiles in poplar leaves after JA/SA treatments and leaf rust pathogen infection. TIFY family genes could be arranged into four groups both after JA/SA treatments and leaf rust pathogen infection. However, we were not able to see a clear rule (data not shown) after analyzing the overlapping TIFY genes between the two classifications. This result suggests

that although most of the TIFY genes can respond to a single elicitor, JA or SA treatment and leaf rust pathogen infection, the mechanisms underlying the response of the TIFY genes to each elicitor are different. This also indicates that gene regulation and signal transduction during plant responses to biotic stress are complicated.

Although there was no rule for the TIFY gene expression profiles in poplar leaves between JA/SA treatments and leaf rust pathogen infection, we were still able to obtain some valuable information by analyzing their expression profiles after phytohormone treatments and leaf rust pathogen infection in poplar. First, a large proportion of TIFY family genes can respond to JA/SA treatments and leaf rust pathogen infection. The expressions of all 20 tested TIFY genes were strongly influenced by JA/SA treatments (**Figure 4**), whereas all 18 tested TIFY genes were strongly influenced by leaf rust pathogen infection. Because TIFY family genes were reported to function in signal transduction when plants encountered various stresses and 65% of its members were found to respond to leaf rust pathogen infection, we speculated that the TIFY gene family in poplar also plays important roles in deploying the defense system against leaf rust via the JA/SA signal transduction pathway. Second, the TIFY genes which showed large changes in expression upon leaf rust pathogen infection could be used as a potential resource to increase resistance to the disease caused by M. larici-populina. The available transgenic platform that has been created in poplar allows overexpression or reduced expression transgenic lines to be generated. For example, the expressions of PtJAZ3, PtJAZ6, PtJAZ9, and PtJAZ15 were largely increased after leaf rust pathogen infection, and we can generate transgenic poplar lines that overexpress these genes (**Figure 5**). In contrast, the expressions of PtJAZ5, PtJAZ7, PtZML3, PtZML5 and PtZML6 were largely reduced after leaf rust pathogen infection, and we can generate transgenic poplar lines with reduced expression of these genes (**Figure 5**). Third, several previous studies showed a number of other genes could be able to respond to leaf rust infection and these genes would play important roles in defensive regulation in polar. For instance, a small number of pathogen-defense genes encoding PR-1, chitinases, and other pathogenesis-related proteins were found to be consistently upregulated throughout the whole leaf rust infection (Miranda et al., 2007); while 1,730 and 416 significantly differentially expressed transcripts were also found in the incompatible interaction between poplar and poplar rust (Rinaldi et al., 2007; Azaiez et al., 2009). Combining the previous reports with our study, we might be able to speculate that TIFY genes showed response to leaf rust infection could interact with genes reported in these studies. Finally, although TIFY family genes have

### REFERENCES


been reported to function in signal transduction when plants encounter various stresses, the detailed mechanisms of how they perform their functions are still not clear, especially in the woody perennial plant poplar. Therefore, more work to understand how TIFY genes perform their functions in poplar should be performed in the future.

In summary, we conducted a genome-wide identification of TIFY genes in poplar. The characteristics of this gene family was carefully investigated. We also analyzed their expression profiles in response to phytohormone treatments and Melampsora larici-populina infection in poplar. The results of this study have provided valuable information for further functional characterisation of TIFY genes in poplar and have also provided candidate genes for the improvement of disease resistance in poplar.

### AUTHOR CONTRIBUTIONS

NW, WX, HY, and PC organised the entire project and performed the experiments and data analysis. NW and JL wrote and edited this manuscript.

### FUNDING

Financial support for this work was provided by the National Natural Science Foundation of China (NSFC accession No.: 31670651) and the Fundamental Research Funds for the Central Universities (No. 520.902-2678436).

### ACKNOWLEDGMENT

The authors are grateful to AJE (American Journal Experts) for improving the English in this paper.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.00493/ full#supplementary-material

#### FIGURE S1 | Phylogeny of the TIFY proteins in the four representative

species. The protein names start with "At" indicate TIFY genes from Arabidopsis; the protein names start with "Os" indicate TIFY genes from rice; the protein names start with "Vv" indicate TIFY genes from grapes; the protein names start with "Pt" indicate TIFY genes from poplar.


defense gene families and gene expression profiling. Crit. Rev. Plant Sci. 28, 309–334. doi: 10.1080/07352680903241063


resistance to rice bacterial blight and its biosynthesis is regulated by JAZ protein in rice. Plant Cell Environ. 37, 451–461. doi: 10.1111/pce.12169


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Xia, Yu, Cao, Luo and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpls-08-00493 April 3, 2017 Time: 14:46 # 11

# Mixed Phenolic Acids Mediated Proliferation of Pathogens *Talaromyces helicus* and *Kosakonia sacchari* in Continuously Monocultured *Radix pseudostellariae* Rhizosphere Soil

Hongmiao Wu1, 2, 3 , Linkun Wu1, 2, 3, Juanying Wang1, 2, 3, Quan Zhu1, 2, 3, Sheng Lin1, 2, 3 , Jiahui Xu1, 2, Cailiang Zheng1, 2, 3, Jun Chen1, 2, 3, Xianjin Qin1, 2, 3, Changxun Fang1, 2, 3 , Zhixing Zhang1, 2, 3, Saadia Azeem1, 2, 3 and Wenxiong Lin1, 2, 3 \*

*<sup>1</sup> Fujian Provincial Key Laboratory of Agroecological Processing and Safety Monitoring, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> Key Laboratory of Biopesticide and Chemical Biology, Ministry of Education, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>3</sup> Key Laboratory of Crop Ecology and Molecular Physiology, College of Life Sciences, Fujian Agriculture and Forestry University, Fuzhou, China*

#### *Edited by:*

*Kumar Krishnamurthy, Tamil Nadu Agricultural University, India*

#### *Reviewed by:*

*Oswaldo Valdes-Lopez, National Autonomus University of Mexico, Mexico Pratyoosh Shukla, Maharshi Dayanand University, India*

> *\*Correspondence: Wenxiong Lin wenxiong181@163.com*

#### *Specialty section:*

*This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology*

*Received: 20 June 2015 Accepted: 03 March 2016 Published: 17 March 2016*

### *Citation:*

*Wu H, Wu L, Wang J, Zhu Q, Lin S, Xu J, Zheng C, Chen J, Qin X, Fang C, Zhang Z, Azeem S and Lin W (2016) Mixed Phenolic Acids Mediated Proliferation of Pathogens Talaromyces helicus and Kosakonia sacchari in Continuously Monocultured Radix pseudostellariae Rhizosphere Soil. Front. Microbiol. 7:335. doi: 10.3389/fmicb.2016.00335* *Radix pseudostellariae* L. is a common and popular Chinese medication. However, continuous monoculture has increased its susceptibility to severe diseases. We identified two pathogenic microorganisms, *Talaromyces helicus* M. (KU355274) and *Kosakonia sacchari* W. (KU324465), and their antagonistic bacterium, *Bacillus pumilus* Z. in rhizosphere soil of continuously monocultured *R. pseudostellariae.* Nine types of phenolic acids were identified both in the rhizosphere soil and in culture medium under sterile conditions. A syringic acid and phenolic acid mixture significantly promoted the growth of *T. helicus* and *K. sacchari*. *T. helicus* could utilize eight types of phenolic acids, whereas *K. sacchari* could only use four phenolic acids. *K. sacchari* produced protocatechuic acid when consuming vanillin. Protocatechuic acid negatively affected the growth of *B. pumilus*. The 3A-DON toxin produced by *T. helicus* promoted the growth of *K. sacchari* and inhibited growth of *B. pumilus* at low concentrations. These data help explain why phenolic exudates mediate a microflora shift and structure disorder in the rhizosphere soil of continuously monocultured *R. pseudostellariae* and lead to increased replanting disease incidence.

Keywords: *Radix pseudostellariae*, monoculture cropping problem, phenolic acids, *Talaromyces helicus*, *Kosakonia sacchari*, *Bacillus pumilus*

### INTRODUCTION

Radix pseudostellariae L. (Caryophyllaceae) is a common and popular Chinese medicine. It contains polysaccharides, ginseng saponins, flavonoids, cyclic peptides, amino acids, and trace elements (Hang and Wang, 2010). High-quality R. pseudostellariae is mainly produced in the ZheRong region of Fujian province in southern China, where soil and climate conditions are favorable for its growth. However, continuously monocultured R. pseudostellariae is prone to severe diseases, which result in reduced biomass of the plant tuberous roots (Lin et al., 2012a,b; Zhao et al., 2015). This phenomenon is known as replanting disease or soil sickness. It has been reported that more than 70% of medicinal plants, especially those with tuberous roots, such as R. pseudostellariae, have been attacked by replanting diseases (Zhang and Lin, 2009). Therefore, it has become the priority to study and overcome the consecutive monoculture problems, especially exhibited in medicinal plant production.

Replanting disease (Wu et al., 2009; Zhang and Wang, 2010), or "Sick Soil Syndrome," is a problem related to replanting in soils where the same species was previously grown. The sick soil syndrome refers to a combination of plant growth dysplasia, pests, and disease problems, which result in reduced yield and quality. It is a consequence of continuous monocropping for many years. Autotoxicity is an intraspecific allelopathy phenomenon, common in monocropping systems, where the plants inhibit their own growth through release of autotoxic chemicals into the soil (Singh et al., 1999; Wu et al., 2015). Terpenoids, phenolics, steroids, alkaloids, and cyanogenic glycosidic exudates are secreted from monocultured crop plant roots (Kato-Noguchi et al., 2002; Kato-Noguchi and Ino, 2003; Kulmatiski et al., 2008). Phenolic acids are important in allelopathy and replanting disease incidence (Bais et al., 2004; Kulmatiski et al., 2008; Wu et al., 2015; Zhao et al., 2015). Phenolic compounds are autotoxins of the plants in monocropping system, such as Rehmannia glutinosa, cucumber, and tobacco (Fang et al., 2010, 2013; Lin et al., 2010; Li et al., 2012; Zhou and Wu, 2012b; Wu et al., 2015). Some studies indicate that phenolic acids change the soil microbial community with possible influences on plant performance (Zhou et al., 2012, 2014). The allelochemicals released by plants might also promote growth of soil-borne pathogens and inhibit beneficial microbes (Kong et al., 2008; Pollock et al., 2011). The details of these phenomena are poorly known, especially in medicinal plants under monoculture regimes.

Most research in crop monocultures has focused on the effects of phenolic acids on the soil microbial community. There is less information on the dynamic changes of phenolics in soil and their effects on specific microbes. Therefore, we identified the types of phenolic acids in soil and their effects on the potential pathogens (Talaromyces helices M. and Kosakonia sacchari W.) and a beneficial microbe (Bacillus pumilus Z.) under laboratory conditions. It was also determined whether the pathogenic microorganisms were able to cause replanting disease in moncultured R. pseudostellariae, and the changes were also examined in the soil microbial community structure and its relationship to replanting disease of R. pseudostellariae under long term continuous monoculture conditions.

### MATERIALS AND METHODS

### The Isolation and Identification of Pathogenic Microorganisms

The pathogenic microorganisms were isolated using a tissue isolation method (Fang et al., 2011). After initial cleaning of the tuber roots of R. pseudostellariae sampled from the monoculture plots, we washed them for 30 min under running water, then used sterile water to wash them once on a sterile operation platform. A clean knife was used to cut the pathogen-infected tuber roots into slices. These slices were soaked in 75% ethanol for 30–40 s, washed twice with sterile water, dipped in 0.1% aqueous mercuric chloride solution for 7–10 min and washed again using sterile water (five times, for 2–3 min each time). The sterilized tissue slices of the infected roots were then cultured on MS medium at 32◦C, after which the pure strains were separated from the mixed bacterial or fungal community in the culture.

### PCR for Bacterial 16S rRNA Genes and Fungal its Regions

PCR amplification of partial bacterial 16S rRNA gene sequences was performed using the primer set 1405f and 456r (Zeng and Zheng, 2011). PCR assays were conducted in a 25-µL volume mix containing 12.5µL of Taq PCR Master Mix (2×) (Transgen, Beijing, China), 1µL of each primer, and 20 ng of purified DNA extracts. PCR conditions were: 94◦C for 5 min; 94◦C for 1 min; 61◦C for 1 min; 72◦C for 90 s, 35 cycles in total, with a final elongation at 72◦C for 10 min. The fungal community sizes were estimated using the primer sets ITS4 and ITS86 (Ferrer et al., 2001). The PCR conditions for the fungal community were: 94◦C for 5 min; 94◦C for 45 s; 51◦C for 45 s; and 7◦C for 1 min. PCR products were separated by electrophoresis on a 1.2% agarose gel and the bands were purified using the Universal DNA Purification Kit (TIANGEN, Beijing, China). The DNA sequences were analyzed using BLAST tools and the NCBI database to determine the species.

### Validation of Microbial Pathogenicity

Based on the tissue culture of R. pseudostellariae, we added 200µL of activator fungi, Kosakonia and B. pumilus (isolated from the rhizosphere of R. pseudostellariae) to the tissue culture. The control received equivalent amounts of PDA broth and LB medium instead. Then, the growth of the tissue culture seedlings co-cultured with the specific pathogenic bacteria was observed.

### Validation of *B. pumilus* Antagonists against *T. helicus*

The amended agar disk diffusion method (Bauer et al., 1966) was used to qualitatively screen the antagonists. Briefly, two strains of B. pumilus were streaked onto a potato dextrose agar (PDA) plate (square plate), and the Talaromyces helicus removed from an actively growing colony margin of T. helicus, was placed in the center of the plate. Plates were incubated for 8 d at 30◦C.

### Quantitative Real-Time PCR (qPCR) of *T. helicus*, *K. sacchari*, and *B. pumilus* Isolated from the *R. pseudostellariae* Rhizosphere Soil

According to the sequences of T. helicus, K. sacchari, and B. pumilus, taxon-specific primers including TH1F/TH1R, saesu-F/saesu-R, and BP-F/BP-R were set (Table S1). The identity and specificity of the primers (the amplifying gene) were confirmed (see Figure S5). Then we used the primers to amplify specific genes through the total DNA extracted from the R. pseudostellariae rhizosphere soil. And we also found these primers could amplify the special fragments corresponding to the T. helicus, K. sacchari, and B. pumilus.

The DNA sequences of T. helicus, K. sacchari, and B. pumilus were amplified using TH1F/TH1R, saesu-F/saesu-R, and BP-F/BP-R from the plasmid and were purified using a TIAN pure Mini Plasmid Kit [TIANGENBIOTECH (BEIJING) CO., LTD]. The plasmids contained the target DNA which can be represented the different microorganisms. We used these plasmids as standard curves to calculate the contents of T. helices, K. sacchari, and B. pumilus, respectively. The concentration of the target DNA was determined using a spectrophotometer (NanoDrop2000c, Thermo Fisher, USA), and was diluted to 1, 0.5, 0.1, 0.05, 0.01, 0.005, and 0.001 ng/µL. qPCR was monitored using a Mastercycler ep realplex (Eppendorf, Germany). Standard curve plotting and melting curve analysis was performed following the qPCR amplification instructions. A standard curve was created by plotting the target DNA concentration against the threshold cycle (Ct)-value exported from the Mastercycler ep realplex. The primer sets TH1F/TH1R, saesu-F/saesu-R, and BP-F/BP-R were evaluated using the established standard curve, and melting curves were determined using qPCR amplification of four replicates with a serial dilution of the target DNA template. The qPCR reactions were performed in 15-µL reaction mixtures (7.5 mL 2×SYBR Green PCR Master Mix, 0.6µL TH1F/TH1R, 0.6µL saesu-F/saesu-R, 0.6µL BP-F/BP-R, and 40 ng DNA made up to a final volume of 15µL with ddH2O). The PCR parameters for T. helicus were as follows: 94◦C for 5 min, followed by 35 cycles of 94◦C for 1 min, 56◦C for 45 s, and 72◦C for 45 s. The K. sacchari PCR parameters were 94◦C for 5 min, followed by 35 cycles of 94◦C for 1 min, 55◦C for 30 s, and 72◦C for 30 s, and the B. pumilus parameters were 94◦C for 5 min, followed by 35 cycles of 94◦C for 1 min, 60◦C for 45 s, and 72◦C for 45 s. After the qPCR run, melting curve analysis was performed to verify the specificity of the amplified product under the following conditions: 95◦C for 15 s, 60◦C for 15 s, followed by an increase to 95◦C over 10 min, and then maintained at 95◦C for 15 s.

### Phenolics Extraction and Determination Soil Phenolics Extraction

Soil samples were collected from the pots containing R. pseudostellaria at different growth stages for each treatment, including consecutive monoculture, newly planted, and control (no plant). Fresh soil was collected on April 22, May 5, June 6, and June 26 in 2014, and the sampled soil was sieved (2 mm mesh) to remove stones and plant residues. Soil phenolic acids of each sample were extracted using previously described methods (Dalton et al., 1987; Zhou and Wu, 2012a). Briefly, 5 g of each soil sample was added to 25 mL of 1 M NaOH and agitated for 24 h on a reciprocal shaker at 30◦C, then spun in a vortex generator for 30 min at maximum speed. The suspension was centrifuged at 10,000 rpm for 10 min, and the liquid supernatant was collected. The filtrate was adjusted to pH 2.5 using 9 M HCl, and extracted five times with ethyl acetate. The resultant extracts were pooled and evaporated to dryness at 35◦C. The residue was dissolved in 5 mL methanol using ultrasound for 5 min and maintained in the dark at 4 ◦C.

### Phenolics Extraction from the Tissue Culture Medium of R. pseudostellariae

The tissue cultures of R. pseudostellariae were incubated for 150, 215, 279, 305, and 355 d. Each treatment contained 60 mL of medium. Then, the used medium from each treatment was collected for extraction of the phenolic acids. Each sample was added to 40 mL of 1 M NaOH and agitated for 90 min in an ultrasonic generator. The other conditions were the same as those mentioned above.

### Phenolics Determination

The methanol solution of the extracts was filtered through 0.22 µm filter membrane for HPLC analysis. The phenolic acids in each sample were determined using a Waters HPLC system (C18 column: Inertsil ODS-SP, 4.6 × 250 mm, 5µm). The mobile phase A was methanol and mobile phase B was 2% glacial acetic acid. The flow rate was kept constant at 0.7 mL/min. Detection was performed at 280 nm. The injection volume was 20µL and the column temperature was maintained at 30◦C. Identification and quantification of phenolic compounds were confirmed by comparing retention times and areas with pure standards.

### The Influence of Phenolic Acids on R. pseudostellariae

Based on the measurement results of the phenolic acids content and the ratio of the mixed phenolic acids detected in R. pseudostellariae rhizosphere soil, phenolic acids standards were used to form a "mixed phenolic acid solution (P1)" that simulated the average ratio of various phenolic acids detected in R. pseudostellariae rhizosphere soil (FY, SF, TF; i.e., gallic acid: coumaric acid: p-hydroxybenzoic acid: vanillic acid: syringic acid: vanillin: ferulic acid: benzoic acid = 3:11:26: 119:14:14:25:28).

In addition, a series of concentration gradients of the mixed phenolic acids were set at 0, 160, 320, 960, and 1280 µmol/L for the tissue culture of R. pseudostellariae. Each treatment had three replicates. Then, the tissue culture was incubated for 50 d. Plant dry mass was measured after oven drying at 70◦C to a constant mass.

### The Influence of Phenolic Acids on the Physiological Characteristics of *T. helices* The Impact of Single and the Mixed Phenolic Acids on the Diameter Growth of the R. pseudostellariae Biotype T. helicus

Soil extract-medium (SEM) preparation: 1 kg of soil and 1 L of double distilled water were shaken for 30 min on the shaker, then sterilized at 121◦C for 15 min. The leachate was filtered through talc in a double suction filter. The filtrate was collected and adjusted to neutral pH and the mother liquor retained for further use. According to method requirements, double distilled water was used to dilute the soil solution with appropriately diluted replicates, and then an appropriate amount of agar (15 g/L) was added to the solution, which was sterilized before use.

Based on the measurements of phenolic acid content in the R. pseudostellariae rhizosphere soil, a series of phenolic acid concentration gradients were set. The concentration gradients were 0, 40, 80, 160, 320, and 960µmol/L. When the 10-fold dilution of the SEM medium was cooled (to about 40–50◦C) after sterilization, appropriate amounts of various stock phenolic solutions were added (dissolved using methanol), after filtration through a 0.22-µm membrane, then immediately poured into the plates. The control received only double distilled water and an equal volume of methanol in order to exclude the effects of methanol on the growth of T. helices. Each treatment was replicated three times. After preparation of the phenolic acid-SEM plates, the activated T. helices spores were cultured in the center of the plate, then placed in a 30◦C constant temperature incubator for 9 d, after which the mycelium diameter was measured.

Based on the phenolic acid content and the ratio of the mixed phenolic acids (P1), a series of phenolic concentrations were established, i.e., 0, 40, 80, 160, 320, and 960 µmol/L. The other conditions were the same as those mentioned above.

### Effect of the Mixed Phenolic Acids on the Toxin Production of the R. pseudostellariae Biotype T. helicus

The dilution ratio of the soil filtrate and its preparation were the same as mentioned above. The concentrations of the mixed phenolic acids were 0, 80, 160, and 960µmol/L. An equal amount of T. helices spores filtrate was added to the SEM liquid medium in the bottle. Then, the bottle was maintained at 30◦C on a 180-rpm thermostatic shaker for 8 d. After culture, HPLC techniques were used to determine the content of two common toxins (3A-DON: 3-Acetyldeoxynivalenol and 15A-DON: 15-O-Acetyl-4-deoxynivalenol). To draw a standard curve, the concentrations of the two types of toxin standards used were 0.5, 1, 2, 5, and 10 ppm. The chromatographic conditions were as follows: chromatographic column: C18 column (Inertsil ODS-SP, 4.6 × 250 mm, 5µm); mobile phase A: acetonitrile; mobile phase B: 0.005% phosphoric acid solution; elution gradient: mobile phase B 95% (0 min) → 30% (9 min) → 0% (18 min) → 0% (23 min) → 95% (23.01 min) → 95% (45 min); oven temperature: 35◦C; detection wavelength: 224 nm; velocity: 0.7 mL/min.

The extraction method for T. helices toxins was as follows. The liquid fermented for 10 d was first filtered using double filter paper and then filtered using a 0.22-µm ultrafiltration membrane. The filtrate (15 mL) was added to a bottle containing the same volume of ethyl acetate, and placed on the shaker for 8 h at 28◦C, 160 rpm for oscillation extraction. This was followed by centrifugation at 10,000 rpm for 3 min at 4 ◦C. The upper layer organic phase was transferred into a new 50-mL centrifuge tube. After concentration by vacuum and drying, 500µL of double evaporated water was added to dissolve the solid residue, then subjected to ultrasound and analyzed immediately using HPLC. The sample toxins were quantitatively analyzed by comparison to the standard curve.

### Detecting the Utilization of Various Phenolic Acids by the *R. pseudostellariae* Biotype *T. helices*

The soil leachate mother liquid (48 mL) was diluted three times and placed into a 250-mL triangle flask. After sterilization and precooling, 0.3 mL of the phenolic acid mixture (coumaric acid, protocatechuic acid, gallic acid, p-hydroxybenzoic acid, vanillic acid, syringic acid, vanillin, ferulic acid, benzoic acid, 4800µmol/L) was added to the flask. These phenolic acids had been detected in the rhizosphere soil of R. pseudostellariae. Then, 250µL of T. helices spore filtrate was added to the flask. All flasks containing the spores were placed in the constant temperature oscillation shaker at 30◦C, 180 rpm. A 1 mL sample of bacterial liquid was collected at 0, 6, 18, 26, 38, 47, 53, 68, 72, and 75 h, filtered through a 0.22-µm ultrafiltration membrane, and then loaded into an HPLC bottle. These samples were analyzed by HPLC using the same chromatographic conditions as previously described.

### Affects of Phenolic Acids on the Physiological Characteristics of *K. sacchari*

### The Affects of Single and Mixed Phenolic Acids on the Growth of K. sacchari

The LB liquid culture medium was diluted six times, placed into glass tubes, and subjected to 121◦C and high pressure sterilization for 20 min. When the culture medium had cooled, an appropriate amount of each phenolic acid stock solution (such as gallic acid, coumaric acid, p-hydroxybenzoic acid, vanillic acid, syringic acid, vanillin, ferulic acid, benzoic acid), which had passed through a 0.22-µm ultrafiltration membrane and K. sacchari liquid (30µL) that had already been activated were added to each tube. Then, all the tubes, which were maintained at 37◦C, were placed on a thermostatic shaker at 200 rpm for 8–10 h. Finally, 200µL of bacterial liquid was transferred to a 96-well enzyme-labeled plate, and a standard enzyme instrument (Thermo Scientific Multiskan MK3, Shanghai, China) was used to determinate the absorbance values at OD 600 nm.

The LB liquid culture medium was diluted six times and added to an appropriate amount of phenolic acid stock solution P1 previously filtered through a 0.22-µm filtration membrane. The final concentration of each phenolic acid was: 0, 30, 60, 120, 240, 480, and 960µmol/L. The other conditions were the same as those mentioned above.

### Detection Conditions for Phenolic Acids used by K. sacchari

Soil leachate mother liquor (48 mL) was diluted three times and placed into a 250-mL triangle flask. After sterilization and cooling, eight types of phenolic acids (coumaric acid, gallic acid, p-hydroxybenzoic acid, vanillic acid, syringic acid, vanillin, ferulic acid, and benzoic acid, 4800µmol/L; 0.3 mL each) were added to the liquor. These phenolic acids were detected in the rhizosphere soil of R. pseudostellariae. Then, 50µL of K. sacchari fluid which had been activated, was added to the flask. All the flasks were placed on the oscillation table at a constant 37◦C, 180 rpm. From then on, At 0, 3, 6, 9, 12, 15, 26, 29, 32, 35, 38, 41, 48 h, 1 mL samples of bacteria liquid were taken, filtered through 0.22µm ultrafiltration membrane, and loaded into HPLC bottles. These samples were immediately stored at 4◦C or analyzed using HPLC with the same chromatographic conditions as noted earlier.

### Detection Conditions for Each Single of Phenolic Acids used by K. sacchari

Every triangle flask contained a single phenolic acid, and the culture time was set at 0, 15, 28, 37, and 47 h. The other conditions were the same as those mentioned above.

### Toxin Production and Autotoxicity Bioassay of *K. sacchari* and *B. pumilus*

The LB liquid culture medium was diluted four times and added to an appropriate amount of single toxin (e.g., 3A-DON, 15A-DON), which had been filtered through a 0.22-µm filtration membrane. The final concentrations of each toxin were: 0, 0.005, 0.01, 0.02, 0.04, 0.08, 0.16, and 0.32 mg/L. Bacterial liquid (60 µL), previously activated, was added to each tube. All tubes were maintained at 37◦C on a thermostatic shaker at 200 rpm for 7 h. Finally, bacterial liquid (200µL) was placed in a 96-well enzymelabeled plate, and a standard enzyme instrument was used to determinate the absorbance values at OD 600 nm.

### The Effect of the Reaction Intermediate, Protocatechuic Acid, on the Beneficial Microorganism *B. pumilus*

The concentrations of protocatechuic acid were: 0, 30, 60, 120, 240, 480, and 960µmol/L. B. pumilus liquid (60µL), previously activated, was added to each tube. The other conditions were the same as those mentioned above.

### Statistical Analysis

Differences among the treatments were calculated and statistically analyzed using the analysis of variance (ANOVA) and the LSD multiple range test (p < 0.05). The Statistical Package for the GraphPad Prism version 5.1 and the Data Processing System (DPS) version 7.05 were used for statistical analysis.

### RESULTS

### Identification of Microorganisms and Validation of Their Pathogenicity

The DNA sequences were analyzed by means of BLAST tools and the NCBI database. The detected microorganisms were K. sacchari (KU324465) and T. helices (KU355274) which were highly pathogenic to the tissue culture plantlets of R. pseudostellariae (Figures S1, S2). However, B. pumilus Z. was not pathogenic to the tissue culture plantlets (Figure S3) and suppressed the mycelial growth of T. helicus when it was co-cultured with the pathogen (Figure S4).

### Dynamics of *T. helicus*, *K. sacchari*, and *B. pumilus* in the Rhizosphere Soil of *R. pseudostellariae*

The qPCR analysis showed a significant increase in the amount of pathogenic T. helicus and K. sacchari in the rhizosphere of R. pseudostellariae as the number of monoculture years increased, especially around the site of infected R. pseudostellariae (SS). This is consistent with the observation that soil-borne diseases become more severe with an increase in the number of monoculture years. However, B. pumilus showed the opposite trend, which tended to increase in 2-year monoculture soil and then significantly decrease in the 3-year monoculture soil (**Figure 1**).

### Component Identification and Analysis of Phenolic Acid Content in *R. pseudostellariae* Rhizosphere Soil and Tissue Culture Medium

HPLC was used to determine phenolic acids in the rhizosphere soil of R. pseudostellariae at different growth stages in different years of monoculture. We identified nine types of phenolic acids in soil. These were gallic acid, coumaric acid, protocatechuic acid, p-hydroxy benzoic acid, vanillic acid, syringic acid, vanillin, ferulic acid, and benzoic acid as shown in **Figure 2**, Figure S6, and **Table 1**. We also found these phenolic acids in the tissue culture medium of R. pseudostellariae (**Figure 3**).

HPLC analysis showed that the phenolic acid levels in the monocultured rhizosphere soil varied with different plant growth stages. For example, most of the phenolic acids fluctuated. They increased initially and then declined (**Table 2** and Figure S7). However, phenolic acid levels tended to increase in the culture

TABLE 1 | Regression equation and detectable limitations.

and third cropping of *R. pseudostellariae*, respectively, grown in pots with the soil from the same plot.


represents vanillic acid; 6, represents syringic acid; 7, represents vanillin; 8, represents ferulic acid; 9, represents benzoic acid; FY, SY, TY: represent the first, second,

*Y, means the concentration of phenolic; x, means the peak area; A, means the concentration of toxins.*

medium as the growth of the tissue culture plantlets increased. Gallic acid, P-hydroxybenzoic acid vanillic acid, and benzoic acid significantly accumulated in the culture medium (**Figure 3**). These phenolic acids did not show autotoxicity toward the culture plantlets of R. pseudostellariae (Figure S8). This implies that microorganisms might be involved in the variability of phenolic acid levels in monocultured rhizosphere soil.

### The Influence of Phenolic Acids on the Physiological Characteristics of the *R. pseudostellariae* Biotype *T. helices*

Mixed phenolic acids significantly promoted the mycelial growth of T. helices. The greatest increase occurred when exposed to the dosage at 160µmol/L of the mixed phenolic acids (**Figure 4**). All treatment concentrations of phenolic acids in the mixture had a positive effect on mycelial growth of T. helicus. The various single phenolic acids had different influences on the mycelial growth. Syringic acid had the greatest positive effect on the T. helicus

at 40µmol/L, whereas vanillin and gallic acid showed variable effects, with a growth promotion effect at a low concentration and a suppressive effect at a high concentration. Hydroxybenzoic acid, ferulic acid, protocatechuic acid, coumaric acid, vanillic acid, and benzoic acid had no significant effect on T. helicus at a low concentration, but they had an inhibitory effect at high concentrations (**Figure 4**). We suggest that not all of the individual phenolic acids can boost the growth of T. helices. Some phenolic acids showed an inhibitory effect. However, the phenolic acids mixed in the same ratio as the result of HPLC analysis for monocultured R. pseudostellariae rhizosphere soil significantly promoted the growth of the pathogenic T. helicus strains.

The soil extracts, which were diluted 10 times and used as the culture medium for T. helicus, contained two types of toxins, 3A-DON and 15A-DON. The concentration of 3A-DON was significantly higher than that of 15A-DON (**Table 1** and Figure S9). However, the production of the 15A-DON toxin can be promoted by T. helicus when exposed to increasing dosages of the mixed phenolic acids, and the toxin content increased sharply, then reached its highest value at 160 µmol/L. The 15A-DON toxin levels were also greater than those in the control group (0µmol/L). The production of 3A-DON also increased with an increase in the treatment concentration of the mixed phenolic acids. This suggests that the mixed phenolic acids were favorable for the mycelial growth and toxin production of the pathogenic bacteria, such as T. helicus (**Figure 5**).

HPLC was used to detect the use of nine phenolic acids by the pathogenic fungus, T. helices, which was detected in the rhizosphere soil of the monocultured R. pseudostellariae. The results showed that T. helicus can use eight types of phenolic acids, i.e., gallic acid, coumaric acid, protocatechuic acid, phydroxy benzoic acid, vanillic acid, vanillin, ferulic acid, and benzoic acid (**Figure 6**). However, syringic acid, which had a stimulatory effect on the pathogenic fungus at a low treatment concentration but an inhibitory effect at a high concentration, could not be used although a stimulatory effect was observed on the mycelial growth of T. helicus at treatment concentrations of more than 40µmol/L.

TABLE 2 | The dynamic changes of phenolic acids in the rhizosphere soil of *R. pseudostellariae* at different growth stages in a continuous cropping system.


↑*, Means an increasing trend; V, means an initial declining trend followed by an increase, whereas,* ∧*, means an increasing trend followed by a decline;* ↓*, means a declining trend.*

### The Influence of Phenolic Acids on the Physiological Characteristics Of *K. sacchari*

Based on the molar proportions of the phenolic acids in the soil, we made a series of mixtures with eight types of phenolic acids at concentrations of 0, 30, 60, 120, 240, 480, and 960µmol/L. These were used to analyze the impact on the growth of K. sacchari. Different concentrations of the mixed phenolic acids had different effects on K. sacchari compared with T. helices. Low concentrations of the phenolic acid mixture had a stimulatory effect on specific pathogens, whereas the reverse was true at high treatment concentrations. Mixed phenolic acids promote the growth of K. sacchari, but only within a specific range of concentrations. Each phenolic acid had distinct effects on the growth of K. sacchari. Syringic acid had the greatest effect on the pathogenic bacterium at a treatment concentration of 60µmol/L. Gallic acid, p-hydroxybenzoic acid, vanillic acid, and benzoic acid promote growth at low concentrations but suppress it at high concentrations. Coumaric acid, protocatechuic acid, vanillin, and ferulic acid had no significant effect at low concentrations (**Figure 7**). Thus, the main allelochemicals of R. pseudostellariae root exudates could have distinctly positive or negative effects on the growth of pathogenic fungi and bacteria, implying that root exudates can exert either positive or negative selection on specific microbes.

Based on the consumption of eight phenolic acids detected in the soil, we found that K. sacchari could only use four types, i.e., gallic acid, coumaric acid, vanillin, and ferulic acid (**Figure 8**). The utilization efficiency of vanillin was the highest of the four compounds. K. sacchari is able to produce protocatechuic acid from consumption of vanillin (**Figure 8**).

### The Influence of Toxin Production and Protocatechuic Acid on the Growth of *K. sacchari* and *B. pumilus*

The 3A-DON toxin, at a low concentration promoted the growth of K. sacchari and inhibited the growth of B. pumilus. In contrast, the 15A-DON toxin had no significant effect on their growth at a low treatment concentration (Figures S10, S11). Protocatechuic acid, at 60µmol/L, had a negative effect on the growth of the beneficial B. pumilus Z. (**Figure 9**).

### DISCUSSION

Previous research related to plant allelopathy or the negative effects of continuous monoculture mainly focused on isolation, identification, and bioassay of allelochemicals, and the evaluation of the direct influence of allelochemicals on the growth of receptor plants (Li et al., 2012). However, the composition and effective concentration of allelochemicals remains controversial. Some believe that the effective concentration of phenolic allelochemicals that can inhibit weeds or stunt crop growth is much higher than that found under field conditions. It is now known that allelochemicals released into the soil can be catabolized, transformed, and processed by microorganisms (Eisenhauer et al., 2012). Allelopathy can operate indirectly between the donor and the recipient plant or between crop residue and new plantings (Kato, 2009; Lin, 2013). Positive or negative effects of plant-plant interactions, such as plant allelopathy, allelopathic autotoxicity, intercropping benefits, or exotic plant invasion, can result from interactions between plants and specific microorganisms, which are often mediated by root exudates (Wang et al., 2013). Plants can transport the carbon fixed by photosynthesis to below ground, where it is released into the soil and provides a source of carbon and energy for microbial growth. Simultaneously, the changes in microbial community structure will affect plant growth both below and above ground (Wardle et al., 2004). Lin et al. (2011) added major rice allelochemicals (p-hydroxy benzoic acid, ferulic acid, salicylic acid, vanillic acid, cinnamic acid) to the soil and found that these phenolic acids decreased by approximately 50–90% from 3 to 7 d. This may be related to microbial decomposition and utilization.

letters are statistically different (LSD-test, *p* < 0.05).

In the present study, we found that many of the phenolic acids produced by tissue culture plantlet roots of R. pseudostellariae accumulated in the culture medium and showed no autotoxicity over the period of plantlet growth (Figure S8). However, these phenolic acids did not increase over the years of monoculture and they did not accumulate in the rhizosphere soil. What is more, the levels of some phenolic acids detected in consecutively monocultured soil were lower than the levels in normal cropping soil, and most of the concerned phenolic acids increased initially and then declined after a several years of continuous cropping. In addition, the content of all the phenolic acids concerned was variable among the different growth stages due to secretions of root exudates, decomposition, synthesis, and transformation of allelochemicals by soil microbes in different monoculture systems. This findings suggest that the consecutive monoculture problems are not due to the direct effects of high phenolic acid concentrations on the receptor plants, but more likely related to indirect effects on plant growth caused by changing microbial flora as shown in **Figure 10**. Zhou and Wu (2012b) found that after 7 years of continuous cucumber cropping, the plant biomass had declined to the lowest level in the study. However, after 7th year of monocropping, the plant biomass gradually increased. The content of total soil phenolics and some phenolic acids (such as ferulic acid, p-hydroxybenzoic, p-coumaric acid, etc.) showed a similar trend, i.e., the content was lowest in the 7th year, after which it gradually increased. This may be the reason that the most serious disease problems result from the rhizosphere interaction involved in shifts of microbial community differentially mediated by root exudates in monocropping system.

T. helices and K. sacchari used in this study were validated as important soil-borne pathogens. We found different phenolic acids differentially mediate the proliferation of the two targets. T. helices and K. sacchari cannot directly utilize syringic acid as carbon or nitrogen sources. But they may play an important role in promoting the growth of pathogenic microorganisms in the rhizosphere soil of R. pseudostellariae. Many studies (Zhang et al., 2008; Wei et al., 2010) have demonstrated that syringic acid functioning as a signal molecule, promotes cell proliferation and regulates the cell cycle. Some researchers also reported that the K. sacchari was isolated from infected stem, root or rhizosphere soil of sugar cane (Zhu et al., 2013; Gu et al., 2014). We have used the quantitative real-time PCR to detect it in the rhizosphere soil of R. pseudostellariae. The result showed a significant increase in the amount of pathogenic K. sacchari in the rhizosphere of R. pseudostellariae as the number of monoculture years increased, especially around the site of infected R. pseudostellariae (SS). K. sacchari was highly pathogenic to the tissue culture plantlets of R. pseudostellariae (Figure S1). And the pathogen could also affect other beneficial bacteria population such as B. pumilus used in this study via its toxin production and its metabolic intermediate. Our results demonstrated that, K. sacchari was able to utilize root exudate, vanillin to produce an intermediate product, protocatechuic acid (PA), which was considered as a better antibacterial substance in previous studies. We also found that PA did not suppress the growth of the pathogenic K. sacchari, but it did have negative effects on the growth of the beneficial, B. pumilus at a low concentration. B. pumilus is a PGPR, which has been successfully used for the biological control of damping-off diseases (Huang et al., 2012). Tadych et al. (2015) also proposed quinic acids could significantly influence virulent facctors of the fruit rot fungi. In our studies, the 3A-DON toxin production of the pathogenic, T. helices, can be promoted by some phenolic acids, especially by the phenolic acids mixed in the same ratio as detected in monocultured rhizosphere soil, then a cascade reaction was triggered off to promote the growth of the other pathogenic fungus, K. sacchari and inhibited the growth of its counterpart, the beneficial bacterium, B. pumilus, (Figure S4). An increase in populations of soil-borne pathogenic fungi (e.g.,

Fusarium oxysporum) is likely responsible for soil sickness (Qi et al., 2009), but the present study also confirmed that the two soil-borne pathogenic microorganisms were also responsible for the monoculture problems. Therefore, it is very complicated by the fact that monoculture problems are involved in an intricate rhizosphere interaction. The imbalanced population structure, with the higher population sizes of T. helices and K. sacchari but lower B. pumilus population influenced by the mixed phenolic acid exudates, may partially account for the soil sickness of R. pseudostellariae.

Our results unveil the underlying mechanism of replanting diseases of R. pseudostellariae popularly in modern agriculture system, especially in nowadays China. The case study deeply illustrates about how root exudates play important roles in the formation of monoculture problems involved in rhizosphere interactions under monocropping regimes. As we know, root exudates are carbon sources for soil microbial growth and also signal substances that function as communication tools between plants and microorganisms in the rhizosphere (Mandal et al., 2010). Peters et al. (1986) found that expression of the nodulation gene (nodD) in nodule bacteria is directly related to legume root secretions of flavonoids and isoflavonoids. Venkatachalam et al. (2012) showed that tomatoes whose leaves were infected by pathogens were able to attract more beneficial bacteria such as B. subtilis to the rhizosphere by regulating root secretion. This was accomplished by increasing the synthesis and release of malic acid, leading to induced systemic resistance allowing plants to better defend against further pathogen infestation. Rousk et al. (2010) also found the pH gradients could change the diversity of the bacterial and fungal communities in an arable soil. Root exudates play an important role in the interaction between plants and microbes. They can have different effects on different microbes, which include both stimulatory and inhibitory influences. For example, corn root secretion of DIMBOA (DIMBOA, 2,4-dihydroxy-7-methoxy-2H-1, 4-benzoxazin-3 (4H)-one) and Ocimum basilicum L. secretion of rosmarinic acid resulted in strong antibacterial activity (Bais et al., 2006; Berendsen et al., 2012). Zhou et al. (2012) found that the allelopathic autotoxicity of coumaric acid from cucumber can significantly promote the growth of the soil pathogen F. oxysporum. Others have found that root exudates of maize, peanuts, watermelons, and American ginseng can significantly promote the growth of pathogens, leading to increased soil-borne disease (Ju et al., 2002; Li et al., 2009; Hao et al., 2010). Benizri et al. (2005) showed that the number of beneficial bacteria was reduced and the number of pathogenic bacteria was increased in continuously cropped peach soil. Our results show the same phenomenon, which is attributed to the changes in soil microbial composition and structure differentially mediated by phenolic acids of root exudates.

This findings provide a clue to open a new avenue for modulating the root microbiome to enhance medicinal plant production and sustainability. Such approaches might include the use of microbial fertilizer and organic amendment application to keep the balance of microbial community in rhizosphere soil in modern agriculture system. And the amendment of the root exudates in R. pseudostellariae can also be used to relieve the problems. Additional work is needed to further understand the intrinsic mechanism of these specific microbial functions in the future.

### REFERENCES


### FUNDING

This work was supported by the National Science Foundation of China (Grant no. U1205021, 81573530, 81303170, 31401950, 31301858), the National Key Basic Research Program of China (Grant no. 2012CB126309), and the Health and Family planning Program of Fujian province (Grant no. WKJ-FJ-34).

### ACKNOWLEDGMENTS

We thank National Science Foundation of China and the National Key Basic Research Program of China for providing the funds used in this work. This work was also supported by Fujian-Taiwan Joint Innovative Center for Germplasm Resources and cultivation of crop [Fujian 2011 Program, (2015), 75].

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.00335

Enterobacter sacchari Zhu et al. 2013 as Kosakonia sacchari comb. nov. Int. J. Syst. Evol. Microbiol. 64, 2650–2656. doi: 10.1099/ijs.0.064709-0


(in Chinese). Chin. J. Ecol. 31, 106–111. doi: 10.13292/j.1000-4890. 2012.0010


glutinosa root exudates under consecutive monoculture. Sci. REP UK 5:15871. doi: 10.1038/srep15871


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Wu, Wu, Wang, Zhu, Lin, Xu, Zheng, Chen, Qin, Fang, Zhang, Azeem and Lin. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Screening of Rhizospheric Actinomycetes for Various In-vitro and In-vivo Plant Growth Promoting (PGP) Traits and for Agroactive Compounds

### Sumaira Anwar, Basharat Ali and Imran Sajid\*

Department of Microbiology and Molecular Genetics, University of the Punjab, Lahore, Pakistan

#### Edited by:

Anton Hartmann, Helmholtz Zentrum München, Germany

#### Reviewed by:

Lei Zhang, Washington State University, USA Zakira Naureen, University of Nizwa, Oman

> \*Correspondence: Imran Sajid imran.mmg@pu.edu.pk

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 14 April 2016 Accepted: 12 August 2016 Published: 29 August 2016

#### Citation:

Anwar S, Ali B and Sajid I (2016) Screening of Rhizospheric Actinomycetes for Various In-vitro and In-vivo Plant Growth Promoting (PGP) Traits and for Agroactive Compounds. Front. Microbiol. 7:1334. doi: 10.3389/fmicb.2016.01334 In this study 98 rhizospheric actinomycetes were isolated from different wheat and tomato fields, Punjab, Pakistan. The isolates were characterized morphologically, biochemically, and genetically and were subjected to a comprehensive in vitro screening for various plant growth promoting (PGP) traits. About 30% of the isolates screened were found to be the promising PGP rhizobacteria (PGPRs), which exhibited maximum genetic similarity (up to 98–99%) with different species of the genus Streptomyces by using16S rRNA gene sequencing. The most active indole acetic acid (IAA) producer Streptomyces nobilis WA-3, Streptomyces Kunmingenesis WC-3, and Streptomyces enissocaesilis TA-3 produce 79.5, 79.23, and 69.26µg/ml IAA respectively at 500µg/ml L-tryptophan. The highest concentration of soluble phosphate was produced by Streptomyces sp. WA-1 (72.13 mg/100 ml) and S. djakartensis TB-4 (70.36 mg/100 ml). All rhizobacterial isolates were positive for siderophore, ammonia, and hydrogen cyanide production. Strain S. mutabilis WD-3 showed highest concentration of ACC-deaminase (1.9 mmol /l). For in-vivo screening, seed germination, and plant growth experiment were conducted by inoculating wheat (Triticum aestivum) seeds with the six selected isolates. Significant increases in shoot length was observed with S. nobilis WA-3 (65%), increased root length was recorded in case of S. nobilis WA-3 (81%) as compared to water treated control plants. Maximum increases in plant fresh weight were recorded with S. nobilis WA-3 (84%), increased plant dry weight was recorded in case of S. nobilis WA-3 (85%) as compared to water treated control plants. In case of number of leaves, significant increase was recorded with S. nobilis WA-3 (27%) and significant increase in case of number of roots were recorded in case of strain S. nobilis WA-3 (30%) as compared to control plants. Over all the study revealed that these rhizospheric PGP Streptomyces are good candidates to be developed as bioferlizers for growth promotion and yield enhancement in wheat crop and can be exploited for the commercial production of different agro-active compounds.

Keywords: plant growth promoting Streptomyces, indole acetic acid (IAA), wheat, 16S rRNA gene sequencing, biofertilizers, agro-active compounds

### INTRODUCTION

Effective farming practices, now-a-days, rely on extensive use of chemical fertilizers in order to enhance plant growth and yield. However, the cost, environmental concerns and the resulting human health hazards due to the inclusion of these chemical fertilizers in food chain are the major limiting factors. Microorganisms have been considered an important source of natural compounds of agro active importance. Use of microbial consortia in the form of bio fertilizers for reduction in the application of chemical fertilizers, pesticides and related agrochemicals, without compromising the plant yield is currently a significant research area in the field of agriculture, microbiology, and biotechnology (Ahmad et al., 2008).

Plant growth promoting rhizobacteria (PGPR) is a group of naturally occurring, free living rhizosphere colonizing bacteria that improve plant growth, increase yield, enhance soil fertility, and reduce pathogens as well as biotic or abiotic stresses (Vessey, 2003; Kumar et al., 2014). PGPR help the plants by producing plant growth phytohormones such as indole acetic acid (IAA), cytokinins, and gibberellins (Marques et al., 2010), solubilization of inorganic phosphate (Jeon et al., 2003), asymbiotic nitrogen fixation (Khan, 2005), antagonistic effect against phytopathogenic microorganisms by producing siderophore, antibiotics, and fungicidal compounds (Lucy et al., 2004; Barriuso et al., 2008; Majeed et al., 2015).

Actinomycetes are present extensively in the plant rhizosphere and produce various agroactive compounds. In the last few years, this group of bacteria, due to its strong antimicrobial potential, and soil dominant saprophytic nature, gained much attention as plant growth promoters (PGP; Franco-Correa et al., 2010). Actinobacteria can actively colonize plant root systems, can degrade a wide range of biopolymers by secreting several hydrolytic enzymes and tolerate hostile conditions by forming spores (Alexander, 1977). Actinobacteria, especially Streptomyces, also exhibit immense biocontrol action against a range of phytopathogens (Wang et al., 2013). Actinobacteria can produce phytohormones (IAA) and siderophore as well as solubilize phosphate and promote plant growth (Jeon et al., 2003). Actinomycetes have been mainly exploited in pharmaceutical industry since 1940s, Whereas, only a few have been developed as commercial products for plant application in agriculture (Minuto et al., 2006). Streptomycetes have been long considered simply as free-living soil inhabitants, but recently the importance of their complex interactions with plants, and other organisms is being uncovered (Seipke et al., 2011).

Interest in the beneficial rhizobacteria associated with cereals has increased recently and several studies clearly demonstrated the positive and beneficial effects of PGPR on growth and yield of different crops especially wheat at different environments under variable ecological conditions (Marques et al., 2010; Mehnaz et al., 2010; Zhang et al., 2012). Wheat (Triticum aestivum) ranked third most abundant cereal crop after maize and rice and covers approximately 30% of the total cereal products worldwide (Fageria and Baligar, 1997). Pakistan is the 18 largest wheat producing country in the world (Tunio et al., 2006) with an average of 20–24 million tons of wheat grown a year. Wheat is grown on major areas in Pakistan but its average yield per hectare is far less than the actual potential (Akhtar et al., 2009). Understanding the diversity and distribution of indigenous actinobacteria in the rhizosphere of particular crops is depended on the knowledge of native actinobacterial populations, their isolation, identification, and characterization. It is therefore mandatory to explore region specific actinobacterial strains that can be used as growth promoters to achieve desired crop production (Deepa et al., 2010). However, in spite of actinobacterial high soil population, secondary metabolite production and capability to endure hostile environments, Streptomyces, and other actinobacteria are unexpectedly under explored for plant-growth promotion, as compared to Pseudomonas or Bacillus spp. (Doumbou et al., 2002).

The main objectives of the present study included: to isolate indigenous actinobacteria from the wheat (Triticum aestivum) and tomato (Solanum lycopersicum) rhizosphere, characterize these isolates on the basis of morphological, and physiological characteristics as well as by 16S rRNA gene sequence analysis, to screen actinobacteria for various plant growth promoting activities (PGPAs), such as IAA production, phosphate solubilization, siderophore production, and in-vitro 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity. The PGP potential of selected isolates was also studied in-vivo under axenic conditions and their effect on wheat growth was investigated. Keeping in mind, there is a lack of data on the Plant growth promotion (PGP) potential of actinomycetes isolated from Pakistan, Therefore, this study provides new and novel information.

### MATERIALS AND METHODS

### Collection and Enrichment of Soil Samples

The soil samples for the isolation of actinobacteria were collected from the rhizosphere of wheat (Triticum aestivum) and tomato (Solanum lycopersicum) plants cultivated in the Punjab province (District: Lahore, Gujranwala, Sheikhupura) Pakistan, during the months of December, 2013–March, 2014. Samples were collected in properly labeled sterile polythene sampling bags. The samples were subjected to physical treatment (heating, 55◦C/60 min) according to the method of Seong et al. (2001) and chemical treatment; CaCO3: soil (1:10 w/w; Oskay, 2009), for the enrichment of actinobateria.

### Isolation and Preservation of Actinobacteria

One Gram of soil sample was mixed in 9 ml of autoclaved water and serially diluted to a final dilution of 10−<sup>3</sup> , 10−<sup>4</sup> , and 10−<sup>5</sup> . The 0.1 ml of each dilution was spreaded on ISP-4 medium (Soluble starch 10 g, CaCO<sup>3</sup> 2 g, (NH4)2SO<sup>4</sup> 2 g, K2HPO<sup>4</sup> 1.0 g, MgSO4.7H2O 1.0 g, NaCl 1.0 g, FeSO4.7H2O 1.0 mg, MnCl2.7H2O 1.0 mg, ZnSO4.7H2O 1.0 mg, Agar 15 g, distilled water 1 L, pH = 7.5 ± 0.2, 28◦C) supplemented with 25µg/ml nalidixic acid and 50 g/ml nystatin as antibacterial and antifungal agents (Taechowisan et al., 2003). Isolated actinobacteria were sub-cultured on ISP-2 medium (Yeast extract 4 g, Glucose 4 g, Malt extract 10 g, Agar15 g, distilled water 1 L, pH = 7.5) and the plates were incubated at 28◦C for 7–12 days. All the isolated strains were preserved in 25% glycerol and kept at −80◦C. After incubation, 98 colonies of actinobacteria were selected on the basis of morphological characters such as distinct color and colony shape. Among them, 30 strains were detected to be exhibiting Plant growth promoting activities in initial in-vitro screening. Finally, six isolates that included WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 which exhibited most significant PGP traits were selected.

### Taxonomic Studies

### Biochemical and Physiological Characterization

A comprehensive morphological, biochemical, and physiological characterization scheme was adopted to determine the taxonomic status of the selected actinobacterial strains. Among the different biochemical characteristics studied, melanin pigment production was determined by following the methods described by Pridham and Lyons (1969) and positive results were indicated by blackening of the media. Carbohydrate utilization was performed by using ISP-9 medium supplemented with different sugars. Decomposition of oxalic acid and other organic acids was performed by using the method of Nitsch and Kutzner (1969). Esculin degradation was detected by blackening of the media after 12 days of incubation was recorded as positive result. Trysoine agar was used for determining tyrosine hydrolysis by actinomycetes (Gordon and Smith, 1953). Xanthine and hypoxanthine utilization is determined by clear zone formation around the colony after 12 days of incubation (Gordon and Smith, 1953). Starch hydrolysis was performed as described by Cowan (1974). After 12 days of incubation, plates were flooded with lugol's iodine and clear zone around colonies were recorded as positive. Urease releases ammonia from urea and the increase in the pH was detected by the indicator phenol red changing from yellow to pink (Gordon et al., 1974), Cell wall type was determined based on the isomers of diaminopimelic acid (DAP) by using the method of Becker et al. (1964). Cultural characteristics such as color of aerial and substrate mycelium and pigmentation of the selected actinobacteria was recorded on ISP-2 medium according to the method of Shirling and Gottlieb (1966). Growth at different pH 5, 6, 7, 8 was noted after 14 days of incubation on the agar plates. Tolerance to temperature was tested at 10, 28, 37, 45◦C and visible growth was recorded as positive result. Agar supplemented with different NaCl Concentration 0, 4, 7, 10, 13% was used for determining NaCl tolerance of actinomycetes.

### In-vitro Screening of Actinobacteria for Their Plant Growth Promoting Activities Colorimetric Analysis of Indole Acetic Acid (IAA) Production

Auxin form different strains of actinomycetes was quantified using the method of Tang and Borner (1979) as described by Ali et al. (2009). All Actinobacteria were grown at 28◦C in ISP-2 liquid medium in triplicates for 7–12 days at 120 rpm on an orbital shaker. Media treatments were supplemented with different concentrations of L-tryptophan (0, 100, 200, 300, 400, and 500µg/ml). Cells were removed from culture medium by centrifugation at 14,000 rpm for 15 min (Sigma 2–5, sigma Laborzentrifugen, Osterode, Germany). The supernatant (1 ml) was mixed with 2 ml of Salkowski's reagent (50 ml, 35% perchloric acid, 1 ml of 0.5 M FeCl<sup>3</sup> solution) and was incubated at room temperature for 30 min in dark. Development of pink or red color indicates IAA production. Optical density was taken at 535 nm by using spectrophotometer (S-300D; R and M Marketing, Hounslow, UK). Standard curve of IAA was used to measure the concentration of IAA produced by the actinobacteria.

### Siderophore Production

The strains were assayed for the siderophore production on the Chrome Azurol S agar according to the method of Alexander and Zuberer (1991). CAS agar plates were prepared and spot inoculated with actinobacterial strains and incubated at 28◦C for 7 days. The colonies producing yellow to orange halos were considered positive for siderophore production.

### Solubilization of Phosphate

Pikovskaya's agar plates were used for qualitative screening of all actinobacterial isolates (Gaur, 1990). Solubilization index (SI) was calculated by using the formula of Edi Premono et al. (1996). King (1932) method was used for quantitative analysis of solubilization of tri-calcium phosphate in liquid medium.

### Ammonia Production

Freshly grown actinobacterial cultures were inoculated into 1 ml of peptone water and incubated at 28◦C for 7–12 days with shaking at 120 rpm. After incubation, 0.5 ml of Nessler's reagent was added in each culture tube. Development of yellow to brown color indicates positive result for ammonia production (Cappuccino and Sherman, 2002).

### Hydrogen Cyanide Production

Producton of hydrogen cyanide by actinobacterial culture was evaluated by adapting the method of Lorck (1948). Actionobacteria were streaked on ISP-2 medium amended with 4.4 glycine/l and whatman filter paper No. 1 dipped in 2% sodium carbonate in 0.5% picric acid for a minute was placed underneath the petri plates lids. Plates were sealed with parafilm and incubated at 28◦C for 7–12 days. Orange to red color of filter was indicative of HCN production.

### Colorimetric Ninhydrin Assay for Screening ACC Utilizing Actinobacteria

Li et al. (2011) method was used. ISP-2 medium was inoculated with actinobacteria, incubated at 28◦C with shaking 200 rpm for 7–12 days. After centrifugation of 2 ml of culture at 14,000 rpm for 5 min, supernatant was discarded, pellet washed with 1 ml of liquid DF- medium, later mixed with 2 ml of ACC substrate containing DF mineral medium. 2 ml of DF-ACC medium without inoculation served as control. All the samples were incubated for 4 days at 200 rpm. After incubation, 1 ml of each bacterial culture was centrifuged at 14,000 rpm for 5 min. 100µl of each supernatant was shifted to another tube and was diluted with 1 ml of liquid DF medium. In 96-well PCR plate, 60µl of each diluted supernatant was mixed 120µl of ninhydrin reagent, covered with parafilm, and placed in boiling water bath for 30 min. DF medium was used as a blank.

The resulted purple color depth was record for visual comparison of actinobacterial strains. 100µl of the reaction mixture was transfer to the microtitre plate in triplicates and absorbance was taken at 570 nm with spectrophotometer. The actinobacterial isolate with visibly reduced color depth and less supernatant absorbance compared to DF-ACC medium without inoculation were considered as ACC utilizing actinobacterial strains.

### In-vivo Screening for Plant Growth Promotion Activities

Actinobacterial isolates were grown on ISP-2 broth at 28◦C for 7–12 days with continuous shaking 200 rpm. Actinobacterial suspensions were prepared according to the method of Errakhi et al. (2007). Wheat seeds were surface sterilized by using the method of Khalid et al. (2004). Sterilized seeds were soaked for 1 h in the suspension and dried under laminar flow hood overnight. For control, seeds were dipped in distilled water only.

### Germination Bioassay

Germination assay was performed by using actinobacterial treated wheat (Triticum aestivum) seeds (three replicates, 5 seeds/plate) were placed in sterilized petri dishes covered with two sheets of filter papers, moistened with 10 ml of sterile distilled water. Water treated control seeds were used. All the petri dishes were incubated in Versatile Environmental Test Chamber (TEMI 850 WISE CUBE) with light intensity of 2000 lux for 16 h daily at 28◦C. After 2 weeks, effect of actinobacterial cultures on root length and number of roots was observed.

### Pot Experiment

Actinobacterial treated eight wheat seeds (as described above) were sown to a depth of 1 cm in plastic pots (12 cm high × 10 cm diameter) filled with sterilized soil. Three replicates were used for actino-bacterial as well as for control treatments. For control, seeds were dipped in distilled water only. Pots were arranged in fully randomized complete block, under standard conditions, in plant growth room (22–26◦C, 16 h light/8 h dark). All pots were watered daily with 10 ml of distilled autoclaved water to achieve moisture level sufficient for seed germination. After 25 days, five plants were removed carefully from the soil, washed with tap water to remove soil particles. Total 15 plants per treatment were considered for statistical analysis. Data was recorded for number of seeds germinated, root and shoot length, root and shoot fresh weight, root, and shoot dry weight (80◦C, 24 h electric oven) and number of roots.

Root colonization potential of inoculated Streptomyces was determined at every 8 days by using serial dilution plating technique on GYM agar and number of viable cells was recorded as colony forming units (CFU) as described in Somasegaran and Hoben (1994).

### Genomic DNA Isolation, PCR Amplification and Sequencing Of The 16s rRNA Gene

Total genomic DNA was isolated according to the CTAB method described by the Liu et al. (2000). Universal Primers 27F 5′ AGAGTTTGATCMTGGCTCAG 3 and 1492 R′ TACGGYTACCTTGTTACGACTT 3′ were used for the PCR amplification of 16S rRNA gene of the selected strains. Agarose gel electrophoresis (1%) was used for analyzing PCR product and remaining mixture was purified by using PCR Purification kit (FavorPrepTM). Purified PCR products were sequenced commercially by GATC Biotech (Germany). The obtained gene sequences were compared with others in the Gen Bank databases using the NCBI Nucleotide BLAST at http://blast.ncbi.nlm.nih. gov/Blast.cgi. Sequences were submitted to NCBI GenBank data base and accession numbers were obtained.

### Statistical Analysis

For all experiments, the data were subjected to statistical analysis using software IBM SPSS Statistics version 21. Data were subjected to analysis of variance (ANOVA) and means separated using Duncan's multiple range test (P = 0.05). The correlation coefficients between bacterial auxin production and L-tryptophan concentrations as well as between bacterial growth traits and plant growth parameters were also calculated (P = 0.01 or P = 0.05).

## RESULTS

### Taxonomic Characteristics of the Selected Rhizopsheric Actinomycetes

Ninty eight actinomycetes were isolated from six different wheat and four different tomato rhizospheric soil samples. All of them were found to be gram positive filamentous rods. On the basis of color of aerial mycelium they were grouped into gray (TA-3, TB-4), yellow (WC-3, WD-3), green (WA-1), and orange (WA-3) color series. Diffusible pigment was produced by the isolate WC-3. All strains contained LL-diaminopimelic acid isomer in their cell wall. Physiologically, most of the actinomycetes isolates were able to utilize different sugars as the carbon source. All strains were able to utilize glucose, xylose, galactose, arabinose, mannose, and mannitol. Strain WA-1 and TA-3 were unable to utilize sucrose, while strain WC-3, and WD-3 were unable to utilize raffinose as their carbon source (**Table 1**). Most of them were able to grow at temperature 28 and 37◦C. None of them were able to grow at temperature 4◦C. Optimum temperature and pH was found to be 28◦C and eight respectively (**Table 2**; Tables S1, S2, Figures S1, S2).

### IAA Production by Selected Actinomycetes

Qualitative analysis of culture supernatant of selected actinomycete isolates revealed production of variable amounts of IAA in the absence and presence of different concentrations of tryptophan (100–500µg/ml). In the absence of L-tryptophan, IAA production was not observed. With increasing concentration of L-tryptophan, the IAA production was increased. For instance, isolate Streptomyces sp. WA-1(r = 0.979, P = 0.01), S. nobilis WA-3 (r = 0.942, P = 0.01),


The symbol, + representing the positive reaction/presence of growth while symbol, − represents the negative reaction/absence of growth.


TABLE 2 | Temperature (◦C), pH and NaCl (%) tolerance of the selected rhizospheric actinomycete strains.

The symbol, + represents the positive reaction/presence of growth while symbol, − represents the negative reaction/absence of growth.

S. kunmingensis WC-3 (r = 0.995, P = 0.01), S. mutabilis WD-3 (r = 0.932, P = 0.01), S. enissocaesilis TA-3 (r = 0.971, P = 0.01), and S. djakartensis TB-4 (r = 0.919, P = 0.05) showed significant positive correlation with increasing L-tryptophan concentrations. Isolate WA-3 produced highest amount of IAA followed by WC-3, TA-3, WD-3, WA-1, and TB-4. The most active IAA producer S. nobilis WA-3, S. Kunmingenesis WC-3 and S. enissocaesilis TA-3 produce 79.5, 79.23, and 69.26µg/ml IAA respectively at 500µg/ml L-tryptophan (**Figure 1**).

### Phosphate Solubilization

Among the selected isolates, six Streptomyces were able to solubilize phosphate by producing clear zones around the colonies after 7 days of incubation. Streptomyces sp. WA-1 showed highest phosphate solubilization index of 2.33 followed by S. djakartensis TB-4 (2.27) > S. enissocaesilis TA-3 (2.25) > S. nobilis WA-3 (2.22) > S. mutabilis WD-3 (2.21) > S. kunmingensis WC-3 (2.20). Quantitatively, the highest concentration of soluble phosphate was produced by Streptomyces sp. WA-1 (72.13 mg/100 ml) and S. djakartensis TB-4 (70.36 mg/100 ml; **Table 3**).

### Siderophore, Ammonia, and HCN Production

Among the selected isolates, six rhizospheric actinomycetes were positive for the ammonia production. Siderophore production was detected in all six isolates on CAS agar media, forming clear orange halo zone around the colonies. (**Figure 2**). All the six isolates were positive for HCN production (**Table 3**). Streptomyces sp. WA-1 and S. djakartensis TB-4 displayed the highest amount of HCN production as depicted by a very deep red color on the filter paper (**Table 3**).

### ACC Deaminase Activity by Actinomycetes

All strains showed variable concentration of ACC-deaminase in the culture media. Strain S. mutabilis WD-3 showed highest concentration of ACC-deaminase 1.9 mmol /l). On the other hand, strain S. djakartensis TB-4, Streptomyces sp. WA-1, S. kunmingensis WC-3, S. nobilis WA-3 produce 1.81, 1.65, 1.59, 1.29 mmol/l concentration of ACC-deaminase, respectively. Strain S. djarkartensis TA-3 produce the lowest concentration (0.71) of ACC deaminase (**Table 3**).

### Seed Germination Assay

Inoculation of wheat seeds with selected actinomycetes stimulated root growth in majority of the strains (**Figure 3**). In case of root length, significant increases were recorded for

S. djakartensis TB-4 (61%), S. Kunmingenesis WC-3 (47%) and S. mutabilis (42%). Increases in number of roots were also observed with S. mutabilis WD-3 (55%), Streptomyces sp. WA-1 (50%), and S. djakartensis TB-4 (41%).

### Root Colonization Potential

Actinobacterial population size was estimated by plate count method on GYM agar at three (8, 16, 24 days) different time intervals. It was observed that all rhizospheric actinobacteria were able to colonize wheat plant roots and displayed persistence in the rhizosphere up to 25 days after inoculation (**Figure 4**). Maximum colonization was recorded between 8 and 16 days post inoculation. Actinobacterial isolate WA-3 showed maximum number of root colonizing colonies at all times as compared to other actinobacterial isolates.

FIGURE 2 | Screening of siderophore producing actinomycete isolates using CAS agar plates after 7 days of growth at 28 ± 2 ◦C. The arrows indicate the halo zone around the colonies of isolate WA-1 (Streptomyces sp.) and TB-4 (Streptomyces djakartensis).



<sup>a</sup>SI, Solubilization index = the ratio of the total diameter (colony + halo zone) to the colony diameter. <sup>b</sup>ACC, 1-aminocyclopropane-1-carboxylate (ACC) deaminase. Different letters on bars indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05). <sup>c</sup>HCN, Hydrogen cyanide. The symbol, + represents the positive reaction/presence of trait while symbol, − represents the negative reaction/absence of trait.

### Plant Growth Promotion Experiment with Selected Actinomycetes

Wheat seeds inoculated with six selected actinomycetes showed variable growth and yield parameters after 3 weeks growth (**Table 4**). Significant increases in shoot length were observed

FIGURE 3 | Effect of actinomycetes spore suspension on seed germination of Triticum aestivum. Bars represents mean ± SE of three replicates (15 plants). Different letters on bars indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05).

with S. nobilis WA-3 (65%), S. djakartensis TB-4 (54%), and S. enissocaesilis TA-3 (53%) as compared to water treated control. Significant increase in root length was recorded with S. nobilis WA-3 (81%), S. djakartensis TB-4 (69%), and S. enissocaesilis TA-3 (51%) as compared to water treated control. Maximum increases in plant fresh weight were recorded with S.nobilis WA-3 (84%) and S. djakartensis TB-4 (75%), increased plant dry weight was recorded in case of S. nobilis WA-3(85%), S. djakartensis TB-4 (66%), and S enissocaesilis TA-3 (60%). In case of number of leaves, significant increases were recorded with S. nobilis WA-3 (27%), S. djakartensis TB-4 (27%), and S. enissocaesilis TA-3 (23%). Significant increase in case of number of roots were recorded in case of S. nobilis WA-3 (30%), S. djakartensis TB-4 (26%), and S. enissocaesilis TA-3 (22%) as compared to water treated control (**Figure 5**).

### Identification of Selected Actinomycetes Strains by 16s rRNA Gene Sequencing

Six of them including WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 were selected based on their ability to produce phytohormone IAA, solubilization of inorganic phosphates, ACC deaminase activity, siderophore, ammonia, and hydrogen cyanide production. For all the selected actinomycetes, single band PCR product of 1.5 kb length were achieved with the universal primers (**Figure 6**). The sequences of 16S rRNA gene were analyzed by comparison with sequences in GenBank through Nucleotide BLAST (http://www.ncbi. nlm.nih.gov/BLAST). After comparison, strains WA-1 showed 98% similarity with Streptomyces sp. while the strains WA-3, WC-3, WD-3, TA-3, and TB-4 showed 99% similarity with Streptomyces nobilis, Streptomyces kunmingenesis, Streptomyces mutabilis, Streptomyces enissocaesilis, Streptomyces djakartensis, respectively. The sequences from strains WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 have been deposited in the GenBank and accession numbers were obtained (**Table 5**).

### DISCUSSION

In this study, a total of 98 actinomycetes strains were isolated from wheat and tomato rhizospheric soil. All of the isolates were


Control , non-inoculated. Data is the mean of 15 plants obtained from three replicates. ± Symbol indicate values for standard error of means. Different letters indicate significant difference between treatments, using Duncan's multiple range test (P = 0.05).

screened for different PGP traits and 30 of them were found to be showing excellent PGP activities in initial screening. Out of these, six most promising actinobacterial strains belonging to the genus Streptomyces were investigated by in-vivo studies. The 16S rRNA gene sequence of these six selected actinomycetes strains including WA-1, WA-3, WC-3, WD-3, TA-3, and TB-4 showed maximum sequence similarity with members of the genus Streptomyces.

Majority of the PGPR actinomycetes synthesize IAA which is responsible for increased number of adventitious roots which help plant to uptake a large volume of nutrients and absorb water, while increased root exudates in turn benefits the bacteria (El-Tarabily, 2008). In our study, the IAA

FIGURE 6 | Gel electrophoresis of PCR products for detection of Actinomycetes isolates. Lane 1: DNA ladder (Fermentas, Germany); lane 2, 3, 4, 5, 6, and 7: positive samples of Streptomyces sp. (WA-1), S. nobilis (WA-3), S. kunmingenesis (WC-3), S. mutabilis (WD-3), Streptomyces enissocaesilis (TA-3), and S. djakartensis (TB-4) respectively (1.5 kb).

production ranges between 10 and 79.5µg/ml and Streptomyces nobilis strain WA-3 was detected as the best strain for the IAA production (79.5 ug/ml) that exceeds the level of previously reported work by Khamna et al. (2009). Abd-Alla et al. (2013) reported Streptomyces sp. CMU-MH021 that could produce 28.5µg/ml of IAA. Several Streptomyces species, such as S. olivaceoviridis, S. rimosus, S. Rochei, S. griseoviridis, and S. lydicus have the ability to produce IAA and improve plant growth by increasing seed germination, root elongation and root dry weight (Mahadevan and Crawford, 1997).

Soil acidification often resulted due to the growth of phosphate-solubilizing bacteria (PSB), which in turn resulted in phosphorus solubilization. Therefore, PSB are well known as solubilizers of inorganic phosphate (Verma et al., 2001). The maximum phosphate solubilization activity was detected in the strain Streptomyces sp. WA-1 (72.13 mg/100 ml). Hamdali et al. (2008) reported 83.3, 58.9, and 39 mg/100 ml phosphate solubilization by Streptomyces cavourensis, Streptomyces griseus, and Micromonospora aurantiaca, respectively.

Almost all of the rhizospheric actinomycetes were also able to produce ammonia and hydrogen cyanide. Marques et al. (2010) recommended that bacteria can synthesize ammonia and supply nitrogen to the host plant. Additionally, over production of ammonia serve as a prompting factor for the virulence of opportunistic plant pathogens. Ammonia production also plays an important role HCN production play an essential role in suppression of plant disease. In this study, all the ammonia and HCN producing isolates belong to the genus Streptmyces. Similarly, Husen et al. (2011) reported Streptomyces sp. LSW05 strain as a potent HCN producer.



Siderophore production is one more feature that stimulates plant growth by forming complex with iron form (Fe 3+) in the rhizosphere making iron unavailable to the phytopathogens. It is suggested by Tan et al. (2009) siderophore production that the production of siderophore is an important factor for phytopathogen antagonism and developing growth of the plant. In our study, we detected the 85.7% isolates positive for siderophore production. Similarly, Khamna et al. (2010) has revealed that Streptomyces CMU-SK 126 isolated from Curcuma mangga rhizospheric soil exhibited high amount of siderophore.

ACC deaminase-producing bacteria have been known to promote plant growth by decreasing ethylene inhibition of various plant processes (Husen et al., 2011). They can increase root growth by lowering endogenous ACC levels (Glick, 2005). Plant roots must be able to perceive and recognize such elicitors in ways similar to the recognition of elicitors from plant pathogens. In fact, plant pathogens might interfere with the action of PGPR by being perceived by similar receptors (Husen, 2003).

The effect of soil microbes on PGP including root development has been reported by Uphoff et al. (2009). In the present study, root elongation assay and pot experiment performed by using wheat seeds inoculated with PGP Streptomyces strains, significantly enhanced the plant growth by increasing plant root length, plant shoot length, dry weight, fresh weight, number of leaves, and number of roots over the un-inoculated control. The Streptomyces strains are extensively reported in the literature for its PGP potential (Nassar et al., 2003; El-Tarabily, 2008; Gopalakrishnan et al., 2011b). As hypothesized earlier, the mechanism by which the Streptomyces enhanced the morphological and yield parameters on both sorghum and rice could be their PGP traits including IAA and siderophore production (direct stimulation of PGP; Gopalakrishnan et al., 2011b).

### REFERENCES


### CONCLUSION

The study revealed that these rhizospheric actinomycetes are potential microbial inoculants because of their intensified PGP activities such as IAA production, phosphate solubilization, siderophore, and HCN production and ACC deaminase production. The strains reported in this study are promising candidates to be developed as commercial biofertilizer formulation and can also be exploited for the production of various agroactive compounds like auxins etc. As such this is the first comprehensive report from Pakistan about the PGP traits and potential agricultural applications of actinomycetes.

### AUTHOR CONTRIBUTIONS

SA: Isolation, biochemical characterization, identification and invitro and in-vivo screening of actinomycetes for multiple PGP traits/hormones. Did all the experimental work. IS: Supervise in the sampling of the rhizospheric soils. Teach how to isolate actinomycetes by using enrichment, help in DNA isolation and PCR purification technique. BA: Co-supervise in performing PGP experiments. Both in-vitro and in-vivo screening of actinomycetes.

### ACKNOWLEDGMENTS

The financial support for this study by Higher Education Commission (HEC) of Pakistan, under indigenous Ph.D. fellowships program is gratefully acknowledged.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01334

mineral fertilizers to improve growth and yield of wheat (Triticum Aestivum L.). Pak. J. Bot. 41, 381–390.


effect on Vigna radiata (L.). World J. Microbiol. Biotechnol. 25, 519–526. doi: 10.1007/s11274-008-9918-9


compounds, indole-3-acetic acid and siderophore production. World J. Microbiol. Biotechnol. 25, 649–655. doi: 10.1007/s11274-008-9933-x


King, J. E. (1932). The colorimetric determination of phosphorus. Biochem. J. 26, 292–297.


strain Actinoalloteichus nanshanensis sp. nov. NEAU 119. Nat. Prod. Res. 27, 1863–1869. doi: 10.1080/14786419.2013.771349

Zhang, J., Liu, J., Meng, L., Ma, Z., Tang, X., and Cao, Y. (2012). Isolation and characterization of plant growth-promoting rhizobacteria from wheat roots by wheat germ agglutinin labeled with fluorescein isothiocyanate. J. Microbiol. 50, 191–198. doi: 10.1007/s12275-012- 1472-3

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Anwar, Ali and Sajid. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Biocontrol Potential of Streptomyces hydrogenans Strain DH16 toward Alternaria brassicicola to Control Damping Off and Black Leaf Spot of Raphanus sativus

Rajesh K. Manhas\* and Talwinder Kaur

Department of Microbiology, Guru Nanak Dev University, Amritsar, India

Biocontrol agents and their bioactive metabolites provide one of the best alternatives to decrease the use of chemical pesticides. In light of this, the present investigation reports the biocontrol potential of Streptomyces hydrogenans DH16 and its metabolites towards Alternaria brassicicola, causal agent of black leaf spot and damping off of seedlings of crucifers. In vitro antibiosis of strain against pathogen revealed complete suppression of mycelial growth of pathogen, grown in potato dextrose broth supplemented with culture supernatant (20% v/v) of S. hydrogenans DH16. Microscopic examination of the fungal growth showed severe morphological abnormalities in the mycelium caused by antifungal metabolites. In vivo studies showed the efficacy of streptomycete cells and culture supernatant as seed dressings to control damping off of Raphanus sativus seedlings. Treatment of pathogen infested seeds with culture supernatant (10%) and streptomycete cells significantly improved seed germination (75–80%) and vigor index (1167–1538). Furthermore, potential of cells and culture supernatant as foliar treatment to control black leaf spot was also evaluated. Clearly visible symptoms of disease were observed in the control plants with 66.81% disease incidence and retarded growth of root system. However, disease incidence reduced to 6.78 and 1.47% in plants treated with antagonist and its metabolites, respectively. Additionally, treatment of seeds and plants with streptomycete stimulated various growth traits of plants over uninoculated control plants in the absence of pathogen challenge. These results indicate that S. hydrogenans and its culture metabolites can be developed as biofungicides as seed dressings to control seed borne pathogens, and as sprays to control black leaf spot of crucifers.

Keywords: Streptomyces hydrogenans DH16, Alternaria brassicicola, Raphanus sativus, biocontrol, culture supernatant

## INTRODUCTION

Alternaria brassicicola is one of the economically important pathogens worldwide with broad host range, and is established and widespread in many countries, including India (Verma and Saharan, 1994; Reis and Boiteux, 2010). It causes black spot disease and damping off of seedlings in Brassica spp. throughout the world and causes huge economic losses (Rimmer and Buchwaldt, 1995). Host

### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Abdullah M. Al-Sadi, Sultan Qaboos University, Oman Lianghui Ji, Temasek Life Sciences Laboratory, Singapore

> \*Correspondence: Rajesh K. Manhas rkmanhas@rediffmail.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 03 July 2016 Accepted: 25 November 2016 Published: 16 December 2016

### Citation:

Manhas RK and Kaur T (2016) Biocontrol Potential of Streptomyces hydrogenans Strain DH16 toward Alternaria brassicicola to Control Damping Off and Black Leaf Spot of Raphanus sativus. Front. Plant Sci. 7:1869. doi: 10.3389/fpls.2016.01869

plants can be affected at all developmental stages. Typical disease signs consist of black lesions on seedlings, leaves, stems, and siliquae resulting in accelerated senescence, premature pod shatter, and shrunken seeds (Mac Kinon et al., 1999; Iacomi-Vasilescu et al., 2004). Seeds infected with the pathogen result in reduced seed germination, photosynthetic potential, seedling vigor and pre- and post-emergence damping-off of seedlings (Nowicki et al., 2012), causing significant reduction in yield quantity and quality.

As a disease management strategy, the most feasible and economical option is the development of resistant brassicaceae crops. Unfortunately, transgenic approach used to develop resistant cultivars has proven to be a failure because of lack of expression of resistance in various varieties of crucifers (Sigareva and Earle, 1997). In many countries protection of crucifer crops from A. brassicicola is achieved using several different families of fungicides including dicarboximides, carbamates, benzimidazoles, and triazoles as seed or foliar treatments. However, in modern agriculture use of fungicides has become unpopular because of several problems related to environmental pollution, toxicity to humans, emergence of resistant strains and detrimental effects on non-target populations (Fox et al., 2007). So, there is a demand for new methods to supplement the existing disease management strategies to accomplish better disease control.

Biocontrol, as a part of integrated pest management, is a feasible replacement to the use of synthetic chemicals for more sustainable agriculture. The adaptiveness of most biocontrol agents to the environment in which they are used, the complexity of the organismal interactions and the involvement of numerous mechanisms of disease suppression by a single microorganism contribute to the belief that biocontrol will be more durable than synthetic chemicals (Sharma and Sharma, 2008).

Among microorganisms, actinobacteria, especially Streptomyces spp. are of utmost value as they are potent producers of bioactive compounds with different biological properties (Prabavathy et al., 2006) and have been established as potential biocontrol agents (Gomes et al., 2000; Ouhdouch et al., 2001). As reviewed by Berdy (2005) nearly 70–75% of secondary bioactive metabolites are isolated from these filamentous bacteria and 60% of the antibiotics that have been developed for agricultural industry are obtained from Streptomyces spp. (Tanaka and Omura, 1993).

Keeping in mind the importance of Streptomyces spp. as potent biocontrol agents, the present study was focused to evaluate in vitro and in vivo potential of Streptomyces hydrogenans strain DH16 (a strong antagonist toward various fungal phytopathogens, Kaur and Manhas, 2014) to control black leaf spot disease and damping off of Raphanus sativus, a major root vegetable crop worldwide with many health benefits.

### MATERIALS AND METHODS

### Microorganisms

Streptomyces hydrogenans DH16 (GenBank: JX123130), a soil isolate having broad spectrum antifungal activity against fungal phytopathogens, was grown on starch casein nitrate agar (SCNA) slants and maintained at refrigeration temperature (4◦C) for laboratory work. Mycelial fragments and spores of streptomycete were preserved in 20% v/v glycerol at −20◦C as stock for future work. A. brassicicola MTCC 2102 was preserved on potato dextrose agar (PDA) slopes at 4◦C.

### Production of Bioactive Metabolites by S. hydrogenans Strain DH16

For production of antifungal metabolites Streptomyces DH16 was grown on SCNA at 28◦C. After 7 days of incubation, growth was scrapped and transferred into the starch casein nitrate broth to develop seed culture. After incubation (48 h), the production medium (starch: 12 g/l; soybean meal: 2.5 g/l; K2HPO4: 1.8 g/l; casein: 0.3 g/l; MgSO4: 0.1 g/l; FeSO4: 0.01 g/l; NaCl: 2.0 g/l; CaCO3: 0.02 g/l) was inoculated with seed culture and fermentation was carried out by incubating at 28◦C at 180 rpm. After 3 days of incubation, filter sterile cell free supernatant was used for further experiments.

### In vitro Antagonistic Activity

In vitro antagonistic activity of 3 days old cell free supernatant was checked against A. brassicicola by well diffusion method (Kaur and Manhas, 2014).

### Suppression of Mycelial Growth of A. brassicicola in Broth

The effect of cell free supernatant of S. hydrogenans strain DH16 on mycelial growth of A. brassicicola was further determined in liquid culture according to Li et al. (2011). The 3 days old culture supernatant at concentrations of 0, 0.5, 1, 5, 7.5, 10, and 20% was supplemented into potato dextrose broth (PDB, 50 ml). Each flask was inoculated with single mycelial disk of 6 mm diameter of test fungus and uninoculated SCN broth served as control. After 7 days of incubation at 28◦C, mycelial dry weight was determined, and growth inhibition was calculated as follow:

```
[(Weight of untreated mycelium − Weight of treated mycelium)]
 Weight of untreated mycelium × 100
```
The experiment was repeated twice with three replicas.

### Effect of Culture Supernatant on Fungal Morphology of A. brassicicola

The effect of cell free culture supernatant of streptomycete on morphology of A. brassicicola was studied microscopically. Mycelium of A. brassicicola was taken from periphery of the inhibition zone around the well (containing culture supernatant of streptomycete) and from control plate and placed on glass slide in a drop of sterile water. The coverslip was placed on the film and then visualized under bright field microscope at 40× (Olympus). Microphotographs were taken using a digital camera.

### Effect of Streptomyces DH16 Culture Supernatant on Spore Germination

To study the effect of culture filtrate on spore germination of A. brassicicola, 100 µl of fungal spore suspension made in PDB (10<sup>5</sup> spores ml−<sup>1</sup> ) were mixed with 100 µl culture filtrate of different concentrations (0, 0.5, 1, 5, 7.5, 10, and 20% v/v; prepared using uninoculated production medium) and incubated at 28◦C. In control, culture filtrate was replaced with 100 µl of SCN broth. After incubation of 8 h, 50 µl of each suspension were placed on sterile glass slide. After placing coverslip, slide was observed under microscope by counting about 50 spores.

### Extracellular Conductivity

The effect of antifungal metabolites on cellular leakage was studied by determining the extracellular conductivity of supernatants obtained from mycelial suspensions (A. brassicicola) treated with culture supernatant of DH16 (Lee et al., 1998). To obtain mycelial growth of fungus. Erlenmeyer flask (250 ml) containing 50 ml of PDB was inoculated with single mycelial disk of 6 mm diameter from 5-day-old PDA plate of test fungus. After 3 days of incubation at 28◦C in PDB, mycelial growth was collected and washed thoroughly with sterile double distilled water. Washed mycelium (3 mg) was then added to flask containing 20 ml of culture supernatant of DH16. Mycelial suspensions were centrifuged at 10,000 rpm for 10 min to obtain supernatants, first immediately after the addition of mycelium, second after 12 h and third after 24 h of treatment. The experiment was repeated three times. Electrical conductivity was measured using a conductivity meter.

### Direct Observations of Antagonistic Effects of Streptomyces DH16 on Spore Germination and Growth of A. brassicicola in Soil Environment

A buried slide technique was used to determine the effect of streptomycete on fungal pathogen, directly in soil environment (Stevenson, 1956). 50 g of soil was sieved, air dried and sterilized in 100 ml glass beaker by autoclaving for 30 min at 121◦C. A 7 ml spore suspension (1 × 10<sup>8</sup> cells) of streptomycete was prepared. Spores were then thoroughly washed with sterile distilled water and inoculated into the sterile soil and incubation of soil was done at 28◦C for 7 days. A 10 ml spore suspension of 5 days old A. brassicicola (1 × 10<sup>5</sup> spores/ml) was mixed with 100 ml of sterile molten agar (1.8%), and 1 ml of it was coated on sterile glass slides. After solidification of the agar layer, slides were carefully inserted vertically into the beakers containing sterile soil alone (control) or DH16 inoculated soil and incubated at 28◦C. At the end of every stated period, i.e., 2, 3, 4, and 6 days, slides were removed and examined immediately under a compound microscope at magnification of 400×. The experiment was conducted twice.

## Efficacy of Culture Supernatant and Cells of Streptomyces DH16 As Seed Treatment against A. brassicicola to Control Damping Off

### Seed Treatment

The biocontrol potential of cell free culture filtrate and cells of S. hydrogenans strain DH16 as seed treatment against A. brassicicola using radish seeds was determined. Seeds were surface sterilized by immersing in sodium hypochlorite (1%) for 10 min. and then washed repeatedly with sterilized distilled water. The sterilized seeds were first artificially infected with the pathogen prior to antagonist treatment. The seeds were immersed for 4 h in fungal spore suspension in presence of 1% carboxymethyl cellulose (CMC; 105–10<sup>7</sup> spores/ml). These pathogen infested seeds were further given second treatment viz. (i) soaked in different concentrations (5, 10, and 20% v/v) of culture supernatant of antagonist/(ii) soaked in cell suspension of antagonist prepared in 1% CMC (107–10<sup>8</sup> /ml).

In another treatment, uninoculated sterilized seeds were treated with (i) 1% CMC only (control), (ii) cells of antagonist only (107–10<sup>8</sup> ), and (iii) cell free culture supernatant only. After 1 h of second treatment, all seeds were dried in laminar flow on a sterile filter paper and used for further experiments.

### Blotter Test

The moistened blotters were first used to determine the effect of antagonist to reduce damping off on radish plants grown from artificially infected seeds. Thirty seeds per treatment were placed in Petri dishes (10 seeds per plate) already lined with moist filter paper and covered loosely with another filter paper. Number of germinated seeds, and healthy and diseased seedlings were recorded after incubation of 7 days at 28◦C in the dark. Seedling vigor (V) was determined by measuring root and shoot lengths and was calculated according to the equation:

$$\mathbf{V} = (\mathbf{L\_s} + \mathbf{L}\mathbf{r}) \times \mathbf{G}$$

Where L<sup>s</sup> is average shoot length in mm and Lr is average root length in mm and G is % germination (Andresen et al., 2015). The experiment was repeated twice.

### In vivo Pot Experiment

Same treatments were given to seeds as described above, except in case of culture supernatant. For pot experiments, culture supernatant at a concentration of 10% was used as it showed significantly better results in blotter test. Seeds were sown in pots containing autoclaved soil with 10 seeds per pot. The pots were kept under natural conditions (month of February, 20 ± 2 ◦C, 14 h light/10 h dark) and were watered daily. Seed germination was recorded after 15 days of sowing on the basis of above ground hypocotyls. Emergence of healthy seedlings, and mean fresh and dry weights of emerged plants were recorded. Disease incidence on radish plants was determined after 28 days on the basis of percentage of diseased seedlings.

The ability of the strain DH16 for root and rhizosphere colonization (seeds treated with Streptomyces cells) was also determined. Roots (from 10 plants) were recovered from the

pots 28 days after planting and were cut into 1 cm pieces and suspended in 1 ml of autoclaved distilled water. Similarly, rhizosphere soil (10 soil samples) adhered to roots was carefully removed, weighed (1 g) and serially diluted. Then 100 µl from root aliquots and soil dilution (10−<sup>4</sup> ) were spread on SCNA plates and cfu/ml were calculated after incubation for 7 days at 28◦C.

### Biocontrol Potential of Culture Supernatant and S. hydrogenans As Foliar Treatment against A. brassicicola to Control Black Leaf Spot

The biocontrol potential of cells of S. hydrogenans and its cell free culture supernatant against A. brassicicola was also studied as foliar treatments using whole plants. Surface sterilized seeds of radish were sown in pots containing autoclaved soil. After 15 days, the leaves were given different treatments (i) inoculation with 10 µl of fungal pathogen (10<sup>5</sup> spores/ml; inoculated control), (ii) co-inoculation of pathogen (10 µl) and culture supernatant of antagonist (10 µl of 10%), (iii) co-inoculation of pathogen and cells of antagonist (107– 10<sup>8</sup> cells/ml), and (iv) non-inoculated healthy control (water only).

In another treatment, cell suspension/culture supernatant of S. hydrogenans was also applied to soil containing uninoculated radish plants.

After 20 days of various treatments, the plants were uprooted, fresh and dry weights of plants were recorded. Disease incidence was determined on the basis of dry weight of plants compared to control plants. The experiment was repeated twice.

### Statistical Analysis

All the experiments were repeated twice and the data (expressed as the mean ± SD) obtained from these experiments were subjected to statistical analysis. Tukey's post hoc test was done with the help of ASSISTAT (7.7 β) to compare the means.

### RESULTS

### Suppression of A. brassicicola in Liquid Broth by Culture Supernatant

Present study demonstrated inhibition of mycelial growth of A. brassicicola grown in broth supplemented with culture supernatant of DH16. In comparison to control, the mycelial dry weight of pathogen in PDB was significantly lowered in the presence of culture supernatant and this suppression of mycelial growth was found to be depended on the concentration of bioactive metabolites present in the culture supernatant. More than 50% inhibition was achieved at concentration of 5% and complete inhibition of mycelial growth occurred at 20% antifungal metabolites (P = 0.05; **Table 1**). Furthermore, the change in pH of the spent medium varied between 6.5 and 7.15, which suggested that the resulted mycelial inhibition was not due to pH change.

TABLE 1 | Effect of filter sterile culture supernatant of Streptomyces hydrogenans DH16 on mycelial growth of Alternaria brassicicola when supplemented in potato dextrose broth at different concentrations.


<sup>a</sup>Means ± SD of two independent experiments, followed by the different letters within a column are significantly different according to Tukeys Test with P ≤ 0.05. <sup>b</sup>Final pH of the spent medium.

### Effect of Culture Supernatant on Fungal Morphology

The inhibition of fungal pathogen by bioactive metabolites in the culture supernatant of the Streptomyces DH16 prompted us to examine the effect of its metabolites on the spore and mycelial structures. Microscopic studies demonstrated severe morphological abnormalities such as hyphal swellings resulting in bulbous structures, granular cytoplasm, leakage of cellular materials, thinning of hyphae, discoloration of hyphae, caused by metabolites present in the culture supernatant. Extracellular metabolites completely inhibited the sporulation along with loss of pigmentation (**Figure 1**).

### Effect of Culture Supernatant on Spore Germination

The effect of culture supernatant on spore germination was studied by incubating spores of A. brassicicola with different concentrations of supernatant. In control, 65% of the spores germinated after 8 h. At lower concentrations of 0.5 and 1%, the germination was not greatly affected as compared to control. However, it was significantly reduced to 25 and 10% at concentrations of 5 and 7.5%, respectively (p ≤ 0.0001). Concentrations of 10 and 20% were found to be completely lethal (no spore germination), and resulted in loss of pigmentation and shrinkage of spores.

### Extracellular Conductivity

Exposure of A. brassicicola to the antifungal metabolites of S. hydrogenans strain DH16 for 12 and 24 h resulted in increased levels of extracellular conductivity as compared to control, which showed leakage of cellular electrolytes from test fungus due to loss of cell wall/cell membrane integrity (**Figure 2**).

### Buried Slide Technique

The results of buried slide technique demonstrated 100% spore germination in control soil after 2 days of incubation (**Figure 3**). However, germination was significantly inhibited (only 7.14%) in streptomycete inoculated soil. With further incubation, germ

FIGURE 1 | Effect of culture supernatant of Streptomyces hydrogenans strain DH16 on Alternaria brassicicola mycelial morphology. (A) Firm control mycelium (B,C) treated hyphae. Arrows indicate abnormalities like distorted mycelial structure, swellings, leakage of cellular material, loss of pigmentation in mycelium.

tubes developed to form long hyphal threads in control soil where as in treated soil, small germ tubes of half the length of spores were formed in germinated spores. After 6 days incubation, complete mycelial structure with new sporulation was seen in control soil whereas in treated soil, germ tubes with loss of pigmentation and high vacuolization were observed.

## Biocontrol Efficiency of Streptomyces DH16 Cells and Culture Supernatant As Seed Treatment

### In vitro Plate Assay

In vitro biocontrol potential of cells and culture supernatant of streptomycete against A. brassicicola was studied as seed treatment in radish. Seed germination, number of healthy seedlings and seedling vigor were found to differ significantly in treated and non-treated seeds (P ≤ 0.05; **Table 2**). In the seeds treated with pathogen alone, the percentage of seed germination and healthy seedlings, fresh and dry weights, and seedling vigor were significantly lower as compared to the uninoculated control seeds. On the other hand, treatment of pathogen infested seeds with culture supernatant at the highest concentration of 20% significantly improved seed germination and seedling vigor to 80% and 1538, respectively and were comparable to control. The percentage of healthy seedlings (90%) and their fresh and dry weights were also significantly higher in treated seeds. Similarly, treatment of pathogen infected seeds with antagonist also significantly improved all the parameters as compared to pathogen infested seeds. Additionally, seeds treated with streptomycete/culture supernatant only were found to be healthier than the uninoculated control seeds. Strain DH16 showed significant stimulatory effect and an increase of 21– 35% over control in various growth parameters of seedlings was observed.

### In vivo Pot Assay

The effect of seed treatment on germination of seeds and growth of emerged seedlings was observed for plants grown in soil. Culture supernatant at concentration of 10% (v/v) was selected for further experimentation, since extracellular metabolites at this concentration were found to be strongly inhibitory to the pathogen as shown in in vitro blotter test. Significant differences in percentages of seed germination, disease incidence, healthy seedlings, and in fresh and dry weights of emerged plants were found between the different treatments (**Figure 4**). The pathogen infected radish seeds when treated with Streptomyces DH16 cells/metabolites, germinated and emerged as healthy seedlings, at significantly (P ≤ 0.001) higher rates (70–95%)

as compared to seeds treated with pathogen only (5–25%). Similarly, disease severity was also significantly reduced in treated seeds in both the treatments. Although the fresh and dry weights of plants emerged from seeds treated with pathogen and antagonists were lower than the water treated control plants but the plants were higher and stronger than the plants emerged from seeds treated with pathogen only. However, the treatment of seeds with S. hydrogenans DH16 cells/ culture supernatant only, significantly enhanced the germination rate and growth of emerged plants with significantly (p ≤ 0.05) higher fresh and dry weights as compared to control plants (**Figure 5**). When serially diluted aliquots of root segments and rhizosphere soil of plants

TABLE 2 | Effect of S. hydrogenans strain DH16 and its metabolites on seed germination, growth of seedlings and seedling vigor in R. sativus during in vitro blotter assays carried out for 7 days


Data represents Means ± SD of two separate tests (total 60 seeds for each treatment) where seeds were artificially infested with pathogen and then given different treatments with streptomycete cells and culture filtrate, different letters with in the column are significantly different according to Tukeys test; p ≤ 0.01.

two experiments with five different treatments of artificially A. brassicicola inoculated seeds, each treatment consist of three plates per experiment with 10 seeds per plate, i.e., total 60 seeds for each different treatment. Error bars represent approximate 95% confidence limit. Same letters on the bar are not significantly different according to Tukeys test (p ≤ 0.01); C, Control (Water only); P, Pathogen only; PE, Pathogen + Extract; PC, Pathogen + Streptomyces cells; SC, Streptomyces cells only.

emerged from streptomycete treated seeds were plated, the cfu counts of 1 × 10<sup>7</sup> /cm and 1 × 10<sup>8</sup> /g soil, respectively, were observed.

### Biocontrol Potential of Culture Supernatant and Streptomyces DH16 As Foliar Treatment

**Figure 6** demonstrates the biocontrol ability of Streptomyces DH16 and its culture supernatant as effective foliar treatments to control black leaf spot disease on radish leaves. In control plants, the symptoms of disease caused by A. brassicicola were clearly visible with 66.81% disease severity and yellowing of leaves and therefore under developed roots. On the other side, symptoms of disease were rarely observed in plants treated with antagonist and its metabolites which significantly reduced the disease severity to 6.78 and 1.47%, respectively (p ≤ 0.01). The fresh and dry weights of treated plants were also significantly higher over the pathogen inoculated plants (p ≤ 0.05). Weight of tap roots (edible part) in treated plants was comparable to the control plants, in contrast to pathogen treated plants where they were not developed properly. As compared to uninoculated control plants, edible tap roots with high fresh weight were obtained in plants where soil was irrigated with spore suspension/culture supernatant of S. hydrogenans DH16 (**Figure 7**).

disease. (A) Percentage of disease incidence (B) fresh weight of plants (C) dry weight of plants (D) weight of root tuber. Error bars represent approximate 95% confidence limit. Same letters on the bar are not significantly different according to Tukeys test (p ≤ 0.01); C, Control (Water only); P, Pathogen only; PE, Pathogen + Extract; PC, Pathogen + Streptomyces cells; SC, Streptomyces cells only.

### DISCUSSION

Biological control of various plant diseases that reduce crop yields has gained the attention of researchers due to its safe and ecofriendly application and therefore, will play vital role in continuous supply of food to the increasing world population. Among various groups of microbial biocontrol agents, actinobacteria especially, number of Streptomyces spp. has been reported as potential biological agents to control various phytopathogenic fungi (Mukherjee and Sen, 2006; Prabavathy et al., 2006; Bressan and Figueiredo, 2008; Li et al., 2011).

This study demonstrated in vitro and in vivo potential of S. hydrogenans strain DH16 to control black leaf spot and damping off of R. sativus caused by A. brassicicola. In the present investigation, in vitro studies revealed inhibition of mycelial growth and spore germination of A. brassicicola in the presence of culture supernatant of S. hydrogenans. With increasing concentration of culture supernatant, suppression of both

FIGURE 7 | Biocontrol potential of S. hydrogenans DH16 and its culture filtrate as foliar application on A. brassicicola causal agent of leaf spot disease in pot experiments. Radish plants 20 days after inoculation with (A) pathogen (B) pathogen + Streptomyces cells (C) pathogen + Streptomyces extract (D) water only (E) Streptomyces DH16 only (F) Streptomyces extract.

mycelial growth and conidial germination increased and almost complete inhibition was observed at concentration of 20%. These results coincides with earlier reports of Prapagdee et al. (2008) and Li et al. (2011) who also studied the effect of culture filtrates of Streptomycesspp., S. hygroscopicus, and S. globisporus JK-1, on the mycelial growth of Colletotrichum gloeosporioides and Sclerotium

rolfsii, and Magnaporthe oryzae, respectively. In contrast, fistupyrone, a metabolite from endophytic Streptomyces sp. TP-A0569, showed inhibitory effects on spores of A. brassicicola but could not inhibit its mycelial growth on PDB and PDA even at concentration of 1000 ppm (Aremu et al., 2003).

Antibiotics from actinobacteria antagonize phytopathogenic fungi by inducing various morphological alterations such as stunting, distortion, swelling, hyphal protuberances in mycelial structure or the highly branched appearance of fungal germ tubes (Gunji et al., 1983). Getha and Vikineswary (2002) observed severe morphological changes such as swellings, distortions, and excessive branching in Fusarium oxysporum f.sp. cubense race 4 caused by extracellular metabolites from S. violaceusniger strain G10. Prapagdee et al. (2008) also reported the absence of C. gloeosporioides conidia as one of the malformations caused by S. hygroscopicus and its sterile culture filtrates.

Similarly, in the present study microscopic observations of fungal mycelia from the margins of the inhibition zones (resulted from culture broth of strain DH16) revealed severe structural alterations in vegetative cells and spores, which indicated that metabolites probably attack the cell wall/cell membrane. These findings further confirmed the antibiosis as the selective mechanism of antifungal activity of strain DH16. One of the antifungal compounds of S. hydrogenans strain DH16 has been purified and characterized to be 10-(2,2-dimethyl-cyclohexyl)- 6,9-dihydroxy-4,9-dimethyl dec-2-enoic acid methyl ester (SH2; Kaur et al., 2016b).

The mechanism of antibiosis is considered to be advantageous in biological control of plant diseases because antimicrobials can diffuse rapidly in nature, and thus, direct contact between the pathogen and antagonist is not indispensable (Hajlaou et al., 1994). In addition, loss of pigmentation (melanin) in hyphae as well as spores of A. brassicicola was also observed. Loss in integrity of cell wall/cell membrane was further confirmed by leakage of cellular materials (electrolytes) indicated by changes in extracellular conductivity.

Organic farming needs alternative seed treatments to eliminate or at least effectively reduce the seed-borne pathogens (Amein et al., 2011). Biological seed treatments with microbial antagonists are attractive alternatives to the chemical pesticides because the latter lead to changes in the metabolic profiles of rhizosphere biodiversity (Correa et al., 2009) whereas the bacterization of seeds with microorganisms does not alter the beneficial rhizosphere bacterial community. Moreover, the seed bacterization method has been proved to be effective as the biocontrol agent can rapidly and extensively grow and cover the surface of the seeds and can protect the plants from invading soil-borne pathogens (Kanini et al., 2013).

In recent years, the use of crude secondary metabolites produced by Streptomyces spp. are also gaining importance in crop protection and these metabolites may be a supplement or an alternative to chemical pesticides (Zacky and Ting, 2013; Harikrishnan et al., 2014). Therefore, in this study both the cells as well as extracellular antifungal metabolites of S. hydrogenans strain DH16 were evaluated for their in vivo biocontrol potential against A. brassicicola. The treatment of pathogen infested radish seeds with streptomycete cells/culture supernatant lead to statistically significant (p ≤ 0.05) improvement in seed germination, seedling vigor and plant weight, in both blotter assay as well as in in vivo pot experiments and thus reduced the frequency rate of damping off of seedlings.

In addition to disease control, cells as well as culture supernatant of Streptomyces DH16 significantly enhanced vigor index and other agronomic parameters (fresh and dry weights) over uninoculated control when applied as seed dressing. These results indicate the plant growth promoting prospect of S. hydrogenans DH16 in the absence of pathogen stress which is to its root colonizing ability as indicated by the adherence of bacterium to the roots and associated soil. The ability to colonize roots is an important trait of plant growth promoting microorganisms for their beneficiary effects (Bouizgarne, 2013) and is also related to the effectiveness of biocontrol activity against pathogens (Bull et al., 1991). To our best knowledge, only one report is available in literature where seed treatment is used to control A. brassicicola. In 1987, Tahvonen and Avikainen reported the biocontrol of seed borne A. brassicicola of cruciferous plants with a powdery preparation of a Streptomyces sp. However, till date no biological seed treatment method has been used commercially.

Furthermore, it is challenging to obtain pathogen free seeds of Brassica spp., due to the lack of foliar fungicides (Kohl et al., 2010). Alternaria species have also been reported to develop resistance against dicarboximides and phenylpyrroles (Iacomi-Vasilescu et al., 2004). Therefore, cells and metabolites of S. hydrogenans strain DH16 were also evaluated as foliar application to control black leaf spot by A. brassicicola. Both the cells as well as metabolites caused reduction of black leaf spot on R. sativus and significantly increased the size and weight of swollen stem as campared to negative control (fungal inoculated plant). In addition, plants irrigated with spore suspension/culture supernatant of streptomycete showed growth enhancement over the untreated control plants which further confirm the PGP potential of the S. hydrogenans DH16. Fistupyrone is the only metabolite reported earlier from Streptomyces sp. which displayed in vivo suppression of black leaf spot caused by A. brassicicola on the seedlings of Chinese cabbage (Igarashi et al., 2000).

The outcomes of the current study clearly indicate the great potential of S. hydrogenans DH16 (a less studied streptomycete) as another potent biocontrol agent which can be used to control both seed borne as well as foliar pathogen, A. brassicicola. The importance of the study lies in the fact that both spores/mycelium and extracellular metabolites of this streptomycete reduced the incidence of A. brassicicola, therefore, this strain might be used as a biofungicide in two forms, one having spores and other containing antifungal metabolites. Additionally, this strain is also having insecticidal (Kaur et al., 2014), nematicidal (Kaur et al., 2016a) and plant growth promoting activities (data communicated) which make it superior over already reported chemical and biofungicides (mycostop, fistupyrone, actinovate, and rhizovit) and thus can also be used as bioinsecticide and bio-fertilizer in addition to biofungicide.

### AUTHOR CONTRIBUTIONS

fpls-07-01869 December 14, 2016 Time: 16:19 # 12

TK was involved in the planning and execution of the research work; analysis and interpretation of the data; manuscript writing following the suggestions of the research supervisor. RM as research supervisor of TK was involved in the design and planning of research work; analysis and interpretation of data; drafting as well as critical editing of the manuscript for intellectual subject matter. Both the authors approved the final version of the manuscript for publication and agreed to be accountable for all aspects of the work in ensuring that questions

### REFERENCES


related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

### ACKNOWLEDGMENTS

We duly acknowledge University Grants Commission (UGC), New Delhi for providing funds to accomplish this work. TK acknowledges the grant of fellowship under UPE (University with Potential for Excellence) scheme of University Grants Commission, New Delhi, India.

aurantiogriseus VSMGT1014 against sheath blight of rice disease. World J. Microbiol. Biotechnol. 30, 3149–3161. doi: 10.1007/s11274-014-1742-9



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Manhas and Kaur. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fmicb-07-01004 June 29, 2016 Time: 11:19 # 1

\*

# Purification and Characterization of a New Antifungal Compound 10-(2,2-dimethyl-cyclohexyl)-6,9 dihydroxy-4,9-dimethyl-dec-2-enoic Acid Methyl Ester from Streptomyces hydrogenans Strain DH16

#### Talwinder Kaur<sup>1</sup> , Amarjeet Kaur<sup>1</sup> , Vishal Sharma<sup>2</sup> and Rajesh K. Manhas<sup>1</sup>

<sup>1</sup> Department of Microbiology, Guru Nanak Dev University, Amritsar, India, <sup>2</sup> Department of Pharmaceutical Sciences, Guru Nanak Dev University, Amritsar, India

#### Edited by:

Gero Benckiser, Retired from Justus-Liebig-Universität Gießen, Germany

### Reviewed by:

Wubei Dong, Huazhong Agricultural University, China Nan Yao, Sun Yat-sen University, China

> \*Correspondence: Rajesh K. Manhas rkmanhas@rediffmail.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 22 April 2016 Accepted: 13 June 2016 Published: 29 June 2016

#### Citation:

Kaur T, Kaur A, Sharma V and Manhas RK (2016) Purification and Characterization of a New Antifungal Compound 10-(2,2-dimethyl-cyclohexyl)-6,9 dihydroxy-4,9-dimethyl-dec-2-enoic Acid Methyl Ester from Streptomyces hydrogenans Strain DH16. Front. Microbiol. 7:1004. doi: 10.3389/fmicb.2016.01004 In agriculture, biocontrol agents have been emerged as safe alternative to chemical pesticides where Streptomyces spp. and their metabolites constitute a great potential for their exploration as potent agents for controlling various fungal phytopathogens. The present study reports an antifungal compound purified from Streptomyces hydrogenans strain DH16, a soil isolate, using silica gel chromatography and semi preparative HPLC. The compound was characterized using various spectroscopic techniques (IR, <sup>1</sup>H and <sup>13</sup>C NMR) and named 10-(2,2-dimethyl-cyclohexyl)-6,9-dihydroxy-4,9-dimethyl-dec-2 enoic acid methyl ester (SH2). Compound (SH2) showed significant inhibitory activity against fungal phytopathogens and resulted in severe morphological aberrations in their structure. Minimal inhibitory and minimal fungicidal concentrations of the compound ranged from 6.25 to 25 µg/ml and 25 to 50 µg/ml, respectively. In vivo evaluation of the compound showed strong control efficacy against Alternaria brassicicola, a seed borne pathogen, on radish seeds. In comparison to mancozeb and carbendazim, the compound was more effective in controlling damping off disease. Additionally, it promoted plant growth with increased rate of seed germination, and displayed no phytotoxicity. The compound retained its antifungal activity after its exposure to temperature of 100◦C and sunlight for 1 h. Furthermore, the compound (SH2) when tested for its biosafety was found to be non-cytotoxic, and non-mutagenic against Salmonella typhimurium TA98 and TA100 strains. This compound from S. hydrogenans strain DH16 has not been reported earlier, so this new compound can be developed as an ideal safe and superior biofungicide for the control of various fungal plant diseases.

Keywords: Streptomyces hydrogenans DH16, antifungal compound, fungal phytopathogens, biofungicide, biocontrol

## INTRODUCTION

fmicb-07-01004 June 29, 2016 Time: 11:19 # 2

Plant pathogens, especially fungi are one of the major threats to agricultural productivity of economically important crops worldwide (Cornelissen and Melchers, 1993). Chemical fungicides are used in current agricultural practices to combat these phytopathogens. However, long term application of these chemicals has resulted in severe negative impacts on environment and human health. Furthermore, their indiscriminate and repeated use has triggered the emergence of resistance in phytopathogens, due to which several important chemical fungicides have lost their efficacy against the resistant pathogens in the field (Yang et al., 2008). These limitations of chemical fungicides and increased public concern for pesticide free food highlight the discovery and development of new safer fungicides (Coloretti et al., 2007).

In recent years, control of plant diseases using microorganisms and their bioactive metabolites has drawn greater attention as better alternative to chemical fungicides. The antifungal antibiotics of microbial origin are safe, broad spectrum, less toxic to host plants, easily biodegradable in the biosphere, thus low residue levels in environment and food. Cycloheximide and streptomycin from Streptomyces griseus were successfully used to control fungal and bacterial diseases in plants, respectively, for the first time (Leben and Keitt, 1954). Since then, many attempts were made to explore various antibiotics from microorganisms for control of plant diseases and some viz. blasticidin S, polyoxin, kasugamycin, validamycin, gopalamycin, dorrigocins, geldanamycin, nigericin, fistupyrone, jiggangmycin, phenyl acetic acid, azalomycin have been developed as fungicides for agricultural use (Hochlowski et al., 1996; Kim and Hwang, 2007).

Streptomyces spp. hold considerable importance in biocontrol of various plant diseases caused by diverse range of plant pests. These bacteria are the largest hub for the antimicrobial agents, and approximately two third of economically important antibiotics developed for agricultural use are from Streptomyces spp. Their use as potent biocontrol agents against phytopathogenic fungi has been reported by various workers (Yuan and Crawford, 1995; Xiao et al., 2002; Kanini et al., 2013; Faheem et al., 2015) and is mainly due to the production of various antifungal compounds (Doumbou et al., 2001; Xiong et al., 2012; Palaniyandi et al., 2013; Nguyen et al., 2015) and cell wall degrading enzymes such as chitinases and glucanases (Gupta et al., 1995; Taechowisan et al., 2003, 2005; Quecine et al., 2008). Streptomycetes have also been used as biofungicide formulations containing live mycelium or spores and their active metabolites. For example, mycostop (containing S. griseoviridis K61), Actinovate and actinoiron (containing Streptomyces lydicus WYEC 108) and Rhizovit<sup>R</sup> (S. rimosus) are commercial biofungicides used to control plant diseases caused by Phytophthora spp., Fusarium spp., Pythium spp., Alternaria brassicicola, Botrytis sp., and Rhizoctonia solani (Tahvonen and Avikainen, 1987; Marten et al., 2000).

Although different antifungal compounds from Streptomyces spp. have been reported but it is just the tip of the ice berg that has been explored. Therefore, in continuous demand for new bioactive metabolites for plant protection, the present study reports the purification, characterization and biological evaluation of a new antifungal compound from S. hydrogenans, a strong antagonist against various fungal phytopathogens. The biosafety of the compound was also evaluated using Ames Mutagenicity and MTT (3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide) cytotoxicity tests.

### MATERIALS AND METHODS

### Microorganisms and Maintenance

Streptomyces hydrogenans strain DH16 (GenBank accession no. JX123130) was isolated from soil procured from Dalhousie (32.53◦ N, 75.98◦ E), Himachal Pradesh, India (Kaur and Manhas, 2014) and maintained on starch casein nitrate agar slants at refrigeration temperature (4◦C). Twenty percent glycerol stocks were also prepared and stored at −20◦C during the study. The three test phytopathogenic fungi viz. Alternaria brassicicola (MTCC 2102), A. solani (MTCC 2101), and Colletotrichum acutatum (MTCC 1037) were obtained from Microbial Type Culture Collection (MTCC), Institute of Microbial Technology (IMTECH), Chandigarh, India. Fusarium moniliforme, Alternaria alternata were isolated in lab. All the fungal cultures were maintained on Potato dextrose agar (PDA) slants at 4◦C.

### Production of Antifungal Metabolites

Production of antifungal metabolites from S. hydrogenans strain DH16 was carried out according to Kaur and Manhas (2014). The actinobacterium was grown on starch casein nitrate agar medium at 28◦C for 7 days and then growth was scrapped and transferred aseptically into the SCN broth to develop seed culture. After 48 h of incubation, the seed culture was inoculated into 250 ml Erlenmeyer flasks containing 50 ml of production medium containing g/L: starch, 12; soyabean meal, 2.5; K2HPO4, 1.8; NaCl, 2; Casein, 0.3; MgSO4, 0.05; FeSO4, 0.01; and CaCO3, 0.02. The fermentation was carried out at 28◦C at 180 × g, and after 3 days of incubation, culture broth was centrifuged at 10000 × g for 20 min at 4◦C to obtain cell free culture supernatant.

### Extraction and Purification of Metabolites

For the recovery of active metabolites, culture supernatant (5 l) was extracted twice with equal volume of ethyl acetate (EA). The organic phase was separated, treated with Na2SO<sup>4</sup> and then concentrated to dryness under vacuum at 45◦C using rotary evaporator (BUCHI Rota vapor R-200). For the purification of antifungal compounds, the resulting solids (1 g) re-dissolved in small volume of methanol were subjected to column chromatography using silica gel (60–120 mesh size; column, 1.0 cm × 35 cm) packed and pre-equilibrated with chloroform. The column was then eluted step-wise with linear gradients of: chloroform/methanol (100:0, 90:10, 80:20,70:30, 60:40, 50:50, 40:60, 30:70, 20:80, 10:90, and 0:100) at a flow fmicb-07-01004 June 29, 2016 Time: 11:19 # 3

rate of 2 ml/min. About 200 ml of each gradient was used for elution, and a total of 88 fractions of 25 ml each were collected and concentrated. Each fraction was then subjected to disk diffusion assay to determine their antifungal activity against A. brassicicola and fractions showing antifungal activity were pooled together and concentrated. Final purification of the active compounds was achieved by preparative RP-HPLC: Shimadzu Microsorb MV, 100 mm × 10 mm ID, 10 µm, flow rate 3 ml/min, gradient Acetonitrile:H2O 50% in 5 min, 50–70% in 15 min and 70–50% in 17 min, and UV detection at 210 nm. All the peaks of chromatogram were collected using a fraction collector coupled with the HPLC system, then concentrated and screened for antifungal activity. Three peaks which showed activity were rechromatographed with the same solvent system to check the purity to homogeneity level.

### Characterization of Active Peak 2

The peak **2** was characterized based on various physicochemical and spectroscopic properties. Appearance, color, odor, and solubility were determined according to the standard procedures. The UV-Visible spectrum was recorded qualitatively on UV-Visible Spectrophotometer (Shimadzu) in the range of 200– 400 nm using acetonitrile as reference solvent.1H NMR and <sup>13</sup>C NMR spectra were recorded in chloroform-d [99.8 atom% D, containing 0.1% (v/v) tetramethylsilane (TMS)] at 25◦C on 500 MHz AVANCE III Bruker spectrometer equipped with a 5 mm double channel solution state probe. The chemical shifts are reported in parts per million (ppm) relative to TMS (δ0.0) used as internal standard. Mass spectrum (HR-MS) was recorded with Bruker MICROTOF II spectrometer. IR spectrum was recorded with Perkin–Elmer FTIR-C92035 Fourier-Transform spectrophotometer in the range 400–4000 cm−<sup>1</sup> by using CHCl<sup>3</sup> as the medium for the preparation of the samples.

## Antifungal Activity of Purified Compound

The antifungal activity of the purified compound SH2 was tested against A. brassicicola, A. solani and C. acutatum, F. moniliforme and A. alternata causing various diseases on diverse host plants. The activity was determined in terms of zone of inhibition by using Kirby–Bauer well diffusion assay (Bauer et al., 1996). Wells of 4 mm diameter were made on PDA plates seeded with the test fungal pathogens. Then aqueous solutions of purified compound, cycloheximide and chemical control agents (carbendazim and mancozeb) were made at concentration of 1 mg/ml and 100 µl of each were added into wells. The diameters of the resultant zones of inhibition were measured in mm after 48–72 h of incubation. Each experiment was performed in duplicates and repeated thrice.

## Effect of Purified Compound on Fungal Morphology

The effect of purified compound **(SH2)** on morphology of A. brassicicola and F. moniliformae was studied microscopically. Mycelia of A. brassicicola and F. moniliforme were taken from periphery of the inhibition zones around the well (containing **SH2**) and from control plate and placed on glass slide in a drop of sterile water. The coverslip was placed on the film and then visualized under bright field microscope at 40× (Olympus). Microphotographs were taken using a digital camera.

### Minimal Inhibitory Concentration (MIC) and Minimal Fungicidal Concentration (MFC)

Minimal inhibitory concentration of EA extract was worked out by 96 well microtitre plate method (Díaz-Dellavalle et al., 2011) using different concentrations (12.5, 25, 50, 100, 250, 500, and 1000 µg/ml) of extract in EA. The spore suspension of test fungus was prepared by scrapping the spores from 5-dayold PDA culture plate with fresh PDB. In 96 well microplates, 100 µl of fungal spore suspension (1 × 10<sup>5</sup> spores/ml) was mixed with 100 µl of extract of different concentrations. Control well contained 100 µl of fungal spore suspension and 100 µl of EA. Control blanks consisted of 100 µl of extract of different concentrations with 100 µl of PDB and other contained 100 µl of PDB and 100 of EA only. The microplates were incubated at 28◦C and readings were taken with microplate reader at 595 nm after 48 h. MIC values were calculated by comparing the growth in wells containing extract to the growth in control wells and is the lowest concentration that resulted in 80% inhibition in growth compared to the growth in control well. Further MFC was determined by plating 20 µl of the broth from each well on fresh PDA plates. The plates were then incubated at 28◦C until growth was visible in the control subculture. The MFC will be that lowest concentration where no visible growth will be observed on plate. The experiments were performed in triplicates. Similarly, the MIC and MFC values for purified compound were determined using different concentrations (6.25, 12.5, 25, 50, 100, and 200 µg/ml).

### Biocontrol of A. brassicicola on Raphanus sativus Seeds by Purified Compound

In vitro and in vivo experiments were used to evaluate the biocontrol potential of compound to control A. brassicicola on surface sterilized radish seeds. The sterilized seeds were first artificially infested with the pathogen by immersing them for 4 h in spore suspension, prepared from 5-day-old A. brassicicola, in presence of 1% carboxy methyl cellulose (CMC; 105–10<sup>7</sup> spores/ml). The pathogen infested seeds were then soaked in 1 mg/ml solution of compound and chemical control agents. After 1 h, all seeds were dried in laminar flow on a sterile filter paper and used for further experiments. Each treatment consisted of three replicates with 10 seeds each.

### In Vitro Blotter Test

The moistened blotters were first used to determine the effect of compound to reduce damping off due to seed-borne A. brassicicola on radish plants grown from artificially infected seeds. Three replicates of 10 seeds per treatment were placed in Petri dishes (10 seeds per plate) already lined with moist filter paper and covered loosely with another filter paper. The number of germinated seeds and healthy and diseased seedlings were recorded after 7 days of incubation at 28◦C in the dark. Seedling vigor (V) was determined by measuring root and shoot lengths of 10 seedlings (selected randomly) and was calculated according to the equation:

$$\mathbf{V} = (\mathbf{L\_s} + \mathbf{L\_t}) \times \mathbf{G}$$

Where L<sup>s</sup> is average shoot length in mm and L<sup>r</sup> is average root length in mm and G is % germination (Andresen et al., 2015).

### In Vivo Pot Experiment

fmicb-07-01004 June 29, 2016 Time: 11:19 # 4

Treated seeds were sown in pots containing sterilized soil with 10 seeds per pot. The pots were kept under natural conditions and were watered daily. Seed germination, emergence of healthy seedlings, mean fresh, and dry weights of emerged plants and seedling vigor were recorded after 15 days of sowing. To determine the dry weight of plants, harvested plants were placed separately on filter papers in oven at 60◦C for 48 h and then weighed using weighing balance.

### Stability of Compound SH2

To determine heat stability, compound **SH2** was heated at 37, 50, 70, and 100◦C for 1 h. Photostability was also tested by exposing the compound separately to sunlight for 1 h. All the treated samples were then checked for residual activity with respect to untreated control against C. acutatum.

### Safety Evaluation

Purified compound **(SH2)** was evaluated for toxicity testing viz. phytotoxicity, mutagenicity, and cytotoxicity.

### Phytotoxicity Testing

Phytotoxicity was checked by treating sterilized seeds with purified compound. In control, the compound was replaced with water. Treated seeds were then sown in sterilized soil and data for important agronomic parameters (seed germination, seedling vigor and weight of plants) were recorded after 10 days.

### Mutagenicity Studies

The mutagenicity of purified compound was determined using Salmonella histidine point mutation assay proposed by Maron and Ames (1983) with slight modifications suggested by Bala and Grover (1989). For toxicity testing, 0.1 ml of bacterial culture and 0.1 ml of compound **SH2** at different concentrations (50, 100, and 250 µg/plate) was added to 2 ml of top agar and poured onto the minimal agar plates followed by incubation at 37◦C for 48 h. To determine the spontaneous reversion which is characteristic of the tester strains (TA98 and TA100), negative control (0.1 ml bacterial culture + 0.1 ml DMSO per plate) was run while 4-Nitro-o-phenylenediamine (20 µg/0.1 ml/plate) and sodium azide (2.5 µg/0.1 ml/plate) were used as a positive control mutagens for strains TA98 and TA100, respectively. After incubation for 48 h, the number of revertant his<sup>+</sup> bacteria colonies were scored. The mutagenic potential of the purified compound was determined by comparing the number of colonies with control plates where no test compound as well as mutagen was added.

### In Vitro Cytotoxicity

The MTT assay was used for determining in vitro cytotoxicity following the method given by Mosman (1983) using Chinese Hamster Ovary (CHO), a non-tumor or normal cell line obtained from National Centre for Cell Science (NCCS), Pune (India). Cells of CHO were grown in tissue culture flask in complete growth medium [Roswell Park Memorial Institute (RPMI)- 1640 medium with 2 mM glutamine, 100 units ml-1 streptomycin and supplemented with 10% Foetal Calf Serum (FCS) and 100 units/ml penicillin] at 37◦C in an atmosphere of 5% CO<sup>2</sup> and 90% relative humidity. The cells at subconfluent stage were harvested from the flask by treatment with trypsin (0.05% in PBS containing 0.02% EDTA) for determination of cytotoxicity. Cells with viability of more than 98%, as determined by trypan blue exclusion, were used for assay. The cell suspension of 1 × 10<sup>5</sup> cells/ml was prepared in complete growth medium for determination of cytotoxicity. Stock solution of purified compound (100 µg/ml) was prepared in sterile filtered DMSO. The compound was serially diluted with complete growth medium containing 50 µg/ml of gentamicin to obtain working test solutions of different concentrations.

The 100 µl of cell suspension (1 × 10<sup>5</sup> cells/ml) were seeded in 96-well tissue culture plates (MicrotestTM, Falcon, USA), followed by addition of 100 µl of each concentration (30, 50, and 100 µg/ml) of **SH2**, after 24 h incubation. Respective control cells were treated with medium only. After 42 h of treatment the spent medium was discarded by inverting the plate on a tissue towel and 100 µl of MTT prepared in non-serum medium (0.5 mg/ml; without FCS) was added to each well. After 2 h of incubation, medium was discarded and blue formazan crystals formed by MTT reaction were dissolved in 100 µl of dimethyl sulfoxide (DMSO, Emplura <sup>R</sup> , Millipore) in each well. The color was read at 540 nm in the Elisa plate reader (Multiscan <sup>R</sup> EX by Labsystems, Finland). The proliferation of cells under treatment was assessed according to following formula:

$$\text{Percentage of refrigeration} = \frac{\text{Absorbance of test}}{\text{Absorbance of control}} \times 100$$

% Growth inhibition = 100 – % Cell growth

### Statistical Analysis

All the experiments were repeated twice and the data (expressed as the mean ± SD) obtained from these experiments were subjected to statistical analysis. Tukey's post hoc test was done with the help of ASSISTAT (7.7 beta) to compare the means.

### RESULTS

### Isolation and Purification of the Compound

The streptomycete DH16 was grown in SCN broth for the production of active antifungal metabolites. After third day of incubation, broth was centrifuged, extracted, and concentrated to yield dark brown residue which was subjected to silica gel column chromatography for isolation of active compounds. Three active fmicb-07-01004 June 29, 2016 Time: 11:19 # 5

fractions (no. 9–11) eluted with chloroform:methanol (90:10) in silica gel chromatography showed antifungal activity and then pooled together based on their similar TLC pattern and concentrated. The pooled fraction was then further fractionated on semi-preparative HPLC and individual peaks were collected. The peak with retention time of 9.61 min showed antifungal activity against the test pathogenic fungus (A. brassicicola). To check the homogeneity of active compound, collected peak was chromatographed under similar conditions and single peak was obtained which indicated the purity of the compound (**Figure 1**).

### Physical Properties of the Purified Compound

The compound was light yellowish in color and was odorless. It was soluble in most of the organic solvents but was sparingly soluble in water.

### Chemical Characteristics of the Purified Compound (SH2) from S. hydrogenans DH16

The compound responsible for antifungal activity was characterized by various spectroscopic techniques such as IR, <sup>1</sup>H and <sup>13</sup>C NMR spectra, and mass spectroscopy. **IR** (CHCl3): νmax (cm−<sup>1</sup> ) = 3441, 3054, 2986, 2685, 2305, 1754, 1606, 1422, 1267, 1151, 908, 896, 752; <sup>1</sup>**H NMR** (500 MHz, CDCl3): δ = 7.45 (d, 1H, J = 5.5 Hz, C2-H), 6.13 (distorted d, 1H, J = 4.2 Hz, C3-H), 5.06–5.03 (m, 1H, C6-H), 3.66 (dist. s, 3H, OCH3), 1.80–1.77 (m, 4H), 1.70–1.65 (m, 4H), 1.52–1.37 (m, 8H), 1.17–1.14 (m, 8H, 2x CH3, C7-H), 0.92–0.85 (m, 8H,); <sup>13</sup>**C NMR** (150 MHz, CDCl3): δ = 173.1 (C = O), 156.2 (C-2), 121.5 (C-3), 83.3, 72.8, 71.7, 41.1, 39.9, 34.2, 33.1, 32.3, 29.8, 29.6, 26.3, 24.9, 23.5, 14.6, 8.2; **MS** (TOF, ESI): m/z:calculated for C21H38O4: 354.2; found: 377.1 [M+ Na]<sup>+</sup> (Supplementary Figures 1–5).

In the proton spectrum proton of C-2 appeared as doublet at δ 7.45 with J = 5.5 Hz and proton of C-3 appeared as distorted doublet at δ 6.13 with J = 4.2 Hz. Proton of remaining carbons in the compound showed multiple resonances by two bond and three bond couplings. <sup>13</sup>C NMR of the compound showed carbonyl resonance at δ 173.1 which is further revealed by IR spectrum showed band at 1754 cm−<sup>1</sup> . Carbon NMR also showed resonances of two olefinic carbons at 156.2 (C-2) and 121.5 (C-3), respectively. All aliphatic carbon resonances of compound also appeared in <sup>13</sup>C NMR below 100 ppm. The alcoholic function is confirmed by IR spectrum band at 3441 cm−<sup>1</sup> and also reveals the other stretching and bending vibrations of functionalities in compound; its mass spectrum showed the molecular ion peak at m/z 377.1 (M+ Na)+, which corresponds to the molecular formula of compound. On the basis of these observations, the purified compound is proposed to be 10-(2,2-Dimethyl-cyclohexyl)-6,9-dihydroxy-4,9-dimethyldec-2-enoic acid methyl ester (**Figure 2**).

### Antifungal Activity of Compound (SH2)

**Table 1** depicts the antifungal spectrum of purified compound at concentration of 100 µg against various fungal pathogens. The results showed that it significantly inhibited the test fungi with inhibition zones in the range of 15–40 mm as against 15–32 mm zones resulted from carbendazim and mancozeb. In case of C. acutatum and Alternaria spp., the purified compound was more effective as compared to chemical control agents. In addition to antifungal activity, the purified compound also showed inhibitory activity against Candida albicans and Bacillus subtilis (data not shown).

### Antifungal Effects of Purified Compound

The effect of purified compound on spore germination and hyphal morphology was studied for A. brassicicola and F. moniliforme. Microscopic observations showed that purified compound significantly inhibited spore germination in both the fungal pathogens as compared to control (p < 0.05). Severe

fmicb-07-01004 June 29, 2016 Time: 11:19 # 6

#### TABLE 1 | Antifungal activity of purified compound SH2 and chemical control agents when used at a concentration of 100 µg.


Means followed by different letters with in a row are significantly different; Tukeys HSD, p ≤ 0.05).

morphological abnormalities such as hyphal swellings resulted in bulbous structures, thinning of hyphae, discoloration of hyphae were also observed. Additionally, purified compound also resulted in pigmentation loss in mycelial structure of A. brassicicola (**Figure 3**). MIC and MFC values of the purified compound and EA extract of streptomycete were determined by 96 well plate method and are shown in

#### TABLE 2 | MIC and MFC values of crude extract and purified compound SH2 against phytopathogenic fungi.


<sup>a</sup>ND, not determined.

**Table 2**. The crude extract showed significant antifungal activity against all the tested fungi with MIC values of 50, 25, 100, 50, and 250 µg/ml for A. brassicicola, C. acutatum, A. solani, A. alternata, and F. moniliforme, respectively. The MIC and MFC values of purified compound were 25 and 50 µg/ml for A. brassicicola and were 6.25 and 25 µg/ml for C. acutatum.

### Biocontrol of A. brassicicola

fmicb-07-01004 June 29, 2016 Time: 11:19 # 7

The results of in vitro as well as in vivo experiments showed ability of compound to control A. brassicicola (**Table 3**). In comparison to chemical control agents, the compound purified from S. hydrogenans strain DH16 was found to be more potent causing significant inhibition of pathogen on seeds and resulting in emergence of healthy seedlings. In in vivo experiments, no germination was observed in seeds, treated with pathogen only. Treatment of pathogen infested seeds with compound improved seed germination and seedling vigor to 71.4% and 1890, respectively, and are comparable to control. On the other hand delayed seed germination of 42 and 14.2% was observed in case of mancozeb and carbendazim treated seeds, respectively. The percentage of healthy seedlings and their fresh and dry weights were also significantly higher in case of seeds treated with compound (p ≤ 0.05).

### Stability of Compound

No loss in antifungal activity of compound was observed after its exposure to temperatures up to 70◦C. However, a decrease of 12.5 and 37.5% in the residual activity was observed after boiling (100◦C for 1 h) and autoclaving (121◦C for 30 min), respectively. The compound was also found to be photostable as only 5% loss was observed in activity against C. acutatum.

### Toxicity of Compound

To work out the biosafety of the purified compound, phytotoxicity, Ames mutagenicity and MTT cytotoxicity tests were carried out. The compound was found to be nonphytotoxic because the seedlings emerged from seeds treated with compound showed increase in all the growth traits (shoot length, root length, seedling vigor). Rather than showing any phytotoxicity, **SH2** triggered as well as enhanced seed germination as compared to water treated seeds. The emerged seedlings were also found to be healthier than the control plants as shown by higher seedling weights. Antifungal compound (**SH2**) was also non-mutagenic at all the concentrations used in the experiment. The numbers of revertant colonies in both the positive controls were numerous; whereas, the numbers colonies of bacteria in the presence of the purified compound were similar to that of spontaneous revertant colonies for TA98 and TA100 (**Table 4**). Further, the compound showed insignificant cytotoxicity, i.e., only 11.6% inhibition (or 88.8% viable cells) against CHO cell line at the highest tested concentration (**Figure 4**).

### DISCUSSION

Streptomyces species produce vast array of antifungal compounds which play important role in biocontrol of various fungal plant diseases. These bacteria constitute major portion of total antibiotics used in agricultural sector and are still great reservoirs of new antibiotics (Tanaka and Omura, 1993). This study further adds to the potential of Streptomyces spp. as unexhausted source of potent antifungal compounds which can be exploited as biofungicides for agricultural use.


TABLE 3 | Invitroandin vivoprotective effect of purified compound SH2 to control A. brassicicolaon seeds

 of R. sativus.

fmicb-07-01004 June 29, 2016 Time: 11:19 # 8



<sup>a</sup>Values are Means ± SD (n = 3) followed by different letters with in a column are significantly different; Tukeys HSD, p ≤ 0.05; <sup>b</sup>Mean number of spontaneous revertants as determined in assays without SH2 and standard mutagens; <sup>c</sup>Mean number of revertants induced by reference mutagens, sodium azide and 4-Nitroo-phenylenediamine (NPD); <sup>d</sup>Mean number of revertants induced by DMSO in the absence of extract and mutagen; <sup>e</sup>Mean number of revertants induced by SH2 at different concentrations.

Here in this work, purification of compound from the culture extracts of S. hydrogenans strain DH16 using various chromatographic techniques resulted in apparently new antifungal compound with broad spectrum activity against different phytopathogenic fungi. This compound is proposed to be 10-(2,2-Dimethyl-cyclohexyl)-6,9-dihydroxy-4,9-dimethyl-dec-2-enoic acid methyl ester on the basis of various spectroscopic techniques.

The importance of the study lies in the fact that it is the first report on production of new potential antifungal compound by this species. The purified compound (**SH2)** showed more activity against phytopathogenic fungi esp. Alternaria spp. and Colletotrichum spp. as compared to cycloheximide and chemical fungicides (carbendazim and mancozeb). Carbendazim exhibited no activity and mancozeb showed weak activity against all the tested Alternaria spp. Similarly, indole-3-carboxylic acid, another bioactive compound from Streptomyces sp. TK-VL\_333 showed better activity than that of mancozeb whereas less activity than carbendazim when tested against F. oxysporum (a wilt pathogen; Kavitha et al. (2010). However, compound **SH2** isolated in the present study was found to be superior to both the fungicides in terms of activity against Fusarium sp. also.

The low MIC and MFC values of purified compound which varied from 6.25 to 200 µg/ml depending upon the sensitivity of test fungi further demonstrated its effectiveness to control the fungal plant pathogens. It showed lowest MIC value (6.25 µg/ml) against C. acutatum and C. gloeosporioides and highest value (100 µg/ml) against F. moniliforme. In contrast, the MIC value of 3-methylcarbazole produced by Streptomyces LJK109 against C. gloeosporioides was found to be high (30 µg/ml; Taechowisan et al., 2012). Hwang et al., 2001 reported the MIC values in the range of 50 to >1000 µg/ml of compounds, phenylacetic acid, and sodium phenylacetate isolated from Streptomyces humidus

strain S5-55 which are again higher than the compound **SH2** of present study.

The biocontrol potential of purified compounds obtained from Streptomyces spp. in controlling different fungal phytopathogens and reducing their disease incidences in in planta experiments has been reported (Matsuda et al., 1998; Hwang et al., 2001; Bordoloi et al., 2002; Kavitha et al., 2010). In current study also, the in vitro and in vivo experiments showed that the compound effectively controlled the development of seed borne damping off of radish seedlings caused by A. brassicicola when used at the concentration of 1 mg/ml whereas the mancozeb showed less control efficacy against the disease at the same concentration. Carbendazim was found to be least effective. In the absence of pathogen, the seedlings emerged from compound treated seeds were found to be more healthier than the seedlings in control. Similarly, SPM5C-1 from Streptomyces sp. PM5 when applied at 500 and 250 µg/ml significantly suppressed the sheath blight disease in rice and also increased the growth parameters as compared to the control in the absence of the pathogen (Prabavathy et al., 2006).

For commercial application of biologically active compounds (to be used as biopesticides/plant growth promoting agent) in agriculture sector, it is important to determine their phytotoxicity. Cycloheximide (isolated from S. griseus), the first compound used to control fungal and bacterial diseases in plants, showed phytotoxicity. Therefore, use of cycloheximide as an agent for plant disease control is restrained because of its toxicity to the host plants (Ford et al., 1958). The application of carbendazim, a systemic fungicide showed phytotoxicity by negatively affecting the plant biomass in Nicotiana tabacum (García et al., 2003). Walia et al. (2014) demonstrated the deleterious impact of mancozeb on soil microflora, nitrification, ammonification, carbon mineralization, soil enzymes, and soil microbial biomass which in turn may result in harmful effects on nutrient uptake and plant growth. However, the antifungal compound purified in present work did not show any phytotoxicity in both in vitro and in vivo experiments. Rather, it enhanced the rate of seed germination and seedling vigor in fmicb-07-01004 June 29, 2016 Time: 11:19 # 9

radish seedlings compared to control plants and therefore can also be used for enhancing plant growth in addition to controlling the pathogens. This data further suggests the superiority of compound **SH2** over chemical fungicides both in terms of activity as well as phytotoxicity.

The ability to tolerate various factors (light, temperature, and pH) in natural environment is very crucial for any agro active compound. Therefore, for commercial application, compound should be thermostable, photostable, and pH stable. The compound obtained from S. hydrogenans strain DH16 was found to be highly thermo and photostable. Prapagdee et al. (2008) and Jayaprakashvel et al. (2010) also reported considerable thermo and photostability of antifungal compounds produced by S. hygroscopicus and Trichothecium roseum MML00l, respectively.

Further toxicity of the compound was tested by Ames mutagenicity test and in vitro cytotoxicity test. Ames test is useful in correlating in vitro bacterial mutagenesis with in vivo carcinogenicity in animals because positive Ames test indicates that the tested chemical is mutagenic and therefore may act as a carcinogen, because cancer is often linked to a mutation. The purified compound **SH2** obtained from S. hydrogenans strain DH16 was found to be potentially bio safe as it did not show any mutagenic response (at all tested concentrations) against S. typhimurium strains TA98 and TA100 in Ames mutagenicity test. The present study gets further credence as the results of MTT assay revealed **SH2** to be non-cytotoxic as 88.8% viable cells of CHO cell line were observed in its presence at the highest tested concentration.

### CONCLUSION

This study demonstrates the purification and characterization of a new heat and photo stable antifungal compound, with plant growth promoting potential, from S. hydrogenans strain DH16, showing more promising activity against a variety of fungal phytopathogens as compared to standard chemical fungicides. The non-phytotoxic, non-mutagenic, and non-cytotoxic nature of the compound suggests that it might serve as a new, safe, and broad spectrum biofungicide to combat serious plant diseases. Therefore, the compound **SH2** can be developed as a better replacement to chemical fungicides as effective plant chemotherapeutic agent.

### REFERENCES


### AUTHOR CONTRIBUTIONS

TK was involved in the planning and execution of the research work; analysis and interpretation of the data; manuscript writing following the suggestions of the research supervisor. VS analyzed, interpreted and characterized the compound on the basis of different spectroscopic techniques and drafted related content of the manuscript. AK provided fungal cultures Fusarium moniliforme, Alternaria alternata, and Alternaria mali; helped in analysing the data and editing of the manuscript. RM as research supervisor of TK was involved in the design and planning of research work; analysis and interpretation of data; drafting as well as critical editing of the manuscript for intellectual subject matter. All authors approved the final version of the manuscript for publication and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

## FUNDING

We duly acknowledge University Grants Commission (UGC), New Delhi for providing funds to accomplish this work.

### ACKNOWLEDGMENTS

TK acknowledges the grant of fellowship under UPE (University with Potential for Excellence) scheme of University Grants Commission, New Delhi, India. We duly acknowledge Dr. Saroj Arora, Professor, Department of Botanical and Environmental Sciences for her help to determine the cytotoxicity of the compound and Shruti Chabba, Research Scholar, Department of Chemistry, Guru Nanak Dev University for help in operating NMR and MS softwares.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2016.01004


phytopathogenic fungus Alternaria spp. Chilean J. Agric. Res. 71, 231–239. doi: 10.4067/S0718-58392011000200008


griseus H7602 against Phytophthora capsici. J. Basic Microbiol. 55, 45–53. doi: 10.1002/jobm.201300820


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Kaur, Kaur, Sharma and Manhas. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fmicb-07-01004 June 29, 2016 Time: 11:19 # 10

# Induced Systemic Resistance against *Botrytis cinerea* by *Bacillus cereus* AR156 through a JA/ET- and *NPR1*-Dependent Signaling Pathway and Activates PAMP-Triggered Immunity in *Arabidopsis*

Pingping Nie1, 2, Xia Li 1, 2, Shune Wang1, 2, Jianhua Guo1, 2, Hongwei Zhao1, 2 \* and Dongdong Niu1, 2 \*

*<sup>1</sup> Department of Plant Pathology, College of Plant Protection, Nanjing Agricultural University, Nanjing, China, <sup>2</sup> Key Laboratory of Integrated Management of Crop Diseases and Pests, Nanjing Agricultural University, Ministry of Education, Nanjing, China*

#### *Edited by:*

*Gero Benckiser, University of Giessen, Germany*

#### *Reviewed by:*

*Akanksha Singh, Central Institute of Medicinal and Aromatic Plants, India Prashant Singh, Lancaster University, UK*

*\*Correspondence:*

*Dongdong Niu ddniu@njau.edu.cn Hongwei Zhao hzhao@njau.edu.cn*

#### *Specialty section:*

*This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science*

*Received: 09 December 2016 Accepted: 08 February 2017 Published: 28 February 2017*

#### *Citation:*

*Nie P, Li X, Wang S, Guo J, Zhao H and Niu D (2017) Induced Systemic Resistance against Botrytis cinerea by Bacillus cereus AR156 through a JA/ET- and NPR1-Dependent Signaling Pathway and Activates PAMP-Triggered Immunity in Arabidopsis. Front. Plant Sci. 8:238. doi: 10.3389/fpls.2017.00238*

Induced resistance response is a potent and cost effective plant defense against pathogen attack. The effectiveness and underlying mechanisms of the suppressive ability by *Bacillus cereus* AR156 to *Pseudomonas syringae* pv. *tomato* DC3000 (*Pst* DC3000) in *Arabidopsis* has been investigated previously; however, the strength of induced systemic resistance (ISR) activity against *Botrytis cinerea* remains unknown. Here, we show that root-drench application of AR156 significantly reduces disease incidence through activation of ISR. This protection is accompanied with multilayered ISR defense response activated via enhanced accumulation of PR1 protein expression in a timely manner, hydrogen peroxide accumulation and callose deposition, which is significantly more intense in plants with both AR156 pretreatment and *B. cinerea* inoculation than that in plants with pathogen inoculation only. Moreover, AR156 can trigger ISR in *sid2-2* and *NahG* mutants, but not in *jar1*, *ein2* and *npr1* mutant plants. Our results indicate that AR156-induced ISR depends on JA/ET-signaling pathway and *NPR1*, but not SA. Also, AR156-treated plants are able to rapidly activate MAPK signaling and *FRK1*/*WRKY53* gene expression, both of which are involved in pathogen associated molecular pattern (PAMP)-triggered immunity (PTI). The results indicate that AR156 can induce ISR by the JA/ET-signaling pathways in an *NPR1*-dependent manner and involves multiple PTI components.

Keywords: *Bacillus cereus* AR156, induced systemic resistance, salicylic acid (SA), jasmonic acid (JA), ethylene (ET)

### INTRODUCTION

Plants are protected from pathogen attack through activation of innate immune system, which is a consequence of co-evolution between plants and their pathogens (Jones and Dangl, 2006). The emergence of pathogens is first detected by pattern recognition receptors (PRRs) that are localized on plant cell membrane. PRRs percept the conserved pathogen identification

**263**

molecules known as pathogen-associated molecular patterns (PAMPs), which include flagellin, lipopolysaccharides, glycoproteins, or chitin (Jones and Dangl, 2006). The interaction between PRRs and PAMPs consequently triggers the socalled PAMP-triggered immunity (PTI), which initiates many immune responses, including oxidative burst, callose deposition, activation of the MAPK (mitogen-activated protein kinase) cascade, and defense-related gene expression (Altenbach and Robatzek, 2007; Schwessinger and Zipfel, 2008). During the course of evolution, some successful pathogens emerged that can successfully infect plants by suppressing PTI. This suppression is achieved by secreting virulent proteins generically termed effectors, which causes effector-triggered susceptibility (ETS) consequently (Speth et al., 2007). In response to effectors, some plants have evolved resistance (R) proteins, which can recognize effectors directly or indirectly, and elicit effectortriggered immunity (ETI). ETI usually is accompanied with a hypersensitive response (HR) at the infection site, which is thought to restrict biotrophic pathogen growth (Chisholm et al., 2006; Jones and Dangl, 2006).

Beyond defense response against an intermediate infection, resistance can be induced by a temporarily prior infection that is effective for a certain period of time, and to a broad spectrum of pathogens temporarily after (Fu and Dong, 2013). The induced defense responses can be activated by pathogen infection, microbial symbiosis, or other elicitation, such as wounding. There are two types of induced resistances that are phenotypically hard-to-distinguished: the induced systemic resistance (ISR) and systemic acquired resistance (SAR). ISR is a systemic resistance induced by some non-pathogenic rhizobacteria that can suppress disease in plants (van Loon et al., 1998). In contrast, SAR is an induced resistance that develops in whole plants in response to a temporally earlier local exposure to a pathogen. In both SAR and ISR, phytohormone signaling pathways, such as salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) are found to play crucial regulatory roles (Glazebrook, 2001).

Induced systemic resistance (ISR) has been reported in many plant species, such as rice, bean, carnation, cucumber, radish, tobacco, tomato, and Arabidopsis, which is effective against a broad spectrum of plant pathogens, ranging from fungi, bacteria, to viruses, and even to insect herbivores (Van der Ent et al., 2008; Pieterse et al., 2014). ISR requires JA and ET signaling pathways and is associated with the expression of the gene encoding plant defensin 1.2 (PDF1.2) (Van Oosten et al., 2008). For example, the rhizobacterial strain Pseudomonas fluorescens WCS417r has been shown to trigger ISR in several plant species and in Arabidopsis, where it functions through JA/ET signaling pathways and in a NPR1-dependent manner (Pieterse et al., 2002). However, dependence on both SA- and JA/ET-signaling pathways is also observed. For example, we previously reported that the ISR mediated by the rhizobacterium Bacillus cereus strain AR156 requires both the SA and JA/ET signaling pathways and NPR1 (Niu et al., 2011). Also, colonization of Arabidopsisroots by Trichoderma atroviride IMI 206040 induces the expression of SA and JA/ET pathways simultaneously to confer resistance against hemibiotrophic and necrotrophic phytopathogens (Salas-Marina et al., 2011).

In most cases, ISR is associated with a potentiated defensive capacity, which is termed "priming". Priming does not cause a direct induction of resistance-related genes, or enhance the production of phytohormones or hormone-responsive genes in systemic tissues. Instead, ISR enhances the sensitivity to hormones rather than their synthesis. Therefore, priming is costeffective in increasing plant resistance and is more efficient in activating defense mechanisms upon pathogen attack (Conrath et al., 2006; Pastor et al., 2013). Beneficial rhizobacteria trigger ISR by priming the plant for potentiated activation of varieties of cellular defense responses, such as oxidative burst (Ahn et al., 2007), cell-wall reinforcement (Heil and Bostock, 2002), defense-related enzymes accumulation (Rahman et al., 2014), and secondary metabolites production (Yedidia et al., 2003).

In our previous study, the beneficial bacterium B. cereus AR156 was demonstrated to trigger ISR in Arabidopsis through both the SA- and JA/ET- signaling pathways, which lead to enhanced resistance to bacterial infection (Niu et al., 2011). As an effort to further dissect the mechanism of AR156-mediated ISR, we are prompted to explore more protective potential of this bacterium. Botrytis cinerea, a necrotrophic fungus causing gray mold disease, is considered an important pathogen around the world. In this study, we show that B. cereus AR156 treatment inhibits B. cinerea infection in Arabidopsis through activation of ISR. The potency of induced protection is lost in jar1, ein2 and npr1 mutant but unaffected in sid2-2 and NahG plants, implicating the JA/ET signaling pathway and NPR1 are required but the SA signaling pathway is dispensable for AR156-induced ISR to B. cinerea infection. We also show that the primed defense in AR156-treated plants is mediated by enhanced activation of multiple PTI defense responses.

## MATERIALS AND METHODS

### Plants and Growth Conditions

Arabidopsis thaliana plants were maintained at 22◦C with a 12-h light/12-h dark photoperiod. Arabidopsis thaliana ecotype Col-0 and mutants were cultivated in vermiculite. All plants were used for experiments when they were 4 weeks old. The mutants used in this study (sid2-2, NahG, jar1, ein2, npr1) were described elsewhere (Staswick et al., 1992; Bowling et al., 1994; Delaney, 1994; Alonso et al., 1999).

### AR156 Treatment, Pathogen Inoculation, and Disease Assays

Bacillus cereus AR156 was grown on Luria-Bertani (LB) agar plates at 28◦C for 24 h. Subsequently, bacterial cells were pelleted by centrifugation and were resuspended in sterile 0.85% NaCl with a final concentration of 5 × 10<sup>8</sup> CFU/ml. For a protection assay, AR156 or corresponding mock (0.85% NaCl) was root-drench applied 7 days prior infection. For fungal infection experiments, 4-weeks-old plants were used. Botrytis cinerea strain B1301 was cultivated on PSA agar medium for 7 days. Spores were collected in B. cinerea infection buffer to prepare the inoculum and adjusted to a final concentration of 1 × 10<sup>6</sup> spores/ml. Inoculation was carried out by depositing a 10 ul droplet on each side of the midvein. Ten inoculated

plants for each genotype were placed in plant growth room maintained at a high humidity. For each time point, at least three biological replicates were analyzed. In planta fungal growth was examined by analyzing the transcript levels of B. cinerea actin gene (BcActin) using primer BcActin-1F (5′ -TCC AAG CGT GGT ATT CTT ACC C-3′ ) and BcActin-1R (5′ - TGG TGC TAC ACG AAG TTC GTT G-3′ ). The Arabidopsis actin gene (AtActin2) amplified by primer AtActin2-1F (5′ -GGC GAT GAA GCT CAA TCC AAA CG-3′ ) and AtActin2-1R (5′ -GGT CAC GAC CAG CAA GAT CAA GAC G-3′ ) was used as an internal control.

### RNA Extraction and qRT-PCR Analysis of Gene Expression

Total RNA was extracted from Arabidopsis leaves with TRIzol Reagent (Invitrogen, San Diego, CA, U.S.A). In brief, 1 ug total RNA was used for cDNA synthesis by using a commercial reverse transcription system (TaKaRa Biotech, Dalian, China). After the cDNA was diluted 10 times, 2 ul diluted cDNA was used for realtime quantitative PCR with the following program: 40 cycles at 95◦C for 30 s, 55◦C for 30 s, and 72◦C for 34 s. Three replications were performed for each sample. The data were normalized with AtActin, and the means of three replications were presented. Primers used in qRT-PCR were listed in Table S1.

### Examination of Hydrogen Peroxide Accumulation and Callose Deposition

Hydrogen peroxide accumulation and callose deposition examination was performed according to previously described procedures (Niu et al., 2011). Briefly, for accumulation of ROS, Arabidopsis leaves from at least three different plants were stained with DAB solution (1 mg of diaminobenzidine per milliliter, pH 3.8) for 8 h dark at 25 to 28◦C. After being cleared with 96 % (vol/vol) ethanol and preserved in 50% (vol/vol) ethanol, hydrogen peroxide was visualized as dark-brown precipitate under the light microscope. For callose deposition, Arabidopsis leaves were immerged in 5 ml of destaining solution (Acetic acid/ethanol = 5:95) (vol/vol) and were infiltrated by applying a vacuum for 5 to 10 min. Leaves were incubated in a 60◦C water bath for 20 to 30 min to clear chlorophyll. The chlorophyll-free leaves were gently rinsed with water and were then soaked in 3 to 5 ml of 0.01% (wt/vol) aniline blue staining solution containing 150 mM K2HPO<sup>4</sup> (pH 9.5) kept in dark for 2 to 4 h. After staining, Arabidopsis leaves were gently rinsed with water and were then mounted on microscope slides that were observed under an epifluorescence microscope with a UV excitation filter. Levels of callose deposition were quantified using Imge J software and expressed relative to total leaf area as described (Luna et al., 2011).

### Protein Extraction and Analysis

Plant tissue was ground in liquid nitrogen and total proteins were extracted using 2 × SDS loading buffer. The samples were resolved on SDS–PAGE gels and transferred onto nitrocellulose membranes. The blots were probed with appropriate antibodies: monoclonal mouse anti-α tubulin (1:4,000 dilution); polyclonal rabbit anti-PR1 (1:2,000 dilution). For MAPK activity assay, sample were analyzed by Western blotting using monoclonal rabbit phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) (D13.14.4E) XP antibodies (Cell Signaling Technology, #4370S, 1:2,000 dilution). For these assays, α-tubulin was used as a loading control.

## RESULTS

### *B. cereus* AR156 Induces an Effective ISR against *Botrytis cinerea* Infection

We previously found in Arabidopsis that B. cereus AR156 could prime the whole plant for an induced resistance to Pst DC3000 infection. In this study, we designed experiments to test whether AR156 is able to induce ISR against B. cinerea as well. Arabidopsis was first pretreated with AR156 or mock (0.85% NaCl) for 7 days in a root-drench application manner, followed by B. cinerea infection. Effect of AR156 on ISR was examined by the plant performance against pathogen infection. Two days after pathogen inoculation, mock-treated plants showed typical symptoms of B. cinerea disease-severe necrosis around inoculating loci, leaves yellowing, and watersoaked spots surrounded by the spores (**Figure 1A**). In contrast, plants with AR156-pretreatment exhibited a significant (P < 0.01) reduction on disease symptoms, manifested by smaller necrosis size, less yellowing, compared with mock-treated plants (**Figure 1B**). As an indicator of the effectiveness of elimination of pathogen infection, fungal hyphae were measured 2 days after inoculation (dpi) by qRT-PCR. In agreement with the reduced disease symptom observed on leaf surface, significantly reduced fungal hyphae were detected inside leaf tissue from AR156 pretreated plants than that with control pre-treatment, indicating AR156 effectively protected Arabidopsis from B. cinerea infection (**Figure 1C**).

### AR156 Induces ISR by Potentiating PR1 Expression, Hydrogen Peroxide Accumulation, and Callose Deposition in Arabidopsis

Establishment of ISR is usually accompanied with potentiated activation of various cellular defense responses against pathogen infection, which was called priming (Conrath et al., 2001). To investigate whether AR156-induced ISR was accompanied with primed defense responses in systemic leaves, we measured the expression level of defense-related protein PR1. With AR156 or control pre-treatment alone, PR1 expression was almost negligible prior pathogen infection, indicating AR156 by itself in not capable of inducing PR1 expression (**Figure 2A**). At 12 hpi, elevated PR1 expression was detected in B. cinerea-infected leaves, indicating the Arabidopsis innate immunity system recognizes and responds rapidly against B. cinerea infection. Moreover, plants with AR156 pre-treatment accumulated much higher PR1 than the control treated plants did. These results indicate that in plants with AR156 pretreatment, the innate immunity system is at a potentiated status so that plants could launch a much accelerated defense response upon infection (**Figure 2A**).

was determined by ratios of *BcActin/AtActin*. A Student's *t*-test was used to determine significant differences between the AR156-treated sample and the control (\*\**P* < 0.01). The means values ± *SD* (*n* = 12) from one representative experiment among three independent repeats are shown.

To investigate the molecular mechanism of AR156-mediated ISR, H2O<sup>2</sup> accumulation and callose deposition pattern were examined in plants inoculated with B. cinerea, with or without AR156 pretreatment. H2O<sup>2</sup> accumulation and callose deposition are two rapid responses elicited by pathogen infection, intensity and rapidity of which are hallmarks of successful immune response. AR156 neither induce the production of H2O<sup>2</sup> at 12 hpi, nor 24 hpi, indicating that AR156 itself is not an elicitor of plant defense (**Figure 2B**). At 12 hpi, B. cinerea infection alone did not induce measurable H2O<sup>2</sup> accumulation, indicating at this stage of infection, plants were not able to deploy an effective defense response yet, in term of H2O<sup>2</sup> accumulation. However, in plants pretreated with AR156, detectable H2O<sup>2</sup> level was observed, indicating AR156 pretreatment potentiated plants for an expiated immune response (**Figure 2B**). At 24 hpi, detectable H2O<sup>2</sup> level was observed in plants inoculated with B. cinerea, indicating an initiation of H2O2-mediated defense response later than 12 hpi but before 24 hpi. In accordance with observation made at 12 hpi, AR156 pretreatment increased the extent of H2O<sup>2</sup> accumulation upon B. cinerea infection (**Figure 2B**). We also compared callose deposition between plants with and without AR156 pretreatment (**Figures 2C,D**). Similar to H2O<sup>2</sup> accumulation, AR156 alone did not induced any detectable callose deposition. With B. cinerea infection, at 12 hpi, callose deposition was detectable, indicating plants were able to perceive the pathogen infection and response efficiently. Moreover, in plants pretreated with AR156, callose deposition was noticeably enhanced, indicating a potentiated defense response due to AR156 pretreatment. In agreement, callose deposition at 24 hpi exhibited similar pattern. Taken together, our results indicate that AR156 primes plants for accelerated and enhanced immune capacity, which is induced only upon pathogen attack and leads to rapidly activated cellular defense responses in systemic tissue.

### AR156-Mediated ISR Is Dependent of the JA/ET Signaling Pathway and *NPR1*

To further elucidate the molecular mechanisms responsible for AR156-triggered ISR, plants defective in different hormone signaling pathways were analyzed. As indicated in **Figure 3A**, 2 days after pathogen infection, B. cinerea caused visible disease symptom on both wild type (Col-0) and sid2-2, NahG, jar1, ein2, npr1 mutant plants, indicating none of the mutant plants exhibited an automatic and effective immune response against B. cinerea infection (**Figure 3A**, control set). Pretreatment with AR156 led to a reduction in disease symptom on Col-0, sid2- 2 and NahG plants, indicting protection mediated by AR156 is still functional in these mutants as effective as in corresponding

deposition were observed under light and epifluorescence microscopes with a UV excitation filter. Relative callose quantities in droplet inoculated leaves of 4-week-old plants. Callose was quantified from digital microscopy photographs (D). Shown are mean areas of callose per leaf relative to total leaf area ± *SD* (*n* = 24). A Student's *t*-test was used to determine significant differences between the AR156-treated sample and the control (\*\**P* < 0.01). Similar results were obtained in three independent repeats.

wild type plants. In contrast, no discernable difference could be observed between the control- and AR156-treated plants on the jar1, ein2 and npr1 mutant plants, indicating the AR156-mediated protection was jeopardized in these mutant plants (**Figure 3A**, AR156 set). In consistence, plants with AR156 pretreatment exhibited a significant (P < 0.01) reduction in necrosis size on Col-0, sid2-2 and NahG plants, but not the jar1, ein2 and npr1 mutant plants, when compared with controltreated plants (**Figure 3B**). Fungal growth in each signaling mutant plants were also examined at 2 dpi. In consistence with the disease symptom observed on leaf surface, dramatically reduced fungal growth was detected in Col-0, sid2-2 and NahG plants. In contrast, this protection was abolished in jar1, ein2 and npr1 mutant plants (**Figure 3C**). These results suggested that AR156-mediated ISR functions by activating the JA/ET signaling pathway and is dependent on NPR1.

It is known that plant innate immunity is modulated by phytohormone signaling networks. To investigate the signaling pathways employed by AR156-mediated ISR, we inspected the transcription of signaling pathway reporter genes, such as PR1, PR2, PR5, and PDF1.2 by RT-PCR. Transcriptions of PR1, PR2, PR5, and PDF1.2 were strongly induced at 48 hpi in Col-0, sid2-2 and NahG mutant plants pretreated with AR156 and inoculated with B. cinerea. Impressively, transcriptions of all these genes were not responsive to B. cinerea infection in the jar1, ein2 and npr1 mutant plants, perfectly conforming to the defective ISR mediated by AR156 (**Figure 4A**). Furthermore, cellular defense responses activated by AR156-mediated ISR was investigated in the above-mentioned five signaling mutant and their corresponding wild type Arabidopsis plants. H2O<sup>2</sup> accumulation and callose deposition were detectable at 12 hpi in Col-0, sid2-2 and NahG plants with AR156 pre-treatment, but not detectable (or to a much lesser degree) in jar1, ein2 and npr1 mutant at the same time point. At 24 hpi, a combination of AR156 pretreatment and B. cinerea inoculation led to more H2O<sup>2</sup> accumulation and callose deposition in the leaves of Col-0, sid2-2 and NahG plants, compared with the control-pretreatment group. At the same time point, H2O<sup>2</sup> accumulation and callose deposition were very weak in jar1, ein2 and npr1 plants in both control- and AR156-treated plants (**Figures 4B–D**).

### AR156-Induced ISR Activates PTI Components

PAMP-triggered immunity (PTI) plays an important role for plant immunity against B. cinerea infection, in which activation of MAPKs is one of the earliest sign of PTI. Therefore, activation of MAPKs could be used as a good indicator for activated PTI, as well as defense responses (Eckardt, 2011; Meng and Zhang, 2013; Singh et al., 2013). To further investigate the relationship between AR156-induced ISR and PTI, antibodies specifically recognize MPK3 and MPK6 were used. In plants with control pretreatment, expression of MPK3 and MPK6 were detected 10 min after B. cinerea inoculation, which was then gradually decrease through 30 to 60 mpi. In contrast, in plants pretreated with AR156, we observed a sustained and gradually increased activation of MPK3 and MPK6 from 10 to 60 min after B. cinerea inoculation, which attained its maximums at 60 min (**Figure 5A**). Our results suggest that the AR156-mediated ISR is associated with an activated and enduring PTI response. MPK3 and MPK6 activation was also examined in sid2-2 and NahG mutant plants pretreated with AR156. Interestingly, sustained activation of MPK3 and MPK6 from 10 to 60 min was also detected in spite of the defect on SID2

(**Figure 5B**) or over-expression of NahG (**Figure 5C**). Our results indicate that MPK3 and MPK6 induction is SA-dependent. This may explain why AR156-mediated ISR is SA-independent.

To further support our hypothesis, we examined other PTI marker gene, such as the flg22-induced receptor-like kinase 1 (FRK1) and WRKY53 (Asai et al., 2002; Singh et al., 2012). qRT-PCR analysis showed that FRK1 and WRKY53 were activated and remained active by B. cinerea alone from 10 to 60 min. However, in AR156-pretreated plants, both the expression level was significantly increased after B. cinerea infection, when compared to the control-pretreated plants (**Figure 5D**). These results confirmed that there is a close association between AR156 mediated ISR and a rapid and sustained activation of PTI.

### DISCUSSION

Induced systemic resistance (ISR) has been recognized as an effective biological control agent that could induce plant defense against a broad range of pathogens. Our precious studies demonstrated that B. cereus AR156 is a plant growth–promoting rhizobacterium that can induce resistance against Pst DC3000 on Arabidopsis and tomato (Niu et al., 2011, 2012). In particular, our study in Arabidopsis showed that the AR156-mediated ISR against the biotrophic pathogen Pst DC3000 is dependent on SA and NPR1, but not the ET/JA signaling pathway. In the current study, we demonstrated that the AR156-mediated ISR is also effective against the necrotrophic pathogen B. cinerea, and this protection is mediated by the ET/JA-, but not the SA-signaling pathway. However, we did identified NPR1 as an indispensable component of this specific ISR, despite the fact that other SA signaling pathway components were not involved. Therefore, our results indicate that the AR156-mediated ISR is effective against both bio- and necro-trophic pathogens, which make it a good candidate for broad-spectrum biological control agent. Meanwhile, our results also indicate that NPR1 could potentially function independent of the SA signaling pathway.

In recent years, many studies have associated PGPRs with improving plant health by enhancing defense against a broad range of pathogens (Pieterse et al., 2014; Rahman et al., 2014; Ma et al., 2016). B. cereus AR156 is a plant growth–promoting rhizobacterium that induces resistance against Pst DC3000 on Arabidopsis and tomato (Niu et al., 2011, 2012). The aim in this study was to explore the potential role of AR156 in eliciting ISR in Arabidopsis against B. cinerea infection. Roots application

#### FIGURE 4 | Continued

value of *AtActin*, which is assigned to 1. Data are presented as the means ± *SD* from three independent experiments and different letters above the columns represent statistically significant differences (*p* < 0.01) between Col-0, *sid2-2*, *NahG*, *jar1*, *ein2* and *npr1* mutant plants. *In situ* detection of accumulation of H2O<sup>2</sup> (B) and callose deposition (C, D) after inoculation with *B. cinerea*. Accumulation of H2O2 and callose deposition in leaves were detected by DAB staining and aniline blue staining, respectively. Relative callose quantities in droplet inoculated leaves of 4-week-old plants. Callose was quantified from digital microscopy photographs. Shown are mean areas of callose per leaf relative to total leaf area ± *SD* (*n* = 24). A Student's *t*-test was used to determine significant differences between the AR156-treated sample and the control (\*\**P* < 0.01).

Frontiers in Plant Science | www.frontiersin.org February 2017 | Volume 8 | Article 238

the AR156-treated sample and the control (\*\**P* < 0.01).

as an internal control. The means values ± *SD* from three independent repeats are shown. A Student's *t*-test was used to determine significant differences between

FIGURE 6 | A proposed model of AR156-mediated ISR against biotrophic and necrotrophic pathogens. When plants are pretreated with AR156, an ISR omnipotent to pathogens with both biotrophic and necrotrophic life styles is induced. When plants in a potentiated immune status is infected by a biotrophic pathogen, such as *Pst* DC3000, both SA and JA/ET signaling pathways are activated. Through a mechanism dependent on *NPR1*, downstream defense-related genes, such as *PR1*, *PR2*, *PR5*, and *PDF1.2* are expressed, and cellular defense responses, such as H2O2 accumulation, callose deposition are activated; when plants are challenged with necrotrophic pathogens, such as *B. cinerea*, only the JA/ET signaling pathways is activated. The necrotrophic-effective ISR is also dependent on *NPR1.* Dashed lings: protective function to other unidentified elicitors.

of AR156 significantly reduced necrosis diameter and inhibited fungal growth on the leaves of Arabidopsis plants (**Figure 1**). Since AR156 only colonized the roots but B. cinerea was a foliar applied, the lack of direct contact between these two parties indicates that it is ISR, instead of a direct limitation of pathogen, leads to the observed resistance to B. cinerea infection.

Under primed condition, the induction of defense-related PR genes following pathogen challenge has been reported in several plant-pathogen interaction (Ahn et al., 2007; Niu et al., 2011). This is also true in ISR against several hemibiotrophic and necrotrophic pathogens, such as the Harpophora oryzae-primed defense genes in the rice–Magnaporthe. oryzae interaction (Su et al., 2013) and Bacillus subtilis-induced PR genes in tomato challenged with Erwinia carotovora subsp. carotovora (Chandrasekaran and Chun, 2016). In this study, we found that PR1 protein expression was stronger in plants with AR156 pretreatment and B. cinerea infection than that in plants with pathogen infection only (**Figure 2A**). This is in consistence with reported results (Su et al., 2013; Chandrasekaran and Chun, 2016) and suggests that the induced PR expression level also contributes to defense against hemibiotrophic and necrotrophic pathogens. Rapid production of cellular defense responses in plant cells, such as quick H2O<sup>2</sup> accumulation and callose deposition (Conrath et al., 2002), has been recognized as hallmark events induced by ISR-triggering bacteria (Ahn et al., 2007; Rahman et al., 2014). H2O<sup>2</sup> and callose accumulation play important roles in plants response to B. cinerea infection (Mengiste, 2012; Schwessinger and Ronald, 2012). We found that AR156-pretreated plants accumulated higher H2O<sup>2</sup> and callose levels than control-treated plants following B. cinerea infection (**Figures 2B–D**), suggesting that AR156 primes plants for accelerated and enhanced disease resistance capacity by activating cellular defense responses in systemic tissue.

Therefore, it interesting to seek how the defense signal is transmitted to remote tissues where the pathogen threaten has not reach yet. We investigated the involvement of SA and JA/ET signaling pathway components, which were previously demonstrated to be involved in the AR156-induced defense responses (Niu et al., 2011, 2012, 2016). AR156-induced ISR was abolished in jar1 and ein2 mutants, suggesting that the defense response is induced by AR156 through JA/ET-signaling pathways. This observation is in consistence with that the JA/ET-signaling pathways are more effective against infections by necrotrophic pathogens. Components involved in the SA signaling pathway were also examined. In our study, both sid2- 2 and NahG plants showed comparable disease symptoms to wild type plants, and similar necrosis diameter and fungal growth in plants pretreated with AR156. SID2 is one of the genes involved in SA synthesis, defect of which leads to reduced cellular SA accumulation. NahG plants are transgenic Arabidopsis expressing an SA hydroxylase (NahG) that degrades SA to catechol. Unaffected ISR in these two independent SA defective plants indicates that the AR156-mediated ISR against B. cinerea is independent of cellular SA level, and pretty much neither of the SA signaling pathway. ISR in NahG transgenic plants and defective of ISR by jar1, ein2 and npr1 have been reported in Arabidopsis previously (Pieterse et al., 1996; Ryu et al., 2003). However, there is a difference between the previous findings and ours that in their experiment, rhizobacterium treatment was not associated with induction of PR gene (van Wees et al., 1999; Verhagen et al., 2004), whereas augmented PR1 protein level was observed in Arabidopsis pretreated with AR156 and then infected with B. cinerea (**Figure 2A**).

Interestingly, we previously demonstrated that in AR156 mediated ISR against the biotrophic Pst DC3000, both the SAand JA/ET-signaling pathways were simultaneously activated (Niu et al., 2011). In contrast, when we analyze the AR156 mediated ISR against the necrotrophic B. cinerea in this study, we concluded that this type of ISR was accompanied by the activation of the JA/ET-signaling pathways, but not the SA signaling pathway. It is generally accepted that biotrophic pathogens, which acquire nutrient supply from live host cells, are more vulnerable to defense through SA-signaling pathway; whereas necrotrophic pathogens, which benefit from host cell death, are better restrained by a JA/ET-dependent defense (Grant and Lamb, 2006). So we speculated that the favorable signal transduction pathway promoted during ISR not only depends on the ISR-inducing strains and the host plants (Pieterse et al., 2002; Choudhary and Johri, 2009; Shoresh et al., 2010), but also on the pathogens the ISR apply to.

Intriguingly, AR156-induced ISR was noticeably jeopardized in the npr1 mutant, the gene of which encodes a redox-sensitive transcriptional regulator of SA-dependent responses. NPR1 also is a mediator of SA-JA cross talk, and a regulator of SAR and ISR (Pieterse et al., 2014). Upon activation by SA, NPR1 acts as a transcriptional coactivator of a large set of PR genes as we observed in our study. This clearly laid a discrepancy between the induction of SA-dependent PRs and the independency on SA synthesis and cellular content, which may suggest a SA-unrelated function of NPR1. Indeed, NPR1 was shown to be required for the SA-independent but JA/ET-dependent ISR triggered by P. fluorescens WCS417r (Pieterse et al., 1998). More and more evidence point to a cytosolic function of NPR1 in JA/ET signaling and ISR (Spoel et al., 2003; Ramirez et al., 2010; Pieterse et al., 2012). It is worthy to note that NPR1 are highly expressed in Arabidopsis roots (Iyer-Pascuzzi et al., 2011), which may imply a potential role in regulating root-associated immune responses including ISR.

Our results also showed that, in AR156-primed Arabidopsis, pathogen infection triggered expression of defense-related genes, and enhanced hydrogen peroxide accumulation and callose deposition (**Figure 2**). However, we found that induced expression of PR1, PR2 and PR5 was also observed in sid2-2 and NahG mutants, which are defective for SA accumulation. This is a surprise to us because PR1, PR2, and PR5 were generally recognized as markers of salicylic acid-dependent disease responses, which should be non-responsive in SAdeficient mutants (Vlot et al., 2009). Our results suggest that in the AR156-mediated ISR against B. cinerea, PR1, PR2 and PR5 were induced through a SA-independent signaling pathway. This is supported by a prior study in which constant MPK3 and/or MPK6 activation causes PR1 induction independent of SA (Tsuda et al., 2013). Taken together, PRs gene activation in sid2-2 and NahG by AR156 pretreatment and B. cinerea infection could be a consequence of activated JA/ET signal pathways and induced MAPKs cascade.

Previous studies already indicate the importance of MAPK signaling in plant defense against infections (Asai et al., 2002; Zipfel et al., 2004). In this study, MAPK activation was detected at 10 min and decreased at 60 min in the leaves of plants only inoculated with B. cinerea, but this induction initiated at about the same time but remained very strong at 60 min in those treated with AR156 and inoculated with B. cinerea (**Figure 4A**). This indicated that AR156-pretreatment induced stronger MAPK activation than plants without pretreatment. MPK3 and MPK6 are positive regulators of plant defense responses controlling ET (Tena et al., 2011; Meng and Zhang, 2013) and JA biosynthesis (Schweighofer and Meskiene, 2008). MPK3 and MPK6 are essential for plant defense against B. cinerea (Ren et al., 2008; Han et al., 2010; Galletti et al., 2011; Mendez-Bravo et al., 2011). This is consistent with our finding that AR156-induced ISR against B. cinerea is mediated by JA/ET-signaling pathways. qRT-PCR analysis of the MAMP-specific early-defense marker genes, such as FRK1 and WRKY53 showed that MAMP-mediated defense responses occur rapidly after treatment with AR156 and inoculation with pathogen (**Figure 5D**), implying that AR156 induces SAR through the activation PTI response.

In our previous study, the AR156-mediated ISR could efficiently protect plants against infections by biotrophic pathogen, such as Pst DC3000. This ISR simultaneously activate the SA- and the JA/ET-dependent signaling pathways, as evident by the induced expression of PR1, PR2, PR5, and PDF1.2 (Niu et al., 2011). In this study, we demonstrated that AR156 was also effective in protecting plant from infection by necrotrophic pathogens, such as B. cinerea. However, in this case, the JA/ET signaling pathways but not the SA signaling pathway is involved. In both cases the induced ISR is associated with similar defense responses, such as activated cellular defense responses, such as H2O<sup>2</sup> accumulation, callose deposition, and expression of some defense related genes. Therefore, we propose a model that AR156-mediated ISR is effective against infection by pathogens with different life cycles. In this model, when plants are pretreated with AR156, ISR is activated by equipping plants with a potentiated immune status that is omnipotent to pathogens with both biotrophic and necrotrophic life styles. When plants in potentiated status are infected by a biotrophic pathogen, such as Pst DC3000, both SA and JA/ET signaling pathways are activated. Through a mechanism dependent on NPR1, downstream defense-related genes, such as PR1, PR2, PR5, and PDF1.2 are expressed, and cellular defense responses, such as H2O<sup>2</sup> accumulation, callose deposition are activated; when plants are challenged with necrotrophic pathogens, such as B. cinerea, only the JA/ET signaling pathways is activated. Once again through a NPR1-dependent mechanism, downstream defense responses are activated, leading to increase resistance (**Figure 6**). However, whether the role of NPR1 is conserved between the SA and JA/ET signaling pathways is unclear to us. Further study is needed to clarify the versatile function of NPR1 in AR156-mediated ISR, which is of great significance in promoting the application of AR156 in crops protection.

### AUTHOR CONTRIBUTIONS

DN and HZ designed the study. PN, XL, and SW performed the experiments. All authors analyzed the data. DN and HZ wrote the manuscript. All authors contributed to the research and approved the final version of the manuscript.

### ACKNOWLEDGMENTS

We thank Profs. Zhou J.M. (National Institute of Biological Sciences, Beijing) and Qi Y.J. (Tsinghua University, Beijing) for kindly providing us with the seeds of Arabidopsis mutants, Prof. Dong H.S. (Nanjing Agricultural University, Nanjing) for Arabidopsis transgenic line NahG, Prof. Wei L.H (Jiangsu Academy of Agricultural Sciences, Nanjing) for Botrytis cinerea strain B1301. This work was supported by the National Natural Science Foundation of China (31501621) to Niu N. and a Fundamental Research Funds for the Central Universities (KYTZ201403) and a PhD Programs Foundation of Ministry of Education of China (B0201300664) to HZ.

### REFERENCES


### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017. 00238/full#supplementary-material


septate endophyte Harpophora oryzae to rice blast disease. PLoS ONE 8:e61332. doi: 10.1371/journal.pone.0061332


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Nie, Li, Wang, Guo, Zhao and Niu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Thyme Oil Reduces Biofilm Formation and Impairs Virulence of Xanthomonas oryzae

Akanksha Singh<sup>1</sup> , Rupali Gupta<sup>1</sup> , Sudeep Tandon<sup>2</sup> and Rakesh Pandey<sup>1</sup> \*

<sup>1</sup> Department of Microbial Technology and Nematology, Central Institute of Medicinal and Aromatic Plants, Council of Scientific and Industrial Research, Lucknow, India, <sup>2</sup> Chemical Processing Department, Central Institute of Medicinal and Aromatic Plants, Council of Scientific and Industrial Research, Lucknow, India

#### Edited by:

Gero Benckiser, Justus-Liebig-Universität Gießen, Germany

#### Reviewed by:

Adam Schikora, Julius Kühn-Institut, Germany Giovanni Di Bonaventura, Università degli Studi "G. d'Annunzio" Chieti-Pescara, Italy

> \*Correspondence: Rakesh Pandey r.pandey@cimap.res.in

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Microbiology

Received: 15 November 2016 Accepted: 29 May 2017 Published: 13 June 2017

#### Citation:

Singh A, Gupta R, Tandon S and Pandey R (2017) Thyme Oil Reduces Biofilm Formation and Impairs Virulence of Xanthomonas oryzae. Front. Microbiol. 8:1074. doi: 10.3389/fmicb.2017.01074 Xanthomonas oryzae pv. oryzae (Xoo), a common bacterial plant pathogen regulates its virulence and biofilm formation attribute via a chemical method of communication. Disabling this mechanism offers a promising alternative to reduce the virulence and pathogencity of the microorganism. In this study, the effect of thyme (THY) oil on Quorum Sensing mediated synthesis of various virulence factors and biofilm formation was analyzed. Treatment of Xoo with 500 ppm THY oil displayed a significant diminution in swimming, swarming, exopolysaccharide and xanthomonadin secretion. However, no effect was observed on bacterial growth kinetics and metabolic activity of the cells. Results were further authenticated by RT-qPCR as significant reduction in motA, motB, and flgE genes was observed upon THY oil treatment. Similarly, the expression of some extracellular enzyme genes such as endoglucanase, xylanase, cellobiosidase, and polygalacturonase was also found to be significantly reduced. However, biochemical plate assays revealed insignificant effect of 500 ppm THY oil on secretion of protease, cellulase, and lipase enzymes. The rpfF gene known to play a crucial role in the virulence of the phytopathogenic bacteria was also significantly reduced in the THY oil treated Xoo cells. HPTLC analysis further revealed significant reduction in DSF and BDSF signaling molecules when Xoo cells were treated with 500 ppm THY oil. Disease reduction was observed in in vitro agar plate assay as lesion length was reduced in THY oil treated Xoo cells when compared with the alone treatment. GC–MS result revealed thymol as the active and major component of THY oil which showed potential binding with rpfF gene. Application of 75 µM thymol resulted in downregulation of gumC, motA, estA, virulence acvB and pglA along with rpfF. The other genes such as cheD, flgA, cheY, and pilA, were not found to be significantly affected. Overall, the results clearly indicated THY oil and its active component Thymol to be a potential candidate for the development of anti-virulence agent which in future when applied in combination with conventional bactericides might not only help in lowering the dose of bactericides but also be successful in curbing the disease progression in rice.

Keywords: biofilm, exopolysaccharide, thyme oil, virulence factors, quorum sensing

## INTRODUCTION

fmicb-08-01074 June 9, 2017 Time: 15:47 # 2

Xanthomonas oryzae pv. oryzae (Xoo), a Gram-negative bacterium, causes bacterial blight, one of the most disastrous diseases of rice know to cause an annual yield losses up to 60% (Niño-Liu et al., 2006). It enters in plant tissue either through wounds or hydathodes and travels to the xylem vessels where it actively multiplies resulting in blight disease on rice leaves. Alternatives for disease control and management strategies are very limited and mainly dependent on the resistant cultivars and usage of antibiotics and chemicals (Mansfield et al., 2012). Given that there are limited effective antibacterial compounds for controlling Xoo, search to develop cost effective and novel strategies that have minimal environmental impact is the need of hour. One such new wrinkle is the use of anti-virulence or anti-quorum sensing (QS) agents that do not kill the bacteria but restrains the production of disease triggering virulence factors (Tay and Yew, 2013).

Like other bacterial genera production of an array of virulence factors such as extracellular polysaccharides (EPS) and various enzymes is under the control of cell-to-cell conversation system (QS). Xoo is known to produce signaling molecules by an intricate regulatory procedure relying on the concentration of QS molecules (Karatan and Watnick, 2009). The process is customarily engaged in governing genes associated with biofilm maturation, motility, competence, and virulence factors production (Xu et al., 2006; Hegde et al., 2009). QS is reported to exist in different bacterial species, with an assortment of diverse signaling molecules like N-acylhomoserine lactones (AHLs), diffusible signal factors (DSF) family signals, oligopeptides and autoinducers-2 (AI-2) (Chen et al., 2002; Barrios et al., 2006; Boon et al., 2008; Deng et al., 2010; Monnet and Gardan, 2015).

Among the different signaling molecules, DSF family signals produced by an array of bacterial pathogens such as Xylella fastidiosa, X. campestris pv. campestris (Xcc), Stenotrophomonas maltophilia including Xoo (Deng et al., 2014; Ryan et al., 2015) permit sensing their fellow natives and thereby coordinate individual task to form resistant biofilm. Key players involved in Xoo regulating the QS pathway include DSF, an intermediate chain fatty acid 'cis-11-methyl-2-dodecenoic acid' and BDSF (cis-2-dodecenoic acid). Basically, coordination and control of QS in Xanthomonads is carried out by the rpf gene cluster (He et al., 2010; Ryan and Dow, 2011).

Among the various factors responsible for eliciting disease symptoms, biofilm and cell degrading enzymes play an important role. Multiple studies have reported that QS-deficient mutants formed thinner and more disorganized biofilms compared to the wild-types (Tomlin et al., 2005). Recently, a study by Li and Wang (2014) reported effective restrainement of citrus canker symptoms caused by X. citri subsp. citri on upon application of foliar-applied biofilm inhibitors D-leucine and 3-indolylacetonitrile (IAN). Similarly, several essential oils were also found to inhibit biofilm formation capacity of Pseudomonas aeruginosa and Staphylococcus aureus (Kavanaugh and Ribbeck, 2012). Keeping in view these findings, we hypothesized that Thymus vulgaris plant oil (THY oil) might provide effective control against blight disease by acting as QS antagonists. To test the hypothesis, the present investigation was thus carried out with the aim to identify the role of THY oil on virulence traits such as biofilm formation, extracellular enzymes production, signaling molecules and consequently its effect on disease development when inoculated on rice plants along with Xoo.

### MATERIALS AND METHODS

### Xanthomonas Strain and Growth Conditions

The X. oryzae pv. oryzae (Xoo) strain AS29 was isolated from susceptible rice cultivar BLB Pusa Basmati 1 (India) and further identified using 16S-rDNA gene sequence. The identified strain was submitted to GenBank database (Accession Number KX010418) and routinely grown on PME (Peptone Malt Extract) (HiMedia, India) plates (0.5% peptone, 3% malt extract, 1.5% bacto agar) at 26 ± 2 ◦C.

### Spectrophotometric Biofilm Assay

The protocol for measuring biofilm production was carried out following the method described by Dunger et al. (2014). Overnight grown XooAS29 culture was inoculated into 2 mL PME broth with 1:1000 dilution so that OD<sup>600</sup> reached 0.6. To the 96-well U-plastic titre plate (SPL Life Sciences, Co., Ltd; South Korea), 1 µL culture was added to 99 µL PME broth in each well. For evaluating the consequence of THY oil, 100–1000 ppm were initially used to assess their effect on biofilm formation. The titre plate was incubated without shaking at 28◦C for 10 h. The incubated cells were then stained with crystal violet (CV) dye for 20 min. The unbound dye was removed by rinsing twice with sterile distilled water (SDW). Finally, the well-bound dye was solubilized in 200 µL dimethyl sulfoxide (100% DMSO) and quantified spectrophotometrically by recording absorbance at 595 nm.

### Microscopic Analysis of Biofilm

Since, 500 ppm oil was most efficient in the aforementioned method; the outcome of this concentration on biofilm (microscopic visualization) in contrast to XooAS29 alone (control) was measured using polyvinyl chloride (PVC) assay (Krzy´sciak et al., 2014). The XooAS29 culture (OD<sup>600</sup> = 0.6) was supplemented into 1 mL of PME broth consisting of 500 ppm THY oil for 14 h at 28◦C. The PVC plate wells were washed twice with SDW to eliminate the buoyant planktonic cells after the incubation time. The bacterial cells adhering to the surface were stained with 0.5% CV and 0.2% methylene blue (MB) dyes, respectively and incubated for 5 min at room temperature. The surplus dye was eliminated by pipetting out and the plates were again kept for air drying at room temperature. The dried plates were further imaged at a magnification of 20X (Olympus BH2, Japan) under a light microscope. The XooAS29 culture was allowed to grow on the cover glass submerged in PME medium in the absence and presence of specific concentration of oil and incubated for 10–12 h. The biofilm formed were further stained with green fluorescent dye 20 mM SYTO-9 and observed under

fluorescent microscope (Leica DMR epifluorescence microscope, Germany).

### Metabolic Activity and Growth Kinetics

In the metabolic activity assay, a stock solution of 0.002% (w/v) resazurin reagent (Sigma, United Kingdom; filter sterilized) was prepared (Mariscal et al., 2009) and stored at −20◦C. XooAS29 culture was grown with and without 500 ppm THY oil in PME broth and incubated for 24 h. The culture was centrifuged and the cells were washed with saline. Individually XooAS29 (100 µL) was added to each well of 96 well titre plate followed by addition of 20 µL resazurin solution. The resazurin fluorescence activity (λex: 560 nm and λem: 590 nm) was measured after 60 min of incubation at 37◦C. To test the growth inhibitory effect of THY oil, XooAS29 was grown in PME broth with and without effective concentrations of THY oil (500 ppm) using a shaking incubator for 8–13 h at 25 ± 2 ◦C and the absorbance was recorded at 600 nm in triplicate samples collected hourly. Xoo cells treated with oil served as treatment while cells grown in the absence of oil was taken as control.

### Swimming, Swarming, and Wetness Assay

XooAS29 inoculum was grown as mentioned above and prior to carrying out the assay the final density was adjusted to 1 × 10<sup>8</sup> CFU mL−<sup>1</sup> . For conducting the motility assays, XooAS29 cells mixed with 500 ppm THY oil were gently inoculated in the center of the solidified PME agar plates. 1 µL suspension droplets were carefully placed in the center of soft swimming plates (3% malt extract, 0.5% peptone, 0.3% agar), and (2) swarm plates (3% malt extract, 0.5% peptone, 0.5% agar), and kept for incubation at 25 ± 2 ◦C. The swimming and swarming traits were evaluated after 24 h by calculating the diameter of the bacterial zone. An equivalent quantity of only XooAS29 inoculum without oil in the respective media served as control.

The wetness assay of the colonies on swarm plates was calculated by the capillary-drop method as delineated by Wang et al. (2005). Capillary tubes of 0.5 mL capacity were mildly positioned on the treatment and control colonies for 30 s and the length of fluid entered was measured in millimeters. Xoo cells treated with oil served as treatment while cells grown in the absence of oil was taken as control. Five tubes were placed on different points on swarm plates for calculating the average and SD values.

### Quantification of Exopolysaccharides (EPS) and Xanthomonadin

For EPS, XooAS29 culture with and without 500 ppm THY oil was grown on PME media for 24 h at 28 ± 2 ◦C. To detect the production of EPS from bacterial cells, the cells were scraped off the plates and resuspended in 15 mL SDW. Centrifugation at 15,000 × g for 12 min was done and the supernatant was collected for further analysis. The collected supernatant was muddled up with 1% (w/v) potassium chloride and double volume of 100% ethanol and further incubated at −20◦C overnight. Again centrifugation was done and the precipitated EPS was left overnight for drying at 55◦C before determining the total dry weight. Expression of the results was done relative to the cell density.

Measurement of xanthomonadin pigment was done as mentioned in the method illustrated by Wang et al. (2015). The XooAS29 cells collected by centrifuging 4 mL broth suspension with and without 500 ppm THY oil was mixed with 1 mL 100% methanol. The concoctions were further incubated in darkness for 10 min kept on rotating shaker followed by centrifugation at 12,000 × g for 8 min to collect the supernatant. The xanthomonadin pigment was estimated by measuring the absorbance at OD<sup>445</sup> and the result was denoted relative to the cell density measured before the assay (OD595).

### Extracellular Various Enzymatic Assays

The fresh colony of XooAS29 strain was grown in 10 mL of PME liquid medium in presence and absence of 500 ppm THY oil at a starting OD<sup>600</sup> of 0.05. After incubation for 24 h, the XooAS29 culture at an OD<sup>600</sup> of 1.8 was centrifuged at 12,000 rpm for 12 min and the supernatant obtained was used for enzymatic plate assays (Yang and Tseng, 1988). The extracellular protease activity was measured by a radial diffusion assay in agar plates containing skimmed milk as substrate with 0.5% skimmed milk and 2% (wt/vol) agar. The media was poured in the plates and wells of 6 mm diameter were cut out with the help of a cork borer. The cell suspension (with or without THY oil) was applied in the well of the plates. Clearing zones around the colony due to the utilization of the substrate were measured after 24 h of incubation at 28 ± 2 ◦C.

Extracellular cellulase activity was calculated as explained previously (Maki et al., 2011) using 2% CMC (Sigma-Aldrich) as substrate. Xylanase activity was also assayed on 1% agarose plates containing 0.5% RBB-Xylan and PME agar media (He et al., 2010). Positive xylanase activity was indicated by production of a white halo around the well in the plate.

For endoglucanase activity, XooAS29 culture with and without oil was pipetted into 6 mm diameter wells cut into CMC agar plates (1% agar, 0.125% CMC in 0.05 M potassium phosphate buffer, pH 6.0). The plates were incubated for 48 h at 28◦C and then developed with Congo red dye (Barber et al., 1997). Extracellular lipase/esterase activity was assayed on PME media containing 0.01% CaCl<sup>2</sup> and 1% Tween 80 (Rajeshwari et al., 2005). The white crystals surrounding colonies was measured in medium containing CaCl<sup>2</sup> and Tween 80 indicating positive lipase activity. All the assays were repeated three times, independently, in triplicates.

### Expression Analysis by Real-Time RT-qPCR

For RNA isolation, single celled colony of XooAS29 with and without 500 ppm THY oil was transferred to 10 mL PME medium at 28 ± 2 ◦C and sampled when the OD<sup>595</sup> reached 2.0 (Xu et al., 2015). RNA was isolated by following the manufacturer's instructions of Trizol (Invitrogen, Carlsbad, CA, United States) method and RNA purity was evaluated

with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, United States). Total RNA was used for synthesizing first-stranded cDNA using c-DNA verso reverse transcriptase (Thermo Scientific) following the instructions mentioned in the kit. The realtime RT-PCR was carried out in 96-well plates using a 7500 fast real-time PCR system (Applied Biosystems) with a Sybr Green master mix (Fermentas) following the manufacturer's instructions (Li and Wang, 2014). The gene expression was done in triplicate in a total volume of 25 µL including a passive reference dye (ROX) (Fermentas). The designing of gene specific primers (**Table 1**) was done on the basis of the available genome sequence of X. oryzae pv. oryzae strain KACC10331. The primers synthesized targeted 21 genes that were earlier recognized to be related to motility, chemotaxis, hydrolytic enzymes, virulence, and DSFs. As an endogenous control DNA gyrase subunit B encoding gene gyrB was used and the relative transcript level was calculated using the method 2−deltadelta CT described by Livak and Schmittgen, 2001.

### Quantification of QS Factors

XooAS29 culture with and without 500 ppm THY oil was individually allowed to grow as mentioned above and sampled when the OD<sup>595</sup> reached 2.0. DSF (cis-11-methyl-2-dodecenoic acid) and BDSF (cis-2-dodecenoic acid) were quantified as illustrated by He et al. (2010) and Xu et al. (2015). Five liters of XooAS29 supernatant with and without 500 ppm THY oil was individually collected by centrifugation at 9,000 rpm for

TABLE 1 | Genes used in quantitative reverse transcription polymerase chain reaction.


40 min at 4◦C. Equal volume of ethyl acetate was used for extracting the obtained supernatant whose pH was adjusted to 4.0. The removal of organic phase was done by rotary evaporator at 40◦C, and the residue obtained was dissolved in 20 mL of methanol. The final dried extract obtained was dissolved in HPLC grade methanol. The signaling molecules present in ethyl acetate extracted samples were detected through prewashed aluminum HPTLC plates (10 cm × 10 cm), coated with silica gel 60- GF254 (Merck). 2 µl of each sample was applied in triplicate on preactivated TLC plate using the Camag Linomat 5 automated TLC applicator equipped with a 100-µl Hamilton syringe. Hexane: diethylether: acetic acid (8:2:0.1) was implemented as mobile phase to develop TLC plate in Camag twin glass chamber. After developing the TLC plate to 90 mm distance, the plate was removed and dried at room temperature for 1 h. Developed plate was documented under 254 and 366 nm, and then transferred into previously saturated iodine chamber to visualize the spot under day light. The plates were then scanned at 270 nm with a Camag TLC Scanner with winCATS 3 software, using the deuterium lamp. The densitograms was recorded and peak area of both samples was observed.

### Virulence Assay on Rice Plants

The leave samples were taken from 40- to 60-day-old susceptible rice cultivar BLB Pusa Basmati 1 plants grown in greenhouse conditions. The leaves were washed twice with SDW followed by clipping tips of leaves with sterile scissors dipped in saturated XOOAS29 (10<sup>9</sup> cells/mL) grown with and without 500 ppm THY oil (Kauffman et al., 1973). As per the adopted method, approximately 10<sup>6</sup> cells are expected to be deposited at the site of pathogen inoculation. The leaves were then placed on the water agar (0.5% (wt/vol) agar) plates and left for 3 days at 28◦C. The lesion lengths were measured by comparing leaves clipped with scissors dipped in XooAS29 alone and with THY oil. The experiment was repeated three times in triplicates.

### Docking of Major Components of Oil with RpfF Protein

The crystal structure of the targeted protein, dsf synthase RpfF (PDB: 3M6M) was selected as receptor and retrieved from the RCSB protein data bank. The elimination of water molecules and selection of single peptide chain of receptor was performed using chimera software. All ligand structures were extracted from NCBI PubChem database in sdf format and converted into pdb format using online tool "Smile Converter." The graphical user interface program "Auto-Dock Tool 4.2" was used in docking simulation. The ligand energy was also optimized by defining the rotable bond, torsion angle and submerging the non-polar hydrogen bonds. Autodock tool demands the precalculated gridbox which was set at 70, 70, and 70 A<sup>0</sup> (x, y, and z) with 0.375 A<sup>0</sup> grid spacing to include all the presented amino acid residues of ligand binding pocket of receptor. Lamarckian genetic algorithm method was applied to perform docking. The default settings were used for all the other parameters. Final docked conformations were clustured using a tolerance of 2 A<sup>0</sup> RMSD and the docking log (dlg) files were analyzed. The best ligandreceptor structure from the docked structures was opted based on the binding energy and number of H-bonds formed between the target and ligand. The results were visualized using visualization tool "Discovery studio 4.5".

### Effect of Thymol on XOOAS29 Biofilm Formation, Growth Kinetics, and Expression of Various Genes

After getting the best interaction with thymol and targeted protein, screening was done to find out the effective concentration of thymol (0, 10, 25, 50, 75, 100, 200, 300, 400, and 500 µM). The selection for the effective concentration was done on the basis of the concentration which showed no growth inhibitory effect on Xoo along with inhibitory outcome on the biofilm formation. The experiment was repeated three times in triplicates.

Also, RT-qPCR was performed to quantify and compare the levels of gene expression for gumC, motA, estA, virulence gene acvB cheD, flgA, cheY, pilA, pglA along with DSF related genes rpfF, rpfC, and rpfG in Xoo when exposed to thymol. qRT-PCR was repeated twice, with four independent biological replicates each time.

### Statistical Analysis

Each experiment was done in three technical replicates and three biological replicates. Mean significant values were determined by Student's t-test using SPSS package (SPSS V16.0, SPSS, Inc., Chicago, IL, United States). Analysis of variance (ANOVA) followed by the post-test (Duncan's Multiple Comparison Test) was also used to analyze significance between more than two treatments. <sup>∗</sup>P < 0.05 and ∗∗P < 0.01 was considered statistically significant.

## RESULTS

fmicb-08-01074 June 9, 2017 Time: 15:47 # 6

### THY Oil Suppresses XooAS29 Biofilm

In the present study, it was attempted to probe whether biofilm formation was reduced in XOOAS29, when treated with THY oil or not. The Xoo strain AS29 (KX010418) was routinely used throughout the experiment. The result of micro titre plate based CV staining revealed a significant decrease in biofilm formation in a concentration dependent manner after 24 h of incubation at 28◦C in THY oil treated XooAS29 cells. As shown in **Figure 1** and Supplementary Figure S1, THY oil at concentrations of 1000 and 500 ppm significantly reduced (1.64 and 1.50 fold respectively) biofilm forming ability of XooAS29.

To further confirm the result obtained from quantitative assay microscopic visualization was done only for the lower concentration of THY oil (500 ppm) significantly inhibiting biofilm formation under light microscope. The images obtained from three different staining methods namely SYTO-9, CV and MB dye further reconfirmed the observation that biofilm production was significantly lessened when XooAS29 cells were

FIGURE 2 | Microscopic observation of specific concentration of THY oil (500 ppm) to suppresses biofilm of X. oryzae pv. oryzae (Xoo) strain AS29. Xoo biofilm was grown in the absence or presence of 500 ppm of THY oil. Only "XooAS29" served as control while "XooAS29+ THY oil" represented the treatment in the figure. (A,B) Fluorescent microscope images; (a) SYTO-9-stained light microscope, (b) CV-stained light microscope images; and (c) MB-stained light microscope (20X, scale bar = 200 µm).

treated with 500 ppm of THY oil than the alone Xoo cultures (control) (**Figure 2**).

### Effect of THY Oil on Metabolic Activity and Growth Kinetics of XooAS29

Since, the aim was to find an anti-virulence agent that only affected the secretion of virulence factors and did not kill the cells, the effect on XooAS29 on both metabolic activity and growth kinetics in the presence and absence of THY oil was investigated. The XooAS29 culture grown in the presence of both the 1000 and 500 ppm concentrations of THY oil had little/no significantly reduced Xoo metabolic activity (**Figure 3A**). Similarly, XooAS29 treated with THY oil (500 ppm) also showed no significant effect on growth of bacterial cells in comparison to the control set (**Figure 3B**).

### THY Oil Suppresses Swimming, Swarming, and Wetness Assay of XooAS29

Since, 500 ppm THY oil didn't have any bactericidal effect on the XooAS29 cells along with significant reduction in biofilm formation; we tested the ability of XooAS29 to swim in low percentage agar medium in the presence of THY oil. After incubation at 28 ± 2 ◦C for 24 h, Xoo AS29 cultured on plates with 500 ppm oil showed (P < 0.05) considerable reduction in swimming and swarming motility (1.19 and 1.24 fold, respectively) as compared to XooAS29 in the control plate. Small swimming and swarming diameters were observed in THY oil treatment which was in sharp contrast with the control sets where large diffused colony was observed on PME plates (**Figures 4a–d**).

In general, it is believed that bacterial cells need a certain amount of wetness for swarming movement on sticky surfaces. Thus, with the capillary-drop method, we evaluated the wetness of XooAS29 colonies grown on swarm plates with and without 500 ppm THY oil. As shown in Supplementary Figure S2, the wetness of the XooAS29 colony was significantly reduced (1.24 fold) in the presence of THY oil, suggesting that swarming defects by THY oil may be due to insufficient production of wetting agents or extracellular materials.

### THY Oil Reduces EPS Synthesis and Xanthomonadin Production of XooAS29

Production of EPS is a key determining factor of biofilm formation and is also correlated with the chemotactic movement in Xoo. Since the results revealed significant inhibition in biofilm formation of XooAS29 treated with 500 ppm of THY oil, we sought to extend our study by evaluating the effect of THY oil on EPS synthesis and xanthomonadin production by Xoo. The XooAS29 culture grown in the presence of 500 ppm THY oil significantly (P < 0.05) reduced (1.35 fold) EPS than the control treatment after 24 h of inoculation (**Figure 5A**). Interestingly, XooAS29 treated with 500 ppm THY oil also significantly (P < 0.05) reduced xanthomonadin which was about 1.23 fold less compared with the control set (**Figure 5B**).

### THY Oil Suppresses Extracellular Hydrolytic Enzymes Production of XooAS29

Several other virulence related phenotypes were assayed by assessing the extracellular enzyme production of XooAS29 both in the presence and absence of THY oil. As shown in **Table 2**, 500 ppm of THY oil was found to significantly reduce endogluconase and xylanase activity (1.55 and 1.94 fold, respectively). However, interestingly non-significant affect was observed in the case of protease, cellulase, and lipase activity as compared to control plate inoculated with only XooAS29.

### Differential Gene Expression of XooAS29 with THY Oil

Previously, we showed that 500 ppm THY oil showed inhibitory effect on biofilm formation, EPS, xanthomonadin, and some of the extracellular hydrolytic enzymes produced by XooAS29. To obtain insight into the pathway by which THY oil exhibited these activities, we assessed the effect of oil on expression of genes significant for motility, EPS, lytic enzymes, virulence factors and DSF signaling in XooAS29 using RT-qPCR. The selected genes included the gum genes

gumC, lytic enymes genes (xylanase, extracellular protease, endoglucanase, cellulase, 1,4-beta cellobiosidase, lytic enzyme, lipase/esterase, and polygalacturonase), DSF related biosynthetic genes rpfF, chemotaxis and motility genes cheD, cheY, motA, and motB, fimbrial protein gene pilA, flagellar genes flgA and flgE, and virulence gene acvB. The results showed that THY oil significantly down- regulated the expression of lytic enzymes genes such as extracellular protease (12%), lipase/esterase (12%), acvB (14%), chemotaxis and motility genes motB (19%), cheD (28%), cheY (11%), and gumC (18%), rpfC (38%), genes. Genes strongly suppressed were related to motility and flagellar biosynthesis genes such as motA (76%), flagellar gene flgE (60%), hydrolytic enzymes such as endoglucanase (75%), xylanase (64%), 1, 4-beta cellobiosidase (67%), and polygalacturonase (65%), along with DSF related gene rpfF (78%). Contrarily, the expression of cellulase, lytic enzyme, flgA and fimbrial protein pilA genes was not found to be significantly affected (**Figure 6**).

### THY Oil Restrains Synthesis of DSF and BDSF Using HPTLC

In Xoo, DSF and BDSF are key signaling molecules controlling virulence in response to cell density. The inhibitory effect of 500 ppm THY on production of both the molecules was also confirmed by HPTLC analysis (**Figure 7**). Separation of DSF and BDSF was attained using a mobile phase comprising of hexane–ethylether–acetic acid (8: 2: 1). Photography of the TLC plate was done at λmax 254 and 366 nm wavelengths in UV absorbance mode using the HPTLC system (CAMAG, Switzerland). The corresponding bands of DSF and BDSF were observed at Rf 5.9 and Rf 7.8, respectively.

### Virulence Assay

In order to gain insight into whether the decrement in the above mentioned genes and traits impaired XooAS29 virulence and disease inciting potential, the virulence assay by leaf clip method was conducted. The results showed that the effective concentration of 500 ppm THY oil enervated the virulence potential of XooAS29 in comparison to the control leaves infected with only Xoo 3 days after pathogen inoculation (**Figure 8**). Water agar plates having rice leaves revealed significant decrement in lesion length (1.57 fold) in THY oil treated rice leaves as compared to only Xoo treated control leaves.

### Gas Chromatographic Identification of the Individual Components of THY Oil

After THY oil was found to be potential in curtailing the pathogencity of XooAS29, our next step was to identify the individual components present in this oil. For this, THY oil was subjected to GC-FID apparatus. Thymol, gamma-terpinene, and para-cymene were found as major components of THY oil being 47.88, 29.61, and 20.15%, respectively. A representative image of the chromatogram is shown in Supplementary Figure S3.

FIGURE 5 | Specific concentration of THY oil (500 ppm) suppresses exopolysaccharide (A) and xanthomonadin activity (B) of X. oryzae pv. oryzae (Xoo) strain AS29. Only "Xoo" served as control while "Xoo+ THY" represented the treatment in the figure. Results are means of three technical replicates and three biological replicates and error bar indicates the standard error. Asterisk indicates <sup>∗</sup>P < 0.05.

TABLE 2 | Effect of THY oil on different hydrolytic enzymes such as protease, cellulase, endogluconase, xylanase, and lipase of X. oryzae pv. oryzae (Xoo) strain AS29.


Results are means of three technical replicates and three biological replicates and error bar indicates the standard error. Asterisk indicates <sup>∗</sup>P < 0.05.

### Rpf F Inhibition by Individual Components of THY Oil Confirmed by Molecular Docking

After identification of the individual components of THY oil as thymol, gamma-terpinene, and para-cymene, we expanded our finding toward to identification of the active component responsible for inhibition of signaling molecules using molecular docking. Silencing of the DSF synthase RpfF protein in X. oryzae pv. oryzae, could be a potential target protein to curb the infection. To study the possible inhibitory action of individual components of THY oil on DSF synthase RpfF protein, we used information available from the RpfF (PDB: 3M6M) to model the three-dimensional (3D) structure of the RpfF protein catalytic domain. Docking of thymol, gamma-terpinene, and para-cymene in the putative RpfF binding pocket indicated that these compounds formed potential hydrogen bonds with some residues of catalytic importance (**Figure 9** and Supplementary Figure S4). However, this structural model revealed a substantial region of interaction between thymol with RpfF protein, a positive regulator of DSF synthesis. The calculated binding energy of thymol, gamma-terpinene, and para-cymene was found to be −6.94, −5.99, and −5.79 kcal/mol, respectively.

### Effect of Thymol on XooAS29 Biofilm Formation, Growth Kinetics, and Selected Gene Expression

XooAS29 biofilm formation with thymol was evaluated using a static biofilm assay. Biofilm formation was reduced by thymol in a concentration dependent manner as shown in **Figure 10A**. The effect of thymol on XooAS29 growth was also evaluated by monitoring the OD of cells cultures at 600 nm (**Figure 10B**). The pattern of XooAS29 growth was not significantly different between the cultures with 0 and 75 µM thymol during the lag, exponential and stationary growth phases (**Figure 10B**).

Previously, we showed that 75 µM thymol inhibited XooAS29 biofilm formation without effective cell growth. As shown in supplementary data, the effect of thymol was evaluated on the expression of various genes responsible for motility, virulence factors and DSF signaling in XooAS29 using RT-qPCR. The results showed that thymol (75 µM) significantly down-regulated the expression of gumC (30%), motA (17%), estA (22%), virulence gene acvB (12%), and pglA (54%)along with rpfF (68%) while other genes such ascheD, flgA, cheY, and pilA, rpfG, and rpfC were not significantly affected by thymol (**Figure 11**).

### DISCUSSION

Pathogenesis of many agriculturally important pathogenic bacteria is under the control of intercellular communication called QS. In QS based cell to cell conversation, bacteria senses change in density of its counterparts in reaction to extracellular signal molecules called autoinducers (AIs) and thereby modifies their gene expression. When a certain threshold level of AIs is accumulated, the bacteria set off a range of biological processes such as motility, hydrolytic enzymes and EPS biosynthesis (von Bodman et al., 2003; Singh et al., 2012). In the present investigation, we hypothesized that the aromatic oils used in traditional medicine system might contain QS blockers. Thus, effect of thyme essential oil isolated from the fresh leaves of T. vulgaris on biofilm formation potential and virulence factors produced by Xoo was examined. Biofilm formation and adherence to solid surface have been previously confederated with the virulence of many plant pathogenic bacteria like Xoo, X. axonopodis pv. citri (Xac) and others (Danhorn and Fuqua, 2007; Lim et al., 2008; Mccarthy et al., 2008; Gottig et al., 2009). However, despite of the significance of biofilm formation in bacterial pathogencity, strategies for developing an antimicrobial therapy are very poorly investigated especially in the field of agriculture. The results revealed that at 500 ppm, THY oil significantly reduced the biofilm formation which

spectra; (B) UV at 254 and 366 nm and (C) Quantification of DSF and BDSF molecules. Control treatment having only Xoo cells is denoted by "C" while THY oil treated Xoo cells are refereed as "T". Results are means of three technical replicates and three biological replicates and error bar indicates the standard error.

FIGURE 8 | Thyme oil effects on rice disease responses to the representative XooAS29. (a) Leaves were inoculated with XooAS29 in the absence or presence of 500 ppm of THY oil using the leaf-clipping method. Only "Xoo" served as control while "Xoo+ THY" represented the treatment in the figure. Photographs were taken at 3 days after pathogen inoculation. (b) Lesion development was examined by measuring lesion length in XooAS29 in the absence or presence of 500 ppm of THY oil. Results are means of three technical replicates and three biological replicates and error bar indicates the standard error. Asterisk indicates <sup>∗</sup>P < 0.05. FIGURE 9 | Molecular docking of the interaction between thymol and rpfF. (A) 3D structure of thymol. (B) Both thymol and ligand contact residues are represented

corroborated well with the reduced EPS and xanthomonadin secretion. Previous investigations have revealed the role of EPS in bacterial pathogenesis as loss of EPS was found to be positively related with loss of virulence (Sutton and Williams, 1970; Dharmapuri and Sonti, 1999). Likewise, Xanthomonadins apart from being the peculiar attribute of the genus safeguards

and (C) binding orientation of thymol in RpfF. The protein is depicted as a ribbon, and secondary structures (i.e., helix, strand, and loop) are shown.

(P < 0.01; Duncan's multiple comparison test).

the pathogen from its own defense mechanism, helps in epiphytic continuance and aids in host infection (Goel et al., 2002; He et al., 2011). In most of the bacteria, EPS production is an important deciding factor for formation of biofilm and hence, an inhibitor of motility (Dow et al., 2003; Jeong et al., 2008; Rosenberg et al., 2013). Interestingly, alleviation in swimming and swarming potential of XOOAS29 by THY oil may thus be possible by the antagonizing action of oil against QS mediated signaling. In a recent study conducted both swimming and EPS production in strain XKK12 were enhanced by exogenous application of DSF thus substantiating the significance of DSF signaling in the processes (Xu et al., 2015). Further, the observations from the gene expression experiment indicated THY oil to have a discrete mechanism responsible for reducing biofilm as expression of a number of biofilm formation related genes pertaining to chemotaxis and motility such as motA, motB, and flgE were significantly decreased. However, at this stage possibility of reduction by other mechanism cannot be ruled out. Interestingly, in a study two compounds D-leucine and 3-indolylacetonitrile (IAN) were found to prevent biofilm formation by X. citri subsp. citri on host leaves and various abiotic surfaces at a concentration lower than the minimum inhibitory concentration (MIC) (Li and Wang, 2014). Likewise, in another study exogenous addition of DSF restored EPS production in the Xoo rpfF knockout mutant indicating a definitive role of DSF in regulating EPS production in Xoo (Jeong et al., 2008; He et al., 2010).

The ability of Xoo to incite the disease depends on other pathways too for controlling the pathogenic web like type II secretion system (T2SS). This system apart from regulating the secretion of EPS also controls the activity of extracellular enzymes such as protease, pectinase, endoglucanase, polygalacturonate lyase, amylase, etc. (Wei et al., 2007; He et al., 2009). Reduced expression of endoglucanase, xylanase, cellobiosidase, and polygalacturonase in our investigation again points out

time polymerase chain reaction (qRT-PCR) analysis. Only "Xoo" served as control while "Xoo+ THY" represented the treatment in the figure. gyrB was used as an endogenous control. Results are means of three technical replicates and three biological replicates and error bar indicates the standard error. Asterisk indicates ∗∗P < 0.01 and <sup>∗</sup>P < 0.05.

the role of THY oil in targeting the virulence arsenal of Xoo. Similar results have been previously reported by Jeong et al. (2008) who showed reduced expression of genes associated with motility, EPS as well as exoenzymes in rpfB, rpfC, rpfF, and rpfG mutants. However, contradictory observations have been recorded by Chatterjee and Sonti (2002) who reported normal EPS and xylanase levels in rpfF mutant. The variability in observations recorded by various groups could possibly be the result of different culture media or Xoo strains used in the studies. In the present investigation, control of rpf genes by some unknown regulatory component cannot be ruled out as decrement in rpfC gene was observed when thyme oil having different active components was applied to Xoo cells. However, rpfC gene expression remained insignificant when Xoo cells were treated with only thymol.

Since the results obtained highlighted the interruption of DSF signaling pathway, we further cross confirmed the results by analyzing the expression of genes involved in synthesis of DSF. The rpfF gene that controls the production of DSF, a signaling molecule regulating the synthesis of EPS and exoenzymes was found to be significantly reduced. Positive effect of rpfF was further verified as X. axonopodis pv. glycines rpfF mutant was observed to produce less CMCase, protease, endo-β-1,4 mannanase, polygalacturonate lyase and pectolytic activity than their wild-type counterparts (Thowthampitak et al., 2008).

HPTLC results further validated the gene expression results as reduced levels of DSF was observed in THY oil treated Xoo as compared to the untreated Xoo cells. In addition to change in the level of DSF QS signaling molecule, our results also indicated a role of THY oil in alleviating the level of BDSF which are reported to be accountable for EPS synthesis (He et al., 2010). Thus, decreased synthesis of EPS by THY oil may infer some part from its positive outcome from DSF and BDSF molecule synthesis. Overall since the results favored the hypothesis of reduction in virulence traits of XOOAS29 by application of THY oil, detached leaf assay for observing the lesion length caused by XOOAS29 in presence and absence of oil was conducted. The oil was found to significantly depreciate the lesion length probably by reducing the biofilm formation and other virulence traits expression. Similar observations were recorded by Li and Wang (2014) as reduced number of lesions was observed on leaves of

### REFERENCES


grapefruit by foliar spray of D- leucine and IAN which impaired the infection causing capacity of X. citri subsp. citri by repressing biofilm formation.

Overall, the results suggest modulation of DSF QS circuits of XooAS29 which probably led to reduced biofilm formation, EPS, virulence and weakened infection (**Figure 12**). The results thus offer an attractive lead for the development of potent alternative for the development of anti-virulence alternative drugs for managing bacterial plant pathogens in agricultural fields. Further, bioinformatics study and wet laboratory experiments have revealed thymol to be a possible player as an anti-virulence agent since significant reduction in biofilm and genes pertaining to motility, chemotaxis, and virulence traits were observed. However, at this stage detailed gene level and knock out studies need to be done to identify the exact mode of action of both thyme oil and its active component thymol.

### AUTHOR CONTRIBUTIONS

AS, RP: conceived and designed the experiments, manuscript writing. AS: performed and conceptualization of the experiments. RG: helped in Real time analysis. ST: GC analysis. RP: critical revision of the manuscript. All authors read and approved the final manuscript.

### ACKNOWLEDGMENTS

The authors are thankful to the Director, CSIR-CIMAP, for his encouragement during course of investigation. AS is grateful to Department of Science and Technology, Government of India for providing fellowship under DST Women Scientist [SR/WOS-A/LS-176/ 2014 (G)].

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2017.01074/full#supplementary-material


polysaccharide production and virulence. FEMS Microbiol. Lett. 179, 53–59. doi: 10.1111/j.1574-6968.1999.tb08707.x


Yang, B. Y., and Tseng, Y. H. (1988). Production of exopolysaccharide and levels of protease and pectinase activity in pathogenic and non-pathogenic strains of Xanthomonas campestris pv. campestris. Bot. Bull. Acad. Sin. 29, 93–99.

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Singh, Gupta, Tandon and Pandey. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Production of ergothioneine by *Methylobacterium* species

#### Kabir M. Alamgir <sup>1</sup> , Sachiko Masuda1, 2, Yoshiko Fujitani <sup>1</sup> , Fumio Fukuda<sup>3</sup> and Akio Tani <sup>1</sup> \*

<sup>1</sup> Group of Plant-Microbe Interactions, Institute of Plant Science and Resources, Okayama University, Okayama, Japan, <sup>2</sup> Advanced Low Carbon Technology Research and Development Program, Japan Science and Technology Agency, Tokyo, Japan, <sup>3</sup> Laboratory of Pomology, Graduate School of Environmental and Life Science, Okayama University, Okayama, Japan

Metabolomic analysis revealed that Methylobacterium cells accumulate a large amount of ergothioneine (EGT), which is a sulfur-containing, non-proteinogenic, antioxidative amino acid derived from histidine. EGT biosynthesis and its role in methylotrophy and physiology for plant surface-symbiotic Methylobacterium species were investigated in this study. Almost all Methylobacterium type strains can synthesize EGT. We selected one of the most productive strains (M. aquaticum strain 22A isolated from a moss), and investigated the feasibility of fermentative EGT production through optimization of the culture condition. Methanol as a carbon source served as the best substrate for production. The productivity reached up to 1000µg/100 ml culture (1200µg/g wet weight cells, 6.3 mg/g dry weight) in 38 days. Next, we identified the genes (egtBD) responsible for EGT synthesis, and generated a deletion mutant defective in EGT production. Compared to the wild type, the mutant showed better growth on methanol and on the plant surface as well as severe susceptibility to heat treatment and irradiation of ultraviolet (UV) and sunlight. These results suggested that EGT is not involved in methylotrophy, but is involved in their phyllospheric lifestyle fitness of the genus in natural conditions.

Keywords: ergothioneine, methylotroph, methanol, *Methylobacterium* species, glutathione, reactive oxygen species, antioxidant

### INTRODUCTION

Methylobacterium species are facultative methylotrophic bacteria that can use both methanol and multi-carbon substances. They are ubiquitous in nature, and metagenomic studies on some plant species (Knief et al., 2011) and rice (Delmotte et al., 2009) showed their predomination on plant surfaces (the phyllosphere). This is believed to occur because plants emit methanol as a by-product of pectin demethylation (Fall and Benson, 1996; Jourand et al., 2005). Since global leaf area is estimated to be ca. 6.4 × 10<sup>8</sup> km<sup>2</sup> (Morris and Kinkel, 2002) and global emission of plant methanol is estimated to be 10<sup>14</sup> grams per year (Guenther et al., 1995), the interaction between plants and Methylobacterium species is of importance to consider plant health, agriculture, and the global cycle of one-carbon compounds as well. In addition to the abundant and extensive study on their methylotrophic metabolism (Vuilleumier et al., 2009), their physiology in the phyllosphere has been gathering a great deal of attention (Gourion et al., 2006), since the phyllosphere is considered to be a harsh environment, in which microorganisms are exposed to UV, temperature shifts, fluctuations in water availability, and limited resources for growth (Vorholt, 2012). Methylobacterium species have been reported to be able to promote plant growth

### *Edited by:*

Gero Benckiser, Retired

#### *Reviewed by:*

Claudia Knief, University of Bonn, Germany Raffaella Balestrini, Consiglio Nazionale delle Ricerche, Italy Hendrik Schäfer, University of Warwick, UK

> *\*Correspondence:* Akio Tani

atani@okayama-u.ac.jp

#### *Specialty section:*

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

*Received:* 18 March 2015 *Accepted:* 12 October 2015 *Published:* 27 October 2015

#### *Citation:*

Alamgir KM, Masuda S, Fujitani Y, Fukuda F and Tani A (2015) Production of ergothioneine by Methylobacterium species. Front. Microbiol. 6:1185. doi: 10.3389/fmicb.2015.01185 (Abanda-Nkpwatt et al., 2006; Tani et al., 2012) due to their ability to synthesize phytohormones (Ivanova et al., 2000; Koenig et al., 2002; Schauer and Kutschera, 2011) and 1-aminocyclopropane 1-carboxylate deaminase, which decreases the ethylene level in plants (Madhaiyan et al., 2006, 2007; Chinnadurai et al., 2009).

The microbial synthesis of low-value products as well as fine chemical compounds using methylotrophs has also been documented. Methanol is a cheap, non-food feedstock that is easily generated from diverse renewable sources such as biogas and synthesis gas. These gases can be derived from methane, which is also abundant and inexpensive. Thus, methanol is a more attractive and advantageous feedstock than sugars and their polymers. In addition, developments in fermentation technology (Bourque et al., 1995; Bélanger et al., 2004) as well as the accumulated knowledge on metabolic pathways (Šmejkalová et al., 2010; Chistoserdova, 2011) have been rendering methylotrophic bacteria as attractive catalysts to synthesize fine chemicals. The production of amino acids and polyhydroxyalkanoates as well as high biomass yield from methanol has been reviewed by Schrader et al. (2009).

In our previous report, we isolated M. aquaticum strain 22A from a hydroponic culture sample of a moss, Racomitrium japonicum (Tani et al., 2012). We have been using the strain as a model for Methylobacterium-plant interaction with respect to plant-growth promotion. We found that strain 22A cells highly accumulate ergothioneine (EGT) through metabolome analysis in this study. EGT is a sulfur-containing, non-proteinogenic amino acid derived from histidine. It was first discovered in ergot fungi and its structure was determined in 1911 (Mann and Leone, 1953). The compound is believed to be synthesized in few organisms, notably actinobacteria, cyanobacteria, and certain fungi (Fahey, 2001; Pfeiffer et al., 2011). The genes for EGT synthesis were first described for Mycobacterium species (Seebeck, 2010). The clustered egtABCDE genes were shown to encode proteins that convert histidine to EGT. EgtD is a methyltransferase that converts histidine to hercynine. EgtB, an FGE (formylglycine generating enzyme) like protein, conjugates γ-glutamylcysteine to hercynine to form γ-glutamylcysteinylhercynine. EgtC, a glutamine amidotransferase, releases glutamate from it to generate S-(β-amino-β-carboxyethyl)ergothioneine sulfoxide. EgtE, pyridoxal 5-phosphate dependent β-lyase, forms EGT. EgtA (γ-glutamylcysteine synthetase) supplies γ-glutamylcysteine for EgtB. Homologous genes for egtB and egtD were found in many eukaryotes and bacteria, including Methylobacterium species. The other genes have also been detected in many eukaryotes (Jones et al., 2014). Thus, EGT synthesis in Methylobacterium species has been predicted, but its productivity and its role in methylotrophy and physiology have not hitherto been investigated.

EGT has been reported to be an important component of cells because of its antioxidant properties (Cheah and Halliwell, 2012). In humans, EGT has been shown to accumulate in various cells and tissues at high concentrations (100µM— 2 mM), although human cells do not produce EGT (Cheah and Halliwell, 2012). EGT is concentrated in mammalian mitochondria, suggesting a functional role in protecting it from mitochondrial superoxide (Paul and Snyder, 2010). EGT is known to absorb ultraviolet (UV) light, which may account for the ability to block UV damage (Bazela et al., 2014). Importantly, in mammalian cells, the EGT transporter ETT/OCTN1 was identified (Gründemann et al., 2005; Grigat et al., 2007), suggesting that EGT is preferentially acquired by the cells. Certain species of mushrooms are distinguished sources of EGT, ranging from 0.4 to 2.0 mg/g (Ito et al., 2011) or 0.08 to 3.78 mg/g (Dubost et al., 2006) in dry weight. In Mycobacterium smegmatis, intracellular and extracellular EGT of 4.1 and 17 pg/10<sup>5</sup> colony-forming units (CFUs), respectively, was reported (Sao Emani et al., 2013). An overproduction system from the fission yeast Schizosaccharomyces pombe was constructed, resulting in 5000-fold higher productivity of 1600µM in intracellular concentration (Pluskal et al., 2014). In cyanobacteria, 0.8 mg/g dry mass was reported (Pfeiffer et al., 2011).

In this study, we revealed the abundant EGT accumulation in Methylobacterium species for the first time and investigated the feasibility of fermentative EGT production from methanol using M. aquaticum strain 22A. Furthermore, we report the phenotype of a mutant deficient for EGT production, and suggest its important role in the phyllospheric lifestyle of the genus.

### MATERIALS AND METHODS

### Microbial Strain and Culture Condition

Mineral medium (MM) composition is presented in Table S1. Different concentrations of methanol, succinate, glucose, or ethanol were used as carbon sources. R2A, Luria-Bertani (LB), and Middlebrook 7H9 (a mineral medium for Mycobacterium species containing sodium citrate and glutamic acid as carbon sources, on which Methylobacterium species grow well when methanol is supplemented) (Middlebrook and Cohn, 1958) were also tested. Kanamycin (25µg/mL) was used when necessary. Methylobacterium species were cultivated at 28◦C. Escherichia coli strains were grown in LB at 37◦C.

A rifampicin-resistant spontaneous mutant of strain 22A (22A-rif) was obtained by streaking wild type on R2A agar containing 2µg/ml rifampicin; its growth on methanol was confirmed to be normal. The frequency of spontaneous mutation was estimated to be 10−<sup>7</sup> . mTn5gusA-pgfp22 (Xi et al., 1999) was used to derive the GFP-expressing kanamycin-resistant strain.

### Metabolome Analysis with Capillary Electrophoresis Time-of-Flight Mass Spectrometry (CE-TOF/MS)

The cells of strain 22A (1.45 × 10<sup>9</sup> cells) grown in liquid MM containing 0.5% methanol for 3 days were harvested, filter-trapped with a 0.4µm pore membrane (millipore), and rinsed with water. The cells were transferred into 2 ml of methanol containing 5µM methionine sulfone and camphor-10-sulfonic acid. The samples were sent to Human Metabolome Technologies Inc. (HMT, Tsuruoka, Japan). A total of 1.6 ml of chloroform and 640µl of water were added to the sample. The sample was then vortexed and centrifuged at 2300 × g, at 4◦C for 5 min. The water layer was taken and filtered by an Ultrafree-MC, 5 kDa (molecular weight) cut-off centrifugal filter device (HMT) to remove proteins. The filtrate was dried, dissolved in 25µl of ultra-pure water, and then analyzed using CE-TOF/MS equipped with an Agilent 6210 TOF/MS (Agilent Technologies, Waldbronn, Germany). Cationic metabolites were analyzed with a fused silica capillary (50µm inner diameter × 80 cm total length) with a commercial cation electrophoresis buffer (Solution H3301-1001, HMT) as the electrolyte. The sample was injected at a pressure of 50 mbar for 10 s (approximately 10 nl). The applied voltage was set at 27 kV. Electrospray ionization-mass spectrometry was conducted in the positive ion mode, and the capillary voltage was set at 4 kV. The spectrometer was scanned from 50 to 1000 m/z (mass-to-charge ratio). Other conditions were the same as those in the cation analysis described previously (Soga et al., 2003). In the same way, anionic metabolites were analyzed with a commercial anion electrophoresis buffer (Solution H3302-1021, HMT). The sample was injected at a pressure of 50 mbar for 25 s (approximately 25 nl). The applied voltage was set at 30 kV. Electrospray ionization-mass spectrometry was conducted in the negative ion mode, and the capillary voltage was set at 3.5 kV. The spectrometer was scanned from 50 to 1000 m/z. Other conditions were as in the anion analysis (Soga et al., 2003).

### EGT and Glutathione Quantification

The cells of strain 22A grown in liquid media were harvested by centrifugation (12,000 × g, 25◦C, 10 min) and washed with 0.85% NaCl. The wet weight of the cells was recorded. Dry weight was obtained by complete drying of the cells at 100◦C when necessary. The intracellular EGT was extracted by heating the cell suspension in water at 94◦C for 10 min. The cell suspensions were vortexed at 1600 rpm for 30 min and centrifuged (14,000 × g, 25◦C, 10 min) to remove cells. The supernatants were filtered (0.2µm) and subjected to EGT quantification using a high performance liquid chromatography (HPLC). For extraction of glutathione, 600µl of methanol was added to the 400µl cell suspension. After mixing, 1 ml of chloroform was added. The sample was sonicated in a sonic bath for 1 min. A water layer of 1 ml was taken and subjected to HPLC analysis.

EGT was quantified by an HPLC equipped with an Asahipak NH2P-50 column (4.6 mm i.d. × 250 mm, Asahi Kasei Co.) attached with a guard column NH2P-50G, in the following conditions: injection volume, 20µl; flow rate, 0.5 ml/min; detection, UV absorbance at 254 nm. Solvent A (50 mM sodium phosphate buffer containing 0.1% triethylamine, pH: 7.3) and solvent B (100 mM NaCl) were used to make a gradient: 0–7 min, 0% B, 7–8 min to 20% B, 8–13 min 20% B, and 13–15 min to 0% B. EGT is eluted at around 6.1 min.

Glutathione was quantified with the same instrument equipped with a reverse-phase µ-BONDAPAK C18 column (4.6 mm i.d. × 250 mm, Waters Co.) attached with a guard column (Inertsil WP300 C18 GL Science). Solvent A (0.1% trifluoroacetic acid) and solvent B (100% acetonitrile) were used in the isocratic solvent system of 30% solvent A at a constant flow rate of 0.75 ml/min. Glutathione was detected by UV absorbance at 210 nm, and was eluted at around 6.1 min (reduced form) and 6.6 min (oxidized form).

### Total Amino Acid Analysis

Strain 22A culture grown in 100 ml MM containing 0.5% methanol for 7 days was separated into two fractions (10 and 90 ml). The former was used for EGT quantification as described above. The latter was used for total amino acid extraction. Cells were collected by centrifugation, and the methanol extract (2 ml) of the cells was dried in vacuo and dissolved in water (5 ml). The sample was applied to Amberlite CR1310NA (4 ml resin, pre-equilibrated with successive 1 M NaOH, water, 1 M HCl, and water). The resin was rinsed with water (50 ml), and amino acids were eluted by 2 M ammonia (40 ml). The eluent was dried in vacuo and dissolved in 5 ml water; 40µl of the sample was analyzed using an amino acid analyzer (Hitachi L-8500B). Authentic EGT could not be quantified since the analysis depends on the ninhydrin reaction in the amino group of amino acids (the amino group is trimethylated in EGT).

### Generation of EGT Mutant and Complemented Strains

Using the amino acid sequences of the egt genes of Mycobacterium species (Seebeck, 2010), egt gene homologs in the strain 22A genome (Tani et al., 2015) were found by basic local alignment search tool (BLAST) analysis. While egtA, C, and E are separately encoded in different loci, egtB and egtD are encoded in tandem in the strain 22A genome and they were subjected to deletion mutagenesis. Each kilobase of the upstream and downstream flanking regions of the egtBD genes was amplified using primers (5′ flanking sequence amplification; egtB\_left\_fw, tcgagctcggtacccatagagcaggctacgctgga and egtB\_left\_rv, catcggat cttccctcatgcg: 3′ flanking sequence amplification; egtD\_right\_fw, cgcatgagggaagatccgatcgtccaccatccggcggcactga and egtD\_right\_rv, ctctagaggatccccggtcgagctccatctccag). The fragments were tandemly cloned into the Sma I site in pK18mobSacB (Schäfer et al., 1994), using the In-Fusion cloning kit (Takara Bio Co.), yielding pK18mob-egtBD. Introduction of the plasmid into strain 22A was done via conjugation using E. coli S17-1. The kanamycin-resistant single-crossover mutant was streaked on R2A medium containing 10% sucrose to deliver doublecrossover mutants (∆egt), which were examined by diagnostic polymerase chain reaction (PCR). ∆egt(mTn5gusA-pgfp22) was generated by introducing mTn5gusA-pgfp22 into ∆egt, and its growth on methanol was confirmed to be normal.

For complementation of the mutation, the PCR fragment amplified from the wild-type genome with egtB\_left\_fw and egtD\_right\_rv, which contains functional egtBD, was cloned into pK18mobSacB. The resultant plasmid was introduced into ∆egt. Kanamycin-resistant colonies were selected for PCR diagnosis. The complemented strain was designated as ∆egtComp. The same plasmid was also introduced into the wild type, and the kanamycin-resistant strain was designated as WTegtDup.

### Phenotypic Assays for ∆*egt*

Strain 22A wild type and ∆egt were grown in MM containing 0.5% methanol to the exponential growth phase. The cells were harvested by centrifugation and washed with fresh media to make a cell suspension of OD600 = 0.5. The cell suspension was used for the following assays, as described previously (Iguchi et al., 2013) with some modifications. (1) Heat shock resistance assay. Cells were spread on R2A agar plates for CFU determination after incubation of the cell suspension (100µl) at 46◦C for 5, 10, 15, and 20 min. (2) UV resistance assay. Cell suspensions (100µl) were exposed to 253 nm UV light (GL-15, Toshiba Co.) for 2 min (distance 20 cm, equivalent to 800µW/cm<sup>2</sup> ), and then diluted and spread on R2A agar plates for CFU determination. (3) Sunlight resistance assay. Methanol-grown cells of ca. 200 CFU were spread on solid MM containing 0.5% methanol prepared in a glass dish. The plates were covered with a glass lid or wrapped with cellophane (thickness, 0.03 mm). The plates were placed on ice and exposed to sunlight at midday (done at IPSR Okayama University on Aug 18, 2015; 12:00 to 15:00; weather, sunny; air temperature, 33◦C; photon flux, 1500µmol s−<sup>1</sup> m−<sup>2</sup> , measured with a quantum sensor, UIZ-PAR-LR, Uizin Co. Japan). After 0 and 3 h of exposure the plates were transferred into an incubator at 28◦C in dark. The temperature of the plates did not exceed 26◦C during the exposure. The colonies formed on the plates were counted after 7 days cultivation. (4) Disk diffusion assay. A cell suspension in 0.75% agar was overlaid on solid MM containing 0.5% methanol and 2% agar. Aliquots (5µl) of 1 M H2O2, 2% methylglyoxal, or 1 M diamide were deposited on filter disks placed at the center of each plate. The diameter of the growth inhibition zone was measured. Student's t-test was used to evaluate the statistical significance of the differences in the bacterial counts.

### Growth of ∆*egt* on Methanol

Strain 22A and ∆egt were cultured in 96-well plates containing 200µl of MM containing methanol. The plates were incubated at 28◦C without shaking. The cell growth (OD600) was measured using a microplate reader (Powerscan HT, DS Pharma Biomedical).

### In planta Competition Experiment

Sterile seeds of Arabidopsis thaliana Col-0 were placed on 1/2 Murashige-Skoog (1/2MS) agar medium containing 3% sucrose, 0.5% (v/v) MS(5) vitamin, and 0.8% agar (Ina Food industry, Co.), pH 5.6, prepared in plastic petri dishes. The rifampicin-resistant strain 22A-rif and kanamycin-resistant ∆egt(mTn5gusA-pgfp22) were inoculated onto the seeds. In the case of single inoculation, the inoculated cell suspension was 5µl per seed, containing ca. 20,000 CFUs. For the competitive condition, a mixture of 2.5µl of each suspension was inoculated. The plants were grown at 23◦C under 16-/8-h light/dark conditions for 27 days. The shoots were excised and put in 500µl water, and homogenized with pestles. The resultant suspension was spread onto selective R2A media containing kanamycin for ∆erg (mTn5gusA-pgfp22) and rifampicin for 22A-rif, for CFU determination.

## RESULTS AND DISCUSSION

### Metabolome Analysis

The methanolic extract of methanol-grown strain 22A cells was subjected to metabolome analysis with CE-TOF/MS. We found that among the major metabolites like serine, glutamine, lysine, arginine, and glycine, EGT was detected as a highly accumulated metabolite (Table S2). However, since the compound was not included in the HMT quantifiable compound library, its quantity was not determined. We did another trial but it was unsuccessful because the internal standard was not detected for unknown reasons. However, the EGT peak area that was identifiable by migration time and mass-to-charge ratio in the trial was the highest among all detected compounds, suggesting the high accumulation of EGT (data not shown).

### Extraction and Quantification of EGT

EGT separation and quantification by HPLC was successfully established, as described in the Materials and Methods section. We used heat treatment (94◦C for 10 min) to effectively extract EGT from the cells without EGT decomposition. Additional heating of the remaining heat-treated cells also gave EGT, the amount of which was as small as 8% compared to that obtained by the first heating. We also varied the temperature of extraction (40–100◦C), and found that 60◦C was sufficient to extract most EGT (100% EGT was extracted compared to 100◦C), and 50◦C was insufficient (only 4.2% was extracted, data not shown). We used 94◦C to ensure complete extraction for the following experiments. Heating of the aqueous EGT solution at 95◦C did not result in apparent decomposition as monitored with HPLC (data not shown). Extraction with methanol resulted in a lowered EGT peak and the appearance of more other peaks in the HPLC chromatogram, suggesting that EGT was unstable in methanol. This might be the reason for the high variance of EGT concentrations detected in the metabolome analysis, where we used methanol for the extraction. We also quantified extracellular EGT in a concentrated spent medium (1-week-old 100 ml MM containing 0.5% methanol, re-dissolved in 1 ml water after completely dried, done in triplicate), but EGT content in the samples was under the detection limit (1µM, injection volume to HPLC was 20µl). Thus, extracellular EGT concentration was lower than 0.01µM, suggesting that EGT is not secreted from the cells.

The simple method of heat treatment eases EGT extraction from the cells and reduces the downstream processing cost for EGT production. Although EGT has been reported to be secreted in Mycobacterium (Sao Emani et al., 2013), it was not secreted in Methylobacterium in this study. The difference in the physiological role of extracellular and intracellular EGT in different microorganisms remains unclear.

### Abundant Accumulation of EGT in Strain 22A

In standard conditions (5 ml MM containing 0.5% methanol, 7 days, 28◦C), cell yield of strain 22A was usually approximately 30 mg (wet weight), equivalent to 6 mg dry cell weight. The EGT yield was usually 10µg (1.6 mg EGT/g dry cell weight). The cell volume of strain 22A was estimated to be 0.65µm<sup>3</sup> (cell dimension was 2.24 × 0.65µm bacilli form); therefore, the intracellular concentration of EGT was estimated to be approximately 183 mM. This intracellular concentration is much greater than the 0.3µM that was detected in fission yeast cells and the 1.6 mM in the P3nmt1-egt1<sup>+</sup> strain overexpressing the EGT synthetic gene (Pluskal et al., 2014). EGT productivity in strain 22A in this condition was comparable to that in the most productive mushroom, Pleurotus osrteatus (oyster mushroom, 3.78 mg/g dry weight) (Woldegiorgis et al., 2014).

As shown in **Figure 1**, the EGT content in the methanolgrown cells was one of the highest among amino acids quantifiable with the amino acid analyzer, again suggesting the unusually high accumulation of EGT.

### *M. aquaticum* Strain 22A is One of the Most EGT-productive Strains Among *Methylobacterium* Type Strains

We examined whether other Methylobacterium species also synthesize EGT. Methylobacterium type strains were grown on 0.5% methanol and intracellular EGT was quantified in the same condition. Through this preliminary screening, we found that most of the type strains synthesize EGT in a range of 0–100µg/g wet weight cells (Figure S1). The productivity is widely distributed in Methylobacterium species and no bias was recognized in the relationship between their productivity and a phylogenetic tree based on their 16S rRNA gene sequences (data not shown). We selected five strains, four of whose genome information was available, and quantified EGT content with higher methanol concentration (3%) in triplicate. As shown in **Figure 2**, M. oryzae DSM18207 was the most productive and strain 22A was the second best. Since the genome information of M. oryzae was not available at that time (Kwak et al., 2014), we decided to use strain 22A for further analysis.

The genes for EGT synthesis have been found in many α-proteobacteria (Seebeck, 2010) and indeed many species of Methylobacterium, Bradyrhizobium, Rhodopseudomonas, etc. harbor the genes. There might be a more productive strain in the class. Since many of them are plant-symbionts, it is tempting to speculate that the EGT found in plants (Ey et al., 2007) might be derived from such symbiotic α-proteobacteria or that plants might also have a different biosynthetic pathway for EGT.

EGT quantification after 7 days of cultivation. The experiment was done in triplicate. The data are presented as the mean ± SD (n = 3). Bar, EGT content (µg/g wet weight cell), and circles, cell yield (mg wet weight).

Moreover, it has been reported that the EGT content in the orchid Gastrodia elata is correlated with the concentration of EGT in its symbiotic fungi, Armillaria mellea (Park et al., 2010).

### Optimization of EGT Production

Strain 22A was cultured in MM containing 0.5% methanol (100 ml) containing different concentrations (0.5, 1, 2, and 3%) of methanol for 38 days and EGT productivity was monitored (**Figure 3**). Within a week, the cultures reached the stationary phase and OD600 increased only gradually afterwards (data not shown). Interestingly, the EGT amount in the culture did not increase after 7 days of cultivation on 0.5% methanol, but it increased when a higher concentration of methanol was used, whereas the OD600 did not increase. Strain 22A could grow in the presence of 3% methanol, but EGT productivity was best when 2% methanol was used. The EGT quantity in the MM containing 2% methanol reached 1000µg/100 ml culture (1.2 mg/g wet weight, 6.3 mg/g dry weight) in 38 days, which overwhelms the productivity of P. ostreatus. EGT accumulated continuously until the late stationary phase, suggesting that EGT is a kind of secondary metabolite that is not involved directly in primary metabolism. Next, we tested other carbon sources, including ethanol, succinate, and glucose, and media of LB or Middlebrook 7H9 supplemented with methanol. EGT productivity was best when methanol was used (**Figure 4**). Succinate did not induce EGT production but glucose and ethanol did, suggesting that EGT production is not only induced by methanol. This result also suggested that EGT is not necessary for growth on methanol.

We tested other nitrogen sources at concentrations equivalent to 30 mM nitrogen, and found that ammonium nitrate, ammonium chloride, and ammonium sulfate gave comparable cell yield and EGT. (Figure S2). Then we changed the

nitrogen source (ammonium chloride) concentration under 0.5% methanol, and found that 0.4–1.2 g/l was the best for production (Figure S3). Concentrations below 0.2 and above 2.0 g/l resulted in poor cell growth. Compared to carbon sources, nitrogen sources were less effective for EGT production.

Histidine and cysteine are the precursors in the EGT biosynthesis pathway (Seebeck, 2010). The supplementation of histidine, cysteine, and their combination did not enhance EGT productivity significantly (Figure S4). However, the combination at 10 mM doubled the productivity, but this concentration of amino acids is too high to be used as a supplement for largescale EGT production. It has been shown that EgtB is dependent on iron (II) for its catalysis (Seebeck, 2010). Thus, we varied the FeSO<sup>4</sup> concentration (0, 10, 17.3 [original concentration], 100, and 1000µM) in MM containing 2% methanol. EGT production was comparable among 10–100µM and decreased in 0 and 1000µM FeSO<sup>4</sup> (data not shown). Methylobacterium species usually possess two methanol dehydrogenases encoded by MxaF and XoxF; the former is calcium-dependent and the latter is lanthanide-dependent (Hibi et al., 2011; Nakagawa et al., 2012; Keltjens et al., 2014). Thus, the effect of lanthanide supplementation on EGT production was investigated under different methanol concentrations. The EGT production and growth were negatively impacted by a higher concentration of lanthanide (Figure S5), suggesting that lanthanide is not effective at enhancing production.

Thus, through the preliminary optimization of cultivation conditions, we found that 2% methanol with His and Cys supplementation at 10 mM was the best condition. Further optimization will be possible by testing other carbon sources and their mixtures, different temperatures, supplementation with other amino acids, and continuous cultivation with maintained pH and dissolved oxygen. Since EGT is considered to be a secondary metabolite, it may be necessary to mimic the condition of an aged culture to enhance the production, the factors of which are currently unknown.

### EGT Genes Found in *Methylobacterium* Species

Genes homologous to egtABCDE in Mycobacterium (Seebeck, 2010; Jones et al., 2014) were found in strain 22A genome (Tani et al., 2015) by BLAST analysis, as well as in the complete genomes in other Methylobacterium species (Table S3). The overall shared identities based on amino acid sequences were 25–46% compared to those in Mycobacterium species. EGT biosynthesis genes were present as single copies in the strain 22A genome. egtB and egtD are found to be encoded tandemly in the largest plasmid (pMaq22A-1) in the strain 22A genome, whereas egtA, C, and E are separately encoded in distant loci in the chromosome (Figure S6). It is reported that the clustered egtABCDE genes are found only in actinomycetes, and that egtB and egtD are under strong selective pressure for genetic clustering (Jones et al., 2014). The cluster is also found in the genomes of M. radiotolerans JCM2831, M. nodulans ORS2060, M. extorquens AM1, and Methylobacterium sp. 4–46, but not in those of M. oryzae CBMB20 and M. populi Bj001. It is unknown whether their EGT productivity is dependent on the gene organization. However, since EgtD and EgtB catalyze the first two steps in EGT synthesis (Seebeck, 2010), the cluster may be important for efficient EGT synthesis. The conservation of five genes in Methylobacterium genomes supported their common ability to synthesize EGT in the genus.

### EGT Is Involved in Resistance to Heat Shock, UV Radiation, H2O2, and Fitness for Growth on Plants

To clarify the function of EGT in Methylobacterium, we constructed an egtBD-deletion mutant (∆egt) via homologous recombination. Furthermore, functional egtBD were introduced into ∆egt for complementation (∆egtComp) and into the wild type as well for gene duplication (WTegtDup). EGT production in ∆egt was completely abolished, and ∆egtComp recovered EGT production with higher production than the wild type (**Figure 5**). Duplication of egtBD resulted in increased synthesis of EGT compared to the wild type. These results indicated that egtBD genes are indeed involved in EGT synthesis and that the synthetic pathway would be the same as the one for Mycobacterium

species. In addition, it was possible to enhance the production by duplicating the genes.

The mutant ∆egt showed a slightly increased growth rate and cell yield on methanol (Figure S7), which was more evident with higher methanol concentration, indicating that EGT is not required for methylotrophic growth. When they were grown on methanol at elevated temperature (31◦C and 34◦C), slightly increased growth of ∆egt was still observed (data not shown). The reason for the better growth would be the biological cost for EGT production. Interestingly, the growth on glucose was not affected by the mutagenesis at all (data not shown), which was congruent with the low productivity of EGT on glucose (**Figure 4**).

∆egt was more sensitive to heat shock than the wild type (**Figure 6A**). In addition to heat shock, UV irradiation severely reduced the viability of ∆egt (**Figure 6B**). Exposure of 100µM EGT solution to UV in the same condition resulted in a marked decrease in the EGT peak (35 and 87% decrease in 5 and 30 min, respectively) and a concomitant increased peak at 10.3 min (unknown compound) in the HPLC chromatogram, suggesting that EGT is degraded by UV (data not shown). Furthermore, when methanol-grown 22A cells were exposed to UV in the same condition for 5, 30, and 60 min, the intracellular EGT content decreased to 31, 0.5, and 0%, respectively (data not shown). These results indicated that EGT protects the cells for survival under UV. These results prompted us to see the effect of sunlight on the mutant survival. It is known that the strongest UV, UV-C (100–290 nm), does not reach to the ground due to the ozone layer. The UV lamp (254 nm) we used for the assay belongs to UV-C. Within the total UV reaching the ground, more than 95% is UV-A (320–400 nm), and less than 5% is UV-B (290–320 nm) that can cause DNA damage and kill bacteria. It is also known that glass is transparent to UV-A but not to UV-B (Duarte et al., 2009), and that cellophane is partly transparent to UV-B (Auras et al., 2004). We found that sunlight through the glass killed the wild type and the mutant at a comparable rate, and that sunlight through the cellophane killed ∆egt more efficiently than the wild type (**Figure 6C**). We also observed that the mutant formed colonies much slower than the wild type after the exposure (data not shown). These results strongly suggested that EGT protects

the cells against sunlight, to which the Methylobacterium cells are exposed in the phyllosphere.

Interestingly, ∆egt showed higher resistance to H2O<sup>2</sup> (**Figure 6D**), but no phenotype in sensitivity to other oxidative stress of methylglyoxal and diamide (**Figures 6E,F**). It is known that Methylobacterium species synthesize glutathione in addition to EGT, and that Mycobacterium species synthesize EGT and mycothiol as sulfur-containing antioxidants. In Mycobacterium smegmatis, mycothiol-deficient mutants showed elevated levels of EGT (Ta et al., 2011), whereas mycothiol level was unchanged in EGT mutant in M. smegmatis (Sao Emani et al., 2013). In contrast, egtA mutant in Streptomyces coelicolor A3(2), which still produces a decreased amount of EGT, produces five-fold more mycothiol compared to the wild type (Nakajima et al., 2015). Thus, in actinobacteria, a defect in one antioxidant may be compensated by an increased amount of another antioxidant. These facts led us to quantify glutathione in ∆egt. The glutathione content (reduced form) in ∆egt grown on methanol was 2.29µg/5 ml culture, while that in the wild type was 2.05µg/5 ml culture, showing a statistically insignificant increase (n = 3 each, p = 0.11, Student's t-test). The oxidized form was not detected. At this time, we cannot attribute the increased H2O<sup>2</sup> resistance of the mutant to increased glutathione content. In the case of egtA mutant in S. coelicolor A3(2) (Nakajima et al., 2015), intracellular EGT content was severely reduced but the secreted EGT was not, and the mutant showed high susceptibility to H2O2. Thus, the resistance to H2O<sup>2</sup> cannot be explained

(C) that was done in quintuplicate, were done in quadruplicate. The data are presented as the mean ± SD (n = 4). Experimental conditions are written in the text.

solely by intracellular EGT content for different bacteria, and it is possible that the strain 22A may contain other antioxidant molecules that compensate for the defect in EGT. A recent paper described the essential role of EGT and mycothiol in lincomycin A synthesis in Streptomyces lincolnensis (Zhao et al., 2015). Thus, it is also possible that Methylobacterium species utilize EGT for synthesis of other unknown secondary metabolites.

Taken together, these results indicate that EGT contributes to resistance to heat shock and UV stress. The antioxidant activity of EGT has been shown in many reports (Cheah and Halliwell, 2012). EGT reacts with the hydroxyl radical at a diffusioncontrolled rate, but not readily with H2O<sup>2</sup> and O<sup>−</sup> 2 . Furthermore, EGT can repair the guanyl radical, which is a product of DNA ionization (Asmus et al., 1996). This could be the reason why ∆egt showed susceptibility to UV irradiation that elicits DNA damage.

Formaldehyde is the central metabolite in methylotrophy and it is not known whether EGT reacts with formaldehyde. Glutathione, on the other hand, is known to spontaneously react with formaldehyde to form S-hydroxymethylglutahione in Paracoccus denitrificans (Goenrich et al., 2002). In formaldehyde dissimilation in Methylobacterium, formaldehyde reacts with dephosphotetrahydromethanopterin (d4HMPT) either spontaneously or enzymatically by formaldehyde-activating enzyme (Fae). In assimilation, methylene tetrahydrofolate (H4F), as the C1 unit for the serine cycle, is produced either spontaneously by the condensation of formaldehyde and H4F or enzymatically through the H4MPT-dependent pathway (Ochsner et al., 2015). Thus, in Methylobacterium, glutathione is not involved in formaldehyde assimilation and dissimilation. However, recalling the case of P. denitrificans, we examined whether EGT reacts with formaldehyde. We mixed 1 M formaldehyde and 100µM EGT, incubated it for 5–24 h at room temperature, and then examined the HPLC pattern. We did not observe any change in peak height or retention time, suggesting that EGT does not spontaneously react with formaldehyde in vitro. This result again indicates that EGT is not directly involved in methylotrophy.

Next, we examined the growth of ∆egt in the phyllosphere. The growth of ∆egt (mTn5gusA-pgfp22) was slightly decreased compared to 22A-rif when they were independently grown on A. thaliana leaves (**Figure 7**). When they were co-inoculated in competitive conditions, the lesser growth of 22A-rif was observed. As previously shown in Figure S7, the biological cost of EGT synthesis might be the reason for the lesser growth of the wild type. In this experimental condition, we used fluorescent lamps, which usually contain little UV. In addition, plastic petri dishes diminish UV. Thus, the experiment was performed in UV-free conditions.

### Exogenous Supplementation of EGT Is Not Effective in the Growth of Wild Type and Plant

Supplementation of EGT did not enhance the growth of the wild type on methanol, and was rather inhibitory (Figure S8A). The cause of this effect is unknown, but it is possible that EGT may inhibit some periplasmic enzymes, to which many of those involved in methanol oxidation belong. It is also unknown whether the cells incorporate EGT via transporters, since EGT is believed to be membrane-impermeable. We found proteins homologous to human OCTN1 (BAA23356.1) encoded in Methylobacterium genomes with less than 27% identity. This should be further examined to determine their involvement in EGT uptake.

The growth of A. thaliana on 1/2 MS agar was not affected by the exogenous supplementation of EGT at the tested concentrations (Figure S8B). Since EGT is exclusively accumulated in the cells and is not secreted extracellularly, exogenous EGT may not occur in Methylobacterium-plant

interactions. Lysed bacterial cells may serve as an EGT source for plants, but there have been no reports on the existence of an active EGT transporter in plants. There are many OCTN1 homologs encoded in the genomes of a wide variety of plants, including cucumber, chickpea, wheat, barley, grapevine, potato, Physcomitrella, and so on, with less than 33% identity. It is currently unknown whether any of them is involved in EGT transport in plants. Since EGT is also found in plants, there must be an EGT transporter in plants; however, the biological role of EGT in plants remains uninvestigated.

In conclusion, we found that Methylobacterium species are able to synthesize abundant EGT. EGT is exclusively accumulated in the cells and is not secreted extracellularly. Using one of the most productive strains, we optimized the culture conditions mainly with respect to carbon and nitrogen sources and supplements. The production level overwhelms that of the most productive mushroom. High cell density cultivation is reported for Methylobacterium species, which can achieve as much as 100 g dry weight cells per liter (Bourque et al., 1995). These established cultivation techniques and the genetic manipulation we showed by gene duplication would be very useful in terms of improving production. Since pure EGT is expensive due to limited sources, we are interested in the further development of more efficient fermentative production using Methylobacterium species and methanol as a cheap feedstock. We identified EGT synthesis genes in strain 22A and generated an EGT synthesis mutant. EGT synthesis was found to be nonessential for methylotrophy. We also found that the mutant was susceptible to UV and heat shock. Importantly, the mutant is more susceptible to sunlight than the wild type. As inhabitants on the plant surface, resistance to UV is essential for their survival. Herein, we provided evidence of the involvement of EGT in the resistance. Since EGT was produced by many Methylobacterium type strains, EGT might be a ubiquitous UV-protectant for the genus.

### REFERENCES


Interestingly, EGT is a precursor for the synthesis of selenoneine, which has a selenoketone structure instead of thiocarbonyl in EGT (Yamashita et al., 2010). Selenoneine is reported to have three orders of magnitude stronger radical scavenging activity against 1-diphenyl-2-picrylhydrazyl than EGT (Yamashita and Yamashita, 2010). The synthesis of selenoneine is suggested to be dependent on selenocysteine (Pluskal et al., 2014). Further study is necessary to examine the possible potential of selenoneine synthesis and its role in Methylobacterium species. In addition, an EGT-degrading enzyme called ergothionase has been reported for E. coli (Wolff, 1962) and Burkholderia (Muramatsu et al., 2013). We found the homologous gene in the 22A genome, the mutant of which may be more productive.

### AUTHOR CONTRIBUTIONS

KA, SM, and AT designed the experiments. KA, SM, YF, FF, and AT performed the experiments. KA, SM, FF, and AT wrote the manuscript.

### ACKNOWLEDGMENTS

This work is supported, in part, by the Japan Science and Technology Agency (JST) A-Step feasibility study stage (#AS251Z00826N), the Noda Institute for Scientific Research, and the Advanced Low Carbon Technology Research and Development Program (ALCA), JST.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2015.01185


Fahey, R. C. (2001). Novel thiols of prokaryotes. Ann. Rev. Microbiol. 55, 333–356. doi: 10.1146/annurev.micro.55.1.333


**Conflict of Interest Statement:** The authors declare that this research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. AT is listed as an inventor on a pending patent application (JP P2014-259232, filing date: December 22, 2014) on the use of Methylobacterium species for EGT production.

Copyright © 2015 Alamgir, Masuda, Fujitani, Fukuda and Tani. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Control of Wilt and Rot Pathogens of Tomato by Antagonistic Pink Pigmented Facultative Methylotrophic *Delftia lacustris* and *Bacillus* spp.

Veeranan Janahiraman<sup>1</sup> , Rangasamy Anandham<sup>1</sup> \*, Soon W. Kwon<sup>2</sup> , Subbiah Sundaram1, 3, Veeranan Karthik Pandi <sup>4</sup> , Ramasamy Krishnamoorthy <sup>1</sup> , Kiyoon Kim<sup>3</sup> , Sandipan Samaddar <sup>3</sup> and Tongmin Sa<sup>3</sup> \*

#### *Edited by:*

Anton Hartmann, Helmholtz Zentrum München, Germany

#### *Reviewed by:*

Oswaldo Valdes-Lopez, National Autonomous University of Mexico, Mexico Wubei Dong, Huazhong Agricultural University, China

#### *\*Correspondence:*

Rangasamy Anandham anandhamranga@gmail.com Tongmin Sa tomsa@chungbuk.ac.kr

#### *Specialty section:*

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

*Received:* 14 July 2016 *Accepted:* 14 October 2016 *Published:* 07 November 2016

#### *Citation:*

Janahiraman V, Anandham R, Kwon SW, Sundaram S, Karthik Pandi V, Krishnamoorthy R, Kim K, Samaddar S and Sa T (2016) Control of Wilt and Rot Pathogens of Tomato by Antagonistic Pink Pigmented Facultative Methylotrophic Delftia lacustris and Bacillus spp. Front. Plant Sci. 7:1626. doi: 10.3389/fpls.2016.01626 <sup>1</sup> Department of Agricultural Microbiology, Agricultural College and Research Institute, Tamil Nadu Agricultural University, Madurai, India, <sup>2</sup> Korean Agricultural Culture Collection, National Academy of Agricultural Science, Rural Development Administration, Jeonju, South Korea, <sup>3</sup> Department of Environmental and Biological Chemistry, Chungbuk National University, Cheongju, South Korea, <sup>4</sup> Department of Plant Pathology, Agricultural College and Research Institute, Tamil Nadu Agricultural University, Coimbatore, India

The studies on the biocontrol potential of pink pigmented facultative methylotrophic (PPFM) bacteria other than the genus Methylobacterium are scarce. In the present study, we report three facultative methylotrophic isolates; PPO-1, PPT-1, and PPB-1, respectively, identified as Delftia lacustris, Bacillus subtilis, and Bacillus cereus by 16S rRNA gene sequence analysis. Hemolytic activity was tested to investigate the potential pathogenicity of isolates to plants and humans, the results indicates that the isolates PPO-1, PPT-1, and PPB-1 are not pathogenic strains. Under in vitro conditions, D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 showed direct antagonistic effect by inhibiting the mycelial growth of fungal pathogens; Fusarium oxysporum f. sp. lycopersici (2.15, 2.05, and 1.95 cm), Sclerotium rolfsii (2.14, 2.04, and 1.94 cm), Pythium ultimum (2.12, 2.02, and 1.92 cm), and Rhizoctonia solani (2.18, 2.08, and 1.98 cm) and also produced volatile inhibitory compounds. Under plant growth chamber condition methylotrophic bacterial isolates; D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 significantly reduced the disease incidence of tomato. Under greenhouse condition, D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 inoculated tomato plants, when challenged with F. oxysporum f. sp. lycopersici, S. rolfsii, P. ultimum, and R. solani, increased the pathogenesis related proteins (β-1,3-glucanase and chitinase) and defense enzymes (phenylalanine ammonia lyase, peroxidase, polyphenol oxidase, and catalase) on day 5 after inoculation. In the current study, we first report the facultative methylotrophy in pink pigmented D. lacustris, B. subtilis, and B. cereus and their antagonistic potential against fungal pathogens. Direct antagonistic and ISR effects of these isolates against fungal pathogens of tomato evidenced their possible use as a biocontrol agent.

Keywords: methylotrophs, induced systemic resistance, pathogenesis-related proteins, biological control, tomato, antagonism

## INTRODUCTION

Tomato (Lycopersicon esculentum) is one of the most popular commercial vegetable crops. In India, it occupies an area of 0.54 million ha with a production of 7.60 million tons (Kumar et al., 2012). Among the pathogens that affect the tomato crop, soil borne fungal pathogens, including Fusarium, Pythium, Rhizoctonia, and Verticillium causing the root rot or dampingoff and wilt affect the quality with yield reduction (Lucas et al., 1997). Among these pathogens, Fusarium oxysporum f. sp. lycopersici is a highly destructive pathogen on both greenhouse and field grown tomatoes. In spite of the promising results shown by chemical treatments in controlling the fungal pathogens, phytotoxicity, and chemical residues are the major problems leading to environmental pollution and human health hazards. Alternatively in the present study, we have tested the possibility of using facultative methylotrophic bacteria which is ubiquitously occurring with intimate association with plants, as a biocontrol agent in controlling wilt and root rot pathogens of tomato.

Methylotrophs consists of a subpopulation of the bacteria which are able to use single carbon compounds (methanol and other C1 carbon compounds) as a sole carbon and energy source. Methylotropic bacteria can take up greenhouse gases and reduce global warming (Iguchi et al., 2015). Methylotrophs are found in both phyllosphere and rhizosphere of the plants and utilize the plant waste methanol as carbon source. Recently, there have been extensive studies on methylotrophs and their ability to promote plant growth were carried out (Madhaiyan et al., 2004; Poonguzhali et al., 2008; Yim et al., 2012). In addition, methylotrophic activity was reported in different bacterial families (Eyice and Schäfer, 2016). Among the facultative methylotrophic (FM) bacteria belonging to Alpha-, Beta-, and Gamma-subclasses of Proteobacteria and Firmicutes, the genus Methylobacterium has been widely studied. These FM bacteria ubiquitously occurring in plants can be isolated using selective media containing methanol as the sole carbon source and identified by their characteristic pink color (Corpe and Basile, 1982; Lidstrom and Chistoserdova, 2002). They are commonly referred as pink pigmented facultative methylotrophic (PPFM) bacteria. They are non-pathogenic and distributed widely in the plant phyllosphere, and have been isolated from more than 100 species of plants, ranging from liverworts and mosses to angiosperms and gymnosperms (Corpe and Basile, 1982).

Among the PPFM, the genus Methylobacterium is one of the dominant genera, which act as a symbiont with plants by consuming methanol, a plant metabolic waste and in turn providing vitamin B12, auxins, cytokinins useful for the plant growth (Madhaiyan et al., 2006). Although the genus Methylobacterium has also been well-documented for their induction of systemic resistance (ISR) in plants against plant pathogens (Madhaiyan et al., 2004, 2006; Indiragandhi et al., 2008), till date to the best of our knowledge there is no report available on the direct antagonistic effect of FM bacteria. Hence in the current study, we tested both direct antagonistic effect of FM bacteria against Fusarium oxysporum f. sp. lycopersici, Sclerotium rolfsii, Pythium ultimum, and Rhizoctonia solani causing tomato root rot and wilt diseases and indirect effect by the induction of ISR and pathogenesis related (PR) proteins in the pathogen challenged tomato plants grown under greenhouse condition.

### MATERIALS AND METHODS

### Source of Bacterial Strains, Fungal Pathogens, and Tomato Seeds

PPFM bacteria were isolated from the phylloplane of crop plants listed in **Table 1** through leaf imprinting technique on solid ammonium mineral salts (AMS) medium (pH 6.8) supplemented with 0.5% methanol and cyclohexamide (30 µg ml−<sup>1</sup> ) and incubated at 28◦C for 3–5 days (Whittenbury et al., 1970). The single colonies with reddish pink pigmentation were picked and purified. The PPFM isolates were grown for 72 h on AMS medium. Gram reaction and other biochemical tests were carried out as described previously (Gerhardt et al., 1981). Plant pathogens (Fusarium oxysporum f. sp. lycopersici MTCC 4356, S. rolfsii KACC 43068, R. solani MTCC 4633, and P. ultimum KACC 40705) were obtained from Microbial Type Culture Collection, Chandigarh, India and Korean Agricultural Culture Collection, Jeonju, Republic of Korea. Fungal pathogens were grown and maintained in potato dextrose agar (PDA) before further use. Tomato cultivar CO4, which is susceptible to both seedling wilt and root rot, was obtained from the Department of vegetables at Horticultural College and Research Institute, Tamil Nadu Agricultural University, Coimbatore, India.

### Antagonistic Assay on Solid Medium

All PPFM bacterial isolates were screened for their ability to inhibit the growth of fungal pathogens on PDA plates by dual culturing technique with three replications (Yoshida et al., 2001). After 120 h incubation at 30 ± 2 ◦C, the distance between the bacterial growth and fungal mycelial growth was measured to find out the antagonistic effect of PPFM bacterial isolates on test pathogenic fungi. The antagonistic effect of volatile antifungal compounds produced by PPFM bacterial isolates against test fungal pathogens was evaluated on PDA medium with three replications after 120 h of incubation at 30 ± 2 ◦C as described by Trivedi et al. (2008). Briefly, PPFM isolates antagonism due to volatile compounds was evaluated by preparing a bacterial lawn on agar plates. After incubation for 24 h, the lid of the plate was replaced by a plate containing an agar block of the test fungus grown on PDA. The two plates were sealed together with parafilm. Control sets were prepared in a similar manner, without PPFM in the bottom plate. Then the Petri dishes were incubated at 28◦C, and the observations were recorded after 5 days. Growth inhibition of the test fungus was calculated in % using the formula: (r1 − r2/r1) ∗ 100, where r1 (a control value) represents the radial growth of the fungus in control sets without, and r2 with bacteria.

### Antagonistic Assay in Liquid Medium

A volume of 1 ml of bacterial culture grown in potato dextrose broth for 72 h (10<sup>9</sup> cfu ml−<sup>1</sup> ) and a disc of test fungus (10 mm) from a well-grown fungal colony on PDA plates were inoculated into 50 ml potato dextrose broth in 250 ml conical flasks and incubated at 30 ± 2 ◦C on a rotary shaker for 120 h. Potato


dextrose broth inoculated only with fungi served as positive control. The broth with fungal mat in various treatments was passed through a pre-weighed Whatman No. 1 filter paper. The filter papers were dried for 24 h at 70 ± 1 ◦C to obtain a constant weight. The percentage reduction in fungal mycelial weight was calculated as described previously by Trivedi et al. (2008).

### Other Antimicrobial Traits Assay

Siderophore, hydrocyanic acid (HCN), and salicylic acid production, β-1,3-glucanase activity and chitinase activity of the isolates were measured as mentioned in **Supporting Information File Materials and Method I**.

### 16S rRNA Gene Sequence Analysis

Bacterial isolates were grown in AMS medium for 72 h and DNA was extracted according to Sambrook et al. (1989). The gene encoding bacterial 16S rRNA was amplified through PCR with forward primer 27f: 5′ -AGAGTTTGATCCTGGCTC AG-3′ and reverse primer 1492r: 5′ -GGTTACCTTGTTACG ACTT-3′ . Nearly, complete 16S rRNA gene sequences of PPFM bacterial isolates from the automatic sequencer were aligned using the integrated SINA alignment tool from the ARB-SILVA website (Pruesse et al., 2007) and bacterial identities deduced by using the EzTaxon-e server (http://eztaxon-e.ezbiocloud.net/) to ascertain their closest relatives (Kim et al., 2012). Bacterial gyrA gene is a potential chromosomal marker for the phylogenetic identification of Bacillus genera. PCR amplification and detection of gyr A gene was given in the **Supporting Information File Materials and Method II**.

### Methanol Utilization and Detection of Methanol Dehydrogenase Gene (mxaF)

Methanol utilization by the isolates was checked according to Sy et al. (2001) in M72 medium. The bacterial suspensions were diluted in M72 medium [optical density (OD) = 0.05], and added with one of the following compounds; methanol (MeOH) (10, 50, 100, or 500 mM), pyruvate (10 mM), or succinate (10 mM). Growth was monitored by measuring OD at 620 nm. Detection of mxaF gene was given in **Supporting Information File Materials and Method III**.

### Hemolytic Activity and Response to Antibiotics

This experiment was conducted to test the production of exotoxins called hemolysin able to destroy the Red Blood Cells (RBC) and hemoglobin by the methylotrophic isolates Delftia lacustris PPO-1, Bacillus subtilis PPT-1, Bacillus cereus PPB-1 in sheep blood agar plates (catalog no. MP1301 Himedia, India). The plates were incubated for 48–72 h at 30◦C and observed for presence of clear zone, greenish brown, and no zone which indicates the complete (β hemolysis), partial (α hemolysis), and no hemolytic (È) activity, respectively. Susceptibility to the antibiotics rifampicin, ampicillin, oleandomycin, chloramphenicol, tetracycline, novobiocin, streptomycin, spectinomycin, kanamycin, trimethoprim, hygromycin, ceftriaxone, gentamycin, cefepime, and amikacin (each at

50 µg ml−<sup>1</sup> ) was tested for D. lacutris PPO-1 on R2A agar plates as per Bauer et al. (1966).

### Plant Growth Chamber Assay

Plant growth chamber experiment was conducted to assess the biocontrol potential of PPFM bacterial isolates on wilt and rot diseases of tomato. Bacterial cultures grown for 72 h were harvested by centrifugation at 10,000 × g for 10 min at 4◦C, washed twice with sterile distilled water and suspended in 0.03 M MgSO<sup>4</sup> solution. Tomato seeds were surface sterilized with 70% ethanol for 1 min, 0.5% NaOCl for 2 min, and washed four times with sterilized distilled water. The surface-sterilized seeds were imbibed in the 5 ml of bacterial suspension (10<sup>9</sup> cfu ml−<sup>1</sup> ) for 2 h, and single seed was sown in a 400 ml plastic pot filled with mixture of sterilized soil and farm yard manure in 3:1 ratio. The pots were subjected to following treatments; T1—B. subtilis PPT-1 + F. oxysporum f. sp. lycopersici, T2—B. cereus PPB-1 + F. oxysporum f. sp. lycopersici, T3—D. lacustris PPO-1 + F. oxysporum f. sp. lycopersici, T4—B. subtilis PPT-1 + S. rolfsii, T5—B. cereus PPB-1 + S. rolfsii, T6—D. lacustris PPO-1 + S. rolfsii, T7—B. subtilis PPT-1 + R. solani, T8— B. cereus PPB-1 + R. solani, T9—D. lacustris PPO-1 + R. solani, T10—B. subtilis PPT-1 + P. ultimum, T11—B. cereus PPB-1 + P. ultimum, T12—D. lacustris PPO-1 + P. ultimum, T13—F. oxysporum f. sp. Lycopersici, T14—S. rolfsii, T15—R. solani, T16—P. ultimum (**Figure S1**). One milliliter of fungal spore suspension (10<sup>6</sup> spore ml−<sup>1</sup> ) was applied to the soil on 30th day after sowing. Ten potted plants were maintained for each treatment. The experiment was conducted in a completely randomized block design with three replications. Pots were placed in growth chambers at 20 ± 1 ◦C, photoperiod starting with 12 h dark followed by 12 h lights (18 µ mol m−<sup>2</sup> s −2 ). Plants were watered with 1 X Hoagland solution. The disease incidence was observed by recording the wilting of leaves leading to plant drying. Randomly 15 plants were selected from each treatment and the number of wilted plants was recorded. The mean wilt disease incidence was expressed in percentage. Disease incidence was recorded up to 30 d after challenge inoculation with fungal pathogens. The percent disease incidence (PDI) was calculated, by using the formula. PDI = Number of plants infected/Total number of plants observed × 100.

## Greenhouse Assay

Tomato seeds were surface sterilized and imbibed in the PPFM bacterial suspension as described earlier. Four seeds were sown in a 20 L earthen pot containing 10 Kg of unsterilized soil. Control seeds were imbibed with 0.03 M MgSO<sup>4</sup> for 2 h. All treatments followed in the plant growth chamber assay were imposed. The experiment was conducted in a completely randomized block design with three replications. Twenty-one days after sowing, PPFM bacterial cells suspended in 0.03 M MgSO<sup>4</sup> (10<sup>9</sup> cfu ml−<sup>1</sup> ) were sprayed over tomato plants with handheld pneumatic sprayer till the plants were completely wetted. Thirty days after sowing, 20 ml of fungal spore suspension (10<sup>6</sup> spore ml−<sup>1</sup> ) was applied into the soil. Initially, F. oxysporum f. sp. lycopersici (1.2 × 10<sup>2</sup> cfu), S. rolfsii (1.5 × 10<sup>2</sup> cfu), R. solani (2.1 × 10<sup>2</sup> cfu), and P. ultimum (1 × 10<sup>2</sup> cfu) were present in 1 g of experimental dry soil. Germination percentage was recorded on 7 days after sowing (DAS). Seedling vigor index was calculated as % germination × seedling length (shoot length + root length) in cm on 15 DAS (Baki and Anderson, 1973). Leaf samples were collected on 0, 3, 5, 7, and 9 days after challenge inoculation of pathogen and frozen immediately at −20◦C for analyses of PR proteins and defense enzymes. The plant height was recorded on 30, 60, and 90 DAS. The crop was harvested at physiological maturity and yield was recorded.

### PR Proteins Analysis

Frozen leaves were ground at 4◦C in an ice-chilled mortar with liquid nitrogen, and the resulting powder was suspended in 100 mM potassium phosphate buffer, pH 7.0 (2:2.5, w/v). Crude homogenates were centrifuged at 8000 × g for 10 min at 4◦C, and the supernatants were kept frozen at −20◦C until use. β-1,3-glucanase was assayed based on the release of reducing sugars from laminarin, as described by Liang et al. (1995). One unit of β-1,3-glucanase activity was determined as the amount of enzyme required to liberate 1 ng of glucose equivalent at 37◦C in 1 h. Chitinase activity was assayed by measuring the release of N-acetyl-glucosamine from prepared colloidal chitin, as described by Singh et al. (1999). One unit of chitinase activity was determined as the amount of enzyme required to liberate 1 n mol of N-acetyl-glucosamine equivalent at 37◦C in 1 h. PAL activity was measured using the procedure described by El-shora (2002). One unit represented the conversion of 1 µ mol of Lphenylalanine to cinnamic acid per min. Oxidative enzymes like PO and PPO were estimated as described previously (Compant et al., 2005). The activity of PO and PPO were expressed as the change in unit of absorbance at 420 nm/g of fresh weight per min. Catalase activity was assayed spectrophotometrically as described by Stern (1937). The activity was measured by monitoring the degradation of H2O<sup>2</sup> using a spectrophotometer at 240 nm over 1 min. expressed in mmol min g−<sup>1</sup> of leaf tissue−<sup>1</sup> .

### Statistical Analysis

The data were analyzed by an analysis of variance (ANOVA) using the general linear model version 9.1; SAS Institute Inc., Cary, NC, USA. Means were compared using the least significant difference (LSD). The significance levels were within confidence limits of 0.05 or less.

### RESULTS

### Antagonistic Potential of PPFM Bacterial Isolates against Fungal Pathogens of Tomato

Twenty PPFM bacteria were isolated from phylloplane of various crop plants (**Table S1**). Among the PPFM bacterial isolates, 12 isolates showed direct antagonism toward tested pathogens. The isolates PPO-1, PPT-1, and PPB-1 exhibited mycelial growth inhibition of Fusarium oxysporum f. sp. lycopersici (2.15, 2.05, and 1.95 cm), S. rolfsii (2.14, 2.04, and 1.94 cm), P. ultimum (2.12, 2.02, and 1.92 cm), R. solani (2.18, 2.08, and 1.98 cm), respectively, in a dual plate assay on day 5 (**Table 1**). All the tested methylotrophic isolates were able to produce siderophores and salicylic acid and none of the isolates produced hydrocyanic acid. The isolates PPO-1, PPT-1, and PPB-1 produced the maximum siderophore (27.94, 27.45, and 25.44 µg ml−<sup>1</sup> ) and salicylic acid (89, 88, and 87 µg ml−<sup>1</sup> ), respectively (**Table 1**). Chitinase and β-1,3-glucanase activities were observed higher in PPO-1 isolate followed by PPT-1 and PPB-1 isolates (**Table 1**).

All PPFM bacterial isolates produced volatile antifungal compound(s), and inhibited the mycelial growth of the tested fungal pathogens in sealed Petridishes. The higher mycelial growth inhibitions were observed in F. oxysporum f. sp. lycopersici (52.84, 51.73, and 50.65%), S. rolfsii (43.34, 42.23, and 41.12%), P. ultimum (57.78, 56.67, and 55.56%), and R. solani (44.45, 43.34, and 42.23%) due to the antagonistic effect of respective PPO-1, PPT-1, and PPB-1 isolates (**Table 2**). All PPFM bacterial isolates showed the reduction in pathogenic fungal biomass in the co-cultured liquid medium. The isolates PPO-1, PPT-1, and PPB-1 showed respective biomass reduction of F. oxysporum f. sp. lycopersici (77.14, 76.57, and 76.00%), S. rolfsii (79.26, 78.45, and 77.64%), P. ultimum (88.08, 87.65, and 85.95%), R. solani (89.81, 86.03, and 85.66%) (**Figure 1**).

### Molecular Characterization and Phylogenetic Analysis

PPFM bacterial isolates PPO-1 (JN 088183), PPT-1 (JN 088185), and PPB-1 (JN 088184) showed the highest 16S rRNA gene sequence similarities to type strains of D. lacustris DSM 21246<sup>T</sup> (100%), B. subtilis KCTC 13429<sup>T</sup> (99.87%), and B. cereus ATCC 14579<sup>T</sup> (99.67%), respectively. The colony morphology of pink pigmented methylotrophic isolates PPO-1, PPT-1, and PPB-1 are shown in **Figure S2**. Bacterial growth was noted on either pyruvate or succinate as the sole carbon source. Also, growth was obtained at a methanol concentration up to 500 mM. Presence of methanol dehydrogenase gene (mxaF) was examined using Methylobacterium extorquens AM1 as reference strain. An amplification product of the expected size, 560 bp, for all the isolates similar to reference strain M. extorquens AM1 were noted which indicated the presence of mxaF. The isolate PPO-1 was Gram negative and isolates PPT-1 and PPB-1 were Gram positive. Isolates PPO-1, PPT-1, and PPB-1 did not exhibit hemolytic activity and isolate PPB-1 grew on mannitol egg yolk polymyxin agar MYP agar forming white precipitation which is the characteristic feature for B. cereus. Presence of housekeeping gene (gyr A) in two isolates of Bacillus sp. (PPT-1 and PPB-1) was examined, PPT-1 showed positive (**Figure S3**). Methylotrophic isolates PPO-1, PPT-1, and PPB-1 were susceptible to the antibiotics such as rifampicin, ampicillin, oleandomycin, chloramphenicol, tetracycline, novobiocin, streptomycin, spectinomycin, kanamycin, trimethoprim, hygromycin, ceftriaxone, gentamycin, cefepime, and amikacin at the concentration of 50 µg ml−<sup>1</sup> . For hemolytic activity the PPFM isolates did not produce any clear or greenish zone around the colonies in sheep blood agar medium, which may indicates that the isolates are non-pathogenic to human.

### Biocontrol Effect of PPFM Bacterial Isolates on Tomato Wilt and Rot Diseases

Under plant growth chamber condition, soil inoculated with fungal pathogens showed 65% disease incidence. Higher diseases incidence was observed in the plants treated with R. solani followed by P. ultimum, F. oxysporum f. sp. Lycopersici, and S. rolfsii. However, the PPFM bacterial isolates inoculated tomato plants when challenged with the fungal pathogens recorded only 15% disease incidence (**Figure 2**). In greenhouse experiment, bacterization of tomato seeds with PPFM isolates D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 significantly increased the germination percentage and the seedling vigor compared to control (**Table 3**). However, pathogen alone treatment has negative impact on seed germination, seedling vigor and plant height. No much variation was observed in plant height among the plants treated with

TABLE 2 | Effect of facultative methylotrophic bacterial volatile antimicrobial compounds on the growth of fungal pathogens.


Values are mean of three replications ± SD (standard deviation). Values in each column followed by same letter(s) are not statistically different by LSD (P ≤ 0.05).

different PPFM isolates and challenge inoculated with pathogens. However, significant variation in plant height was observed between PPFM challenge-inoculated with pathogen treatment and pathogen alone treated plants (**Figure S4**). D. lacustris PPO-1 inoculated tomato plants when challenged with R. solani, S. rolfsi, and F. oxysporum f. sp. lycopersici recorded the higher fruit yield of 9.35, 9.08, and 8.90 fruits plant−<sup>1</sup> , respectively (**Table 3**).

In this study, D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 inoculations induced significant protection in tomato plants against the tested pathogens. PPFM isolates D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 inoculated tomato plants when challenge-inoculated with F. oxysporum f. sp. lycopersici, S. rolfsii, P. ultimum, and R. solani, increased PR proteins over uninoculated controls. The higher activities of β-1,3 glucanase, chitinase, PO, PPO, PAL, and catalase enzymes were recorded on 5th day after challenge inoculation with pathogens (**Table 4**). Increase in PO enzyme activity was observed in tomato plants inoculated with PPFM isolates D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 and challenged with F. oxysporum f. sp. lycopersici, S. rolfsii, P. ultimum, and R. solani. The activities of PPO, PAL, chitinase, and

different letter(s).

solani, T16—P. ultimum. Each value represents means of three replicates per treatment. Error bars indicate ± standard error (SE). In the bar, significant differences according to LSD at 0.05% levels are indicated by different letter(s).

catalase enzymes also had a similar trend (**Table 4**). In particular, D. lacustris PPO-1 inoculated tomato plants when challenge inoculated with R. solani, P. ultimum, S. rolfsii, and F. oxysporum f. sp. lycopersici exhibited significant increase in β-1,3-glucanase activity (**Table 4**). Delftia lacustris PPO-1 challenge inoculated with pathogen exhibited higher chitinase and catalase activity compared to other treatments.

### DISCUSSION

Members of pink-pigmented facultative methylotrophs occupies different habitats such as leaf surface, soil, water, and air. The plant-methylotroph association could be attributed to the unique ability of these bacteria to grow at the expense of methanol, a cell wall pectin degradative product from plants. PPFM has the ability to oxidize methanol using the methanol dehydrogenase enzyme. The mxaF gene encodes the large subunit of this enzyme, which is key in methylotrophic metabolism. In recent years, interaction study have shown that PPFM bacteria can increase plant protection against phytopathogen attack (Benhamou et al., 2000). In the current study, PPFM bacteria were isolated from various crop plants and used as a bio-control agent against tomato phytopathogens.

Bio-control ability of the PPFM isolates against phytopathogens were tested using three methods. One is by dual culture assay in which the PPFM isolate and pathogens were grown in same plate. In that the isolates PPO-1, PPT-1, and PPB-1, exhibited higher antagonistic potential against F. oxysporum f. sp. lycopersici, S. rolfsii, P. ultimum, and R. solani. The inhibitory effect of fungal growth observed in the current study in the dual culturing assay may be attributed to certain diffusible antifungal metabolites of the methylotrophic isolates (Montealegre et al., 2003). In the second method, the effect of PPFM volatile compound against phytopathogen was tested. The results revealed that PPFM bacterial isolates produced volatile antifungal compound which inhibited the growth of test fungal pathogens. In third method, both PPFM isolates and phytopathogens were simultaneously grown in liquid medium as co-cultures. The reduction in fungal biomass due to inhibitory effect of antagonistic PPFM was compared with fungal cultures grown in medium without antagonistic PPFM. Earlier a non-methylotrophic Delftia lacustris possessing nitrogen fixing trait capable of suppressing the growth of rice pathogens (Xanthomonas oryzae pv. oryzae, R. solani, and Pyricularia oryzae) was reported by Han et al. (2005).

The PPFM isolates capable of producing siderophopres proves their competitive advantage in the natural ecosystem by limiting the supply of iron and essential trace elements to the fugal pathogens as previously suggested by Indiragandhi et al. (2008). The inhibitory effect on the fungal pathogens tested may also be attributed to the salicylic acid production capability of PPFM isolates, as already evidenced in Methylobacterium oryzae CBMB20 inoculated Pseudomonas syringae pv. tomato challenged tomato plants (Indiragandhi et al., 2008). However, PPO-1, PPT-1, and PPB-1 methylotrophic bacterial isolates not producing HCN, eliminated the possibility of blocking the cytochrome oxidase system in fungal pathogen (Trivedi et al., 2008). Presence of chitinase and β-1,3-glucanase activities were previously reported in Delftia sp., B. subtilis and B. cereus (Chen et al., 2004; Jørgensen et al., 2009; Liang et al., 2014). Chitin and glucans those are responsible for the rigidity of fungal cell walls, thereby destroying cell wall integrity due to secretion of chitinase and β-1,3-glucanase limiting the growth of the pathogen (Chen et al., 2004; Jørgensen et al., 2009; Liang et al., 2014).

Methylotrophs oxidize methanol to formaldehyde via a key enzyme methanol dehydrogenase and the use of structural gene mxaF as a functional probe for methylotrophs is well-explained (Sy et al., 2001). Therefore, to evaluate the presence of methanol oxidation genes PCR amplification was performed with nondegenerate primers corresponding to highly conserved regions of mxaF gene sequences. Presence of mxaF gene and utilization of methanol clearly indicated the methylotrophic nature of isolates PPO-1, PPT-1, and PPB-1.

In the current study, none of the tested methylotrophic isolates exhibited hemolytic activity. This test was applied to investigate the potential pathogenicities of clinical and environmental isolates for plants and humans (Bevivino et al., 2002). In a previous study, Shin et al. (2012) observed the hemolytic behavior and resistance to antibiotics such as cefepime, amikacin, and gentamycin in opportunistic pathogenic D. lacustris. In the present study D. lacustris PPO-1 did not exhibit hemolytic activity and susceptibility toward various antibiotics. Shin et al. (2012) isolated D. lacustris from blood and bile fluids of human whereas D. lacustris PPO-1 was isolated from



Values are mean of three replications ± SD (standard deviation). Values in each column followed by same letter(s) are not statistically different by LSD (P ≤ 0.05). \*F. oxysporum—F. oxysporum f. sp. lycopersici.

phylloplane of onion. Clinical Burkholderia cepacia isolates generally exhibited resistance to larger numbers of antibiotics and a higher degree of resistance to the single antibiotics than the environmental isolates (Bevivino et al., 2002). Clinical strains of B. cepacia secrete cytotoxic factors that allow macrophage and mast cell death in the presence of external ATP which appears to be lower in the environmental strain (Melnikov et al., 2000).

Hence, it is concluded that D. lacustris PPO-1 is probably not a pathogen. However, suitable molecular finger printing techniques should be devised in future to differentiate pathogenic clinical D. lacustris from environmental isolates. Chun and Bae (2000) demonstrated the use of gyrA sequences (coding for DNA gyrase subunit A) for accurate classification of B. subtilis and related taxa. In this study, isolate PPT-1 showed the amplified product size of 1000 bp which corresponds to the gyrA gene and confirmed as Bacillus subtilis. Though facultative methylotrophy has been documented in diverse heterotrophic genera (Hung et al., 2011), to the best of our knowledge, this is the first study to report the methylotrophy in Delftia lacustris. Nevertheless, the occurrence of pink pigmented B. subtilis was earlier reported, their methylotrophy was not documented (Oppong et al., 2000). However, the methylotrophy was reported only in non-pigmented B. methylicus, B. methanolicus, and B. methylotrophicus (Arfman et al., 1992; Madhaiyan et al., 2010). Hence, our study also first reports the methylotrophic pink pigmented B. subtilis PPT-1 and B. cereus PPB-1.

In the current study, the reduction of root rot and wilt disease incidences in the D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 inoculated pathogen challenged tomato plants may be attributed to their direct antagonistic effect producing diffusible and volatile antibiotics as well as indirect induction of β-1, 3-glucanase, chitinase, PO, PPO, PAL, and catalase. ISR in several crops has been correlated with a two fold increase in activity of pathogenesis related PO and chitinase proteins in PGPR inoculated plants when challenged with pathogens (Nielsen et al., 1998; Xue et al., 1998; Nandakumar et al., 2001). Previously, several studies reported the induction of ISR like chitinase, PAL, β-1,3-glucanase, peroxidase, and PPO in rice, peanut, and tomato plants due to methylobacterial inoculation against various pathogens (Madhaiyan et al., 2004, 2006; Indiragandhi et al., 2008). Similarly, Choudhary and Johri (2008) explicated the mechanisms of Bacillus species as inducers of systemic resistance and Kloepper et al. (2004) reported induction of systemic resistance due to inoculation of Bacillus spp. in several crops including tomato. These defense proteins have the potential to hydrolyze the major components of fungal cell walls viz, chitin and β-1,3-glucans (Ren and West, 1992). The ubiquitous plant colonizing methylobacteria shown to be highly resistant to various abiotic stresses (Romanovskaya et al., 1996) could have enabled them for their effective epiphytic colonization in plants compared to other biocontrol agents.

### CONCLUSIONS

The finding of the present study not only supports the use of these pigmented facultative methylotrophic isolates as a



potential biocontrol agents against tomato root pathogens but also expands our current knowledge on the taxonomy of facultative methylotrophic bacteria. Further field level performance testing of these methylotrophic isolates will confirm their biocontrol efficacy against root pathogens of tomato.

### AUTHOR CONTRIBUTIONS

VJ and VK—Isolated bacterial culture and conducted pot culture experiment; RA—conceived and designed the experiments; VJ and RA—analyzed the data; SK—Sequenced the bacterial culture; SUSU, TS, RK, KK, and SS—Manuscript preparation and editing.

### ACKNOWLEDGMENTS

This study was supported by Basic Science Research Program through the National Research Foundation (NRF) funded by the Ministry of Education, Science and Technology (2015R1A2A1A05001885), Republic of Korea.

### REFERENCES


### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016. 01626/full#supplementary-material

Figure S1 | Treatment details used in this experiments. Filled arrow indicates that the respective column contains the bacterial treatment. Dotted arrows indicates the challenge inoculation of pathogen.

Figure S2 | Colony morphology of pink pigmented facultative methylotrophic isolates in Ammonium mineral salts medium supplemented with 0.5% methanol incubated at 28 ± 2 ◦ C for 5 days. (A) Bacillus subtilis PPT-1; (B) Bacillus cereus PPB-1; (C) Delftia lacustris PPO-1.

Figure S3 | Amplification of *gyr A* gene. L1, Negative control; L2, Methylotrophic isolate PPB-1 L3, Methylotrophic isolate PPT-1; L4, Positive control (Bacillus subtilis MTCC 121); M, Marker (0.5–10 kb DNA Ladder).

Figure S4 | Influences on methylotrophs on plant growth and diseases control in pot culture experiment. T1—PPT-1 + F. oxysporum f. sp. lycopersici, T4—PPT-1 + S. rolfsii, T7—PPT-1 + R. solani, T12—PPO-1 + P. ultimum, T13—F. oxysporum f. sp. lycopersici T14—S. rolfsii, T15—R. solani, T16—P. ultimum.

Table S1 | Screening for antagonistic facultative methylotrophs against plant pathogens.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Janahiraman, Anandham, Kwon, Sundaram, Karthik Pandi, Krishnamoorthy, Kim, Samaddar and Sa. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Assessment of Culturable Tea Rhizobacteria Isolated from Tea Estates of Assam, India for Growth Promotion in Commercial Tea Cultivars

#### *Jintu Dutta1, Pratap J. Handique2 and Debajit Thakur1\**

*<sup>1</sup> Microbial Biotechnology Laboratory, Life Sciences Division, Institute of Advanced Study in Science and Technology, Guwahati, India, <sup>2</sup> Department of Biotechnology, Gauhati University, Guwahati, India*

#### *Edited by:*

*Kumar Krishnamurthy, Tamil Nadu Agricultural University, India*

#### *Reviewed by:*

*Biswapriya Biswavas Misra, University of Florida, USA Lei Zhang, North Carolina State University, USA*

> *\*Correspondence: Debajit Thakur debajitthakur@yahoo.co.uk*

#### *Specialty section:*

*This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology*

*Received: 11 September 2015 Accepted: 27 October 2015 Published: 10 November 2015*

#### *Citation:*

*Dutta J, Handique PJ and Thakur D (2015) Assessment of Culturable Tea Rhizobacteria Isolated from Tea Estates of Assam, India for Growth Promotion in Commercial Tea Cultivars. Front. Microbiol. 6:1252. doi: 10.3389/fmicb.2015.01252* In the present study, 217 rhizobacterial isolates were obtained from six different tea estates of Assam, India and subjected to preliminary *in vitro* plant growth promotion (PGP) screening for indole acetic acid (IAA) production, phosphate solubilization, siderophore production and ammonia production. Fifty isolates showed all the PGP traits and five isolates did not exhibit any PGP traits. These 50 potential isolates were further analyzed for quantitative estimation of the PGP traits along with the aminocyclopropane-1-carboxylate (ACC) deaminase, protease and cellulose production. After several rounds of screening, four rhizobacteria were selected based on their maximum ability to produce *in vitro* PGP traits and their partial 16S rRNA gene sequence analysis revealed that they belong to *Enterobacter lignolyticus* strain TG1, *Burkholderia* sp. stain TT6, *Bacillus pseudomycoides* strain SN29 and *Pseudomonas aeruginosa* strain KH45. To evaluate the efficacy of these four rhizobacteria as plant growth promoters, three different commercially important tea clones TV1, TV19, and TV20 plants were inoculated with these rhizobacteria in greenhouse condition and compared to the uninoculated control plants. Though, all the rhizobacterial treatments showed an increase in plant growth compared to control but the multivariate PCA analysis confirmed more growth promotion by TG1 and SN29 strains than the other treatments in all three clones. To validate this result, the fold change analysis was performed and it revealed that the tea clone TV19 plants inoculated with the *E. lignolyticus* strain TG1 showed maximum root biomass production with an increase in 4.3-fold, shoot biomass with increase in 3.1-fold, root length by 2.2-fold and shoot length by 1.6-fold. Moreover, two way ANOVA analysis also revealed that rhizobacterial treatment in different tea clones showed the significant increase (*P <* 0.05) in growth promotion compared to the control. Thus, this study indicates that the potential of these indigenous plant growth promoting rhizobacteria isolates to use as microbial inoculation or biofertilizer for growth promotion of tea crops.

Keywords: tea rhizosphere, plant growth promoting rhizobacteria, PGP traits, 16S rRNA, tea growth promotion

### INTRODUCTION

The rhizosphere is the narrow dynamic zone of soil influenced by plant roots where intense plant-microbe interaction is found. These microorganisms can have beneficial effects on the plant health like plant growth and nutrition in agro-ecosystems (Philippot et al., 2013). There are different distinct microbial communities present in the rhizosphere and one of them is well known as plant growth promoting rhizobacteria (PGPR). The PGPR is a group of beneficial soil bacteria associated with the plant roots which can promote plant growth both directly and indirectly (Glick, 1995). The PGPR may influence the plant growth directly by nitrogen fixation, different phytohormones production, phosphate solubilisation and sequester iron by siderophore production while indirectly stimulate the plant growth by producing antifungal metabolites by preventing different phytopathogens (Glick and Bashan, 1997). Therefore, PGPR are very important in improving plant growth and development or abiding in multiple biotic and abiotic stresses, hence the use of PGPR can help in developing eco-friendly practices for sustainable agriculture. A diverse range of bacterial genera such as *Arthrobacter, Azospirillum*, *Azotobacter*, *Bacillus*, *Pseudomonas*, *Klebsiella*, *Burkholderia*, *Erwinia*, *Flavobacterium*, *Micrococcous*, *Enterobacter*, *Xanthomonas*, *Chromobacterium*, *Serratia,* and *Caulobacter* have been documented to promote plant growth (Bhattacharyya and Jha, 2012; Bal et al., 2013).

In the recent years, PGPR got more attention and it has been used as potent biofertilizers (Richardson et al., 2009; Compant et al., 2010). The extensive use of chemical fertilizers has harmful effect on soil health by destabilizing soil fertility and beneficial microbial population (Kalia and Gosal, 2011). In Asia, most of the agricultural fields are highly fertilized with enormous quantities of chemical fertilizers and herbicides for enhancing crop production. Despite its efficiency, the long-term applications of such fertilizers have proved to be perilous to soil health as well as the human and also reduced the crops quality (Islam et al., 2013). Therefore, alternative biotechnological approaches are adapted in different agriculture practices to not only increase the crop production and plant growth, but also to maintain soil health (Fernando et al., 2005).

Though PGPR has been reported previously on different agricultural crops like rice (Sudha et al., 1999), tomato (Mena-Violante and Olalde-Portugal, 2007), wheat (de Freitas, 2000), maize (Biari et al., 2008), canola (Naderifar and Daneshian, 2012) but research of PGPR on tea plants is still poorly explored especially in the Northeast region of India. Tea plant [*Camellia sinensis* (L.) O. Kuntze] of family Theaceae plantation is one of the oldest organized practices in India with massive plantation in the Northeast corner of the agroclimatic belt. Assam is the largest tea producer state in India and it is divided into two main parts; Brahmaputra Valley and Barak Valley (Arya, 2013). The productivity of tea is decreased due to intensive application of chemical fertilizers for a prolonged period (Chakraborty et al., 2013). Hence there is a need to explore the indigenous microflora associated with the tea rhizosphere soil to not only reduce the use of chemical fertilizer but also for the benefit of plant and soil health.

The present study was designed to evaluate the potential of tea root associated bacteria isolated from six different tea estates of Assam, India for plant growth promotion (PGP). The most efficient rhizobacterial isolates were selected on the basis of their *in vitro* PGP experiments and characterized by 16S rRNA gene sequencing. The efficacy of these selected rhizobacterial isolates to use as potential biofertilizer was further evaluated by greenhouse experiment using three Tocklai vegetative (TV) tea clones TV1, TV19, and TV20 which are most extensively cultivated for commercial tea production in the tea estates of Northeast India. During the past years starting from 1949, 31 TV series tea clones and 153 location specific garden series clones have been released by Tocklai Tea Research Institute (TTRI), Tea Research Association (TRA), Jorhat, Assam, India to the tea industry for commercial cultivation. In Northeast India, 60% of the total tea growing areas are covered by planting materials released by TTRI (Das et al., 2012). TV1, TV19, and TV20 are the most popular TV series tea clones grown by the tea estates in Northeast India in terms of productivity and quality of made tea.

### MATERIALS AND METHODS

### Sample Collection and Isolation of Rhizobacteria

Tea rhizosphere soil samples were collected from six different tea estates of Assam, India, i.e., Sonapur tea estate (26◦06 56.40 N91◦58 33.18E), Khetri tea estate (26◦06 53.81N 92◦05 27.74E), Tocklai tea growing area (26◦45 18.40N94◦13 16.92E), Difaloo tea estate (26◦36 29.41N 93◦35 03.96E), Teok Tata tea estate (26◦36 29.41N 94◦25 42.59E) and Hathikuli tea estate (26◦34 55.94N 93◦24 43.15E) during January to April, 2013. The soil samples were collected from 5 to 30 cm depth along with the tea roots. To collect soil samples total area of each tea estate was divided into five blocks. In each block, samples were collected randomly from selected five adult tea plants containing roots and roots adhered soil. The samples collected from each block were mixed to make one composite sample resulting in a total of five composite samples. Soil samples were carried to the laboratory in sterile plastic bags stored at 4◦C to process for bacterial isolation. Bacterial isolation of collected samples was completed within 5 days of collection.

For rhizobacterial isolation, 1 g of the soil with roots were suspended in 100 ml of saline solution (NaCl 9 gl<sup>−</sup>1) and then kept in shaking condition (200 rpm) at 30◦C for 30 min. The soil suspension was then serially diluted up to 10–6 using sterile saline. Before plating, the samples were agitated at maximum speed using the vortex. An aliquot of 100 μl of each dilution was evenly spread over the surface of isolation media namely nutrient agar (NA), pseudomonas agar (PA), azotobacter media (AM) and azospirillum agar (AS) (HiMedia, India). Plates were incubated at 30◦C for 12–24 h. The rhizobacterial colonies appeared on the plates were counted. Further, the isolates were subculture and purified to store at −80◦C in 15% glycerol for a longer period. The gram staining of the bacterial isolates were performed and observed under light microscope (40X, Motic BA410 trinocular Microscope).

## Screening for *In vitro* Plant Growth Promoting (PGP) Traits

### Phosphate Solubilisation

To determine the phosphate solubilising activity, rhizobacterial isolates were spotted onto Pikovskaya's agar media (HiMedia, India). After 3 days of incubation at 28 ± 2◦C, isolates that induced clear zone around the colonies were considered as positive (Katznelson and Bose, 1959). Quantification of tri-calcium phosphate solubilization was carried out using ammonium-molybdate reagent (Fiske and Subbarow, 1925) in liquid Pikovskaya's medium. Absorbance values were obtained using the calibration curve with KH2PO4 at 650 nm for each isolates in 96-well microtiter plate by using multimode reader (Varioskan flash, Thermo Scientific, USA). The pH variation in the medium during the growth of each isolate was also observed.

### Indole Acetic Acid (IAA) Production

The production of IAA was determined by both qualitative and quantitative assay as described by Gordon and Weber (1951). The bacterial isolates were grown in minimal salt (MS) medium amended with 100 μg ml−<sup>1</sup> L-tryptophan for 72 h on the orbital shaker. The MS medium contained (per liter) 1.36 g KH2PO4, 2.13 g Na2HPO4, 0.2 g MgSO4.7H2O and trace elements. To measure the amount of IAA produced, bacterial supernatant was added to Salkowski's reagent (35% HClO4 containing 10 mM FeCl3) in 1:2 ratio. After 25 min, the samples were read at 530 nm absorbance by using 96-well microtiter plate in multimode reader (Varioskan flash, Thermo Scientific, USA). Development of pink or red color indicates IAA production. Concentration of IAA produced by the bacterial isolates was compared to the standard curve of commercial IAA (Sigma–Aldrich, USA).

### Siderophore Production

The siderophore of the bacterial isolates was determined by using chrome azurol S (CAS) agar plate assay. Briefly, inoculum (5 μl) of bacterial isolates were spotted onto the CAS agar plates and incubated at 28 ± 2◦C for 72 h. Siderophore production was assessed on the basis of change in color of the medium from blue to orange (Schwyn and Neilands, 1987). Quantitative estimation of siderophores was performed by CAS-shuttle assay in which 0.5 ml of culture supernatant was mixed with 0.5 ml of CAS reagent (Payne, 1994). Absorbance was measured at 630 nm by using 96-well microtiter plate in multimode reader (Varioskan flash, Thermo Scientific, USA) against a reference consisting of 0.5 ml of uninoculated broth and 0.5 ml of CAS reagent. Siderophore content in the aliquot was calculated by using the following formula:

%*siderophore units* = *Ar* − *As/Ar* × 100

Where, Ar = absorbance of the reference at 630 nm (CAS reagent) and As = absorbance of sample at 630 nm.

### Ammonia Production

Ammonia production was determined both qualitatively and quantitatively as described by Cappuccino and Sherman (1992). For estimation, freshly grown bacterial isolates were inoculated in test tubes with 10 ml of peptone water and incubated for 48 h at 28 ± 2◦C. After incubation, 1 ml of each culture was transferred to1.5 ml microtubes and 50 μl of Nessler's reagent [10% HgI2; 7% KI; 50% aqueous solution of NaOH (32%)] was added in each microtube. The development of faint yellow color indicates a small amount of ammonia and deep yellow to brownish indicates maximum production of ammonia. The production of ammonia was measured at 450 nm using standard curve of ammonium sulfate in the range of 0.1–5 μmol ml<sup>−</sup>1by multimode reader (Varioskan flash, Thermo Scientific, USA).

### Aminocyclopropane-1-carboxylate (ACC) Deaminase Activity

Aminocyclopropane-1-carboxylate deaminase acitivity was performed by qualitative assay using Dworkin and Foster (DF) salt medium (Dworkin and Foster, 1958) with 3 mM filter sterilized ACC as a sole nitrogen source (Sigma–Aldrich, USA). The 3 mM ACC solution was spread over the agar plates and bacterial isolates were spot inoculated on it. The bacterial growth on the plates was observed after 2 days incubation at 28◦C (Naik et al., 2008).

### Protease Activity

The protease activity was determined using skim milk agar medium which contains (per liter) 5 g pancreatic digest of casein, 2.5 g yeast extract, 1 g glucose, 7% skim milk solution and 15 g agar. Rhizobacterial isolates were spot inoculated and after 2 days incubation at 28◦C, proteolytic activity was identified by clear zone around the cells (Smibert and Krieg, 1994).

### Cellulase Activity

The rhizobacterial isolates were screened for cellulase production by plating onto M9 MS medium (HiMedia, India) agar amended with (per liter) 10 g cellulose and 1.2 g yeast extract. The bacterial colonies were observed after 8 days of incubation at 28◦C. The clear halos formed by the colonies considered as positive for cellulase production (Cattelan et al., 1999).

### Molecular Identification of Potent PGPR

For molecular identification of the potent strains, PCR amplification of 16S rRNA gene was performed using universal primers PA (5 -AGAGTTTGATCCTGGCTCAG-3 ) and 1492R (5 -GGTTACCTTGTTACGACTT-3 ) (Weisburg et al., 1991). For PCR reaction, a single bacterial colony was picked up with a sterile toothpick and mixed in the PCR cocktail.PCR cocktail (40 μl) contained 1XTaq DNA polymerase buffer, 1 U of Taq DNA polymerase, 0.2 mM of each dNTP, 1.5 mM MgCl2, 0.2 μM of each primer and a single bacterial colony for template DNA. Amplifications were performed in thermocycler (Proflex PCR system, Applied Biosystems, USA) programmed with an initial denaturation at 94◦C for 7 min, followed by 35 cycles of 30 s at 94◦C, 30 s at 54◦C and 1 min at 72◦C with an extension of 72◦C for 7 min. 5 μl aliquot of each amplification product was electrophoresed on a 1.7% agarose gel in 1XTAE buffer at 50 V for 45 min stained with ethidium bromide. PCR products were visualized under BioDoc-ItTM Imaging System (UVP, Cambridge, UK). The PCR product was purified by GenElute


TABLE 1 | The location, soil type and number of bacteria isolated from tea rhizosphere soil samples collected from the tea estate of Assam.

PCR clean up kit (Sigma–Aldrich, USA) and sent to Xcelris Genomics for sequencing located at Ahmedabad, India. The bacterial strains were identified by BLAST analysis. Calculation of the level of sequence similarity was performed using EzTaxon server 2.1 (Chun et al., 2007) and submitted to Genbank.

The sequences of 16S rRNA gene along with their closest homology sequences were aligned with the assist of multiple sequence alignment program CLUSTAL W (Higgins et al., 1992). The reference sequences were obtained from the Genbank database. Pairwise evolutionary distances were computed with the help of Kimura's 2 parameter model (Kimura, 1980). The phylogenetic tree was constructed by neighbour-joining (NJ) method using MEGA 6 program with bootstrap values based on the 1000 replications of the original dataset (Felsenstein, 1985).

## Plant Growth Promoting Experiment

### Plant Material

Six months old vegetatively propagated tea clones were obtained from TTRI, Tea Research Association, Assam, India and maintained in the greenhouse in polythene sleeves of size 15–17.7 cm layflat, 20–25 cm long and 150 gage thick. Three different types of tea clones, viz., TV1, TV19, and TV20 were used for the PGP experiments. TV1 is Assam-China hybrid standard clone, TV 19 is Cambod hybrid yield clone and TV 20 is Cambod hybrid standard clone (Das et al., 2012).

### Preparation of Bacterial Inoculums and Treatment

To prepare bacterial inoculum, the bacterial isolate was cultured in a 500 ml flask containing 200 ml nutrient broth (HiMedia, India) and allowed to grow aerobically in shaking incubator at 180 rpm for 48 h at 30◦C. The bacterial suspension was then diluted in sterilized distilled water to a final concentration of 108–109 cfu ml−1. The resulting suspensions were used to treat tea plants under greenhouse conditions. In greenhouse, the PGP experiment was performed in completely randomized design on three commercial tea clones TV1, TV19, and TV20. For each clone there were five treatments and an untreated control and each treatment contained three plant replicates. The bacterial inoculums (108–109 cfu ml−1) were applied in the form of soil drenching once in a month for three times. The soil of the tea sleeves were non-sterile. The plants were harvested after 6 months of inoculation. Shoot and root elongation, biomass production and number of tea leaves were compared with the uninoculated control plants.

### Data Analysis

The total isolates producing different PGP traits profile is represented as Venn diagram using the multiple dataset analysis feature of Vennture software (Martin et al., 2012). Plant growth promoting experiment of three different tea clones were conducted in a completely randomized design. The principal component analysis (PCA, is calculated based on eigen values and eigen vectors in a matrix) were performed on the datasets to evaluate the relationship between the samples to analyze the bacterial treatment for PGP. Moreover, fold change analysis was also performed in MetaboAnalyst 3.0 software (Xia et al., 2015) to compare the absolute value changes between two group means and output values are plotted in log2 scale, so that same fold change (up/down-regulated) will have the same distance to the zero baseline.

Further, the plant growth promoting datasets were subjected to two-way ANOVA using triplicate value to evaluate the significant difference in each plant growth promoting parameter between treated/inoculated and control (untreated/uninoculated plants) across the all three clones. The statistical analysis was performed with the SPSS software (SPSS 18.0, SPSS Inc., Chicago, IL, USA).

### RESULTS

### Isolation of Rhizobacteria from Tea Rhizosphere Soil

The rhizobacteria were isolated from tea rhizosphere samples in different isolation media used. A total of 217 bacteria were isolated from six different tea estates located in Assam. The number of rhizobacteria isolated from each tea garden and their cfu range were summarized in the **Table 1**. Moreover, the soil texture of soil samples collected was identified according to the USDA (2010) soil classification with their pH.

### Screening of Rhizobacteria for PGP Traits

All 217 isolates were preliminary screened for different PGP traits out of which 106 isolates showed phosphate solubilizing ability, 65 isolates showed siderophore production, 66 isolates were positive for indole acetic acid (IAA) and 164 were positive for ammonia production. However, based on the preliminary screening, 50 isolates showed positive result for all the four traits, i.e., siderophore production, phosphate solubilisation, IAA production, and ammonia production and five isolates did not exhibit any PGP traits. Hence, out of 217 isolates 212 isolates exhibited atleast one PGP trait and the data was depicted by the Venn diagram representation (**Figure 1**). The 50 isolates which showed all the PGP traits were selected as potential rhizobacteria and further estimated for siderophore production, IAA production, phosphate solubilisation and ammonia production. All the PGP traits tested for 50 isolates are summarized in **Table 2**.

### Estimation of Phosphate Solubilisation

For phosphate solubilisation, the isolates were observed up to 5 days from incubation and there was significant decrease in the pH level of the media. Most of the isolates showed good phosphate solubilising ability but 22 isolates were able to solubilize 150 μg ml−<sup>1</sup> or more calcium phosphate in the medium (**Figure 2**). Isolates KH45, TT6, TG1, and SN29 showed significantly higher phosphate solubilisation than the other isolates.

### Estimation of IAA, Siderophore and Ammonia Production

Out of the 50 isolates, 12 isolates produced 20 μg ml−<sup>1</sup> or more IAA after 72 h of incubation with 100 μg ml−<sup>1</sup> supplement of L-tryptophan. Siderophore production was estimated for all the selected 50 isolates. Only five isolates were able to produce above 30% of siderophore units. Most of the isolates showed ammonia production in the range of 4–4.5 μmol ml<sup>−</sup>1. Isolate KH49 showed the maximum ammonia production with 4.9 μmol ml−<sup>1</sup> among all the isolates. All the isolates producing IAA, siderophore, and ammonia are represented in **Figures 3–5**.

### Screening of ACC Deaminase, Protease, and Cellulase Assay

All the 50 isolates were further screened qualitatively for ACC deaminase, protease and cellulase production, out of which 29 (58%) isolates showed positive results for ACC deaminase activity, 25 (50%) isolates showed protease production and 21 (42%) isolates showed cellulase production. The results are shown in **Table 2**.

### Selection of Potent PGPR for Greenhouse Plant Growth Promoting Experiment

Four isolates from *in vitro* PGP screening, KH45, TT6, SN29, and TG1 showed higher amount of IAA production among all the isolates out of which TG1 produced maximum IAA which was 92.5 μg ml<sup>−</sup>1. These four isolates also showed significantly higher phosphate solubilisation as depicted in **Table 2**. Five isolates producing more than 30% units of the siderophore were KH45, TT6, SN29, HK23, and TG1. Though the isolate HK23 produced 33.8% siderophore unit but it produced less amount of IAA and lower levels of phosphate solubilisation compared to the other four isolates which were recorded for the best production of siderophore. Moreover, HK23 could not show positive result for all the PGP traits tested. Isolate KH49 showed the best ammonia production with 4.9 μmol ml<sup>−</sup>1. Isolate KH49 also produced less amount of IAA, siderophore and phosphate solubilisation compared to the isolates KH45, TT6, SN29, and TG1. Moreover, the isolate KH49 did not show positive result for all the *in vitro* traits tested. On the basis of these PGP screening, the isolates KH45, TT6, SN29, and TG1 were considered as most potent isolates as all the four showed very prominent results in all the PGP traits tested for this study. The efficacy of these four isolates was further evaluated by greenhouse plant growth promoting experiment on tea plants.

### Molecular Identification of the Potent PGPR

In order to determine the identity of four most potent PGP isolates, 16s rRNA gene was amplified by colony PCR method and sequences were obtained from Xcelris Genomics sequence service provider. The sequences were then compared using BLAST tool. Based on BLAST analysis of the 16S rRNA gene homology, the isolates were identified as *Pseudomonas aeruginosa* strain KH45, *Enterobacter lignolyticus* strain TG1, *Bacillus pseudomycoides* strain SN29 and *Burkholderia* sp. strain TT6. The sequences were deposited in the Genbank database under the accession numbers KJ767521, KJ767522, KJ767523, and KJ767524 (**Table 3**). A neighbor-joining dendogram was generated using the sequences from the four rhizobacteria and the representative sequences from the databases (**Figure 6**).

#### TABLE 2 | Screening of selected rhizobacterial isolates for *in vitro* plant growth promoting traits.


<sup>a</sup>*,*b*,*c*,*d*Values are mean of three replicates* <sup>±</sup> *SD.* <sup>e</sup>*,*f*,*<sup>g</sup> *"–" means showed no activity and "*+*" means showed activity present.*

### Plant Growth Promoting Experiment

The PCA analysis of the five different treatments for greenhouse PGP showed that the two groups of treatments TG1 and SN29 clustered together and other three groups of treatments TT6, KH45, and consortia appeared with the control in all the three TV1, TV19, and TV20 clones (**Figures 7A–C**).

To validate the PCA results, fold change analysis was performed to analyze the level of increase in the plant growth parameters of bacterial inoculated plants compared to the uninoculated control plants. In all the three tea clones TV1, TV19, and TV20, the bacterial inoculum of *E. lignolyticus* strain TG1 showed significant increase in number of fold change compared to the other bacterial inoculums. TG1 showed maximum root biomass production with an increase in 4.3-fold, shoot biomass with increase in 3.1-fold, root length by 2.2-fold and shoot length by 1.6-fold in TV19 clone. However, TG1 showed maximum number of leaves in TV20 clone by increase in 1.7-fold compared to the control plants. The graph of fold change analysis and their log values are provided as a Supplementary File 1: Figures S1A–O and Tables S2A–O.

Further, two-way ANOVA analysis was carried out to analyze the effect of treatment and clone on the growth parameters. In this study, three clones, i.e., TV19 as yield clone and TV1 and TV20 as standard clones were evaluated in greenhouse PGP experiment. All the plants inoculated with the isolates showed significant difference at *P <* 0.05 in all the growth parameters.

The tea plants of TV19 clone inoculated with *E. lignolyticus* strain TG1 showed maximum increase in the shoot and root length and biomass production (**Figures 8A–G**; see Supplementary File 2: Figures S2A–E and Supplementary File 3: Table S3). There was no significant difference in number of leaves comparing among all the three tea clones.

### DISCUSSION

In the present investigation, 217 rhizobacterial isolates from the tea rhizosphere soil were isolated and screened for different *in vitro* PGP traits out of which 50 isolates showed excellent *in vitro* PGP activity and five isolates did not exhibit any PGP traits analyzed. Further, based on the quantitative analysis of 50 isolates for wide array of PGP traits, four most promising rhizobacterial strains belongs to different genera were selected for further study. The 16S rRNA gene analysis of these four bacteria TG1, TT6, SN29, and KH45 has revealed 99% sequence similarity with *E. lignolyticus*, *Burkholderia* sp., *B. pseudomycoides* and *P. aeruginosa,* respectively. These bacterial isolates showing various *in vitro* PGP attributes multiple beneficial mechanisms of action and plant growth promoting efficacy was further evaluated in greenhouse experiment. These selected isolates showed excellent *in vitro* PGP activity like phosphate solubilisation, IAA production, siderophore production, and ammonia production. One of these beneficial characters is IAA synthesized by PGPR which is responsible for the increased adventitious root number and length and also the root volume by which roots can provide a large amount of nutrients to the plant and its higher root


TABLE 3 | Molecular identification of 16S rRNA gene, accession numbers and their origin of four selected potent rhizobacterial strains for greenhouse plant growth promotion (PGP).

exudates in turn benefits the bacteria (Ramos et al., 2008). In our study, *E. lignolyticus* strain TG1 was recorded as the best strain for IAA production (92.5 μg ml<sup>−</sup>1) that exceeds the IAA levels of previously reported works (Yasmin et al., 2007; Vasconcellos et al., 2010; Ribeiro and Cardoso, 2012). The PGPR, *E. cloacae* in rice (Mehnaz et al., 2001), *Enterobacter* sp. in sugarcane (Mirza et al., 2001) showed IAA production. Moreover, this strain solubilises inorganic phosphate up to 298.2 μg ml−<sup>1</sup> with significant decrease in pH. The solubilised inorganic phosphate is a promising attribute for the selection of bacteria capable of increasing available phosphorus in the rhizosphere. Liu et al. (2014) demonstrated that the most common and effective acids involved in the inorganic phosphate solubilisation are gluconic acid, lactic acid, malic acid, succinic acid, formic acid, citric acid, malonic acid, and tartaric acid. In our study, the solubilisation of calcium was observed due to significant decrease in pH of the culture medium which was probably responsible for production of organic acids by the bacteria. Phosphate solubilisation and siderophore production by different free-living rhizospheric bacteria, *Bacillus*, *Azotobacter,* and *Pseudomonas* was documented by Ahmad et al. (2008). Siderophore production is another important attribute for PGP. Rhizobacteria synthesize and release siderophore which are low molecular weight molecules that bind Fe3<sup>+</sup> that can chelate Fe3<sup>+</sup> and make it less available for other species in the microbial community of the rhizosphere. PGPR have been demonstrated to enhance the PGP very efficiently by producing extracellular siderophores which can control several plant diseases by depriving the pathogen of iron nutrition thus resulting in increased crop yield (O'Sullivan and O'Gara, 1992). The strain *B. pseudomycoides* showed the highest siderophore production (39.8%) among the four selected PGPR. Yu et al. (2011)reported, siderophore producing *B. subtilis* CAS15 strain showed growth promotion by stably colonized on pepper roots. Ammonia production also plays an important role

in plant growth by accumulation of nitrogen and helps in increase of root and shoots growth and biomass production (Marques et al., 2010). ACC deaminase production of PGPR elicits the growth promotion by decreasing the level of ethylene production. Though ethylene is an essential metabolite for plant growth but in different stress condition the ethylene level is increased drastically which could negatively influence the plant growth. ACC is an immediate precursor molecule for ethylene production and the bacteria producing ACC deaminase could hydrolyse ACC into ammonia and α-ketobutyrate instead of ethylene (Ahemad and Kibret, 2014). Cellulase and protease production by PGPR could also be possible plant growth promoting factors. Cellulose is found most abundantly in plant biomass which can be decomposed by cellulase enzyme produced by microbes (Lynd et al., 2002). Moreover, protease producing microbes can also act as a biocontrol agent on protein cell wall bearing pathogens. These cellulase and protease producing microbes play a significant role in decomposition of organic matter, nutrient mineralization, and PGP (Lima et al., 1998). The bearing of these traits implies that these microbes have potential to use as biofertilizer for crop yield.

All the four isolates selected in this study for greenhouse experiment on the tea plants showed significant increase in all the growth parameters evaluated. The individual inoculum of the bacterial isolates showed most promising effect on the plants rather than the consortia except strain KH45. The multivariate PCA analysis of the greenhouse parameters showed relation between the treatment and growth parameters of the plants. The PCA analysis of the plants inoculated with the *E. lignolyticus* strain TG1 and *B. pseudomycoides* strain SN29 showed significantly higher PGP compared to the other isolates. To validate these results, the fold change analysis was performed to analyze the level of fold change of growth parameters of inoculated plants compared to the uninoculated control plants. The fold change analysis revealed that the *E. lignolyticus* strain TG1 exhibited highest increase in fold change in root and shoot length and biomass production of TV19 clone compared to the control. The TG1 showed 4.3-fold increase in root biomass production and 2.2-fold increase in root length which could be correlated with the *in vitro* production of IAA. Rhizobacterial IAA is an important metabolite which helps in root development and we have found that the isolate TG1 showed maximum *in vitro* IAA production. The rhizobacterial IAA also increases root exudation by loosening the plant root cell walls which inturn helps in the rhizobacterial colonization and growth (Glick, 2012). These results implied that the isolate TG1 helps in tea root development and it could be used as a potential biofertilizer agent in tea crop fields. Further, the two-way ANOVA analysis also justified the PGP results analyzed by PCA and fold change analysis.

(

The effect of treatments on the PGP in different tea clones exhibited significant (*P <* 0.05) increase in growth promotion of tea plants except the number of leaves and the isolate TG1 showed the best growth promotion among all the isolates tested.

The genus *Enterobacter* was previously reported for PGP. *E. arachidis* was reported as a novel species of genus *Enterobacter* from rhizosphere soil of groundnut showing PGP activity (Madhaiyan et al., 2010). *E. asburiae* was isolated from mustard rhizosphere soil which was reported as fungicide tolerant and endowed with PGP traits (Ahemad and Khan, 2010). Madhaiyan et al. (2013) reported the complete genome sequence of *Enterobacter* sp. strain R4-368, an endophytic N-fixing γ-proteobacterium isolated from surface sterilized roots of *Jatropha curcas* L. showing strong growth-promoting activity by increasing plant biomass and seed yields. Also, *E. radicincitans* DSM16656 isolated from the phyllosphere of winter wheat has been shown to promote the growth of several plant species (Witzel et al., 2012). Although, there were several reports in genus *Enterobacter* as PGPR but there has been no evidence reported for the *E. lignolyticus* as plant growth promoting agent in the tea crop field.

### CONCLUSION

In the present study, effective isolates of rhizobacteria from tea rhizosphere soil was selected which could be a useful component of sustainable agricultural management. Although, all four rhizobacteria have been demonstrated for their PGP potential in tea plant, isolate *E. lignolyticus* strain TG1 was found to have superiority over other isolates in terms of growth

### REFERENCES


promotion. Hence, these indigenous rhizosphere associated soil microbial inhabitants with wide array of PGP activity could be beneficial for tea plantation of Northeast India. However, further experiments are needed to determine the effectiveness of these rhizobacterial isolates under different field conditions to study the nature of interaction with other soil native microflora and the host plant.

### AUTHOR CONTRIBUTIONS

DT supervised the research work and guided the experimental design. PJH provided the research work suggestion. JD performed the laboratory and field experiments, analysed the data. DT and JD prepared the manuscript.

### ACKNOWLEDGMENTS

This work was supported by the Department of Biotechnology (DBT), Govt. of India under RGYI scheme (Grant No. BT/PR6011/GBD/27/379/2012). The authors wish to thank Director, IASST, Guwahati, Assam, India, for providing facilities for this work.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal*.*frontiersin*.*org/article/10*.*3389/fmicb*.* 2015*.*01252


the rhizosphere of banana. *Microb. Ecol.* 56, 492–504. doi: 10.1007/s00248-008- 9368-9


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2015 Dutta, Handique and Thakur. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Calcite Dissolution by Brevibacterium sp. SOTI06: A Futuristic Approach for the Reclamation of Calcareous Sodic Soils

#### S. M. Tamilselvi <sup>1</sup> , Chitdeshwari Thiyagarajan<sup>2</sup> and Sivakumar Uthandi <sup>1</sup> \*

*<sup>1</sup> Biocatalysts Lab, Department of Agricultural Microbiology, Tamil Nadu Agricultural University, Coimbatore, India, <sup>2</sup> Department of Soil Science and Agricultural Chemistry, Tamil Nadu Agricultural University, Coimbatore, India*

#### Edited by:

*Kumar Krishnamurthy, Tamil Nadu Agricultural University, India*

### Reviewed by:

*Balasubramanian Ramakrishnan, Indian Agricultural Research Institute, India G. Selvakumar, ICAR- Indian Institute of Horticultural Research, India*

> \*Correspondence: *Sivakumar Uthandi usivakumartnau@gmail.com*

#### Specialty section:

*This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science*

Received: *01 July 2016* Accepted: *21 November 2016* Published: *08 December 2016*

#### Citation:

*Tamilselvi SM, Thiyagarajan C and Uthandi S (2016) Calcite Dissolution by Brevibacterium sp. SOTI06: A Futuristic Approach for the Reclamation of Calcareous Sodic Soils. Front. Plant Sci. 7:1828. doi: 10.3389/fpls.2016.01828* Assessing the ability of soil microorganisms to dissolute poorly soluble native calcite to supply Ca2<sup>+</sup> is a new area to be explored in reclaiming sodic soils by supplying adequate Ca2<sup>+</sup> and reducing the recurrent sodicity. Hence, the present study aimed to isolate a calcite dissolving bacteria (CDB) from calcareous sodic soils and to understand the mechanism of calcite dissolution. Of the 33 CDB isolates recovered from the calcareous sodic soils of Tamil Nadu (Coimbatore, Ramnad, and Trichy), 11 isolates were screened for calcite dissolution based on titratable acidity. 16S rRNA gene sequence analysis of the three best isolates *viz.,* SORI09, SOTI05, and SOTI06 revealed 99% similarity to *Bacillus aryabhattai*, 100% to *B. megaterium,* and 93% to *Brevibacterium* sp., respectively. Among them, *Brevibacterium* sp. SOTI06 released more Ca2<sup>+</sup> (3.6 g.l−<sup>1</sup> ) by dissolving 18.6% of the native calcite. The spectral data of FTIR also showed reduction in the intensity of calcite (55.36–41.27) by the isolate at a wave number of 1636 cm−<sup>1</sup> which confirmed the dissolution. Besides producing organic acids (gluconic acid and acetic acid), *Brevibacterium* sp. SOTI06 also produced siderophore (91.6%) and extracellular polysaccharides (EPS, 13.3 µg. ml−<sup>1</sup> ) which might have enhanced the calcite dissolution.

Keywords: calcite dissolution, Brevibacterium sp., in-vitro analysis, calcareous soils, sodicity reclamation

## INTRODUCTION

Soil degradation due to sodicity is the widest stress observed worldwide since the presence of high Na<sup>+</sup> concentration increases the inter particulate distances by enhancing the repulsive forces and therefore causes dispersion and loss of porosity, which consequently results in undesirable soil structure and reduced water permeability in the soil profile. Many of these soils are highly deficient in plant nutrients due to high pH, exchangeable Na+, carbonates and bicarbonates, as a consequence crop production in these soils is also very poor (Murtaza et al., 2013; Tazeh et al., 2013). Hence, reclamation of these soils necessitates the removal of excess soluble Na<sup>+</sup> from the soil to facilitate better crop growth.

Most of the sodic soils are calcareous in nature contains inherent or precipitated sources of Ca2<sup>+</sup> in the form of calcite within the soil profile and such soils are widely spread in arid and semi arid regions. Calcite dissolution results in the release of Ca2<sup>+</sup> ions to the soil solution (Qadir et al., 2007) which replace Na<sup>+</sup> as detailed below (Qadir et al., 2005).

$$\text{2Na}^+ - \text{Clay } + \text{Ca}^{2+} \Leftrightarrow \text{Ca}^{2+} - \text{Clay } + \text{ 2Na}^+ \tag{1}$$

Therefore, reclamation of calcareous sodic soils is possible when suitable amendments were identified and used at appropriate amounts. Generally amelioration of these soils has been achieved through the application of chemical amendments like gypsum (Abdel-Fattah, 2012; Cucci et al., 2012) as a direct source to supply sufficient Ca <sup>2</sup><sup>+</sup> for exchanging Na+. However, high cost and recurrent sodicity necessitates in finding out alternate sources and strategies. Phyto-remediation, a low cost technology involving different crops like kallar grass, sesbania, cotton, and halophytes like Aster sp., Atriplex sp., and Plantago sp. (Murtaza et al., 2009; Hasanuzzaman et al., 2014) helps to certain extent in lowering the sodicity but requires suitable plants, several growing seasons, and act only at limited depths (USEPA, 2000). Recently, microbial mediated calcite dissolution is gaining acceptance to reduce the sodicity.

However, most of the calcite dissolution mechanism has been studied without microorganisms (MacInnis and Brantley, 1992; Newton and Manning, 2002; Cucci et al., 2012) and only a very few reports have focused on the calcite dissolution by microorganisms (Lüttge and Conrad, 2004; Li et al., 2005; Jacobson and Wu, 2009; Subrahmanyam et al., 2012; Cacchio et al., 2014). Several mechanisms were reported for the extent of calcite dissolution such as acidification (Whitelaw et al., 1999) by producing organic acids (Goldstein, 1995; Fasim et al., 2002; Chen et al., 2006), inorganic acids (Hopkins and Whiting, 1916), chelating substances (Liermann et al., 2000; Yoshida et al., 2002), EPS (Yi et al., 2008), etc. Despite many reports on the mechanism of calcite dissolving microorganisms, it mainly centered around the production of organic acids like acetic acid, lactic acid, propionic acid, pyruvic acid, and succinic acid (Garcia-Pichel, 2006; Sulu-Gambari, 2011), enzymes like phosphatase (Ehrlich et al., 2008), EPS (Bissett et al., 2011) but none of them revealed the quantitative data on calcite dissolution. Hence, the present investigation aimed to isolate, identify an efficient CDB and measure their in-vitro calcite dissolution ability with an intention of using them for bio-remediating the calcareous sodic soils.

### MATERIALS AND METHODS

### Materials

Organic acids were from Sigma-Aldrich, India (Bengaluru) and other organic, inorganic analytical grade chemicals and agarose were from HI-Media Laboratories Pvt. Ltd. (Mumbai). Molecular biology chemicals were from New England Biolabs (Gurgaon, India) and Takara India (New Delhi).

### Media and Cultivation Conditions

Unless and otherwise stated all the culture conditions were performed in 100 ml of DB (Devenze-Bruni) medium in 250 ml Erlenmeyer flasks (with final OD600 nm of 0.1) containing CaCO<sup>3</sup> (5 g.l−<sup>1</sup> ) and incubated at 30◦C under shaking at 120 rpm for 24 h. The cell free culture supernatant obtained by centrifugation at 8000 g for 15 min was used for analysis of pH, TA, Ca2+, CaCO3, CO2<sup>−</sup> 3 , HCO<sup>−</sup> 3 , acid phosphatase, organic acid, EPS, biofilm, and siderophore.

## Isolation, Screening, and Identification of Calcite Dissolving Bacteria

### Soil Sampling and Enrichment

Calcareous sodic soil samples collected from three districts of Tamil Nadu, India viz., Coimbatore (Altitude of 411 m above mean sea level, 11.0◦N latitude and 76.9◦E longitude), Ramnad (Altitude of 2 m, 9.3◦N latitude and 78.8◦E longitude), and Trichy (Altitude of 85 m, 10.7◦N latitude and 78.7◦E longitude), showed the free CaCO<sup>3</sup> concentration of 7.2, 7.6, and 7.8%, respectively and were stored at 4◦C. In order to isolate CDB, 100 g of each soil was enriched with 1% CaCO<sup>3</sup> individually and incubated for 2 weeks. Along with enriched soil samples, native, or initial soil samples were also used for the isolation of CDB.

### Isolation and Screening of CDB Isolates

The CDB were isolated from both enriched and initial soil samples by serial dilution and plating technique using DB agar medium consisting of g.l−<sup>1</sup> Glucose 5; Yeast extract 1; Peptone 1; K2HPO<sup>4</sup> 0.4; MgSO<sup>4</sup> 0.01; NaCl 5; (NH4)2SO<sup>4</sup> 0.05; CaCO<sup>3</sup> 5 and Agar 20 (Cacchio et al., 2004). The CD positive isolates picked based on clear zone formation around the colony were further confirmed by point inoculation onto the same medium. The solubilization index (SI) of the individual isolates was determined by measuring the ratio of the clear zone and colony size on DB agar plate by using the following formula:

$$\text{Solubility index} = \frac{\text{Clear zone} + \text{Colony size}}{\text{Colony size}} - \text{---[F1]}$$

$$\{\text{Mihalache et al., 2015}\}$$

Secondary screening of positive isolates was carried out by calculating titratable acidity (TA) from 24 h old cultures grown in DB liquid medium. One milliliter of the cell free culture supernatant was titrated against 10 mM NaOH in the presence of phenolphthalein indicator until the appearance of pink color (Whitelaw et al., 1999).

**Abbreviations:** −, Absence; +, Presence; µl, Microliter; 16S rRNA, Ribosomal Ribo Nucleic Acid; Amp-X-gal-IPTG, Ampicillin, 5-bromo-4-chloro-3-indolyl-β-D-galactopyranosid and Isopropyl β-D-1-thiogalactopyranoside; Ar, Absorbance of reference; As, Absorbance of sample; FT-IR, Fourier Transformation Infra-Red spectrophotometer; C, Control ; Ca2+, Calcium ions; CaCO3, Calcium carbonate; CAS, Chrome Azurol S; CD, Calcite dissolution; CDB, Calcite dissolving bacteria; cm, Centimeter ; CO2<sup>−</sup> 3 , Carbonate ; CS, Colony size; CZ, Clear zone; DB, Devenze-Bruni; DNA, Deoxyribo Nucleic Acid; EPS, Exopolysaccharide; g, Gravity; g.l−<sup>1</sup> , Gram per litre; H2SO4, Sulphuric acid; HCl, Hydrochloric acid; HCO<sup>−</sup> 3 , Bicarbonate; HPLC, High Performance Liquid Chromatography; IAA, Indole acetic acid; M, Molarity; min, Minute; ml, Millilitre; Na+, Sodium ion; NaCl, Sodium chloride; NaOH, Sodium hydroxide; nm, Nanometer; OD, Optical density; PCR, Polymerase chain reaction; rpm, Rotation per minute; SI, Solubilization index; SOCE, Sodic soil Coimbatore district Enriched sample; SOCI, Sodic soil Coimbatore district Initial sample; SORE, Sodic soil Ramnad district Enriched sample; SORI, Sodic soil Ramnad district Initial sample; SOTE, Sodic soil Trichy district Enriched sample; SOTI, Sodic soil Trichy district Initial sample; T, Treated; TA, Titratable acidity; UV-VIS, Ultra violet-visible; Zn, Zinc; ZnO, Zinc oxide.

### Identification of CDB Isolates by 16S rRNA

Total genomic DNA of the selected isolates were extracted and purified using the method described by Clark (2013). CDB isolates were identified by amplification of 16S rRNA gene using 27F (5′ AGAGTTTGATCCTGGCTCAG 3′ ) and 1492R (5′ GGTTACCTTGTTACGACTT 3′ ) primers with the PCR conditions of initial denaturation at 95◦C for 10 min followed by 35 cycles of denaturation at 94◦C for 30 s, annealing at 55◦C for 30 s and extension at 72◦C for 1 min, followed by a final extension at 72◦C for 15 min in a thermo cycler (BioRad, USA). Then, the PCR products were cloned into the pGEMT vector and transformed into chemically competent E. coli DH5α cells (Sambrook et al., 1989). Positive clones were selected based on bluewhite screening from Amp-X-gal-IPTG plates and further confirmed by colony lysis PCR using M13 forward (5′ GTAAAACGACGGCCAGT 3′ ) and reverse primers (5′ AACAGCTATGACCATG 3′ ). The positive clones were sequenced [Bioserve Biotechnologies (I) Pvt. Ltd., Hyderabad, India]. 16S rRNA gene sequence obtained for each clone was aligned and compared with available sequences of bacterial lineage using Ez Taxon-e (http://eztaxon-e.ezbiocloud.net/). A phylogenetic tree was constructed using MEGA 6 program (Tamura et al., 2013) and their grouping sequence was based on confidence values obtained by bootstrap analysis of 1000 replicates.

### Surface Attachment of Brevibacterium sp. SOTI06

### Biofilm (Planktonic) Formation

One day old Brevibacterium sp. SOTI06 culture (0.1 ml) was taken into 96 well micro titre plate, covered and incubated at 30◦C for 24 h. After incubation, the plates were washed thoroughly with sterile distilled water and air dried. One hundred and fifty microliters of 0.1% crystal violet was added to each well and incubated for 45 min. The excess stain was removed by sterile distilled water and air dried. Subsequently, 200µl of 95% ethanol was added to each well and plates were incubated for 10–15 min. Contents of each well were mixed and 125µl of the crystal violet/ethanol solution was transferred to a separate clear bottom well and optical density was measured at 600 nm using micro plate reader (Molecular Devices LLC, USA; Djordjevic et al., 2002).

### EPS Production

Brevibacterium sp. SOTI06 was cultured in DB liquid medium supplemented with CaCO<sup>3</sup> (5 g.l−<sup>1</sup> ) and incubated at 30◦C for 24 h at 120 rpm. The culture was centrifuged at 4000 g for 15 min and the pellet was used for estimation of EPS by suspending the pellet with 5 ml distilled water and 5 ml 0.1 N KOH. The contents were boiled at 100◦C for 10 min. After cooling, the suspension was neutralized with 1M HCl and 1 ml of suspension, was mixed with 5 ml Anthrone reagent and the intensity of color was measured at 620 nm in UV-VIS spectrophotometer (Systronics, India, DuBois et al., 1956).

## Calcite Dissolution (CD) Potential of Brevibacterium sp. SOTI06

### Estimation of Dissolution

In order to quantify CD ability of Brevibacterium sp. SOTI06 grown in DB liquid medium, cell free culture supernatant obtained at periodical intervals were subjected to the analysis of calcium (Jackson, 2005), calcium carbonate (Piper, 1944), carbonates, bicarbonates (Richards, 1954), phosphatase (Tabatabai and Bremner, 1969), protein concentration (Bradford, 1976), pH, and TA.

### Quantification of Organic Acid Production

Organic acid production was estimated from 24 h old culture by injecting 30µl of 0.2µm filtered cell free supernatant in HPLC with a UV detector set at 210 nm. The organic separation was carried out on Cosmosil packed column (Nacalai Tesque, Japan) with 10.8% Acetonitrile in 0.0035 M H2SO<sup>4</sup> as mobile phase at a flow rate of 0.6 ml.min−<sup>1</sup> (Chen et al., 2006). The data integration and analysis was done using Autochrom software. HPLC grade organic acids kit (No.47264 from Sigma Aldrich, USA) was used as standards.

### Analysis of CD by ATR-FT-IR

FT-IR spectrum of CaCO<sup>3</sup> in the spent medium by Brevibacterium sp. SOTI06 was recorded in JASCO FT-IR 6800 fitted with diamond enabled Attenuated Total Reflectance (ATR) sample holder and a DLaTgs detector and compared with CaCO3. The wavelength range was from 400 to 4000 cm−<sup>1</sup> . Spectral measurements were done in triplicates and 64 scans were recorded for all samples at a 4 cm−<sup>1</sup> resolution.

### Siderophore Production

Siderophore production of Brevibacterium sp. SOTI06 was observed by point inoculation with fresh culture onto Chrome Azural S (CAS) agar plate and incubated for 48 h at 30◦C (Schwyn and Neilands, 1987), which was further confirmed by broth assay. The assay was carried out by mixing the culture supernatant (0.5 ml) with 0.5 ml CAS reagent and the absorbance was measured at 630 nm against a reference consisting of uninoculated liquid medium. Siderophore content was estimated using the formula:

$$\text{Per cent siderophore units} = \text{(Ar} - \text{As/Ar)} \times 100 - -- - \text{[F2]}$$

$$\{\text{Payne, 1994}\}$$

Where, Aris the absorbance of reference and As is the absorbance of sample.

### Statistical Analysis

All the data were subjected to statistical analysis in Microsoft Excel (Windows 2007) add-in with XLSTAT version 2010.5.05 (XLSTAT, 2010).

### RESULTS

### Isolation, Screening, and Identification of CDB Isolates

A total of 33 isolates (17 from native and 16 from enriched soils) showing clear zone (**Figure 1**) in DB medium was evaluated for calcite solubilization index (SI) which varied from 0.37 to 6.67. Among the three soils, SI values were higher with the isolates from Trichy soil than in Coimbatore and Ramnad soils. Higher SI values were observed in native isolates (0.88–6.67) compared to isolates from enriched soils (0.37–2.33). Among the isolates, SOTI06 showed maximum calcite SI (6.67) followed by SORI01 (3.70) and SOTI05 (2.10) which were from initial soils. On the other hand, maximum SI of 2.33 was observed in enriched isolate SOCE29. The least SI was recorded for the isolates SOCE22 and SOCE33 (**Figure 2**). Top 11 isolates having the highest SI were evaluated for TA production ability. Among them, eight isolates produced TA in the range of 0.05–0.12 g.l−<sup>1</sup> and three isolates, SORI09, SOTI05, and SOTI06 produced maximum TA of 0.81, 0.60, and 1.41 g.l−<sup>1</sup> , respectively (**Figure 3**).

Identification of the three promising isolates based on 16S rRNA gene sequence revealed that SOTI06 showed 93% similarity to Brevibacterium halotolerans DSM 8802 as their closest organism. SOTI05 showed 100% similarity to Bacillus megaterium NBRC 15308 and SORI09 showed 99% similarity to Bacillus aryabhattai B8W22, respectively (**Figure 4**). Genbank accessions for 16S rRNA gene sequence of these isolates, SOTI06, SOTI05, and SORI09 were KX443712, KX443711, and KX443710, respectively.

### Surface Attachment of Brevibacterium sp. SOTI06

Microbial mediated calcite dissolution starts with surface attachment of the bacteria by means of biofilm formation and EPS production subsequently the mineral dissolution by secreting organic acids, siderophore, and phosphatase. Brevibacterium sp. SOTI06 was able to form higher amount of biofilm when supplemented with CaCO<sup>3</sup> than medium without CaCO<sup>3</sup> which was evidenced with the increase in OD600 nm of former (0.21) than later (0.16). Similarly,

FIGURE 1 | Clear zone formation by Brevibacterium sp. SOTI06. The bacterium forms a clear zone around the colony on DB medium in the presence of CaCO3 (A) indicating calcite dissolution was compared with control plate (B).

the production of EPS was higher (13.3µg.ml−<sup>1</sup> ) in the medium supplemented with CaCO<sup>3</sup> than in control (4.39µg.ml−<sup>1</sup> ).

### Calcite Dissolution Estimation of Dissolution

The calcite dissolution behavior of Brevibacterium sp. SOTI06 was estimated over a period of 5 days by measuring the pH, TA, phosphatase, CaCO3, Ca2+, CO2<sup>−</sup> 3 , and HCO<sup>−</sup> 3 content in the medium. The results revealed that pH decreased gradually from 8.02 to 5.72 until 4th day and there after increased to 6.60 on 5th day. On contrary, TA and phosphatase activity showed an increasing trend from 0 to 4th day and decreased later. Production of TA started on 1st day (0.93 g.l−<sup>1</sup> ) and almost doubled on 2nd day (1.63 g.l−<sup>1</sup> ), further an increment in TA was noticed up to 4th day (1.95 g.l−<sup>1</sup> ) and suddenly dropped to 1.33 g.l−<sup>1</sup> at 5th day. But, the phosphatase activity was linearly increased from 1st day (19.7 U.ml−<sup>1</sup> ) onwards, reaching maximum up to 4th day (91.8 U.ml−<sup>1</sup> ) and reduced to 70.7 U.ml−1on 5th day (**Figure 5A**). The protein concentration was also increased from 1st day (10.0 g.l−<sup>1</sup> ) to 4th day (34.4 g.l−<sup>1</sup> ) and declined on 5th day (26.7 g.l−<sup>1</sup> ).

The supplemented calcium carbonate content was slowly decreased from 5.0 to 4.07 g.l−<sup>1</sup> with simultaneous increase in calcium and bicarbonate concentrations. The release of Ca2+into the solution was higher than bicarbonate ions. A minimal amount of calcium (0.04 g.l−<sup>1</sup> ) and no bicarbonates were released on 0th day and thereafter, the release of calcium content was higher until 5th day (3.60 g.l−<sup>1</sup> ). However, the bicarbonate content was increased until 4th day reaching the maximum of 0.65 g.l−<sup>1</sup> and decreased later. Overall, Brevibacterium sp. SOTI06 was capable of dissolving 18.6% of calcite within 5 days of incubation (**Figure 5B**).

### Organic Acid Production

Supplementation of CaCO<sup>3</sup> to Brevibacterium sp. SOTI06 resulted in the production of organic acids such as gluconic acid, acetic acid, fumaric acid, and phytic acid. Among the secreted organic acids, gluconic acid was the predominant one (3.24 mg.ml−<sup>1</sup> ) followed by acetic acid (3.17 mg.ml−<sup>1</sup> ). A minimal amount of phytic acid (10µg.ml−<sup>1</sup> ) and fumaric acid (7µg.ml−<sup>1</sup> ) was also recorded in the medium enriched with CaCO<sup>3</sup> (**Figure 6**). Conversely, the medium without CaCO<sup>3</sup> resulted in lesser production of acetic acid (0.92 mg.ml−<sup>1</sup> ) and fumaric acid (0.25µg.ml−<sup>1</sup> ) whereas; the release of predominant gluconic acid was not observed (data not given).

### FT-IR Analysis

FT-IR spectra of calcite dissolution by Brevibacterium sp. SOTI06 was compared with uninoculated control (**Figure 7**) and the spectral data showed changes in vibration and alteration of structure with reduced intensity (55.36–41.27%) which confirmed the calcite dissolution by bacterium (1636 cm−<sup>1</sup> ). Further, the presence of additional two new peaks at wave number of 1222 and 1370 cm−<sup>1</sup> with strong OH groups was observed in treated sample (**Table 1**).

### Siderophore Production

The development of yellow halo around the colonies in CAS plate was confirmed by broth assay showed that Brevibacterium sp. SOTI06 produced siderophore both in CaCO<sup>3</sup> amended as well as unamended liquid medium was evidenced by a change of color from blue to yellow (**Figure 8**). But, their production was higher in amended medium registering 91.6 per cent siderophore units than the control (88.6%).

### DISCUSSION

Salt affected soils are wide spread in many arid and semiarid regions which are the major constraints for agricultural

substitutions/1000 bases.

pH was decreased from day 0 to 4. Whereas, the Ca2<sup>+</sup> and HCO<sup>3</sup> content increased over a period and a decrease in CaCO<sup>3</sup> content indicated that dissolution occurred by the bacterium. Means of three replicate values plotted and error bars indicate the standard error.

expansion and productivity. The main reason for the increased sodicity is due to faulty irrigation and drainage practices which leads to soil degradation and ultimately reduces crop yield (Sumner, 1993; Sharma and Rao, 1998; Haynes and Hamilton, 1999; Gharaibeh et al., 2011). In order to reduce the sodicity in calcareous soils, the native calcite need to be dissolved to release adequate Ca2<sup>+</sup> so as to replace the Na<sup>+</sup> ions, which can be leached out through irrigation (Oster, 1982; Shainberg et al., 1989; Qadir and Oster, 2002). Microbial mediated calcite dissolution studies are very sparse in the literature

for instance, calcite and dolomite dissolution was studied in Shewenella oeindeisis MR1 (Davis et al., 2007), Bacillus subtilis, and Burkholderia fungorum (Friis et al., 2003; Jacobson and Wu, 2009). Recently, Brevibacterium sp. was isolated from Krast caves and reported their calcite dissolution ability (Sonntag, 2015). Hence, it is imperative to develop calcite dissolving microbes and understanding its mechanisms of dissolution so as to reclaim the calcareous sodic soils effectively. In this contest, the present


) Functional group Bond Intensity Mode

TABLE 1 | FTIR spectrum of Brevibacterium sp. SOTI06.

Sample Wavenumber (cm−<sup>1</sup>

FIGURE 8 | Siderophore production by Brevibacterium sp. SOTI06. The siderophore production was estimated by CAS assay with reference (A) in the presence (B) and absence (C) of CaCO3. The change of color from blue to yellow indicated the siderophore production.

study on isolation, screening and identification of CDB and understanding the mechanism underpinning calcite dissolution is significant.

The present investigation indicated SOTI06 as the best isolate based on the calcite dissolving ability and titratable acidity. The 16S rRNA gene sequence of the newly isolated bacterial isolate SOTI06 was analyzed to establish its phylogenetic relationship, which showed only 93% similarity with B. halotolerans strain DSM 8802 suggesting that this isolate might be a new one and needs further systematic and taxonomical studies to reveal its novelty.

Calcite dissolution is regulated by a wide range of molecules like organic acids, amino acids and these molecules inhibit calcite growth thereby promoting dissolution (Teng et al., 2006). The growth, planktonic form of biofilm formation and EPS production are the mechanisms by which microorganisms attached to the mineral surface (Banfield et al., 1999; Kraemer, 2004; Peacock et al., 2004; Buss et al., 2007; Yi et al., 2008; Shirvani and Nourbakhsh, 2010; Parrello et al., 2016) and helps dissolution. In the present study, the CDB isolate Brevibacterium sp. SOTI06 produced considerable amount of EPS and planktonic form of biofilm in the presence of calcite suggesting its possible attachment to dissolute calcite as evidenced by Bissett et al. (2011).

A reduction in pH as induced by the production of TA by Brevibacterium sp. SOTI06 determines the solubility of minerals (Whitelaw et al., 1999; Ogbo, 2010; Barroso and Nahas, 2013). The trend of decrease in calcium carbonate content and increase in Ca2<sup>+</sup> supply from first day onwards indicates that the dissolution process initiated upon inoculation as evidenced from the gluconic acid production by Brevibacterium sp. SOTI06 and suggests that the gluconic acid might be the predominant one involved in calcite dissolution. The increase in calcium carbonate content on fifth day might be due to precipitation and these results are in accordance with Subrahmanyam (2013). In multicellular organisms like sponges, the cellular attachment on mineral surface, penetration, and dissolution of calcareous substrates are mediated by many enzymatic activities particularly carbonic anhydrase and acid phosphatases (Kreitzman and Fritz, 1970). In the present CD experiment, the enhanced acid phosphatase activity coupled with drop in pH, CaCO3, and release of Ca2<sup>+</sup> explains the role of this enzyme on supplying calcium through effective calcite dissolution.

The FT-IR results suggest that bacterial dissolution might have altered the structure of calcite and resulted in vibration change. Since the Brevibacterium sp. SOTI06 secreted gluconic acid and other acids, which might have facilitated the release of Ca2<sup>+</sup> from calcite and results in overall mass reduction. Such reduction was evident in the FT-IR spectra of treated samples and also the results with measurement of Ca2<sup>+</sup> supported this phenomenon. The presence of two additional peaks with strong OH groups might be attributed to acids secreted by Brevibacterium sp. SOTI06. From the FT-IR study, the dissolution behavior and acid secretion of Brevibacterium sp. SOTI06 was confirmed.

Though, the present study showed calcite dissolution of the isolate under in vitro condition, the potential of this bacterium is yet to be evaluated in detail under in vivo condition to remediate the calcareous sodic soils.

### CONCLUSION

The present study reported a calcite dissolving Brevibacterium sp. SOTI06 with a potential to dissolute 18% calcite with a simultaneous release of Ca2<sup>+</sup> ions under in vitro conditions. Gluconic acid production, biofilm formation, production of siderophore, and EPS by Brevibacterium sp. SOTI06 might be the possible mechanisms attributed to the dissolution of calcite.

### REFERENCES


### AUTHOR CONTRIBUTIONS

SU and CT conceived the idea and designed experiments. ST conducted the experiments, analyzed the data and helped in drafting the manuscript. SU finalized the results after compiling data and completed the manuscript preparation.

### ACKNOWLEDGMENTS

Fellowship support from Department of Biotechnology, Ministry of Science and Technology, Government of India, New Delhi for the scheme "Bioremediation of degraded calcareous Sodic and Saline-Sodic soils" (BT/PR7187/BCE/8/935/2012) offered to ST is gratefully acknowledged. The authors also acknowledge the additional financial support by the Ministry of Human Resource Development (MHRD-FAST CoE) (F.No.5-5/2014-TSVII), GOI, New Delhi.


column-built leached soil-limestone karst systems. Appl. Soil Ecol. 29, 274–281. doi: 10.1016/j.apsoil.2004.12.001


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The handling Editor declared a shared affiliation, though no other collaboration, with the authors and states that the process nevertheless met the standards of a fair and objective review.

Copyright © 2016 Tamilselvi, Thiyagarajan and Uthandi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Expression of Zinc Transporter Genes in Rice as Influenced by Zinc-Solubilizing Enterobacter cloacae Strain ZSB14

Selvaraj Krithika and Dananjeyan Balachandar\*

Department of Agricultural Microbiology, Tamil Nadu Agricultural University, Coimbatore, India

Zinc (Zn) deficiency in major food crops has been considered as an important factor affecting the crop production and subsequently the human health. Rice (Oryza sativa) is sensitive to Zn deficiency and thereby causes malnutrition to most of the rice-eating Asian populations. Application of zinc solubilizing bacteria (ZSB) could be a sustainable agronomic approach to increase the soil available Zn which can mitigate the yield loss and consequently the nutritional quality of rice. Understanding the molecular interactions between rice and unexplored ZSB is useful for overcoming Zn deficiency problems. In the present study, the role of zinc solubilizing bacterial strain Enterobacter cloacae strain ZSB14 on regulation of Zn-regulated transporters and iron (Fe)-regulated transporter-like protein (ZIP) genes in rice under iron sufficient and deficient conditions was assessed by quantitative real-time reverse transcription PCR. The expression patterns of OsZIP1, OsZIP4, and OsZIP5 in root and shoot of rice were altered due to the Zn availability as dictated by Zn sources and ZSB inoculation. Fe sufficiency significantly reduced the root and shoot OsZIP1 expression, but not the OsZIP4 and OsZIP5 levels. Zinc oxide in the growth medium up-regulated all the assessed ZIP genes in root and shoot of rice seedlings. When ZSB was inoculated to rice seedlings grown with insoluble zinc oxide in the growth medium, the expression of root and shoot OsZIP1, OsZIP4, and OsZIP5 was reduced. In the absence of zinc oxide, ZSB inoculation up-regulated OsZIP1 and OsZIP5 expressions. Zinc nutrition provided to the rice seedling through ZSB-bound zinc oxide solubilization was comparable to the soluble zinc sulfate application which was evident through the ZIP genes' expression and the Zn accumulation in root and shoot of rice seedlings. These results demonstrate that ZSB could play a crucial role in zinc fertilization and fortification of rice.

Keywords: metal transporter, rice, zinc solubilizing bacteria, zinc uptake, ZIP genes

## INTRODUCTION

Zinc (Zn) is a critical micronutrient responsible for several cellular functions in plant and its deficiency causes decrease in plant growth and yield. Zn deficiency in major food crops, apart from yield loss, reduced the Zn content of grains and subsequently causes serious problems in human nutrition. Rice, the staple diet for more than 560 million people of world, is one of

#### Edited by:

Gero Benckiser, University of Giessen (Retired), Germany

#### Reviewed by:

Zakira Naureen, University of Nizwa, Oman Chong Zhang, University of Maryland Baltimore County, USA

> \*Correspondence: Dananjeyan Balachandar dbalu@tnau.ac.in

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 28 December 2015 Accepted: 21 March 2016 Published: 06 April 2016

#### Citation:

Krithika S and Balachandar D (2016) Expression of Zinc Transporter Genes in Rice as Influenced by Zinc-Solubilizing Enterobacter cloacae Strain ZSB14. Front. Plant Sci. 7:446. doi: 10.3389/fpls.2016.00446

**337**

the "most sensitive" crops for Zn nutrition. Nearly 50% of the rice-growing soils are under Zn-deficient and rice grown on these soils generally produce low yield with poor nutritional quality. For incidence, Zn-sufficient rice had about 40 mg/kg of Zn in its grains, whereas Zn-deficient rice accumulated less than 10 mg/kg of grain-Zn (Wissuwa et al., 2008). Soil pH especially alkalinity, low organic carbon, high carbonates (calcareous) and low redox potential of wetland primarily limit the Zn availability for rice. However, agricultural intensification, imbalanced nutrients and neglected micronutrient application further worsen the Zn-deficiency problem in rice. Indeed, zinc deficiency in rice is becoming one of the public health problems through malnutrition in many rice-based food adopting countries of Asia (Cakmak, 2008).

For rice, soil and foliar application of highly soluble zinc sulfate (ZnSO4) is a common Zn-fertilization to correct the Zn deficiency. However, the applied Zn got precipitated as hydroxides, carbonates, phosphates and sulfides as dictated by physico-chemical properties of the soil and resulted with very low fertilizer use efficiency (1–5%). Alternatively, exploring the soil bacteria, capable of solubilizing inorganic Zn and thereby increasing the availability for crop assimilation, is a viable option to achieve the objective of correcting the Zn deficiency and thereby overcoming the zinc malnutrition in human (He et al., 2010; Mäder et al., 2010). Several zinc solubilizing bacteria (ZSB) were characterized from tropical and temperate soils to provide plant available Zn (Hafeez et al., 2013). For example, Gluconacetobacter from sugarcane, Bacillus and Pseudomonas from soybean, rice and wheat capable of solubilizing zinc compounds such as oxide, carbonate, and phosphate were reported earlier (Saravanan et al., 2011). These ZSB strains produce variety of low molecular weight organic acids, particularly gluconic acid, dissolute the insoluble Zn; reduce the pH of the soil solution and thereby increase the plant available zinc (Hafeez et al., 2013). Inoculation of these bacteria enhanced the Zn uptake of rice (Vaid et al., 2014), maize (Goteti et al., 2013), wheat (Rana et al., 2012), green gram (Sharma et al., 2012), and soybean (Ramesh et al., 2014). Few studies also confirmed the ability of ZSB for biofortification in rice (Vaid et al., 2014) and wheat (Ramesh et al., 2014). However, their full potential to mitigate the zinc deficiency and to increase the grain-Zn is not yet explored due to poor understanding of microbe-soil-plant interactions.

The soil available zinc (Zn2+) is taken up by root membrane transport mechanisms in rice which include phytosiderophores (Bashir et al., 2010) and Zn-regulated transporters and iron (Fe)-regulated transporter-like protein (ZIP) family (Guerinot, 2000). In rice, several ZIPs including OsIRT1, OsIRT2, OsZIP1, OsZIP3, OsZIP4, OsZIP5, OsZIP7, and OsZIP8 were reported to be responsible for Zn uptake from soil, translocation within root and from root to shoot as well as for storage in grains (Ramesh et al., 2003; Ishimaru et al., 2005, 2006; Yang et al., 2009; Lee et al., 2010a,b). OsITR1 and OsITR2 are responsible for transport of Fe2<sup>+</sup> from rhizosphere to root with less affinity to Zn (Ishimaru et al., 2006). OsZIP1, OsZIP3, OsZIP4, OsZIP5, and OsZIP8 are rice plasma membrane Zn transporters and are induced by Zn deficiency (Ramesh et al., 2003; Ishimaru et al., 2005; Yang et al., 2009; Lee et al., 2010a; Suzuki et al., 2012). The expression of most of the well-studied rice ZIP genes (OsZIP1, OsZIP4, OsZIP5) was controlled by the availability of divalent cations such as Zn2+, Fe2+, Cu2+, Mn2<sup>+</sup> (Bughio et al., 2002; Ishimaru et al., 2005; Lee et al., 2010a). Few studies also confirmed that these transporter genes' expression varied between root and shoot tissues of rice (Ishimaru et al., 2011). Similarly, Chen et al. (2008) reported the differential expression pattern of ZIP genes (OsZIP1, OsZIP3, and OsZIP4) between Zn-efficient and Zn-inefficient cultivars of rice. These ZIP genes varied their expression levels at different growth stages of rice from germination to grain filling (Ishimaru et al., 2011). The plant growth promoting rhizobacteria (PGPR) upon colonizing the roots, acidify the rhizosphere through organic acids and produce siderophores which facilitate the trace elements' uptake by the crop plants. However, no attempts were made so far to elucidate the role of these zinc solubilizing PGPR strains to regulate the expression of metal transporter genes in the root. Understanding the interaction between rice plant and Zn solubilizing PGPR in terms of Zn transporter genes' expression would help to alleviate the Zn deficiency as well as to improve the Zn fortification. In the present work, we have reported the root and shoot ZIP genes' expression pattern of rice seedlings upon inoculating with a potential ZSB (Enterobacter cloacae strain ZSB14) under controlled condition. Our results suggest that the ZSB in rhizosphere of rice roots may regulate ZIP genes' expression either directly or indirectly through Zn availability.

### MATERIALS AND METHODS

### Bacterial Strain and Culture Condition

Enterobacter cloacae strain ZSB14, isolated and characterized from rhizosphere of rice, capable of solubilizing insoluble Zn compounds viz., ZnO (24.05 µg/ml of soluble Zn), ZnCO<sup>3</sup> (19.37 µg/ml) and Zn3(PO4)<sup>2</sup> (6.06 µg/ml) was used for this study. In order to maintain the Zn solubilizing potential of the strain, The culturing was routinely done in Bunt and Rovira medium containing 0.1% ZnO with and without agar (1.5%; Bunt and Rovira, 1955) at 30◦C in an incubator (Lab Companion, USA).

### Rice Culture and ZSB Inoculation

Rice (Orzya sativa) cultivar Co51 of Tamil Nadu Agricultural University, Coimbatore was used for this experiment. De-husked healthy seeds were surface sterilized with sodium hypochlorite with 5% available chlorine for 10 min followed by five washes with sterile distilled water. The seeds were soaked in sterile distilled water for over-night for sprouting. Uniformly sprouted seeds were placed (10 seeds per plate) on Fe sufficient (Fe+) modified Hoagland medium (5 mM KNO3, 2 mM MgSO4, 2 mM Ca(NO3)2, 2.5 mM KH2PO4, 70 µM H3BO3, 1 µM MnCl2, 0.5 µM CuSO4, 10 µM NaCl, 0.2 µM Na2MoO4, 50 µM FeEDTA, 1.0 g/l MES buffer pH 5.8; 40 mM Sucrose; 8.0 g/l plant agar). For Fe-deficient (Fe−) condition, modified Hoagland medium lacking FeEDTA was used. Both under

Fe<sup>+</sup> and Fe−, five treatments for Zn nutrition were adopted viz., (i) no-zinc control; (ii) soluble Zn as ZnSO<sup>4</sup> (5 µM); (iii) sparingly soluble ZnO (10 µM); (iv) ZnO with ZSB inoculation; (v) ZSB inoculation alone. ZnSO<sup>4</sup> (5 µM) or ZnO (10 µM) was supplemented directly in modified Hoagland medium depending upon the treatment and seeds were placed. The plants were grown in a growth chamber with 12 h light (200 mole/m<sup>2</sup> /s) at 28◦C. After 7 days of growth, the rice seedlings were inoculated with ZSB strain depending upon the treatment. For this, the strain ZSB14 was cultivated in Bunt and Rovira medium added with ZnO to achieve a final Zn concentration of 0.1% at 30◦C till reached a final concentration of approximately 10<sup>11</sup> colony forming units (cfu) per ml. The bacteria were pelletized by centrifugation at 5000 g for 20 min at room temperature and cell pellets were re-suspended in 10 mM MgSO<sup>4</sup> and centrifuged. This operation was repeated and afterward the cell pellets were re-suspended in 10 mM MgSO4. The bacterial titer was adjusted to the OD<sup>600</sup> of 0.05 (10<sup>8</sup> cfu per ml) and 20 µL of bacterial suspension was then applied on each root of 7-days-old seedlings, right below the hypocotyl. After additional 7 days of incubation, the seedlings were removed carefully from the plates and assessed for ZIP gene expression.

### RNA Preparation and Real-Time RT-PCR Analysis

Total RNA from shoot and root of rice was extracted separately by following the procedure of Oñate-Sánchez and Vicente-Carbajosa (2008). The residual genomic DNA in the RNA preparation was digested with RNAse-free DNase I (New England Biolabs) until no amplicons were obtained when using RNA preparations directly in the PCR reaction with the primers for the actin gene (OsACT1). The primer details are provided in **Table 1**. Subsequently, complementary DNA (cDNA) was synthesized from 3 µg of DNA-free total RNA using Revert Aid H minus reverse transcriptase (Thermo Scientific) by primering with oligo d(T)<sup>18</sup> (Invitrogen) in a 40-µL reaction mixture according to the manufacturer's instruction. Real-time PCR was performed in Roche Lightcycler 480II (Roche, Switzerland) to quantify the transcripts of OsZIP1, OsZIP4, and OsZIP5 (Primer details in **Table 1**) using SYBR Green (SYBR Premix ExTaq, Tli RNase H Plus, Takara) as the detection system. The constitutively expressed OsACTIN1 gene was amplified as the reference gene. Changes in expression were calculated by relative quantification (11Ct) method (Livak and Schmittgen, 2001) using threshold cycle (Ct) values of target and reference genes. For all real-time RT-PCR analyses, three biological replicates and two technical replicates were used. The size and intensity of amplified fragments were confirmed by gel electrophoresis.

### Determination of Zn Content in Rice Plant

Rice seedlings washed until free from agar medium were ovendried at 70◦C for 5 h and digested with 15 ml triple acid mixture (nitric, sulfuric, and perchloric acid in the ratio of 9:2:1) for overnight. The volume of cooled digest was made up to 25 ml using deionized double distilled water and the dilutions were used for Zn estimation using atomic absorption spectrophotometer (GBS Scientific, Australia) at the wavelength of 213.86 nm.

### Statistical Analyses

All the data were subjected to statistical analysis with software, Microsoft Excel for Windows 2007 add-in with XLSTAT Version 2010.5.05 (XLSTAT, 2010). Statistically significant differences between the treatments were analyzed using analysis of variance (ANOVA) and Duncan's Multiple Range Test (DMRT) at 5% significance level.

### RESULTS

### OsZIP1 Expression

Both the Fe-levels and Zn-treatments significantly influenced the expression of OsZIP1 (**Figure 1**). The 7-days-old Fe<sup>−</sup> rice roots had higher copies of OsZIP1 transcripts than those of Fe<sup>+</sup> seedlings. Under Fe<sup>−</sup> condition, ZnO addition recorded highest OsZIP1 transcripts followed by no-Zn control. Under Fe<sup>+</sup> condition, no-Zn control recorded lowest OsZIP1 transcripts followed by ZnSO<sup>4</sup> while, ZnO recorded the maximum relative expression among the treatments. Addition of ZnO in the medium up-regulated OsZIP1 expression both in Fe<sup>−</sup> and Fe<sup>+</sup> conditions (**Figure 1A**). When ZSB-inoculated seedlings were assayed for OsZIP1 expression after 7 days (14 days-old seedlings), the expression pattern of root OsZIP1 was different than the earlier (**Figure 1B**). The ZSB inoculation substantially reduced the OsZIP1 expression both in Fe-deficient and Fesufficient rice roots in the presence of ZnO. When ZnO was not in the medium, ZSB inoculated rice seedlings had maximum OsZIP1 transcripts for both Fe-deficient and Fe-sufficient conditions. Under Fe<sup>+</sup> condition, ZnSO<sup>4</sup> amendment recorded higher OsZIP1 transcripts than No-Zn and ZnO + ZSB, but lower than ZnO and ZSB. Under Fe<sup>−</sup> condition, the same trend was noticed with the exception of ZnSO<sup>4</sup> and ZnO with at par levels.

The OsZIP1 levels of shoot were nearly 10-fold higher than root in the 7-days-old seedlings before exposure to ZSB inoculation (**Figure 1C**). However, the Fe-sufficient shoots did not show any significant difference within the Zn-treatments for the level of OsZIP1 transcripts, while the Fe-deficient shoots showed significant difference between them. The no-Zn controls and ZnSO<sup>4</sup> had lowest shoot OsZIP1 while ZnO amended Fe<sup>−</sup> rice recorded maximum expression. After 7-days of ZSB inoculation, OsZIP1 levels had significant different in the Zn-treatments. The ZSB inoculation considerably reduced the OsZIP1 expression of the rice shoot (**Figure 1D**). When comparing the ZnO and ZnO + ZSB, nearly 50% reduction in OsZIP1 expression was recorded due to ZSB inoculation in both Fe<sup>−</sup> and Fe<sup>+</sup> conditions. In shoot also, the OsZIP1 expression was reduced in the presence of Fe in the medium.

### OsZIP4 Expression

The expression of OsZIP4 gene was significantly influenced by different Zn-treatments but not by Fe<sup>−</sup> levels (**Figure 2**). Both

#### TABLE 1 | Primers used for quantitative real-time reverse transcription PCR.


Fe<sup>+</sup> and Fe<sup>−</sup> rice responded similar pattern of OsZIP4 expression in the root and shoot of rice seedlings. The root of 7-days-old rice seedlings recorded significantly highest OsZIP4 transcripts due to ZnO followed by no-Zn control under Fe-deficient and sufficient conditions (**Figure 2A**). The ZnSO<sup>4</sup> amendment down-regulated the OsZIP4 significantly than other treatments. When the ZSB strain was inoculated, the root OsZIP4 showed remarkable difference of expression after 7 days. Irrespective of treatments, the level of relative expression of OsZIP4 at 14 daysold seedling had been increased nearly 10-fold than 7 daysold plants (**Figure 2B**). Among the rice seedlings exposed to different amendments, ZnO addition significantly increased the root-OsZIP4 transcripts followed by ZnSO4, while the ZnO + ZSB and ZSB alone significantly reduced the expression. The no-Zn controls did not show any variations in their root OsZIP4 transcript levels between two observations.

indicate the standard error. Values followed by the same letter in each panel are not significantly different from each other as determined by DMRT (p ≤ 0.05).

With reference to shoot OsZIP4, the level of expression remained same between Fe<sup>+</sup> and Fe<sup>−</sup> seedlings after 14 days of incubation (**Figures 2C,D**). The shoot of 7-days-old rice seedlings before ZSB inoculation exposed to ZnO had significantly higher OsZIP4 transcripts that ZnSO<sup>4</sup> and no-Zn controls of both Fedeficient and sufficient rice plants (**Figure 2C**). When ZSB was inoculated on 7th day, the OsZIP4 significantly reduced in ZnO + ZSB treatment to a tune of 77 and 88% for Fe-sufficient and deficient rice plants, respectively as compared to ZnO treatment (**Figure 2D**). Irrespective to Fe-levels, ZnO amended uninoculated plants remained constant level of expression for both the assessments; whereas ZnSO<sup>4</sup> amended plants increased their OsZIP4 expression levels to ninefold after 7 days of additional incubation.

### OsZIP5 Expression

The transcripts of OsZIP5 were strongly found in shoots and weakly in roots of 7-days-old rice seedlings (**Figure 3**). Like OsZIP4, OsZIP5 also did not show significant response to Fe levels. In the roots of 7-days-old rice seedlings, the no-Zn and ZnO amended rice seedlings showed significantly higher levels of OsZIP5 in both Fe-sufficient and deficient conditions (**Figure 3A**). The ZnSO<sup>4</sup> amended rice plants had the least expression of OsZIP5 in their roots. After ZSB inoculation and 7 days incubation, the pattern of OsZIP5 expression was different than those of before inoculation. After additional 7 days of incubation, the rice seedlings exposed to ZnO alone had nearly 60-fold increased OsZIP5 transcripts in both Fe<sup>+</sup> and Fe<sup>−</sup> conditions (**Figure 3B**). However, ZSB inoculation alone also induced OsZIP5 in Fe<sup>+</sup> and Fe<sup>−</sup> rice roots but significantly lower than ZnO amendment. The ZnO + ZSB inoculation, ZnSO<sup>4</sup> and no-Zn controls had least OsZIP5 transcripts in their Fe<sup>+</sup> and Fe<sup>−</sup> roots.

The response of rice shoot OsZIP5 was similar to that of root but with twofold increased levels than roots (**Figures 3C,D**). There was no significant difference between Fe<sup>+</sup> and Fe<sup>−</sup> plants in terms of shoot OsZIP5 expression. The ZnSO<sup>4</sup> amendment in the medium down-regulated the OsZIP4 in Fe<sup>+</sup> and Fe<sup>−</sup> shoots, whereas no-Zn controls and ZnO amendments had significantly higher copies of OsZIP5 transcripts (**Figure 3C**). When ZSB was inoculated to their respective treatment plants, there was significant effect found due to ZSB inoculation. ZnO + ZSB inoculation significantly down-regulated the OsZIP5 to a tune of 91 and 95% for Fe<sup>+</sup> and Fe<sup>−</sup> plants, respectively as compared

14th day expression levels. Fe+, Fe sufficient condition; Fe−, Fe deficient condition. Control, No-zinc control; ZnSO<sup>4</sup> at 5 mM; ZnO at 10 mM; ZSB, Zinc solubilizing bacteria (Enterobacter cloacae strain ZSB14) inoculation on 7th day. Relative mRNA abundance of OsZIP5 was quantified and normalized with OsACTIN1 gene on 7th day and 14th day. Data from real-time RT-PCR experiments were analyzed according to the 2−11Ct method. Means of six replicate values plotted, errors bars indicate the standard error. Values followed by the same letter in each panel are not significantly different from each other as determined by DMRT (p ≤ 0.05).

to ZnO amended rice shoots (**Figure 3D**). The ZSB inoculation without ZnO up-regulated the OsZIP5 expression in shoots after 7-days of incubation. The no-Zn and ZnSO<sup>4</sup> also had significantly higher levels of OsZIP5 transcripts than ZnO + ZSB treatment.

### Leave Chlorosis of Rice Seedlings

We examined the role of Fe and ZSB-bound Zn availability on metal uptake of rice (Fe and Zn) in terms of chlorosis of leaves. The color intensity of rice leaves after 14-days of exposure to various Zn treatments under Fe<sup>+</sup> and Fe<sup>−</sup> conditions showed significant difference (**Figure 4**). The Fe<sup>+</sup> condition made rice leaves with dark intensity while the Fe<sup>−</sup> showed chlorosis. Under Fe<sup>+</sup> condition, ZnO induced the chlorosis of leaves, while with ZSB inoculation, the chlorosis was reduced. The ZnSO<sup>4</sup> and ZSB alone did not show any chlorosis at all. Under Fe<sup>−</sup> condition, no-Zn and ZnO showed severe yellowing, while the ZnSO<sup>4</sup> and ZSB had less chlorosis.

### Zn Content of Rice Seedlings

The root and shoot Zn content of rice seedlings before ZSB inoculation (7th day) was significantly influenced by Zn amendments and also due to Fe conditions. In 7-days-old

seedlings, ZnSO<sup>4</sup> recorded 198 and 280% higher root Zn and 108 and 121% higher shoot Zn than Fe<sup>+</sup> and Fe<sup>−</sup> no-Zn controls, respectively (**Table 2**). ZnO also increased the root and shoot Zn of rice seedlings than no-Zn controls, which were trivial



Values are mean (± standard error; n = 3) and values followed by the same letter in each column are not significantly different from each other as determined by DMRT (p ≤ 0.05). Fe (+), Fe sufficient condition; Fe (−), Fe deficient condition; Control, No-zinc control; ZSB, Zinc solubilizing bacteria (Enterobacter cloacae strain ZSB14) inoculation on 7th day.

(14–18%) as compared to ZnSO4. The Zn uptake measured as zinc content of rice seedlings after ZSB inoculation (14th day) was also significantly influenced by Zn sources. Fe<sup>−</sup> condition increased the Zn contents of root and shoot significantly than Fe<sup>+</sup> for Zn treatments (ZnSO<sup>4</sup> and ZnO + ZSB) but not for no-Zn and ZSB alone controls at 14th day (**Table 2**). Among the various treatments enforced, the ZnSO<sup>4</sup> recorded maximum root and shoot Zn contents (198.25 and 154.26 mg/g for Fe<sup>−</sup> and 145.15 and 125.37 mg/g for Fe<sup>+</sup> respectively). The ZnO + ZSB treated plants recorded 157.46 and 124.13 mg/g of Zn in root and shoot, respectively under Fe-deficient condition, and 87.24 and 90.16 mg/g for Fe-sufficient conditions. The ZnO alone treated plants had very little increase of Zn content as compared to no-Zn control. ZSB inoculation alone had at par root and shoot Zn levels as that of no-Zn controls for Fe<sup>+</sup> and Fe<sup>−</sup> rice seedlings.

### DISCUSSION

Exploiting the ZSB for alleviating Zn-deficiency as well as for Zn-fortification in food grains like rice could be a promising agronomical approach to minimize the Zn-deficiency in human being. Keeping in view the unambiguous benefits of ZSB (Hafeez et al., 2013), through the present investigation, we reported that ZSB inoculation to rice could alter the expression of zinc transporting genes of rice based on the Zn solubilization and thereby regulate the uptake of zinc.

The ZIP family transporters are well-characterized and are suggested to be the primary uptake system for Zn in plants (Guerinot, 2000; Mäser et al., 2001). Most of these ZIP genes are induced by Zn deficiency (Ramesh et al., 2003; Ishimaru et al., 2005; Chen et al., 2008) and their expression pattern varied between root and shoot system. OsZIP1 was shown to be expressed higher levels in roots than shoots under Zn-deficient condition (Ramesh et al., 2003; Ishimaru et al., 2005). Chen et al. (2008) observed that OsZIP1 was up-regulated in Zn-deficient roots, but no visible transcripts detected in shoots of both Znefficient and Zn-inefficient rice genotypes. In contrast to these, Ramegowda et al. (2013) found that OsZIP1 over-expressing transgenic finger millet showed higher expression of OsZIP1 in leaves under Zn-sufficient condition. In the present work also, we found higher expression of OsZIP1 in shoot than root in 7 days-old rice and the OsZIP1 expression was influenced by Fe availability apart from zinc. The rice grown for 7-days under Fesufficient condition had relatively lower OsZIP1 transcripts than those plants grown in Fe-deficient condition. Among the two Zn-treatments, sparingly soluble ZnO up-regulated the OsZIP1 as compared to highly soluble ZnSO<sup>4</sup> before ZSB inoculation. This is in accordance with the earlier findings that the zinc abundance reduced the root OsZIP1 expression (Ramesh et al., 2003; Ishimaru et al., 2005). However, in the present work, when the rice plants grown with ZnO had highest OsZIP1 expression in their roots after 7 days which was higher than no-zinc control. This implies that the sparingly soluble ZnO could not supply the available Zn in the growth medium of rice and subsequently cause more stress than no-Zn condition. Further investigations are needed to understand how the ZnO induced the ZIP transporters higher than no-Zn condition. However, when ZSB was inoculated on 7-days-old rice seedling, considerable reduction in OsZIP1 expression was noticed in both Fe<sup>+</sup> and Fe<sup>−</sup> root and shoot of rice. This might be due to the ZSB-mediated Zn solubility and availability in the medium as well as the ZSB-mediated rhizospheric effects. Interestingly, ZSB inoculation increased the root and shoot OsZIP1 expression even in the absence of Zn.

In the present work, Fe<sup>+</sup> and Fe<sup>−</sup> conditions did not alter the expressions of OsZIP4 and OsZIP5 as that of OsZIP1 which is in accordance with the earlier works (Ishimaru et al.,

2005, 2007; Lee et al., 2010a). OsZIP1 is primarily associated with metal uptake from rhizosphere (Ramesh et al., 2003), while OsZIP4 and OsZIP5 are involved in the translocation of Zn with in the plant (Ishimaru et al., 2005) might be the reason, why these genes are not regulated due to Fe levels. The previous works confirmed that OsZIP4 in Zndeficient rice was expressed in meristem and vascular bundles of roots and shoots and is responsible for Zn translocation to various plant parts that require Zn (Ishimaru et al., 2011). As the transporters involving in metal uptake from soil may have non-specific uptake of the ions such as Zn, Fe, Cu, Cd, Mn from soil to the root, these genes' expression was regulated based upon the affinity of the metals. However, the transporters responsible for translocation of metals within the plant had less impact of other metal species. For example, the transporters OsZIP4, OsZIP5, and OzZIP8 responsible for Zn translocation in rice are not influenced by Fe+, while the OsZIP1 and OsITR1 responsible for Zn and Fe uptake from soil respectively, were also influenced by other metals (Lee and An, 2009). The present results are in accordance with these findings. In the present work, ZnSO<sup>4</sup> in the growth medium made Zn sufficient condition and thereby reduced the OsZIP4 expressions in both root and shoot. When ZnO was amended, the relative OsZIP4 expression was significantly higher than no-Zn control which means that the addition of ZnO cause more stress to the rice than no-Zn. When the ZSB was inoculated on 7th day and incubated for additional 7-days, the relative expression of OsZIP4 got varied in those treatments which imply that rhizosphere colonization of ZSB either directly or indirectly regulates ZIP genes of rice. Compare to ZnO treatment, ZnO + ZSB reduced the OsZIP4 expression revealed that the ZSB-mediated solubilization of ZnO enhanced the uptake of Zn and thereby reduced the Zn deficiency. The down-regulation of OsZIP4 found in rice shoot due to ZSB inoculation implies that the ZSB-bound Zn release has been effectively translocated to the shoot system also. Compare to no-Zn control, ZSB inoculation in the absence of ZnO up-regulated root OsZIP4 but down-regulated the shoot OsZIP4. However, compare to OsZIP1, the ZSB-mediated regulation of OsZIP4 was relatively low. Hence, further investigation is needed to understand this variation between ZIP transporters' response for ZSB inoculation.

OsZIP5 is a plasma membrane-bound transporter responsible for Zn translocation within the rice plant. Expression of OsZIP5 is mainly regulated by Zn levels and Zn deficient condition upregulated the expressions in both shoot and root (Lee et al., 2010a). Over-expression of OsZIP5 over-expressed plants showed sensitive to excess Zn, while the OsZIP5 knock-out plants had high Zn tolerance (Ishimaru et al., 2011). In the present experiment, the expression pattern of OsZIP5 was differed from OsZIP1 and OsZIP4 in several treatments. Before ZSB inoculation, no-Zn and ZnO applied rice plants, which are suffered with Zn deficiency had maximum OsZIP5 expression both in root and shoot. ZnSO<sup>4</sup> in the medium down regulated the expression of root and shoot OsZIP5. When ZSB was inoculated to ZnO and no-Zn plants, OsZIP5 was in low copies in ZnO amended plants, while No-Zn but ZSB inoculated plants had significantly higher transcripts. This result implies that ZSB had significant influence on OsZIP5 by providing soluble Zn from ZnO while in the absence of Zn, ZSB up-regulated OsZIP5 as that of OsZIP1. Hence, it is clear from these experiments that ZSB had direct impact on OsZIP1 and OsZIP5 and for OsZIP4, the regulation is dependent of Zn-availability due to the functioning of ZSB.

Several previous studies also confirmed that ZSB inoculation enhanced the exchangeable Zn in the soil or rhizosphere of crops through organic acid production and enhanced microbial processes and subsequently improved the Zn uptake (Oburger et al., 2009; Ramesh et al., 2014; Shakeel et al., 2015). As supportive to these findings, in the present work, the Zn content of shoot and root of ZnO + ZSB inoculated rice plants was higher than no-Zn, ZnO alone and ZSB alone plants, but lower than ZnSO<sup>4</sup> amended plants. This was further evident from the observation on the chlorosis of rice leaves in the present experiment (**Figure 4**). Fe and Zn sufficient conditions did not show any chlorosis, while ZnO induced the chlorosis implies that ZnO may affect the Fe uptake along with Zn. Hence, inoculation of ZSB in the root zone improved the Zn uptake and translocation within the plant and thereby increased the Zn contents of root and shoot compared to no-Zn and ZnO alone controls.

### CONCLUSION

In the present investigation, we proved that the inoculation of ZSB under controlled condition can able to regulate some of the Zn-regulated transporters family genes and thereby controlled the Zn uptake in rice seedlings. Zn sufficient condition created by ZnSO<sup>4</sup> down regulated OsZIP4 and OsZIP5 both in root and shoot of rice. The application of sparingly soluble ZnO as Zn source created severe Zn related stress to the rice, which up-regulated all the ZIP genes. Upon inoculation of ZSB, the expression levels of OsZIP1, OsZIP4, and OsZIP5 were reduced. In the absence of Zn source, ZSB inoculation could regulate OsZIP1 and OsZIP5 but not the OsZIP4. These results are evident that the ZSB inoculation as PGPR could regulate the Zn uptake and translocation in rice plant and thereby zinc fortification in rice grains.

### AUTHOR CONTRIBUTIONS

The experiments were planned and executed together by SK and DB. SK undertook the data analysis. The interpretation of results and manuscript preparation were done by DB.

### FUNDING

The corresponding author (DB) thanks Ministry of Human Resource Development, New Delhi, India for providing financial support through Centre of Excellence in Frontier Areas of Science and Technology on Microbes to Feed the World: Plantmicrobe interaction to boost the agricultural production (F. No. 5-5/2014 – TS VII).

### ACKNOWLEDGMENTS

We thank Prof. Dr. Corne M. J. Pieterse, Dr. Peter A. H. M. Bakker, and Dr. Roeland L. Berendsen of Plant

### REFERENCES


Microbe Interactions Laboratory, Department of Biology, Utrecht University, Utrecht, The Netherlands for sufficient knowledge given through HRD training to carry out the present experiments.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Krithika and Balachandar. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Root Associated Bacillus sp. Improves Growth, Yield and Zinc Translocation for Basmati Rice (Oryza sativa) Varieties

### Muhammad Shakeel, Afroz Rais, Muhammad Nadeem Hassan and Fauzia Yusuf Hafeez \*

Department of Biosciences, COMSATS Institute of Information Technology, Islamabad, Pakistan

Plant associated rhizobacteria prevailing in different agro-ecosystems exhibit multiple traits which could be utilized in various aspect of sustainable agriculture. Two hundred thirty four isolates were obtained from the roots of basmati-385 and basmati super rice varieties growing in clay loam and saline soil at different locations of Punjab (Pakistan). Out of 234 isolates, 27 were able to solubilize zinc (Zn) from different Zn ores like zinc phosphate [Zn<sup>3</sup> (PO4)2], zinc carbonate (ZnCO3) and zinc oxide (ZnO). The strain SH-10 with maximum Zn solubilization zone of 24 mm on Zn<sup>3</sup> (PO4)2ore and strain SH-17 with maximum Zn solubilization zone of 14–15 mm on ZnO and ZnCO3ores were selected for further studies. These two strains solubilized phosphorous (P) and potassium (K) in vitro with a solubilization zone of 38–46 mm and 47–55 mm respectively. The strains also suppressed economically important rice pathogens Pyricularia oryzae and Fusarium moniliforme by 22–29% and produced various biocontrol determinants in vitro. The strains enhanced Zn translocation toward grains and increased yield of basmati-385 and super basmati rice varieties by 22–49% and 18–47% respectively. The Zn solubilizing strains were identified as Bacillus sp. and Bacillus cereus by 16S rRNA gene analysis.

### Edited by:

Anton Hartmann, Helmholtz Zentrum München, German Research Center for Environmental Health, Germany

#### Reviewed by:

Biswapriya Biswavas Misra, University of Florida, USA Oswaldo Valdes-Lopez, National Autonomus University of Mexico, Mexico

\*Correspondence:

Fauzia Yusuf Hafeez fauzia@comsats.edu.pk

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 21 August 2015 Accepted: 03 November 2015 Published: 18 November 2015

#### Citation:

Shakeel M, Rais A, Hassan MN and Hafeez FY (2015) Root Associated Bacillus sp. Improves Growth, Yield and Zinc Translocation for Basmati Rice (Oryza sativa) Varieties. Front. Microbiol. 6:1286. doi: 10.3389/fmicb.2015.01286 Keywords: bio fertilizer, PGPR, rhizobacteria, zinc, rice, super basmati, basmati-385, 16S rRNA gene

### INTRODUCTION

Rice is an important cereal crop growing across the world and is a major staple food of Asian population. In current scenario of intensive farming, soils are continuously depleting in macro and micro nutrients especially in wheat-rice cropping system (Rana et al., 2015). Among micro nutrients, Zn is a highly essential nutrient required throughout the life cycle of rice (Holler et al., 2014).

Farmers use chemical fertilizers to make up the deficiency of essential nutrients and thus achieve high yields. Irrational use of chemical fertilizers has led to severe environmental problems especially contamination of underground water due to leaching and pollution of atmosphere through gaseous emissions (Lockhart et al., 2013; Wu and Ma, 2015). Moreover, fluctuations in price and nonavailability of chemical fertilizers due to energy crisis and other reasons pose a major constraint in sustainable crop production. Under these circumstances, plant growth promoting rhizobacteria (PGPR) may offer a valuable alternative to chemical fertilizers.

This is because PGPR live freely in soil, colonize plant roots aggressively and establish symbiotic association with plants. Existence of PGPR with the plant roots is generally classified by two environments viz; rhizosphere and endosphere. Rhizosphere represents the soil volume under the direct influence of root while endosphere represents the internal tissue of root (Timm et al., 2015). The strains inhabiting rhizosphere and endosphere are called rhizobacteria and endophytes respectively.

PGPR enrich soil with major plant nutrients such as nitrogen (N) by fixing it from the atmosphere, phosphorous (P), and potassium (K) by solubilizing them from the soil (Patel et al., 2015; Pii et al., 2015; Zahid et al., 2015). They also assist in bioavailability of Zn by solubilizing it from various ores like Zn<sup>3</sup> (PO4)2, ZnCO3, and ZnO (Abaid-Ullah et al., 2015; Sirohi et al., 2015). In addition to provide macro and micronutrients to the plants, PGPR also protect them from pathogens. They suppress the activity of pathogens by producing numerous antifungal metabolites like siderophores, hydrolytic enzymes, and antibiotics (Chowdhury et al., 2015). Therefore, they could be utilized as an alternative to the chemical fertilizers and fungicides and hence may ensure sustainable agriculture production, environmental safety and lower production cost.

This argument is supported by earlier reports where inoculation of plants with PGPR has resulted in the improved nutrition, vigorous plant growth and high yield (El-Sayed et al., 2014; Majeed et al., 2015). Many effective strains have been formulated as biofertilizers. The use of registered biofertilizers and microbial technologies has become a widely accepted strategy in the current intensive agricultural practices prevailing throughout the world (Shen et al., 2015). The PGPR strains identified so far belong to genus Pseudomonas, Ochrobacterum, Bacillus, Azosperillum, Azotobacter, Rhizobium, Stenotrophomonas, Serratia, and Enterobacteria (Hassan et al., 2010; Ma et al., 2011; Abaid-Ullah et al., 2015). Highly positive effects of these species on the growth of various crops such as sugarcane, maize, wheat, rice, canola, sunflower and other vegetables have been observed both in vitro and in vivo under variable climatic conditions (Hassan et al., 2011; El-Sayed et al., 2014). For example, Kosakonia radicincitans increased dry weight and N content of yerba mate by 183 and 30% (Bergottini et al., 2015). Wheat yield was improved by 9% upon inoculation with a consortium of Bacillus thuringiensis and Serratia sp. (Abaid-Ullah et al., 2015; Pereg and McMillan, 2015).

In addition to an increase in yield, PGPR also significantly affect the nutrients uptake by plants. This property of rhizobacteria has drawn the attention of researchers to exploit them in cereals bio fortification. Ramesh et al. (2014) reported that Zn solubilizing strains of Bacillus aryabhattai improved Zn mobilization in wheat and soybean. Recent studies conducted at our laboratory (Abaid-Ullah et al., 2015) reveiled that certain strains of Serratia sp., Pseudomonas sp., and Bacillus sp., enhanced Zn translocation toward wheat grains by 7–12% compared to that of chemical Zn (Abaid-Ullah et al., 2015).

The plant–microbe interaction is quite complex phenomena and effect of this interaction on growth and physiological processes occurring in the life of plant has been explored in detail. It has been widely reported that there is high diversity in origin and function of PGPR and their growth promoting potential may be highly specific to certain soils, plant species, genotypes and cultivars (Lucy et al., 2004; Mehta et al., 2015; Zahid et al., 2015). Hence, a thorough investigation of native bacterial communities, their population and characteristics is required to assess the diversity of indigenous bacteria and their distribution in the rhizoplane of certain crops (Bulgarelli et al., 2013; Piromyou et al., 2013).

Effects of PGPR strains vary for different crops growing in various soil types under variable climatic conditions. Therefore, it is necessary to cultivate region-specific microbial strains for the development of suitable bio inoculum to obtain maximum yield and nutrient content of a specific crop (Farag et al., 2013; Habibi et al., 2014). In view of these facts, present study was designed to isolate indigenous bacterial strains from the rice endosphere growing on clay loam and saline soil at different locations. These strains were screened in vitro for Zn solubilization potential from different Zn ores. The potent strains were characterized for other plant growth promoting traits and inoculated to rice varieties, basmati 385 and super basmati to assess their potential to enhance yield and translocate Zn under net house conditions. The potent strains were identified by 16S rRNA gene analysis.

### MATERIALS AND METHODS

### Sample Collection and Isolation of Bacteria

Representative plants of five rice varieties viz; super basmati, basmati-385, shaheen basmati, kainat basmati, and basmati-515 growing in two types of soil clay loam and saline were sampled from different locations of Punjab. Three to four fields located in the radius of one kilometer growing in same soil type were identified per location. Four-five plants of variable vigor were selected from each field, uprooted with bulk rhizospheric soil, and pooled up to make a representative sample. The samples were placed individually in paper bags, labeled and transported to lab. Bacterial endophytes were then isolated by following the method of Surette et al. (2003) with certain modifications. Briefly, roots of each plant were separated and thoroughly washed with tap water to remove any adhering soil. The root tissues of each sample were mixed thoroughly and then surface disinfected by a 3 min treatment with commercial bleach (5.25% available chlorine), transferred to a 3% hydrogen peroxide solution for 3 min and finally rinsed three times with sterile milli-Q water followed by air drying in sterile filter paper under the safety cabinet. One gram of the roots was crushed and grinded in sterilized mortar and pestle. The potentially endophytic bacteria were isolated by serial dilution plating of sterilized crushed root on Luria Bertani (LB) agar plates (Hassan et al., 2010). LB agar plates were incubated at 28 ± 2 ◦C for 24–36 h. The individual colonies appearing on the plates were picked and purified by re streaking on LB agar plates. The purified strains were preserved in 20% glycerol at −80◦C.

### Screening of Zn solubilizing Bacteria

Potential of isolates to solubilize Zn from various ores such as zinc sulfide (ZnS), ZnO, Zn3(PO4)2, and ZnCO<sup>3</sup> was tested on Bunt and Rovira agar medium (Bunt and Rovira, 1955). The bacterial strains were grown in LB overnight. Five microliter of each bacterial suspension having optical density (OD) normalized to 0.5 was inoculated on specific plates containing 0.1% of the respective ore. The inoculated plates were incubated at 30◦C for 36–96 h. Appearance of halo zone around the colonies indicated their potential to solubilize Zn which was estimated by measuring the zone diameter. There were three biological replicates and the experiment was repeated twice.

### Morphological and Biochemical Characterization of Potent Zn Solubilizing Strains

Two strains exhibiting maximum potential to solubilize Zn from various ores were selected for morphological and biochemical characterization. For each test, the strains were freshly grown in LB broth overnight and normalized to the OD-600 of 0.5 before inoculation on respective plate.

### Gram Reaction and Antibiotic Resistance

The Zn solubilizing strains were characterized for Gram reaction and intrinsic resistance to antibiotics following the method of Vincent (1970). Briefly, the strains were inoculated on agar plates and spreaded by swab. Antibiotic discs of levofloxacin (5µg), streptomycin (10µg), piperociline (100µg), amoxyciline (10µg), tetracycline (30µg), kanamycine (30µg), vanlomycine (30µg), and minocycline (30µg) were placed on the plate of each strain and incubated at 28 ± 2 ◦C for 24–48 h. Inhibition of bacterial growth was observed around each antibiotic disc and strains were designated as highly resistant, moderate resistant, moderate susceptible and highly susceptible on the basis of inhibition zone diameter as recommended by the manufacturer.

### Phosphorus (P) and potassium (K) Solubilizing Activity

Ability of strains to solubilize the major plant nutrients (P, K) were tested on Pikovskaya agar containing tricalcium phosphate as insoluble phosphate source and Aleksandrov agar having potassium aluminum silicate as source of insoluble inorganic potassium respectively (Pikovskaya, 1948; Kumar et al., 2012). Each bacterial culture was spot inoculated in the center of respective agar plates. The plates were incubated at 28 ± 2 ◦C for 7–10 days and observed for the appearance of halo zone around the colonies. Size of the zone diameter around the colonies provided a semi quantitative potential of P and K solubilization of the strains. The experiment was repeated twice with three biological replicates.

### Determination of Indole-3-acetic Acid and Siderophores

Indole-3-acetic acid (IAA) production by the Zn solubilizers was qualitatively determined by growing the strains on LB agar plates supplemented with 100µg mL−<sup>1</sup> of tryptophan (Shrivastava and Kumar, 2011). A 0.5 cm deep cavity of 1–2 cm diameter was made by sterile cork borer. A 100µL of freshly grown culture was inoculated in each cavity. The inoculated plates were incubated at 28 ± 2 ◦C. After 16 h incubation, the bacterial colonies were removed from the cavities with the help of sterile cotton swab and approximately 200µL of IAA reagent consisting of (1 mL of 0.5 M FeCl<sup>3</sup> mixed in 50 mL of 35% HClO4) was added in the cavity and observed for change in color. Appearance of pink halo zone around the cavity indicated production of IAA by the bacterial culture. Qualitative production of siderophores by the bacterial strains was detected on the Chrome-azurol S (CAS) medium (Schwyn and Neilands, 1987). Each bacterial culture was inoculated separately on CAS agar plates and incubated at 28 ± 2 ◦C for 72 h. The plates were observed for the change in color i.e., orange to yellow and zone diameter was measured to estimate the production of siderophores semi quantitatively.

### Antagonism against Economically Important Pathogens of Rice

Two economically important pathogens of rice, Pyricularia oryzae causing rice blast and Fusarium moniliforme causing bakanae disease of rice were used as test strains to study the antagonistic potential of potent Zn solubilizers. The antagonism was tested by dual culture assay as described by Spence et al. (2014) with certain modifications. The fungal disc of diameter 6 mm was placed at the center of potato dextrose agar (PDA) petri plate. The bacterial culture was spotted at a distance of 4 cm from fungal disc. LB broth instead of bacterial culture was used in mocked (control). The PDA plates were sealed with parafilm and incubated at 28 ± 2 ◦C for 8–10 days. The mycelial diameter of fungus growing out from the edge of bacterial colony was measured. Percentage inhibition of fungal mycelium was calculated by using the following formula:

%inhibition = [(C − T) × 100)/C]

Where C = mycelium diameter (cm) of the fungus growing in the control plate and T = mycelium diameter (cm) of the fungus growing in the bacterial treated plates. The experiment was repeated twice with three biological replicates each time.

## Determination of HCN (Hydrogen Cyanide)

Production of HCN was determined by following the method of Miller and Higgins (1970) with certain modifications. The bacterial cells were inoculated on LB agar plates amended with 4.4 g glycine L−<sup>1</sup> . A piece of filter paper having diameter 6 cm was dipped in solution consisting of 0.5% picric acid, 1% Na2CO<sup>3</sup> and placed in the upper lid of each petri plates. The plates were wrapped with parafilm and incubated at 28 ± 2 ◦C for 48–72 h. The change in color of filter paper from yellow to brown was used as indicator for HCN production.

## Determination of Hydrolytic Enzymes

Hydrolytic enzymes production such as protease, cellulase, and glucanase were detected on the agar plates containing skim milk, carboxy methylcellulase and laminarin respectively (Hassan et al., 2011; Kumar et al., 2012; Abraham et al., 2013). The bacterial strains were inoculated on the respective agar plates and incubated at 28 ± 2 ◦C for 4–7days. Development of halo zone around the colonies indicated enzyme production. Zone of exo β-1, 3-glucanase was observed after staining with congo red (Nagpure et al., 2014).

### Molecular Identification of Zn solubilizing Bacteria

Zn solubilizing bacteria were identified at molecular level by sequencing 16S rRNA gene. The genomic DNA of bacterial strains was extracted by CTAB extraction (Wilson, 1987). A 1500 bp 16S rRNA gene was amplified by using the primers P1 (5′ -AGAGTTTGATCCTGGTCAGAACGAACGCT-3′ ) and P6 (TACGGCTACCTTGTTACGACTTCACCCC - 3′ ) as described by Tan et al. (1997). PCR reaction mixture consisting of 10– 15 ng DNA, 1.5 mM MgCl2,1 X PCR buffer, 200µM of each, dATP, dCTP, dGTP, and dTTP (Fermentas), 10 mM of each primer, and 1.0–1.5 U of Taq polymerase (Fermentas) was amplified in thermocycler (Peq lab Germany) with the amplifying conditions; initial denaturation at 95◦C for 5 min, 25 cycles (94◦C for 1 min, 56◦C for 1 min, 72◦C for 1.75 min) followed by final extension at 72◦C for 5 min (Tan et al., 1997; Hassan et al., 2010). The amplified 16S rRNA gene was analyzed on 1% agarose gel and compared with 1 kb DNA ladder (Fermentas). Specific band of 16S rRNA gene was eluted from the gel and purified by using the Gel Extraction Kit (Qiagen). The purified PCR product was sequenced commercially by Macrogen Inc. (Korea). The 16S rRNA gene sequence was annotated, analyzed on BLAST and identified on the basis of closest homologous strain.

### Effect of Zn solubilizing Bacteria on Rice Plants

Effect of Zn solubilizing strains inoculation on plant growth, yield and grain Zn concentration on two rice varieties basmati 385 and basmati super was examined in cleaned earthen pots (20 cm × 30 cm) under net house conditions. The pots were filled with 5 Kg sterilized clay loam soil. NPK fertilizer was applied at the rate of 40, 30, and 20 mg kg−<sup>1</sup> of soil in the form of urea, single super phosphate and potassium sulfate respectively. Phosphorous (P) and K were applied in single dose before sowing the plants while N was applied in three split doses. The experiment was laid out in a completely randomized design (CRD) with three replications per treatment. Zinc sulfate (ZnSO4) at the rate of 7.5 mg kg <sup>−</sup><sup>1</sup> of soil was used as Zn in respective treatments. There were eight treatments viz. T1 = uninoculated plants (Negative control), T2 = Zn (positive control), T3 = Zn solubilizing strains SH-10, T4 = Zn solubilizing strains SH-17, T5 = Consortium of Zn solubilizing strains SH-10 and SH-17, T6 = Zn solubilizing strains SH-10 + Zn, T7 = Zn solubilizing strains SH-17 + Zn, T8 = Consortium of Zn solubilizing strains + Zn.

Roots of 30 days old rice plants obtained from nursery were surface sterilized and transferred aseptically to the pots. The Zn solubilizing rhizobacteria were inoculated as soil drenching near the plant roots after 2 days of seedling transplant. The bacterial strains were grown in LB broth in a 250 mL Erlenmeyer flask on a shaking incubator at 100 rev min−<sup>1</sup> , 28 ± 2 ◦C for overnight. The cells were pelleted by centrifugation and dissolved in 0.85% saline with a cell OD = 0.45 (∼10<sup>9</sup> CFU mL−<sup>1</sup> ). One mL of this cell suspension was applied near the root of each seedling. Sterile saline without bacteria was applied in negative control. The pots were kept in net house during the months of July to October (Natural season of crop). The plants were irrigated when needed until maturity. A second dose of bio inoculants was applied after 45 days of 1st inoculation. The plants were harvested during the month of October and observed for all agronomic traits like number of tillers, plant height, panicle length, thousand grain weight and yield except the leaf chlorophyll content which was measured at anthesis stage. Three leaves per plant were randomly selected and chlorophyll content of each leaf was measured by SPAD meter (Minolta, Tokyo, Japan) from different places (Ranganathan et al., 2006). Plant height was measured from ground level to the tip of panicle by using a measuring rod. The number of tillers were counted by uprooting each plant. The plants of each pot were threshed to separate grains and straw which were weighed separately. Average values of respective parameters were computed and expressed per pot.

### Determination of Grain Zn Content and Zn Translocation Index (ZTI)

Plants of four treatments viz un-inoculated plants (T1), Zn (T2), consortium of strains (T5) and consortium of strains along with Zn (T8) were selected for determining the Zn content in shoot and grains. The Zn analysis was carried-out commercially by the Nuclear Institute for food and Agriculture (NIFA), Peshawar, KPK, Pakistan. Zn translocation index (ZTI) toward rice grains was calculated by using the formula (Rengel and Graham, 1996).

ZTI = [Zn concentration in grains/Zn concentration in shoot]×100

### Statistical Analysis

The data were subjected to analysis of variance using statistical package Genstat 9.2 (VSN International Ltd., Hemel Hempstead, Hertfordshire,UK, Abaid-Ullah et al., 2015). The differences among various treatment means were compared using the Fisher's protected least significant differences test (LSD) at probability level (P ≤ 0.05) (Steel and Torrie, 1980).

### RESULTS

### Prevalence of Zn solubilizing Bacteria in Rice Endosphere

A total of 234 isolates were obtained from different rice varieties growing at varying locations. Number of isolates in rice endosphere were found to be variable i.e., 3–4 isolates per sample (**Figure 1**). Twenty seven isolates solubilized Zn either from one or more Zn ores with a solubilization zone of 1–24 mm (**Figure 2**, Table S1). Distribution of Zn solubilizers associated with rice endosphere was highly variable i.e., 0–18% per sample depending upon the variety and soil conditions (**Figure 1**). Two strains SH-10 and SH-17 showing maximum solubilizing zone on respective Zn ores were designated as potent Zn solubilizers and selected for further studies.

### Morphological and Plant Growth Promoting (PGP) Traits of the Potent Zn Solubilizers

The Zn solubilizers were Gram positive rod. The strain SH-10 resisted all the antibiotics except levofloxacin (5 µg), tetracycline (30µg), vanlomycine (30µg), and minocycline (30µg) while strain SH-17 showed resistance against all the antibiotics. The strains solubilized P and K with a solubilization zone of 38–46 mm and 47–55 mm respectively but did not produce IAA and HCN. They also inhibited the growth of P. oryzae and F. moniliforme by 22–30% and produced the antifungal metabolites protease, cellulase and glucanase. The strain SH-10 did not produce siderophores (**Table 1**).

### Effect of Zn Solubilizers on Yield and Yield Components of Rice Varieties Basmati-385

Zn solubilizing strains significantly increased the yield and yield components of rice variety basmati-385 (**Table 2**). Maximum effect on yield (45.8 g/pot) and yield components i.e., plant height (97.1 cm), number of tillers (31.4/pot), panicle length (20.6 cm),



Values are mean of three replicates and bearing different letters in the same row are significantly different from each other according to the analysis of variance (p < 0.05).

\*,\*\*\*The values are solubilization zones in mm.

\*\*The values are percent inhibition of fungal mycelium.

chlorophyll content (34.5) was observed in co-inoculation of the Zn solubilizers along with the Zn followed by the Zn treatment (**Table 2**). The strain SH-17 and Zn resulted a 35.4 g grain yield/pot which was statistically at par with that of consortium of Zn solubilizers and Zn (45.8 g grain yield/pot) and only Zn treatment (34.5 g grain yield/pot).

### Super Basmati

Effect of Zn solubilizing bacteria on the yield and yield components of rice variety super basmati was significant (**Table 3**). Highest grain yield/pot (38.6 g) and yield components i.e., plant height (100.9 cm), tillers/pot (33.8), chlorophyll content (34.9) and 1000 grain weight (23.2 g) were observed in super basmati plants treated with the consortium of strains and Zn followed by that of treated with strain SH-10 and Zn with grain yield/pot (32.3 g), plant height (92.3 cm), tillers/pot (30.6), chlorophyll content (32.9), and 1000 grain weight (21.4 g). Moreover, effect of treatments i.e., strain SH-10 and Zn, SH-17 and Zn and only Zn on basmati rice was statistically same but different from that of other treatments (**Table 3**).

### Zn Translocation Index of Rice Varieties

Effect of Zn solubilizers on the ZTI of both varieties was similar (**Table 4**). Highest ZTI was observed in the rice plants treated with consortium of strains (ZTI = 1.6–1.7) followed by the plants treated with only Zn (ZTI = 1.3–1.4) or consortium of strains and Zn (ZTI = 1.3–1.4). The lowest ZTI was observed in the uninoculated plants (ZTI = 0.9–1.1). This clearly shows the role of Zn solubilizers in Zn translocation toward rice grains.

### Molecular Identification of Potent Zn Solubilizers

The potent Zn solubilizers were identified as Bacillus sp. and Bacillus cereus on the basis of 16S rRNA gene analysis. The sequences of strains were submitted to NCBI Gene Bank database under accession numbers KT380823 and KT380824.

### DISCUSSION

PGPR inhabit wide range of crops growing under varying agricultural practices and commonly used as bio inoculants. In addition to their synergistic effect on plant's growth and yield, they have strong potential to enhance the Zn content of cereals (Sharma et al., 2013; Wang et al., 2014; Abaid-Ullah et al., 2015). Utilization of such PGPR to enhance (Zn) content of rice grains could be a promising strategy to minimize the Zn deficiency in human beings. Keeping in view the specific advantages of indigenous strains such as host adaptability and field efficacy, certain potent strains were screened in vitro and in vivo to enhance the growth, yield and Zn content of rice varieties.

Among the various potentially endophytic isolates associated with rice varieties, a significant difference in the number of Zn solubilizers was observed in different varieties and soil types. Maximum Zn solubilizers were enumerated from the endosphere of variety super basmati growing in clay loam soil while minimum strains were obtained from the endosphere of basmati kainat. However, in saline soil growing plants, prevalence of Zn solubilizers in the endosphere of all rice varieties was almost similar. The significant difference in the number of Zn solubilizers among different varieties and soil types may be due to the fact that plant microbe interaction is highly dependent on soil conditions and plant genotype (Schreiter et al., 2014; Sugiyama and Yazaki, 2014; Belimov et al., 2015). Variation in quantity and composition of microbes associated with the rhizosphere and endosphere of different plants, species and even varieties within same species have already been well documented (Beneduzi et al., 2013; Lagos et al., 2014; Ling et al., 2014). In a recent study, Hameed et al. (2015) has reported that the diversity of bacteria inhabiting rice endosphere and their distribution as well as PGP characteristics were dependent on multiple factors such as host's genotype, soil characteristics and nutrients. The PGPR strains have been screened from the rhizosphere of rice grown in different countries (de Souza et al., 2013) but less attention has been paid to explore the Zn solubilizers associated with rice, a crop which grows in flooded conditions and numerous factors affect the Zn availability in such conditions (Lefèvre et al., 2014; Abaid-Ullah et al., 2015).

In this study, a large number of potentially endophytic isolates were recovered from the rice endosphere but a very few strains depicted Zn solubilization potential. The strains capable to solubilize maximum Zn from respective ores were further tested for their morphological traits like Gram's reaction, cell shape and colony morphology. These traits are essential to recognize specific bacterial strain and tentative identification (Mohamad et al., 2014).

A potent strain exhibiting multiple PGPR traits must be able to resist extreme environmental conditions so that it may survive and maintain optimum population throughout the life cycle of specific crop. Competitive ability of strain to survive in environment is strongly correlated with its intrinsic antibiotic resistance. The Zn solubilizing strains resisted most


 <




Control, non-inoculated; Zinc, Zinc sulfate (ZnSO4) at the rate of 7.5 mg kg−<sup>1</sup> of soil.

Values are mean of three replicates and bearing different letters in the same column are significantly different from each other according to the analysis of variance (p < 0.05).

of the important antibiotics which depicted their ability to tolerate the environmental stress. PGPR especially belonging to genus Bacillus sp. are able to survive in adverse environmental conditions due to their ability to form endospores and change the fatty acid patterns depending on variable colonizing niches (Checinska et al., 2015; Diomande et al., 2015). Thus, the Zn solubilizers screened in this study could maintain their population in rice rhizosphere throughout the crop cycle.

Phosphorous (P) and K are the macro nutrients required by rice and the other field crops for growth and optimum yield (Saleque et al., 2013; Wang et al., 2013; Damon et al., 2014; Zorb et al., 2014). Uptake of these macro (N, P, K) and micronutrients (Zn) by the plants from soil is mutually dependent (Bouain et al., 2014). They experience a complex processes in soil and exhibit dynamic equilibrium between insoluble and soluble forms under the influence of soil pH. This equilibrium could be affected by the acid secretion and other activities of soil microbiota, thereby enhancing their availability to plant roots for absorption (Saravanan et al., 2004). Zn application enhanced the uptake of macronutrients (N, P, K) in rice and influenced the biological properties of soil (Pooniya et al., 2012). These facts further advocate the worth of Zn solubilizers screened in this study as they also solubilize the mineral P and K.

Ability of rhizobacteria to solubilize Zn, K, and (PO4) −3 is dependent on secretion of acids such as gluconic acids, lactic acid, malic acid and oxalic acid etc (Estrada et al., 2013; Zhang and Kong, 2014; Abaid-Ullah et al., 2015). These three elements [Zn, K, and (PO4)−<sup>3</sup> ] are released from their insoluble compounds by following the same basic mechanism of acidification and thus correlate the metabolism of each other.

A PGPR could be an ideal candidate for the development of bioformulation provided it possesses multiple characteristics for plant growth promotion. The Zn solubilizers suppress growth of economically important pathogens of rice P. oryzae and F. moniliforme and produce several secondary metabolites involved in antagonistic activity of the PGPR. These metabolites include siderophores which chelate iron and deprive the pathogen from an important source of nutrition, thereby inhibiting the pathogen by creating a competitive environment (Aznar and Dellagi, 2015). The strains also produce hydrolytic enzymes like glucanase, cellulase and protease which hydrolyze the various components of cell wall, thereby paralyzing the pathogen which ultimately leads to death of the pathogen (Nagpure et al., 2014). Thus, Zn solubilizing rhizobacteria exhibit multiple PGP traits in vitro which are similar to the earlier findings (El-Sayed et al., 2014; Abaid-Ullah et al., 2015). Suppression of pathogens by Zn solubilizers augment their potential as an effective bioinoculant because they can protect the plants from devastating diseases along with providing nutrition (Abaid-Ullah et al., 2015). In certain bacteria, the properties of Zn solubilization and pathogen suppression has been found to be interlinked. As reported by Saravanan et al. (2007), presence of solubilized Zn in the culture filtrate enhanced the antagonistic activity of Gluconacetobacter diazotrophicus. The siderophores produced by the antagonistic bacteria chelate iron and play major role in their antagonistic activity. On the other hand, Fe <sup>+</sup><sup>3</sup> oxidation under aerobic soil conditions limits the Zn availability (Gao, 2007). This fact advocates the production of siderophores as a mechanism adopted by PGPR for enhancing Zn availability.

In vivo evaluation of the strains exhibiting multiple plant growth promoting properties in laboratory conditions is necessary to develop an effective bio inoculum. Hence, the potent strains were inoculated on two rice varieties, basmati 385 and super basmati grown in net house under the natural growth conditions. The strains inoculated in consortium along with chemical Zn significantly enhanced the yield and related parameters in present research. Effect of PGPR on rice yield has already been reported (Estrada et al., 2013; Ji et al., 2014) but they synergize the effect of chemical Zn on rice basmati varieties, is being firstly reported. As the strains showed their potential to increase yield and yield related traits in vivo they may serve as useful bio inoculum.

In addition to increase yield, the Zn solubilizing rhizobacteria also enhanced Zn translocation to the rice grains in similar way as that of chemical Zn. These findings were similar to earlier studies where PGPRs translocated Zn toward rice grains (Tariq et al., 2007; Sharma et al., 2014; Vaid et al., 2014; Wang et al., 2014). The Zn translocation toward rice grains may depend upon the ability of rhizobacteria to enhance Zn availability by executing multiple mechanisms such as mineralization, solubilization and induction of physiological processes in rice involved in Zn uptake just like the induction of systemic resistance in rice against pathogens (Lucas et al., 2014). However, it needs to explore in future studies. There was no significant increase in Zn translocation toward rice grains when consortium of Zn solubilizers was inoculated along with the chemical Zn. This may be due to the inherent Zn uptake potential of the tested rice varieties, i.e., basmati 385 and super basmati. These findings suggest potential use of strains in rice bio fortification with Zn.

The potent Zn solubilizers capable to enhance growth, yield and grain Zn content of rice were identified as Bacillus sp. and Bacillus cereus. These findings are similar to that of earlier reports in which zinc solubilizers have been found to be recurrent among various bacterial taxa (He et al., 2011; Abaid-Ullah et al., 2015). Certain strains of B. cereus are opportunistic human pathogens but numerous strains isolated from plant rhizoplane exhibit PGPR traits and their potential as bio fertilizer is well documented (Niu et al., 2011; Chun Juan et al., 2012; Ramesh et al., 2014; Chowdhury et al., 2015). Thus, the Bacillussp. capable to solubilize nutrients, produce siderophores and antagonize the pathogens by different mechanisms could be used as effective bio inoculants.

### REFERENCES


### CONCLUSION

The Zn solubilizing bacteria associated with indigenous host enhance the growth, yield and Zn content by providing nutrition as well as disease protection. Such strains could be ideal candidate to develop bioformulation to improve not only rice yield but also produce Zn fortified rice grains. Moreover, these findings could help researchers to explore the mechanisms involved in PGPR mediated Zn translocation in cereals. The Zn solubilizing strains reported in this study would be made available upon request as per institutional material transfer agreement policy.

### ACKNOWLEDGMENTS

We would like to thank Higher Education Commission (HEC), Pakistan for providing funds under NRPU grant no. 20-1982, National institute for food and agriculture (NIFA), Peshawar for Zn analysis of rice samples, Mr. Sajid (Lab attendant) for his help in growing rice plants and all the technical staff involved in maintaining the net/green house. We are highly thankful to Dr. Samina Nadeem (Professor, Department of humanities, CIIT Islamabad) and Dr. Saqib Mumtaz (Assistant Professor, Department of biosciences, CIIT Islamabad) for improving the English of manuscript.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2015.01286

toward a better understanding of their homeostasis interaction. J. Exp. Bot. 65, 5725–5741. doi: 10.1093/jxb/eru314


and related rhizobia. Int. J. Syst. Bacteriol. 47, 874–879. doi: 10.1099/00207713- 47-3-874


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2015 Shakeel, Rais, Hassan and Hafeez. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# High Level of Nitrogen Makes Tomato Plants Releasing Less Volatiles and Attracting More Bemisia tabaci (Hemiptera: Aleyrodidae)

Md. Nazrul Islam1,2,3† , Abu Tayeb Mohammad Hasanuzzaman1,2,4† , Zhan-Feng Zhang1,2 , Yi Zhang1,2 and Tong-Xian Liu1,2 \*

<sup>1</sup> State Key Laboratory of Crop Stress Biology for Arid Areas, Northwest A&F University, Yangling, China, <sup>2</sup> Key Laboratory of Integrated Pest Management on the Loess Plateau of Ministry of Agriculture, Northwest A&F University, Yangling, China, <sup>3</sup> Agrochemical and Environmental Research Division, Institute of food and radiation Biology, Atomic Energy Research Establishment, Dhaka, Bangladesh, <sup>4</sup> Vertebrate Pest Division, Bangladesh Agricultural Research Institute, Gazipur, Bangladesh

Tomato (Solanum lycopersicum) production is seriously hampered by the infestation of the sweetpotato whitefly, Bemisia tabaci MEAM 1 (Middle East-Asia Minor 1). The infestation behavior of the whiteflies could be affected by the quantity of plant released volatile organic compounds (VOCs) related to nitrogen concentrations of the plant. In this study, we determined the infestation behavior of B. tabaci to the tomato plants that produced different levels of VOCs after application of different levels of nitrogen with a wind tunnel and an olfactometer. We also analyzed the VOCs released from nitrogentreated tomato plants using solid phase microextraction and gas chromatography-mass spectrometry. The results revealed that the production of eight VOCs (β-pinene, (+)- 4-carene, α-terpinene, p-cymene, β-phellandrene, α-copaene, β-caryophyllene, and α-humulene) was reduced after the plants were treated with high levels of nitrogen. However, more whiteflies were attracted to the tomato plants treated with high levels of nitrogen than to the plants treated with normal or below normal levels of nitrogen. These results clearly indicated that nitrogen can change the quality and quantity of tomato plant volatile chemicals, which play important roles in B. tabaci host plant selection.

Keywords: Bemisia tabaci, tomato plant, nitrogen, plant volatiles, SPME, GC-MS

## INTRODUCTION

Tomato, Solanum lycopersicum L., is an important vegetable in the world (Olaniyi et al., 2010; Bhowmik et al., 2012). There are many pests that cause both qualitative and quantitative losses of tomato (Lange and Bronson, 1981; Bari and Sarder, 1998). The sweetpotato whitefly, Bemisia tabaci (Genn.) (Hemiptera: Aleyrodoidea) is a major pest insect of vegetables, broadleaf field crops, and ornamentals in the tropics and sub-tropics of the world and in the protected environments of other areas (Liu, 2007). It is also considered one of the most important pests of tomato in the tropical and sub-tropical regions, causing heavy losses to crops by direct feeding and by transmitting geminiviruses (Toscano et al., 2002; Inbar and Gerling, 2008).

#### Edited by:

Gero Benckiser, University of Giessen, Germany

#### Reviewed by:

Ming Wang, UC Riverside, USA Stefanie P. Glaeser, University of Giessen, Germany

### \*Correspondence:

Tong-Xian Liu txliu@nwsuaf.edu.cn

†These authors have contributed equally to this work.

> Received: 31 October 2016 Accepted: 16 March 2017 Published: 31 March 2017

#### Citation:

Islam MN, Hasanuzzaman ATM, Zhang Z-F, Zhang Y and Liu T-X (2017) High Level of Nitrogen Makes Tomato Plants Releasing Less Volatiles and Attracting More Bemisia tabaci (Hemiptera: Aleyrodidae). Front. Plant Sci. 8:466. doi: 10.3389/fpls.2017.00466

Many abiotic factors have been shown to influence the emission of VOCs that affect the host preference of insect pests, their colonization and life histories. Nitrogen is an abiotic factor that could affect the emission of volatiles from crop plants and further affect the behaviors of insects (Berenbaum, 1995; Duffey and Stout, 1996; Veromann et al., 2013; Han et al., 2014). As tomato is a high valued commercial crop, tomato growers have a tendency to over-use fertilizers (Locascio et al., 1992). Many studies have showed that excessive nitrogen application above the standard recommendation often increases nitrogen leaching, causing soil and water pollution, and increasing cost (Zotarelli et al., 2007; Sieling and Kage, 2010; Engström et al., 2011). Most plant feeding insects have a capability to search for the plants with high nitrogen content (Southwood, 1973). Although nitrogen fertilization rarely affects herbivores directly, it can change or alter morphological, physiological, and biochemical characters of host plants and increase food quality for herbivores (Bernays, 1990; Simpson and Simpson, 1990). In general, nitrogen content of a host plant is considered as an indicator of nutritional quality and a factor influencing host plant selection by plant feeding insects (Mattson, 1980). Several studies showed that nitrogen application modifies plant biochemical contents and pest resistance against herbivores. For example, Han et al. (2014) found that sub-optimal nitrogen supply is not favorable for the survival and development of the tomato leafminer Tuta absoluta (Meyrick), and this may be due to an increase of leaf chemical defense system and decrease in leaf nutritional value. In another experiment, Chen et al. (2008) found that cotton plants with high nitrogen were preferred for oviposition by the female beet armyworm Spodoptera exigua (Hübner). Moreover, Bentz et al. (1995a,b) found that protein-nitrogen content was linearly increased in the leaves of the poinsettia (Euphorbia pulcherrima Willd. et Kl.) plant with increasing level of nitrogen application, and B. tabaci host selection was also linearly increased with increasing of nitrogen content in the plants.

The composition and levels of plant VOCs might be significantly affected by nitrogen fertilization (Chen et al., 2010) and consequently affecting their attractiveness to pests. The relation between the levels of fertilizer application and the emission of plant volatiles depends on the plant species either there can be a positive, negative or no relation between them. For instance, Jang et al. (2008) observed a reduction in jasmonic acid levels in rice plants when receiving high amounts of nitrogen in all three cultivars tested. Van Wassenhove et al. (1990) also confirmed that constitutive volatile chemicals extracted from celery significantly decreased with increasing of high levels of mineral and/or organic nitrogen fertilizers. On the other hand, Gouinguenè and Turlings (2002) found that a lower quantity of volatile compounds was released from unfertilized corn plants (Zea mays L. var. Delprim) compared to those receiving a complete nutrient solution. Therefore, the quantitative differences of VOCs released from different plant species can vary when plants receive different levels of fertilization. However, the effects of nitrogen on the pattern of release of VOCs might be system- or species-specific, and there might be a correlation between nitrogen application, volatiles production, and host plant preference of insect pests.

Chemical communication between host plants and herbivores mostly depends on the herbivore and plant species. It is also based on multiple compounds of the plants (Blight et al., 1997). It is well known that many herbivorous insect pests choose their hosts based on visual modalities (optical characteristics of plants), semiochemical stimuli (plant volatile compounds), or both (Bernays and Chapman, 1994a; Schoonhoven et al., 2005; Cook et al., 2007a,b; Hasanuzzaman et al., 2016). In a study, Bleeker et al. (2009) found that monoterpenes and sesquiterpenes released from tomato plants stimulated a response from receptors on the antennae of B. tabaci and these terpene volatiles played an important role in a free-choice bioassay. Ying et al. (2003) also demonstrated that B. tabaci can distinguish different types of host plant volatiles without any visual references.

Host plant volatiles can act as repellents or attractants for herbivores (Unsicker et al., 2009; Dicke and Baldwin, 2010; Mumm and Dicke, 2010). The defensive and nutritional chemistry of the plant leaf is one of the factors that influences host choice and fitness of herbivore insects (Bernays and Chapman, 1994b). Fertilization and allelochemicals can influence selection of hosts by pest population (Slosser et al., 2004; Chau et al., 2005; Germano et al., 2011). In this study, we hypothesized that nitrogen fertilization levels affect the abundance of B. tabaci, and that this potential modification is related to changes in volatile compound bouquets of the plants. We tested our hypothesis by applying different doses of a nitrogenous fertilizer to tomato plant and observed host preferences of the whiteflies through olfactory responses, and determined the plant volatiles emitted from the treated tomato plants. The results of this study deliver evidences that the composition of the volatile compounds from tomato plant is associated with the nitrogen fertilization and this influences host plant selection by B. tabaci.

### MATERIALS AND METHODS

### Plant Growth and Fertilization

Tomato plant [Solanum lycopersicum L., cv. 'Gan Liang Mao Fen 802 F1' seeds (Xian Qinshu Seeds Company Limited, Xian, Shaanxi, China)] were purchased from a local market. The seeds were sown in small plastic pots (7 cm × 7 cm × 8 cm) containing perlite as a plant-growing medium. The plants were grown inside insectaries where environmental conditions maintained at 25 ± 1 ◦C with 60 ± 5% RH and 16:8 h L:D at a light intensity of 1400–1725 lux. All plants were nurtured in insect-proof mesh cages (60 cm × 60 cm × 60 cm).

We used four levels of nitrogen for growing tomato plants namely, 5 mM (T1), 10 mM (T2), 15 mM (T3), and 20 mM (T4) nitrogen where 5 mM was considered as a below normal level, 10 mM a normal level, and 15 mM and 20 mM were as a high level. Wahle and Masiunas (2003) and Wang et al. (2007) reported that the growth and yield of tomato was highest at near about 10 mM nitrogen levels but the tomato growth rate was limited with below 5 mM of nitrogen solution. Nutrient

solutions for growing tomato plants were prepared with 5 mM nitrogen from a commercial fertilizer, 'Kang Pu Jin' with 20-20- 20 of N-P2O5-K2O + Mg + TE (COMPO Expert GmbH, Krefeld, Germany), which was used as a standard control in all treatments. For other three treatments, i.e., T2, T3, and T4, the source of additional nitrogen was urea (Sigma U 1250, purity > 99.5%). The nutrient combination of above mentioned fertilizer was N-P2O5-K2O (20%-20%-20%), Mg (0.3%) and other necessary trace elements, including S (0.8%), B (0.01%), chelated Cu (0.04%), chelated Fe (0.1%), chelated Mn (0.1%), chelated Zn (0.04%), and Mo (0.003%). The pH of the nutrient solutions was corrected at 6.0–6.5 if needed to add concentrated HCl or NaOH. The nutrient solutions were prepared with de-ionized water (Millipore, RiosTM 5).

Different levels of nitrogen containing fertilizer solutions were applied to tomato plants following the technique of Stout et al. (1998) with slight modification. Tomato seeds were sown in the small plastic pots (7 cm × 7 cm × 8 cm) containing perlite as a plant-growing medium and watered with de-ionized water (Millipore, RiosTM 5) up to 8 days after seed sowing. The plants were randomly assigned as T1, T2, T3, and T<sup>4</sup> treatments for 5, 10, 15, and 20 mM nitrogen, respectively. From 9th to 14th day, all plants were fertilized daily with 20 ml of 5 mM nitrogen prepared nutrient solution to avoid severe changes in the level of nitrogen applied to a plant. From days 15 to 20, all plants were fertilized daily with 20 ml of 10 mM nitrogen prepared nutrient solution except T<sup>1</sup> treatment, which were fertilized daily with 5 mM nitrogen prepared nutrient solution. From days 21 to completion of the experiment, T<sup>3</sup> and T<sup>4</sup> treatments were fertilized daily with 20 ml of 15 mM and 20 mM nitrogen prepared nutrient solutions, respectively. At that time, T<sup>1</sup> and T<sup>2</sup> treatments were fertilized daily with 20 ml of 5 mM and 10 mM nitrogen prepared nutrient solutions, respectively, as followed in previous fertilization. To avoid water stress, additional deionized water was added as necessary. Thirty-five days old intact tomato plants were used for all experiments, i.e., wind tunnel and Y-tube bioassays, determination of total nitrogen and volatile compounds containing four to five fully expanded leaves (Supplementary Figure S1).

### Insect Rearing

The whiteflies, B. tabaci MEAM1 (Frohlich et al., 1999), were mass-reared in walk-in insectaries in large screen cages (60 cm × 60 cm × 60 cm) on eggplant Solanum melongena L. (Solanaceae) cv. 'Zichangqie' which was grown in 15 cm diameter plastic pots containing 5:1:1 by volume of peat moss, perlite and vermiculite. The insectaries were maintained at 25 ± 1 ◦C with 60 ± 5% RH and a 16:8 h L:D at a light intensity of 1400–1725 lux, which was similar to the experimental environments in which tomato plants were grown.

### Wind Tunnel Bioassays

Response of adult B. tabaci toward different levels of nitrogen applied tomato plants was tested in a plexiglass wind tunnel (200 cm × 70 cm × 70 cm) which was set up in a small controlled environmental room (temperature 25 ± 1 ◦C, RH 60 ± 5%). A blower pulled the air, and the airflow rate was adjusted at 22 cm s−<sup>1</sup> . Three 18 W fluorescent lamps were set up at the take-off point in the flight chamber. The plastic pot except the plant was covered with polythene bags (EasyOne Oven Bags, Reynolds Kitchens, Lake Forest, Illinois, USA) and aluminum foil to minimize plant-growing media produced volatiles before using the plants. In this test, at a time two different nitrogen doses applied intact tomato plants were placed next to each other on the upwind end of the arena maintaining 30 cm distance to observe that the male or female B. tabaci would fly upwind in the presence of both visual and volatile plant cues. A petri dish containing 100 male or female B. tabaci was kept inside the wind tunnel maintaining a short-distance (1-m) from the two plants. Thereafter, B. tabaci were released there to make a choice their suitable host. The B. tabaci were released at 12:00 h, and 24 h later, the individuals were carefully counted to investigate their preference between two nitrogen treatments in the presence of both visual and olfactory cues. Numbers of whitefly adults were counted 24 h later because it took the adults some time to make a choice and settle down in the wind tunnel against wind flow. Before mass releasing of the whiteflies, five individuals were pre-released to ensure that the mass release does not affect the responses of the whiteflies to the odor sources. Each experiment was conducted three times. The wind tunnel was wiped thoroughly with 70% ethanol and air was pulled through the wind tunnel for at least 2 h before it was re-used.

### Olfactory Choice Test with Y-Tube Olfactometer

A Y-tube olfactometer was used to evaluate the behavioral response of B. tabaci to nitrogen applied tomato plant volatile compounds, following the procedure as previously described by Akol et al. (2003) and Saad et al. (2015) with some modifications. The transparent glass made Y-tube olfactometer consisted of 8 cm long base with 0.8 cm internal diameter, two 8 cm lateral branches at a 60◦ angle from each other. The lateral branches were individually connected to two 3 L glass container with odorless tubes and each glass container contained volatiles releasing one intact tomato plant from each nitrogen treatments. The plant was kept in the glass container for 1 h before sampling for exiting impure volatiles from the system. Plant containing glass containers were covered with thick paper so that B. tabaci individual cannot receive visual cues from the plants. Charcoal filtered humidified and purified air was provided at 100 ml min−<sup>1</sup> to both branches of the Y-tube via odor sources using a vacuum pump (Beijing BCHP Analytical Technology Institute, China) for circulating the volatile organic compounds. The air flow was adjusted and measured by an inline flowmeter (LZB-3WB, Changzhou, Jiangsu, China). Male and female B. tabaci were tested separately. Adult whiteflies were starved for 2 h and then released within 0.5–1.0 cm of the base of the Y-tube with a small PCR tube and their responses assessed for 10 min. A B. tabaci adult that walked into one of the lateral branches of the Y-tube at least 5 cm, stayed a minimum of half a minute and did not return was considered as a positive, responsive individual. If an individual did not make a decision within 10 min, they were

excluded from the results, and considered as non-responsive insect. To eliminate lighting bias, a 20 W fluorescent light was placed vertically 0.5 m over the Y-tube olfactometer. The positions of two lateral branches of the Y-tube were inverted 180◦ after every five insects tested. Each olfactory test was repeated two times for each combination of stimulus pairs, and each replicate consisted of 25 B. tabaci adults tested individually, i.e., a total of 50 B. tabaci adults for each treatment were assayed. The experiment was carried out between 12.00 and 16.00 h in a controlled environment maintained at 25 ± 1 ◦C and 60 ± 5% relative humidity. To minimize plant-growing media produced volatiles, the pot except the plant was covered with the same polythene bags and aluminum foil as described previously. The experimental Y-tube and all glassware were washed with soap with tap water, then distilled water, and finally sterilized with 70% ethyl alcohol before oven drying (120◦C for 3 h) to reduce the contamination risk of previous tested odors.

### Determination of Nitrogen from Tomato Plants

Tomato plants (leaves with stem) were taken separately from each treatment and oven dried at 65◦C for 4 days. The dried plant samples were ground with a mortar and a pestle, taken in poly bag and preserved it at room temperature before analysis. Total nitrogen from tomato plant tissues was measured using the Kjeldahl method. Exactly 0.1 g sample was digested with 4 ml of H2SO<sup>4</sup> at 420◦C for 1 h along with catalyst of K2SO<sup>4</sup> and CuSO<sup>4</sup> at a ratio of 9:1. FOSS 8400 automatic Kjeldahl apparatus (FOSS Analytical AB, Sweden) was used to analyze total nitrogen from tomato plant tissues. Five samples from each treatment were used to measure the quantity of nitrogen.

### Volatile Compound Collection from Nitrogen Applied Tomato Plants

Different levels of nitrogen-treated tomato plants released dynamic headspace volatiles were collected with solid phase micro extraction (SPME) fiber coated with poly dimethyl siloxane-divinyl benzene (PDMS-DVB, 65 µm) purchased from Supelco (Bellefonte, PA, USA). The volatile collection procedure is shown in **Figure 1**. Previously, many researchers collected VOCs in headspace static using the SPME method by placing a plant in a closed chamber, and this method consequently increased the temperature and relative humidity in the chamber, which could affect the normal physiological process in volatile chemicals emissions (Gouinguenè and Turlings, 2002; Egigu et al., 2014). In our experiment, VOCs were collected using the headspace dynamic SPME method. Here, a continuous airflow went through a 35-days old plant which was confined inside a glass jar (3 L) that minimized the variation of temperature and humidity inside the glass jar. Again, the same polythene bags and aluminum foil were used to avoid plant-growing media produced volatiles as described previously. On the top of the lid of the glass container was a large hole that closed with a glass made cover containing a bend opening (4 mm in diameter) which was used to insert the SPME fiber. At the bottom of the glass container was another hole for ventilation. A vacuum pump was used for circulating the air which was purified with activated charcoal and thereafter a Tenax TA adsorbent. The flow rate of the air was maintained at 200 ml min−<sup>1</sup> with a flow meter. The SPME fiber was conditioned at 250◦C for 30 min in a gas chromatograph injection port according to the guideline of the manufacturer. The plant was kept in the glass container for 1 h before sampling for exiting impure volatiles from the system. The SPME needle

was inserted into the opening of the glass container and extended the fiber to absorb dynamic headspace plant volatiles which escaped through the opening of the glass container for 1 h. After absorption of the volatiles, the SPME needle was directly inserted into a gas chromatograph-mass spectrometer (GC-MS) thermal desorption port as soon as possible, and thereafter the fiber was extended and kept 5 min for desorption of volatile substances.

### Analysis of Volatile Compounds

The system consisted of a GC (TRACE 1310, Thermo Fisher Scientific, Waltham, MA, USA) that was used for the separation of volatile chemicals and an MS (ISQ Single Quadrupole MS, Thermo Fisher Scientific, Waltham, MA, USA) used for their detection, identification, and quantification. The thermally desorbed VOCs were separated in a 30 m × 0.25 mm × 0.25 µm film thickness fused silica capillary column (Zebron, ZB-5 MS Ui). Programming splitless injector temperature was maintained at 250◦C whereas MS transfer line and ion source temperatures were maintained at 280◦C. The purified helium (99.999%) was used as a carrier gas and the flow rate was maintained at 1.0 ml min−<sup>1</sup> with constant mode. The initial GC oven temperature was set to 40◦C for 4 min. The oven temperature was increased from 40 to 250◦C at a rate of 8◦C min−<sup>1</sup> and held for 5 min. The MS was operated in an electron ionization (EI) mode. The ion energy and emission current were maintained at 70 eV and 25 µA, respectively. The Xcalibur program (Ver. 2.1, Thermo Electron Corporation, USA) was used for data acquisition which was performed in a total ion chromatogram (TIC) with mass range from 33 to 500 amu. The identification of separated compounds was carried out with NIST 2008 (National Institute of Standards and Technology, Washington, DC, USA) database. Kovats retention index (KI) was calculated for each constituent in relation to a mixture of n-alkanes standard (Van Den Dool and Kratz, 1963), and C7–C<sup>40</sup> (Sigma–Aldrich, Louis, MO, USA). The data were matched to previously published data (**Table 1**). The peak area of each component of the volatiles was the relative quantity (Fang et al., 2013).

### Statistical Analyses

IBM SPSS statistics version 19 (Chicago, IL, USA) was used to conduct all statistical analyses. Data produced from Y-tube olfactometer choice bioassays were analyzed by X 2 test. Wind tunnel bioassay data were analyzed by paired t-test. Tomato plant volatiles and nitrogen were analyzed through one-way analysis of variance (ANOVA); means were separated by the Tukey test. Correlation was used to identify a possible relationship between different levels of nitrogen and volatile constituent production of tomato plants. In all cases, means were considered significant at P < 0.05 level.

### RESULTS

### Two-Way Choice Tests Conducted in a Wind Tunnel

Different levels of nitrogen-treated tomato plants were tested in the wind tunnel to investigate their attractiveness to B. tabaci in the presence of both visual and olfactory cues. The results of these dual choice bioassays are presented in **Figure 2** for B. tabaci females and **Figure 3** for B. tabaci males. The attractiveness of B. tabaci females was found different in 5, 10, 15, and 20 mM N treated tomato plants. The results revealed that significantly more B. tabaci females were found in 15 mM N (t = 7.317, P < 0.05; **Figure 2B**) and 20 mM N (t = 5.034, P < 0.05; **Figure 2C**) than in 5 mM N treated tomato plants. When given an option to choose 10 mM N versus 15 mM N treated plants, B. tabaci females preferred 15 mM N to 10 mM N treated plant (t = 4.508, P < 0.05;

TABLE 1 | Molecular weight, mass peak (m/z), retention time (RT), calculated Kovats indicies (CKI) and tabulated Kovats retention indices (TKI) of the volatile compounds identified from intact tomato plants after four nitrogen treatments (see Figure 6 for relative amounts of volatile organic compounds indicated by peak areas found from different levels of nitrogen-treated tomato plants).


**Figure 2D**). When 10 mM N and 20 mM N were offered, B. tabaci females preferred 20 mM N to 10 mM N treated plant (t = 6.263, P < 0.05; **Figure 2E**). However, B. tabaci females did not show significant preference between 5 mM N and 10 mM N (t = 1.199, P = 0.353; **Figure 2A**) and between 15 mM N and 20 mM N treated tomato plants (t = 1.609, P = 0.249; **Figure 2F**).

The data from the wind tunnel dual choice test showed that adult B. tabaci males did not have significant preference between 5 mM N and 10 mM N (t = 0.122, P = 0.914; **Figure 3A**), between 5 mM N and 15 mM N (t = 2.592, P = 0.122; **Figure 3B**), between 5 mM N and 20 mM N (t = 0.610, P = 0.604; **Figure 3C**), between 10 mM N and 15 mM N (t = 1.131, P = 0.375; **Figure 3D**), between 10 mM N and 20 mM N (t = 1.113, P = 0.382; **Figure 3E**), and between 15 mM N and 20 mM N (t = 0.355, P = 0.757; **Figure 3F**) treated tomato plants.

### Olfactory Bioassay with Y-Tube Olfactometer

The preferences of B. tabaci were assayed in the Y-tube olfactometer toward volatile blends released from different levels of nitrogen-treated tomato plants, and the results are shown in **Figure 4A** for B. tabaci females and **Figure 4B** for B. tabaci males. Adult B. tabaci females showed significant preference among the VOCs of the plants treated with various levels of nitrogen including 5, 10, 15, and 20 mM N. Significantly more B. tabaci females were attracted to 15 mM N (χ <sup>2</sup> = 8.333; P < 0.01) than 5 mM N when 5 mM N and 15 mM N were offered, and to 20 mM N (χ <sup>2</sup> = 5.333; P < 0.05) when 5 mM N and 20 mM N were offered. Similarly, B. tabaci females preferred 15 mM N when 10 mM N and 15 mM N treated tomato plants were offered (χ <sup>2</sup> = 5.000; P < 0.05), and preferred 20 mM N (χ <sup>2</sup> = 5.488; P < 0.05) when 10 mM N and 20 mM N treated tomato plants were offered. However, B. tabaci females did not show significant preference between 5 mM N and 10 mM N (χ <sup>2</sup> = 0.556; P = 0.456), and between 15 mM N and 20 mM N (χ <sup>2</sup> = 1.043; P = 0.307) treated tomato plants (**Figure 4A**).

However, in the choice tests, B. tabaci male adults did not show any significant preference among the VOCs from the plants treated with different levels of nitrogen (5 mM N and 10 mM

N: χ <sup>2</sup> = 1.333; P = 0.248; 5 mM N and 15 mM N: χ <sup>2</sup> = 0.364; P = 0.546; 5 mM N and 20 mM N: χ <sup>2</sup> = 1.140; P = 0.286; 10 mM N and 15 mM N: χ <sup>2</sup> = 0.000; P = 1.000; 10 mM N and 20 mM N: χ <sup>2</sup> = 2.689; P = 0.101; and 15 mM N and 20 mM N: χ <sup>2</sup> = 0.184; P = 0.668) (**Figure 4B**).

### Amounts of Nitrogen in Tomato Plants

The results of total nitrogen in different levels of nitrogen-treated tomato plants are shown in **Figure 5**. Significant variation of total nitrogen in tomato plants (leaves with stem) were found due to different levels of nitrogen application. Tomato plants grown in 15 mM (T3) and 20 mM (T4) nitrogen levels had significantly higher percentage of total nitrogen than the plants grown in 5 mM (T1) and 10 mM (T2) nitrogen (F3,<sup>16</sup> = 12.163, P < 0.001). There was no significantly difference of percentages of total nitrogen in T<sup>1</sup> and T<sup>2</sup> treatments. Similarly, no significant difference of percentages of total nitrogen was observed between T<sup>3</sup> and T<sup>4</sup> treatments (**Figure 5**). However, different levels of nitrogen affected plant weight (Supplementary Figure S2).

### Volatile Organic Compounds Identified from Intact Tomato Plants after Application of Different Levels of Nitrogen, Quantity of Volatile Constituents and Correlation between Nitrogen Levels

Effects of nitrogen on tomato plant VOCs are shown in **Figures 6A–P**, and GC-MS chromatograms of volatiles are presented in **Figures 7A–D**. Sixteen VOCs were identified from the tomato plants with nitrogen fertilization treatments. Besides plant emitted VOCs, some compounds are generally related with the analytical system such as phthalates or siloxanes, as well as compounds related with earth's atmosphere such as benzene and toluene (Warneke et al., 2001; Jansen et al., 2008) were not included in the list. The quantities of eight VOCs emitted from tomato plants, including heptanal (F3,<sup>12</sup> = 2.461, P = 0.113; **Figure 6A**), α-pinene (F3,<sup>12</sup> = 2.367, P = 0.122; **Figure 6B**), myrcene (F3,<sup>12</sup> = 1.993, P = 0.169; **Figure 6D**), nonanal (F3,<sup>12</sup> = 0.220, P = 0.881; **Figure 6I**),

δ-elemene (F3,<sup>12</sup> = 2.007, P = 0.167; **Figure 6J**), longifolene (F3,<sup>12</sup> = 0.177, P = 0.910; **Figure 6L**), α-cedrene (F3,<sup>12</sup> = 0.733, P = 0.552; **Figure 6M**), and farnesan (F3,<sup>12</sup> = 0.178, P = 0.909; **Figure 6P**), were not significantly varied with increasing nitrogen treatments. The quantities in the remaining eight identified volatiles differed significantly among the N levels applied, including β-pinene (F3,<sup>12</sup> = 8.863, P < 0.01; **Figure 6C**), (+)-4-carene (F3,<sup>12</sup> = 7.853, P < 0.01; **Figure 6E**), α-terpinene (F3,<sup>12</sup> = 5.290, P < 0.05; **Figure 6F**), p-cymene (F3,<sup>12</sup> = 5.875, P < 0.05; **Figure 6G**), β-phellandrene (F3,<sup>12</sup> = 14.110, P < 0.001; **Figure 6H**), α-copaene (F3,<sup>12</sup> = 16.683, P < 0.001; **Figure 6K**), β-caryophyllene (F3,<sup>12</sup> = 29.783, P < 0.001; **Figure 6N**), and α-humulene (F3,<sup>12</sup> = 9.029, P < 0.01; **Figure 6O**). Of the eight significantly varied VOCs including β-pinene, (+)-4-carene, α-terpinene, p-cymene, β-phellandrene, α-copaene, β-caryophyllene, and α-humulene, no significant differences were found between 5 mM (T1) and 10 mM (T2) nitrogen-treated plant volatiles except for β-pinene, p-cymene, and α-humulene. The tomato plants at 15 mM (T3) and 20 mM (T4) nitrogen produced significantly lower levels of β-pinene, (+)-4-carene, α-terpinene, p-cymene, β-phellandrene, α-copaene, β-caryophyllene, and α-humulene as compared with

the tomato plants grown at the 5 mM (T1) and 10 mM (T2) nitrogen. Of the eight significantly varied VOCs, no significant differences were observed between 15 mM (T3) and 20 mM (T4) nitrogen-treated plant volatiles except for α-terpinene and α-humulene. The amounts of significantly varied eight VOCs identified are well and negatively correlated with the N levels in the tomato plants, the higher the N levels, the lower of the VOCs identified (r = −0.883 to −0.967).

### DISCUSSION

In our study, the tomato plants treated with below normal and normal nitrogen levels of nitrogen (T<sup>1</sup> and T2) had similar plant nitrogen contents. Similarly, the tomato plants treated with higher nitrogen levels (T<sup>3</sup> and T4) showed similar plant nitrogen contents (**Figure 5**). Han et al. (2014) reported that optimal nitrogen-treated tomato plants showed statistically similar leaf nitrogen content to those treated with high nitrogen, but the amount of tomato leaf nitrogen was statistically higher than the plants treated with low levels of nitrogen input. However, our results demonstrated that the nitrogen induced VOCs emitted from tomato plants influenced the behavioral responses of B. tabaci female to host plants. For instance, in the wind tunnel bioassay, B. tabaci females preferred higher levels of nitrogentreated tomato plants (**Figure 2**). The Y-tube olfactometer tests were also provided strong evidence that B. tabaci females chose the VOCs from T<sup>3</sup> and T<sup>4</sup> plants without visual or physical contact with the plants (**Figure 4A**). In the free-choice experiment, the whiteflies could make a decision to choose their host plant with different morphological characteristics, volatile constituents of the plants, or both. However, in the olfactometer experiments, the whiteflies chose the plants only based on the VOCs from the plant, not the morphological characters. Addesso and McAuslane (2009) reported that without volatile compounds, the pepper weevil (Anthonomus eugenii) adults were more inclined to move downwind or remain stationary than to move upwind in the wind tunnel. Therefore, volatile compounds play a critical role to choose the host plant. In our study, the whiteflies showed similar preferences to the VOC from the plants treated with different amounts of nitrogen in both the wind tunnel and the olfactometer experiments. These results are consistent with the findings of Saad et al. (2015) who found that the host plant preference of B. tabaci was similar in the free-choice and olfactometer experiments.

The wind tunnel and Y-tube olfactometer bioassays revealed that the females and males of B. tabaci had different responses to the VOCs emitted from the plants treated with different amounts of nitrogen. For instance, the male adults of B. tabaci did not show any preference to the VOCs emitted from the plants treated with different amounts of nitrogen (**Figures 3**, **4B**). It is generally believed that females play a greater role in finding plants for oviposition for offspring development than males (Zu Dohna, 2006). Moreover, this host finding behavioral difference between the sexes supported the general pattern where many female insects are more attracted by host plant odors than males (Raguso et al., 1996; Zhang et al., 1999; Das et al., 2007; Szendrei and Rodriguèz-Saona, 2010). Saad et al. (2013) and Zheng et al. (2013) reported that B. tabaci and Aleurodicus dispersus females significantly preferred chili plant odors and hexanol isomers, respectively, in an olfactometer assay whereas the males did not show significant responses. Sex-specific olfactory responses and host selection of B. tabaci based on volatile compounds are comparatively poorly studied. It is possible that volatile chemicals emitted by host plants are assumed to mainly affect females (Finch, 1980), whereas pheromones stimulates higher responses from male insects than from female insects (Li and Maschwitz, 1985). Therefore, further studies on the differences between females and males of B. tabaci are needed.

The results of our wind tunnel and olfactometer experiments showed that higher amounts of nitrogen receiving plants, for example T<sup>3</sup> and T4, were highly preferred by B. tabaci females compare to T<sup>1</sup> and T<sup>2</sup> nitrogen-treated plants, indicating that the behavioral response of B. tabaci females could be attributed quantitative differences of nitrogen induced volatile compounds released by the plants. In the volatile analysis test, we found that highly attractive plants emitted significantly less amount of monoterpenes (β-pinene, (+)-4-carene, α-terpinene, β-phellandrene, and p-cymene) and sesquiterpenes (α-copaene, α-humulene, and β-caryophyllene) from the intact tomato plants. This result indicated that production of monoterpenes and sesquiterpenes, generally decreased with the increase of nitrogen levels in tomato plants (**Figures 5**, **6**), and these terpene volatiles could influence the behavioral response of B. tabaci. This result is in agreement with those reported by Tuomi et al. (1994); Lee et al. (2005), and Chen et al. (2008), who found that plants grown at a high level of fertilizer had a lower terpene concentration than the plants grown in a low level of fertilizer. Similarly, Fretz (1976) reported that a low level of terpene was found in Juniperus horizontalis after additional nitrogen fertilizer

application. It has been found that host plant resistance of B. tabaci is related to the optimal nutrient (nitrogen) content of the plants. Reddy and Rao (1989) and Bentz et al. (1995b) reported that nutrient levels (nitrogen) supplied to a plant can increase its nutrient quality, such as, increase of leaf nitrogen content could interfere the natural resistance mechanism of the host plant to insects. Nitrogen may influence nutritional values and semiochemicals of plants and also behavioral characteristics of herbivores. For instance, Jauset et al. (1995) and Zaini et al. (2013) showed that B. tabaci populations were higher when crops were provided with higher levels of nutrients than low levels of nutrients. Our results also showed that B. tabaci female preferred the plants with high levels of nitrogen. Therefore, we think that high levels of nitrogen-treated tomato plants exhibited

to B. tabaci. The amount of plant VOCs released by individual plants can vary with abiotic factors that impact plant's physiology.

different plant defense mechanism that become more attractive

These variations may induce different defense mechanisms of the plants to herbivores, and further affect the infestation performance and behavior of insect pests (Gonzales et al., 2002). In our study, the tomato plant treated with normal and below normal levels of nitrogen emitted high amounts of volatile monoterpenes and sesquiterpenes compared to the plants treated high levels of nitrogen. Monoterpenes and sesquiterpenes increased with the decrease of nitrogen level in tomato plants are consistent with the carbon-nutrient balance hypothesis where predicted secondary metabolites production will be affected in case of lack of any nutrients (Gershenzon and Croteau, 1991). Growth rate of the plant will be reduced due to low nutrient availability but carbohydrate accumulation continues by constant photosynthesis due to carbon availability. Subsequently, accumulation of carbohydrate will lead to synthesis of constitutive secondary compounds like terpenoids (Herms and Mattson, 1992; Gershenzon, 1994). Therefore, we think that T<sup>1</sup> and T<sup>2</sup> nitrogen-treated tomato

P = 0.05 (Tukey test; one-way ANOVA)

plants produced high amount of insect repelling terpenes, especially, monoterpenes and sesquiterpenes than T<sup>3</sup> and T<sup>4</sup> nitrogen-treated tomato plants, which is the possibly reason to move a higher number of B. tabaci females to T<sup>3</sup> and T<sup>4</sup> nitrogen-treated tomato plants in both wind tunnel and olfactometer experiments. These terpene volatiles released from plants in low amount are often reduced in defense against herbivores (Bleeker et al., 2009; Fang et al., 2013). A number of monoterpenes and sesquiterpenes have been reported to display repellent property to B. tabaci. For example, virus infected tomato plant was significantly susceptible to B. tabaci and that plant had a significant lower concentration of the volatiles α-pinene, limonene, 4-carene, thymine, β-phellandrene, caryophyllene, α-cedrene, β-cedrene, and α-humulene than the healthy plant (Fang et al., 2013; Luan et al., 2013). Similarly, significantly less amount of monoterpenes (e.g., p-cymene, 1, 8-cineole) and sesquiterpenes (e.g., α-copaene, β-cedrene) emitting healthy plants were more preferred by B. tabaci females than infested plants (Saad et al., 2015). Our results also showed that the concentration of certain terpenes decreased with the increase of nitrogen levels in tomato plants that was significantly attractive to B. tabaci as shown in the wind tunnel and olfactometer experiments. Zhang et al. (2004) and Yang et al. (2010) conducted a similar olfactory bioassay of B. tabaci with a vertical olfactometer and showed that ginger oil extract repelled adult B. tabaci. A mixture of volatile constituents, including monoterpenes (p-cymene, α-terpenene, β-phellandrene, 1,8 cineole, camphene), sesquiterpenes, alcohols, and aldehydes, were associated to the repellent properties of an essential oil (Owolabi et al., 2007; Ukeh et al., 2009; Sa-Nguanpuag et al., 2011; Nampoothiri et al., 2012). Similarly, Simmons and Gurr (2005) and Bleeker et al. (2009) reported that B. tabaci prefer cultivated tomato plants to wild tomato plants, and their work showed that wild tomato plants released higher levels of terpenes, such as p-cymene, α-terpinene, γ-terpinene and phellandrene, which act as a repellent to B. tabaci. Besides monoterpenes, the antennae of the whitefly are also able to detect certain sesquiterpenes, such as zinziberene and curcumene even in low concentrations. Olson et al. (2009) reported that Spodoptera exigua larvae preferred a higher level of nitrogen receiving induced cotton plants during their growth to recommended level of nitrogen-treated plant, because of lower amounts of terpene volatiles (e.g., ocimene, linalool, indole) emitted

from the higher level of nitrogen receiving cotton plants in compare to recommended level of nitrogen receiving plant. These results support the conclusion that nitrogen plays a role in the release of terpenes from tomato plants, which affect host preference of B. tabaci.

This experiment provide a valuable message that high levels of nitrogen hampers the defenses of tomato plant against B. tabaci by decreasing the quantity of different volatile organic compounds, including β-pinene, (+)-4-carene, α-terpinene, p-cymene, β-phellandrene, α-copaene, β-caryophyllene, and α-humulene. Further study is needed to evaluate each of these volatiles for its effect on the behavior of B. tabaci to find out which has the most adverse effect on B. tabaci as a repellent or an attractant. This finding can be used in future to identify B. tabaci repellents and attractants that could be used as tools of IPM of B. tabaci and other whitefly pests.

### AUTHOR CONTRIBUTIONS

Conceived and designed the experiments: MI, AH, T-XL. Performed the experiments: MI, AH, YZ, and Z-FZ. Analyzed the data: MI, AH, and YZ. Contributed reagents/materials/analysis tools: MI, AH, YZ, and Z-FZ. Wrote the paper: MI, AH, and T-XL.

### REFERENCES


### FUNDING

Funding of this research was partly fulfilled by the following grants: the National Basic Research Program of China (973 Project No. 2013CB127600), National Natural Science Foundation of China (No. 31272089), and China Agriculture Research System (No. CARS-25-B-06)

### ACKNOWLEDGMENTS

We greatly appreciate Prof. James Ridsdill-Smith (University of West Australia, Perth) for critical review and editing of this manuscript. Cordial thanks are expressed to N. Chen, Y.-H. Liu, and H.-H. Cao for their valuable technical instruction and data collection. We are glad for the support of all students and staff in the Key Laboratory of Applied Entomology, Northwest A&F University at Yangling, Shaanxi, China.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.00466/ full#supplementary-material



at greenhouse scale. Ann. Appl. Biol. 154, 441–452. doi: 10.1111/j.1744-7348. 2008.00311.x



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer SG and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.

Copyright © 2017 Islam, Hasanuzzaman, Zhang, Zhang and Liu. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Unraveling Aspects of Bacillus amyloliquefaciens Mediated Enhanced Production of Rice under Biotic Stress of Rhizoctonia solani

Suchi Srivastava<sup>1</sup> , Vidisha Bist<sup>1</sup> , Sonal Srivastava<sup>1</sup> , Poonam C. Singh<sup>1</sup> , Prabodh K. Trivedi<sup>2</sup> , Mehar H. Asif <sup>2</sup> , Puneet S. Chauhan<sup>1</sup> and Chandra S. Nautiyal<sup>1</sup> \*

<sup>1</sup> Division of Plant Microbe Interactions, Council of Scientific and Industrial Research (CSIR)-National Botanical Research Institute, Lucknow, India, <sup>2</sup> Gene Expression Lab, Council of Scientific and Industrial Research (CSIR)-National Botanical Research Institute, Lucknow, India

Rhizoctonia solani is a necrotrophic fungi causing sheath blight in rice leading to

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Tapan Kumar Adhya, KIIT University, India Zakira Naureen, University of Nizwa, Oman

\*Correspondence: Chandra S. Nautiyal nautiyalnbri@lycos.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 23 December 2015 Accepted: 18 April 2016 Published: 06 May 2016

### Citation:

Srivastava S, Bist V, Srivastava S, Singh PC, Trivedi PK, Asif MH, Chauhan PS and Nautiyal CS (2016) Unraveling Aspects of Bacillus amyloliquefaciens Mediated Enhanced Production of Rice under Biotic Stress of Rhizoctonia solani. Front. Plant Sci. 7:587. doi: 10.3389/fpls.2016.00587 substantial loss in yield. Excessive and persistent use of preventive chemicals raises human health and environment safety concerns. As an alternative, use of biocontrol agents is highly recommended. In the present study, an abiotic stress tolerant, plant growth promoting rhizobacteria Bacillus amyloliquefaciens (SN13) is demonstrated to act as a biocontrol agent and enhance immune response against R. solani in rice by modulating various physiological, metabolic, and molecular functions. A sustained tolerance by SN13 primed plant over a longer period of time, post R. solani infection may be attributed to several unconventional aspects of the plants' physiological status. The prolonged stress tolerance observed in presence of SN13 is characterized by (a) involvement of bacterial mycolytic enzymes, (b) sustained maintenance of elicitors to keep the immune system induced involving non-metabolizable sugars such as turanose besides the known elicitors, (c) a delicate balance of ROS and ROS scavengers through production of proline, mannitol, and arabitol and rare sugars like fructopyranose, β-Dglucopyranose and myoinositol and expression of ferric reductases and hypoxia induced proteins, (d) production of metabolites like quinazoline and expression of terpene synthase, and (e) hormonal cross talk. As the novel aspect of biological control this study highlights the role of rare sugars, maintenance of hypoxic conditions, and sucrose and starch metabolism in B. amyloliquefaciens (SN13) mediated sustained biotic stress tolerance in rice.

Keywords: Rhizoctonia solani, B. amyloliquefaciens, biological control, sheath blight, plant growth promotion

## INTRODUCTION

Rice is globally the second most important cereal after wheat catering to the calorific and nutritional needs of more than 40% of the global population. Many biotic stresses hamper rice production, specifically fungal diseases and cause huge economic losses. Among different fungal diseases, rice sheath blight caused by Rhizoctonia solani is a major production constraint causing annual yield losses up to 25–40% (Lee and Rush, 1983). The disease manifests initially as water soaked lesions on sheath of lower leaves and moves up the plant infecting both sheaths and

leaves by joining the lesions (Lee and Rush, 1983; Kumar et al., 2009; XiaoXing et al., 2013). Conventional methods of introducing resistance to disease involve selection breeding, molecular breeding (XiaoXing et al., 2013; Hossain et al., 2014; Vasudevan et al., 2014) and development of transgenics through mapping and expressing different genes (Datta et al., 2001; Kalpana et al., 2006; Yadav et al., 2015). While the conventional breeding techniques are constraint by requirement of long time, development of transgenics becomes a matter of acceptance and propagation in many countries (Gewin, 2003). Therefore, quick alternatives used for disease management focuses on extensive use of fungicides, which creates concern about environmental health, pathogen resistance, and escalating costs (Slaton et al., 2003). Other alternatives include use of various plant extracts, microbial based products, and nutritional amendments for controlling the disease (Kumar et al., 2009; Carvalhais et al., 2013).

In the context of increasing concern for food and environmental safety, use of biocontrol agents and plant growth promoting rhizobacteria (PGPR) for reducing agrochemical inputs in agriculture is considered as potentially sustainable means to control the disease (Herman et al., 2008; Srivastava et al., 2012; Nautiyal et al., 2013; Chowdhury et al., 2015). Microorganisms capable of directly antagonizing fungal pathogens by competing for the niche and essential nutrients, or by producing fungitoxic compounds (biofungicides) and inducing systemic acquired resistance are promising environment friendly methods for crop-management (Herman et al., 2008; Carvalhais et al., 2013; Nautiyal et al., 2013; Tóth and Stacey, 2015). Molecular studies on pathogenesis and stress related genes in rice cultivars have generated volumes of data and knowledge suggesting various signaling pathways and their regulation to play key roles in the crosstalk between plant and biotic/abiotic stresses for plant protection (Fujita et al., 2006; Zheng et al., 2013; Sayari et al., 2014). A lot of molecular and chemical cross talk is known to occur between a plant and the interacting microbe (de Souza et al., 2016). However, there may be unconventional mechanisms working in latency that may have a holistic effect in maintaining plant health. Since mutualistic plant–microbe associations are known to impart physiological and molecular benefits, they may be the constant source of plant health stimulant (Carvalhais et al., 2013; Tóth and Stacey, 2015). This interaction/cross talk of plant with a pathogen and a PGPR are though overlapping to some extent becomes specific depending on the nature of the interacting microbe at later stages (Pauly et al., 2006; Tóth and Stacey, 2015). Yet, there also exist a condition of tripartite interaction when a pathogen attacks a bacteria (biocontrol or PGPR) treated plant. We hypothesize a delicate balance between the pathways followed by the plant in presence of a pathogen or a PGPR and some latent mechanisms to combat disease incidence and growth promotion. This largely unexplored multivariate interaction among Plant–PGPR–Pathogen is the subject of the present study. The study presents a detailed analysis of molecular and physiological mechanisms modulated by plant growth promoting strain Bacillus amyloliquefaciens to explore the unconventional mechanisms employed by the rice plant against sheath blight.

### MATERIALS AND METHODS

### In Vitro Fungus–Bacteria Interaction Assays

Biocontrol efficacy of B. amyloliquefaciens SN13 (Nautiyal et al., 2013) under in vitro condition was determined by growing the pathogen R. solani and PGPR SN13 on Potato Dextrose Agar (PDA) plates using dual culture technique and in different combinations in Czapek dox broth medium. The treatments were SN13 added 1 day before, SN13 and R. solani inoculated simultaneously and SN13 was added 1 day after R. solani inoculation. Culture supernatant after 3, 5, 7, and 10 days of inoculation was used for determination of protease, cellulase, β-1,3-glucanase and chitinase activity.

Proteolytic assay was carried out using 100 µl supernatant using casein as substrate and measured in terms of liberated tyrosine at 660 nm (Frey-Klett et al., 2011). Cellulolytic activity was determined by adding 100 µl supernatant to 1 ml of 0.5% carboxymethyl cellulose (CMC) prepared in 0.5 M phosphate buffer (pH 7.0) followed by 1 h incubation at 37◦C. β-glucanase and chitinase assays were carried out by the co-incubation of 0.5% laminarin (prepared in 0.5 M phosphate buffer pH 7.0) and colloidal chitin in 50 mM acetate buffer along with the culture supernatant in 1:1 ratio at 37◦C for 1 h. The liberated oligosaccharides were measured using dinitrosalicylic acid (DNS) at 540 nm.

### Greenhouse Studies

Role of B. amyloliquefaciens NBRISN13 in suppressing sheath blight disease in rice was assessed under greenhouse conditions. Susceptible rice cultivar Narayan and the virulent isolate of sheath blight pathogen R. solani were used in all the experiments. Soil application of B. amyloliquefaciens NBRISN13 and transplantation of 15 days old seedlings was carried out as described earlier (Nautiyal et al., 2013). Plants were inoculated with the pathogen R. solani at maximum tillering stage (45 days after transplantation) by placing sclerotia wrapped in moist absorbent cotton at the lowest inner sheath of the main tiller and the plants were covered with inverted polythene bags to maintain humidity for 48 h to ensure pathogen infection (Marshall and Rush, 1980; Zheng and Wang, 2011). Treatments were (a) SN13 (b) R. solani and (c) SN13 + R. solani (d) control. Observations were recorded for (a) development and severity of sheath blight symptoms at 15, 30, and 45 days post inoculation (dpi) (b) microarray analysis and GCMS profiling at 45 dpi.

### Histological Assays

Plants harvested at 15 dpi were used for microscopic examinations. The leaf samples were fixed in 70% ethanol:acetic acid (3:1) for 12 h followed by two washings for 1 h each in 70% ethanol. The fixed samples were preserved in 70% ethanol and stored at 4◦C until use. The outermost covering of the stem was stained with 0.1% trypan blue to study surface colonization of the pathogen. Application of stain was followed by heat treatment in a microwave at the high energy level (∼500 watts) for 10 s in flooded condition. The sample was washed and destained in 70%

ethanol for 10 min and observed under microscope. To study the starch distribution in stem, hand cut transverse sections were stained with Lugol's solution (0.01% KI) which turned the starch granules blue to black. Brown colored granules were indicative of inhibition of KI and starch reaction due to presence of organic acids or phenolic compounds such as ascorbic, gallic, and chlorogenic acid as suggested earlier (Kurata and Yamamoto, 1998).

### Sugar, Chlorophyll, and Proline Content in Greenhouse Grown Plants

Sugar estimation in greenhouse grown plants was performed using 0.2 g of fresh leaves crushed with 80% methanol as per the protocol of Dubois et al. (1956). Total chlorophyll and proline content were determined as described earlier by Nautiyal et al. (2013).

### Defense Enzyme Assays

Defense enzyme assays, i.e., superoxide dismutase (SOD), ascorbate peroxidase (APX), and catalase (CAT) were performed after crushing 1.0 g of plant tissue in extraction buffer containing 0.1 M potassium phosphate buffer, 0.1 mM EDTA, 1% polyvinylpyrrolidone (PVP), PMSF (protease inhibitor), and dithiothreitol.

Superoxide dismutase activity was determined by measuring the inhibition of photochemical reduction of nitroblue tetrazolium (NBT; Beauchamp and Fridovich, 1971). APX activity of the samples was quantified as per the method of Nakano and Asada (1981). The enzyme activity was expressed as µmoles of ascorbate oxidized (ε = 2.8 mM−<sup>1</sup> cm−<sup>1</sup> ) per minute per gram fresh weight. Catalase activity was determined at 25◦C according to Aebi (1984). The enzyme activity was expressed as µmoles of H2O<sup>2</sup> degradation per minute per gram of fresh weight. All enzyme activities were converted into Units mg−<sup>1</sup> protein.

### Cell Wall Degrading Enzyme Assays

Rice leaves were homogenized in sodium phosphate buffer (pH 7.0) containing 20 mM cysteine HCl, 20 mM EDTA and 0.5% triton X-100 and different cell wall degrading enzymes (CWDE) like polygalacturonase (PG), pectin methyl esterase (PME) and endopectate lyase (PL) and β-glucosidase (β-gluc) activity in the homogenate was measured as described earlier (Nautiyal et al., 2013).

### Hormone Analysis

Different hormones such as gibberellic acid (GA), indole-3-acetic acid (IAA), abscisic acid (ABA), and salicylic acid (SA) were analyzed by modified method of Wu and Hu (2009) using HPLC. Primary stock solutions (1 mg ml−<sup>1</sup> ) of GA, IAA, ABA, and SA were prepared and diluted (0.5–50 µg ml−<sup>1</sup> ) in methanol. All solutions were stored at −20◦C until use.

Leaf tissues were homogenized with 70% methanol with constant overnight stirring at 4◦C and evaporated on rotavapor under vacuum. The remaining aqueous phase was adjusted to pH 8.5 with phosphate buffer (pH 8.5) and partitioned with ethyl acetate thrice. Ethyl acetate phase was removed and aqueous phase was adjusted to pH 2.5 using HCl and partitioned with diethyl ether. Diethyl ether fraction was evaporated to dryness and dissolved in 2 ml of methanol and used for HPLC analysis.

Qualitative and quantitative analysis for separation of all compounds was achieved by HPLC-PDA with a LC-10 system comprising LC-10AT dual pump system, SPD-M20A PDA detector, and rheodyne injection valve furnished with a 20 µl sample loop (Shimadzu, Japan). Compounds were separated on a 250 mm × 4.6 mm, i.d., and 5 µm pore size Merck RP-C18 column protected by guard column containing the same packing. Mobile phase was isocratic, consisting of 30 mM orthophosphoric acid in HPLC-grade water (component A) and acetonitrile (component B) in 70:30 ratios. HPLC was run at the flow rate of 0.8 ml min−<sup>1</sup> for 30 min. Data was recorded and analyzed on different wavelength viz. IAA at 265 with 12.50 retention time (RT), ABA, and GA at 208 with 14.43 and 6.31 RT and SA at 280 nm with a RT of 15.92. Data was integrated by Shimadzu class VP series software after comparing with standards, results were mean values of three replicates (**Supplementary Figure S2**).

### GC–MS Analysis

Gas chromatography–mass spectrometry (GC–MS) analysis of the rice leaves was performed using methanolic extract of rice leaves as described earlier (Fukusaki et al., 2006). Fresh rice leaves (1 g) were extracted in 5 ml of methanol and water mixture (2.5:1v/v). The polar phase was lyophilized to dryness and TMS derivatives were prepared. The derivatized mixture was analyzed on Gas Chromatography–Mass Spectrometry (GC– EIMS) on a Thermo Fisher TRACE GC ULTRA coupled with DSQ II Mass Spectrometer using TR 50MS column (30 m × 0.25 mm ID × 0.25 µm, film thickness). Conditions used were – carrier gas: Helium; flow rate: 1 ml min−<sup>1</sup> ; injector temperature: 230◦C; oven temperature: started from 70◦C, (hold time 5.0 min) to 290◦C with ramp of 5◦C min−<sup>1</sup> (hold time 5 min); sample injection: split mode (1:50); injection volume: 1 µl; ion source temperature: 220◦C; transfer line temperature: 300◦C and ionization: electron impact mode at an ionization voltage of −70 eV. Mass range was used from m/z 50 to 650 amu (**Supplementary Figure S3**). Identification of individual compounds was carried out by comparison of their mass spectra with those of the internal reference mass spectra library (NIST/Wiley).

### Microarray Analysis

Total RNA was isolated from liquid N<sup>2</sup> frozen leaf blade tissues of each treatment using the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany). Three independent replicated microarray analysis experiments were carried out using Affymetrix GeneChip <sup>R</sup> Rice Genome Arrays (Gene Expression Omnibus platform Accession No. GPL2025). Preparation of targets, arrays hybridization, etc., were carried out according to manufacturer's instructions (Affymetrix, USA) and analyzed using Affymetrix GeneChip Operating Software (GCOS version 1.3). Normalization and differential expression analysis was carried out using dChip software (DNA-chip Analyser). Significant differentially

expressed genes were selected after a combined criterion of >twofold at P < 0.05 in the t-tests after normalization at different level.

### Real Time PCR Analysis

The same total RNA was used for first strand cDNA synthesis using maxima H minus first strand cDNA synthesis kit (Fermentas, Thermo scientific) as per manufacturer's instructions. Real-time PCR for randomly selected genes using actin as an internal reference gene on an Agilent technology, Stratagene Mx3000P Quanti TectTM SYBR <sup>R</sup> Green PCR kit (Qiagen) (Supplementary Table S1). The reactions were performed using the cycle conditions of an initial denaturation at 94◦C for 5 min, followed by 35 cycles of 94◦C for 30 s, 60◦C for 30 s, and 72◦C for 30 s. After obtaining ct-value for each reaction, the fold change was calculated by using delta–delta ct method.

### Statistical Analysis

Plants were harvested 45 dpi. Values are means ± SE of 12 replicates from each treatment of rice. Different biochemical assays were performed at n = 3. Tukey's multiple HSD test at P = 0.05 was used to evaluate significance of the data and indicated with different letters. Principal component analysis (PCA) for metabolic differentiation among the treatments was analyzed using Statistica 7.0. All the comparisons were made as compared to control.

### RESULTS

### Biocontrol Efficacy of B. amyloliquefaciens under In Vitro Conditions

Biocontrol efficacy of B. amyloliquefaciens (SN13) against necrotrophic pathogen R. solani initially evaluated as dual culture method was further validated by co-incubating the R. solani and SN13 under in vitro conditions. Interaction studies showed that co-incubation results in ∼50% decrease in fungal dry mass irrespective of the combination strategies used (i) SN13 added 1 day before R. solani inoculation (ii) SN13 and R. solani inoculated simultaneously and (iii) SN13 added 1 day after R. solani inoculation (**Figure 1A**). SN13 showed profuse growth in all the three combinations (**Figure 1A**). Mycolytic enzymes assessed as a known mechanism of biocontrol showed synergistically enhanced protease (∼70– 120%) and chitinase (∼2–21%) activity when SN13 and R. solani were co-inoculated as compared to R. solani or SN13 alone. Among the three combinations pre-inoculation of SN13 in SN13 + R. solani combination was found to enhance protease and chitinase activity by 120 and 21%, respectively (**Figure 2**). Present study also reports reduction in glucanase and cellulase activity (2–96%) in R. solani + SN13 combinations. Higher protease and chitinase activity probably proves the antifungal activity of the strain NBRISN13 associated with their secretory systems for induced immune response.

### Biocontrol Efficacy of

### B. amyloliquefaciens against R. solani in Rice

The disease incidence observations were taken in terms of number, diameter, and length of the spots and their distance from the girth. The lesions observed were of necrotrophic type and were increasingly high in the R. solani treatment (**Figure 1B**). The number and size of spots consistently increased till 45 days post infection (dpi) in the R. solani treatment (9–11 in number and 2.8–3.8 mm in size). Reduction in disease incidence in terms of number and size in SN13 + R. solani treatment manifested its efficacy as a biocontrol agent under natural conditions (**Table 1**).

The physical growth parameters of the rice plants showed 15.25% reduction in shoot length on R. solani infection which was increased in SN13 + R. solani treatment by 23.72% as compared to control. The reduction in dry weight (36.29%) observed in R. solani alone treatment was overcome by 24.19% in SN13 + R. solani treatment. However, SN13 alone treatment was found to increase the dry mass by 40.64% as compared to control (**Figure 1B**).

### Histological Studies

A profuse colonization of stem surface with hyphae penetrating the stem tissues was observed in R. solani treatment. Whereas, SN13 primed plants showed very sparse R. solani colonization, restricted stem surface in SN13 + R. solani treatment (**Supplementary Figure S1**).

Programmed cell death (PCD), an important phenomenon results in aerenchyma formation which is more progressive under stress conditions. Large sized aerenchyma are observed in outer sheath which are decreasing in size toward inner layers. The air spaces in aerenchyma, encircled in **Figure 3** were observed to increase in presence of R. solani with an average area ranging from 18000 and 94000 µm<sup>2</sup> as compared to the control (6000–8500 µm<sup>2</sup> ). SN13 treatment reduced/delayed formation of aerenchyma in the middle sheath which ranged from 0 and 8000 µm<sup>2</sup> and was nearly absent in the innermost layer. The observations from stem sections stained with IKI showed loss of iodine reactivity with starch granules in SN13 treatment showing brown colored cells, whereas black colored starch granules were distinctly observed in R. solani and SN13 + R. solani treatments.

### Effect of B. amyloliquefaciens Inoculation on Physiological Parameters of Rice Plants

Physiological status of SN13, R. solani, and SN13 + R. solani treated plants was determined by measuring chlorophyll, sugar, and proline content. Significant changes in Chl A and total chlorophyll content was observed with time. Chlorophyll content decreased progressively in all the treatments, however, with different rates. The per cent reduction with time was maximum in R. solani treatment (50%) was followed by SN13 (36.8%) and SN13 + R. solani (7.6%). Higher sugar content in shoots of SN13 treated plants was observed over the period of 15–45 dpi. At the time of harvest the sugar content of SN13 + R. solani treated leaves was found to be highest by 320% and less in

individual treatments. The pathogen and PGPR alone were found to increase proline content by 70 and 114%, respectively. However, SN13 + R. solani did not show change in proline content with respect to control (**Table 1**). All the comparisons were made as compared to control.

### Modulation in Defense Enzyme Activities

Defense enzymes were assayed in rice leaves after 15 and 45 dpi for CAT, APX, and SOD activity. These activities were found to be higher in R. solani treatment by 39, 11, and 12%, respectively. The activities were more pronounced at 15 dpi as compared to 45 dpi (**Figure 4B**). SN13 imparted unstressed condition to the plant which maintained slightly higher or unaltered defense enzyme activities in presence and absence of pathogen treatments (SN13, SN13 + R. solani) as compared to control.

### Modulation in Cell Wall Enzyme Activities

Polygalacturonases, PME, and β-glucosidases (β-gluc) activities were measured in rice plants at 15 and 45 dpi. There was 11 and 179% increase in the activities of PME and β-glucosidase, except PG (46% decrease) in the presence of R. solani at 45 dpi (**Figure 4A**). However, at 15 dpi, PME and PG activities were increased by 1800 and 45%, respectively, and accompanied by 37% reduction of β-glucosidase activity in R. solani treatment. SN13 + R. solani treatment at 45 dpi showed reduction in PME

activity by 100% whereas PG and β-gluc activity were enhanced by 33 and 77%, respectively, but at 15 dpi all the three activities were higher in SN13 + R. solani treatment as compared to control (**Figure 4A**).

### Hormone Accumulation

HPLC analysis of plant hormones IAA, GA3, SA, and ABA showed differential accumulation in the treatments emphasizing the involvement of hormonal cross talk during tripartite interaction of PGPR–pathogen and rice (**Figure 5**). Reduced accumulation of SA in R. solani treatment (21.3%) was found to be enhanced by the presence of SN13 (SN13 + R. solani) by 36.8%. IAA content was found to be higher by 450% in SN13 + R. solani treatment which was otherwise unaltered in other treatments. R. solani treatment was found to have GA3 and ABA nearly twice the amount as in control. ABA concentration, approximately 10 times higher than that of control was observed in SN13 + R. solani treatment.

### Effect of B. amyloliquefaciens and R. solani Treatment on Metabolic Profiling of Rice

Gas chromatography–mass spectrometry analysis depicted the metabolic modulations in rice leaves treated in response to the interactions with SN13 and R. solani (**Figure 8**; **Supplementary Figures S3** and **S6** and Supplementary Table S2). The defense responsive phenolic acids, sugar, and their analogs and alkaloids were found to be differentially accumulated in R. solani, SN13 and combination treatments. R. solani infection led to higher accumulation of several defense responsive metabolites which were otherwise less accumulated in SN13 and SN13 + R. solani treatments. R. solani infection lead to enhanced or exclusive production of propionic acid, succinic acid, and quinoline which are known members of the phenylpropanoid pathway. ROS quenching metabolites pyrazole, imidazole, and chlorothiophene were triggered in R. solani treatment in response to defense. Besides, SAR inducers, quinoline, an alkaloid plant secondary metabolite was induced in response to R. solani infection (**Figure 8**; **Supplementary Figure S6**; Supplementary Table S2). Several sugars and sugar alcohols were observed in enhanced amounts during R. solani infection which included glycerol, arabitol and mannitol (**Figure 8**). Significant reductions as compared to control were observed with respect to concentration of turanose, mannopyranose, and sucrose during pathogen treatment, which were further enhanced in SN13 + R. solani treatment (Supplementary Table S2). Fructopyranose, glucopyranose, and myoinositol are the discriminating sugars accumulating higher in SN13 + R. solani treatment (**Figure 8**). Enhanced level of glucose/fructose ratio in SN13 + R. solani may be because of stimulated breakdown of sucrose into glucose and fructose. Some discriminating factors like quinazoline, a defense induced alkaloid were specifically accumulated in higher amounts in SN13 + R. solani treatment (**Figure 8**).

Principal component analysis analysis of 20 metabolites from rice shoots in four treatments (CON, SN13, R. solani,



Different letters on bar diagram show significant differences at p < 0.05 using Tukey's test. nd, not detected.

SN13 + R. solani) which distributed at 84.25 and 12.52% on factor 1 and 2, respectively, infer the distinct pattern among four treatments. Results clearly reveal that metabolic profiles of SN13 and SN13 + R. solani are closer to each other as compared to control and R. solani (**Supplementary Figure S6**). From these results it is interpreted that inoculation of SN13 altered the metabolic profiles in R. solani infected rice plants resulting in amelioration of the biotic stress.

Metabolites based PCA clearly show a wide distribution among some metabolites from the four treatments that are distributed at 61.41 and 25.46% on factor 1 and 2, respectively. Nine metabolites, i.e., propanoic acid (Pro), silanamine (Sil), 1H-imidazole (Imi), succinic acid (Suca), glycerol (Gly), arabitol (Ara), mannitol (Man), 2,5-di-butyl-3 chlorothiophene (Dbcn), quinoline (Qui) were grouped together on the plot, predicting that they remained unaltered in the four treatments. Other 11 metabolites, i.e., 1H-pyrazole (Pyr), fructopyranose (Fru), D-fructose (Dfr), D-glucose (Glu), mannopyranose (mnp), β-D-glucopyranose (Glp), myoinositol (Myo), sucrose (Suc), turanose (Tur), 1,2-benzene dicarboxylic acid (Ben), quinazoline (Qun) were distinctly separated on the plot (**Figure 8**).

### Effect of B. amyloliquefaciens and R. solani Treatment on Expression Profiling

Differentially expressed genes through microarray analysis were identified by dchip analysis. The comparisons were considered as control vs. SN13, control vs. R. solani and control vs. SN13 + R. solani. While comparing control vs. SN13, only 63 genes showed significant differential expression with p-value

FIGURE 3 | Stem cross section showing effect of B. amyloliquefaciens (SN13) on aerenchyma formation in rice, 15 dpi. "First' and 'second' indicate the sheath layer from core and the encircled area shows the extent of aerenchyma formation. CON, control; SN13, biocontrol agent; R. solani, pathogen; dpi, days post infection. Magnification 40×.

of >0.005. Of these, most of the genes were down-regulated and only 18 genes showed >onefold expression. Significantly expressed genes were: putative pullulanase precursor, α-amylase, 4-α-glucanotransferase, α-glucan phosphorylase isozyme, α-amylase, catalytic domain containing protein, universal stress protein domain containing protein, NADPH-dependent FMN reductase domain containing protein, inositol oxygenase, putative, Os1bglu4-β-glucosidase-like protein without signal sequence and hsp20/α-crystalline family protein (**Figure 6** and **Supplementary Figures S4** and **S5**). R. solani treatment resulted in a larger number of genes (92) being differentially expressed. As many as 19 genes were >twofold up-regulated and many of the up-regulated genes were clearly related to defense and pathogen related stress. The highly up-regulated genes were: oxidoreductase, aldo/keto reductase family protein, OsWAK14-OsWAK receptor-like protein kinase, ankyrin repeat domain containing protein, matrix attachment region binding protein, peptide transporter PTR2, retrotransposon protein, sulfate transporter, putative dehydration-responsive element-binding protein, ethylene-responsive transcription factor and putative serine/threonine protein kinase. In SN13 + R. solani treatment, a large number of genes (90) were differentially regulated significantly, however, none of the genes were >1.5-fold up-regulated and only three genes were >1.5-fold down-regulated (**Supplementary Figure S5**). There were only 10 genes that up-regulated > onefold. Interestingly the stress responsive genes were significantly downregulated. The major genes up-regulated were: phospholipase D, auxin response factor, galactosyltransferases, etc. The genes were categorized based on functional pathways involving primary metabolism, stress responses and hormone biosynthesis. Many genes in different categories show very contrasting expression profiles. The arrows in **Figure 6** show those genes in SN13 + R. solani which has been restored to control levels as compared to R. solani alone treatments.

Validation of microarray results through RT-PCR analysis of randomly picked genes (marked as blue arrow in **Figure 6**) was performed using gene specific primers. The Analysis of genes related to cell wall disintegration, limit dextrinase (Os04g08270), 4-α-glucan transferase (Os07g43390), isoamylases (Os08g40930), and inositol oxygenase (Os06g36560) showed up-regulation in R. solani treatment as compared to control. Up-regulation of isoamylases and 4-α-glucan transferase in SN13 + R. solani treatment was also observed. In a further support to this Os03g61780, coding for glucan endo-1,3-β-glucosidase showed

marginal overexpression, contrary to the microarray result (−1.2 fold down-regulation) in fungus treatment as compared to control. Universal stress protein domain (Os02g53320) was another gene up-regulated in SN13 as compared to control (**Figure 7**). The pathogenesis related Bet V gene (Os12g36830) was down-regulated in R. solani and SN13 treatment whereas it was further down-regulated in SN13+R. solani treatment. Downregulation of Bet V gene was very well correlated in both the experiments.

Further, overexpression of serine threonin protein kinase (Os 01g66860) of MAPKinases family by fivefold and phospholipase D (Os06g40180) involved in signaling of plants' multiple defense response showed up-regulation by threefold in SN13 as compared to control and twofold as compared to R. solani. Gene responsible for ROS balancing, Os08g35210 coding for ferric reductases of plasma membrane NADPH oxidase family (Nox) was downregulated (0.5-fold) in SN13 + R. solani treatment (**Figure 7**). However, genes coding for peroxidase precursor (Os01g73200)

and glutathione S-transferase (Os10g38590) were found to be upregulated in R. solani alone treatment as compared to control and further down-regulation of these genes in SN13 + R. solani treatment was observed, which does not correlate with the microarray results.

Up-regulation of gibberellin 20-oxidase (Os01g66100) and a DREB homolog (Os02g45450), genes in SN13 (∼2.5-fold) and R. solani treatment (∼1.5-fold) has been found up-regulated in present study which shows further similar expression as compared to control in SN13 + R. solani treatment. The gene (Os04g26910) encoding an oxidoreductase showed concomitant increase in R. solani, SN13 and SN13 + R. solani treatment predicting increased auxin as a defense response.

Matrix attachment binding protein (Os12g20410) existing in a co-repressor/co-activator complex was up-regulated in fungus (3.2-fold) and bacterial treatment (2.1-fold) but was downregulated in the SN13 + R. solani treatment. Role of stomatal closure was also observed by the down-regulation of subtilisin homolog (Os01g58280) in R. solani and SN13 and similar level of expression of this gene in SN13 + R. solani was observed as a stress reliever (**Figure 7**). Differential accumulation of metabolite observed was also correlated with the overexpression of terpene synthase (Os02g02930) in SN13 (threefold) well correlated with the accumulation of quinazoline due to beneficial interaction.

### DISCUSSION

The severity of sheath blight disease in rice depends on many external and internal factors such as plant vigor, presence of beneficial microbes and nutritional status of the soil (Herman et al., 2008; Srivastava et al., 2012; Carvalhais et al., 2013). Present study elucidates the effect of a bioinoculant (SN13), a pathogen (R. solani) and their interaction (SN13 + R. solani) on the network of events active at a relatively later stage of infection in rice plants, unlike in other studies where immediate events are discussed. The chain of events involved participation of the plant's machinery related to enhanced elicitation, ROS scavenging, hormonal cross talk, sugar signaling, and secondary metabolism. Correlations among different events have been discussed in this study.

Bacillus amyloliquefaciens (SN13) antagonism of R. solani under in vitro conditions mediated by mycolytic enzymes was increased in SN13 + R. solani which was probably triggered by presence of R. solani as reported earlier (Smitha and Singh, 2014). Effect of this interaction was evident under greenhouse conditions in terms of reduced fungal colonization or biocontrol (Frey-Klett et al., 2011) and presence of a constant source of elicitors.

SN13 imparted health and vigor to the plants with or without R. solani and enhanced carbon assimilation can be related well with the higher dry mass, chlorophyll content, and starch accumulation (**Table 1**; **Figure 3**). Induction of PG and PME after 15 days of infection in R. solani treatment probably resulted in cell wall loosening through degradation and desertification of pectins (Raiola et al., 2011; Bellincampi et al., 2014). Down-regulation of PME and up-regulation of PG and β-gluc activity (oligosaccharide release) in SN13 primed plants probably enhanced deliverance of elicitor responsible for the induced systemic response. However, this activity declines with time as

supported by the down-regulated genes in microarray data and also attenuated enzyme activities after 45 dpi (**Figure 4A**).

A consistency in CWDE in R. solani infected plants, is congruous with the fact that pathogens first invade the cell wall through differential expression of limit dextrinase (Os04g08270), 4-α-glucan transferase (Os07g43390), isoamylases (Os08g40930), and inositol oxygenase (Os06g36560) which have role in cell wall expansion, starch and sucrose metabolism and oxidation of myoinositol to glucuronic acid (Trafford et al., 2013) was also correlated well with myoinositol content. Transcriptional and translational correlation between myoinositol concentration and inositol oxygenase (Os06g36560) exists in SN13 + R. solani and SN13 treatments probably maintains a high glucuronic acid concentration for elicitation of defense (**Figures 5, 6** and **8**; Supplementary Table S2; Vera et al., 2011). Higher myoinositol might result in hypersensitive response by inducing local cell death and suppressing effector triggered immunity through R gene (Os11g46070) and disease resistance protein interaction (Os06g06850; **Figure 6**) (Bruggeman et al., 2015). Furthermore proteases from SN13 may be inducing nitrogen limitation for fungal partner and induce immune response as reported for proteases from pathogenic sources during sheath blight disease (Suarez et al., 2005; Zheng et al., 2013). The observations clearly demonstrate role of SN13 in employing plants' system to elicit defense that not only maintains the cell wall integrity but also resists pathogen invasion by degrading fungal cell wall polysaccharides, as evident earlier (Nakano et al., 2013).

A stable and balanced intracellular redox is maintained by the plant's defense mechanism by differentially regulating various enzymes and/or metabolites (Hare and Cress, 1997). Modulation of proline levels by SN13 and R. solani in different treatments affirmed as an indicator of ROS status; lowered proline and ROS in SN13 and higher in SN13 + R. solani treatments shows competence of the bacteria to maintain ROS balance as per requirement of the plant (**Table 1**) (Hare and Cress, 1997). An induced level of ROS quenching enzymes such as CAT, APX and

SOD in R. solani treatment an indicator of high intracellular redox is in concordance with the up-regulation of genes coding for peroxidase precursor (Os01g73200) and glutathione S-transferase (Os10g38590). Controlled intracellular redox in SN13 + R. solani is in accordance to the down-regulation of defense responsive genes and enzymes. Down-regulation of NADPH oxidase family such as ferric reductase (Os08g35210), a key producer of ROS in SN13 + R. solani (0.5-fold) emphasizes relatively unstressed conditions (**Figure 6**) as stated earlier (Nakano et al., 2013). These cell wall interactions and elicitation lead to kinase mediated signaling, as evident by overexpression of serine-threonin protein kinase (Os01g66860) of MAPKinases (fivefold) and phospholipase D (by threefold in SN13 as compared to control) that functions in response to ROS accumulation. The results are in accordance to prior report where enhanced level of these genes was associated with pathogen resistance (Sun et al., 2014). Differential accumulation of ROS quenchers like pyrazole, imidazole, and chlorothiophene in SN13 + R. solani as compared to R. solani treatment also insist SN13 mediated ROS balancing. The sustained plant

defense may thus be attributed to the other unconventional mechanisms.

Levels of metabolizable sugars (glucose/fructose) in SN13 + R. solani may be responsible for inducing auxin, however, their enhanced production in R. solani and SN13 alone may also be associated with increased ROS production and involvement of stress mediated signaling as also evident from microarray and RT-PCR analysis of auxin response factor 9 (Os4g36054; **Figure 7**) (Hamzehzarghani et al., 2005; Li et al., 2011; Sairanen et al., 2012; Nafisi et al., 2015; Naseem et al., 2015). Decreased SA accumulation in R. solani treatment corresponds well with prior reports of rice–necrotroph interaction (Bari and Jones, 2009). Increased SA level in SN13 treatments reinforces the role of SN13 in priming the defense response active both in presence and absence of pathogen. Higher level of ABA and GA in SN13 + R. solani treatment propose ABA and GA mediated defense regulation in plants as evident in microarray data and earlier reports against Pythium irregulare and Magnaporthe grisea (**Figures 5** and **6**) (Ton and Mauch-Mani, 2004; Schmidt et al., 2008; Parker et al., 2009; Zheng et al., 2013; Nafisi et al., 2015). The delayed aerenchyma formation (probably due to low ABA and GA) and less starch content in parenchyma cells (anatomical results) emphasize SN13 mediated delayed apoptosis of parenchyma cells even under high oxidative state for improved stress tolerance through the overexpression of universal stress protein domain (Os02g53320, coding for ENOD18) having role in adjusting hypoxic conditions through differential hormone signaling and accumulation of ROS quenchers like pyrazole, imidazole, and chlorothiophene (Sauter et al., 2002).This may be argued as a probable mechanism of biocontrol by PGPR in plants and opens a new avenue of further verification.

The interplay of hormones and metabolites employed by SN13 primed plants when challenged with a pathogen lead to a stark difference in the accumulation of sugars, their alcohols and pyranoses as compared to earlier report of defense elicitation (Zhang et al., 2015). These differences were observed as widely separated metabolites on the PCA plot which may be used as predictive marker metabolites to understand the effects of PGPR mediated amelioration of biotic stresses (**Figure 8**; **Supplementary Figure S6**). The high levels of glycerol, mannitol, and arabitol cause the quenching of ROS generated by plant, thereby acting in plant defense and also act as nutrient for R. solani as supported by Zhang et al. (2015) (**Figure 8**; **Supplementary Figure S3** and Supplementary Table S2). Enhanced amounts of pyranoses in SN13 + R. solani provided a pool of sugars which are easily inter-convertible to usable sugar when required but not available for ROS generation (**Figure 8**). Pyranoses are also known to be the cyclic sugars that will polymerize for the formation of storage molecules like starch. Therefore, it may be hypothesized that

one of the mode of action of SN13 primed plants to combat an infection is by depriving the pathogens for readily usable forms of sugars (**Figure 8**). A reduction in sucrose/hexose (glucose + fructose) ratio in R. solani (0.56) and SN13 (0.75) and further enhancement in SN13 + R. solani (0.79) treatment is in accordance to the report that cell wall invertases play crucial role for determining the pathways, differently triggered by pathogen and PGPR through fine regulation of sucrose/hexose ratio in an ABA dependent manner (**Figures 5** and **8**) (Bolouri-Moghaddam et al., 2010; Li et al., 2011; Tauzin and Giardina, 2014). Up-regulation of α-amylases (Os08g40930) and 4-α-glucanotransferase (Os07g43390) in SN13 + R. solani treatment correlates with decreasing sucrose (GC MS results) and starch content (anatomical results) showing importance of SN13 in modulating starch and sucrose metabolism (**Figures 3** and **8**; Supplementary Table S2).

Differences in yet another sugar, turanose, a nonmetabolizable sucrose analog is of interest since its active involvement during plant pathogen interaction has not been reported earlier to the best of our knowledge. Reduced accumulation of turanose in R. solani (∼90%) intends pathogen mediated inhibition of defense signaling pathway by catabolizing turanose to produce osmolytes, glutamate and the dipeptide N-acetylglutaminyl glutamine amide (Mchunu et al., 2013) (**Figure 8**; **Supplementary Figure S3** and Supplementary Table S2). The negative correlation observed between turanose and proline in R. solani treatment may be due to its catabolism to glutamate which is precursor for proline. The observation opens a new area of investigation in determining role of turanose in plant–microbe interactions.

Exclusive production of propionic and succinic acid of TCA cycle during R. solani interaction may indicate a compensatory response to the requirement for a high metabolic flux during plant pathogen interaction to cope up the increasing energy consumption (Alteri et al., 2015). Accumulation of tryptophan dependent quinoline, an alkaloid of phenylpropanoid pathway emphasizes the switch between primary to secondary metabolism during pathogen interaction (Iriti and Faoro, 2009). Overexpression of terpene synthase (Os02g02930; involved in terpene biosynthesis) is correlated well with various defense mechanisms and improved biological control (Tholl et al., 2011). The threefold up-regulation in SN13 is well correlated with the accumulation of antibacterial like benzene dicarboxylic acid and quinazoline resulting in sustained defense response (**Supplementary Figure S3**).

Up-regulation of genes related to biosynthetic processes, catalytic activity and transferase activity during R. solani infection are indication of plant's agitated condition. Differentially expressed genes in SN13 are characteristic of plant–bacterial interaction and down-regulation of defense related genes emphasize plant's rather unstressed environment. The GO terms of SN13 + R. solani are comparable to plants treated only with SN13which is also evident in the PCA plot with metabolites (**Supplementary Figure S6**).

From the above discussion it is evident that SN13 primed and unprimed plants responded differently upon R. solani challenge. The bioinoculant SN13 induced a network of events which resulted in long term priming of the plants against the infection. Direct confrontation of the pathogen enhanced elicitation of immune response was through the maintenance of differential physiological and metabolic status. ROS modulation, either through accumulation of quenchers (mannitol, arabitol, and phospholipases) or through generation of ROS inhibitors was observed. SN13 primed plants exhibited rare sugars (pyranoses and turanose), which eventually compromised the fungal growth and elicited defense response using MAPK signaling and fortification of metabolites like terpenes and quinazoline. Auxin responsive hormonal cross talk in SN13 induced resistance enhancing plant immunity in ISR mediated SAR in SA, ABA, and benzene dicarboxylic acid dependent manner where expression of JA (Os03g55800) and ethylene responsive transcription factor (Os03g09170) remained unaltered (**Figure 6**).

### CONCLUSION

The present study shows a sustained tolerance toward R. solani infection by SN13 primed plants, that may be attributed to unconventional mechanisms mediated through (a) general maintenance of plant health and vigor, (b) involvement of bacterial mycolytic enzymes, (c) cell wall modification and sustained maintenance of elicitors, (d) a delicate balance of ROS and ROS scavengers that helped in both resisting the pathogen infection and protecting its own cells and, (e) production of metabolites especially non-metabolizable sugars and secondary metabolites.

### AUTHOR CONTRIBUTIONS

SuS, VB, SoS, PS, did all the experiments. PKT and MA, assisted in analysis of Microarray experiment. SuS and PS, wrote the MS. PSC, PKT, and CSN reviewed the manuscript.

### ACKNOWLEDGMENTS

The study was supported by project "RootSF–BSC 0204" funded by Council of Scientific and Industrial Research (CSIR), New Delhi, India. Authors are thankful to Dr. S. K. Raj, Rtd. Chief Scientist, CSIR–NBRI, Lucknow for critical editing of the MS. Thanks are due to Dr. Anil Sharma and Dr. Abhishek Niranjan, CIF facility CSIR–NBRI, for help with GC–MS and HPLC analysis.

## SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016.00587

FIGURE S1 | Light micrograph showing colonization of rice leaves by R. solani during infection in absence (C,E,G) and presence of a biocontrol/PGPR SN13 (D,F,H) on stem surface in different treatments; control (A, CON), Bacillus amyloliquefaciens NBRISN13 (B, SN13), pathogen treatment (R. solani) and combination of both (SN13 + R. solani) after 15 dpi. Magnification 10×; A–D: 4×; E–F: 10×; G–H: 40×.

FIGURE S2 | HPLC chromatogram of standards, 1 = gibberellic acid, 2 = indole acetic acid, 3 = abscisic acid, 4 = salicylic acid and of Control; R. solani; SN13 and SN13 + R. solani treatments of rice leaves after 15 days of infection.

FIGURE S3 | Comparative ion electropherograms of metabolites in methanolic leaf extracts of rice through GC–MS analysis in absence (CON) and presence of sheath blight fungi (R. solani) and PGPR B. amyloliquefaciens (SN13) and combination of both (SN13 + R. solani).

FIGURE S4 | Expression profiling and hierarchical clustering of differentially expressed genes obtained through microarray data when comparisons were made between (A) Control\_Fungus (CON\_R. solani),

### REFERENCES


(B) Control\_Bacteria (CON\_SN13) and, (C) Control\_Bacteria + Fungus (CON\_SN + R. solani).

FIGURE S5 | Functional classification of the up-regulated (a) and down-regulated (b) Oryza sativa genes in terms of their Gene ontology (GO terms), relative to their representation in the genome of rice plant, involved in a plant, PGPR, and pathogen interaction.

FIGURE S6 | Principal component analysis (PCA) of rice treatments differentiating control, pathogen infected (R. solani), Bacillus amyloliquefaciens treated (SN13), and pathogen + B. amyloliquefaciens (SN13 + R. solani) treatments based on GC–MS metabolic profiles of leaves.

quantitative resistance in wheat cultivars against fusarium head blight. Physiol. Mol. Plant Pathol. 66, 119–133. doi: 10.1016/j.pmpp.2005.05.005


community in rice during salt stress. Plant Physiol. Biochem. 66, 1–9. doi: 10.1016/j.plaphy.2013.01.020


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Srivastava, Bist, Srivastava, Singh, Trivedi, Asif, Chauhan and Nautiyal. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Interactions between cranberries and fungi: the proposed function of organic acids in virulence suppression of fruit rot fungi

Mariusz Tadych<sup>1</sup> \*, Nicholi Vorsa<sup>2</sup> , Yifei Wang<sup>1</sup> , Marshall S. Bergen<sup>1</sup> , Jennifer Johnson-Cicalese<sup>2</sup> , James J. Polashock <sup>3</sup> and James F. White Jr. <sup>1</sup>

<sup>1</sup> Department of Plant Biology and Pathology, Rutgers University, New Brunswick, NJ, USA, <sup>2</sup> Philip E. Marucci Center for Blueberry and Cranberry Research and Extension, Rutgers University, Chatsworth, NJ, USA, <sup>3</sup> Genetic Improvement of Fruits and Vegetables Laboratory, United States Department of Agriculture-Agriculture Research Service, Philip E. Marucci Center for Blueberry and Cranberry Research and Extension, Chatsworth, NJ, USA

#### Edited by:

Gero Benckiser, Justus-Liebig-Universität Gießen, Germany (Retired)

#### Reviewed by:

Raffaella Balestrini, Consiglio Nazionale delle Ricerche, Italy Zonghua Wang, Fujian Agriculture and Forestry University, China

#### \*Correspondence:

Mariusz Tadych, Department of Plant Biology and Pathology, Rutgers University, 59 Dudley Road, New Brunswick, NJ 08901-8520, USA tadych@aesop.rutgers.edu

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 24 March 2015 Accepted: 29 July 2015 Published: 14 August 2015

#### Citation:

Tadych M, Vorsa N, Wang Y, Bergen MS, Johnson-Cicalese J, Polashock JJ and White JF Jr. (2015) Interactions between cranberries and fungi: the proposed function of organic acids in virulence suppression of fruit rot fungi. Front. Microbiol. 6:835. doi: 10.3389/fmicb.2015.00835 Cranberry fruit are a rich source of bioactive compounds that may function as constitutive or inducible barriers against rot-inducing fungi. The content and composition of these compounds change as the season progresses. Several necrotrophic fungi cause cranberry fruit rot disease complex. These fungi remain mostly asymptomatic until the fruit begins to mature in late August. Temporal fluctuations and quantitative differences in selected organic acid profiles between fruit of six cranberry genotypes during the growing season were observed. The concentration of benzoic acid in fruit increased while quinic acid decreased throughout fruit development. In general, more rot-resistant genotypes (RR) showed higher levels of benzoic acid early in fruit development and more gradual decline in quinic acid levels than that observed in the more rot-susceptible genotypes. We evaluated antifungal activities of selected cranberry constituents and found that most bioactive compounds either had no effects or stimulated growth or reactive oxygen species (ROS) secretion of four tested cranberry fruit rot fungi, while benzoic acid and quinic acid reduced growth and suppressed secretion of ROS by these fungi. We propose that variation in the levels of ROS suppressive compounds, such as benzoic and quinic acids, may influence virulence by the fruit rot fungi. Selection for crops that maintain high levels of virulence suppressive compounds could yield new disease resistant varieties. This could represent a new strategy for control of disease caused by necrotrophic pathogens that exhibit a latent or endophytic phase.

Keywords: benzoic acid, bioactivity, cranberry fruit rot disease, pathogenicity, quinic acid, reactive oxygen species, resistance, Vaccinium

### Introduction

Many pathogens possess a latent phase where they grow within tissues of hosts without causing harm or resulting in expression of disease symptoms (Luo and Michailides, 2001; Sauer et al., 2002; Vega et al., 2010; O'Connell et al., 2012; Tadych et al., 2012; Delaye et al., 2013). It is only after this period of latent development that disease expression may become evident. Our previous study (Tadych et al., 2012) suggested that the majority of cranberry fruit rot necrotrophic fungi possess a latent phase in which they grow inside cranberry fruit asymptomatically. Fruit rot fungi may coexist in developing ovaries as endophytes or they may exist alone in ovaries. After this phase of symptomless development, usually shortly before or at fruit maturation, disease expression in the form of rot may become evident.

Fungal diseases, particularly the cranberry fruit rot disease complex, have been serious problems limiting fruit production from the beginning of commercial cultivation of American cranberry (Vaccinium macrocarpon Aiton) (Halsted, 1889; Stevens, 1924; Shear et al., 1931; Oudemans et al., 1998; Tadych et al., 2012). Among the most common fungi causing cranberry fruit rot disease are Coleophoma empetri (Rostr.) Petr., Colletotrichum acutatum J. H. Simmonds, Colletotrichum gloeosporioides (Penz.) Penz. & Sacc., Fusicoccum putrefaciens Shear, Phomopsis vaccinii Shear, N. E. Stevens & H. F. Bain, Phyllosticta vaccinii Earle and Physalospora vaccinii (Shear) Arx & E. Müll (Oudemans et al., 1998; Polashock et al., 2009; Tadych et al., 2012).

Defensive mechanisms against pathogens in many animals and plants involve the direct action of reactive oxygen species (ROS), such as superoxide (O<sup>−</sup> 2 ), hydroxyl radical (OH• ), and hydrogen peroxide (H2O2) (Foyer and Harbinson, 1994; Wu et al., 1997; Missall et al., 2004; Silar, 2005). It has been shown that ROS are generated as anti-pathogen agents and as warning signals to adjacent host cells, triggering other host defensive reactions (Lamb and Dixon, 1997; Wojtaszek, 1997). Pathogens often trigger an increase in ROS called "oxidative burst," which results in the accumulation of ROS in tissues of the plant proximal to the pathogen (Apel and Hirt, 2004). The accumulation of ROS may cause damage to cells by peroxidizing lipids and disrupting structural proteins, enzymes, and nucleic acids, and may subsequently lead to cell death (Apel and Hirt, 2004).

Previous research has associated ROS secretion by fungal necrotrophs with induction of cell death and necrosis in host tissues (Álvarez-Loayza et al., 2011; Heller and Tudzynski, 2011). The linkage between fungal ROS secretion and initiation of the hypersensitive response in host plant tissues provides a target for identification of natural plant constituents that will prolong the non-destructive latent phase of the cranberry rot fungi.

Many bioactive compounds can function as constitutive or inducible barriers against microbial pathogens, and bioactive compound composition can change in response to microbial attack (Dixon and Paiva, 1995; Grayer and Kokubun, 2001; Miranda et al., 2007; Carlsen et al., 2008; Koskimäki et al., 2009; White and Torres, 2010; Oszmiañski and Wojdył, 2014). Cranberry fruit are known to be rich sources of nutrients and bioactive compounds, including phenolics, flavonoids, sugars, organic acid, etc., (Fellers and Esselen, 1955; Schmid, 1977; Coppola et al., 1978; Mäkinen and Söderling, 1980; Hong and Wrolstad, 1986; Zuo et al., 2002; Zheng and Wang, 2003; Cunningham et al., 2004; Shahidi and Naczk, 2004; Vvedenskaya et al., 2004; Singh et al., 2009; Neto and Vinson, 2011), any of which could have activity against rot-inducing fungi (Marwan and Nagel, 1986a,b; Cushnie and Lamb, 2005). Previous research suggests that fungi that cause cranberry fruit rot disease colonize surface layers of cranberry ovaries early in flower development (Zuckerman, 1958; Tadych et al., 2012) and induce disease in mature fruit tissues possibly by secretion of ROS into fruit, resulting in a cascade of events in fruit tissues that leads to cell death and fruit rot. According to this model, suppression of growth and ROS secretion by fungi will result in suppression of rot disease. We hypothesize that fruit rot resistant selections of cranberry are resistant to rot due to organic acid constituents that enable them to suppress growth and ROS production by cranberry fruit rot fungi. We further hypothesize that levels of organic acids may change as fruit mature, leading to a release of ROS suppression and increase in fungal growth and disease incidence in fruit. Objectives for this study were: (1) to identify naturally occurring chemicals in cranberry fruit that suppress growth of cranberry fruit rot fungi, (2) to determine whether secretion of ROS by these fruit pathogens could be stimulated or inhibited by these cranberry constituents, (3) to investigate the organic acid profiles in the fruit of rot-resistant and rotsusceptible cranberry genotypes at intervals throughout fruit maturation that may coincide with fruit rot occurrence.

### Materials and Methods

### Reference Compounds and Other Chemicals

Agarose (A6013), L-Alanine (A7627), Benzoic acid (242381), 3,3′ -diaminobenzidine (D5905), Folic acid (F7876), Formic acid (F0507), D-(–)-Fructose (F0127), D-(+)-Glucose (BDH0230), Glycine (G7126), Horseradish peroxidase (P6782), DL-Malic acid (M0875), D-Mannitol (M4125), N-Z-Soy<sup>r</sup> Peptone (P1265), Pectin from apple (P8471), Phosphoric acid, D-(–)-Quinic acid (138622), Starch from rice (S7260), and Sucrose (50389) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Acetonitrile (AX0145) was obtained from EMD Millipore (Billercia, MA), Citric acid anhydrous (A940) from Fisher Scientific (Fair Lawn, NJ), Sorbitol (V045-07) from J. T. Baker – Mallinckrodt Baker, Inc. (Phillipsburg, NJ) and Proflo Premium Quality Cottonseed-derived Protein Nutrient (069061) from Trades Protein, Southern Cotton Oil Company (Memphis, TN). All solvents were of HPLC grade and water was of Milli-Q quality (Millipore Corp., Bedford, MA).

### Fungal and Plant Material Used

Cranberry fruit rot fungi used in this study were collected from infected cranberry ovaries in 2009 at the Philip E. Marucci Center for Blueberry and Cranberry Research and Extension of Rutgers University located in Chatsworth, New Jersey (39◦ 42′ 50.75"N, 74◦ 30′ 33.07"W; altitude 12 m) as described by Tadych et al. (2012) and stored in −80◦C until this study. After removing from freezer the fungi were regrown first in potato dextrose broth (PDB) and then re-cultured on potato dextrose agar (PDA) at room temperature (22◦C) for 10 days.

Fruit of the American cranberry used in the present study were represented by four rot-resistant, i.e., 'US88-1', 'US88-30', 'US88- 79', 'US89-3', and two rot-susceptible, i.e., 'Mullica Queen' (MQ) and 'Stevens' (ST), genotypes growing at the Philip E. Marucci Center for Blueberry and Cranberry Research and Extension at Rutgers University (Vorsa and Johnson-Cicalese, 2012). The plants were planted in randomized block design with 54 rows and 12 columns with one genotype per plot established in 2009. The experimental plots from which the plant material was collected were not fungicide treated and were cultivated in an identical way during the 2012 and 2013 growing season.

On July 11, 2012 green cranberry fruit were collected from US88-79 and ST genotypes, at the phenological stage of "small fruit" (Brown and McNeil, 2006) and transferred to the laboratory. The fruit were ground in liquid nitrogen and added to the basal medium (0.5% agarose + 0.5% glucose) to reach a final concentration of 10% fruit tissue in the medium (**Table 1**) and tested for the impact on growth and ROS secretion of cranberry fruit rot fungi. In addition, in order to test the influence of sterilization on antimicrobial activity of the fruit, in one case raw fruit tissue was added to an already autoclaved medium [raw rot-resistant genotype green berries – raw resistant (RR); raw rotsusceptible genotype green berries – raw susceptible (RS)], and in another case the fruit tissue was added to a medium before medium autoclaving [autoclaved rot-resistant genotype green berries – autoclaved resistant (AR); autoclaved rot-susceptible genotype green berries – autoclaved susceptible (AS)].

TABLE 1 | Cranberry constituents screened for inhibitory effects on fungal growth and hydrogen peroxide production by selected cranberry fruit rot fungi.


\*Concentration of a particular compound found in cranberry fruit (as previously reported: Fellers and Esselen, 1955; Schmid, 1977; Coppola et al., 1978; Mäkinen and Söderling, 1980; Hong and Wrolstad, 1986; Zuo et al., 2002; Zheng and Wang, 2003; Cunningham et al., 2004; Shahidi and Naczk, 2004; Vvedenskaya et al., 2004; Singh et al., 2009; Neto and Vinson, 2011); N/A, not applicable.

\*\*Concentration of a particular compound tested (added to the basal medium—0.5% agarose and 0.5% glucose).

To study the within-season variation of organic acids (OA) in the fruit, material for each plant-genotype was collected from 3 plots in late July (Jul 26), two times in August (Aug 6 and Aug 23), and two times in September (Sep 5 and Sep 16). After harvesting, the berries were transferred to the laboratory, immediately frozen at −80◦C and stored until analyzed (within 6 months of harvest). From each sample harvested from an individual plot approximately 2 g (±1 g depending on fruit availability) of berries were randomly selected for chemical analysis.

### Effect of the Tested Compounds on the Radial Growth of Selected Fungal Pathogens

Sixteen compounds previously reported (**Table 1**; Fellers and Esselen, 1955; Schmid, 1977; Coppola et al., 1978; Mäkinen and Söderling, 1980; Hong and Wrolstad, 1986; Zuo et al., 2002; Zheng and Wang, 2003; Cunningham et al., 2004; Shahidi and Naczk, 2004; Vvedenskaya et al., 2004; Singh et al., 2009; Neto and Vinson, 2011) as naturally occurring in cranberry fruit and four combinations of the green cranberry fruit tissues representing rot-resistant and rot-susceptible genotypes were incorporated into basal medium to study their effects on fungal radial colony growth. Plates (85 mm diam.) containing 20 ml of a media with test compounds (**Table 1**) were prepared by incorporating filtersterilized compounds into a basal medium after autoclaving. A 10-mm plug taken from edge of an actively-growing 10 day-old culture of four selected isolates of cranberry fruit rot fungi, i.e., Coleophoma empetri, Colletotrichum gloeosporioides, Phyllosticta vaccinii and P. vaccinii growing on PDA, was used to inoculate the plates containing each of the media. The basal medium was used as a control. The plates were incubated at 22◦C in darkness. The radial growth of each fungus on plates was measured 10 days after inoculation along perpendicular axes. The study was performed with three replicates for each fungus/medium combination.

### Detection of H2O<sup>2</sup> Secreted into Agar by Fungi

To visualize secretion and accumulation of ROS as H2O<sup>2</sup> into agar, the plates with the 10-day-old fungal colonies were stained by flooding plates with 6 ml of 100 mM potassium phosphate buffer, pH 6.9, 2.5 mM 3,3′ -diaminobenzidine tetrachloride (DAB) and 5 purpurogallin units ml−<sup>1</sup> of horseradish peroxidase (Type VI-A), swirled to cover the entire surface, and incubated at room temperature for 10 h (Pick and Keisari, 1980; Munkres, 1990). To stop the color development the plates were rinsed in sterile dH2O. Visible ROS reaction zones were measured and each plate was photographed.

### Extraction and Quantification of Total Benzoic Acid, Citric Acid, Malic Acid, and Quinic Acid in Developing Cranberry Fruit Extraction of Organic Acids

Fruit thawed for 1 h at 22◦C were homogenized using Waring™ Laboratory Blenders (Model 31BL91 with MC-1 Mini Container, Waring Commercial, Torrington, CT) prior to analyses. Homogenized fruit samples were diluted with dH2O in ratio approximately 1:10 and mixed by shaking. The mixture was then sonicated (Branson 3510, Branson Ultrasonics Corporation, Danbury, CT) for 10 min, transferred and left on a stirrer at 60 rpm for 10 min in a water bath (Precision 2870, Thermo Electron Corporation, Waltham, MA) with temperature of 90◦C. The suspension was filtered with Whatman filter paper and 1 ml of the supernatant was centrifuged (AccuSpin Micro 17R, Fisher Scientific, Osterode, Germany) at 13,300 g for 10 min at 4◦C. From each centrifuged sample 300µl of supernatant was transferred into an HPLC vial and analyzed for organic acids.

### High-performance Liquid Chromatography (HPLC) Analysis

Aqueous extracts obtained from cranberry fruit were analyzed for identification and quantification of organic acids (OA) using a Dionex <sup>R</sup> HPLC system (Dionex, Sunnyvale, CA) equipped with AS50 Autosampler, AS50 Thermal Compartment, PDA-100 Photodiode Array Detector and GP40 Gradient Pump. For all standards and extracts, a Waters Atlantis dC18 column (5.0µm particle size; 100A; 250 mm length × 4.6 mm ID; Waters Co., Milford, MA) with a security guard cartridge (Phenomenex, Inc., Torrance, CA) was used; the column temperature was 25◦C. A binary solvent system with solvent A: 0.5% phosphoric acid in water and solvent B: 0.5% phosphoric acid in acetonitrile was used with isocratic elution of 0% B from 0 to 11 min; linear gradient of 0 to 20% B from 11 to 13 min; 20 to 60% B from 13 to 18 min; 60 to 80% B from 18 to 20 min; isocratic elution of 80% B from 20 to 25 min; linear gradient of 80 to 20% B from 25 to 28 min; 20 to 0% B from 28 to 30 min and isocratic elution of 0% B from 30 to 40 min at a flow rate of 0.6 ml min−<sup>1</sup> with 20µl sample injection volume.

Chromatograph peaks were identified taking into account the retention time and the UV-Vis absorption spectra of the peaks with those of corresponding standards. Photodiode Array Detector was monitored at three wavelengths, 210 nm for citric (CA) and malic (MA) acids, 214 nm for quinic acid (QA), and 230 nm for benzoic acid (BA). Data acquisition and processing were performed using Dionex Chromatography Software—Chromeleon Client version 6.80 (Dionex, Sunnyvale, CA). Quantification of all the organic acids was based on a standard curve prepared with BA, CA, MA, and QA. The contents of organic acids were expressed as milligrams of organic acid per gram of fresh weight (fw). The samples were analyzed in duplicates.

### Calibration Curves

Concentration of organic acids in cranberry ovary extractions was determined based on calibration curves. The calibration curves were constructed using the standard solutions. Known amounts of BA, CA, MA, and QA adding to double distilled water resulted in stock solutions and their serial dilutions were used to prepare standard solutions. Except for BA where 16 standard concentrations were prepared, eight standard concentrations were prepared for the other three acids and analyzed by HPLC in triplicate. Calibration curves were generated by plotting peak area (mAU) against acid concentration (mg ml−<sup>1</sup> ) with linear regression analysis.

### Statistical Analyses

Results were expressed as mean ± standard error of the mean (SEM). Significant differences (α = 0.01) between means were estimated by use of analysis of variance (ANOVA), General Linear Model (GLM) followed by the Ryan-Einot-Gabriel-Welsch Q multiple range test. The effects of plant-genotype and within-season variation were investigated with different dates analyzed independently. The data analysis was generated using SAS/STAT software, Version 9.3 of the SAS System for Windows (SAS Institute Inc., Cary, NC, USA, 2011).

### Results

### Effect of the Cranberry Compounds on the Radial Growth of Selected Cranberry Fungi

The results from in vitro screening of compounds previously identified in cranberry fruit (**Table 1**) on growth of selected cranberry fruit rot fungi, Coleophoma empetri, Colletotrichum gloeosporioides, Phyllosticta vaccinii, and Physalospora vaccinii, are shown in **Figures 1A–D**. Whereas, many of the compounds had no effect on the respective fungi, effects of some of the compounds differed greatly among the species. Starch (STA), peptone (PEP), and addition of raw susceptible (RS), autoclaved susceptible (AS), and autoclaved resistant (AR) tissue of green fruit significantly stimulated (p < 0.0001; α = 0.01) growth of Coleophoma empetri, Phyllosticta vaccinii, and Physalospora vaccinii (**Figures 1A,C,D**). Protein (PRO) also significantly stimulated (p < 0.0001; α = 0.01) growth of Phyllosticta vaccinii (**Figure 1C**). Alanine (ALA), benzoic acid (BA), and quinic acid (QA) significantly inhibited (p < 0.0001; α = 0.01) growth of Coleophoma empetri (**Figure 1A**), whereas addition of amino acids (ALA, GLY), organic acids (BA, CA, FA, MA, QA), pectin (PEC) and any kind of cranberry green fruit tissue (RR, AR, RS, and AS) in the medium significantly inhibited (p < 0.0001; α = 0.01) growth of Colletotrichum gloeosporioides compared to that of the control medium (**Figure 1B**) but did not inhibited other fungi.

### Detection of H2O<sup>2</sup> Secreted into Culture Media by Fungi

Most of the compounds tested had no effects or stimulated H2O<sup>2</sup> secretion by fungi (**Table 1**; **Figures 1E–H**, **2**). However, we identified chemical compounds of cranberry fruit that inhibited H2O<sup>2</sup> secretion by the fungi.

Addition of BA and QA into the medium completely inhibited H2O<sup>2</sup> secretion by Coleophoma empetri, Colletotrichum gloeosporioides, Phyllosticta vaccinii, and Physalospora vaccinii (**Figures 1E–H**, **2**). Addition of citric acid (CA) and malic acid (MA) also significantly (p < 0.0001; α = 0.01) reduced secretion of H2O<sup>2</sup> into the medium by Colletotrichum gloeosporioides and Phyllosticta vaccinii (**Figures 1F,G**), and by Colletotrichum gloeosporioides, Phyllosticta vaccinii, and Physalospora vaccinii, respectively (**Figures 1F–H**, **2**). However, folic acid (FA) and PEC added to the medium significantly (p < 0.0001; α = 0.01) stimulated production of H2O<sup>2</sup> by Phyllosticta vaccinii (**Figure 1G**). Medium amended with glucose showed significant (p < 0.0001; α = 0.01) reduction of H2O<sup>2</sup> secretion by

FIGURE 1 | Growth (green) of Coleophoma empetri (A), Colletotrichum gloeosporioides (B), Phyllosticta vaccinii (C), and Physalospora vaccinii (D) and secretion of hydrogen peroxide (red) into the media by Coleophoma empetri (E), Colletotrichum gloeosporioides (F), Phyllosticta vaccinii (G), and Physalospora vaccinii (H), respectively. CTR, control (dark green or dark red); ALA, alanine; GLY, glycine; BA, benzoic acid; CA, citric acid; FA, folic acid; MA, malic acid; QA, quinic acid; FRU, fructose; GLU, glucose; SUC, sucrose; PEC, pectin; STA, starch; MAN, mannitol; SOR, sorbitol; PEP, N-Z-Soy® Peptone; PRO, Proflo Premium Quality Cottonseed protein; RR, raw rot-resistant genotype green berries; AR, autoclaved rot-resistant green berries; RS, raw rot-susceptible cranberry green berries; AS, autoclaved

growth of colonies (mm; along perpendicular axes) ± standard error of the mean or reactive oxygen species (ROS) reaction zone (mm) ± standard error of the mean of hydrogen peroxide (3,3′ -diaminobenzidine tetrachloride/horseradish peroxidase staining) secreted into the media in triplicates (N = 3). The same letters are not significantly different (P < 0.01; α = 0.01) as determined by the Ryan-Einot-Gabriel-Welsch Q (REGWQ) multiple range test. Organic acids (benzoic and quinic acids) show inconsistent suppression of growth but consistent suppression of ROS in all rot fungi tested. Amino acids, sugars, polysaccharides, sugar alcohols, and proteins often increase fungal growth and show no effect or increase ROS secretion by fruit rot fungi.

susceptible cranberry green berries. Values are the average of radial

FIGURE 2 | Cranberry fruit rot fungi Coleophoma empetri (1), Colletotrichum gloeosporioides (2), Phyllosticta vaccinii (3), Physalospora vaccinii (4) grown on basal medium (0.5% agarose and 0.5% glucose; control) before staining (A), basal medium (0.5% agarose and 0.5% glucose; control) (B), and basal medium with: mannitol (C), malic acid (D), benzoic acid (E), or quinic acid (F); and then stained with 3,3′ -diaminobenzidine tetrachloride/horseradish peroxidase to visualize hydrogen peroxide secretion into the media. Red pigment around fungal colonies on stained control, mannitol and malic acid media indicates high production of hydrogen peroxide. Benzoic acid and quinic acid inhibit hydrogen peroxide production in all rot fungi tested.

Phyllosticta vaccinii (**Figure 1G**). Addition of both, STA and PEP to the media significantly (p < 0.0001; α = 0.01) reduced of H2O<sup>2</sup> secretion by Coleophoma empetri (**Figure 1E**) but stimulated significantly (p < 0.0001; α = 0.01) secretion H2O<sup>2</sup> by Colletotrichum gloeosporioides (**Figure 1F**); medium amended with STA significantly (p < 0.0001; α = 0.01) reduced secretion of H2O<sup>2</sup> by and Phyllosticta vaccinii (**Figure 1G**). Media amended with ALA, glycine (GLY), sucrose (SUC), or PEC significantly stimulated (p < 0.0001; α = 0.01) H2O<sup>2</sup> secretion by Physalospora vaccinii compared to that of the control medium (**Figure 1H**). Presence of RR significantly (p < 0.0001; α = 0.01) inhibited secretion of H2O<sup>2</sup> into the medium by all tested fungi (**Figures 1E–H**); addition of RS, AR, and AS significantly (p < 0.0001; α = 0.01) inhibited secretion of H2O<sup>2</sup> by Phyllosticta vaccinii (**Figure 1G**), while addition of RS significantly (p < 0.0001; α = 0.01) inhibited secretion of H2O<sup>2</sup> into the medium by Physalospora vaccinii (**Figure 1H**).

### Quantification of Organic Acids in Developing Cranberry Fruit

Our results showed that the levels of BA, QA, and CA in developing cranberry fruit were significantly different depending on a cranberry genotype and a fruit development stage (**Table 2**).

### Benzoic Acid

Comparing the plant genotypes within the collection date revealed a 5.7-, 4.0-, 1.4-, 2.5-, and 6-fold variation in BA content between the genotypes collected at a particular collection date (**Table 2**). At the first collection date (July 26), a significantly higher (p < 0.0003; α = 0.01) level of BA was found in the genotype US88-79 than in the US88-1, US89-3, and ST genotypes. On August 6 significantly higher (p < 0.0009; α = 0.01) BA content was found again in the genotype US88-79 and US88-30 than in ST. Starting with the collection date of August 23, due to severe fruit rot occurrence, there were no more fruit of the rot-susceptible genotypes (MQ and ST) available for further analysis; the number of fruit of the rot-resistant genotype US88- 1 was limited as well. On both collection dates, August 23 and September 5, there was no significant variation in BA content between all analyzed rot-resistant genotypes (RR). At the last collection date (September 15), content of BA in fruit of the genotype US88-30, US88-79, and US89-3 was significantly higher (p < 0.0004; α = 0.01) than in the genotype US88-1. During the collection period (from July 26 to September 15), the mean level of BA in the developing fruit increased 29.4-fold, from 0.0044 to 0.1295 mg g−<sup>1</sup> (**Table 2**; **Figure 3A**).

Analysis of BA content in the first two collection dates showed significant differences between investigated genotypes and collection dates; a significantly higher (p < 0.0001; α = 0.01) level of BA was found in fruit of the genotype US88-79 than in fruit of ST and US88-1 genotypes (**Figure 4A**). Concentration of BA increased over 300% (**Figure 4B**) and was significantly higher (p < 0.0001; α = 0.01) in the second collection date (August 6) than in the first collection date (July 26).

### Citric Acid

The level of CA at the first collection date (July 26) varied and was significantly higher (p < 0.0001; α = 0.01) in the US88- 1, US88-30, and US89-3 genotypes than in MQ (**Table 2**). There were no significant differences in CA content found in fruit of all analyzed genotypes at the second (August 6) and the last (September 15) collection date. On August 23 (p < 0.0003; α = 0.01) and September 5 (p < 0.0001; α = 0.01) significantly higher CA concentration was found in the genotype US89-3 and in the genotype US88-1 than in all other analyzed genotypes, respectively. However, the result obtained for the genotype US88- 1 was based on one extract sample only. As the fruit developed, the mean level of CA in the cranberry fruit first increased from 8.897 mg g−<sup>1</sup> on July 26 to 10.949 mg g−<sup>1</sup> on August 23 and then gradually decreased to 9.811 mg g−<sup>1</sup> by the end of the growing season (September 15) (**Table 2**; **Figure 3B**).


TABLE 2 | Concentrations (mg g−<sup>1</sup> fresh weight) of benzoic acid (BA), citric acid (CA), malic acid (MA), and quinic acid (QA) in developing fruit of six cranberry genotypes collected during growing season.

Results are expressed as mean ± standard error of the mean (mg g−<sup>1</sup> fw); N = 6 unless indicated otherwise; MQ, Mullica Queen; ST, Stevens; R, rot-resistant genotype; S, rotsusceptible genotype; NS, no sample available due to fruit rot.

\*Values with the same letters within a compound and within a column indicate that the genotypes are not significantly different (P < 0.01; α = 0.01) as determined by the Ryan-Einot-Gabriel-Welsch Q (REGWQ) multiple range test.

\*\*No significant genotype by collection date interaction found but there was significant genotype and collection date effects (P < 0.01; α = 0.01).

The genotype US88-30 was found to contain significantly higher (p < 0.0001; α = 0.01) level of CA in young fruit (the first two collection periods) than MQ and US88-79 genotypes (**Figure 4C**) and there were no significant differences observed between the collection dates (**Figure 4D**).

### Malic Acid

There was no significant genotype by collection date interaction for MA content in fruit of the investigated genotypes, comparing to all collection dates (**Table 2**; **Figure 3C**). Statistical analysis revealed that during the season significantly (p < 0.0001; α = 0.01) higher concentration of MA was found in MQ genotype than in the genotype US88-79 and ST (**Figure 5A**). However, fruit of the MQ genotype, due to the severe fruit rot, were only available for analysis for the first two collection dates (July 26 and August 6). The average concentration of MA (when comparing all investigated genotypes combined for each date) increased as the season progressed (**Figure 5B**) and reached significantly (p < 0.0001; α = 0.01) higher concentration (5.114 mg g−<sup>1</sup> ) on the final collection date (September 15).

Young fruit of the MQ genotype had a significantly (p < 0.0001; α = 0.01) higher MA content than the young fruit of the genotype US88-30, US89-3, US88-79, and ST while comparing to the first two collection dates (**Figure 4E**). However, there were no significant differences in MA content between the two collection dates found (**Figure 4F**).

### Quinic Acid

The content of QA in all analyzed samples varied from 14.6 to 31.1 mg g−<sup>1</sup> (**Table 2**). There was a 1.3- to 1.5-fold variation found in the level of QA between the investigated genotypes at each collection date. On July 26, a significantly (p < 0.0001; α = 0.01) higher level of QA was found in the genotype US89- 3 compared to US88-1 and US88-79 genotypes. At the second collection date significantly (p < 0.0008; α = 0.01) higher QA

content was found again in the US89-3 genotype than in the genotype US88-79 and ST. A significantly (p < 0.0004; α = 0.01) higher level of QA was found in the genotype US89-3 on August 23 than in the other three genotypes analyzed. On September 5, the genotype US89-3 showed significantly (p < 0.0002; α = 0.01) higher QA content than the genotype US88-1 and US88- 30. At the last collection date, content of QA was significantly (p < 0.0019; α = 0.01) higher in fruit of the genotype US89-3 and US88-30 than in US88-1 and US88-79 genotypes. However, genotype US88-30 was represented by two samples only. On

average, the mean level of QA in fruit decreased about 30% as the fruit mature, from 26.496 mg g−<sup>1</sup> on July 26 to 18.625 mg g−<sup>1</sup> on September 15 (**Table 2**; **Figure 3D**).

In comparing the concentrations of QA in the young fruit during the first two collection dates, the genotype US89-3 was found to contain significantly (p < 0.0001; α = 0.01) higher levels of QA than fruit of genotypes US88-79, US88-1, and ST (**Figure 4G**). The levels of QA in fruit showed a tendency to fall between these two dates and were significantly (p < 0.0001; α = 0.01) lower on August 6 (**Figure 4H**).

### Discussion

Most of the tested cranberry compounds (**Table 1**) were found to increase, stimulate or only slightly inhibit H2O<sup>2</sup> secretion, when compared to the control (**Figures 1E–H**). BA and QA, when added to the medium, completely inhibited H2O<sup>2</sup> production. In addition, green fruit extract from unautoclaved cranberry fruit of RR significantly inhibited H2O<sup>2</sup> secretion compared to that from the more rot-susceptible genotypes and/or the control (**Figures 1E–H**). Autoclaving of fruit resulted in loss of inhibitory activity; autoclaving may have caused denaturation or degradation of QA and other inhibitors.

Our data show that levels of QA in cranberries of the six tested genotypes decreased throughout fruit development (**Table 2**) and the RR showed a more gradual decline in QA levels during the first 2 weeks of our study than that observed in the more rot-susceptible genotypes (**Figure 3D**). At the same time, the content of BA in fruit increased as the fruit developed. In general, more RR showed higher levels of BA early in fruit development, i.e., in the first 2 weeks of our study (**Table 2**; **Figure 3A**). The levels and trends for BA and QA we found to be consistent in two growing seasons (unpublished data). Chemical and physical factors are thought to account for differences often observed in disease resistance among different fruit development stages (Prusky, 1996). Kalt and McDonald (1996) found that contents of CA and QA in Lowbush blueberry (Vaccinium angustifolium Aiton) were lower in overripe compared to underripe berries, while the concentration of MA was at similar levels among all maturity groups. These authors (Kalt and McDonald, 1996) also reported significant cultivar differences in CA, QA, but not MA. Observed decreases in content of QA and increases in BA levels in cranberry fruit as the season progresses may be related to increasing levels of fruit infection by fungi (Tadych et al., 2012). The high content of QA in cranberry fruit might indicate its importance as a chemical defense compound (Grayer and Kokubun, 2001) and perhaps also as a precursor for antimicrobial secondary metabolites (Weinstein et al., 1961; Boudet, 1980; Hawkins et al., 1993; Richards et al., 2006; Pero et al., 2008; Dewick, 2011; Tzin et al., 2012; Ghosh et al., 2014).

### Mechanism of Action

Our experiments were focused on understanding the effects of cranberry compounds on pathogenicity behavior (e.g., ROS secretion) of the cranberry fungi. However, earlier research has associated H2O<sup>2</sup> and other ROS secretion by fungal necrotrophs with the initial trigger of localized cell death and necrosis of host tissues (Govrin and Levine, 2000). Prevention of ROS secretion by fungi may prevent the initiation of the hypersensitivity response in fruit tissues. The likely mechanism of action for the ROS suppression effect of BA and QA is through inhibition of oxidase and oxygenase enzymes that are responsible for production of reactive oxygen. Among the oxidase enzymes of fungi are laccases that play roles in pathogenesis, detoxification, polyphenolic, and lignin degradation (Claus, 2004; Solomon et al., 2014). Generalized inhibitory activity to oxidase enzymes suggests the potential that plant produced ROS-suppressive compounds like BA may also inhibit oxidases in plants. BA is well documented to have a direct inhibitory effect on the cyclooxygenases(COX) of animals (Marnett and Kalgutkar, 2004; Corazzi et al., 2005). One of the known cyclooxygenases is COX-2, an inducible isoform of cyclooxygenase enzyme responsible for the production of pro-inflammatory prostaglandins in inflamed and neoplastic tissues. In animals the inhibition of the COX-2 enzyme by BA results in reduced ROS production with a consequent reduction in inflammation. Because BA has this effect, it is generally considered to be an anti-inflammatory. Among plant oxygenases are pathogen-induced oxygenases (PIOX) (Sanz et al., 1998). PIOX enzymes produce ROS defensively in response to invasion by pathogens. The PIOX enzymes of plants have been shown to have considerable homology to COX-2 in animals (Jahabbakhsh-Godehkahriz et al., 2013). Because of these similarities, it seems likely that oxygenase inhibitors, like BA and QA, may have ROS suppressive effects in both fungus and host tissues, essentially suppressing both fungal and plant secretion of ROS and preventing the hypersensitive response in the host.

### Growth Inhibitory Action of Benzoic Acid and Quinic Acid

Benzoic acid naturally occurs in both plant and animal tissues with low or no toxicity evident. However, it is known to inhibit some bacteria and fungi by reducing respiration (Warth, 1991). Because of this effect, BA and its salts, calcium benzoate, potassium benzoate, and sodium benzoate, are used at levels ranging from 0.03 to 0.3% as preservatives in food products, beverages, dentifrices, cosmetics, and pharmaceuticals to prevent decomposition by microbial growth (Krebs et al., 1983; FAO/WHO Food Standards, 2013).

Benzoic acid is known as one of the simplest of phytoalexins (Harborne, 1983; Grayer and Kokubun, 2001). The resistance of immature apples to Neonectria ditissima (Tul. & C. Tul.) Samuels & Rossman (syn. Nectria galligena Bres.) is thought to be related to the presence of BA in apples, which is accumulated by the apple after fungal infection (Brown and Swinburne, 1971, 1973; Seng et al., 1985). Exogenous applications of BA in vitro (at 9 mM concentration) completely inhibited the growth of Bipolaris oryzae (Breda de Haan) Shoemaker (the casual agent of rice brown spot disease) and under field conditions (at 20 mM), significantly reduced both disease severity and incidence in plant leaves as well as led to a significant increase in grain yield (Shabana et al., 2008). Similarly, BA at a concentration of 20 mM significantly reduced growth and spore germination of Fusarium oxysporum Schlect. emend. Snyd & Hans, Fusarium solani (Mart.) Sacc. and Rhizoctonia solani Khun (Shahda, 2000). Diversity and successional changes in populations of fungi in

### References

Álvarez-Loayza, P., White, J. F. Jr., Torres, M. S., Balslev, H., Kristiansen, T., Svenning, J.-C., et al. (2011). Light converts endosymbiotic fungus to pathogen, influencing seedling survival and niche-space filling of a common tropical tree, Iriartea deltoidea. PLoS ONE 6:e16386. doi: 10.1371/journal.pone.00 16386

cranberry fruits were observed as the fruits developed and the season progressed (Tadych et al., 2012). The fungi possibly stimulate biochemical responses in the fruit leading to synthesis of BA in the fruit. To the best of our knowledge the role of BA as a phytoalexin in cranberry was never documented, although its function as an antifungal compound is well known. As this study shows, young, green cranberry fruits did not accumulate BA, but its content gradually increased as the season progressed.

Özçelik et al. (2011) found that QA may also act as an antimicrobial agent. However, other studies were contrary and indicated that QA alone did not have inhibitory effects, or even stimulated growth of various microorganisms (Clague and Fellers, 1934; Valle, 1957; Sokolova, 1963; Kallio et al., 1985; Bartz et al., 2013). Growth rate of an organism is not necessarily an indication of its pathogenicity and virulence. Although QA stimulated growth of Rhizoctonia solani, it significantly reduced fungal production of plant growth regulators belonging to phenylacetic acid metabolic complex, and as a result, suppressed disease development on tomato plants (Bartz et al., 2013). In our study, QA and BA added to the medium did not entirely suppress growth of the fruit rot fungi, consistent with previous observations, but what might be more significant for fruit rot disease development, they completely inhibited H2O<sup>2</sup> production and its secretion into the medium (**Figures 1**, **2E,F**) and may have inhibited secretion of other virulence factors.

### Conclusions

Based on our studies, further examination of organic acids for their virulence inhibition effects seems warranted. To reduce plant disease it may be a viable strategy to select crop plants that maintain higher levels of organic acids or other potential virulence suppressors through plant development. We propose that organic acids and other compounds should be examined as potential modulators of virulence in fungi and defensive reaction in hosts.

### Acknowledgments

We thank Dr. Edward Durner for assistance in statistical analyses. This work was supported in part by the United States Department of Agriculture Specialty Crop Research Initiative (SCRI) 2008-51180-04878 (NV) grant, the British Columbia Cranberry Growers Association, Ocean Spray Cranberries, Inc., the New Jersey Agricultural Experiment Station, USDA NIFA Multistate Project W3147 and the Rutgers Center for Turfgrass Science.

Apel, K., and Hirt, H. (2004). Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu. Rev. Plant Biol. 55, 373–399. doi: 10.1146/annurev.arplant.55.031903.141701

Bartz, F. E., Glassbrook, N. J., Danehower, D. A., and Cubeta, M. A. (2013). Modulation of the phenylacetic acid metabolic complex by quinic acid alters the disease-causing activity of Rhizoctonia solani on tomato. Phytochemistry 89, 47–52. doi: 10.1016/j.phytochem.2012.09.018


cranberry ovaries from flower to mature fruit: diversity and succession. Fungal Divers. 54, 101–116. doi: 10.1007/s13225-012-0160-2


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2015 Tadych, Vorsa, Wang, Bergen, Johnson-Cicalese, Polashock and White. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Enhancing Neoplasm Expression in Field Pea (Pisum sativum) via Intercropping and Its Significance to Pea Weevil (Bruchus pisorum) Management

Abel Teshome<sup>1</sup> \*, Tomas Bryngelsson<sup>1</sup> , Esayas Mendesil<sup>2</sup> , Salla Marttila<sup>2</sup> and Mulatu Geleta<sup>1</sup>

#### Edited by:

Gero Benckiser, Justus-Liebig-Universität Gießen, Germany

#### Reviewed by:

Oswaldo Valdes-Lopez, National Autonomus University of Mexico, Mexico Vijai Kumar Gupta, NUI Galway, Ireland

#### \*Correspondence:

Abel Teshome abelito.teshome@gmail.com; abel.teshome@slu.se

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 25 February 2016 Accepted: 28 April 2016 Published: 18 May 2016

#### Citation:

Teshome A, Bryngelsson T, Mendesil E, Marttila S and Geleta M (2016) Enhancing Neoplasm Expression in Field Pea (Pisum sativum) via Intercropping and Its Significance to Pea Weevil (Bruchus pisorum) Management. Front. Plant Sci. 7:654. doi: 10.3389/fpls.2016.00654 <sup>1</sup> Department of Plant Breeding, Swedish University of Agricultural Sciences, Alnarp, Sweden, <sup>2</sup> Department of Plant Protection Biology, Swedish University of Agricultural Sciences, Alnarp, Sweden

Neoplasm formation, a non-meristematic tissue growth on young field pea (Pisum sativum L.) pods is triggered in the absence of UV light and/or in response to oviposition by pea weevil (Bruchus pisorum L.). This trait is expressed in some genotypes [neoplastic (Np) genotypes] of P. sativum and has the capacity to obstruct pea weevil larval entry into developing seeds. In the present study, 26% of the tested accessions depicted the trait when grown under greenhouse conditions. However, UV light inhibits full expression of this trait and subsequently it is inconspicuous at the field level. In order to investigate UV light impact on the expression of neoplasm, particular Np genotypes were subjected to UV lamp light exposure in the greenhouse and sunlight at the field level. Under these different growing conditions, the highest mean percentage of Np pods was in the control chamber in the greenhouse (36%) whereas in single and double UV lamp chambers, the percentage dropped to 10 and 15%, respectively. Furthermore, when the same Np genotypes were grown in the field, the percentage of Np pods dropped significantly (7%). In order to enhance Np expression at the field level, intercropping of Np genotypes with sorghum was investigated. As result, the percentage of Np pods was threefold in intercropped Np genotypes as compared to those without intercropping. Therefore, intercropping Np genotypes with other crops such as sorghum and maize can facilitate neoplasm formation, which in turn can minimize the success rate of pea weevil larvae entry into developing seeds. Greenhouse artificial infestation experiments showed that pea weevil damage in Np genotypes is lower in comparison to wild type genotypes. Therefore, promoting Np formation under field conditions via intercropping can serve as part of an integrated pea weevil management strategy especially for small scale farming systems.

Keywords: Bruchus pisorum, field pea, intercropping, neoplasm, pea weevil, Pisum sativum

## INTRODUCTION

fpls-07-00654 May 13, 2016 Time: 15:0 # 2

Abiotic factors like light, water and nutrients have a major influence on the phenotype of crops which in turn influences the multi-trophic interactions of the crop with ecological and economic implications (Dicke and Hilker, 2003). Some genotypes of field pea (Pisum sativum ssp. sativum L.) produce neoplasm on its pods when grown under greenhouse conditions. This phenomenon is a non-meristematic tissue growth on the surface of young pods in response to absence of UV light (Nuttall and Lyall, 1964; Dodds and Matthews, 1966). This trait can also be triggered by pea weevil (Bruchus pisorum L.) as a direct response to oviposition (Berdnikov et al., 1992; Doss et al., 2000). However, according to Doss et al. (1995), neoplastic (Np) tissue growth triggered by pea weevil oviposition is morphologically different from that caused by absence of UV light. Neoplasm in field pea is a result of mutation of a gene at Np (neoplasm) locus, in which a mutant allele (Np) is dominant over a wild type allele (np) (Nuttall and Lyall, 1964). Despite the fact that the expression of this trait is controlled by the dominant (Np) allele, its penetrance is influenced by the genotype (homozygosity: Np/Np vs. heterozygosity: Np/np), and the level of UV light intensity or humidity (Nuttall and Lyall, 1964; Burgess and Fleming, 1973; Doss et al., 2000). This trait has also been reported in other Pisum species like P. elatius and P. humile under greenhouse conditions (Dodds and Matthews, 1966).

Organic compounds extracted from pea weevil that were referred to as bruchins were reported to trigger neoplasm formation when applied on young pea pods of particular genotypes (Doss et al., 1995; Oliver et al., 2000; Doss, 2005). Doss (2005) reported that treating Np pea pods with bruchins lead to upregulation of genes that are known to be involved in defense metabolic pathways. In general, most studies conducted to date show that this trait is an induced response triggered by both abiotic and biotic stresses. Comparison of Np and wild type genotypes in their resistance to pea weevil revealed a lower average pea weevil damage in neoplasm producing genotypes under field (Doss et al., 2000) and greenhouse conditions (Teshome et al., 2015), which suggests the importance of this treat in pea weevil managment.

Pea weevil is the major menace in field pea production in Ethiopia and elsewhere (Pesho et al., 1977; Clement et al., 2002; Seyoum et al., 2012). Currently, chemical pesticide spraying and after harvest fumigation are the only options for pea weevil containment (Horne and Bailey, 1991; Baker, 1998). Despite the potential role of neoplasm formation in pea weevil management, little attention has been given to it until now. This is partly due to the fact that the penetrance of the trait under field conditions is inconspicuous and/or inconsistent (Nuttall and Lyall, 1964; Doss et al., 2000). In the present study, the influence of UV light on the expression of neoplasm under greenhouse conditions was studied in parental and F<sup>1</sup> hybrids of Np and wild type (non-Np) genotypes. In addition, intercropping Np genotypes was investigated, if the shade provided by intercropping could enhance Np expression in Np genotypes under field conditions. Furthermore, both Np and wild type genotypes were screened for pea weevil resistance under greenhouse conditions.

## MATERIALS AND METHODS

### Impact of UV Light on Neoplasm Formation

Neoplasm producing genotypes were identified from various field pea accessions during greenhouse screening experiments for pea weevil resistance in 2012 and 2013 (Teshome et al., 2015). The Np genotypes used in the present study are hereafter referred to as Np genotypes. The Np genotypes used in this study were 32433A, 203084A, 235899A, 237065A, 226037A, 226037B, 226037C, 226037D, and 226037E. Two separate experiments were carried out to study the effect of UV light on neoplasm formation. Primarily, Np genotypes were tested both under greenhouse and field conditions. Additionally, F<sup>1</sup> hybrids produced from crosses of wild type genotype pollen recipient and Np pollen donor parents were tested for neoplasm formation under greenhouse conditions.

All plants were grown in 2 l plastic pots in a greenhouse chamber at 22◦C and a minimum of 12 h light. Before flowering, plants were moved into chambers (2.54 m<sup>2</sup> ) covered with a lightproof plastic sheet. Five chambers were used for this experiment with all having two cool white Fluorescent lamps, Sylvania Luxline plus 58W. One of the five chambers was used as a control chamber. The remaining four chambers had UV lamps, two of which had a single UV lamp (3 U 15 W UV light bulb) and the remaining two having double UV lamps (2 × 3 U 15 W UV light bulbs) each. The plants in the experimental chambers were exposed to UV light for 12 h from 6:00 pm to 6:00 am. Each genotype was represented by a minimum of three and a maximum of six replicates. Pods were harvested at maturity and individual pods of each genotype were assessed for neoplasm formation. Based on the level of neoplasm formation, pods from each plant were categorized into two different groups, low and high. The low score was given when there was sparse coverage of Np tissue on both sides of the pods. A high score was given when there was a conspicuous neoplasm formation that covered most of the pod. A similar protocol was used for the F<sup>1</sup> hybrids as for the parental Np genotypes although they were only tested in double UV lamp and control chambers.

### Intercropping of Np Genotypes with Sorghum

Field experiments were carried out at Alnarp, Sweden in 2013 and 2014 to investigate the effect of shading provided by the canopy of sorghum on neoplasm expression. In this experiment, Np genotypes were intercropped with sorghum (Sorghum bicolor L. Moench.). Sorghum and pea genotypes were planted in different rows, in which each pea row had adjacent rows of sorghum on both sides. The distance between pea plants in a row was 5 cm and likewise in-between sorghum plants. The distance between rows was 10 cm. The total area of the plot was 2 m<sup>2</sup> with two replications. Sorghum plants were also grown along the borders of the plots. There were five blocks in total with two blocks with intercropping and another two without intercropping and the last block with shading but without intercropping.

### Scanning Electron Microscopy (SEM)

Morphological and anatomical characteristics of Np pods, which were harvested at maturity from plants grown in the greenhouse, were examined under SEM. Small pieces of the pods from genotype 226037B were fixed with 2.5% glutaraldehyde in 0.1 M Na-phosphate buffer pH 7.2 overnight at +4 ◦C, washed with the same buffer 3 × 15 min, dehydrated in graded series of ethanol and critical-point dried (CPD 020, Balzers, Lichtenstein). The samples were attached on the sample stubs with double-sided tape and sputter coated with gold and palladium 3:2 mix (JFC-1100, JOEL, Tokyo, Japan), and examined in a SEM (435 VP, LEO Electron Microscopy Ltd, Cambridge, UK) with 10 kV.

### Greenhouse Screening of Np and Wild Type Genotypes for Pea Weevil Resistance

Artificial infestation was done in insect rearing cages (60 cm × 60 cm × 120 cm, MegaView Science Co Ltd, Taiwan). The plants were moved into cages when they started to flower. Six plants were placed in each cage and each genotype was intermixed with different genotypes in consecutive experiments. Newly emerged adult pea weevils from seeds of a previous pea weevil screening study were used for artificial infestation. The weevils were kept at 4◦C until they were released. In order to balance the sex ratio and ensure successful mating, the sex of the weevils was determined as described by Bousquet (1990) ahead of release. Twenty-five pairs of naive male and female pea weevils were released in each cage as soon as the first flower was detected. Pods were harvested at maturity and stored at room temperature. Three months after harvest, damage assessment was carried out on seeds of each genotype. Percent seed damage (PSD) was calculated based on pea weevil damage symptoms as described in Teshome et al. (2015). All plants used in this experiment were grown in a similar manner as in the UV light experiment.

### Data Analysis

All percentages of Np pods and PSD were arcsine transformed to homogenize variances and ensure normal distribution. A one way analysis of variance model was used to compare the proportion of neoplasm formation in Np genotypes when grown under different greenhouse and field conditions. In addition, pair-wise comparisons between the control, i.e., greenhouse normal lamp condition and all other conditions were carried out based on posthoc t-test with P-values adjusted using the single-step method (Hothorn et al., 2008). A significance level of 5% was used for ANOVA and multiple comparison test. All analysis were carried on R version 3.2.2 (R Core Team, 2015).

### RESULTS

Among 19 accessions used for resistance screening against pea weevil in the greenhouse, five accessions showed consistent neoplasm expression (**Table 1**). Neoplasm formation in these genotypes was clear and distinct (**Figure 1A**). The scanning electron micrographs also revealed a distinct outgrowth on TABLE 1 | Comparison of average performance of neoplastic (Np) and non-Np genotypes originating from different accessions against pea weevil attack in greenhouse experiments.


<sup>N</sup>refers to accessions with consistent Np formation, the rest are wild type (without Np formation). PF, Pisum fulvum genotype; PSD, percent seed damage.

the outer surface of pods of Np genotypes (**Figure 1B**). The remaining 15 field pea accessions as well as P. fulvum did not show any Np growth in repeated greenhouse experiments (**Table 1**).

Particular Np genotypes that have shown consistent Np formation in repeated greenhouse trials were exposed to UV light to investigate UV influence on neoplasm formation. The highest percentage of Np pods, 36%, was recorded when genotypes were grown in the control chamber under greenhouse conditions and the least, 7%, when the replicates were grown in the field without intercropping. The mean percentage of Np pods in the field with intercropping was threefold of the mean Np pods without intercropping (**Table 2**). In addition, the median of the percentage of Np pods with intercropping was also higher than the median of the percentages recorded for the single and double UV light exposed chambers under greenhouse conditions (**Figure 2**).

The comparison of the mean percentage of Np pods under greenhouse UV light exposure and field conditions revealed a marginally significant difference (P = 0.05). Further post-hoc test revealed that the mean percentage of Np pods without intercropping was significantly different from the control group in the greenhouse (**Table 2**). The Np genotypes grown under single or double UV lamps in the greenhouse or intercropped in the field were not significantly different from the control group in their mean percentage of Np pods. Under intercropping conditions, three genotypes 235899A, 237065A, and 22603B scored 30% or higher percent of Np pods. Interestingly, 203084A scored the highest percentage of Np pods (42.9%) in the field with

TABLE 2 | Pair-wise comparison of percent neoplasm formation on pods of selected genotypes grown under different greenhouse and field conditions with percent neoplasm formation on pods of same genotypes grown under normal light greenhouse condition (control).


Arc mean, arcsine transformed mean for ANOVA. The arcsine and original mean of percent neoplasm formation of the control was 0.37 and 0.36, respectively. <sup>∗</sup>Significantly different.

intercropping but produced low percentage of Np pods in the control chamber (data not shown).

genotype 226037B; (C) pea weevil eggs oviposited on neoplastic (Np) pod in greenhouse screening.

All F1 hybrids produced by crossing non-Np (used as pollen recipient) and Np (used as pollen donors) produced neoplasm under greenhouse conditions (**Figure 3**). In both control and UV chambers, the highest percentage of Np pods was recorded for the F<sup>1</sup> hybrid 32018-20 × 226037-2S with 100 and 77.8% Np pods, respectively. The least percentage of Np pods was scored for the 32397-6 × 226037D hybrid which was 50% in the control chamber. In most F<sup>1</sup> hybrids, the percentage of Np pods decreased significantly when grown under double UV lamp conditions (**Figure 3**).

In general, Np accessions scored low average PSD in comparison to susceptible checks. The least average PSD among Np accessions was recorded for genotype 32433 which was 12% and the highest, 35%, for 203084 (**Table 1**). Among the wild type accessions, 231277, 32426, and 208459 recorded low PSD whereas 32397 (susceptible check) scored the highest average PSD (42%). After the initial screening that included both Np and wild type accessions, selected Np genotypes, a P. fulvum accession from NordGen and a Np F<sup>1</sup> hybrid were tested for pea weevil resistance under greenhouse conditions. The Np genotypes scored less PSD in three consecutive greenhouse experiments with few notable exceptions (**Table 4**). The PSD of most Np genotypes was lower than the average seed damage all of plants in the same cage. For example, genotype 237065A scored 30% PSD at the first screening despite the mean PSD per cage and percent of plants with infested seeds per cage was 51 and 100%, respectively. Relatively lower PSD values were also observed for this genotype in the second and third round of screenings using seeds from the same generation. The Np F<sup>1</sup> hybrid of 32397-3 × 226037A also scored a relatively low PSD in two consecutive experiments. On the contrary, 203084A, which is an Np genotype, scored high PSD values in all experiments. This genotype scored low percentage of Np pods in a control chamber in the greenhouse (**Table 3**). The P. fulvum genotype scored no seed damage in two consecutive experiments (**Table 4**). Analysis of variance (ANOVA) of PSD of Np genotypes, F<sup>1</sup> hybrids and the P. fulvum line gave a highly significant variation, P = 0.001 (**Table 4**).

### DISCUSSION

Neoplasm formation is an infrequent phenomenon that occurs in certain genotypes of P. sativum in the absence of UV light, for example under greenhouse conditions (Nuttall and Lyall, 1964). **Figure 1A** shows Np tissue growth in Np genotype 226037B when grown under greenhouse conditions. In the present study, 26% of the tested P. sativum accessions depicted the trait (**Table 1**). According to Berdnikov et al. (1992), only 2.3% of the assessed Ethiopian germplasm collections showed neoplasm formation. The high percentage of Np accessions observed in the present study is most likely due to preselection of these accessions from a pool of collections used for resistance screening against pea weevil (Teshome et al., 2015).

The present study clearly showed that neoplasm formation on Np genotypes is conspicuous. However, the level of neoplasm varies among Np genotypes and growing conditions. A similar trend was reported in an oviposition preference study by Mendesil et al. (2016), where the level of neoplasm formation was different among Np genotypes. This study revealed that the highest proportion of pods with neoplasm was recorded



<sup>a</sup>Non-Np genotype. PSD, percent seed damage; MPSDC, mean percent seed damage per cage; PPISC, percent of plants with infested seeds per cage.

TABLE 4 | ANOVA comparison of mean arcsine transformed PSD of genotypes tested in greenhouse screening.


∗∗Highly significant.

when Np genotypes were grown in the control chamber under greenhouse conditions. In the control chamber, 36% of the pods showed neoplasm formation. On the contrary, when the same genotypes were exposed to single and double UV lamps, the proportion of Np pods was reduced. Furthermore, when these genotypes were grown under field conditions, the percentage of Np pods dropped to only 7%. This result is consistent with previous findings that reported a negative influence of UV light on neoplasm formation (Nuttall and Lyall, 1964; Dodds and Matthews, 1966).

Previous studies showed that the oviposition of female pea weevil on pods of Np genotypes triggers the expression of Np gene (Berdnikov et al., 1992; Hardie, 1992; Doss et al., 2000). However, the type of neoplasm formed on Np pods in the absence of UV light or upon oviposition by pea weevil is morphologically different (**Figures 1A,C**). Host plants usually trigger a series of responses to either prevent further oviposition or to reduce the success rate of deposited eggs (Hilker and Meiners, 2002; Fatouros et al., 2005; Meiners et al., 2005). According to Doss et al. (2000), neoplasm formation triggered by oviposition can impede the entry of newly hatched larvae and hence minimize infestation rate. Furthermore, the additional mass of cells on the epidermal layer of Np pods could potentially upset the behavior of the gravid female pea weevil when choosing site of oviposition. Oviposition preference experiments showed that Np genotypes had a reduced rate of oviposition as compared to wild type genotypes (Mendesil et al., 2016). Despite this trait being pertinent in pea weevil resistance, its expression is attenuated by UV light and hence less effective against pea weevil under field conditions (Nuttall and Lyall, 1964; Snoad and Matthews, 1969; Doss et al., 1995).

Nuttall and Lyall (1964) and Doss et al. (1995) detected neoplasm formation on shaded pods grown in the field. The present study has also showed an increase in neoplasm formation when the pods are shaded from direct sunlight at the field level (data not shown). However, mechanical shading is inconvenient for periodic application and resource and time consuming in the case of small-scale farming systems. On the other hand, the idea to shade the pods of Np genotypes with the canopy of taller and branching sorghum plants resulted in a significant increase in neoplasm formation. The proportion of Np pods with intercropping was three fold higher than without it at the field level (**Table 2**). The fact that neoplasm formation can be enhanced with intercropping, as shown in this study, indicates that intercropping can be implemented as part of an integrated pest management approach against pea weevil.

The fact that Np genotypes scored a relatively low PSD in three consecutive experiments under greenhouse conditions suggests intercropping as a viable approach in pea weevil management. Doss et al. (2000) reported that Np genotypes are less susceptible to pea weevil in comparison with wild type genotypes under field conditions. Similar results were also reported by Teshome et al. (2015) in an experiment conducted for screening field pea germplasm for resistance against pea weevil. Hence, enhancing Np formation with intercropping could be a way forward to minimize pea weevil damage at field level. Intercropping could also result in release of non-host volatiles that can adversely affect the pea weevil's capability to locate its host and oviposit. According to Ali et al. (2007), intercropping field pea with different crops reduces susceptibility to Ascochyta blight and pea aphid (Acyrthosiphon pisum) infestation. Hence, intercropping is a silver bullet management option in field pea production that is environmentally benign, cost effective and requires minimum skill-set for application.

In order to successfully use neoplasm formation in field pea as part of integrated pea weevil management, the trait needs to be bred into locally adapted varieties. The experiment conducted to determine the heritability of neoplasm in field pea in the present study through crossing Np genotypes with wild type genotypes showed that all F<sup>1</sup> hybrids produced neoplasm under greenhouse conditions suggesting that the Np allele at Np locus is dominant

over wild type and the inheritance of the trait is according to the principle of Mendelian genetics, in line with previous studies (Nuttall and Lyall, 1964; Dodds and Matthews, 1966). However, similar to what was observed in the parental Np genotypes, the exposure of the F<sup>1</sup> hybrids to double UV lamps results in a significant reduction in neoplasm formation (**Figure 3**), which signifies the importance of avoiding direct sunlight for effective expression of this trait under field conditions. The interspecific hybrids of Np field pea genotypes and P. fulvum also depicted neoplasm formation under greenhouse conditions. P. fulvum is known to have enhanced resistance against pea weevil as compared to cultivated P. sativum varieties (Hardie et al., 1995; Clement et al., 2002; Byrne et al., 2008). In the present study, both the Np genotypes and the P. fulvum line included in the artificial infestation experiment scored comparatively low PSD (**Table 4**). Therefore, developing field pea varieties through crossing Np genotypes with P. fulvum could result in pyramiding of resistance genes with different modes of action against pea weevil. Such varieties could have sustainable resistance and could easily be augmented by integrated pest management techniques like intercropping and trap cropping.

### REFERENCES


### AUTHOR CONTRIBUTIONS

TB and MG secured the funding. AT and MG conceived and designed the study. AT, MG, EM, and SM collected the data. AT and MG performed data analysis. AT wrote the manuscript with help of MG, TB, EM, and SM.

### FUNDING

The authors would like to thank the Swedish International Development Agency (Sida) for funding this research project. We would also like to acknowledge Ethiopian Institute of Biodiversity and NordGen, for providing field pea accessions used in the present study.

### ACKNOWLEDGMENTS

We would also like to thank Tiny Motlhaodi for providing the sorghum accessions used in the intercropping trial and Kerstin Brismar for her help with microscopy.


of the 17th Annual Conference. 26-27 November 2010 Invasive Plant Pests Threatening Ethiopian Agriculture, Ethiopianed, ed. B. Mulatu (Addis Ababa: Plant Protection Society of Ethiopia), 52–66.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Teshome, Bryngelsson, Mendesil, Marttila and Geleta. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Induction of Systemic Resistance against Aphids by Endophytic Bacillus velezensis YC7010 via Expressing PHYTOALEXIN DEFICIENT4 in Arabidopsis

Md. Harun-Or-Rashid, Ajmal Khan, Mohammad T. Hossain and Young R. Chung\*

Division of Applied Life Science (BK21 Plus), Plant Molecular Biology and Biotechnology Research Center, Gyeongsang National University, Jinju, South Korea

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

William Underwood, USDA-ARS, USA Zonghua Wang, Fujian Agriculture and Forestry University, China

> \*Correspondence: Young R. Chung yrchung@gnu.ac.kr

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 29 October 2016 Accepted: 03 February 2017 Published: 15 February 2017

#### Citation:

Rashid MH, Khan A, Hossain MT and Chung YR (2017) Induction of Systemic Resistance against Aphids by Endophytic Bacillus velezensis YC7010 via Expressing PHYTOALEXIN DEFICIENT4 in Arabidopsis. Front. Plant Sci. 8:211. doi: 10.3389/fpls.2017.00211 Aphids are the most destructive insect pests. They suck the sap and transmit plant viruses, causing widespread yield loss of many crops. A multifunctional endophytic bacterial strain Bacillus velezensis YC7010 has been found to induce systemic resistance against bacterial and fungal pathogens of rice. However, its activity against insects attack and underlying cellular and molecular defense mechanisms are not elucidated yet. Here, we show that root drenching of Arabidopsis seedlings with B. velezensis YC7010 can induce systemic resistance against green peach aphid (GPA), Myzus persicae. Treatment of bacterial suspension of B. velezensis YC7010 at 2 × 10<sup>7</sup> CFU/ml to Arabidopsis rhizosphere induced higher accumulation of hydrogen peroxide, cell death, and callose deposition in leaves compared to untreated plants at 6 days after infestation of GPA. Salicylic acid, jasmonic acid, ethylene, and abscisic acid were not required to confer defense against GPA in Arabidopsis plants treated by B. velezensis YC7010. Bacterial treatment with B. velezensis YC7010 significantly reduced settling, feeding and reproduction of GPA on Arabidopsis leaves via strongly expressing senescence-promoting gene PHYTOALEXIN DEFICIENT4 (PAD4) while suppressing BOTRYTIS-INDUCED KINASE1 (BIK1). These results indicate that B. velezensis YC7010-induced systemic resistance to the GPA is a hypersensitive response mainly dependent on higher expression of PAD4 with suppression of BIK1, resulting in more accumulation of hydrogen peroxide, cell death, and callose deposition in Arabidopsis.

Keywords: Bacillus velezensis YC7010, aphid, Arabidopsis, induced systemic resistance, PAD4

### INTRODUCTION

Plants are usually challenged by various herbivorous insects in their natural environments. They have to develop diverse defense responses to protect themselves against attacks from different insects including aphids. Green peach aphid (GPA), Myzus persicae, a phloem sap feeding insect, has a wide range of hosts. It causes severe yield losses. It is also involved in the transmission of several plant viral diseases (Kennedy et al., 1962; Matthews, 1991; Blackman and Eastop, 2000).

These aphids have highly modified stylets that enable them to enter sieve elements and secrete gelling and watery saliva to a certain extent during probing and feeding. Salivation of aphid plays a major role in successful colonization. It is also associated with its virulence (Tjallingii, 2006). During infestation of plants by aphids, the major defense related plant hormones are salicylic acid (SA), jasmonic acid (JA), ethylene (ET), and abscisic acid (ABA). They are reported to be involved in the induced systemic resistance (ISR) of many plants against aphids (Morkunas and Gabry´s, 2011). The SA signaling pathway is activated to induce resistance in a number of plant species by aphid feeding (Moran et al., 2002; Zhu-Salzman et al., 2004; Coppola et al., 2013). On the contrary, Pegadaraju et al. (2005) have reported that SA signaling is not involved in the defense response of Arabidopsis against aphids. It has been reported that exogenous application of JA to a tomato plant can induce systemic defense against potato aphid (Cooper and Goggin, 2005). The population of GPA is increased in ET-insensitive Arabidopsis mutant ein2, indicating that ET can also confer resistance to aphid (Kettles et al., 2013). Resistance of Arabidopsis to aphid also depends on ABA biosynthesis and signaling (Kerchev et al., 2013).

Plant defense and cell death pathways induced by pathogens and insects are often regulated by certain plant hormones to elicit the accumulation of hydrogen peroxide (H2O2) and callose deposition (De Vos et al., 2005; Ahn et al., 2007; Zhou et al., 2009). In pathogen-infected plants, the production of reactive oxygen species (ROS, e.g., H2O2), and consequentially the onset of cell death referred to as 'hypersensitive response' (HR) can lead to systemic resistance (Durrant and Dong, 2004; Jones and Dangl, 2006; Singh et al., 2016). ROS and local cell death are also major defense mechanisms used by plants to protect themselves against phloem sap feeding GPA (Lei et al., 2014). Arabidopsis PHYTOALEXIN DEFICIENT4 (PAD4) is especially crucial for its defense against aphids. Elevated expression of PAD4 can deter GPA from settling on plants or feeding from the sieve elements (Louis et al., 2012; Lei et al., 2014). In addition, it has been reported that PAD4 can stimulate premature leaf senescence, resulting in elevated expression of a subset of SENESENCE ASSOCIATED GENES (SAG) characterized by chlorophyll loss and cell death in GPA infested plants (Pegadaraju et al., 2005, 2007). The expression of PAD4-dependent constitutive expression of SAG13 can confer hyper-resistance of Arabidopsis to GPA infestation (Louis et al., 2010). Recently, it has been reported that the molecular mechanism of PAD4 resistance against aphid is reliant on interaction with BOTRYTIS-INDUCED KINASE1 (BIK1). The mutant bik1 can induce resistance to aphids through ROS production, cell death and leaf senescence. Such bik1 induced resistance is dependent on the expression of PAD4. However, BIK1 overexpression can make Arabidopsis plants more susceptible to aphid infestation (Lei et al., 2014). It has been shown that BIK1 can control plant defense against aphids by negatively regulating PAD4 expression (Louis and Shah, 2014).

BOTRYTIS-INDUCED KINASE1, a receptor-like cytoplasmic kinase (RLCK), is directly phosphorylated by BRASSINOSTEROID INSENSITIVE1-ASSOCIATED RECEPTOR KINASE (BAK1) and associated with FLAGELLIN-SENSITIVE2 (FLS2)/BAK1 complex in modulating pathogen-associated molecular patterns (PAMPs) mediated signaling (Lu et al., 2010; Zhang et al., 2010; Liu et al., 2013). Detection of PAMPs or microbe-associated molecular patterns (MAMPs) by particular transmembrane pattern recognition receptors is responsible for basal plant defense response collectively referred to as MAMP-triggered immunity (MTI) (Boller and Felix, 2009; Monaghan and Zipfel, 2012). The most characterized PAMP/MAMP receptors are receptor like kinases (RLKs), among which FLS2 and EF-TU RECEPTOR (EFR) can recognize bacterial elongation factor EF-Tu (Gómez-Gómez and Boller, 2000; Zipfel et al., 2006). Upon binding to their cognate MAMPs, FLS2, or EFR is associated with BAK1 (another RLK) during bacterial interaction with host (Chinchilla et al., 2007). For successful colonization of rhizobacteria and beneficial interaction with host plants, it is necessary to suppress MTI (Gutjahr and Paszkowski, 2009; Zamioudis and Pieterse, 2012). Pseudomonas fluorescens WCS417, one of plant growth promoting rhizobacteria (PGPR), has been shown to be able to suppress flagellin-triggered MTI responses and induce callose depositions during colonization in Arabidopsis (Millet et al., 2010). Callose deposition on sieve plates of rice plants can affect phloem transportation. It plays an important role in preventing brown planthopper (BPH) from ingesting the phloem sap (Hao et al., 2008). Several PGPR have been reported to use ISR to protect plant against pathogens. However, few studies have reported on ISR used by PGPR against insects (Zehnder et al., 1997; Lugtenberg and Kamilova, 2009; Pieterse et al., 2012). The main mechanisms of these bacteria involved in ISR upon pathogen infection or insect infestation include HR-type reactions, elevated cell wall or apoplastic peroxidase activity, callose deposition, and H2O<sup>2</sup> accumulation (Conrath, 2006; Valenzuela-Soto et al., 2010; Niu et al., 2011; Rahman et al., 2015). Recently, these PGPR have been used for plant growth promotion, stress tolerance and biocontrol agents for insects and plant pathogens (Phi et al., 2010; Van de Mortel et al., 2012; Chung et al., 2015; Hossain et al., 2016; Zebelo et al., 2016). Some endophytic PGPR inhabiting the interior of host plants have shown ISR activity against insects (de Oliveira Araujo, 2015).

Recently, we have reported that novel endophytic strain of B. oryzicola YC7010 isolated from rice roots can inhibit the growth of important fungal and bacterial pathogens of rice such as Fusarium fujikuroi and Burkholderia glumae via antibiotic production and ISR (Chung et al., 2015; Hossain et al., 2016). The novelty of this species is now on debate and the name for this species was suggested to be changed as B. velezensis (Dunlap et al., 2016). The objective of this study was to determine whether B. velezensis YC7010 could induce systemic resistance against GPA in Arabidopsis and elucidate its underlying mechanism in terms of enhancing the expression of PAD4 to activate cellular defense responses.

### MATERIALS AND METHODS

fpls-08-00211 February 13, 2017 Time: 11:54 # 3

### Plant Materials, Growth Conditions, and Aphid Rearing

Wild-type Arabidopsis ecotype Columbia-0 (Col-0), NahG and mutants sid2, jar1, ein2-1, abi2-2, pad4, bik1, and bik1pad4 were used in this experiment. Seeds were sterilized with 70% (v/v) ethanol for 5 min followed by treatment with 1.2% (v/v) sodium hypochlorite (NaOCl) for 5 min. They were then washed with sterile distilled water. After sterilization, seeds were kept at 4◦C for 48 h and grown on 0.5x Murashige and Skoog (MS) agar media supplemented with 1% (w/v) sucrose. Agar plates were then kept horizontally in a plant growth chamber for growing seedlings. Seedlings at 10 days old were transferred into pots containing 100 g soils autoclaved twice for 20 min at 120◦C. Phloem sap-feeding GPA were cultured on cabbage (Brassica oleracea) and maintained in an environmental chamber with a long day photoperiod (16 h of light and 8 h of darkness) at 22◦C with a light intensity of 100 µmol m−<sup>2</sup> s −1 . All experiments were performed in the growth chamber.

### Bioassay for Induced Systemic Resistance to Aphid by Bacteria

For bacterial inoculation of Arabidopsis plants, strain B. velezensis YC7010 (KACC18228, Jeonju, Korea) was cultivated in onetenth strength tryptic soy broth (1/10 TSB, BactoTM, Sparks, MD, USA) at 28◦C for 48 h on a rotary shaker (160 rpm). Cells were then harvested by centrifugation at 6,000 × g for 15 min and suspended in a buffer solution (10 mM MgSO4) to adjust to 2 × 10<sup>7</sup> colony forming unit (CFU) ml−<sup>1</sup> for use. Seedlings of Arabidopsis at 4 weeks old were drenched with 10 ml bacterial suspension on the rhizosphere in each pot. Equal volume of 10 mM MgSO<sup>4</sup> was drenched as control. No-choice and choice tests were performed to assess ISR to aphid after bacterial treatment. For the no-choice tests, five second-instar nymphs were placed at 5 days after bacterial inoculation. Seven days after infestation, total aphid population (adults and nymphs) on each plant was recorded. Each treatment had eight replicates. For the choice tests, 35 adult aphids were released at an equal distance between bacterial inoculated and uninoculated plants of different genotypes. At 24 h after releasing, the number of adult aphids settled on each plant was recorded. For each comparison, eight pairs of plants were used.

### Determination of Aphid Feeding Activity

Aphid feeding activity can be determined by measuring the amount of produced honeydew. To measure honeydew production, line split Whatman filter papers were placed under Arabidopsis plants of both treated and untreated plants infested by 30 adult aphids. To avoid absorbance of water from soil, a plastic membrane was placed under the filter paper. The filter papers used to collect honeydew at 3, 4, and 5 days after infestation of aphids were soaked in 0.1% (w/v) ninhydrin solution in acetone and dried in a 65◦C oven for 30 min. Purple spots were shown when honeydew was stained by ninhydrin (Kim and Jander, 2007). To quantify honeydew stains, the stained filter papers were cut into pieces and extracted with 1 ml of 90% (v/v) methanol for 1 h at 4◦C with continuous agitation. The absorbance of the supernatant was measured at 500 nm after centrifugation at 6,000 × g for 1 min. Methanol (90%) was used as a blank (Nisbet et al., 1994).

### Measurement of Hydrogen Peroxide Content

The concentrations of H2O<sup>2</sup> in bacterial treated plants were measured at 0, 12, 24, 48, and 72 h after aphid infestation using a spectrophotometer. Treated leaf tissues (200 mg) were homogenized with 3 ml phosphate buffer (50 mM, pH 6.8) containing 1 mM hydroxylamine (catalase inhibitor). The mixture was centrifuged at 6,000 × g for 25 min and 3 ml of the supernatant was mixed with 1 ml 0.1% (v/v) titanium sulfate in 20% (v/v) H2SO4. The absorbance of the supernatant of the mixture was then determined at 410 nm after centrifugation at 6,000 × g for 15 min. Extinction co-efficient (0.28 µmol−<sup>1</sup> cm−<sup>1</sup> ) was used to calculate H2O<sup>2</sup> content (Jana and Choudhuri, 1982).

### Histochemical Analyses of Cellular Defense Responses

For histochemical analyses of cellular defense responses, H2O<sup>2</sup> accumulation, cell death, and callose deposition in Arabidopsis leaves were observed. Four weeks old Arabidopsis plants were treated with the same bacterial strain used for bioassay of ISR. A total of 40 aphids were placed on both treated plants and untreated control plants at 5 days after bacterial treatment. To visualize H2O<sup>2</sup> accumulation, collected leaves at 6 days after infestation of aphids were vacuum infiltrated with 3,3' diaminobenzidine (DAB) solution (1 mg ml−<sup>1</sup> of DAB in pH 3.5 water). Under dark condition, leaves were immersed in the same solution overnight followed by destaining with 95% (v/v) ethanol until clear. They were preserved in 50% (v/v) ethanol. A dissecting microscope was used to capture the image of accumulated H2O<sup>2</sup> in the leaves.

To visualize cell death of treated leaves, trypan blue staining was performed. Trypan blue was dissolved at a concentration of 0.125 mg ml−<sup>1</sup> in lactophenol solution (phenol:lactic acid:glycerol:water [1:1:1:1]). Leaves were boiled in this staining solution for 1 min and destained in 95% (v/v) ethanol after cooling. Cell death images of leaves were captured with an Olympus Provis AX70 microscope at 10× magnifications.

To detect callose deposition, aniline blue staining was performed using published protocol (Clay et al., 2009). Buffer solution containing 10% (v/v) formaldehyde, 5% (v/v) acetic acid, and 50% (v/v) ethanol was used for fixation of Arabidopsis leaves at 37◦C overnight. Fixed leaves were washed in 95% ethanol several times until clear, rinsed twice in water and then stained for 4 h or longer in the dark with 0.01% (w/v) aniline blue in 150 mM K2HPO<sup>4</sup> (pH 9.5). Callose deposition was visualized with an Olympus Provis AX70 microscope at 10 × magnifications under UV illumination equipped with a broad band DAPI filter set. To quantify callose accumulation in the Arabidopsis leaves, images were subjected to intensity analysis using the image processing software IMAGEJ.

### Determination of Bacterial Population

In order to determine the bacterial population of B. velezensis YC7010 in Arabidopsis roots during plant growth, 10 ml of cell suspension (2.0 × 10<sup>7</sup> CFU/ml) was drenched into Arabidopsis seedlings in a pot containing 100 g soils. Roots pieces (100 mg) were collected from treated seedlings at 0, 1, 2, 4, and 8 days after inoculation. Collected samples were surfacesterilized with 1.2% (v/v) NaOCl for 5 min followed by 70% ethanol treatment for 5 min. Finally, they were washed with sterile distilled water several times. The sterilized samples were homogenized in a buffer solution (10 mM MgSO4) using a sterile pestle and mortal. The aliquots were appropriately diluted before plating onto 1/10 TSB agar media supplemented with chloramphenicol (40 µgml−<sup>1</sup> ) (Hossain et al., 2016). Plates were incubated at 28◦C for 48 h. The CFU/g of fresh roots was counted.

### Measurement of Chlorophyll Content

To measure chlorophyll content, bacterial suspension of B. velezensis YC7010 and 40 aphids were used to treat 4 weeks old Arabidopsis plants using the protocols described earlier. Leaves were then collected at 6 days after aphid infestation of treated and untreated plants. Sample tissue (10 mg) was crushed with 1 ml of 80% (v/v) acetone in a glass grinder followed by centrifugation at 13,000 × g for 5 min. The absorbance of the supernatant was measured at wavelengths of 646.8 and 663.2 nm using a spectrophotometer. Total chlorophyll content was calculated using the following formula (Lichtenthaler, 1987): Total chlorophyll (mg/fresh weight or fw) = (7.15 <sup>∗</sup> A663.<sup>2</sup> + 18.71 <sup>∗</sup> A646.8)/1000/fw of leaves.

### Quantitative RT-PCR

Bacterial suspension of B. velezensis YC7010 and 40 aphids were used to treat 4 weeks old Arabidopsis plants as described previously. Leaves were collected at 0, 12, 24, 48, and 72 h after aphid infestation of treated and untreated plants. Collected samples were frozen and ground in liquid nitrogen to a fine powder. Total RNA was extracted by using RNA extraction kit (Qiagen RNeasy Plant Mini Kit) and used to synthesis of complementary DNA using QuantiTect Reverse Transcription Kit according to the manufacturer's instruction. Using SYBR Green Master Mix (Bio-Rad), quantitative reverse transcription RT-PCR reactions were performed according to the manufacturer's protocol. Primers for candidate genes were designed using Primer3 software. Primer sequences are provided in Supplementary Table S1. To quantitatively determine the accumulation levels of genes, 2−11Ct method (Livak and Schmittgen, 2001) was used. Expression of genes was normalized against Arabidopsis UBIQUITIN10 (AT4G05320) as an internal reference. All experiments were repeated three times and each real time PCR sample was run in triplicates.

### Statistical Analysis

All data were subjected to analysis of variance (ANOVA). Mean differences were estimated by Tukey's honestly significant difference (HSD) using statistical software SPSS 17 (SPSS Inc., Chicago, IL, USA) and Sigma plot (version 12). Statistical significance was considered when P value was less than 0.05.

### RESULTS

### B. velezensis YC7010 Induces Systemic Resistance to Aphids in Arabidopsis

To determine whether B. velezensis YC7010 could induce resistance against aphids, both no-choice and choice tests were performed after drenching bacterial suspension to the rhizosphere of Arabidopsis. Such bacterial treatment significantly (P < 0.01) reduced the number of aphids than untreated control (**Figures 1A,B**). In the no-choice test, the number of aphids on bacterial treated plants was 20.75 per plant, which was significantly (P < 0.01) less than 35.50 on untreated plants. In the choice test, the number of aphids on bacterial treated plants was 11.38 per plant, which was also significantly (P < 0.01) lower than 22.00 on untreated plants. In the no-choice test and the choice test, the numbers of aphids on treated plants were reduced by 41.55 and 48.27%, respectively, by bacterial treatment when compared to those without bacterial treatment. In agreement with these results, aphids on bacterial treated plants showed significantly less excretion of honeydew, indicating less food intake at all time points at 3, 4 or 5 days after aphid infestation (**Figures 1C,D**).

### B. velezensis YC7010-Induced Systemic Resistance Is Dependent on Hydrogen Peroxide Production, Cell Death, and Callose Deposition in Arabidopsis

The effect of bacterial treatment on the accumulation of H2O2, cell death and callose deposition was observed at 6 days after aphid infestation. The accumulation of H2O<sup>2</sup> and the level of H2O<sup>2</sup> production at 12, 24, 48, and 72 h after infestation (HAI) of aphids in bacterial inoculated plants were significantly higher than those in untreated plants. The highest content of H2O<sup>2</sup> was 8.12 µmol g−<sup>1</sup> fw at 24 HAI in bacterial treated leaves. It was decreased to 6.00 µmol g−<sup>1</sup> fw at 48 HAI. However, this was still significantly higher than that in untreated plants after aphid infestation (**Figures 2A,B**). More cell death and callose deposition were detected in bacterial treated Arabidopsis plants comparing to those of untreated plants after aphid infestation (**Figures 2C–E**). A combination of bacterial treatment and aphid infestation led to significantly more H2O<sup>2</sup> production, cell death and callose deposition, which might have contributed to effective defense response against aphids.

### B. velezensis YC7010-Induced Systemic Resistance Is Independent of SA, JA, ET, or ABA in Arabidopsis

To evaluate whether hormones such as SA, JA, ET, and ABA might play a role in B. velezensis YC7010 induced resistance to aphids, Arabidopsis plants such as wild ecotype Col-0, NahG, sid2,

jar1, ein2-1 and abi2-2 were used in choice and no-choice tests. In both no-choice and choice tests, root drenching with bacterial suspension resulted in significant (P < 0.05) reduction in the number of aphids without significant difference among the Col-0, NahG and mutants (sid2, jar1, ein2-1 or abi2-2) (**Figures 3A,B**). These results showed that B. velezensis YC7010 could confer resistance to aphids in all treated Arabidopsis plants regardless of mutation. Accumulation of H2O<sup>2</sup> and cell death were also observed in all bacterial treated Col-0, NahG, and mutants. Both H2O<sup>2</sup> accumulation and cell death were found in bacterial treated Col-0, NahG, sid2, jar1, ein2-1, or abi2-2 plants. However, H2O<sup>2</sup> accumulation and cell death were not detected in untreated control plants at 6 days after aphid infestation (**Figures 3C,D**). Collectively, these results indicate that the reduction in the number of aphids, H2O<sup>2</sup> accumulation and cell death induced by B. velezensis YC7010 is not dependent on SA, JA, ET, or ABA in Arabidopsis.

### B. velezensis YC7010-Induced Aphid Resistance Is Dependent on Interactions among PAD4, BIK1, and SAG13

To investigate whether B. velezensis YC7010-induced aphid resistance was dependent on PAD4 associated with BIK1, we examined aphid performance on wild type Col-0 and mutants (pad4, bik1, and bik1pad4) of Arabidopsis treated with B. velezensis YC7010 (**Figure 4**). In both no-choice and choice tests, no significant difference in the number of aphids between bacterial treated and untreated all mutants was found. However, significantly less number of aphids was found on bacteria treated Col-0 Arabidopsis than that on untreated control Arabidopsis (**Figures 4A,B**). On the other hand, the number of aphids on bacterial treated or untreated pad4 and bik1pad4 mutants was higher than that on bik1 (**Figure 4A**). Interestingly, in the no-choice test, aphid growth was suppressed in both treated

and untreated bik1 mutants. However, in the choice test, the number of aphids was not significantly different on these mutants regardless of bacterial treatment (**Figure 4B**). Accumulation of H2O2, cell death and callose were observed on bacterial inoculated Col-0, bik1 and untreated bik1 mutants after aphid infestation. However, none of these responses was detected on bacterial treated or untreated control pad4 or bik1pad4 mutants even with aphid infestation (**Figures 4C–E**). These results suggest that PAD4 is required for H2O<sup>2</sup> accumulation, cell death and callose deposition in Arabidopsis. More H2O<sup>2</sup> accumulation, cell death and callose deposition were found on bacterial treated Col-0 and bik1 plants treated with or without B. velezensis YC7010. Therefore, we investigated the expression patterns of PAD4 and BIK1 genes in Col-0 plants. At every time point, the expression level of PAD4 was higher in B. velezensis YC7010 treated plants than that in untreated plants (**Figure 5A**). The highest expression level of PAD4 was found in bacterial treated plants at 48 HAI of aphids. On the contrary, the expression

by one-way ANOVA. Means with different letters were significantly different (P < 0.05).

level of BIK1 was higher in untreated plant compared to that in bacterial treated plants with aphid infestation (**Figure 5B**). As PAD4 can stimulate premature leaf senescence in aphids-infested Arabidopsis plants, we investigated whether the expression level of senescence associated SAG13 gene was affected by B. velezensis YC7010 in Col-0 plants. The expression level of SAG13 was higher in bacterial treated plants than in untreated plants at all time points after aphid infestation (**Figure 5C**). The highest expression was found in bacterial treated plants at 72 HAI of aphids. Lower chlorophyll content was found in most bacterial treated plants (except pad4, bik1, and bik1pad4) compared to that in untreated control Col-0 plants (Supplementary Figure S1). The contents of chlorophyll in the leaves of pad4, bik1, and bik1pad4 mutants in treated and untreated plants were similar to each other. The chlorophyll content in bik1 mutant was lower than that in pad4 or bik1pad4 mutant (treated or untreated), although the difference was not statistically significant. Bacterial count in the roots of bik1 mutant was higher than that of Col-0 (**Figure 6**),

number of aphids on the leaves was counted at 7 days later after bacterial inoculation. (B) Choice test on different genotypes. The number of settled aphids was counted at 24 h after releasing 35 adults between two plants of the treated plants and control plants. (C) Representative leaf images of 3,3'-diaminobenzidine staining (H2O<sup>2</sup> indicator). (D) Trypan blue staining (cell death indicator). Untreated control (top) or treated plants (bottom) at 6 days after aphids infestation. Data were analyzed by one-way ANOVA. Means with different letters were significantly different (P < 0.05).

indicating that root colonization of B. velezensis YC7010 was suppressed by BIK1 in Arabidopsis plants.

### DISCUSSION

Arabidopsis roots colonized by certain beneficial PGPR can activate ISR response that is effective against a broad range of plant pathogens (Van Oosten et al., 2008; Van Wees et al., 2008; Van der Ent et al., 2009). Likewise, some nonpathogenic rhizobacteria or endophytic bacteria commonly found inside roots also can enhance plant resistance against insect pests (Pineda et al., 2010; de Oliveira Araujo, 2015). These bacteria with ability to enhance plant growth and development have the potential to be utilized for biological control of numerous insect pests. We have previously identified an endophytic B. velezensis YC7010 with anti-microbial, plant growth-promoting and systemic resistance-inducing activities

(Chung et al., 2015). In this study, our results demonstrated that root drenching of Arabidopsis with B. velezensis YC7010 suspension resulted in the establishment of an ISR against GPA, regardless of test methods (choice or no-choice tests, **Figure 1**). Up to date, few reports have been published on tritrophic interactions among bacteria, insects and host plants. Reports on ISR by endophytic bacteria is especially limited. Van de Mortel et al. (2012) have reported that colonization of Arabidopsis roots by non-pathogenic rhizobacteria can induce resistance against lepidopteran insect herbivore Spodoptera exigua, in agreement with our results. Less honeydew excretion by GPA also indicates less food intake from bacterial treated plants.

Inoculation of B. velezensis YC7010 enhanced H2O<sup>2</sup> production, cell death, and callose deposition with aphid infestation in Arabidopsis (**Figure 2**). This indicates that ISR by this bacterial strain might be due to cellular responses, resulting in early cell death and callose deposition as HR. Insect feeding induced oxidative stress is an important component of plant defense to attacking insects. ROS detoxification may decrease antioxidant levels but increase toxic oxidation products in corn earworm infested soybean plants (Bi and Felton, 1995). Increased level of H2O<sup>2</sup> and other oxidative products of ROS in plants can directly damage the midgut of insects and inhibit their growth. High mortality of insects by consumption of artificial diets containing H2O<sup>2</sup> also supports the effect of ROS on the suppression of GPA (Liu et al., 2010). Another study has shown that higher level of H2O<sup>2</sup> accumulation in rice (Oryza sativa) can enhance the resistance against phloem sap sucking BPH (Nilaparvata lugens) (Zhou et al., 2009). In addition, cellular accumulation of H2O<sup>2</sup> can lead to plant cell death which may act as HR to GPA (Hoeberichts and Woltering, 2003). Cell death is considered as a plant defense factor against aphid by manipulating host nutritional quality in plant–microbe interactions (Goggin, 2007). Callose deposition is also an important defense mechanism that prevents insects from ingesting phloem sap (Hao et al., 2008). Microbes-mediated ISR is often associated with accumulation of H2O2, cell death, and deposition of callose in plant–pathogen interactions (Conrath et al., 2002; Jacobs et al., 2011). However, no such report has been published on three-way interaction of rhizobacteria (B. velezensis

YC7010), host plants and aphids through ISR associated with these defense responses. It is possible that H2O2-induced cell death in infested and adjacent cells could limit photoassimilates flow to the feeding sites, which can move aphids away from their feeding sites. No discrete accumulation of H2O<sup>2</sup> and callose deposition was detected in bacterial treated or untreated leaves before aphid infestation. However, accumulation of H2O<sup>2</sup> and

callose deposition were observed in leaves infested with GPA 6 days later. On the other hand, more H2O<sup>2</sup> accumulation was found only in bacterial treated plants with GPA infestation at different time points (**Figure 2**). Similarly, accumulation of H2O<sup>2</sup> and callose deposition was observed in Arabidopsis treated with rhizobacteria or inoculated by pathogens (Ahn et al., 2007). Primed plants show faster and/or stronger activation of cellular defenses when subsequently challenged by pathogen or insect attack, resulting in enhanced level of resistance (Conrath, 2011). These results suggest that B. velezensis YC7010 can induce priming responses against GPA in Arabidopsis, which is partially similar to results of previous studies, supporting the perception that ISR by beneficial microbes is commonly based on defense priming (Pieterse et al., 2014).

Our results showed that B. velezensis YC7010-mediated ISR to aphids were not dependent on SA, JA, ET, or ABA that were not associated with H2O<sup>2</sup> accumulation or cell death in bacterial treated Arabidopsis plants (**Figure 3**). In other previous studies on aphid defense mechanisms, SA is not essential, while JA, ET, and ABA are not mainly involved in defense of Arabidopsis plants against aphid either (Pegadaraju et al., 2005; Lei et al., 2014). On the contrary, PAD4 was found to be required for induction of B. velezensis YC7010-mediated resistance to aphid in this study. In addition, root drenching with B. velezensis YC7010 enhanced the expression of PAD4 in Arabidopsis plants after aphid infestation (**Figures 4** and **5A**). Aphid feeding is well known to induce the expression of PAD4 which is required for cell death-mediated resistance. It has been reported that transgenic plants with overexpression of PAD4 can enhance their resistance against GPA more than wild type Arabidopsis Col-0 (Pegadaraju et al., 2005, 2007). In this study, the loss of BIK1 function promoted the induced resistance against aphid in the no-choice test without bacterial treatment. HRs such as H2O<sup>2</sup> accumulation and cell death were observed in both bacterial

treated and untreated plants. However, they were absent in bik1pad4 mutant, suggesting that BIK1 might not directly repress, but indirectly modulate cell death pathway through PAD4. Root drenching with B. velezensis YC7010 suppressed the expression of MTI related gene BIK1 (**Figure 5B**). However, the number of B. velezensis YC7010 in the root was higher in bik1 mutant than that in wild type Col-0, indicating that suppression of BIK1 might contribute to stable colonization of B. velezensis YC7010 (**Figure 6**). Similar to our results, it has been reported that root inoculation with Bacillus cereus AR156 can actively block immune responses in Arabidopsis roots in order to establish a compatible interaction with the host which is important for root colonization by bacteria (Niu et al., 2011). Lei et al. (2014) have also shown that PAD4 expression is much higher in bik1 mutant to suppress aphids. These results collectively demonstrate that root drenching of B. velezensis YC7010 can effectively suppress the growth of aphids in leaves via expression of PAD4 which depends on the suppression of BIK1 in Arabidopsis.

The colonization of Arabidopsis roots by B. velezensis YC7010 resulted in enhanced expression level of SAG13 as well as chlorophyll loss in leaves upon aphid infestation (**Figure 5C**; Supplementary Figure S1). Senescence-associated processes have negative effect on aphid growth. For example, a gall aphid induced premature senescence in Pistacia palaestina trees has been shown to be correlated with reduced performance of another aphid feeding on the same leaflet (Inbar et al., 1995). However, PAD4 stimulates the premature senescence of leaves which can confer resistance to aphids (Pegadaraju et al., 2005). These results suggested that root drenching of B. velezensis YC7010 can elevate the expression of PAD4 and activate premature leaf senescence which is involved in resistance to aphid.

In summary, B. velezensis YC7010 root treatment could induce primed systemic resistance against GPA in Arabidopsis. Root colonization of Arabidopsis by B. velezensis YC7010 suppressed the expression level of BIK1, resulting in higher expression level of PAD4 and SAG13 in bacterial treated plants. Enhanced expression of PAD4 triggered more rapid H2O<sup>2</sup> accumulation, cell death, and callose deposition in bacterial treated Arabidopsis than in untreated plants after aphid infestation. To the best of

### REFERENCES


our knowledge, this is the first report on the mechanism of ISR against aphid by an endophytic PGPR. In this aspect, the results of this study will be helpful for developing environmental friendly management strategies for insect pests using beneficial endophytic bacteria.

### AUTHOR CONTRIBUTIONS

This study was designed by YC, MR, AK, and MTH. All experiments in this study were performed by MR. Data was analyzed by MR, AK, and MTH. The manuscript was written by YR and MR.

### FUNDING

This work was carried out with the support of "Cooperative Research Program for Agriculture Science and Technology Development (PJ01104901)" funded by Rural Development Administration, Republic of Korea. MR, AK, and MTH were supported by a scholarship from the BK21 Plus Program, the Ministry of Education, Republic of Korea.

### ACKNOWLEDGMENTS

We thank Jian-Min Zhou (Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China) and Tesfaye Mengiste (Department of Botany and Plant Pathology, Purdue University, USA) for providing Arabidopsis mutants bik1 and bik1pad4. We also thank Monica Höfte (Department of Crop Protection, Ghent University, Belgium) for giving good suggestions and comments on the manuscript.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.00211/ full#supplementary-material



green peach aphids via PHYTOALEXIN DEFICIENT4. Plant Physiol. 165, 1657–1670. doi: 10.1104/pp.114.242206



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Rashid, Khan, Hossain and Chung. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Mechanisms and Characterization of *Trichoderma longibrachiatum* T6 in Suppressing Nematodes (*Heterodera avenae*) in Wheat

Shuwu Zhang<sup>1</sup> , Yantai Gan2, 3, Weihong Ji <sup>4</sup> , Bingliang Xu<sup>1</sup> \*, Baohong Hou<sup>1</sup> and Jia Liu<sup>1</sup>

<sup>1</sup> Biocontrol Engineering Laboratory of Crop Diseases and Pests of Gansu Province, College of Plant Protection, Gansu Agricultural University, Lanzhou, China, <sup>2</sup> Gansu Provincial Key Laboratory of Arid Land Crop Science, Gansu Agricultural University, Lanzhou, China, <sup>3</sup> Swift Current Research & Development Centre, Agriculture and Agri-Food Canada, Swift Current, SK, Canada, <sup>4</sup> Human-Wildlife Interactions Research Group, Institute of Mathematical and Natural Sciences, Massey University, Auckland, New Zealand

#### *Edited by:*

Gero Benckiser, Justus Liebig Universität Giessen, Germany

#### *Reviewed by:*

Lei Zhang, Washington State University, United States Ömür Baysal, Mugla University, Turkey ˘

Andrew Bent contributed to the review of Lei Zhang

> *\*Correspondence:* Bingliang Xu xubl@gsau.edu.cn

#### *Specialty section:*

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

*Received:* 31 August 2016 *Accepted:* 11 August 2017 *Published:* 15 September 2017

#### *Citation:*

Zhang S, Gan Y, Ji W, Xu B, Hou B and Liu J (2017) Mechanisms and Characterization of Trichoderma longibrachiatum T6 in Suppressing Nematodes (Heterodera avenae) in Wheat. Front. Plant Sci. 8:1491. doi: 10.3389/fpls.2017.01491 Heterodera avenae is an important soil-borne pathogen that affects field crops worldwide. Chemical nematicides can be used to control the nematode, but they bring toxicity to the environment and human. Trichoderma longibrachiatum has been shown to have the ability to control H. avenae cysts, but detailed microscopic observations and bioassays are lacking. In this study, we used microscopic observations and bioassays to study the effect of T. longibrachiatum T6 (TL6) on the eggs and second stage juveniles (J2s) of H. avenae, and investigate the role of TL6 in inducing the resistance to H. avenae in wheat seedling at physiological and biochemical levels. Microscopic observations recorded that TL6 parasitized on the H. avenae eggs, germinated, and produced a large number of hyphae on the eggs surface at the initial stage, thereafter, the eggs were completely surrounded by dense mycelia and the contents of eggs were lysed at the late stage. Meanwhile, the conidia suspension of TL6 parasitized on the surface of J2s, produced a large number of hyphae that penetrated the cuticle and caused deformation of the nematodes. TL6 at the concentration of 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> had the highest rates of parasitism on eggs and J2s, reflected by the highest hatching-inhibition of eggs and the mortality of J2s. In the greenhouse experiments, wheat seedlings treated with TL6 at 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> had reduced H. avenae infection, and increased plant growth significantly compared to the control. The cysts and juveniles in soil were reduced by 89.8 and 92.7%, the juveniles and females in roots were reduced by 88.3 and 91.3%, whereas the activity of chitinase and β-1, 3-glucanase, total flavonoids and lignin contents in wheat roots were increased significantly at different stage after inoculation with the eggs and TL6 conidia in comparison to the control. Maximum activity of chitinase and β-1, 3-glucanase were recorded at the 20th and 15th Days after inoculation with TL6 and thereafter it declined. The maximum contents of total flavonoids and lignin were recorded at the 35th and 40th Days after inoculation with TL6. After being stained with the rapid vital dyes of acridine orange (AO) and neutral red (NR), the frozen and infected eggs and J2s of H. avenae changed color to orange and red, respectively, while the color of eggs and J2s in control group did not change. Therefore, our results suggest that TL6 is potentially an effective bio-control agent for H. avenae. The possible mechanisms by which TL6 suppresses H. avenae infection are due to the direct parasitic and lethal effect of TL6 on the eggs and J2s activity, and the induced defense response in wheat plants together.

Keywords: *Trichoderma* spp., *Heterodera avenae*, eggs and second stage juveniles, parasitic and lethal effects, biological control, antioxidative defense system

### INTRODUCTION

Plant parasitic nematodes are one of the most important pathogens causing plant diseases, affecting the growth, yield and quality of crops, which results in economic losses (Bird and Kaloshian, 2003; Wei et al., 2014). It is estimated that the annual cost due to parasitic nematodes is about US \$157 billion in crop production worldwide (Abad et al., 2008). Among pathogenic nematodes, Heterodera glycines, Heterodera avenae, Heterodera schachtii, Globodera rostochiensis, Globodera pallida, and Meloidogyne spp. are most economically important (Barker and Noe, 1987; Nicol and Rivoal, 2008). H. avenae causes"Molya" disease in wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.). In India, H. avenae infection causes yield loss in wheat by 47.2% and in barley by 87.2% (Rivoal and Cook, 1993). In China, with the increase of wheat growing areas, the impact of H. avenae has become a serious concern and innovative management measures are required to combat the disease (Nicol et al., 2007; Li et al., 2010). Since first reported in Hubei in 1989, H. avenae is now widely distributed and has been recorded in Hebei, Henan, Beijing, Shanxi and other provinces. It is estimated that more than one million hectares of wheat is affected each year, with the yield loss up to 50% (Peng et al., 2009; Riley et al., 2010; Yuan et al., 2010). In cereal crops, H. avenae can feed on the crop roots and cause root wound, providing opportunity for Rhizoctonia solani infection and seedling rot (Peng et al., 2009).

Chemical and biological nematicides are currently used to control nematode infection in agricultural crops (El-Alfy and Schlenk, 2002; Huang et al., 2014). For example, the nematicides Carbofuran, Etho-prophos, and Miral have about 90% efficacy in controlling of H. avenae, but these nematicides are toxic to the beneficial microorganisms in soil (Sergio, 2011). Moreover, some of the nematicides are difficult to degrade in soil which can cause pollution of underground water and the environment (Jatala, 1986). Some synthetic nematicidal agents are not species specific and can produce toxins that kill other symbiotic organisms in rhizosphere (Sergio, 2011). Because of the limitations of chemical control, biological control has been considered as an alternative. A great number of potential biological control microorganisms have been isolated, including fungi (predatory fungus, wireworm parasitic fungus, egg parasitic fungus, poisonous fungus and mycorrhizal fungus) (Crump et al., 1983; Sergio, 2011), bacteria (Chen and Dickson, 1996), and protozoon (Becker et al., 1988). Some biological control microorganisms such as Paecilomyces lilacinus and Pochonia chlamydosporia have been tested under field conditions (Sun et al., 2006). However, the effectiveness of these microbial agents is limited, as the microorganism P. lilacinus usually parasitizes nematode eggs and the rate of infection is related to the duration of the infection (Leij De et al., 1992; Bonants et al., 1995). Pochonia chlamydosporia usually infects eggs and females and the ability to colonize plant rhizosphere differs with plant species (Oclarit and Cumagun, 2009), thus, it is ineffective in parasitizing the eggs of root nematodes (Sun et al., 2006). It is imperative to identify and develop effective bio-control agents for preventing and controlling plant parasitic nematodes.

Trichoderma spp. is one of the important groups of rhizosphere fungi widely distributed in soil. Some species, such as T. hatzianum and T. viride have been proven to be good antagonists against some soil-borne plant pathogens such as Rhizoctonia spp., Sclerotium spp., Fusarium spp., and Pythium spp. (Bourne et al., 1996; Samuels, 1996; Deng et al., 2007). Trichoderma formulations have been commercialized in some countries, i.e., T. harzianum T22 (Topshield) in the United States (Louzada et al., 2009) and the strain of T. harzianum T39 (Trichodex) in Israel (Elad, 2000). In our previous studies, we found that the strain of T. longibrachiatum T6 (TL6) has a remarkable effect on alleviating the adverse effects of abiotic stress on wheat seedling growth and development (Zhang et al., 2016). Further, we found the fermentation broth of T. longibrachiatum has a great potential to be used as a biocontrol agent against H. avenae cysts (Zhang et al., 2014a,b). The spore suspension of T. longibrachiatum was found highly effective against M. incognita (Zhang et al., 2015). However, the process of the conidia suspension of TL6 in controlling H. avenae eggs and second stage juveniles (J2s) has not been reported. Little information is available in regard to the effectiveness of TL6 in controlling nematodes, especially on the parasitic and inhibitory effects on the eggs and J2s of H. avenae. In the present study, we (i) assessed the infection process of TL6 on eggs and J2s of H. avenae and (ii) determined the mechanisms and the effectiveness of the conidia suspension of TL6 in the control of H. avenae by greenhouse experiments and in vitro tests at physiological and biochemical levels.

### MATERIALS AND METHODS

Experiments were carried out at the Laboratory of Plant Pathology, College of Plant Protection, Gansu Agricultural University. The soil samples were collected from a wheat field in Xingyang, Henan province, China. H. avenae cysts were isolated from the soil and obtained using the "Flotation separation"

**Abbreviations:** AO, Acridine orange; FO, Fusarium oxysporum; J2s, Second stage juveniles; N-AcG, N-acetyl glucosamine; NR, Neutral red; PDA, Potato dextrose agar; TL6, Trichoderma longibrachiatum T6.

method, and sterilized with 1% NaOCl for 1 min, and then were gently washed six times with sterile water to remove NaOCl (Long et al., 2012). The cysts were then crushed to obtain eggs using a tissue grinder. The surface of eggs was sterilized with 1% NaOCl for 30 s and washed six times with sterile water. The final concentration of the eggs was prepared to 2 ± 1 per 10µl and 100 ± 5 per 50µl of sterile water. For obtaining the fresh hatched J2s, the sterilized eggs were collected on the size of 25µm mesh sieve and transferred to the Petri dishes containing sterilized tap water. The hatched fresh J2s of H. avenae were collected every day by incubating eggs in extraction Petri dishes at 20◦C for 14 Days, and stored in sterilized tap water at 4◦C. The concentration of the J2s for inoculation was then prepared to 2 ± 1 per 10µl and 100 ± 5 per 50µl of sterile water.

### Fungal Inoculum Preparation

Trichoderma longibrachiatum T6 (TL6) was isolated from a rhizisphere soil of a forest site nearby Tianshui, Gansu. Some basic tests were conducted at the Plant Pathology Laboratory of Gansu Agricultural University and it was reported that the strain has no hazardous effects to the environments (Zhuang et al., 2006). The strain of TL6 has since been collected at the China General Microbiological Culture Collection Center, in Beijing, with the patent number (CGMCC No.13183). For the present experiments, the strain of TL6 was cultured on potato dextrose agar (PDA) in Petri dishes for 6 Days at 25◦C. The conidia suspension of TL6 was prepared according to the methods described by Zhang et al. (2014b). Final conidia suspension of different densities, 1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> of TL6 conidia per ml, were prepared and stored at 4◦C.

Fusarium oxysporum (FO) was obtained from the Plant Pathology Laboratory, Gansu Agricultural University and cultured on potato dextrose agar (PDA) medium at 25◦C for 7 Days, and filtered into sterilized beakers. The spore concentration was determined using a hemacytometer, and diluted with sterile water to a final concentration of 1.5 × 10<sup>7</sup> conidia per ml.

### Effects of Different Concentrations of TL6 on Egg Hatching of *H. avenae*

This experiment was replicated three times. The treatments included five different concentrations of TL6 (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) and the two controls with one treated with sterile water and the other treated with F. oxysporum suspension. For each treatment and each replicate, surface sterilized eggs of 100 ± 5 suspended in sterile water were placed in each Petri dish, and counted under a stereomicroscope. After confirmation of the number of eggs in each plate, 5 ml of TL6 suspension were added to the plates for each of the five TL6 concentration treatments, whereas 5 ml of sterile water or F. oxysporum suspension were added to the control plates. Hatching (number of J2s emerged from eggs) and parasitism were recorded on the 8th Day after the treatment. Percentages of inhibition and parasitism were calculated according to Gao et al. (1998).

$$\text{PPE (\%)} = \text{(NEPET/TNTE)} \times 100\tag{1}$$

Where PPE represents percentages of parasitism on eggs, NEPET number of eggs parasitized in each treatment and TNTE the total number of test eggs.

RPIEH (%) = (NJHSWCG − NJHET)/NJHSWCG × 100 (2)

Where RPIEH represents relative percentages of inhibition of eggs hatching, NJHSWCG number of J2s hatched in sterile water control group and NJHET number of J2s hatched in each treatment.

### Effects of Different Concentrations of TL6 on J2s of *H. avenae*

This experiment was replicated three times. The treatments included five different concentrations of TL6 and the two controls with one treated with sterile water and the other treated with F. oxysporum suspension. For each treatment and each replicate, surface sterilized J2s of 100 ± 5 suspended in sterile water were placed in each 6-well sterilized cell culture plate, and counted under a stereomicroscope. After confirmation of the number of J2s in each plate, 3 ml of TL6 suspension were added to each well for each of the five TL6 concentration treatments, whereas 3 ml of sterile water or F. oxysporum suspension were added to the control plates.

After 24, 48, and 72 h of treatment, the remaining J2s were prodded with a needle and those that did not respond were considered dead. The percentages of immotile nematodes were calculated (Meyer et al., 2004). Linear regression was used to evaluate the effect of different concentrations of TL6 on the corrected motility rate of J2s of H. avenae (Khan et al., 2006). Mortality and the corrected mortality of H. avenae J2s were calculated using the equations described by Zhang et al. (2015).

$$\text{M (\%) = (NDJET/TNTJET) \times 100\tag{3}$$

Where M represents mortality, NDJET number of dead J2s in each treatment and TNTJET the total number of test J2s in each treatment.

$$\text{CM} \text{(\%)} = \text{(MET} - \text{MSWCG)} / \text{(1} - \text{MSWCG)} \times 100 \quad \text{(4)}$$

Where CM represents corrected mortality, MET the mortality in each treatment and MSWCG the mortality in sterile water control group.

The parasitic effect of different concentrations of TL6 on the J2s of H. avenae was observed every 2 Days after the treatment from the 6th to 14th Day, and percentages of parasitism were calculated using the equation of Zhang et al. (2015).

$$\text{\{PPSSJ} (\%) = \text{\{NSSJPET/TNTSSJ\}} \times 100\tag{5}$$

Where PPSSJ represents percentages of parasitism on second stage juveniles, NSSJPET number of second stage juveniles parasitized in each treatment and TNTSSJ the total number of test second stage juveniles.

### Microscopic Observation of the Infection Process of TL6 on Eggs and J2s of *H. avenae*

This experiment was replicated six times with the seven treatments the same as those described above. For each treatment and each replicate, 10µl (2 ± 1 eggs) of surface sterilized eggs suspended in sterile water were placed in each sterilized Petri dish of 3.5 cm in diameter. After that the suspension of TL6 (990µl) was added to each sterilized Petri dish of 3.5 cm in diameter for each of the five TL6 concentration treatments, whereas sterile water or F. oxysporum suspension (990µl) was added to the two separate control plates. The process of infection was observed using a stereomicroscope. The daily observations started on the 2nd Day after the treatment until the 12th Day. In each day of the observation, the detailed information was recorded for the same eggs.

Following the same protocol of the observation on eggs, we observed the infection of TL6 on J2s for each of the seven treatments and each of the six replicates. Ten µl of suspension containing 2 or 3 J2s were pipetted into each sterilized Petri dish of 3.5 cm in diameter, and TL6 suspension (990µl) was then added to each culture plate for each of the five TL6 concentration treatments, whereas the same amount of sterile water or F. oxysporum suspension was used in the two controls, respectively. The daily observations started on the 2nd Day after the treatment until the 12th Day. In each day of the observation, the detailed information was recorded for the same J2s.

### A Rapid Method for Assaying the Viability of Nematodes

Two vital dyes were used to determine the viability of H. avenae eggs and J2sin vitro. Acridine orange (AO, Sigma A6014, US) and neutral red (NR, Solarbio-N8160, Amresco-E89) were dissolved in sterilized distilled water and the dyeing was added to the incubation solution to make (i) 0.01% concentration of AO and (ii) 0.01% concentration of NR. This experiment included two treatments and replicated six times. The first treatment was under ultra-low temperature freezer (−80◦C) for 15 min, and the second treatment was with the suspension (1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> ) of TL6. The detailed treatment procedures are as follow:

For the first treatment, the suspensions of surface sterilized eggs or J2s (2 ± 1 per 10µl) in 1.5 ml centrifuge tubes were placed in ultra-low temperature freezer (−80◦C) for 15 min, and then moved to a water bath (60◦C) for 1 min to thaw quickly. The surface sterilized eggs with the similar stage of growth or J2s were maintained at 4◦C as the control. Also, 10µl of eggs or J2s in each treatment were placed into each well of sterilized cell culture plate (96 well) in the treatment group, and 90µl of the dye solutions were added to each 96 well cell culture plate, and mixed thoroughly until the dyes fully diffused. The eggs were stained for 15 min whereas J2s were stained for 30 min before being rinsed with sterilized distilled water. The viability of eggs and J2s was identified using an optical microscope.

For the second treatment, the surface sterilized eggs or J2s (2 ± 1 per 10µl) were placed into each 96 well sterilized cell culture plate, and then the suspension (1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> ) of TL6 (90µl) was added to each plate in the treatment group and sterilized water (90µl) was added to the control. The eggs and J2s in the treatment and the control group were stained at Day 8 of incubation treatment. The dye solution of 100µl was added to each 96 well sterilized cell culture plate, and mixed thoroughly until the dyes fully diffused. The eggs were stained for 15 min whereas J2s were stained for 30 min before being rinsed with sterilized distilled water. The viability of eggs and J2s was identified using an optical microscope.

### Bio-control Experiment in Greenhouse

This experiment included seven treatments and replicated three times. The wheat cultivar Yongliang 4, susceptible to H. avenae, was used in the experiment. Wheat seeds were surface sterilized with 1% sodium hypochlorite (NaOCl, 5 min) and sown in 15 cm diameter pots containing 500 g sterile soil (silty clay: sand = 3:1 v/v). Ten seedlings per pot were grown in a greenhouse with air temperature of 25◦C ± 0.5, and supplemental day/night lighting of 16/8 h. Pots were irrigated daily with sterilized distilled water which enabled the relative humidity to be maintained around 65%. The experiment was arranged using a completely randomized design. Each replicate had 18 pots, allowing multiple sampling (described below). When seedlings reached 10 cm in height, about 15 Days after sowing, each pot was inoculated with 1,500 ± 100 eggs of H. avenae. Ten Days after the inoculation, seedlings were inoculated with 20 ml of TL6 conidia suspension. The treatments included the five concentrations (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) of TL6 conidia and the two controls (i) seedlings inoculated with eggs but not with TL6 (replaced with sterilized distilled water), and (ii) seedlings not inoculated with eggs or TL6 (Mockinoculated plants (normal)).

The fresh roots of wheat seedlings were sampled six times at 5 or 10 days intervals (i.e., 5, 10, 15, 20, 30, and 40 Days) after inoculation with eggs and TL6 conidia suspension. Chitinase and β-1, 3-glucanase activity, the contents of lignin and total flavonoids were assayed:

### Chitinase and β-1, 3-Glucanase Activity

For the determination of enzyme activity, the enzyme extracts of the fresh roots were prepared following the methods of El Ghaouth et al. (2003) with some modifications. Fresh roots of 1 g were homogenized in 5 ml of buffer that contained 50 mM sodium acetate (pH 5.0), and then the homogenate was filtered and centrifuged at 10,000 g for 30 min. The supernatant was retained and used as the enzyme extract to assay the chitinase and β-1, 3-glucanase activity. The experiments were repeated six times.

Chitinase and β-1, 3-glucanase activity were assayed according to the method of El-Katatny et al. (2001). For the activity of chitinase, 50 mM sodium acetate buffer (pH 5.0) and 0.2 ml of 0.5% colloidal chitin were added to 0.5 ml of the supernatant of the wheat seedling enzyme extracts. The activity of chitinase was measured by assaying the quantity of reducing sugar as described by Adney and Baker (2008). The absorbance of the reaction mixture was determined at 585 nm using the spectrophotometer (UV), and the quantities of reducing sugar were calculated from a calibration curve with N-acetyl glucosamine (N-AcG) as the standard. The activity of chitinase was expressed as nmol N-AcG mg−<sup>1</sup> protein min−<sup>1</sup> .

Activity of β-1, 3-glucanase (expressed as nmol glucose mg−<sup>1</sup> protein min−<sup>1</sup> ) was determined and performed by using 5% laminarin as a substrate. The reaction mixture was mixed with 0.5 ml of extract, 50 mM sodium acetate buffer (pH 5.0) and 0.2 ml of 5% laminarin, and thereafter was incubated in a water bath at 37◦C for 20 min. The activity of β-1, 3-glucanase was determined by assaying the quantity of reducing sugar in the reaction mixture. The absorbance was read at 540 nm by a colorimetric assay, and a calibration curve was prepared by using glucose as the standard (El-Katatny et al., 2001).

### Lignin and Total Flavonoids

The contents of lignin in wheat seedling roots were measured following the method of Lee et al. (2007), with minor modifications. The dry weight of 0.5 g root samples were homogenized with distilled water and 95% ethanol twice to remove the metabolites of soluble sugar. Thereafter, the supernatant of extraction was discarded after being centrifuged at 10,000 g for 5 min, and the insoluble residue was left to dry at 45◦C overnight. The dried samples were washed with 1 ml of acetyl bromide with acetic acid (1:3, v/v), and then left at 70◦C for 30 min. After that, the supernatant was discarded and the precipitate was re-suspended in 0.36 ml of NaOH (2 M) and 0.04 ml of hydroxylamine hydrochloride (7.5 M), and then acetic acid was added to the mixture to make up the final volume to 10 ml when it cooled down to room temperature. The absorbance of the supernatant was measured at 280 nm after the final reaction mixture was centrifuged at 1,000 g for 5 min (Lin and Kao, 2001), and the contents of lignin were calculated according to the linear calibration curve with lignin as the standard (Lee et al., 2007). All treatments were repeated six times.

The reaction mixtures of wheat seedling fresh roots were extracted following the procedure of Hertog et al. (1992). The contents of total flavonoids were measured as gallic acid using Folin Ciacalteau reagent (Ragazz and Veronese, 1973) with a modification. Fresh roots samples of 1 g were crushed to 5 ml of 1% hydrochloric acid-methanol for 24 h, and then the extracts were diluted to 25% of original concentration. After that, the diluted extracts (0.2 ml) were mixed with 0.5 ml of Folin Ciacalteau reagent and 5 ml of distilled water. After a 5 min reaction, the reaction mixture was neutralized with 3 ml of 20% Na2CO3, vortexed and left for 30 min at room temperature. The absorbance was measured at 725 nm using a spectrophotometer. The total flavonoids were expressed as mg of gallic acid per g of wheat seedling fresh roots. The experiment was repeated six times.

### Growth Traits of Wheat Seedlings and Nematode Populations in the Rhizosphere

Sixty-five Days after sowing, plant height, root length, shoot and root fresh weights of wheat seedlings inoculated with eggs and the different concentrations of TL6 were measured. The number of cysts, females and juveniles were recorded and assessed in both soil and roots after inoculation with eggs and different concentrations of TL6. Heterodera avenae cysts in the soil were extracted from 200 g of soil samples per pot using "Flotation separation" method (Long et al., 2012), and H. avenae juveniles in the soil were extracted from 20 g of soil samples per pot using centrifugation technique (Castillo et al., 2006). Nematode root densities were assessed from 2 g of root sub-samples of each plant (Sharon et al., 2007).

### Statistical Analysis

In the study, replicated observations were made randomly and independently of each other and they were in a normal distribution with common variances, thus, the ANOVA assumption was generally met. Our treatments was essentially one factor only, therefore, one-way ANOVA was performed to determine the treatment effect using SPSS Version 16.0 (SPSS Inc., Chicago, IL). The significant differences between the treatments were considered at the level of P < 0.05. Fisher's least significant difference (LSD) values were computed using standard error and T-values of adjusted degrees of freedom. Linear regression was used to determine the relationship between the corrected mortality (%) and the corresponding values of the concentrations of the conidia suspension of TL6. The percentage values were log-transformed prior to statistical analysis.

### RESULTS

### Microscopic Observation of the Infection Process of TL6 on *H. avenae* Eggs

Overall, the different concentrations of conidia suspension of TL6 (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) had the different inhibitory and parasitic effects on the hatching of H. avenae eggs, with higher concentrations of TL6 presenting a stronger and more significant inhibitory and parasitic effects (**Figure 1**). At the concentration of 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> , a large number of conidia of TL6 adhered and surrounded the surface of eggs at the early stages (at Day 2) of infection (**Figure 1Ai**), and the conidia germinated and produced a large number of hyphae parasitized on the surface at Day 8 (**Figure 1Aii**). The content of eggs was dissolved by the metabolites of TL6 at Day 12 (**Figure 1Aiii**). At the concentration of 1.5 × 10<sup>6</sup> conidia ml−<sup>1</sup> , many conidia adhered to the surface of eggs at Day 2 (**Figure 1Bi**). With the increase of incubation time, the conidia parasitized and grew on the surface of eggs. Fewer conidia germinated hyphae and grew on the eggs at Day 8 (**Figure 1Bii**). The germinated hyphae penetrated into the eggs shell and dissolved the content and eggs shell at Day 12 (**Figure 1Biii**). As the TL6 concentration was reduced to 7.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> , only a few conidia adhered and parasitized on the surface of eggs at Day 2 of inoculation (**Figure 1Ci**). The hatched nematodes body and the eggs shell were parasitized with the hyphae at Day 8 (**Figure 1Cii**), and were dissolved at Day 12 (**Figure 1Ciii**). With the TL6 concentrations were reduced further to 3.0 × 10<sup>5</sup> and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> , only a few conidia surrounded the eggs at Day 2 after inoculation (**Figures 1Di,Ei**). However, the eggs were parasitized with the conidia and dense mycelium at 3.0 × 10<sup>5</sup> conidia ml−<sup>1</sup> at Day 8 (**Figure 1Dii**), while the dense mycelium

parasitized, penetrated and grew on the eggs at 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> at Day 8 (**Figure 1Eii**). The content and shell of eggs were completely dissolved at the concentration of 3.0 × 10<sup>5</sup> conidia ml−<sup>1</sup> of treatment at Day 12 (**Figure 1Diii**), whereas the parasitized dense mycelium was grown and penetrated into the eggs shell, and the contents of eggs were dissolved after inoculated with the concentration of 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> at Day 12 (**Figure 1Eiii**). In contrast, the strain of F. oxysporum in the control had a lower inhibitory and parasitic effect on the hatching of H. avenae eggs. Some conidia surrounded the eggs at Days 2 and 8 (**Figures 1Fi,ii**), but only a few of them geminated hyphae and grew on the surface of eggs at Day 12 (**Figure 1Fiii**). No inhibitory and parasitic effects were found in the sterile water control. No mycelium was formed and the embryonic development was normal in the sterile water control group (**Figures 1Gi,ii**) and also nematodes begun to hatch at Day 12 (**Figure 1Giii**).

### Effects of Different Concentrations of TL6 on *H. avenae* Eggs

With the increase of TL6 concentrations (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ), the inhibitory and parasitic effects on the H. avenae eggs hatching rates were increased, while the strain of F. oxysporum in the control had a lower inhibitory and parasitic effect, and the sterile water control had no inhibitory and parasitic effect (**Tables 1, 2**).

There was a trend that the number of eggs parasitized increased with the increase of TL6 concentrations, and the two

#### TABLE 1 | Effects of different concentrations of TL6 on parasitism of H. avenae eggs.


Data are means ± SE in Experiment. More than three hundreds surface sterilized eggs were tested in each treatment. Values in columns followed by different letters are significantly different at P < 0.05 based on Fisher's LSD test. FO-1.5 × 10<sup>7</sup> represents eggs treated with the spore suspension of F. oxysporum at the concentration of 1.5 × 10<sup>7</sup> spores ml−<sup>1</sup> . Control represents eggs treated with sterile water.

TABLE 2 | Effects of different concentrations of TL6 on relative percentages of hatching inhibition of H. avenae eggs.


Data are means ± SE in Experiment. More than three hundreds surface sterilized eggs were tested in each treatment. Values in columns followed by different letters are significantly different at P < 0.05 based on Fisher's LSD test. FO-1.5 × 10<sup>7</sup> represents eggs treated with the spore suspension of F. oxysporum at the concentration of 1.5 × 10<sup>7</sup> spores ml−<sup>1</sup> . Control represents eggs treated with sterile water.

highest concentrations of 1.5 × 10<sup>7</sup> and 1.5 × 10<sup>6</sup> conidia ml−<sup>1</sup> resulted in highest percent parasitism (P < 0.01; **Table 1**). This effect increased with the increase of the treatment days. In contrast, no fungi parasitized the eggs in the sterile water control. In vitro, with the increase of the TL6 concentrations, the relative percentages of inhibition of eggs hatching were increased (**Table 2**). The relative percentages of eggs hatching-inhibition were highest at Day 8 (100.0%) and Day 9 (99.2%) and the effect decreased with the treatment days (P < 0.01). Among the five concentrations of TL6, the highest percentages of egg hatching inhibition were obtained with the highest concentrations at 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> (P < 0.01).

### Microscopic Observation of the Infection Process of TL6 on J2s of *H. avenae*

The different concentrations of conidia suspension of TL6 (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) had the different parasitic and lethal effects on the J2s nematodes (**Figure 2**). Microscopic examination observed that when the newly hatched J2s were treated with the conidia suspension of TL6, most J2s died, and the conidia adhered at or parasitized on the surface at Day 2 at 1.5 × 10<sup>7</sup> (**Figure 2Ai**) and 1.5 × 10<sup>6</sup> conidia ml−<sup>1</sup> (**Figure 2Bi**), whereas the J2s were treated with lower concentrations (7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) became dull, stiff, and showed a wave-like distortion at Day 2, and even a small number of dead J2s were adhered or parasitized by the conidia of TL6 (**Figures 2Ci,Di,Ei**). At the concentration of 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> , a large number of conidia of TL6 surrounded or adhered to the surface of J2s at Day 8 (**Figure 2Aii**), and the surface of J2s was completely parasitized by the conidia of TL6 at Day 12 (**Figure 2Aiii**). At the concentration of 1.5 × 10<sup>6</sup> conidia ml−<sup>1</sup> , a few conidia germinated hyphae and parasitized on the dead nematodes surface at Day 8 (**Figure 2Bii**), and even the dead nematodes were dissolved by the parasitized hyphae and conidia of TL6 at Day 12 (**Figure 2Biii**). At the concentration of 7.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> , the dead nematodes were parasitized by the conidia and hyphae of TL6 at Day 8 (**Figure 2Cii**). The body of J2s was completely parasitized by the dense mycelium, and even some conidia started reproduction on the parasitic sites of J2s body. The J2s began to dissolve at Day 12 (**Figure 2Ciii**). As the TL6 concentration was reduced to 3.0 × 10<sup>5</sup> conidia ml−<sup>1</sup> , a few conidia parasitized on the J2s surface, and even germinated hyphae and penetrated into the J2s body at Day 8 (**Figure 2Dii**), while the parasitized site reproduced a few number of conidia and hyphae which grew on the surface at Day 12 (**Figure 2Diii**). At the concentration of 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> , the conidia germinated a small number of hyphae and parasitized on the J2s surface at Day 8 (**Figure 2Eii**). However, a large number of hyphae penetrated into the integument and the parasitic sites became shrunk by the dense mycelium at Day 12 (**Figure 2Eiii**). On the contrary, the strain of F. oxysporum had no significant inhibitory and parasitic effect on the J2s of H. avenae, with just a few spores surrounded the J2s at Days 2 and 8 (**Figures 2Fi,ii**), and only a few geminated hyphae parasitized on the surface of J2s at Day 12 (**Figure 2Fiii**). The activity, body′ s color and shape of J2s in the control group remained intact after inoculated with the sterile water (**Figures 2Gi,ii,iii**).

### Effects of Different Concentrations of TL6 on the J2s of *H. avenae*

With the increase of TL6 concentrations (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ), the percentages of parasitism on J2s of H. avenae increased significantly. The percentages of parasitism reached 88.7% after 14 Days of treatment with the concentration of TL6 at 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> (P < 0.01), while the strain of F. oxysporum in the control had lower parasitic effect, and the sterile water control had no parasitic effect. Also, with the increase of incubation days, the percentages of parasitism on J2s of H. avenae increased significantly after treated with different concentrations of TL6 (**Table 3**).

Different concentrations of TL6 showed a significant lethal effect on the activity of J2s, and the highest concentration of 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> resulted in highest mortality and corrected mortality (P < 0.01). Moreover, there was a significant linear relationship between TL6 and the J2s corrected mortality, and the increased concentrations of TL6 increased the rates of corrected mortality of J2s. The mortality or corrected mortality increased with the increase of the treatment time when treated with different concentrations of TL6. In contrast, the strain of F. oxysporum in the control had lower lethal effect on the activity of J2s at 1.5 × 10<sup>7</sup> spores ml−<sup>1</sup> , and the sterile water control had no significant lethal effect on the J2s mortality regardless of the TL6 concentration (**Table 4**).

### Viability of Nematodes

Compared to that of the controls, the color of frozen and parasitized nematodes was significantly different after stained with the dye of AO or NR, respectively. After stained for 15 min, the color of frozen eggs changed to orange (**Figures 3Aii,iii,iv**) or red (**Figures 3Avi,vii,viii**), while the color of eggs in control group did not change significantly after stained with AO (**Figure 3Ai**) or NR (**Figure 3Av**) for 15 min. Meanwhile, the eggs color changed to orange or red after infected with the conidia suspension of TL6 and stained with AO or NR for 15 min. Especially, the color of stained eggs changed significantly when infected with the conidia (**Figures 3Bii,vi**) or the mycelium (**Figures 3Biii,iv,vii,viii**) of TL6, but the color of eggs in control group did not change (**Figures 3Bi,v**).

The color of frozen J2s changed to orange (**Figures 4Aii,iii,iv**) or red (**Figures 4Avi,vii,viii**) after stained with AO or NR, but the color of stained J2s in control group did not change (**Figures 4Ai,v**). However, the color of stained J2s significantly changed after treated or parasitized with the conidia suspension of TL6. Especially, the color changed quickly and significantly when the conidia (**Figures 4Bii,vi**) or the mycelium (**Figures 4Biii,iv,vii,viii**) surrounded and parasitized on the surface of some J2s. In contrast, the color in the control did not change (**Figures 4Bi,v**).

#### TABLE 3 | Effects of different concentrations of TL6 on parasitism of H. avenae J2s in vitro.


Data are means ± SE in Experiment. More than three hundreds J2s were tested in each treatment. Values in columns followed by different letters are significantly different at P < 0.05 based on Fisher's LSD test. FO-1.5 × 10<sup>7</sup> represents J2s treated with the spore suspension of F. oxysporum at the concentration of 1.5 × 10<sup>7</sup> spores ml−<sup>1</sup> . Control represents J2s treated with sterile water.

TABLE 4 | Effects of different concentrations of TL6 on the activities of H. avenae J2s in vitro.


Data are means ± SE in Experiment. More than three hundreds J2s were tested in each treatment. Values in columns followed by different letters are significantly different at P < 0.05 based on Fisher's LSD test. FO-1.5 × 10<sup>7</sup> represents J2s treated with the spore suspension of F. oxysporum at the concentration of 1.5 × 10<sup>7</sup> spores ml−<sup>1</sup> . Control represents J2s treated with sterile water.

### The Symptoms of Wheat Seedling Inoculated with Eggs of *H. avenae* and Different Concentrations of TL6 in Greenhouse Experiments

Sixty-five Days after sowing, an average of 89.7% of wheat seedlings leaves showed etiolation, stunted and turned yellow after inoculated with the eggs of H. avenae, and even averagely 82.6% of leaves wilt readily (**Figure 5Ag**) compared with mockinoculated wheat seedlings (**Figure 5Aa**). However, the wheat seedlings inoculated with eggs of H. avenae grew normally when treated with high concentrations (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , and 3.0 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) of TL6 (**Figures 5Ab–e**). For low concentration of treatment (1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ), an average of 23.3% of wheat seedlings leaves showed the symptoms of slight etiolation (**Figure 5Af**) compared to the control (inoculated with eggs but not with TL6; **Figure 5Ag**).

Abnormal symptoms of roots included decreased number of roots, highly branched and hyperplasia short lateral roots,

FIGURE 3 | The morphological characteristics of the eggs of H. avenae were stained by different dyes in vitro. Where (A) the eggs were frozen in ultra-low temperature freezer (−80◦C); (i,v) represent the eggs were not frozen but stained by AO and NR, respectively; (ii–iv) represent the eggs were frozen and stained by AO, and (vi–viii) represent the eggs were frozen and stained by NR; (B) the eggs were treated with 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> of TL6; (i,v) represent the eggs were not treated with TL6 but stained by AO and NR, respectively; (ii–iv) represent the eggs were treated with TL6 and stained by AO, and (vi–viii) represent the eggs were treated with TL6 and stained by NR. Bars are in the unit of 100µm in all the cases. The arrows name is detailed in the footnote of Figure 1.

and disorder and entanglement of root system after inoculated with eggs of H. avenae (**Figure 5Bg**). In contrast, these were not observed in mock-inoculated plants (**Figure 5Ba**). Moreover, the eggs of H. avenae infected roots of wheat seedlings grew normally while treated with the high concentrations (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , and 7.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) of TL6 (**Figures 5Bb–d**). For low concentrations of treatments (3.0 × 10<sup>5</sup> and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ), an average of 26.5% of wheat seedlings roots appeared slightly decreased in each treatment (**Figures 5Be,f**), but the symptoms of roots have not appeared highly branched and hyperplasia compared to the control (inoculated with eggs of H. avenae but not with TL6; **Figure 5Bg**).

### Effect of Different Concentrations of TL6 on Wheat Seeding Growth and *H. avenae* Numbers in Potting Soil and Roots

Sixty-five Days after sowing, the plant height, root length and fresh weight of wheat seedlings decreased after inoculated with the eggs of H. avenae compared to mock-inoculated plants (P < 0.01), while application of different concentrations of TL6 (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) significantly increased the growth of wheat seedlings, and the effect was more pronounced with the increased concentrations of TL6, compared to the control. The relative growth rates of wheat seedling increased significantly when treated with TL6 at 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> , and the plant height, root length, shoot and root fresh weight increased by 66.7, 159.3 (**Figure 6A**), 169.1, and 170.2% (**Figure 6B**), respectively, compared to the control (P < 0.01). Meanwhile, compared to the mock-inoculated plants, treatments with the concentrations lower than 7.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> had no significant effect on the growth of wheat plants, but had a positive effect on their growth in comparison to the control (**Figure 6**).

Sixty-five Days after sowing, the number of nematodes in the soil and roots of wheat seedling decreased after the combined application of different concentrations of TL6 and the eggs of H. avenae. Among the five concentrations of TL6, the highest concentration of 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> significantly decreased the number of cysts (89.8%) and juveniles (92.7%) of H. avenae in soil (P < 0.01), and the number of juveniles (88.3%), and females (91.3%) in roots after the wheat seedlings inoculated

FIGURE 4 | The morphological characteristics of the J2s of H. avenae were stained by different dyes in vitro. Where (A) the J2s were frozen in ultra-low temperature freezer (−80◦C); (i,v) represent the J2s were not frozen but stained by AO and NR, respectively; (ii–iv) represent the J2s were frozen and stained by AO, and (vi–viii) represent the J2s were frozen and stained by NR; (B) the J2s were treated with 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> of TL6; (i,v) represent the J2s were not treated with TL6 but stained by AO and NR, respectively; (ii–iv) represent the J2s were treated with TL6 and stained by AO, and (vi–viii) represent the J2s were treated with TL6 and stained by NR. Bars are in the unit of 100 µm in all the cases. The arrows name is detailed in the footnote of Figure 1.

with eggs of H. avenae, compared to the control (P < 0.01) (**Table 5**).

### Determination of Chitinase and β-1, 3-Glucanase Activity

Compared to the mock-inoculated treatment, the activity of chitinase and β-1, 3-glucanase in wheat seedling roots were increased significantly after inoculation with the eggs of H. avenae (P < 0.01). However, a higher increase in chitinase and β-1, 3-glucanase activity was observed after the wheat seedlings treated with different concentrations of TL6 (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) and eggs of H. avenae. The activity of chitinase and β-1, 3-glucanase reached its maximum at the 20th and 15th Days after inoculation with TL6 and thereafter it declined. In the wheat seedlings roots, the maximum activity of chitinase and β-1, 3-glucanase increased from 0.4 to 43.9% (**Table 6**), and 6.1 to 32.0% (**Table 7**) for the wheat seedlings inoculated with the different concentrations of TL6 and eggs at the 20th and 15th Days, respectively, compared to the control. Specifically, the chitinase and β-1, 3-glucanase activity significantly increased by 43.9 and 32.0% in wheat seedlings roots after the combined inoculation with the eggs and highest concentrations of TL6 at 1.5 × 10<sup>7</sup> conidia ml−<sup>1</sup> , compared to the control (P < 0.01).

### Measurement of Total Flavonoids and Lignin Contents

Total flavonoids in the roots of wheat seedlings increased significantly as results of application of different concentrations of TL6 (1.5 × 10<sup>7</sup> , 1.5 × 10<sup>6</sup> , 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> ) and inoculation with eggs of H. avenae in comparison to the mock-inoculated plants. Specially, application of higher concentrations of TL6 presented higher total flavonoids in the roots of wheat seedlings, but the lowest concentration increased the total flavonoids contents slightly, compared to the control. The maximum content of total flavonoids in wheat seedling roots was recorded at the 30th Day after application of different concentrations of TL6 and inoculation with eggs of H. avenae (P < 0.01). Moreover, the total flavonoids in the roots of wheat seedlings increased after inoculation with the eggs of H. avenae alone, compared to that of mock-inoculated plants (**Table 8**).

the eggs of H. avenae in greenhouse experiments. Data are reported as means of replicates. Values in columns followed by different letters are significantly different at P < 0.05 based on Fisher's LSD test. Small bars represent the standard errors of the means. Normal (mock-inoculated) represents seedlings neither inoculated with eggs nor with TL6. Control represents seedlings inoculated with eggs but not with TL6.

The contents of lignin in wheat seedling roots increased after inoculated with the eggs of H. avenae alone, while application of different concentrations of TL6 significantly increased the lignin contents in wheat seedling roots, compared to the mockinoculated wheat seedlings (P < 0.01). There was a trend that the content of lignin in wheat seedling roots increased consistently with the increase of TL6 concentration or incubation days. The highest increase in the contents of lignin was achieved after application of TL6 at 1.5 × 10<sup>7</sup> and 1.5 × 10<sup>6</sup> conidia ml−<sup>1</sup> , followed by 7.5 × 10<sup>5</sup> , 3.0 × 10<sup>5</sup> , and 1.5 × 10<sup>5</sup> conidia ml−<sup>1</sup> , respectively. Compared to the control, the highest increase in the contents of lignin by 34.9% and 32.2% at the 40th Day after inoculated with eggs and TL6 at the concentrations of 1.5 × 10<sup>7</sup> and 1.5 × 10<sup>6</sup> conidia ml−<sup>1</sup> , respectively (**Table 9**).

## DISCUSSION

Previous studies report that the nematode of H. avenae, a root pathogen of cereals, is found in more than 31 wheat growing countries causing significant economic yield losses, particularly in those countries where rainfed cereal systems predominate (Nicol et al., 2003). Our previous studies have demonstrated that T. longibrachiatum is a well-known bio-control agent against several plant pathogens, which including the fungi and H. avenae cysts without hazardous effects to the environment (Zhang et al., 2014a,b), but little information is available regarding the process of the conidia suspension of TL6 infecting the same eggs or J2s at different time points, and also a rapid method to assay the viability of H. avenae in vitro. To our knowledge, this study is the first to discover the detailed process of the conidia suspension of TL6 infecting the same eggs or J2s in each day of the observation by microscope, and also find the vital dyes of AO and NR were considered as the rapid method to assay the viability of H. avenae. Meanwhile, the possible mechanisms for the potential of TL6 as a bio-control agent against the eggs and J2s of H. avenae were assessed in both greenhouse experiments and TABLE 5 | Effects of different concentrations of TL6 on the populations of nematodes in wheat seeding rhizosphere after inoculated with the eggs of H. avenae in greenhouse experiments.


Mean ± SE in a column followed by different letters are significantly different at P < 0.05, based on Fisher's LSD test. Mock-inoculated represents seedlings neither inoculated with eggs nor with TL6. Control represents seedlings inoculated with eggs but not with TL6.

TABLE 6 | Effects of different concentrations of TL6 on the activity of chitinase in wheat seedling roots after inoculated with the eggs of H. avenae in greenhouse experiments.


Mean ± SE in a column followed by different letters are significantly different at P < 0.05, based on Fisher's LSD test. Mock-inoculated represents seedlings neither inoculated with eggs nor with TL6. Control represents seedlings inoculated with eggs but not with TL6.

in vitro tests. Interestingly, the consistent results from the two assessment facilities indicate that T. longibrachiatum T6 (TL6) has the potential to be used as an effective bio-control agent in suppressing of H. avenae, and the possible mechanisms are due to the direct parasitic and lethal effect of TL6 on the eggs and J2s activity, and the induced defense response (accumulation of defense compounds and up-regulation of enzymes) in wheat plants.

In vitro studies showed that the conidia suspension of TL6 had strong inhibitory and parasitic effects on the eggs of H. avenae and J2s. Moreover, the parasitic effect of the conidia suspension of TL6 on eggs of H. avenae was faster than the effect on their cysts of H. avenae and the J2s of M. incognita, compared with the previous studies (Zhang et al., 2014b, 2015). Also, the lethal effect on the J2s was faster than the inhibitory effect on the eggs in our present study. To study the potential of Trichoderma species in controlling of nematodes, Yang et al. (2010) evaluated the fungal filtrates of 329 Trichoderma strains against Panagrellus redivivus and Caenorhabditis elegans, and found lower nematicidal effect of the strain of T. longibrachiatum against either nematode. This may be related to the role of mutual recognition of the metabolites (an active compound of acetic acid) produced by the conidia suspension of TL6 and the chemicals contained by eggs shell and the body wall of nematodes (Djian et al., 1991). Previous study show that parasitism is one of the modes of action of Trichoderma species against M. javanica, but the percentages of parasitism on the eggs and J2s of M. javanica were very low (Sharon et al., 2009). In our present study, we found the percentages of parasitism on the eggs and J2s of H. avenae were higher than the parasitic effect on the eggs and J2s of M. javanica in vitro, compared with the previous study. It may be related to the contact opportunity of the conidia suspension of TL6 with the eggs shell and the parasites, or the species of Trichoderma strains (Sharon et al., 2009). Moreover, previous studies revealed TABLE 7 | Effects of different concentrations of TL6 on the activity of β-1, 3-glucanase in wheat seedling roots after inoculated with the eggs of H. avenae in greenhouse experiments.


Mean ± SE in a column followed by different letters are significantly different at P < 0.05, based on Fisher's LSD test. Mock-inoculated represents seedlings neither inoculated with eggs nor with TL6. Control represents seedlings inoculated with eggs but not with TL6.

TABLE 8 | Effects of different concentrations of TL6 on the contents of total flavonoids in wheat seedling roots after inoculated with the eggs of H. avenae in greenhouse experiments.


Mean ± SE in a column followed by different letters are significantly different at P < 0.05, based on Fisher's LSD test. Mock-inoculated represents seedlings neither inoculated with eggs nor with TL6. Control represents seedlings inoculated with eggs but not with TL6. FW represents fresh weights.

that P. lilacinus was an opportunistic fungus, and it usually parasitizes on nematode eggs and the percent parasitism was related to the length of time when it contacts with eggs (Leij De et al., 1992; Bonants et al., 1995; Oclarit and Cumagun, 2009). Similarly, P. chlamydosporia usually infects the eggs and females, and the ability to colonize different nematodes varies greatly (Bourne et al., 1996). However, the findings from the present study revealed that the strain of TL6 not only parasitized or colonized on the eggs, but also parasitized the J2s of H. avenae.

Previous studies have demonstrated that some nematophagous fungi, such as the egg-parasitic fungi Paecilomyces spp. (Huang et al., 2004) and Pochonia spp. (Tikhonov et al., 2002), have the ability to trap nematodes, infect nematode eggs, and suppress the hatching of juveniles, thereby reducing nematodes population. In the present study, our results showed that the inhibitory effect of TL6 (1.5 × 10<sup>5</sup> spores ml−<sup>1</sup> ) on the hatching of eggs was stronger than the inhibitory effect of P. lilacinus and P. chlamydosporia on the hatching of individual eggs (41.3 and 36.0%) of Meloidogyne spp. at 1.0 × 10<sup>5</sup> spores ml−<sup>1</sup> (Sun et al., 2006). Such differences may be related to the strains, the mechanism of infection and the host of bio-control fungi (Stirling, 1991; Kerry, 1997; Jeffries et al., 2003; Tchabi et al., 2010). Our results also revealed that the highest parasitic and lethal effect of F. oxysporum on the eggs and J2s of H. avenae were lower than 37.0%, and significantly lower than the parasitic and lethal effect of TL6. In a similar study, Yan et al. (2011) demonstrated that the average control efficacy of five isolates of Fusarium sp. on M. incognita was 35.7%.

To our knowledge, this study is the first in the scientific literature to report and document the detailed process of the


TABLE 9 | Effects of different concentrations of TL6 on the contents of lignin in wheat seedling roots after inoculated with the eggs of H. avenae in greenhouse experiments.

Mean ± SE in a column followed by different letters are significantly different at P < 0.05, based on Fisher's LSD test. Mock-inoculated represents seedlings neither inoculated with eggs nor with TL6. Control represents seedlings inoculated with eggs but not with TL6. DW represents dry weights.

conidia suspension of TL6 infecting the same H. avenae eggs or J2s in each day of the observation by microscope, and the direct parasitism was probably the important mode of action of TL6 in suppressing of H. avenae eggs and J2s growth and development. Zhang et al. (2014b) have been reported that the process of H. avenae cysts infestation by T. longibrachiatum in vitro, but there are no related information regard to the process of TL6 infecting H. avenae eggs or J2s, especially the process of TL6 infecting the same eggs or J2s of H. avenae at different days. Also, our results indicate that TL6 may produce the metabolites and enzymes that cause the physiological disorder of eggs, dissolve nematodes body, and severely affect their physical vitality (Khambay et al., 2000; Sharon et al., 2007; Gortari and Hours, 2008; Pau et al., 2012). However, this needs to be further investigated. Moreover, similar results were reported that the fermentation broth of P. lilacinus was found to contain acetic acid that inhibited the eggs growth (Djian et al., 1991). The parasitism and inhibition of cysts through the increased extracellular chitinase activity serve as the main mechanism with which the fermentation broth of T. longibrachiatum against H. avenae cysts (Zhang et al., 2014b).

Some species of insecticides and nematicides may kill the nematodes eggs, but eggs death cannot be determined unless the contents are visibly damaged (Donald and Niblack, 1994), and also some fungal biological control agents in eggs can be detected only after mycelium has proliferated within the eggs (Zetsche and Meysman, 2012). Moreover, it remains a challenge to count live and dead organisms in a concentrated sample of debris in recent years (Fischel, 1908). Staining procedures have been used for over 100 years to differentiate living cells from dead cells (Crippen and Perrier, 1974). Vital stains acridine orange and tetrazolium red for differentiating viable and nonviable eggs of H. glycines provide a useful tool for laboratory and greenhouse tests (Donald and Niblack, 1994). Less toxic vital stain of neutral red has been successfully applied to copepods (Elliott and Tang, 2009), and recently, the applicability to natural field assemblages of zooplankton has been further promoted (Yan et al., 2011). For the first time in the literature, our study determined the details of the viability of eggs and J2s of H. avenae after treated with biological control strain of TL6 or under ultra-low temperature conditions (−80◦C). We showed that the vital dyes of AO and NR are rapid, effective, and efficient in the assay of the viability of H. avenae.

Our earlier study showed that the fermentation broth of T. longibrachiatum applied at appropriate concentrations suppressed H. avenae cysts development and infection, and at the same time promoted the growth of wheat that was inoculated with the cysts of H. avenae and T. longibrachiatum (Zhang et al., 2014a). Leij De and Kerry (1993) reported that the number of root-knot nematodes in soil was reduced by 90% after application of P. chlamydosporia. An interesting finding from the present study revealed that a high efficiency of TL6 in suppressing of H. avenae development and infection, and promoting wheat seedling growth in comparison to the previous studies (Leij De and Kerry, 1993; Zhang et al., 2014a). Meanwhile, an added value from the present study was that in the study of nematodes, eggs can be used as the inoculum nematodes.

Furthermore, a number of studies have been reported on microorganisms mediated induction of resistance in different plant species (Lopez and Hernández, 2014). However, there are only a few reports on the application of plant growth promoting microorganisms for the induction of resistance in wheat against H. avenae. Especially there are no reports about the colonization of wheat roots by the strain of TL6 inducing the activity of chitinase and β-1, 3-glucanase, and increasing the content of total flavonoids and lignin resistance to H. avenae infection. At our knowledge, the current study is firstly revealed that wheat inoculated with antagonistic bio-control strain of TL6 under H. avenae infection condition induces resistance enzymes (chitinase and β-1, 3-glucanase) and defense compounds (total flavonoids and lignin) responsible for a reprogrammed metabolic cascade related to plant defense and signaling. Our results also provide an insight into the possibilities of applying the biocontrol strain of TL6 for managing H. avenae disease (Harman et al., 2004; Bakker et al., 2007; Singh et al., 2014), which would be eco-friendly means to manage H. avenae disease and contributes to the maintenance of plant and soil health.

Chitinase and β-1, 3-glucanase are the two important pathogenesis-related (PR) proteins in plant tissue and the accumulation of these can contribute to plant defense response against the pathogen infection (Kauffmann et al., 1987; Poonam et al., 2013). We discovered that the activity of chitinase and β -1, 3-glucanase significantly increased when the wheat seedlings were inoculated with the strain of TL6 compared to un-inoculated plants, and that the high concentrations of TL6 remarkably promoted the chitinase and β-1, 3-glucanase activities. The maximum activities of chitinase and β-1, 3 glucanase were recorded at the 20th and 15th Days after the combined application of different concentrations of TL6 and eggs in soil. Guzmán-Valle et al. (2014) also revealed that the activity of glucanase, chitinase and peroxidase in bulbs and roots of different varieties of onion (Allium cepa L.) were increased after application of T. asperellum and Sclerotium rolfsii. Our results were in the line of previous studies which demonstrate that the interaction between Trichoderma species and plant roots induces the increase in enzyme activity and that the magnitude of the activity is related to the infection of soilborne plant pathogens (Harman, 2006; Salas-Marina et al., 2011; Singh et al., 2013). Moreover, a number of previous studies have demonstrated that root colonization by T. harzianum strains has been shown to increase the level of resistance-related enzymes (Howell et al., 2000; Evans et al., 2003), and especially root colonization by T. longibrachiatum significantly increased the specific activities of resistance-related enzymes (peroxidase, polyphenol oxidase and phenylalanine ammonia lyase) to induce the resistance of wheat against H. avenae infection (Zhang et al., 2014a). Singh et al. (2016) reported that wheat seedlings co-inoculated with Bacillus amyloliquefaciens B-16 and T. harzianum UBSTH-501 increased the activity of phenylalanine ammonia lyase, peroxidase, chitinase, β-1, 3-glucanase and other enzymes related to induce systemic resistance responses against Bipolaris sorokiniana. Unlike previous studies, our present study discovered the resistance-related enzymes (chitinase and β -1, 3 glucanase) increased in wheat after inoculated with the strain of TL6 and H. avenae eggs, which may serve as the main mechanism responsible for TL6 against H. avenae.

Flavonoids are an important group of secondary metabolites in plants that can function as an "inducer" that induces the resistance of plants against pathogen infections (Bahraminejad et al., 2008). Lignin is a complex polymer of hydroxylated and methoxylated phenylpropane units, and also cell wall lignifications play a significant role in the incorporation of lignin into plant cell wall so that improving plant resistance to the pathogen invasion. In the present study, we discovered the eggs of H. avenae infection significantly induced the contents of total flavonoids and lignin in the roots of wheat seedlings compared to the un-inoculated plants, and the most notable increase in the contents of total flavonoids and lignin occurred after the combined application of different concentrations of TL6 and eggs in soil at different stages. The maximum contents of total flavonoids and lignin were recorded at the 30th and 40th Days after the combined application of different concentrations of TL6 and eggs in soil. In studying faba bean (Vicia faba L.), Abd El-Rahman and Mohamed (2014) found that the application of benzothiadiazole and T. harzianum combination increased total flavonoids (3.1 and 2.9-fold) and lignin content (81.3% and 59.5%) when the faba bean was infected by the pathogens of Botrytis faba or B. cinerea in comparison to the untreated control. Moreover, previous studies demonstrated that the induction of host plants with resistance by Trichoderma spp. has been described to be associated with activation of disease resistance mechanisms and production of a wide range defense compounds mainly including peroxidase, phenylalanine ammonia lyase, flavonoids (Karthikeyan et al., 2006; Magro et al., 2009; Govindappa et al., 2010; Nianlai et al., 2010), and the accumulation of lignin and pectin in plants cell walls (Al-Hakimi and Al-Ghalibi, 2007; Nianlai et al., 2010). Our results are in agreement with these alterations in plants which may contribute to the resistance by stopping pathogen invasion or slowing down the penetration process, thus allowing the activation of further defense mechanisms in plants (Sticher et al., 1997). Meanwhile, our present results indicate that the changes of lignin contents in wheat seedlings roots may be related to resistance in incorporating of lignin into plant cell wall is bound to make it more resistance to the nematodes invasion. Also, lignified cell walls could also constitute a defense barrier preventing free nutrient movement and therefore help to starve the nematodes (Sticher et al., 1997). However, a more detailed mechanism for the induction of host plants with resistance to plant parasitic nematodes by the strain of TL6 need to be solved in the future studies.

In addition, Gupta et al. (2017) reported that a consortium of bioinoculants of B. megaterium, T. harzianum ThU and Glomus intraradices significantly induced the total phenolic and flavonoid contents in chamomile (Matricaria recutita L.) resistance to root-knot nematode infection. Similarly, Singh et al. (2016) found higher amounts of phenolic acids (gallic acid and ferulic acid) were accumulated in wheat leaves after coinoculated with B. amyloliquefaciens B-16 and T. harzianum UBSTH-501, compared to the individually inoculated and uninoculated control plants. Finally, our results indicate that roots colonization by plant growth promoting microorganism of TL6 induces the accumulation of plant defense compounds and upregulation of enzymes to establish symbiotic interactions as well as to fight against pathogens during the pathogen infection, which in lined with the findings of Harman et al. (2004) and Singh et al. (2013), who showed that colonization of plants with the bio-agents of T. harzianum, B. amyloliquefaciens and other plant growth promoting microorganisms can induce the plant against pathogens.

### CONCLUSION

Our study suggests that the conidia suspension of TL6 has a broad prospect on the prevention and control of H. avenae eggs and J2s in vitro and greenhouse, and has the potential to be used as an effective bio-control agent for H. avenae. The main mechanisms of the strain of TL6 against the eggs and J2s of H. avenae were due to (i) the direct parasitic and lethal effect of TL6 on the activity of the eggs and J2s development, and (ii) the promoting effect on the wheat growth and development, and improving chitinase and β-1, 3-glucanase activities, and the flavonoids and lignin contents in plants resistance to H. avenae infection. In addition, AO and NR can be considered as the rapid vital dyes to assay the viability of H. avenae. Further studies are required to develop effective methodology with which active agents in nematicidal activity can be identified and isolated. Also, studies are needed to determine the functionality of biological nematicide genes in the strain of TL6 and the persistence of the strain in various life stages when interacting with nematodes.

### AUTHOR CONTRIBUTIONS

SZ and WJ designed the experiments with the help of BX. SZ and JL performed the microscopic observation of the infection

### REFERENCES


process of TL6 on H. avenae eggs and J2s. SZ, YG, and BH performed the greenhouse experiments and analyzed the data, with the help of BX. SZ and YG wrote the manuscript, and revised and approved the final manuscript with the help of BX.

### ACKNOWLEDGMENTS

This work was supported by Fostering Foundation for the Excellent Ph.D. Dissertation of Gansu Agricultural University (project YBPY2014002); International Scientific and Technological Cooperation of Gansu Province (project 1604WKCA010) and Hall of Gansu Province Farming Herd Biology Technology (project GNSW-2013-19). The authors are grateful to Dr. Alejandro Calderón-Urrea (Professor of Developmental Biology, California State University, Fresno) and Dr. Yingyu Xue (Professor of Plant Pathology, Gansu Agricultural University, China) for their helpful discussion and critical review of the manuscript.

Chen, S. Y., and Dickson, D. W. (1996). Fungal penetration of the cyst wall of Heterodera glycines. Phytopathology 86, 319–327. doi: 10.1094/Phyto-86-319


on the Tibetan Plateau, Qinghai, China. Australas. Plant Path. 39, 424–430. doi: 10.1071/AP10084


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Zhang, Gan, Ji, Xu, Hou and Liu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# SEC-Translocon Dependent Extracytoplasmic Proteins of Candidatus Liberibacter asiaticus

Samiksha Prasad† , Jin Xu† , Yunzeng Zhang† and Nian Wang\*

Citrus Research and Education Center, Department of Microbiology and Cell Science, Institute of Food and Agricultural Sciences, University of Florida, Lake Alfred, FL, USA

Citrus Huanglongbing (HLB) is the most destructive citrus disease worldwide. HLB is associated with three species of the phloem-limited, gram-negative, fastidious α-proteobacteria: Candidatus Liberibacter asiaticus (Las), Ca. L. americanus (Lam), and Ca. L. africanus (Laf) with Las being the most widespread species. Las has not been cultured in artificial media, which has greatly hampered our efforts to understand its virulence mechanisms. Las contains a complete Sec-translocon, which has been suggested to transport Las proteins including virulence factors into the extracytoplasmic milieu. In this study, we characterized the Sec-translocon dependent, signal peptide containing extracytoplasmic proteins of Las. A total of 166 proteins of Las-psy62 strain were predicted to contain signal peptides targeting them out of the cell cytoplasm via the Sec-translocon using LipoP, SigalP 3.0, SignalP 4.1, and Phobius. We also predicated SP containing extracytoplasmic proteins for Las-gxpsy and Las-Ishi-1, Lam, Laf, Ca. L. solanacearum (Lso), and L. crescens (Lcr). For experimental validation of the predicted extracytoplasmic proteins, Escherichia coli based alkaline phosphatase (PhoA) gene fusion assays were conducted. A total of 86 out of the 166 predicted Las proteins were experimentally validated to contain signal peptides. Additionally, Laspsy62 lepB (CLIBASIA\_04190), the gene encodes signal peptidase I, was able to partially complement the amber mutant of lepB of E. coli. This work will contribute to the identification of Sec-translocon dependent effector proteins of Las, which might be involved in virulence of Las.

#### Keywords: citrus HLB, Liberibacter, Virulence Factors, Sec pathway, secretion

### INTRODUCTION

Citrus Huanglongbing (HLB) is the most destructive disease for citrus industry worldwide. HLB is associated with three species of the phloem-limited, gram-negative, fastidious α-proteobacteria: Candidatus Liberibacter asiaticus (Las), Ca. L. americanus (Lam), and Ca. L. africanus (Laf; Capoor et al., 1967; Jagoueix et al., 1996; Bové,, 2006). Liberibacters are vectored by two psyllid species, Diaphorina citri Kuwayama (ACP) or Trioza erytreae (Del Guercio; Halbert and Manjunath, 2004). Las is the widest spread and most virulent species and so far is the only one reported in the US (Gottwald et al., 2010; Wang and Trivedi, 2013).

Besides HLB, Liberibacters are also known to cause many other plant diseases (Jagoueix et al., 1994; Teixeira et al., 2005; Hansen et al., 2008; Liefting et al., 2008; Raddadi et al., 2011) For

### Edited by:

Bernd Honermeier, University of Giessen, Germany

#### Reviewed by:

Liliana M. Cano, University of Florida, USA Zonghua Wang, Fujian Agriculture and Forestry University, China

#### \*Correspondence:

Nian Wang nianwang@ufl.edu †These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 15 August 2016 Accepted: 28 November 2016 Published: 20 December 2016

#### Citation:

Prasad S, Xu J, Zhang Y and Wang N (2016) SEC-Translocon Dependent Extracytoplasmic Proteins of Candidatus Liberibacter asiaticus. Front. Microbiol. 7:1989. doi: 10.3389/fmicb.2016.01989

example, Ca. L. solanacearum (Lso; Liefting et al., 2009) has been known to cause Zebra chip of potato and to infect peppers and tomatoes. On the other hand, Ca. L. europaeus has been suggested as an endophyte rather than a pathogen (Raddadi et al., 2011). With except L. crescens (Lcr), which was originally isolated from mountain papayas (Leonard et al., 2012; Fagen et al., 2014a), most other Liberibacters have not been cultured in artificial media, therefore, traditional molecular and genetic analyses are difficult to apply. This has greatly hampered our efforts to understand the virulence mechanisms of Las. So far, most insights of the HLB biology and Las pathogenicity are derived from the genome sequences of Las and other related Liberibacters including Las, Lam, Laf, Ca. Lso, and Lcr (Duan et al., 2009; Lin et al., 2011; Leonard et al., 2012; Fagen et al., 2014b; Wulff et al., 2014).

One of the most important virulence factors of bacterial pathogens is the presence of protein secretion systems, which secrete proteins, called effectors, into host cells. Interestingly, Las contains a complete General Secretory Pathway (GSP/Sectranslocon), but lacks the Sec-dependent type II (T2SS) and type V (T5SS) secretion systems and type III (T3SS) secretion system (Duan et al., 2009). The Sec machinery facilitates the majority of protein transport across the cytoplasmic membrane and is essential for bacterial viability (Segers and Anné, 2011). The Sec pathway is also critical for secretion of important virulence factors by certain bacterial pathogens, e.g., Phytoplasma, a bacterial pathogen residing in the phloem similarly as Las.

Bacterial proteins translocated exclusively by the Sectranslocon are synthesized initially as protein precursors in the cytoplasm, containing signal peptide (SP) sequences of approximately 20–30 amino acid residues at the amino-terminal (Economou, 1999). Proteins containing these SP have a similar architecture and are normally cleaved by signal peptidases: (i) a basic "n region" at the amino terminus, which is about 5–8 amino acids long and is characterized by the presence of basic residues. The net positive charge of this region is known to be crucial for interaction with the negatively charged surface of the inner membrane (Rehm et al., 2001; Palmer and Berks, 2012). (ii) a hydrophobic "h region" in the middle, about 8–12 amino acids long. It is composed largely of non-polar amino acids. This region has a high propensity for alpha-helical formation, a conformation that may facilitate interaction with the interior of the bilayer (Berks, 1996; Palmer and Berks, 2012) and (iii) a polar "c region" or cleavage region about 6 amino acids long at the carboxyl terminus. This region is involved in signal peptidase recognition and cleavage, which is usually required to achieve final folding and localization of the exported proteins (Tuteja, 2005; Palmer and Berks, 2012). The characteristic tripartite amino acid composition in the SP sequences of Sec-translocon dependent pre-proteins is particularly useful to distinguish proteins containing SP (Pugsley, 1993). Numerous dedicated bioinformatics tools are available for predicting the potential localization and eventual destination of the proteins based on the protein sequence (Andersson and von Heijne, 1994).

We hypothesized that Sec-translocon serves as a potent system for the transportation of Las proteins into the extracytoplasmic milieu, which can be identified by the presence of signal peptide sequence. We comprehensively identified Sec-dependent cytoplasmic proteins containing SP in Las and other sequenced Liberibacters using four well-adopted algorithms, and validated the bioinformatic predictions for SP-containing Sec-dependent cytoplasmic proteins in Las using the Escherichia coli-based PhoA assay.

### MATERIALS AND METHODS

### Prediction of Sec-Dependent Extracytoplasmic Proteins of Liberibacters

The entire annotated genome of Las strain Psy62 (taxid: 537021, GenBank accession no. CP001677; Duan et al., 2009). Las strain gxpsy (taxid: 1174529, GenBank accession no. CP004005; Lin et al., 2013); Las strain ishi-1 (taxid: 931202, GenBank accession no. NZ\_AP014595; Katoh et al., 2014); Lam strain São Paulo (taxid: 1261131, GenBank accession no. CP006604; Wulff et al., 2014); Laf strain PTSAPSY (taxid: 1277257, GenBank accession no. CP004021; Lin et al., 2015); Lso strain ZC1 (taxid: 658172, GenBank accession no. CP002371; 16) and Lcc strain BT-1 (taxid: 1215343, GenBank accession no. CP003789; Leonard et al., 2012) were screened to identify the genes encoding proteins containing SP. The SP prediction was conducted using the following online algorithms: LipoP server 1.0 (Juncker et al., 2003), Phobius (Käll et al., 2004, 2007), SignalP version 3.0 (Bendtsen et al., 2004), and SignalP version 4.1 (Petersen et al., 2011). The screening was performed with default settings of the algorithms for gramnegative bacteria.

### Ortholog Cluster Homology Analysis of SP Containing Proteins

Genome-wise orthologous gene clustering among the seven strains were performed using Get\_homologs program (ver. 20140311) with parameters: -M, -e 0, -E 0.01 and -S 60 (Contreras-Moreira and Vinuesa, 2013). The ANIm values between genomes were calculated using the NUCmer algorithm v3.1integrated in Jspecies v1.2.1 (Richter and Rossello-Mora, 2009). The orthologous relationship of the identified SP positive genes were determined based on the orthologous gene clusters generated by Get\_homologs. Manual curation was performed for the genes whose original annotation was not proper. A total of 596 clusters of orthologs were generated in this analysis across the seven genomes. The hierarchical clustering of the seven Liberibacters was conducted based on gene presence and absence matrix of the orthologous clusters. Dendro UPGMA<sup>1</sup> was used to generate the UPGMA tree with Jaccard coefficient. A total of 100 bootstrap replicates were prepared, and the values of >50% at each node was noted as a percent value.

### Gene Specific Primer Design

Gene specific forward and reverse primers for each of the 166 predicted SP containing extracytoplasmic proteins of Las-psy62

<sup>1</sup>http://genomes.urv.es/UPGMA/

strain were designed for amplification of the full-length gene (excluding the stop codon; Supplementary Table S9). The melting temperature and GC content of the primers were calculated<sup>2</sup> . The primers were designed to incorporate appropriate restriction enzyme sites at the 5<sup>0</sup> and 3<sup>0</sup> ends of the resultant amplicons (Supplementary Table S9).

### Las Genomic DNA Extraction

fmicb-07-01989 December 16, 2016 Time: 14:37 # 3

Huanglongbing symptomatic leaves from citrus groves of Citrus Research and Education Center (CREC), University of Florida, Lake Alfred, Florida were collected and washed with sterilized double distilled water and the midrib section of the leaf was used for extraction of Las genomic DNA. DNA was extracted using the Wizard Genomic DNA purification kit (Promega).

### Alkaline Phosphatase (PhoA) Assays

Gene specific forward and reverse primers were used for amplification of the Las genes. The resultant amplified PCR products were digested with the cognate restriction enzymes (NEB) and subsequently purified by Wizard SV Gel and PCR Clean-Up System (Promega). The fragments were then subjected to ligation with pJDT1-SDM-1 vector using T4 DNA ligase (NEB) to obtain an in-frame gene fusion with phoA. The amplified Las genes do not contain the stop codon, and the phoA is truncated without its SP sequence for in-frame fusion purpose. The E. coli chemically competent strain of JM105 (Promega) was used for transformation.

The transformants were selected on LB agar plates containing 100 µg/mL Ampicillin. The transformants were tested for PhoA activity on LB agar plates containing 90 µg/mL 5- BCIP as chromogenic substrate. To block endogenous phosphatase activity, 75 mM Na2HPO<sup>4</sup> was added. SP presence was indicated by blue colonies, whereas lack of PhoA activity was signified by the white colonies. The plasmids from PhoA positive colonies were purified and sequenced with primers adjacent to the location of insertion (5<sup>0</sup> -CAG GAA ACA GCT ATG AC-3<sup>0</sup> ; 50 -CGC TAA GAG AAT CAC GCA GAG C-3<sup>0</sup> as forward and reverse primers, respectively) for confirmation. The empty pJTD1-SDM-1 vector transformed JM105 competent cells were used as a negative control.

### Multiple Sequence Alignment

The DNA sequences of the lepB gene encoding the SPase I in E. coli and Las strains were retrieved from National Center of Biotechnology Information (NCBI). Multiple sequence

<sup>2</sup>http://www.endmemo.com/bio/tm.php

alignment was conducted using Multiple Sequence Comparison by Log Expectation (MUSCLE) with default settings. For the phylogenetic tree and identity matrix of the sequences, the ClustalO (Clustal Omega) version 2.1 at default settings was used.

### Screening for Complementation with Las lep Gene

The E. coli K-12 MG1655 wild type (IT42: lep) and amber mutant (IT41: lep9, or 1lep) strains were grown from stocks received from Dr. Inada at Kyoto University, Japan on LB plates at 37◦C overnight with 20 µg/mL tetracycline as selection marker. Single colonies were picked for further studies. The Las lepB gene was amplified, flanked by appropriate restriction sites (HindIII and SpeI) for insertion into the pBBR1mcs5 vector. The amplified fragment was digested with appropriate restriction enzymes (NEB) and purified with Wizard <sup>R</sup> SV Gel and PCR Clean-Up System (Promega). The construct was subjected to ligation with the pBBR1mcs5 vector with T4 DNA ligase (NEB). The chemically competent strain of E. coli JM105 (Promega) was used for transformation. The resultant plasmid transformed into the amber mutant of E. coli (lep−) by electroporation. The three strains: E. coli wild type (WT), E. coli amber mutant (1lep) and E. coli amber mutant complimented with Las\_psy62 lep (1lep::lepLas) were grown at 32 and 42◦C to assess the bacterial growth.

### RESULTS

### Prediction of SP Containing Proteins for Liberibacters

A total of 166 proteins were predicted to contain signal peptides in Las-psy62, comprising 15% of the total annotated proteins of Las using LipoP, SigalP 3.0, Signal P4.1, and Phobius (**Tables 1** and **2**). The four tools use distinct algorithms for signal peptide prediction and complement each other, thus the merged list from the four tools comprehensively represented the potential signal peptide containing proteins in Las-psy62 (**Figure 1**). LipoP server 1.0 also categorized proteins into lipoprotein and nonlipoprotein.

Prediction of SP containing proteins was performed for six more Liberibacter strains including two more Las strains, another two species also causing HLB (Lam and Laf), one non-citrus pathogenic species Lso and one non-pathogenic relative species Lcc (**Table 2**, Supplementary Tables S1–S6). Lasgxpsy and Las-ishi-1 were predicted to have a total of 168 (Supplementary Table S1) and 164 (Supplementary Table S2) SP

TABLE 1 | Prediction of signal peptide containing extracytoplasmic proteins in Las-psy62 and PhoA assay results.



TABLE 2 | Prediction of signal peptide containing extracytoplasmic protein predictions in different species and strains of Liberibacter.

containing extracytoplasmic proteins, respectively. Lam contains 133 (Supplementary Table S3) predicted SP containing proteins, whereas Laf contains 141 (Supplementary Table S4). For Lso 171 SP containing proteins were predicted (Supplementary Table S5). A total of 214 putative SP containing proteins were predicated for Lcc (Supplementary Table S6).

### Orthologous Cluster Homology Analysis of SP Containing Proteins

Orthologous relationship between the identified putative extracytoplasmic proteins of the seven sequenced Liberibacters was determined (**Figure 2**, Supplementary Table S7). 596 orthologous clusters were formed when the threshold identity 60% and coverage 75% was applied. This analysis allowed us to compare the predicted SP containing extracytoplasmic proteins of different strains and species of Liberibacter. Interestingly, this phyletic tree based on the distribution of the SP positive proteins among the seven strains is consistent with the maximumlikelihood phylogenetic trees reconstructed using 16S rRNA

gene sequences (Fagen et al., 2014a), indicating the gain and loss history of these SP positive proteins was convergent with the evolution history of the relevant genome background.

Only 17 predicted extracytoplasmic proteins are homologous between the seven Liberibacters (**Table 3**, Supplementary Table S7). Amongst the six infectious Liberibacters, i.e., Las, Lam, Laf, and Lso, 45 SP containing proteins were predicted. Totally 151 SP containing proteins were shared among the three strains of Las (Supplementary Table S8). 73, 60 and 45 SP containing homologous proteins were shared by Laf, Lso, and Lam, respectively, to Las.

### Using E.coli as a Model to Indirectly Validate the Predicated SP Containing Proteins with PhoA Assay

To experimentally validate the presence of SP in the predicted SP containing proteins in Las-psy62 strain, PhoA assay was conducted using E.coli as a model since SP is highly conserved among different bacteria (Ammerman et al., 2008). Gene specific primer sets for each gene encoding the predicted proteins were designed (Supplementary Table S9). The amplified DNA sequence encoding the putative SP containing protein was inserted upstream of the phoA without SP in frame. Out of the 166 predicted proteins, 86 proteins (52%; **Table 1** and Supplementary Table S10) were PhoA positive and turned dark blue at the presence of bromo-4-chloro-3-indolyl phosphate (BCIP; **Figure 3**), suggesting that they contain a SP in their sequences that can direct them to translocate outside of the cytoplasm via the Sec pathway. The empty PJDT1-SDM-1 was used as a negative control, which did not result in color changes. Fifty one predicted proteins were PhoA negative whereas 29 predicted proteins could not be determined experimentally (Supplementary Table S10).

### SPase I Is Conserved in E. coli and Las Strains

Type I signal peptidase (SPase I) is responsible for cleaving off the amino-terminal signal peptide from proteins that are secreted across the bacterial cytoplasmic membrane (Paetzel, 2014). We further test whether SPase I is conserved in Las and E.coli. Multiple sequence alignment was conducted for SPase I of

#### TABLE 3 | Common homologs between the seven different species and strains of Liberibacter<sup>∗</sup> .


<sup>∗</sup>Proteins which are PhoA positive

E. coli strain K-12 substrain MG1655 (EO53\_04950); Las-psy62 (CLIBASIA\_04190); Las-gxpsy (WSI\_04025) and Las-Ishi-1 (CGUJ\_04190) strains (Supplementary Figure S1). The identity for the SPase I of the three Las strains is 100%, whereas the Las SPase I protein shares 34% identity and 52% similarity with that of E. coli.

We further tested whether Las lepB gene which encodes SPase I could complement the E. coli amber mutants of lepB. The lepB amber mutant of E. coli (1lep) displays temperature sensitivity, leading to conditional lethality at 42◦C, but not at 37◦C (Paetzel, 2014). At 37◦C, the WT, 1lep and the complimented 1lep:lepLas strains showed similar growth. At 42◦C, the WT and 1lep:lepLas strains displayed growth, whereas the 1lep strain was unable to grow (**Figure 4**). It is noteworthy that 1lep:lepLas grew slower than the wild type E.coli strain, which indicates that Las lepB could partially complement the lepB mutant of E.coli.

### DISCUSSION

The signal peptide is an important protein-sorting signal that targets its passenger protein for transportation out of the cytoplasm in prokaryotes (Von Heijne, 1990). Many methods have been used for predicting signal peptides, including SignalP (Nielsen et al., 1997; Nielsen and Krogh, 1998; Bendtsen et al., 2004; Petersen et al., 2011), PrediSi (Hiller et al., 2004), SPEPlip (Fariselli et al., 2003), Signal-CF (Chou and Shen, 2007), Signal-3L (Shen and Chou, 2007), signal-BLAST (Frank and Sippl, 2008), Phobius (Käll et al., 2004), LipoP (Juncker et al., 2003) and Philius (Reynolds et al., 2008). All the prediction methods have limited ability to discriminate between signal peptides and N-terminal transmembrane helices. The common characteristic of signal peptides and N-terminal transmembrane helices is hydrophobic. Transmembrane helices usually have

homolog of Las partially restores the growth of the mutant strain at 42◦C. WT: E. coli wild type (IT42: lepB), E.coli1lep: E. coli (IT41: lep9) amber mutant and E.coli1lep:lepLas: E. coli (IT41: lep9) amber mutant complimented with Las lepB gene.

longer hydrophobic regions. Transmembrane helices do not have cleavage sites that are associated with signal peptides. However, the cleavage-site pattern alone is not sufficient to distinguish the two types of sequence. Consequently, each method has its pros and cons and both false positives and false negatives were reported for each prediction method (Heng Choo et al., 2009). Among them, SignalP, Phobius, and LipoP use distinct algorithms for prediction and complement each other. Specifically, Phobius combined transmembrane protein topology and signal peptide predictor, thus generating superior prediction in differentiating signal peptides from transmembrane helices. In addition, LipoP using hidden Markov model (HMM) can distinguish between lipoproteins (SPaseII-cleaved proteins), SPaseI-cleaved proteins, cytoplasmic proteins, and transmembrane proteins (Juncker et al., 2003). On the other hand, SignalP and most prediction programs are only trained on SPaseI-cleaved proteins (Nielsen et al., 1997; Nielsen and Krogh, 1998; Bendtsen et al., 2004; Petersen et al., 2011). Thus, we combined SignalP 3.0, SignalP 4.1, Phobius, and LipoP for prediction of SPcontaining extracytoplasmic proteins in Liberibacters. In spite of the potential false positive and false negative predictions, it is believe the prediction is still useful since 87 to 96% accuracy have been reported for the various programs (Juncker et al., 2003; Heng Choo et al., 2009). The overlapping prediction results of SignalP 3.0 and 4.0, Phobius, and LipoP will likely to be accurate, but with false negative, whereas the overall predication results will likely remove false negative results, but with false positives. Thus, experimental confirmation is critical for the in silico predication of SP-containing extracytoplasmic proteins in Liberibacters.

Since Las has not been cultivated in media, we have used E.coli as a model to indirectly validate the predicated SP containing proteins with PhoA assay. Out of the 166 proteins predicted, 86 proteins were PhoA positive tested in E. coli, suggesting that they contain a SP in their sequences that can direct them to be translocated outside of the cytoplasm via the Sectranslocon. PhoA assay using E.coli as a model has been used to experimentally test SP-containing proteins in multiple bacteria including Pseudomonas aeruginosa (Lewenza et al., 2005),

Helicobacter pylori (Bina et al., 1997), Bacillus subtilis (Payne and Jackson, 1991), Actinobacillus actinomycetemcomitans (Mintz and Fives-Taylor, 1999; Ward et al., 2001), Mycobacterium tuberculosis (Wiker et al., 2000), Streptococcus pneumoniae (Pearce et al., 1993), Vibrio cholerae (Taylor et al., 1989), Staphylococcus aureus (Williams et al., 2000) and Rickettsia typhi (Ammerman et al., 2008). A heterologous system could be used to test the secretion of SP-containing proteins by the Sec pathway is because that the SP and Sec apparatus are conserved. The Las Sec apparatus contains SecB, Ffh, SecE, SecD/F, YidC, YajC, SecY, and SecA which share 28–50% identity and 52–70% similarity with their counterparts in E.coli. The majority of signal peptides are cleaved by signal peptidase I which is encoded by lepB and shares 34% identity and 52% similarity with its counterpart in E.coli (Supplementary Figure S1). Type II signal peptides, which are associated with lipoproteins are cleaved by signal peptidase II. The signal peptidase of Las shares 37% identity and 57% similarity with that of E.coli. Las lepB could partially complement the lepB amber mutant of E.coli (**Figure 4**). The aforementioned evidence suggests that the PhoA assay using E.coli as a model will provide strong experimental support of confirmation of SP. Furthermore, among the 86 PhoA positive proteins, many are associated with the cell envelope including outer membrane proteins (e.g., OmpA/MotB, and Omp19), flagellar proteins, Type IV pilus proteins, proteases, dehydrogenases, hydrolase, monophosphatase, monooxygenase, ATPase, ABC transporters, periplasmic binding proteins, translocation protein, and nodulation related efflux protein (Supplementary Table S10), which further support the reliability of PhoA assay. Additionally, we need to point out that 29 predicated SP containing proteins were not determined in this study. Most of them are due to failure of amplification despite repeated attempts. Thus it is likely that more predicted SP containing proteins can be experimentally verified.

Remarkably, significantly high number of hypothetical proteins (47) were PhoA positive in E. coli, which is intriguing and certainly suggests the need for further investigation. Additionally, 36 SP containing proteins have been shown to be highly expressed in planta compared to in psyllids whereas eight are highly expressed in psyllids compared to in planta (Supplementary Table S11) (Yan et al., 2013), which suggest that those proteins might play critical roles for Las adapts to its living in the two hosts. In addition, CLIBASIA\_04040 contains four known domains out of which two motifs (PF09487: HrpB2 and PF05758: Ycf1) have been shown to be involved in virulence in other plant pathogens, e.g., P. syringae and animal pathogens, e.g., Yersinia. How the SP-containing hypothetical extracytoplasmic proteins contribute to the virulence of Las remains to be explored.

As Las possesses a highly reduced genome size (1.23-Mb), presence of the Sec-translocon suggests the Sec-translocon and its substrates play important roles for Las and other Liberibacters. A total of 166 proteins were predicted to contain SP in Las-psy62 whereas 168 and 164 SP-containing extracytoplasmic proteins were predicated for Las-gxpsy and Las-ishi-1, respectively. The three Las strains from USA, China and Japan show high uniformity in their Sec dependent extracytoplasmic proteins with 151 overlapping in all three. This is consistent with the high ANI values (99.85–99.94%) of the three strains. The similarity in Sec dependent extracytoplasmic proteins and ANI indicate that the Las strains in US, China and Japan have not undergone extensive evolution changes despite the graphical separation. However, significant differences were observed between Las, Laf, and Lam even though they all cause HLB. Only 45 Sec dependent extracytoplasmic proteins showed homology between them. The significant difference in Sec dependent extracytoplasmic proteins in Las, Laf, and Lam might contribute to the virulence and/or adaption difference among the three Liberibacter species with Las being the most widely spread species.

### CONCLUSION

We predicted SP-containing extracytoplasmic proteins for Las, Lam, Laf, Lso, and Lcr. Eighty six Las proteins has been experimentally confirmed to be SP-containing extracytoplasmic proteins using PhoA assay with E.coli as a model. Our study has provided insight into the potential function of certain SP-containing hypothetical proteins of Las. Our data also showed that Las lepB gene can partially complement the E.coli lepB amber mutant. Due to the importance of Sec-translocon and its substrate, suppression of the Sec secretion system by developing antimicrobials targeting suitable targets, e.g., SecA, has the potential to inhibit HLB progression (Akula et al., 2011).

### AUTHOR CONTRIBUTIONS

NW, SP, JX, and YZ initiated the project and designed experiments. SP, JX, YZ, and NW performed all experiments and data analysis. NW, SP, JX, and YZ wrote the manuscript. NW supervised the project.

### FUNDING

This project has been supported by Florida Citrus Research and Development Foundation.

## ACKNOWLEDGMENTS

We thank Drs. M. Sayeedur Rahman and Abdu Azad, University of Maryland School of Medicine, for the gift of the pJDT plasmid for the PhoA assay. We thank Dr. Inada at Kyoto University, Japan for the gift of the lepB mutant of E.coli.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb.2016. 01989/full#supplementary-material

### REFERENCES

fmicb-07-01989 December 16, 2016 Time: 14:37 # 8



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Prasad, Xu, Zhang and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Diaphorina citri Induces Huanglongbing-Infected Citrus Plant Volatiles to Repel and Reduce the Performance of Propylaea japonica

Yongwen Lin1,2,3,4† , Sheng Lin1,2,4,5† , Komivi S. Akutse1,2,4,5, Mubasher Hussain1,2,3,4 and Liande Wang1,2,3,4 \*

<sup>1</sup> State Key Laboratory of Ecological Pest Control for Fujian and Taiwan Crops, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> Plant Protection College, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>3</sup> Key Laboratory of Biopesticide and Chemical Biology, Ministry of Education, Fuzhou, China, <sup>4</sup> Key Laboratory of Integrated Pest Management for Fujian-Taiwan Crops, Ministry of Agriculture, China, Fuzhou, China, <sup>5</sup> Institute of Applied Ecology, Fujian Agriculture and Forestry University, Fuzhou, China

#### Edited by:

Anton Hartmann, Helmholtz Zentrum München, Germany

### Reviewed by:

Bryan Bailey, United States Department of Agriculture, USA Yi Li, Peking University, China

#### \*Correspondence:

Liande Wang liande\_wang@126.com †These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 02 September 2016 Accepted: 12 December 2016 Published: 26 December 2016

### Citation:

Lin Y, Lin S, Akutse KS, Hussain M and Wang L (2016) Diaphorina citri Induces Huanglongbing-Infected Citrus Plant Volatiles to Repel and Reduce the Performance of Propylaea japonica. Front. Plant Sci. 7:1969. doi: 10.3389/fpls.2016.01969 Transmission of plant pathogens through insect vectors is a complex biological process involving interactions between the host plants, insects, and pathogens. Simultaneous impact of the insect damage and pathogenic bacteria in infected host plants induce volatiles that modify not only the behavior of its insect vector but also of their natural enemies, such as parasitoid wasps. Therefore, it is essential to understand how insects such as the predator ladybird beetle responds to volatiles emitted from a host plant and how the disease transmission alters the interactions between predators, vector, pathogens, and plants. In this study, we investigated the response of Propylaea japonica to volatiles from citrus plants damaged by Diaphorina citri and Candidatus Liberibacter asiaticus through olfactometer bioassays. Synthetic chemical blends were also used to determine the active compounds in the plant volatile. The results showed that volatiles emitted by healthy plants attracted more P. japonica than other treatments, due to the presence of high quantities of D-limonene and beta-ocimene, and the lack of methyl salicylate. When using synthetic chemicals in the olfactory tests, we found that Dlimonene attracted P. japonica while methyl salicylate repelled the predator. However, beta-ocimene attracted the insects at lower concentrations but repelled them at higher concentrations. These results indicate that P. japonica could not efficiently search for its host by using volatile cues emitted from psyllids- and Las bacteria-infected citrus plants.

Keywords: ladybird beetle, volatiles, huanglongbing disease, insect vector, D-limonene, methyl salicylate

## INTRODUCTION

Plants are constantly damaged by sap-feeding insects that not only create physical damage while probing but also transmit pathogens that may cause serious disease in plants. Insect-borne pathogens can induce changes in the traits of their primary hosts as well as their vectors that eventually affect the frequency and nature of interactions between hosts and vectors (Eigenbrode et al., 2002; Purcell, 2003). Simultaneous damage by insects as well as pathogenic bacteria induces infested host plants to secrete volatiles that modify not only the behavior of its insect vector but also their natural enemies.

Infected plants have been reported to attract insects infested with pathogens more than a healthy plant (Mann et al., 2009). For example, cucumber mosaic virus (CMV) infected plants emit some chemicals that can manipulate the behavior (increase the attractiveness) of the vector aphids, Myzus persicae (Sulzer) and Aphis gossypii Glover (Hemiptera: Aphididae) (Mauck et al., 2010). In another study, the phytopathogenic bacterium, Candidatus Liberibacter asiaticus (Rhizobiales: Rhizobiaceae) (Las) infected citrus was found to alter the release of specific headspace volatiles and attract more Diaphorina citri Kuwayama (Hemiptera: Psyllidae) vectors (Mann et al., 2012). In addition, Las has the ability to manipulate the propensity of psyllids movement and consequently satisfy the vector to promote their spread (Martini et al., 2015). However, de Oliveira et al. (2014) found that the bird cherry-oat aphid, Rhopalosiphum padi (Linnaeus) (Hemiptera: Aphididae), which is a vector of cereal yellow dwarf virus (CYDV), a pathogen of wheat, became more vulnerable to attacks by the parasitoid wasp, Aphidius colemani Viereck (Hymenoptera: Braconidae) when carrying CYDV. This indicates that plant pathogens alter the metabolites of infected host plants (Rogers and Bates, 2007; Mauck et al., 2010; Van Den Abbeele et al., 2010; Davis et al., 2012; Mann et al., 2012).

Studies have shown that parasitoids are more sensitive to volatiles emitted by infected plants. To attract more insect vectors, Las induces citrus plants to emit more methyl salicylate (Mann et al., 2012). However, simultaneous attack of citrus plants by Las and psyllids, change both the profile and quantity of volatiles secreted (Hijaz and Killiny, 2016). Interestingly, methyl salicylate, produced by Las-infested host plants, is not only used by the bacterial pathogen to manipulate the behavior of its insect vector to promote its own proliferation, but also used by an effective ectoparasitoid of D. citri, Tamarixia radiata (Waterston) (Hymenoptera: Eulophidae), as a cue to trace its host insect (Martini et al., 2014).

As a consequence, plant pathogens not only regulate the metabolism of host plants, but also influence their vectors, while parasitoids become more sensitive to new metabolites. Polyphagous insect predators are also attracted by herbivoreinduced-plant volatiles (Ninkovic et al., 2001; Gencer et al., 2009). It is known that parasitoids are more sensitive to and are attracted by the induced volatiles than their predators (Li et al., 2014). However, little is known about the mechanism that regulates this difference. Therefore, in this study, we used ladybird beetles as models to address the question how ladybird beetles respond to volatiles emitted from host plants that are damaged by insect vectors as well as are infected by plant pathogens.

Las is a phytopathogenic bacterium that is spread by an insect vector, the Asian citrus psyllid, D. citri. It is a gram-negative, fastidious, phloem-limited bacterium that causes huanglongbing (HLB) disease in citrus (Roossinck, 2011). Ladybird beetles are known as one of the polyphagous insect predators of D. citri in citrus orchards (Michaud et al., 2004). In this study, we investigated the response of a common insect vector Propylaea japonica (Thunberg) (Coleoptera: Coccinellidae), in the crop systems of China, to volatiles from citrus damaged by both D. citri and Las bacteria. Synthetic chemicals were also used to determine the active compounds in plant volatile blends that affect P. japonica.

### MATERIALS AND METHODS

### Plants

Sour orange (Citrus aurantium L.) seeds were cultured in an insect-proof greenhouse at 28◦C, 40% RH and L16: D8 for 4 months. The 4-month-old seedlings were infested by grafting them with four pieces of bud wood sticks from a PCR-positive HLB source. The infection was determined and confirmed using PCR as described by Tatineni et al. (2008). Infected (4 month post inoculation) and uninfected 8-month-old citrus plants were used in this study.

### Insect Colonies

Propylaea japonica adults and larvae were collected from the fields of Fuzhou, Fujian, China (26.08 N, 119.28 W) in 2014, and reared on D. citri that were maintained on Murraya paniculata L. (Jack). Five-day-old adult P. japonica, which emerged from the established colony (third generation) were selected for experiments. Diaphorina citri used for rearing the ladybird beetles and for the experiments were obtained from a colony established in 2014 from field populations in Ganzhou, Jiangxi, China (25.85 N, 114.92 W); and 2- to 3-day-old adults were selected for the tests. This D. citri culture was maintained on uninfected sour orange plants in a clean 50 cm × 50 cm × 50 cm netted cage in an insect-proof greenhouse maintained at 28◦C, 40% RH, and L16: D8. Psyllids reared on the infested citrus plants were sampled to assess Las bacteria infection by PCR, and the results were positive.

### Olfactometer Assays of Propylaea japonica

The response of P. japonica to odor sources was studied in two-choice tests with a closed system Y-tube olfactometer, using the method described by Gencer et al. (2009) with a slight modification. Five treatment combinations were used to test the attraction of the chemicals to P. japonica with an olfactometer: (1) healthy citrus (HC) vs. Las-infected citrus (LC); (2) healthy citrus infested with psyllids D. citri (HCfP) vs. Las-infected citrus infested with psyllids (LCfP); (3) HC vs. HCfP; (4) LC vs. LCfP; and (5) psyllids only vs. fresh air only. For the feeding damage treatments or citrus infestation with the psyllids, 50 adult male and 50 adult female psyllids were placed on plants for 24 h before assays (Mann et al., 2012; Martini et al., 2014). This duration of feeding damage has been proven sufficient to release herbivore-induced volatiles from citrus plants (Mann et al., 2012). Plant samples were placed in 35 cm tall × 25 cm wide domeshaped volatile collection chambers as described by Lin et al. (2016). For the olfactometer experiment, purified air at the rate of 300 ml min−<sup>1</sup> was blown into the chambers from the inlet, and then along with headspace and flowed out of the outlet, then went into the arm of the Y-tube (**Figure 1**). A cotton ball was placed at the end of each arm of the Y-tube to block the released ladybird

beetle. Adult P. japonica were released individually at the entry of the Y-tube and were allowed 300 s to exhibit a behavioral response (Martini et al., 2015). Fifty adult P. japonica (male:female ratio 1:1) were individually released into the Y-tubes for each of the five treatments as defined above. Y-tubes were cleaned with hot water (above 60◦C) after every set of five insects was tested. Different sets of plant treatments were used to test 50 adult P. japonica, and the selected adult ladybird beetles were starved for 24 h before testing. In total, there were five replicates in this experiment.

### Headspace Collection and Volatiles Analysis

Headspace was used to collect volatiles for 24 h with a collecting system described by Pineda et al. (2013). Four treatments were used as described above; HC, LC, HCfP, and LCfP. For the feeding damage treatments, 50 adult male and 50 adult female psyllids were placed on plants for 24 h prior to the bioassays (Mann et al., 2012; Martini et al., 2014). All the treatment plants were placed in clean glass jars inside the collecting chamber as mentioned above. Fresh air was blown into the jar at the rate of 300 ml min−<sup>1</sup> and flowed out of the headspace through a glass tube filled with absorbent, 200 mg Tenax TA (60/80 mesh; Grace-Alltech, Deerfield, MA, USA). After 24 h collection, the herbivore-induced volatiles, Las-induced volatiles and herbivore-Las-induced volatiles (HBIPVs), which were trapped by Tenax TA, were immediately eluted into 1 ml methenyl trichloride and kept at −20◦C until analysis.

The herbivore-induced volatiles, Las-induced volatiles and HBIPVs were analyzed with a gas chromatograph (Agilent Technologies 7890B GC System)-mass spectrometer (Agilent Technologies 5977A MSD) (GC-MS) aligned with HP-5 column (30 mm × 0.25 mm, i.d., 1.0 µm film thickness, Agilent). The GC-MS oven temperature was programmed from 40◦C (5-min hold) to 280◦C at the rate of 10◦C min−<sup>1</sup> . The column effluent was ionized by electron impact ionization at 70 eV. Mass spectra were acquired by scanning from 35 to 350 m/z with a scan rate of 5.38 scans s−<sup>1</sup> .

Compounds were identified by using the deconvolution software AMDIS (version 2.64, NIST, USA) in combination with NIST 05 and Wiley seventh edition spectral libraries and by comparing their retention indices with those from the literature. Additional experiments were conducted prior to the bioassays and the results are compiled in the Supplementary documents (Supplementary Experiments 1, 2, and 3 dataset).

### Propylaea japonica Response to Synthetic HBIPVs Source

As shown in Supplementary Table 3, D-limonene, methyl salicylate and beta-ocimene volatile compounds collected from the headspace have a significant correlation with the olfactory response of P. japonica and therefore were considered for single and blends bioassays. For the experiments, the synthetic version of the three chemicals mentioned above (methyl salicylate, beta-ocimene, and D-limonene) were purchased from Macklin Biochemical Co. Ltd. (Shanghai, China) with approximately 97– 99% purity.

For testing single dosage, D-limonene was dissolved in triethyl citrate (TEC) to 50, 25, 17, 10, 1, and 0.1 nmol ml−<sup>1</sup> ; methyl salicylate was dissolved in TEC to 0.03, 0.06, 0.15, 0.18, 0.32, and 0.64 nmol ml−<sup>1</sup> ; and beta-ocimene was dissolved in TEC to 0.04, 0.08, 0.20, 0.45, 1.00, and 2.00 nmol ml−<sup>1</sup> concentrations. The range of dilutions used for the bioassays was based on the

Supplementary Experiment 2 mentioned above (Supplementary Tables 4 and 5). However, for testing blends, these three chemical compounds were mixed in TEC at different concentrations as follows: (1) blend HC: <sup>D</sup>-limonene 24.75 nmol ml−<sup>1</sup> and beta-ocimene 1.00 nmol ml−<sup>1</sup> ; (2) blend LC: D-limonene 17.32 nmol ml−<sup>1</sup> , methyl salicylate 0.06 nmol ml−<sup>1</sup> and beta-ocimene 0.45 nmol ml−<sup>1</sup> ; (3) blend HCfP: D-limonene 10.13 nmol ml−<sup>1</sup> , methyl salicylate 0.15 nmol ml−<sup>1</sup> , and betaocimene 0.20 nmol ml−<sup>1</sup> ; and (4) blend LCfP: D-limonene 1.06 nmol ml−<sup>1</sup> , methyl salicylate 0.18 nmol ml−<sup>1</sup> , and betaocimene 0.08 nmol ml−<sup>1</sup> (Supplementary Table 5).

Then, 1 ml of each suspension solution was applied to a small cotton roll and placed in the collecting chamber. Purified air at the rate of 300 ml min−<sup>1</sup> was pumped/blown into chambers from the inlet, and then along with volatile and flowed out from the outlet, through the arm of the Y-tube. At the same time, fresh air was forced/pumped directly into the other arm of the Y-tube with a pump at the rate of 300 ml min−<sup>1</sup> (**Figure 1**). Adult P. japonica were released individually at the entry of the Y-tube and allowed 300 s to exhibit a behavioral response as described above (Martini et al., 2015). Y-tubes were cleaned with hot water (above 60◦C) after every set of five insects was tested. Different sets of blends (blend HC, blend LC, blend HCfP, and blend LCfP) and single dosages as defined above were used to test 50 adult P. japonica, and selected adult ladybird beetles were starved for 24 h prior to the bioassay test, and the experiment was conducted five times.

### Data Analysis

For the Y-tube olfactometer assays, we performed chi-squared tests on the mean values of five replicates. In the Supplementary Experiment, we performed Pearson Correlation to determine the correlation between the olfactory responses of P. japonica to citrus volatile blends. We used SPSS 21.0 for overall statistical analyses.

## RESULTS

### Olfactometer Assays of P. japonica

The results obtained from the olfactometer assays that tested the behavior of P. japonica to the different volatiles emitted from damaged citrus plants are shown in **Figure 2** with different effects on the predator. HC treatment was more attractive to P. japonica than treatments LC (χ <sup>2</sup> = 12.89, df = 4, P = 0.024) and HCfP (χ <sup>2</sup> = 11.63, df = 4, P = 0.040). Conversely, treatment LCfP was less attractive to P. japonica than treatments LC (χ <sup>2</sup> = 9.85, df = 4, P = 0.043) and HCfP (χ <sup>2</sup> = 17.31, df = 4, P = 0.004) (**Figure 2**).

### Headspace Collected Volatiles and Analysis

The various volatiles collected from the headspace system from the damaged and infested citrus plants were detected and quantified by GC-MS. Three major chemical compounds (Dlimonene, methyl salicylate, and beta-Ocimene) were identified as shown in **Table 1**. All three compounds were found in the various treatments in variable quantities, except methyl salicylate, which was absent in the HC treatment (**Table 1**). There was significant difference (F2, <sup>2</sup> = 65.01, P = 0.002) in the emitted quantity of D-limonene between the treatments. The amounts of D-limonene in treatments HC (13.45 ± 0.99) and LC (10 ± 0.62) did not vary but were considerably higher than in treatments HCfP (6.26 ± 0.73) and LCfP (0.70 ± 0.03) (**Table 1**). Similarly, there was significant difference (F2, <sup>2</sup> = 165.86, P = 0.0001) between the treatments in the quantities of methyl salicylate produced, which was higher in LCfP (0.097 ± 0.006) compared to HCfP (0.055 ± 0.001) and LC (0.028 ± 0.001). In addition, significantly high amounts of beta-ocimene was obtained in HC (0.62 ± 0.04) compared to other treatments (F2, <sup>2</sup> = 58.84,

#### TABLE 1 | Quantities (mean ± SE) of major compounds from the headspace of treated citrus plants.

fpls-07-01969 December 22, 2016 Time: 18:13 # 5


HC, healthy citrus; LC, Las-infected citrus; HCfP, healthy citrus infested with psyllids; LCfP, Las-infected citrus infested with psyllids; and ND, no detected compound. Means followed by different alphabets within a column represent significant differences at P < 0.05.

P = 0.002); while the lowest quantity was emitted in LCfP (0.051 ± 0.037) (**Table 1**).

### Propylaea japonica Responses to Synthetic Compounds

Three compounds, D-limonene, methyl salicylate and betaocimene, among the emitted volatiles correlated significantly with the olfactory responses of P. japonica, while the remaining were not (Supplementary Table 3). Therefore, D-limonene, betaocimene and methyl salicylate were chosen for the bioassays to further study the responses of P. japonica.

Single dosages and blends of D-limonene, methyl salicylate, and beta-ocimene were used as odor sources for the olfactometer assays. The responses of P. japonica to the three synthetic compounds are shown in **Figures 3** and **4**.

In the single dosage treatments, D-limonene concentrations of 50, 25, 17, and 10 nmol ml−<sup>1</sup> were significantly more attractive to P. japonica than fresh air (χ <sup>2</sup> = 12.67, df = 4, P = 0.013; χ <sup>2</sup> = 11.05, df = 4, P = 0.026; χ <sup>2</sup> = 9.85, df = 4, P = 0.043; χ <sup>2</sup> = 9.73, df = 4, P = 0.045, respectively). However, no significant differences were observed between fresh air and concentrations 0.1 nmol ml−<sup>1</sup> (χ <sup>2</sup> = 0.99, df = 4, P = 0.912) and 1 nmol ml−<sup>1</sup> (χ <sup>2</sup> = 1.00, df = 4, P = 0.910) (**Figure 3**). Similarly, in the single treatment of methyl salicylate, concentrations 0.64, 0.32, 0.18, and 0.15 nmol ml−<sup>1</sup> had significantly high repellent effects on P. japonica compared to fresh air (χ <sup>2</sup> = 10.45, df = 4, P = 0.033; χ <sup>2</sup> = 10.45, df = 4, P = 0.033; χ <sup>2</sup> = 9.95, df = 4, P = 0.041; χ <sup>2</sup> = 11.35, df = 4, P = 0.023, respectively) while no significant differences were observed between fresh air and 0.03 nmol ml−<sup>1</sup> (χ <sup>2</sup> = 5.00, df = 4, P = 0.288) and 0.06 nmol ml−<sup>1</sup> (χ <sup>2</sup> = 2.51, df = 4, P = 0.644) concentrations (**Figure 3**). In the single treatment of beta-ocimene, concentrations 0.04, 0.08, and 0.2 nmol ml−<sup>1</sup> showed significantly high attractive effect (χ <sup>2</sup> = 11.78, df = 4, P = 0.019; χ <sup>2</sup> = 9.69, df = 4, P = 0.046; χ <sup>2</sup> = 11.53, df = 4, P = 0.021, respectively) to P. japonica, while 1.0 nmol ml−<sup>1</sup> (χ <sup>2</sup> = 13.91, df = 4, P = 0.008) and 2.0 nmol ml−<sup>1</sup> (χ <sup>2</sup> = 10.29, df = 4, P = 0.036) were significantly repellent to the predator compared to fresh air and the three remaining concentrations (0.04, 0.08, and 0.2 nmol ml−<sup>1</sup> ) (**Figure 3**). However, there was no significant difference (χ <sup>2</sup> = 1.51, df = 4, P = 0.825) between 0.45 nmol ml−<sup>1</sup> and fresh air (**Figure 3**).

For the blend treatments, the results showed that the HC blend was more attractive to P. japonica than blends LC (χ <sup>2</sup> = 12.20, df = 4, P = 0.016) and HCfP (χ <sup>2</sup> = 10.06, df = 4, P = 0.039). Conversely, LCfP was less attractive to P. japonica than LC (χ <sup>2</sup> = 11.29, df = 4, P = 0.024) and HCfP (χ <sup>2</sup> = 13.44, df = 4, P = 0.009) (**Figure 4**).

infested with psyllids; LCfP, Las-infected citrus infested with psyllids.

### DISCUSSION

Plants effectively use their secondary metabolites to protect themselves against herbivores and pathogens (Francis et al., 2004; Zhu and Park, 2005; Boone et al., 2008; Halitschke et al., 2008; Unsicker et al., 2015). At the same time, natural enemies use these induced compounds to seek their hosts. When insect vectors feed on pathogen infected plants, the defense mechanisms of

the natural enemies, plants as well as the vectors are disturbed, due to changes in the synthesis of plant volatiles (Francis et al., 2004; Sullivan, 2005; Boone et al., 2008; Unsicker et al., 2015). In this study, we found that D. citri-infected citrus emitted less Dlimonene and beta-ocimene, compared to citrus damaged by the vector insect as well as infected by Las; these plants produced high quantities of methyl salicylate. This finding suggested that the plant pathogen, Las, interacts with its vector D. citri and regulates volatiles emission from the host plant.

In the first olfactometer test using citrus with multiple treatments, we found that volatiles emitted from HC plants attracted the most P. japonica, while volatiles from citrus damaged by the vector as well as Las were less attractive to the vector. This result is inconsistent with previous findings from the olfactometer tests conducted in wasps (Gouinguene et al., 2003; D'Alessandro and Turlings, 2005; Dannon et al., 2010; Xu et al., 2014; Mutyambai et al., 2015; Takemoto and Takabayashi, 2015). It suggests that HC plants can produce volatiles that may attractant P. japonica, but these volatiles may be down-regulated by Las and psyllids through volatile profiles modification. In addition, our results also indicated that, the two pests (Las bacteria Candidatus Liberibacter asiaticus and D. citri) may induce citrus host plants to emit repellent chemicals for P. japonica. Simultaneous insect damage and pathogenic bacterial infection of citrus host plants have been reported to induce volatiles that modify not only the behavior of its insect vector (Mann et al., 2012) but also their natural enemies (Li et al., 2014). A number of studies have demonstrated that HIPVs emitted by infected plants attract more obligatory parasites or wasps (Francis et al., 2004; Sullivan, 2005; Zhu and Park, 2005; Boone et al., 2008; Halitschke et al., 2008; Zhang et al., 2009; Unsicker et al., 2015). The results of this study showed that infection of plants with pathogens through insect vectors may alter the balance and amounts of volatiles emitted by the infected plants thus reducing their attraction to a facultative predator, P. japonica.

Among the major compounds emitted from treated citrus plants, D-limonene, methyl salicylate and beta-ocimene had significant correlation with the olfactory responses of P. japonica, while others did not (Supplementary Table 3). In the bioassay responses of P. japonica, we found that synthetic D-limonene was attractive to P. japonica while synthetic methyl salicylate repelled it. Besides, synthetic beta-ocimene had dual effects by attracting the predator at lower concentrations and repelling it at high concentrations, regardless of whether the citrus plants were damaged by psyllids and/or Las. These three synthetic compounds were therefore found to have a significant impact on P. japonica behavior particularly in the search for prey on infested citrus plants. This result suggests that D-limonene and beta-ocimene may act as attractive cues, while methyl salicylate could serve as a repellent signal for P. japonica. D-limonene is one of most significant emitted volatiles of citrus and was previously identified to attract predatory beetles (Wheeler et al., 2002; Francis et al., 2004; Himanen et al., 2005; Zhu and Park, 2005; Heil and Silva Bueno, 2007; Wei et al., 2007; Boone et al., 2008; Halitschke et al., 2008; Zhang et al., 2009; Hare and Sun, 2011; Unsicker et al., 2015). Contrary to this, some studies also demonstrated that methyl salicylate could be used as an attractant cue for some beetle species (Francis et al., 2004; Boone et al., 2008; Halitschke et al., 2008; Zhang et al., 2009; Unsicker et al., 2015). Variations in the emitted quantities of these three major compounds may therefore explain the reduction in the attractiveness of Las-infected and/or psyllidsinfested citrus host plants to P. japonica, compared to healthy plants. By comparing the three synthetic chemicals, D-limonene, beta-ocimene and methyl salicylate, we found that D-limonene is more attractive to P. japonica at concentrations 1nmol or less. On the contrary, methyl salicylate repelled the beetle, and the effect was more robust at higher concentrations. This suggests that the reduction in P. japonica attraction in treated citrus plants could have been caused by D-limonene (decreasing) and methyl

salicylate (increasing). Interestingly, beta-ocimene showed both attractive and repellent activity at different concentrations and consequently, could be used together with D-limonene to increase P. japonica attraction in the management of D. citri.

It is interesting that psyllids induced volatiles from LC were less attractive to P. japonica than volatiles emitted from a HC plant since it is know that many insect infested host plants emit HIPVs to attract natural enemies, especially parasitoid wasps. One reason in the reduction of P. japonica prey performance might be due to changes in the physical constitution of psyllids; i.e., modifications caused by Las bacteria might not support the feeding and development of P. japonica. Since P. japonica is generalist ladybird beetle (Halitschke et al., 2008), it may exercise choice feeding on other prey and not feed on infected psyllids.

It seems that the vector and Las developed a strategy through long-term evolution to deceive infested citrus plants into repelling the ladybird beetles. This could be a mechanism by which herbivores fight their predators through avoidance behavior. It therefore suggests that the ladybird beetle, P. japonica, might not be a good biological candidate to be used to control vectors in some situations. On the other hand, some studies have shown that the volatiles from different cultivars, and the emission from resistant cultivar were more attractive to natural enemies than those from susceptible cultivars (Krips et al., 1996; Michereff et al., 2011). This may be due to the variation in volatile profiles therefore the combination of predator/prey/pathogens in different plant genetic backgrounds needs to be studied. Further studies are also warranted to understand why P. japonica reacts differently to the three emitted compounds. These findings may promote the application of volatiles and specifically D-limonene in biocontrol strategies.

### REFERENCES


### AUTHOR CONTRIBUTIONS

Conceived and designed the experiments: YL, SL, and LW. Performed the experiments: YL and MH. Analyzed the data: YL, MH, and KA. Contributed reagents/materials/analysis tools: YL, SL, and KA. Wrote the paper: YL, MH, KA, and LW.

### FUNDING

This project is funded by National Natural Science Foundation of China (31371998), Key Projects of Science and Technology of Fujian Province (2016N0005) and a Research Fund for the Doctoral Program of Higher Education of China (20123515110003).

### ACKNOWLEDGMENTS

The authors are grateful to Prof. Minsheng You (Institute of Applied Ecology, FAFU, China) for his kind cooperation. This project was funded by the National Natural Science Foundation of China (31371998) and Key projects of Science and Technology of Fujian Province (2016N0005). The authors also thank LetPub (www.letpub.com) for its linguistic assistance during the preparation of this manuscript.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2016.01969/ full#supplementary-material


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Lin, Lin, Akutse, Hussain and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpls-07-01969 December 22, 2016 Time: 18:13 # 8

fpls-08-01597 September 20, 2017 Time: 15:22 # 1

# Pre-infestation of Tomato Plants by Aphids Modulates Transmission-Acquisition Relationship among Whiteflies, Tomato Yellow Leaf Curl Virus (TYLCV) and Plants

#### Xiao L. Tan1,2,3† , Ju L. Chen<sup>2</sup>† , Giovanni Benelli<sup>4</sup> , Nicolas Desneux<sup>5</sup> , Xue Q. Yang<sup>6</sup> , Tong X. Liu<sup>3</sup> \* and Feng Ge<sup>1</sup> \*

#### Edited by:

Gero Benckiser, Justus-Liebig-Universität Gießen, Germany

#### Reviewed by:

Yuanyuan Song, Fujian Agriculture and Forestry University, China Yaobin Lu, ZheJiang Academy of Agricultural Sciences, China

#### \*Correspondence:

Tong X. Liu txliu@nwsuaf.edu.cn Feng Ge gef@ioz.ac.cn

†These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 05 December 2016 Accepted: 31 August 2017 Published: 22 September 2017

#### Citation:

Tan XL, Chen JL, Benelli G, Desneux N, Yang XQ, Liu TX and Ge F (2017) Pre-infestation of Tomato Plants by Aphids Modulates Transmission-Acquisition Relationship among Whiteflies, Tomato Yellow Leaf Curl Virus (TYLCV) and Plants. Front. Plant Sci. 8:1597. doi: 10.3389/fpls.2017.01597 <sup>1</sup> State Key Laboratory of Integrated Management of Pest Insects and Rodents, Institute of Zoology, Chinese Academy of Sciences, Beijing, China, <sup>2</sup> State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing, China, <sup>3</sup> State Key Laboratory of Crop Stress Biology for Arid Areas and the Key Laboratory of Crop Pest Management on the Losses Plateau of Ministry of Agriculture, Northwest A&F University, Yangling, China, <sup>4</sup> Department of Agriculture, Food and Environment, University of Pisa, Pisa, Italy, <sup>5</sup> INRA (French National Institute for Agricultural Research), UMR 1355-7254 Institute Sophia Agrobiotech, CNRS, Université Côte d'Azur, Sophia Antipolis, France, <sup>6</sup> Key Laboratory of Economical and Applied Entomology of Liaoning Province, College of Plant Protection, Shenyang Agricultural University, Shenyang, China

Herbivory defense systems in plants are largely regulated by jasmonate-(JA) and salicylate-(SA) signaling pathways. Such defense mechanisms may impact insect feeding dynamic, may also affect the transmission-acquisition relationship among virus, plants and vectoring insects. In the context of the tomato – whitefly – Tomato Yellow Leaf Curl Virus (TYLCV) biological model, we tested the impact of pre-infesting plants with a non-vector insect (aphid Myzus persicae) on feeding dynamics of a vector insect (whitefly Bemisia tabaci) as well as virus transmission-acquisition. We showed that an aphid herbivory period of 0–48 h led to a transient systemic increase of virus concentration in the host plant (root, stem, and leaf), with the same pattern observed in whiteflies feeding on aphid-infested plants. We used real-time quantitative PCR to study the expression of key genes of the SA- and JA-signaling pathways, as well as electrical penetration graph (EPG) to characterize the impact of aphid pre-infestation on whitefly feeding during TYLCV transmission (whitefly to tomato) and acquisition (tomato to whitefly). The impact of the duration of aphid pre-infestation (0, 24, or 48 h) on phloem feeding by whitefly (E2) during the transmission phase was similar to that of global whitefly feeding behavior (E1, E2 and probing duration) during the acquisition phase. In addition, we observed that a longer phase of aphid pre-infestation prior to virus transmission by whitefly led to the up-regulation and down-regulation of SA- and JA-signaling pathway genes, respectively. These results demonstrated a significant impact of aphid pre-infestation on the tomato – whitefly – TYLCV system. Transmission and acquisition of TYLCV was positively correlated with feeding activity of B. tabaci, and both were mediated by the SA- and JA-pathways. TYLCV concentration

**457**

during the transmission phases was modulated by up- and down-regulation of SA- and JA-pathways, respectively. The two pathways were inconsistent during the acquisition phase; SA- related genes were up-regulated, whereas those up- and down-stream of the JA pathway showed a more complex relationship. These findings enhance our understanding of plant – herbivore – virus interactions, which are potentially important for development of ecologically sound pest and pathogen management programs.

Keywords: TaqMan Real-Time PCR, systemic induced defense, Bemisia tabaci, Myzus persicae, electrical penetration graph, jasmonic acid, salicylic acid, signaling pathways

### INTRODUCTION

fpls-08-01597 September 20, 2017 Time: 15:22 # 2

In recent decades, there have been many studies on the ecological and physiological mechanisms underlying plantmediated interactions among herbivore arthropods and plant pathogens, particularly regarding plant defense responses (Al-Bitar and Luisoni, 1995; Beale et al., 2006; Bleeker et al., 2011; Baldin et al., 2013; Mouttet et al., 2013). Although it is well established that infection by pathogenic microorganisms may indirectly influence the behavior of the subsequent herbivore insects of the plant (Beale et al., 2006; Goggin, 2007; Mouttet et al., 2011), there has been less work on the effect of insect herbivore infestation on plant pathogens (Czosnek et al., 1988; Mouttet et al., 2011, 2013). The defense responses induced in plants after incidence of herbivory is central to the longstanding co-evolutionary system of host plant, pathogenic microorganism and herbivore arthropod (including vectors). Furthermore, these responses often impact prevalence of disease and pests, as well as the degree of damage to plants (Mayer et al., 2002; Cory and Hoover, 2006; Huot et al., 2013). However, there are key gaps in understanding of host plant – vector insect – pathogen interaction systems. For example, it is unclear how phytohormone-mediated plant defenses which are induced by insect infestations or virus transmission may influence behaviors of virus-carrying insects. Thus, understanding the mechanisms by which virus transmission occurs in plants showing herbivore-induced defense may be helpful for development of novel pest management strategies.

There have been numerous studies describing various defense systems that are induced in plants in response to damage caused by pathogens and herbivore arthropods (Karban and Baldwin, 1997; Huot et al., 2013). Key signaling pathways identified as regulators of plant defense are Jamonate (JA) for insect infestations and Salicylate (SA) for pathogen infestations, although the relationship between gene expression and function can be complex (Thaler et al., 2010). Further, understanding signal interactions within a plant – insect – pathogen system requires consideration of ecological context and the temporal sequence of interactions, such as the effect that an initial herbivore attack might have on the system during subsequent infestations. These types of interactions are more representative of incidents in natural and agricultural settings, and thus must be studied to develop more effective solutions for pest/pathogen management (Agrawal et al., 2000; Howe and Jander, 2008; Taggar and Gill, 2016).

Tomato Yellow Leaf Curl Virus is a cosmopolitan geminivirus which can infect dicotyledons, including tomato (Goggin, 2007; Heil, 2008), and is usually transmitted by the globally invasive whitefly Bemisia tabaci (Green and Ryan, 1972; Rubinstein and Czosnek, 1997). The transmission of TYLCV by B. tabaci causes huge losses in tomato yield in many regions, including China (Pan et al., 2012). The acquisition of TYLCV by B. tabaci occurs when a non-infected whitefly feeds on the phloem of a TYLCVinfected plant, and transmission occurs when a virus infected B. tabaci feeds on a previously healthy plant (Czosnek et al., 2002). It has been shown that within this defense system, a prior attack by aphid Myzus persicae can influence the feeding strategy and population dynamics of subsequent herbivores (Tan et al., 2014). It is unclear whether TYLCV transmission and acquisition are also affected by the feeding behavior of whiteflies, or whether these are influenced by aphid pre-infestation.

Plants infected by TYLCV may attract more whiteflies, and generally create conditions suiting colonization and suppression of the virus (Stout et al., 2006; Luan et al., 2013). Previous research has shown that plant defense responses induced by aphids are similar to those induced by pathogens (Thompson and Goggin, 2006). Aphid attacks can activate the SA-signaling pathway and suppress the JA-signaling pathway of the plant, and consequentially attract more whiteflies (via JA) and suppress TYLCV (SA) (Zarate et al., 2007; Verhage et al., 2010). Increasing this attraction can lead to aggregation of whiteflies and increase their feeding, both of which may increase TYLCV transmission and acquisition. Therefore, evaluation of the SA- and JA-pathway under aphid pre-feeding may help us to understand the mechanism by which induced defense is regulated.

In order to better understand the interactions in the plant- herbivore-virus system under conditions of sequential herbivory, we tested the hypothesis that pre-infestation by aphid M. persicae affects the expression of hormone-related genes and the SA- and JA-signaling pathways, consequentially impacting feeding behavior of subsequent whiteflies and thus the transmission and acquisition of TYLCV. TaqMan Real-Time PCR has been used for rapid and efficient quantitative detection of organismal DNA. It has proven highly sensitive for detecting DNA viruses in insect vectors and plants (Lei et al., 1987; Kunkel and Brooks, 2002), and there have been reports on using this technology for detection of TYLCV in tomato (Liu et al., 2007). We use Real-Time PCR to quantify TYLCV in infected tissues of tomato plant as well as in B. tabaci. We tested if pre-infestation with M. persicae (i) affect the feeding behavior of B. tabaci, (ii) modulate TYLCV transmission and acquisition between host plant and B. tabaci, and (iii) impact the expression of SAand JA-signaling pathway genes (and potential link with vector feeding behavior and virus transmission and acquisition).

### MATERIALS AND METHODS

### Virus, Insects, and Plants

fpls-08-01597 September 20, 2017 Time: 15:22 # 3

Tomato plants, Solanum lycopersicum L. (var. Baofen-F1, 2008, Changfeng Institute of Vegetable, Lintong, Xi'an, Shaanxi, China) were cultivated in plastic trays (50.0 × 25.0 × 15.0 cm), with eight plants per tray. Seedlings, about 4–5 cm in height, were transplanted into plastic pots (20 cm in depth and 15 cm in diameter) and placed in clean cages (60 × 60 × 60 cm, plastic frame, screened with 120-mesh nylon yarn net). Plants used in all experiments were approximately 30 cm in height with 4–6 true, fully expanded leaves. The experiments were conducted in an environmental chamber at 25 ± 2 ◦C, 65 ± 5% relative humidity and photoperiod of 16/8 h of light/dark with artificial lighting of 3500 lux. About 15,000 M. persicae individuals were collected from pepper plants (Capsicum annuum L. var. Jingyuan New Prince, provided by Qing County Modern Agricultural Technology Promotion Center, Hebei Province) in a greenhouse on the campus of Northwest A&F University, Yangling, Shaanxi (116◦ 220 4200E and 39◦ 590 5800N) in March 2011. The aphids were maintained on tomato plants under the laboratory conditions described above. All tested aphids were offspring (>five generations) of the original collected specimens.

Bemisia tabaci whiteflies (507 males and 631 females) of Middle East-Asia Minor 1 (MEAM 1) species were collected from tomato plants in a greenhouse and their identity confirmed by mitochondrial DNA marker analysis as described by Chu et al. (2007). They were reared on cotton plants under the same conditions described for green peach aphids during March and April 2011. Newly emerged adult whiteflies were used in all experiments after they had occupied the tomato plants for more than five generations. TYLCV-infected tomato plants were provided by the Natural Enemy Application and Research Laboratory, Institute of Plant and Environment Protection, Beijing Academy of Agriculture and Forestry Sciences, Beijing, China. The plants were cultivated in the State Key Laboratory of Crop Stress Biology of Northwest A&F University using ventilated culturing cages (60 × 60 × 60 cm, plastic frame, screened with 120-mesh nylon yarn net) containing 100 pairs of whitefly adults. The whiteflies were moved to the cages containing 15 tomato plants with 4–6 leaves each. TYLCVinfected plants were tested after 20 days. Bemisia tabaci specimens were obtained from cotton leaves from which pseudo pupae had been cut off and the petiole had been inserted in a glass bottle filled with water and placed in a plastic cage (13 × 13 × 30 cm; 100-mesh nylon yarn net door on one side). Non-viruliferous whiteflies were maintained on cotton plants (Gossypium arboreum cv. SN407) that had been identified as non-infected hosts for TYLCV. Newly emerged Non-viruliferous adult whiteflies were used in experiments after they were moved from cotton plants and had occupied the tomato plants for more than five generations. Newly emerged whiteflies were collected and released onto the TYLCV-infected tomato plants in ventilated culturing cages for a duration of 48 h, after which the whiteflies were used in the experiments; previous studies indicate that the viral concentration in whitefly peaks after 12–48 h of feeding on tomato plants infected with TYLCV (Mason et al., 2008; Luo et al., 2010; Lucas-Barbosa et al., 2011).

### DNA Extraction and Virus Detection

Leaves of five TYLCV-infected tomato plants were ground into powder in a pre-chilled motor in liquid nitrogen. Genomic DNA was extracted from 40 mg powder using the WizardTM Genomic DNA purification kit (Promega Corporation, Madison, WI, United States) according to manufacturer's instructions. Genomic DNA from healthy leaves of non-infected tomato plants were used as control. The extracted DNA was used as a template for PCR reaction or kept at −20◦C for later use.

Based on the complete genomic sequence of TYLCV (GenBank accession number: NC\_004005.1), two primer pairs: F1 (5<sup>0</sup> -CCAATAAGGCGTAAGCGTGTAGAC-3<sup>0</sup> ) and R1 (5<sup>0</sup> -A CGCATGCCTCTAATCCAGTGTA-3<sup>0</sup> ), F2 (5<sup>0</sup> -TCCCCTCTAT TTCAAGATAACAGAACA-3<sup>0</sup> ) and R2 (5<sup>0</sup> -TCTGAGGCTGTAA TGTCGTCCA-3<sup>0</sup> ), and TaqMan probe (5<sup>0</sup> -FAM- CCCAATGCC TTCCTG-MGB-3<sup>0</sup> ) were designed by GeneCore Bio Technologies Co. Ltd. (Shanghai, China) to amplify 543 and 161 bp regions of the TYLCV gene. Actin (ACT) gene of B. tabaci (GenBank accession number: AF071908) and S. lycopersicum (GenBank accession number: AB199316) were used as internal controls, and 174 and 191 bp regions of BtACT and SlACT genes were amplified using F3 (5<sup>0</sup> -TCCTTCCAGCCATCCTTCTTG-3 0 ) and R3 (5<sup>0</sup> -CGGTGATTTCCTTCTGCATT-3<sup>0</sup> ) (Li et al., 2013) and F4 (5<sup>0</sup> -TGGTCGGAATGGGACAGAAG-3<sup>0</sup> ) and R4 (50 -CTCAGTCAGGAGAACAGGGT-3<sup>0</sup> ) (Chen et al., 2017) as primer pairs, respectively.

The PCR reactions were conducted in a C1000 Thermal Cycler (BioRad, Hercules, CA, United States) using F1 and R1 as primer pair in 30 µL of mixture containing 1 µL of template DNA, 1 µL of each primer (10 µM), 3 µL of 10 × PCR buffer (150 mM Tris–HCl, 500 mM Tris–HCl, pH 8.0 with 25 mM MgCl2), 1 µL of dNTP Mix (2.5 mM), 0.5 µL of Premix Ex Taq DNA polymerase (5 U/µL, Takara Biotechnology Co. Ltd., Dalian, China) and 22.5 µL of double distilled H2O. The PCR conditions were 10 min at 94◦C, followed by 40 cycles of 1 min at 94◦C, 1 min at 57◦C and 2 min at 72◦C and a final extension at 72◦C for 10 min. The PCR products were separated on a 2% agarose gel and purified using the Biospin Gel Extraction Kit (Bioer Technology Co. Ltd., Hangzhou, China) as described by the manufacturer. The recycled PCR products were ligated into the pMD-19 T vector (Takara) and transformed into DH5α cells. After blue-white selection, positive clones were sequenced by the Shanghai Sunny Biotech Co. Ltd., China. The sequenceverified clones were cultivated and plasmids were extracted. The concentration of extracted plasmid DNA was measured using a UV-Vis spectrophotometer (NanoDrop 2000 c). Copy number was determined according to the equation: copy number/mL = 6.02 × 10<sup>23</sup> (cn/mol) × plasmid concentration (g/mL)/MW (g/mol). The standard curve of TYLCV coat protein gene was obtained by using serial 10-fold diluted plasmids fpls-08-01597 September 20, 2017 Time: 15:22 # 4

(9.16 × 10<sup>10</sup> to 9.16 × 10 copies) as templates, and by using F2 and R2 as primer pairs and TaqMan fluorescence probe (**Figure 1**). Amplifications were performed in 25 µl volumes containing 12.5 µL of Premix Ex Taq (Probe qPCR), 2 µL of template, 0.5 µL of each 10 µM F2 and R2 primers using thermocycler conditions as follows: 95◦C for 10 s, followed by 40 cycles of 10 s at 95◦C and 30 s at 58◦C. Then, the linear equation of log concentration versus Ct curve was generated, where Ct values were plotted from 10-fold serial dilutions of the plasmid DNA. The estimation of copy number in samples was performed by computing the estimates of linear regression coefficients. The quantifications of DNA samples were calculated based on the fluorescence (1Rn) values. All samples were run in duplicate by TaqMan Real-Time PCR assay for accuracy of results.

High specificity was determined for primers (**Figure 1**). PCR amplification showed that the product of TYLCV DNA corresponded to a 161 bp band, and lacking primer dimers. The data for the estimation of a standard curve for absolute quantitative and TaqMan Real-Time PCR assays were obtained and analyzed by an iQTM5 Multicolor Real-Time PCR Detection System. Thus, sample Ct values were obtained automatically. Ct values and the logarithm of TYLCV DNA copy number

DNA marker. Standards (copy numbers per µl of target DNA): 1, 9.16 × 1010; 2, 9.16 × 10<sup>9</sup> ; 3, 9.16 × 10<sup>8</sup> ; 4,9.16 × 10<sup>7</sup> ; 5, 9.16 × 10<sup>6</sup> ; 6, 9.16 × 10<sup>5</sup> ; 7, 9.16 × 10<sup>4</sup> ; 8, 9.16 × 10<sup>3</sup> ; 9, 9.16 × 10<sup>2</sup> ; 10, 9.16 × 10<sup>1</sup> ; 11, 9.16 × 10<sup>0</sup> . (B) Standard curve analysis: the relationship between TYLCV concentration and corresponding Ct value is shown by linear regression equation of Ct (x) transformed into log (starting quantity).

(log) had a linear relationship. The logarithmic average of TYLCV DNA copy number (log) was calculated through a standard curve, giving y = −3.4015x + 42.465, slope −3.4015, R <sup>2</sup> = 0.9973 from each sample of TYLCV, an amplification efficiency (E) of 96.8%, and showed a robust linear relationship (**Figure 1B**).

### Impact of Aphid Pre-infestation Duration on TYLCV Transmission and Acquisition by Whiteflies on Tomato Plants

We assessed pre-infestation effects on viruses, separately for the processes of transmission and acquisition, and each by varying the duration of pre-infestation. We sequentially added aphids then whiteflies, then measured virus concentration. For transmission, Each tomato plant with 4–5 expanded leaves was infested separately with 60 fourth-instar aphid nymphs, then the aphids were removed using a soft brush after 12, 24, or 48 h. Aphids were randomly assigned to each plant. Sixty viruliferous adult whiteflies were introduced throughout the tomato plant (with sufficient space for flight in the mini cage) and fed for 24 h after removal of aphids. The whiteflies were removed using an aspirator. A group of five tomato plants (including roots) was collected, the roots washed with sterile distilled water, and stored at −80◦C. Muniyappa et al. (2000) showed that the virus can be transmitted from whiteflies that have fed on TYLCVinfected plants for 24 h, to uninfected plants after exposure of only 20 min. Jiu et al. (2006) reported that the shortest time of transmission of TYLCV to plants by whiteflies was 15–30 min. Thus, in the present study, a 24 h feeding time was thought sufficient to transmit TYLCV to tomato plants. The experiment was repeated five times, each time with five tomato plants. Control tomato plants were not infested with aphids. Leaves of five TYLCV-infected tomato plants were ground into powder in a pre-chilled motor in liquid nitrogen. Virus concentration was assessed by DNA extraction and PCR; the TaqMan PCR amplification system followed the methods described above. For virus acquisition, we also measured TYLCV for B. tabaci via feeding on virus-infested tomato plants. Viruliferous whitefly were first used to transmit virus to the tomato plant, followed by aphid preinfestation and subsequent whitefly exposure. Tomato plants treated as above were individually placed in a square plastic cage (13 × 13 × 30 cm), and 60 viruliferous adult whitefly introduced using an aspirator. The whitefly adults were removed after feeding for 24 h, and then the plants were pre-infested with aphids as above. Fifty newly emerged non-viruliferous whitefly adult females were introduced into the cage to acquire virus from the infected tomato plant after removing the aphids. After feeding for 24 h, 40 whiteflies per plant from each group of five tomato plants were collected and stored at −80◦C. Control tomato plants were not infested with aphids. The experiment was replicated five times, each time with five different tomato plants. For virus detection in whitefly, DNA was extracted using the WizardTM Genomic DNA purification kit (Promega, Madison, WI, United States) from 200 whiteflies of a group of five tomato plants, which were ground into homogenate in 600 µL of nuclear pyrolysis liquid. The extracted DNA was diluted 10 times for fpls-08-01597 September 20, 2017 Time: 15:22 # 5

TaqMan PCR assay or saved at −28◦C for later use. PCR amplification system and response procedures were as reported above.

### Effects of Whitefly Feeding Behavior on Virus Transmission and Acquisition

### Tomato Plants Were Treated as Outlined in Section "Impact of Aphid Pre-infestation Duration on TYLCV Transmission and Acquisition by Whiteflies on Tomato Plants"

Each tomato plant with 4–5 expanded leaves was infested separately with 60 fourth-instar aphid nymphs, then the aphids were removed using a soft brush after 24 or 48 h. Aphids were randomly assigned to each plant. Because the concentration of TYLCV was similar for aphid feeding of 12 and 24 h, this test used tomato plants infested with aphids for 24 or 48 h. The experiment was repeated three times, each time with eight whiteflies recorded.

For transmission, whitefly feeding during the virus transmission phase was recorded. Feeding behavior was assessed for newly emerged viruliferous whiteflies from TYCLV infected tomato plants. For acquisition, whitefly probing behavior during virus acquisition (from tomato) was recorded for newly emerged whiteflies from non-infected tomato.

Whitefly feeding was monitored using the electrical penetration graph (EPG) method (Liu et al., 2013). In EPG, the insect and its host plant creates an electrical circuit that is completed when the mouthparts of the phloem-feeding insect penetrate the plant. The resulting electrical signals are amplified and digitized. Fluctuations in voltage and electrical resistance are recorded on computer, and can be matched to specific feeding events. Prior to recording, whiteflies were immobilized by placing them in a freezer for several minutes, then waiting 1 h without feeding. The recording electrode of the EPG system (a thin gold wire of length 2 cm and diameter 10 µm), was connected to the whitefly's dorsum with silver paint. Electrode adhesion was conducted under a microscope, and then the whitefly was put on the abaxial surface of the leaf and allowed to crawl freely. The wired whiteflies were then connected to the Giga-8 probe input, and another electrode was inserted into the soil at the base of the tomato plant. The whole system was placed into an electrically grounded Faraday cage to guard against external electrical noise. The EPG signals were digitized with a DI710-UL analog-to-digital converter, and the output was acquired and stored with PROBE3.4 software. The data were analyzed subsequently with STYLET 2.0 software. Phases of feeding behavior are described by 14 EPG parameters related to the pathway phase (C waveform, F waveform and potential drops), phloem phase (E waveform) and xylem phase (G waveform). Phloem subphase 1 (E1: salivary secretion into sieve elements) and phloem subphase 2 (E2: phloem ingestion) were extracted from each recording and compared among treatments. 12 h of EPGs were continuously recorded for each replicate. All experiments were carried out under artificial light (1,500 lx) with a 16 h light/8 h dark regime and at 25◦C ± 2 ◦C, at 70% relative humidity (RH).

### Aphid Pre-infestation Effects on Tomato Defense-Related Genes before Transmission and Acquisition

Each treatment combination was replicated three times with eight plants for biological replicates, and each biological replicate contained three technical replicates. The expression of target genes involved in the JA-signaling [lipoxygenase D (LOXD), proteinase inhibitor II (PI-II)] and SA-signaling [phenylalanine ammonia lyase (PAL), and β-1,3-glucanase (PR2b)] pathways were determined.

The RNeasy Mini Kit (Qiagen, Valencia, CA, United States) was used to isolate total RNAs from tomato leaves (100-mg samples) before viruliferous whitefly and non-infected whitefly feeding, and 1 µg of RNAs were used to generate the cDNAs. The mRNA amounts of seven target genes were measured by real-time quantitative PCR.

Specific primers for each gene were designed from the tomato plant expressed sequence tag sequences using PRIMER5 software (**Table 1**). The reaction mixture contained the following reagents: 1 µL of first-strand cDNA, 2 µL of reaction buffer, 1 µL of MgCl<sup>2</sup> (50 mM), 0.3 µL of SYBR Green, 0.2 µL of PLATINUM Taq polymerase, 0.5 µL of dNTPs (10 mM) and 10 µM of each primer in a total volume of 20 µL. Reactions were carried out on the Mx 3500P detection system (Stratagene), and PCR was performed under the following conditions: 2 min at 95◦C; followed by 40 cycles of 30 s at 95◦C and 30 s at 60◦C. Following the real-time quantitative PCR, a melting curve was generated by gradually increasing the temperature to 95◦C to test the homogeneity of PCR products. Relative standard curves for the transcripts of every gene were prepared with linear gradient (five-fold) cDNA as template and were included within each real-time quantitative PCR. The relative level of each target gene was standardized by comparing the copy numbers of target mRNA with SlACT and SlTUB CT values, which remained constant under different treatment conditions. The fold-changes of target genes were calculated using the 211C<sup>T</sup> method. All PCR runs were performed with three technical replicates. The three

TABLE 1 | Overview of the target genes used in this study, showing their GenBank accession numbers and the primer pair used for qRT-PCR.


<sup>∗</sup>SlTUB and SlActin are reference genes.

independent biological replicates, each containing eight plants, were analyzed.

### Data Analysis

fpls-08-01597 September 20, 2017 Time: 15:22 # 6

One-way analysis of variance (ANOVA) was used to analyze the data of log copy number of TYLCV DNA for the effects of duration of infestation by M. persicae on the transmission and acquisition of TYLCV, whitefly probing profiles (EPG) and defense gene expressions. ANOVA was performed with IBM SPSS statistics version 20.0 (SPSS, Chicago, IL, United States). An independent t-test was used to test the effects of aphid feeding duration on defense gene expression during transmission and acquisition. All data were checked for normality and equality of residual error variances, and were transformed (log or square root) where appropriate to satisfy assumptions of ANOVA. General Linear Mixed Model (GLMM) was used to analyze the effects of the duration of aphid pre-infestation, plant position and their interaction on TYLCV. In both cases, when a significant effect was found, Duncan's test was used as post hoc test to compare the log copy number of TYLCV DNA (P < 0.05).

### RESULTS

### Impact of Aphid Pre-infestation Duration on TYLCV Transmission and Acquisition by Whiteflies

The GLMM tests highlighted significant differences in TYLCV DNA concentrations for different durations of aphid preinfestation (F = 12.622, df = 3176, P < 0.05), and among different tomato plant tissues (F = 4.107, df = 2177, P < 0.05). Further, there was a significant interaction between duration of aphid pre-infestation, tomato plant tissues, and concentration of TYLCV DNA (F = 7.208, df = 6, P < 0.05) (**Figure 2**). The concentration of TYLCV DNA in B. tabaci after feeding on virus-infected tomato was influenced by duration of aphid pre-infestation (F = 3.769, df = 3176, P < 0.05; **Figure 3**).

### Aphid Pre-infestation Duration Effects on Feeding Behavior of Whiteflies during Transmission and Acquisition Phase Transmission

Twenty-four hour aphid infestation significantly increased the time that viruliferous whiteflies spent in the salivation phase [E1] (F = 3.586, P = 0.033, df = 2, 69) and duration of phloem feeding [E2] (F = 4.296, P = 0.017, df = 2, 69; **Figure 4A**). By contrast there was no significant effect on the total duration of probing by whiteflies (F = 2.658, P = 0.077, df = 2, 69; **Figure 4A**). For 48-h aphid infestation, the duration of E2 was similar to plants not infested with aphids; whereas significant differences are observed at 48 h for the duration of E1 and total duration of probing, compared to 0 h.

FIGURE 2 | TYLCV DNA concentrations in different tomato plant parts after feeding by TYLCV-infested Bemisia tabaci according to various Myzus persicae pre-infestation duration treatments. The error bars are standard errors of the mean. Colored letters indicate significant differences of DNA concentration among different aphid pre-infestation durations in each tomato plant part, and '<sup>∗</sup> ' indicates significant differences among three plant parts in the same pre-infestation duration according to Duncan's test at P = 0.05.

### Acquisition

Pre-infestation with M. persicae significantly impacted several aspects of whitefly feeding on virus-infected leaves (E1: F = 4.836, P = 0.011; E2: F = 6.173, P = 0.003; and total duration of probing: F = 3.669, P = 0.031; all df = 2, 69; **Figure 4B**). 24 h of aphid infestation significantly increased E1, E2 and the total duration of probing by whiteflies. For 48-h aphid infestation, the durations of E1, E2 and of probing were similar to plants not infested with aphids. This trend was similar to the virus concentration in whiteflies with increasing aphid feeding duration on viruliferous tomato plants (**Figure 4B**).

### Effects of Aphid Pre-infestation on Defense Genes in Tomato Prior to Transmission and Acquisition

The relative mRNA expressions of PAL (before transmission: F = 9.213, P = 0.0001, df = 2, 69; and before acquisition: F = 4.727, P = 0.012, df = 2, 69) and PR2b (before transmission: F = 19.881, P = 0.0001, df = 2, 69; and before acquisition: F = 2.553, P = 0.085, df = 2, 69) positively correlated with aphid feeding time prior to transmission by whiteflies (**Figure 5**). Similarly, prior to acquisition by whiteflies, expressions of PAL and PR2b in plants with 24 h of aphid feeding were higher than for those without aphids, but similar to those with 48 h of aphid feeding. Moreover, expressions of both genes were higher prior to acquisition by whiteflies than prior to transmission by whiteflies: PAL for 0 h, t = −2.924, df = 33.022, P = 0.006; for 24 h, t = −4.709, df = 35.63, P = 0.0001; and for 48 h, t = −2.132, df = 35.117, P = 0.04; and PR2b for 0 h, t = −6.654, df = 23.494, P = 0.0001; for 24 h, t = −4.547, df = 26.518, P = 0.0001; and for 48 h, t = −3.111, df = 35.117, P = 0.003 (**Figure 5**).

fpls-08-01597 September 20, 2017 Time: 15:22 # 7

In contrast, the relative mRNA expressions of JA-signaling genes LOXD (before transmission: F = 12.418, P = 0.00001, df = 2, 69; and before acquisition: F = 12.566, P = 0.00001, df = 2, 69) and PI-II (before transmission: F = 20.617, P = 0.00001, df = 2, 69; and before acquisition: F = 2.611, P = 0.081, df = 2, 69) were significantly reduced with increase in aphid feeding time, except for PI-II in plant prior to acquisition by whiteflies. The expression of LOXD in plants was lower prior to acquisition than prior to transmission by whiteflies, except for during an aphid infestation period of 24 h (LOXD for 0 h, t = 2.497, df = 46, P = 0.016; for 24 h, t = 1.666, df = 45.831, P = 0.103; and for 48 h, t = −2.741, df = 28.652, P = 0.01). However, expression of PI-II was lower in the non-infested plant prior to acquisition but not prior to transmission (PI-II, 0 h: t = 2.304, df = 46, P = 0.026; 24 h: t = −0.532, df = 30.834, P = 0.598; 48 h: t = −4.889, df = 27.284, P = 0.00001) (**Figure 5**).

### DISCUSSION

This study is the first investigation on impact of herbivore preinfestation on virus transmission and acquisition by subsequent herbivores, in terms of key interactions such as feeding behavior and plant defenses. We demonstrated that virus transmission and acquisition between tomatoes and vector whiteflies under aphid pre-infestation follow two different mechanisms. The SA- and JA-signaling pathways had different roles and the feeding behavior of whiteflies dominated the acquisition process. We believe that this overlooked issue will be of wide ecological interest, since knowledge of the effects of herbivorous feeding activity on the acquisition and transmission of viruses can shed light on complex plantmediated relationships among herbivores, arthropod vectors and pathogens (Mehta et al., 1994; Messina et al., 2002; Beale et al., 2006; Mithöfer and Boland, 2012). Feeding behavior (salivation/phloem ingestion/duration of feeding) of viruliferous and non-infected B. tabaci was also influenced by different aphid pre-infestation times, in contexts of both transmission and acquisition. Furthermore, we found that the expression of two SA-signaling pathway genes may increase when aphid pre-infestation period is prolonged, and expression of the related JA-signaling pathway is reduced, especially during the transmission phase. Previous studies have shown that attack by aphids can inhibit the JA-signaling pathways and consequentially reduce negative impacts of herbivory (Papayiannis et al., 2009, 2010; Giordanengo et al., 2010). Plant defense responses induced by sap-sucking herbivores or pathogens are often similar, since both responses can act to resist both insects and pathogens (Thaler et al., 2002; Murugan and Dhandapani, 2007). The defense signaling pathways relying on SA and JA may vary with injury time (Preston et al., 1999; Nombela et al., 2009). Tobacco plants infected with TMV developed SAR, while damage response via JA was not normally activated, possibly because infection by TMV increased SA levels, inhibiting the JA pathway (Ohgushi, 2005). However, this might reduce JA/ET-associated insect resistance (Salzman et al., 2005).

### TYLCV Transmission/Acquisition

The pattern in transmission of virus from viruliferous B. tabaci to non-infected tomato was consistent among different plant organs tested (attacked leaves, stems, and roots), as well as under different durations of aphid pre-infestation. When measured directly on the aphid-attacked leaf, the amount of TYLCV initially increased, and was followed by a decline, with extension of period of pre-infestation. The stem and the root showed similar tendencies; the pre-infestation was found to stimulate systemic resistance, resulting in a more effective response to subsequent attacks. Generally, herbivore feeding action leads to a diverse and synergistic systemic defense, performed across different tissues of the plant (Pico et al., 1999; Schilmiller and Howe, 2005; Jung et al., 2009). The defense pathways may be induced in tissues away from the position of attack, although this reaction is somewhat delayed (Kessmann et al., 1994). In this study, TYLCV DNA concentration in roots showed a belated increase, which demonstrated that a period of time is required for systemic transmission of TYLCV within tomato plants. Thus, a threshold period is needed for transmission of injury messages within the plant and for producing the defense compounds responsible for defense (Sauge et al., 2002). The acquisition of TYLCV by whitefly from TYLCV-infected tomato plants also displayed a tendency to rise and then decline. When increasing duration of aphid infestation beyond the 12 h threshold, the TYLCV concentration in whiteflies began to decrease. Aphids feeding on tomato plants for sufficient time can induce a plant defense response which inhibits whiteflies from acquiring TYLCV. Our result indicated that extension of the pre-infestation phase may increase virus transmission and acquisition by whitefly in the short term. This result might be explained by feeding efficacy, as influenced by the plant induced defense response.

### Feeding Behavior of Bemisia tabaci According to Aphid Pre-infestation

fpls-08-01597 September 20, 2017 Time: 15:22 # 8

As our results show, the feeding behavior strategies of either TYLCV infected B. tabaci on non-infected tomato (virus transmission) or non-infected whitefly on viruliferous plants (virus acquisition) were both influenced by duration of pre-infestation by M. persicae. We found that the duration of phloem ingestion (E2) of B. tabaci may increase at 24 h and decrease at 48 h, with increasing of duration of pre-infestation. However, the total probing duration for whitefly was inconsistent between transmission/acquisition treatments. The transmission and acquisition of virus between whitefly and tomato plant may be due to different feeding strategies; for transmission of virus from whitefly to plant, net virus infestation is associated with salivation (E1). Thus, as indicated by the amount of TYLCV DNA in the host plant (**Figure 2**), virus infection in tomato may have a corresponding increase with salivation by viruliferous B. tabaci, although unaffected by total duration of probing. In contrast, the acquisition of virus by whitefly from infected plants was mainly related to phloem ingestion (E2). Clearly, our results showed similar responses to variation in aphid pre-infestation between the amount of TYLCV in whitefly and the duration of their phloem ingestion (**Figures 3**, **4**). It seems that virus transmission/acquisition is influenced by the whitefly's different feeding strategies. The feeding of whiteflies may be inhibited by the plant defense response triggered by aphid pre-infestation. Empirical reports indicate that defensive materials (like JA, SA and its derivants) may suppress phloem ingestion (Kehr, 2006). Thus, the non-viruliferous whitefly showed a reduced ability to combat this response, and struggle to maintain normal phloem

fpls-08-01597 September 20, 2017 Time: 15:22 # 9

ingestion feeding time. Interestingly, the virus infection may be quite specific in the way it regulates feeding behavior, with the whitefly enhancing only key processes, particularly salivation, which is vital for virus transmission.

### Gene Expression Related to SA- and JA-Signaling Pathways

Defense response to herbivore or plant pathogens occurs mostly in the Salicylic Acid (SA) and Jasmonic Acid (JA) signaling pathways (Lamb et al., 1989). Feeding by aphids induce plant responses similar to those induced by pathogens (Czosnek et al., 2002; Thompson and Goggin, 2006). Aphid infestation may activate the SA-signaling pathway and suppress the JA-signaling pathway, which will attract whiteflies and suppress TYLCV (Verhage et al., 2010). SA activation and JA suppression both inhibit viruses and improve feeding behavior of herbivore insects, respectively (Zarate et al., 2007). Our results showed that extending the period of aphid preinfestation may reduce expression of the genes involved in the JA-signaling pathway, and increase expression of the genes involved in the SA-signaling pathway. According to our results, it seems that upregulating the SA pathway and suppressing the JA pathway with aphid pre-infestation may directly impact the feeding strategies of the whitefly, and consequentially, the transmission and acquisition of TYLCV between whitefly and tomato plant. Similarly, there have been several reports showing that induced plant defense responses are stronger against pathogens compared to insect herbivores (Ghanim et al., 2001; Tan and Liu, 2014). Ohgushi (2005) reported that infestation by tobacco mosaic virus can induce systemic acquired resistance, while not activating the JA signaling pathway. Generally, the SA pathway is activated by pathogens, and the JA pathway is induced by insects (Salzman et al., 2005). Previous studies have shown that attack by sap-sucking herbivores can induce plant responses similar to those induced by pathogen infestations, inhibiting the JA-signaling pathways and consequentially reducing negative impacts on herbivores (Papayiannis et al., 2009, 2010; Giordanengo et al., 2010). Plant defense responses induced by herbivores or pathogens are often similar, since both responses can act to resist both insects and pathogens, or to minimize interference between herbivores and pathogen defenses (Thaler et al., 2002; Murugan and Dhandapani, 2007). Tomato plants infested by phloem-feeding insects (e.g., aphids) evoked JA- and SA-signaling pathways to co-regulate the expression of plant defense genes (Muniyappa et al., 2000). The antagonism between SA and JA impacts feeding behavior of whiteflies and viral propagation. In the transmission test, the SA and JA was dominant during the whole process. However, during the acquisition test, the JA-signaling pathway was the key factor affecting virus acquisition, and the feeding behavior of whiteflies dominated the whole process. SA activity may deter the transmission of pathogens, but inhibition of JA might attract herbivorous insects (Chu et al., 2007). The results showed that a decrease in JA combined with an increase in SA led to a net positive effect of aphid infestation on increased feeding of whiteflies, outweighing inhibition of TYLCV transmission.

Furthermore, the defense responses of these two pathways may vary with duration of injury (Preston et al., 1999; Nombela et al., 2009). Our results not only showed a temporal component to variation in associated gene expression, but also showed that indirect systemic responses (variation in amount of TYLCV across plant organs) also vary according to duration of preinfestation. Generally, herbivore feeding leads to systemic defense in the plant, enacted in a concerted manner across tissues and organs (Polston et al., 1990; Pico et al., 1999). Thus, further evaluation of defense-related gene expression induced by insect feeding on different plant parts is required, and how this is impacted by pre-infestation.

Generally, we have confirmed the hypothesis that aphid preinfestation impacts transmission and acquisition of TYLCV between tomato and B. tabaci. Pre-infestation showed different respective influences to the feeding strategies during the transmission or acquisition processes, while associated with opposing trends in expression of the SA- and JA- pathways. Based on current understanding of the interactions between the SA- and JA- pathways, particularly cross-inhibition, we believe that the mechanism delineating the complex and precise interaction among these organisms is worthy of further research. Notwithstanding, due to the economic importance of tomato crop, our study might have implications in the development of novel strategies for the regulation of plant pathogens and insect vectors.

### AUTHOR CONTRIBUTIONS

XT, JC, TL, and FG designed the study. All authors drafted the manuscript. XT, GB, ND, and XY implemented the analyses. All authors gave final approval for publication.

## ACKNOWLEDGMENTS

We thank Dr. T. Giordani (University of Pisa, Italy) for reading and commenting on an earlier version of the manuscript, and thanks Dr S. Wang and Dr. H. X. Xu for providing part of the material and methods. This work was supported by the National Basic Research Program of China (973 Program) (grant No. 2013CB127605), the Special Fund for Agroscientific Research in the Public Interest (grant Nos. 201303024 and 201303108), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB11050400), the National Nature Science Fund of China (grant No. 31370438), the R&D Special Fund for Public Welfare Industry (Agriculture 201303019) and the State Key Laboratory of Integrated Management of Pest Insects and Rodents (Grant No. ChineseIPM1611). We are grateful for the assistance of all staff and students in the Key Laboratory of Applied Entomology, Northwest A&F University at Yangling, Shaanxi, China.

### REFERENCES

fpls-08-01597 September 20, 2017 Time: 15:22 # 10


fpls-08-01597 September 20, 2017 Time: 15:22 # 11

euphorbiae) induce variable resistance or susceptibility responses. Bull. Entomol. Res. 99, 183–191. doi: 10.1017/S000748530800 6214


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Tan, Chen, Benelli, Desneux, Yang, Liu and Ge. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Harnessing Host-Vector Microbiome for Sustainable Plant Disease Management of Phloem-Limited Bacteria

Pankaj Trivedi<sup>1</sup> \*, Chanda Trivedi<sup>1</sup> , Jasmine Grinyer<sup>1</sup> , Ian C. Anderson<sup>1</sup> and Brajesh K. Singh1,2

<sup>1</sup> Hawkesbury Institute for the Environment, Western Sydney University, Penrith South, NSW, Australia, <sup>2</sup> Global Centre for Land Based Innovation, Western Sydney University, Penrith South, NSW, Australia

Plant health and productivity is strongly influenced by their intimate interaction with deleterious and beneficial organisms, including microbes, and insects. Of the various plant diseases, insect-vectored diseases are of particular interest, including those caused by obligate parasites affecting plant phloem such as Candidatus (Ca.) Phytoplasma species and several species of Ca. Liberibacter. Recent studies on plant– microbe and plant–insect interactions of these pathogens have demonstrated that plant–microbe–insect interactions have far reaching consequences for the functioning and evolution of the organisms involved. These interactions take place within complex pathosystems and are shaped by a myriad of biotic and abiotic factors. However, our current understanding of these processes and their implications for the establishment and spread of insect-borne diseases remains limited. This article highlights the molecular, ecological, and evolutionary aspects of interactions among insects, plants, and their associated microbial communities with a focus on insect vectored and phloemlimited pathogens belonging to Ca. Phytoplasma and Ca. Liberibacter species. We propose that innovative and interdisciplinary research aimed at linking scales from the cellular to the community level will be vital for increasing our understanding of the mechanisms underpinning plant–insect–microbe interactions. Examination of such interactions could lead us to applied solutions for sustainable disease and pest management.

Keywords: pathogens, phytoplasma, Candidatus Liberibacter species, insects, biocontrol, microbial communities

### INTRODUCTION

Plant pathogenic bacteria cause serious diseases for many major agriculture crops and fruit trees throughout the world (Vidhyasekaran, 2002), costing billions of dollars in damage annually (Pimentel et al., 2001). Of the various plant diseases, insect vectored diseases caused by obligate parasites of plant phloem are of particular interest (Bové and Garnier, 2003). These include the large and diverse group of Candidatus (Ca.) Phytoplasma species (transmitted by various hemipteran species including leafhoppers) and several species of Ca. Liberibacter (transmitted by the hemipteran species, psyllids). The fastidious nature of the members within Ca. Phytoplasma

#### Edited by:

Kumar Krishnamurthy, Tamil Nadu Agricultural University, India

#### Reviewed by:

Bhim Pratap Singh, Mizoram University, India Abdullah M. Al-Sadi, Sultan Qaboos University, Oman

\*Correspondence: Pankaj Trivedi p.trivedi@westernsydney.edu.au

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 17 July 2016 Accepted: 07 September 2016 Published: 30 September 2016

#### Citation:

Trivedi P, Trivedi C, Grinyer J, Anderson IC and Singh BK (2016) Harnessing Host-Vector Microbiome for Sustainable Plant Disease Management of Phloem-Limited Bacteria. Front. Plant Sci. 7:1423. doi: 10.3389/fpls.2016.01423

and Ca. Liberibacter hampers efforts to explore their epidemiology, the genetic mechanisms for disease manifestation, and for devising suitable control/prevention measures (Wang and Trivedi, 2013; Bertaccini et al., 2014). Infection by both groups of pathogens is often fatal, causing devastating damage to agricultural production around the world (Strauss, 2009; Munyaneza et al., 2010; Al-Sadi et al., 2012; Munyaneza, 2012; Wang and Trivedi, 2013). For example, phytoplasma epidemics among coconut palms have destroyed the livelihoods of many people in Africa and the Caribbean (Strauss, 2009). Huanglongbing (HLB) disease caused by Ca. Liberibacter spp. [including Ca. L. asiaticus (Las), Ca. L. africanus, and Ca. L. americanus) has had a devastating effect on the citrus industry worldwide (Wang and Trivedi, 2013). A relative, Ca. L. solanacearem causes zebra chip disease in potato, stunting and chlorosis in solanaceous species and foliage discoloration in carrots (Munyaneza et al., 2010; Munyaneza, 2012).

In recent years movement of propagative plant material and vegetable products has allowed the spread of both pest and pathogens around the world and their establishment in new areas where the conditions for disease development may be more favorable than in the area of origin (Wang and Trivedi, 2013). In addition, diseases transmitted by insects are expected to increase in frequency and spread to different localities due to global warming and climate change as future climates are advantageous to the cold-sensitive vectors (Hogenhout et al., 2008). Therefore, the development of robust and environmentally sustainable pest and pathogen control methods will become more important in the future.

Much of our understanding of the molecular mechanism governing decisions between compatibility or defense in host– pathogen interactions come from the studies that incorporate "single species and monoculture"; typically reduced to one plant interacting with one experimentally added pathogen. This "reduced complexity" approach forms the basis of the "disease triangle" paradigm which conceptualizes the interaction between the host, pathogen, and environment by providing a framework used to explain disease causation factors (Francl, 2001). However, in nature microbes live in constant association with other microbial species, directly or indirectly interacting and creating multispecies communities (Sagaram et al., 2009; Bulgarelli et al., 2013; Turner et al., 2013; Chaparro et al., 2014). A paradigm shift that employs a broad community level view toward the evolution and ecology of plant pathogenic bacteria is now being considered. This view has the potential to provide new directions in disease control measures by unearthing the hidden ecology and pathogenic potential through a mechanistic understanding of pathogen interactions with their host and associated microbial communities (**Figure 1**).

Hosts (insect and plants) as well as the environment (such as soil) consist of complex and diverse microbiome that interacts with the respective micro-environments (**Figure 1**). The dynamic interaction of host and its associated microbiome together with environmental microbiota provides benefit to the host in terms of growth and fitness (Berg et al., 2014; Lebeis, 2014). Within the host and the environment, different niche habitats provide variable conditions for the development of specific microbiome (**Figure 1**). For example, microbial community differs between different sized aggregates in soil (Trivedi et al., 2015) and different tissues/parts in hosts (Edwards et al., 2015; Fonseca-García et al., 2016). Obligate endophytic pathogens that are vectored by insects reside in specific parts in their insects during various life stages (Douglas, 2015). In both plant and insect, pathogens interact with the microbiome of specific tissue and influence changes in host responses (Weiss and Aksoy, 2011). These host–environment– microbiome–pathogen interactions are influenced by climate, land use, management practices, and other environmental factors (soil properties, nutrient status, etc.; Mueller and Sachs, 2015). Selection of particular set of these environmental factors can affect the microbiome that will influence the outcome of pathogen infection. Similarly manipulation of host associated microbiome can lead to the development of novel disease management practices.

Although the unique features of phytoplasmas and Ca. Liberibacter spp. have long made them a subject of interest, the difficulty of in vitro culture has hindered their molecular characterization. In recent years the availability of genome sequences for several phytoplasma strains (Oshima et al., 2004; Bai et al., 2006; Kube et al., 2008; Tran-Nguyen et al., 2008), and Ca. Liberibacter spp. (Duan et al., 2009; Lin et al., 2011,

FIGURE 1 | Interactions between the environment, the hosts (insects and plants), their associated microbiome and obligate endophytic pathogens transmitted by insects (e.g., Candidatus (Ca.) Phytoplasma species and several species of Ca. Liberibacter). Microbiome associated with hosts and environment is shown by different colored circles where the size of the circle represents greater numbers and diversity of the associated microbiome. Overlapping circles within hosts and environment represents different niches with specific microbiomes. Red colored arrow represents pathogen movement between plants and insects. Different color of pathogen in insect and plant is indicative of differential host adaptation strategies. Factors that influence host-environment-microbiome-pathogen interactions are shown in blue circles.

2015) have contributed significantly to our understanding of the biology of these pathogens. Analysis of these genomes has revealed that both groups have a very small genome (530–1350 kb for phytoplasma's and 1190–1260 kb for Ca. Liberibacter spp.) and lack intact metabolic pathways involved in the biosynthesis of various fatty acids, sterols, amino acids, and nucleotides (Oshima et al., 2004; Tran-Nguyen et al., 2008; Duan et al., 2009; Wang and Trivedi, 2013). Consistent with their intracellular nature, phytoplasma and Ca. Liberibacter spp. lacks type III and type IV secretion systems (except for one type IVB system in some phytoplasma) as well as typical free-living or plant colonizing extracellular degradative enzymes. Although metabolic genes are scarce, the genomes of both groups of pathogens contain many genes encoding transporter systems, suggesting that these pathogens rely heavily upon nutrients and metabolites extracted from their host (Oshima et al., 2004; Bai et al., 2006; Kube et al., 2008; Tran-Nguyen et al., 2008; Duan et al., 2009; Wang and Trivedi, 2013). Considering the limited metabolic capacity, it is remarkable that they can interact with their hosts from two different kingdoms (Plantae and Animalia) and successfully colonize highly dissimilar environments. The consumption of metabolites by the pathogen greatly disturbs the metabolic balance of the host cell, causing disease symptoms. These altered conditions result in significant changes in the structure and function of stable multispecies communities associated with the host, wherein the augmented gene pool and the combined metabolic repertoire can influence pathogen survival and disease manifestation (Hosni et al., 2011).

This article explores the interactions of the phloem limited and insect vectored plant pathogens, Phytoplasma's and Ca. Liberibacter spp. [mainly Liberibacter asiaticus (Las)] with their insect and plant host and their associated microbial community. We describe: (1) interactions of the insect associated microbial community with the pathogen(s); (2) differential gene expression that enables adaptation of pathogens to different hosts; (3) the modulation of host response by pathogens for their own transmission; (4) fluctuations in the structure and function of the plant associated microbiome in response to pathogen infection. We further highlight the potential for beneficial microbes within each plant and insect microbial community to be developed as biocontrol agents for the sustainable management of diseases caused by phytoplasma's and Ca. Liberibacter spp.

### INSECT-ASSOCIATED MICROBIAL COMMUNITY AND ITS INTERACTION WITH PATHOGENS

The vascular tissues of plants are generally deficient in essential nutrients, therefore sap-feeding insects rely exclusively on their associations with bacterial symbionts to supplement their dietary needs (including amino acids, lipids, and vitamins; Buchner, 1965; Bourtzis and Miller, 2003). Studies have shown that sap feeding insects such as aphids and psyllids have significantly less microbial diversity as compared to xylophagous and leaf feeder insects (Ishii et al., 2013; Sugio et al., 2014). For these insects, most of the associated microbes are obligate or primary and facultative or secondary symbionts that are specifically associated with these different groups of sap feeders. The presence or absence of these bacteria could affect the competency of the insect vector to transmit pathogens or the life history traits of the insects themselves. For example obligate intracellular bacteria Wolbachia that are presumably found in up to 66% of all insects (Hilgenboecker et al., 2008) manipulate host reproduction by inducing cytoplasmic incompatibility, parthenogenesis, feminization, and male-killing (Stouthamer et al., 1999). Quantifying the presence of obligate endosymbionts and understanding the variety of facultative endosymbionts these insects utilize, may provide insights into the transmission of pathogens.

The Ca. Liberibacter asiaticus (Las) concentration within the insect was found to have a strong negative relationship with an endosymbiont residing in the syncytium of the mycetocyte (Fagen et al., 2012). Interestingly, the population of another bacteriocyte-associated bacteria, mycetocyte endosymbiont, was unaffected by Las acquisition. The variable effect of Las on endosymbiotic bacteria may be caused by its irregular distribution within the host causing certain bacteria to be displaced but not others (Fagen et al., 2012). Las titer had a positive relationship with the endosymbiotic community composed of Wolbachia which has been shown to alter host insect gene expression that creates a favorable intracellular environment for its growth (Hussain et al., 2011). A comparable mechanism may lead to increased Wolbachia and related increases in Las populations within its vector Asian citrus psyllid (ACP). Finding an increase in Wolbachia titer with Las infection indicates a more complicated mechanism than simple replacement of indigenous endosymbionts by Las. Ishii et al. (2013) reported strikingly complex endosymbiotic microbiota of the Macrosteles leafhoppers that vectored two genetically distinct phytoplasma's. The microbiome of these leafhoppers included two obligate endosymbionts, "Ca. Sulcia muelleri" and "Ca. Nasuia deltocephalinicola," and five facultative endosymbionts, Wolbachia, Rickettsia, Burkholderia, Diplorickettsia, and a novel bacterium belonging to the Rickettsiaceae. The highly complex endosymbiotic microbiota suggested ecological interactions between the obligate endosymbionts, the facultative endosymbionts, and the pathogenic phytoplasma's within the same host insects that may affect the competence of insect vector. The role of insect-associated microbes in altering the transfer rate of pathogens has not yet been reported (Sugio et al., 2014). Filling this knowledge gap is key to understanding disease epidemiology and for improving disease control strategies.

### PATHOGENS MODULATE GENE EXPRESSION DURING TRANSMISSION IN DIFFERENT HOSTS

In order to proliferate and cause disease, insect vectored pathogens have to switch between the diverse environments of plants and insects (Chatterjee et al., 2008; Yan et al., 2013). These different environments have a dramatic effect on bacterial gene expression; specific genes whose products assist in survival are

activated, whereas non-essential gene products in a particular environment are deactivated (Chowdhury et al., 1996). It has been suggested that virulence factors are expressed at different stages of the infection process and are dictated by the changing microenvironment of the host (Chowdhury et al., 1996).

The number of Las genes up-regulated in plants was higher when compared to the insect vector (Yan et al., 2013), while an opposite trend was observed for phytoplasma (Oshima et al., 2011; Makarova et al., 2015). One possible reason for this difference is the co-evolution of the pathogen, plant host and insect vector. Ca. Liberibacter spp. evolved from an ancestor in the plant-associated Rhizobiaceae family whereas phytoplasma's are closely related to an animal-associated Mycoplasma or Acholeplasma spp. (Oshima et al., 2013). It can be proposed that Ca. Liberibacter spp. and Phytoplasma's would have undergone adaptive, diversifying, and reductive evolutionary processes that would have made them more suitable for their interactions with specific plants and insects, respectively. The intimate association of Ca. Liberibacter spp. with plants as endophytes predisposes them to frequent encounters with herbivorous insects, providing ample opportunities to evolve alternative associations with insects (Nadarasah and Stavrinides, 2011). Similarly, associations between phytoplasma and insects provide opportunities for alternate associations with different plant species.

The expression levels of several transporter genes were differentially expressed between hosts in Aster Yellows phytoplasma witches' broom (AY-WB) and Ca. Phytoplasma asteris OY-M strain of phytoplasma and Las. Zinc transporter genes were upregulated in insects for both the phytoplasma species, whereas for Las they were highly expressed in plants (Oshima et al., 2011; Yan et al., 2013; Makarova et al., 2015). For both groups of pathogens, multidrug efflux pumps were upregulated in plants (Oshima et al., 2011; Yan et al., 2013; Makarova et al., 2015) demonstrating host-specific genetic expression in order to adapt between two hosts.

After inoculation into the plant phloem by the insect, the pathogen encounters a change in osmolarity and must protect itself from dehydration and loss of turgor. Phytoplasma and Las deal with the problem of osmolarity through different mechanisms. For both pathogens, the genes for dealing with osmolarity were up-regulated in plants when compared to insects (Oshima et al., 2011; Yan et al., 2013; Makarova et al., 2015). The Las gene proX, involved in the transport of the most common osmoprotectants glycine betaine, was up-regulated in plants compared to their insect vectors psyllids (Yan et al., 2013). In AY-WB and Ca. Phytoplasma asteris OY-M the MscL channel that senses mechanical stretching of the membrane was significantly up-regulated when compared to the insect vector (Oshima et al., 2011; Makarova et al., 2015).

Analyses of the phytoplasma and Ca. Liberibacter spp. genomes have identified glucanase, serralysins, and hemolysinlike proteins as possible virulence factors (Bai et al., 2006; Duan et al., 2009; Wang and Trivedi, 2013). Hemolysins are bacterial toxins that act to form a transmembrane channel within the membrane of susceptible cells causing the leakage of ions, water, and low molecular weight molecules into the host cell (Gouaux, 1998). The expression of putative hemolysin genes were upregulated in the insects in AY-WB phytoplasma while in Las these genes were upregulated in plants (Yan et al., 2013; Makarova et al., 2015). Although no differential expression of this gene was found in another phytoplasma OY-M, it has been suggested that hemolysin may be involved in virulence and insect transmission in other insect transmitted plant pathogens (Wang et al., 2012). In Las, genes encoding a secreted metalloprotease serralysin were highly expressed in plants compared to the insect vector (Yan et al., 2013). Serralysin is postulated to aid bacterial survival in plants by modifying plant defense and nutrient uptake. Interestingly, many candidate secreted effectors proteins of Las postulated to modulate cellular functions for disease progression (Pitino et al., 2016) were upregulated in plants (Yan et al., 2013).

In Las, genes that encode enzymes involved in glycolysis were up-regulated in plants (Yan et al., 2013). The high expression of glycolysis-associated genes in plants indicates that Las can use glucose acquired from the host plant to generate energy for intracellular growth. In contrast, phytoplasma glycolytic genes were not differentially expressed in both hosts (Oshima et al., 2011; Makarova et al., 2015). In both AY-WB and OY-M strains, genes relevant to the malate and pyruvate pathways were markedly up-regulated in insects. Malate can be utilized as a sole source of energy by phytoplasma (Kube et al., 2012) and might serve as an important source of energy for phytoplasma's when colonizing the leafhopper vector (Makarova et al., 2015).

Overall the studies on the gene expression profiles of phytoplasma and Las in plants and insects have reported a dramatic response to the diverse host environments when compared with environmental changes in other bacteria (Oshima et al., 2011). In general, responses were markedly different for Las and phytoplasma suggesting species and host-specific adaptations to the distinct environment of plant and insects. However, our understanding of the exact roles of these genes in host switching remain largely unknown. A clear understanding of the molecular basis of host switching can unlock the possibility for the development of novel methods in pest control for insect transmissible pathogen diseases.

### PATHOGENS ALTER PLANT PHYSIOLOGY AND MORPHOLOGY TO ATTRACT VECTORS

It is common for the insect vectored pathogens to manipulate plant–insect interactions to enhance their own dissemination via effects on: (a) the quality of the primary host as a resource for the vector (Mauck et al., 2010), or; (b) the production of hostderived cues that mediate vector attraction (Lefèvre et al., 2006). Infection and subsequent disease manifestation in plants change the plant architecture and/or physiology that enhance both vector recruitment to infected plants and subsequent dispersal of the pathogen to healthy plants.

Effector proteins secreted by different species of phytoplasma induced the production of many leaves and stems in their hosts, creating a characteristic bushy appearance and converting plants into more attractive hosts for egg-laying and reproduction of leafhopper vectors (Hoshi et al., 2009; MacLean et al., 2011; Sugio

et al., 2011). For example phytoplasma AY-WB produces a novel effector SAP54 that degrades the highly conserved transcription factors of the MADS-box family involved in flower development leading to the generation of sterile plants (MacLean et al., 2014). These sterile plants, which form leaves rather than flowers, are more attractive to leafhoppers. Similarly another virulence effector from OY-M, tengu-su inducer (TENGU) disrupts the auxin signaling pathway and induces dwarfism and witches' broom symptoms that attract more insects (Hoshi et al., 2009; Sugio et al., 2011; Minato et al., 2014). The characteristic yellowing symptoms in Liberibacter and phytoplasma infected plants (Wang and Trivedi, 2013; Bertaccini et al., 2014) can also play a role in insect vector attraction.

Pathogens are also known to induce plant responses that modify behavior of the insect vector by altering the olfactory cues through changes in volatile and non-volatile secondary metabolites that insects use to locate suitable host plants (Orlovskis et al., 2015). For example, using a multitrophic system consisting of a phytoplasma [Ca. Phytoplasma mali (Ca. P. mali)], a host tree (Malus domestica), and a phloem-feeding insect (Cacopsylla picta), Mayer et al., 2008a,b) showed that infected apple plants released higher amounts of the sesquiterpene β-caryophyllene than uninfected plants. Newly hatched adults of C. picta were attracted by the odor of infected apple trees. Las infected plants produce significantly more methyl salicylate and less methyl anthranilate and D-limonene as compared to non-infected plants (Mann et al., 2012). Methyl salicylate was attractive to psyllids, while methyl anthranilate did not affect their behavior suggesting that odorants mediate psyllid preference. This apparent pathogen-mediated manipulation of vector behavior may facilitate pathogen spread. Feeding on citrus by ACP adults also induced release of methyl salicylate, suggesting that it may be a cue to reveal the location of conspecifics on host plants (Mann et al., 2012). Similar processes have been documented in other complex patho-systems (Shapiro et al., 2012) suggesting that detailed insights on the mechanisms driving such effects will have far-reaching implications both for basic ecology and for the management of disease processes in natural and agricultural settings. Further characterization of the infochemicals that are induced by plant pathogens to attract the vectors will lead to the development of new traps (such as sticky traps) for monitoring or even mass trapping of vectors for pest control.

### IMPACT OF PATHOGENS ON PLANT ASSOCIATED MICROBIAL COMMUNITIES

Plants are associated with an astounding number and variety of microbes that interact with their hosts with different degrees of dependencies including competition, commensalism, mutualism, and parasitism (Garbeva et al., 2004; Bulgarelli et al., 2013; Philippot et al., 2013). It has been postulated that the disruption of multi-trophic plant–microbe–environment interactions under the influence of invading pathogen(s) will cause community reorganization and changes in local feedback interactions (**Figure 2**). However, there is a paucity of synthesized knowledge on the extent to which such community shifts may occur, the dynamics of these changes and the putative effects regarding the microbial mediated ecological functions (Trivedi et al., 2010, 2012). As the diversity and stability of plantassociated microbial communities heavily influence soil quality, plant production, and ecosystem processes, fluctuations in microbial community structure could have serious implications in ecosystem sustainability (**Figure 2**).

The structure of plant associated bacterial community changes in response to a variety of processes, and these shifts have been suggested to impact various ecosystem processes (e.g., nutrient recycling, decomposition) and/or the outcome of host– pathogen interactions (e.g., growth of pathogens, release of plant growth promoting rhizobacteria; Emmert and Handelsman, 1999; Trivedi et al., 2011; Philippot et al., 2013). Also, the interactions between plant-associated microbial communities and pathogens are not well understood, and our knowledge of the intimacy and decisiveness of such associations with respect to the behavior and survival of participating organisms is still in its infancy (Trivedi et al., 2010, 2012).

In general, pathogen triggers a cascade of reactions in plants, leading to the synthesis of defensive compounds which in turn enable it to withstand pathogen attack either directly (e.g., by structural or physiological modifications) or by mediating different plant signaling pathways (Lichtenthaler, 1998). The altered conditions after the pathogen attack could have variable effects on the survival and proliferation of different groups of plant-associated microbes. For example, infection by Las (Sagaram et al., 2009; Trivedi et al., 2010) and phytoplasma (Bulgari et al., 2011, 2014) has been reported to restructure the endophytic microbial community of their respective hosts. Pathogen infection caused a decrease in the overall bacterial diversity in the infected host (Trivedi et al., 2010; Bulgari et al., 2011). Interestingly, the abundance of bacteria belonging to Sphingobacterium was increased in the plants infected from Las and Phytoplasama (Trivedi et al., 2010; Bulgari et al., 2011). In general, it appears that infection by plant pathogens restructures the microbial community; many species show reduced levels, are not detected or are replaced by other indigenous populations, which can better tolerate/adapt to the stress condition and interact closely with the pathogen.

Interestingly obligate endophytic pathogens have been reported to restructure the native microbial community even when a direct competition effect is lacking. For example, significant changes in the microbial community structure of rhizosphere soil samples were observed in Las infected citrus (Trivedi et al., 2012). In this case, alteration in plant physiology leading to quantitative and qualitative changes in partitioning the photo-assimilates was the primary cause of the shift in microbial diversity of the diseased host. Typical rhizosphere inhibiting bacteria such as those belonging to Proteobacteria were significantly reduced in the infected plants suggesting that rhizosphere bacteria react more strongly to changes in plant physiology and exudation induced by pathogen infection.

In recent years, several reports have demonstrated profound shifts in the structure and composition of plant-associated

on ecosystem functions are not well understood. Using Las and citrus huanglongbing as a model for pathogen–disease interactions that involve the blockage of vascular tissues, Trivedi et al. (2012) have reported that the introduction of pathogens into natural ecosystems perturbs the stability of the microbial community, thus affecting biogeochemical cycles that regulate soil fertility and ecosystem functions. Using comprehensive functional micro-array "GeoChip 4.0" authors showed that HLB disease has significant reduced abundance of functional guilds involved in key processes involved in nutrient cycling such as nitrogen cycling, carbon fixation, phosphorus utilization, metal homeostasis, and resistance. As the diversity and stability of the plant-associated microbial communities heavily influence soil and plant quality and ecosystem processes (Nannipieri et al., 2003; Garbeva et al., 2004), erosion of microbial diversity could have serious implications on the agro-ecosystem sustainability. In addition shrinking genetic and functional diversity in response

microbial communities (Araújo et al., 2002; Reiter et al., 2002; Trivedi et al., 2010, 2011, 2012; Bulgari et al., 2011) in response to pathogen infection, however, the implication of these shifts to pathogen infection, will compromise the capacity of adaptive responses to further perturbation. These results pointed toward the beyond yield effect of plant diseases on ecosystem processes and suggested that in the long term, these fluctuations might have important implications for the productivity and sustainability of agro-ecosystems.

### EXPLOITING HOST–MICROBE–PATHOGEN INTERACTIONS FOR DISEASE MANAGEMENT

### The Potential of Insect Associated Microbiome for Pest Management

Currently, the management of diseases caused by phytoplasma and Ca. Liberibacter species is commonly based on the control of the insects, i.e., by spraying various insecticides, and on practices where the removal of symptomatic plants is undertaken (Tiwari et al., 2011; Wang and Trivedi, 2013). It is well

fpls-07-01423 September 28, 2016 Time: 15:55 # 6

recognized that the use of chemical insecticides as the main control strategy is not sustainable, and is known to have negative side-effects, including both environmental and biological effects (Qureshi and Stansly, 2007; Tiwari et al., 2011; Orduño-Cruz et al., 2015). Based on information on the insect associated microbial community Crotti et al. (2012) have proposed a "Microbial Resource Management (MRM)" that foresees the proper management of the microbial resource present in a given ecosystem in order to solve practical problems through the use of microorganisms. Some first steps of MRM applications have been already carried out on insect vectors, with the aim of defining the microbial community composition and functionality within the insects (Marzorati et al., 2006; Miller et al., 2006; Crotti et al., 2012). Researchers have reported various potential biological control bacteria associated with insect vectors that can provide opportunities for controlling these economically important vectors, either through potential paratransgenesis or cytoplasmic incompatibility (Powell and Tabachnick, 2014). The final aim is to propose a biocontrol approach based on the management of the microbial symbiont associated with the vector in order to counteract directly the pathogen or to reduce the vector competence.

Efforts are underway to develop mycoinsecticides for the biocontrol of ACP by the use of single-spore high-virulence strains of endophytic fungi Isaria javanica (Ayala-Zermeño et al., 2015; Gallou et al., 2016). This species has been reported to be associated with ACP and has been described as a pathogen of Lepidoptera, Coleoptera (Samson, 1974; Shimazu and Takatsuka, 2010) and the greenhouse whitefly of the order Hemiptera (Scorsetti et al., 2008). Application of conidial based formulations of endophytic fungi Metarhizium anisopliae, Isaria fumosorosea and Hirstuella citriformis resulted in high mortality of vectors of Ca. Liberibacter spp. such as ACP or Bactericera cockerelli (Tamayo-Mejía et al., 2014; Orduño-Cruz et al., 2015). Although the speed of kill caused by an entomopathogenic fungus is not comparable with that of a chemical insecticide, entomopathogenic fungi are known to reduce the feeding activity of infected hosts (Avery et al., 2009), resulting in reduced pathogen transfer, but this needs further experimental confirmation (Orduño-Cruz et al., 2015). Further research is

BOX 1 | Microbiome Engineering to Improve Host Performance and Health. Recent breakthroughs in sequencing technologies have provided concrete evidence that the number of microbial cells and the sum of their genetic information are numerically dominant than that of their host. Microbiotas and their hosts interact in a manner that affects the fitness of the holobiont (host genome+microbiome) in many ways, including its morphology, development, physiology, resistance to disease, growth performance, and stress tolerance. Taken together, these interactions characterize the holobiont as a single and unique biological identity. Since that microbiome can adjust more rapidly and by more processes than the host genome to environmental dynamics (including disease progression), it plays fundamental role in the adaptation and fitness of the holobiont (Rosenberg and Zilber-Rosenberg, 2016). Mueller and Sachs (2015) have proposed a novel approach to improve animal and plant fitness by artificially selecting upon microbiomes, thus engineering evolved microbiomes with specific effects on host fitness. The host-mediated microbiome engineering approach selects upon microbial communities indirectly through the host and leverages host traits that evolved to influence microbiomes (Mueller and Sachs, 2015). Evidence that microbiome can be optimized for disease resistance by the application of phytohormones that activate defense responses is also available (Lebeis et al., 2015). Generating host-mediated artificial selection of microbiomes may be a cheaper way to help curb plant and animal

diseases rather than pesticides and antibiotics, or creating genetically modified organisms. Sheth et al. (2016) have highlighted emerging in situ genome engineering toolkit to manipulate microbial communities with high specificity and efficacy over a range of specificities and magnitudes. Plant ecological engineering (e.g., integrating plant breeding with microbiome selection) has enormous potential to manipulate host microbiome in order to enhance effectiveness of diseases management.

required before the true potential of controlling insect vectors by biocontrol agents can be realized. Given the efficacy of biocontrol agents is reported to influenced by a range of parameters such as type of formulations, time and mode of applications and environmental and climate conditions, developing whole microbiome approach can potentially provide better disease control. However, this would require the development of effective tools to manipulate microbiome of the vector.

### The Potential of Plant Associated Microbiome for Increasing Plant Performance and Disease Resistance

Plant-associated microbes which improve the fertility status of soil and contribute in augmenting overall plant growth and health known as Plant Growth Promoting Microbes (PGPM) are receiving increased attention for use as microbial inoculants in agriculture (Estrada-De Los Santos et al., 2001; Choudhary and Johri, 2009; Lugtenberg and Kamilova, 2009; Trivedi et al., 2011) (**Figure 3**). These microbes support plant health and growth by various mechanisms that include nutrient solubilisation and fixation, production of plant hormones, stress relief, and suppression of plant pathogens by induction of plant defenses, production of antibiotics, and/or out-competition of pathogens (Rosenblueth and Martínez-Romero, 2006) (**Figure 3**). To increase field efficiency of microbial inoculation workers have advocated to screen "eco-specific strains" that are acclimatized to a particular set of environmental conditions (Trivedi and Pandey, 2008; Trivedi et al., 2011). This favors efficient establishment and survival of the introduced bacteria leading to increased performance and also does not affect the preexisting balance among indigenous populations.

It has been noticed that certain trees (called escape plants) may survive in heavily infected areas under heavy load of pathogen and vector (Sagaram et al., 2009; Trivedi et al., 2011). Because these escape plants have the same genotype as susceptible plants and have developed under similar edaphic and climatic conditions, a possible explanation for the lack of disease symptoms may lie in the nature of the microbial community associated with these plants. In previous studies, it has been documented that specific endophytic bacterial communities are associated with these escape plants (Sagaram et al., 2009; Bulgari et al., 2011, 2014; Trivedi et al., 2011). Some of the bacteria isolated from these escape plants showed typical traits BOX 2 | Priority Challenges. Microbiome approach to manage vector mediated plant disease has enormous potential but to achieve this goal, there are some key challenges that need to be solved by integrated fundamental and applied research. These priority challenges include:


of potential biocontrol agents (Bulgari et al., 2011; Trivedi et al., 2011). Isolation frequency of bacterial strains showing multiple beneficial activities was higher in escape/healthy as compared to Ca. P. mali or Las infected plants (Bulgari et al., 2011; Trivedi et al., 2011). These isolates belonged to Pseudomonas, Bacillus, and Lysinibacillus species and have been previously developed as a carrier based bio-inoculant to increase plant productivity and health of various plant species (Trivedi and Pandey, 2008; Trivedi et al., 2008).

The research on screening effective biocontrol agents against obligate endophytic pathogens such as phytoplasma and Ca. Liberibacter spp. is hampered due to the unavailability of proper in vitro screening systems that provides repeatable and reliable results in shorter periods of time. The widely used dual culture technique could not be applied to screen bacteria antagonistic to these obligate endophytes due to our inability to culture these bacteria. Trivedi et al. (2009) have developed a method to quantify viable Las with the aid of ethidium mono-azide (EMA) and subsequent qPCR that can differentiate live from dead cells. The EMA-qPCR assay was optimized for screening potential biocontrol bacteria effective against Las (Trivedi et al., 2011). The selected novel isolates are further being tested in planta and in field conditions to determine whether they could be used in management of HLB.

Beneficial soil-borne microbes can induce an enhanced defensive capacity in above-ground plant parts to protect plants

against insect herbivores (**Figure 3**). This induced systemic resistance (ISR) triggered by soil-borne microbes is often not associated with enhanced biosynthesis of plant hormones that are important for defense against insect herbivores, nor with massive changes in defense-related gene expression. Instead, beneficial soil-borne microbes prime the plants for enhanced defense that is characterized by a faster and stronger expression of defense responses activated upon insect attack, resulting in increased resistance to the insects, and/or decrease in pathogen proliferation (Pieterse et al., 2013). Very recently, the concept of inducing enhanced resistance to phytoplasma with beneficial bacteria has been evaluated using Chrysanthemum as a model organism (Gamalero et al., 2010; D'Amelio et al., 2011; Musetti et al., 2011). Results showed that pretreatment with Pseudomonas putida S1Pf1Rif decreases the negative effects on plant growth infected with chrysanthemum yellows phytoplasma (CYP), but had no effect on CYP viability and proliferation (Gamalero et al., 2010). Co-inoculation of P. putida S1Pf1Rif and mycorrhizal fungi Glomus mosseae BEG12 resulted in a slightly increased resistance and a delay of symptoms in CYP infected and nonresistant plants (D'Amelio et al., 2011). G. mosseae could also reduce symptoms of the stolbur phytoplasma causing Bois noir in grapevine and tomato (Lingua et al., 2002). Musetti et al. (2011) demonstrated that the endophyte (Epicoccum nigrum) treatment induced ultrastructural changes both in C. roseus tissues and in the pathogen and these changes were associated with a lower titer of phytoplasmas in the host plant. Soil-borne microbes can also induce the production of plant hormones such as salicylic acid, which plays a role in plant defense against insect herbivores with a piercing/sucking feeding mode, such as ACP.

Many experiments have demonstrated the growth stimulation of plant crops in the greenhouse, resulting in increased yield parameters and in the control on pathogenic organisms, however, the replication of successful results of PGPMs applications under field conditions has been limited (Antoun and Prévost, 2005; Trivedi and Pandey, 2008; Choudhary and Johri, 2009; Trivedi et al., 2011). The inconsistency in performance (Bashan, 1998; Pandey et al., 1998) may be due to a number of factors but the most important of these are likely to be the differences in the establishment and survival of introduced bacteria (Bashan, 1998; Pandey et al., 1998; Trivedi et al., 2011). Workers have emphasized that understanding ecology, survival, and activity of PGPM's is a key for their successful field application (Pandey et al., 1998; Trivedi et al., 2005; Trivedi and Pandey, 2008). Formulation of multi-stains of PGPM's with a broader spectrum of microbial weapons to stimulate plant growth or provide protection against diseases are reported to be more efficient in field conditions compared with single strains (Compant et al., 2005; Gopalakrishnan et al., 2015). Appropriate screening and the application of molecular tools to understand and manage the plant and insect associated microbiome can lead to new products or novel disease management strategies (**Box 1**). One of the key requirements to attain this goal includes a better understanding of interactions between host microbiome and pathogens and to identify key interactions that reduce survival/proliferation of pathogens in host or vectors. This fundamental knowledge can then pave the way to develop new products or tools for sustainable disease management.

## CONCLUSION

The research progress to better understand the interactions between obligate endophytic pathogens belonging to Ca. Phytoplasma and Ca. Liberibacter species and their hosts and vector has moved slowly because of the inability to isolate these fastidious bacteria on culture media. Studies of plant– pathogen and insect–pathogen interactions are taking advantage from high-throughput techniques and also from the constant improvement of genome sequencing and annotations of both microbes and their hosts (Mitter et al., 2013). However, there is lack of application of these techniques in the area of interaction between host, pathogens and biocontrol agents. Even the most intimate association between the pathogen and its host in the natural environment, whether occurring at the epiphytic or endophytic phase are influenced by a myriad of microbes that are intimately associated with plants or insects. Although it is well documented that various groups of microbes can increase plant productivity in several important crops or defend against pathogen attacks, there are significant challenges that need to overcome in order to harness host associate microbes for sustainable disease management (**Box 2**). Emerging technologies (e.g., next-generation sequencing, new in vitro screening tools) combined with well-defined controlled experiments based on evolutionary and ecological theories will facilitate better fundamental understanding on the interaction of pathogens and host associated microbiomes. Furthermore, research on the host-associated microbial community, and its variability, would provide insights into the ecological behavior of pathogenic bacteria in the context of surrounding microorganisms present in the same niches. Such knowledge on multi-trophic microbiome interactions has potential to be harnessed for development of more effective and sustainable management of vector-borne plant diseases.

## AUTHOR CONTRIBUTIONS

All authors listed, have made substantial, direct, and intellectual contribution to the work, and approved it for publication.

## ACKNOWLEDGMENTS

We acknowledge funding support Grains Research and Development Corporation (project UWS00008) and Australian Research Council (project DP13010484).

### REFERENCES

fpls-07-01423 September 28, 2016 Time: 15:55 # 10



and microbe invasions. Agric. Ecosys. Environ. 84, 1–20. doi: 10.1016/S0167- 8809(00)00178-X


(Hemiptera: Psyllidae). Ann. Entomol. Soc. Am. 102, 297–305. doi: 10.1603/ AN10128


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Trivedi, Trivedi, Grinyer, Anderson and Singh. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Bensulfuron-Methyl Treatment of Soil Affects the Infestation of Whitefly, Aphid, and Tobacco Mosaic Virus on Nicotiana tabacum

Renyi Li<sup>1</sup> , Saif Ul Islam<sup>1</sup> , Zujian Wu<sup>1</sup> \* and Xiujuan Ye1,2 \*

<sup>1</sup> Key Laboratory of Plant Virology of Fujian Province, Institute of Plant Virology, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> Key laboratory of Biopesticide and Chemical Biology, Ministry of Education, Fujian Agriculture and Forestry University, Fuzhou, China

Bensulfuron-methyl (BSM) is widely used in paddy soil for weed control. BSM residue in the soil has been known to inhibit the growth of sensitive crop plants. However, it is unknown whether BSM residue can affect the agrosystem in general. In this study, we have found significant effects of BSM on the infestation of Bemisia tabaci, Myzus persicae, and Tobacco mosaic virus (TMV) in Nicotiana tabacum. The soil was treated with BSM before the pest inoculation. The herbicide-treated tobaccos showed resistance to B. tabaci, but this resistance could not be detected until 15-day postinfestation when smaller number of adults B. tabaci appeared. In M. persicae assay, the longevity of all development stages of insects, and the fecundity of insects were not significantly affected when feeding on BSM-treated plants. In TMV assay, the BSM treatment also reduced virus-induced lesions in early infection time. However, the titer of TMV in BSM treated plants increased greatly over time and was over 40-fold higher than the mock-infected control plants after 20 days. Further studies showed that BSM treatment increased both jasmonic acid (JA) and salicylic acid (SA) levels in tobacco, as well as the expression of target genes in the JA and SA signaling pathways, such as NtWIPK, NtPR1a, and NtPAL. NtPR1a and NtPAL were initially suppressed after virus-inoculation, while NtRDR1 and NtRDR6, which play a key role in fighting virus infection, only showed up- or were down-regulated 20 days post virus-inoculation. Taken together, our results suggested that BSM residue in the soil may affect the metabolism of important phytohormones such as JA and SA in sensitive plants and consequently affect the plant immune response against infections such as whitefly, aphids, and viruses.

Keywords: herbicide, Bemisia tabaci, Myzus persicae, Tobacco mosaic virus, jasmonic acid, salicylic acid

## INTRODUCTION

Herbicides are nowadays widely used to control weeds in order to reduce the cost of labor (Powles, 2014). Properties such as low-toxicity, environmental-friendly, and high-selectivity of the herbicides are desirable (Busi et al., 2013). Bensulfuron-methyl (BSM), which was developed in 1970s, is a herbicide belonging to the sulfonylurea class (Saeki and Toyota, 2004) and used in paddy

#### Edited by:

Gero Benckiser, University of Giessen, Germany

#### Reviewed by:

Zonghua Wang, Fujian Agriculture and Forestry University, China Javier Veloso, University of Giessen, Germany

#### \*Correspondence:

Zujian Wu wuzujian@126.com Xiujuan Ye xiujuanye2004@gmail.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Plant Science

Received: 24 August 2016 Accepted: 12 December 2016 Published: 26 December 2016

#### Citation:

Li R, Islam SU, Wu Z and Ye X (2016) Bensulfuron-Methyl Treatment of Soil Affects the Infestation of Whitefly, Aphid, and Tobacco Mosaic Virus on Nicotiana tabacum. Front. Plant Sci. 7:1970. doi: 10.3389/fpls.2016.01970

**480**

fields (Lin et al., 2012). The mechanism of BSM-mediated weed killing involves the inhibition of acetolactate synthase (ALS) and the biosynthesis of branched-chain amino acids (Saeki and Toyota, 2004). This herbicide shows high-selectivity and also appears harmless to the Poaceae crops such as rice and maize (Saika et al., 2014). However, over-utilization can have negative impact on farming by creating herbicide-resistant weeds (Powles and Yu, 2010; Walsh and Powles, 2014) and damaging sensitive crops such as tomatoes and tobaccos (Lin et al., 2012).

Both Bemisia tabaci (Gennadius) and Myzus persicae (Sulzer) are serious pests affecting the production of tobaccos and other crops (Kollenberg et al., 2014; Santos-Garcia et al., 2015; Cao et al., 2016; Kang et al., 2016). Also maize fields have been reported to be damaged by the colonization of biotype B whiteflies (Quintela et al., 2016). B. tabaci not only causes disintegration of chloroplasts, sooty mold by honeydew, and lower carotenoid in many plants, especially the ones belonging to Solanaceae (Masuda et al., 2016), but carries and transmits several species of Geminiviruses which can lower crop yields tremendously (Shi et al., 2014a). Owing to fast evolution, B. tabaci nowadays has diverse biotypes worldwide among which the B and Q types can cause severe crops losses (Santos-Garcia et al., 2015). Green peach aphid M. persicae, which is another kind of important pest affecting tobaccos, has a broad spectrum of hosts all over the world (Cao et al., 2016; Kang et al., 2016; Kayser et al., 2016). The pests can both deprive plants of nutrients and transmit cucumber mosaic viruses or other pathogens (Zhao et al., 2015), resulting in immense losses of crop yield (Cao et al., 2016). Tobacco mosaic virus (TMV) is widely spread around the world, and it can affect a broad spectrum of hosts such as tobaccos, tomatoes, and potatoes, leading to economical losses.

The interaction between plants and pests/pathogens is affected by various factors such as chemicals, light, salt, temperature, and also biological factors (Shavit et al., 2013; Shi et al., 2014b; Su et al., 2016). In this regard, a number of studies have shown that the interaction depends on the endogenous levels of jasmonic acid (JA) and salicylic acid (SA) in host plants. Shi et al. (2014b) investigated three signal pathways involving JA, SA, and proteinase inhibitor (PI) in the plant-whitefly interaction and discovered that non-viruliferous and viruliferous B-type can induce different defense pathways in host plants. In addition, the infestation of Chinese cabbage Brassica by aphids can strongly induce the production of SA, but exogenous application of SA doesn't have any effect on the performance of aphids (Cao et al., 2016). It has been founded that the replication of TMV can be inhibited by the application of SA (Conti et al., 2012). In the interaction between viruses and phytohormones, RNA-dependent RNA polymerase (RDR) family is involved in the establishing of the resistance in host plants. In MtRDR1 transgenic tobaccos, MtRDR1 can function as RDR6 and can help the host plants to recover from the symptom caused by virus infection in the presence of SA (Lee et al., 2016).

Jasmonic acid and SA accumulate in host plants via a complex network of response and biosynthesis pathways (Turner et al., 2002). Wound induced protein kinase (WIPK) plays a key role in the JA production induced by wound or biting of herbivores and can suppress SA accumulation (Turner et al., 2002). Phospholipase D (PLD), lipoxygenase (LOX), allene oxide synthase (AOS), allene oxide cyclase (AOC), OPDA reductase (OPR) are involved in the biosynthesis of JA from chloroplast membrane lipids in plants (Turner et al., 2002). In the biosynthesis of SA, phenylalanine ammonia lyase (PAL) catalyzes the first step of the biosynthesis of SA (Shah, 2003; Chen et al., 2009), and SA can also induce the expression of pathogenesis-related gene (PR; Turner et al., 2002; D'Maris et al., 2010). In addition, isochorismate synthase (ICS) that was discovered in bacterial has been shown to be the key enzyme in the biosynthesis pathway of SA (Shah, 2003; Chen et al., 2009).

So far, few studies have directly addressed the herbicide residue problem in contaminated soils and the effect on susceptible crops. In this study, we focus on the ecological effect of BSM on tobaccos and on the interactions between the plants and pests/viruses. We also investigate physiological changes involving JA and SA signaling pathway induced by the herbicide residue.

### MATERIALS AND METHODS

### Plant Materials, Insects, and Viruses

Seedlings of Nicotiana tabacum K326 were germinated under the following conditions: 28◦C, 16 h of light and 8 h of dark. We used 0.25 mg·kg−<sup>1</sup> as the most appropriate concentration of BSM to mimic the herbicide residue in the soil (Dry Weight, DW; Saeki and Toyota, 2004). Ten days post chemical-treatment, the symptoms caused by the herbicide were clearly visible. Plants not treated with BSM were used as controls. Treated and untreated (control) plants were then used for different experiments, i.e., pests infestation, pests feeding, Virus-induced symptom-evaluation, TMV inoculation RNA extraction, RTqPCR assay, and JA and SA contents determination.

Non-viruliferous B. tabaci, M. persicae, and TMV strains isolated from Fujian Province (China) were kept on different plants of N. tabacum in the laboratory to be used for the investigation.

### B. tabaci Infestation

In the experiments, plants were tested pairwise. One plant has been treated with BSM in the soil for 10 days, and the other was an untreated control. Each pair of the plants were infested with 50 adults of non-viruliferous whitefly, who could choose any plant to feed on. We chose to calculate the number of adults on the third to fifth leaves at 1, 3, 5, 10, 15, and 20 days post the infestation because the majority of adults chose to feed on those leaves. At the same time the size of corresponding leaves was measured in order to calculate the density of insects by the equation below:

$$\mathcal{S} = \, 0.6345ab$$

S, The size of the leaf; a, The length of the leave; b, The width of the leaf.

This leaf area calculation method was based on the previous research (Sun et al., 2006) then we made the modification according to the shape of the leaves we tested.

Totally 18 pair of plants were replicated to test, so we collected the data of B. tabaci adults number from 54 pairs of leaves. The density of eggs deposited by adults on corresponding leaves was also calculated 3 days post the infestation.

### M. persicae Feeding

fpls-07-01970 December 23, 2016 Time: 9:59 # 3

Leaves of the plants treated using the method mentioned in section "Plant Materials, Insects, and Viruses" were harvested and cut into the pieces of 1.5 cm<sup>2</sup> × 1.5 cm<sup>2</sup> . Then each piece of leaf was placed in a dish together with a moist piece of filter paper in order to keep the leaf fresh. Then one aphid nymph born within the last 5 h was placed on the leaf. The length of each nymph stage and adult stage, and the number of offspring was recorded by observation in a fixed time (3:00 p.m.) each day. The leaves for the feeding were changed every second days. Fourty bio-replicates were used. The longevity of each instar of nymphs and adults was recorded as well as the number of offspring laid by the adults.

### Symptoms Evaluation for TMV Assay

For the TMV assay a mechanical inoculation method was used. First we extracted the juice via grinding the virusinfected plant tissues contained TMV with 1 mL PBS buffer. Then we rubbed the same volume of juice we acquired on each inoculated leaves with emery. Finally we sprayed water on the inoculated leaves in order to wash away the emery. After the virus-inoculation, the average time before appearance of infection symptoms was recorded in order to figure out whether the herbicide residue can accelerate or slow down the progress of virus-infection. Consequently, disease severity was evaluated 5, 10, 15, and 20 days after the inoculation, according to Zhao et al. (2007) with the following modifications.

We divided the symptom into five degrees:

0 – no symptom;

1 – slight symptom including slight mosaic or slight deformation of no more than 1/3 of the leaves;

2 – 1/3 to 1/2 of leaves showed of mosaic or deformation;

3 – 1/2 to 2/3 of leaves had TMV correlated symptoms;

4 – more than 2/3 of leaves had TMV correlated symptoms.

Thirty bio-replicates were sued in this experiment. In addition to the above, all leaves were photographed for the records so that it is possible to do a re-evaluation and also make comparisons with other experiments.

### RNA Extraction and Reverse Transcription

A fixed amount of leaf samples (0.15 g) were harvested and ground in liquid nitrogen 10 days post the treatment of herbicide residue. Total RNA was extracted using the Tiangen plant RNA extraction prep kit (No. DP-432) following the manufacturers protocol and samples were collected in 1.5 mL Eppendorf tubes and kept at −80◦C for the further analysis.

Total RNA was used for reverse transcription according to the instruction for the Promega reverse transcripts kit (No. A5001). The resulting cDNA used later as template was kept at −20◦C.

### Determination of TMV Viral Replication

Tobacco mosaic virus inoculation was described in section "Symptoms Evaluation for TMV Assay," while the method for total RNA extraction was given in section "RNA Extraction and Reverse Transcription." We used RT-qPCR method was used to compare the amount of TMV in different treated plants. A SYBR Green kit (Promega, Product No. A6002) was used for this experiment. The 2−11Ct method was used to quantify the amount of TMV. NtACTIN (GenBank No. U60495.1) was used as housekeeping gene for data normalization. The TMV sequence was downloaded from NCBI (GenBank No. AF395127.1). Twelve biological replicates (each biological replicate includes three technical replicates) were performed for this assay.

### Endogenous JA and SA

Jasmonic acid and SA extraction and determination protocols has been described in earlier studies (Pan et al., 2010). High performance liquid chromatography and mass spectrometry (LC-MS) methods were used for the experiments. The conditions used for the analysis was described by Pan et al. (2010). This assay was completed at Institute of Genetics and Developmental Biology, Chinese Academy of Science.

The RT-qPCR method and the 2−11Ct were used to determine the expression of genes located in signal pathways (Turner et al., 2002; Shah, 2003; Chen et al., 2009; Katou et al., 2013). NtWIPK (Access No. D61377.1), NtPDF1.2 (Access No. JQ654634.1), NtLOX (Access No. X84040.1), NtPLD (Access No. AF223573.1), NtAOS (Access No. AB778304.1), NtAOC (Access No. AJ308487.1), NtOPR (Access No. AB233416.1) were used for analysis to reflect the endogenous JA levels in plants exposed to BSM treatment. The housekeeping gene NtActin (Access No. U60495.1) in N. tabacum was used as internal reference to correct the result. NtPR1a (GenBank No. X12737.1), NtPAL (GenBank No. AB008199.1), NtSABP (GenBank No. AY485932.1), NtICS1 (GenBank No. AY740529.1), and NtSIPK (GenBank No. U94192.1) was used to represent the endogenous SA response to soil BSM treatment. NtACTIN (GenBank No. U60495.1) and NtG6PDH (GenBank No. AJ001772.1) were used as housekeeping genes for data normalization. NtRDR1 (GenBank No. XM\_016610634.1) and NtRDR6 (GenBank No. FJ966891.1) were studied as viral infection related genes.

### Data Analyses

In pest-infestation experiment, students' t-tests were used to compare the density of adults on the leaves belonging to BSM treatment to controls at different time points. The same method was used to compare the egg-depositing by the adults, the symptom levels caused by TMV inoculation, TMV amount in plants, and the level of endogenous JA and SA as well as the expression of their responsive and biosynthetic genes. The statistical analysis was performed using of the statistical package SPSS19.0 (SPSS Inc., Chicago, IL, USA).

We used "Age-stage, Two-sex life" table method to evaluate the lifespan of M. persicae feeding on different types of leaves (Chi and Liu, 1985; Chi, 1988). The results included the longevity of

each insect stage, the morality of nymphs, net reproductive rate (R0), gross reproduction rate (GRR), mean fecundity per female adult (F), intrinsic rate (rm), finite rate of population (λ), mean generation time (T), and longevity, the meaning of these results can be referred to the study performed by Huang and Chi (2013).

### RESULTS

### BSM Residue Alters the Interaction between Host Plants and Whitefly B. tabaci

First, we tested the effects of different concentrations of BSM treatment of soil (2.0, 1.0, 0.25, 0.1, 0.05, and 0 mg·kg−<sup>1</sup> as the control) on tobaccos for B. tabaci infestation and found that 0.25 mg·kg−<sup>1</sup> is a suitable concentration for further experiments, as recommended in the fields applications. Higher concentrations could kill the plants, while the lower concentrations showed no noticable effect on susceptible plants for B. tabaci infestation.

Bensulfuron-methyl-induced symptoms such as malformed leaves and chlorosis on treated plant tissue became apparently before the infestation of pests. Both of the densities of adult B. tabaci feeding on two different treated tobaccos (**Figure 1**) reduced quickly. The difference of densities of adult B. tabaci between control and treated plants were not significant during the early infection, but were increasingly apparent after 10 days, with F = 4.385, df = 106, p = 0.014 at 15 days and F = 6.361, df = 106, p = 0.013 at 20 days post-infestation, as BSM treatment reduced the infestation of B. tabaci.

We also counted the number of eggs on correlated leaves deposited by adult B. tabaci and determined the breeding preference of whitefly adults. But there was no significant difference in breeding preference with or without BSM-treatment according to the students' t-test (F = 1.818, df = 106, p > 0.05).

### Effect of BSM on M. persicae Infestation of Plants

Aphids such as M. persicae fed on tobacco leaves similar to the whitefly experiments described above, but our result showed no significant difference in the longevity of each stage of M. persicae feeding on BSM-treated and control leaves, also feeding on BSMtreated plants had no remarkable impact on the morality of nymphs (**Table 1**).

Then we determined the longevity and the fecundity of the insects and we also found no significant difference between the groups feeding on BSM-treated and healthy plants (**Table 2**). All the results shown in the table were acquired using the method of Chi and Liu (1985) and Chi (1988).

### BSM Treatment Reduce the Symptom of TMV Infection

Tobacco mosaic virus was inoculated on tobacco plants after 10 days of BSM treatment, and viral lesions appeared at about the same time as the control plants without BSM treatment, although the lesion was less severe as evidenced by lighter color (**Figure 2A**). But after the time point of 10 days post virus-inoculation, TMV-caused lesion on BSM-treated plants

TABLE 1 | The longevity and the nymph-morality of M. persicae feeding on leaves from BSM-treated and control tobaccos.


In this table, fecundity value = 1, time unit = 1, n = 40. The data consists of mean value ± SE.

TABLE 2 | he lifetable of M. persicae feeding on leaves of BSM-treated and control tobaccos.


In this table, fecundity value = 1, time unit = 1, n = 40. The data consists of mean value ± SE.

was not less severe any more. This result was in keeping with the evaluation of the symptom levels. BSM-treated plants also had lower level of virus-symptoms compared to the plants not treated by herbicide. As the development of the virus-induced symptom, and the accumulation of the virus, there was no significant difference in the viral lesions and symptoms if TMV was inoculated 15 or 20 days after BSM treatment (**Figure 2B**).

### BSM Treatment Reduces Early RNA Replication

Next we determined the effect of BSM treatment on TMV RNA replication upon infection of the susceptible tobacco strain K326, in comparison to control plants without BSM treatment. In this regard, we used RT-qPCR to determine the level of coat protein (CP) to reflect virus replication in these tobacco plants, and found that early viral replication was significantly lower in BSMtreated tobacco leaves (**Figure 3A**). Interestingly, the situation changed during later viral infection, with dramatic increase of viral RNA replication in BSM-treated plants at 15- and 20-day post-infection, indicating that BSM treatment only delayed viral RNA replication rather than directly blocking the process per se.

Since our result showed that BSM residue in the soil strongly affected viral RNA replication in the plants, we wondered if lower BSM concentrations were also effective. To this end, we performed the same TMV experiments as described above except with fivefold lower BSM concentration (0.05 mg·kg−<sup>1</sup> ).

Our result showed no significant difference in early viral replication (10 days post-infection) with or without BSM treatment (F = 24.413, df = 22, p > 0.05), while this low BSM concentration remained effective in boosting late viral RNA replication (20 days post-infection) albeit to a lesser extent than higher BSM concentrations (**Figure 3B**).

### BSM Treatment Induces JA and SA Production, As Well As the Expression of Responsive and Biosynthetic Genes

Jasmonic acid and SA signal transduction pathways are important for plant immune system to response to pathogen infections. Among the responsive and biosynthetic genes, we found that BSM treatment up-regulated only NtWIPK gene in the JA pathway (F = 12.558, df = 4, p = 0.024; **Figure 4A**), but upregulated the expression of two genes, NtPAL (F = 3.910, df = 4, p = 0.009) and NtPR1a (F = 4.375, df = 4, p < 0.001), in the SA pathway (**Figure 4B**).

We determined the levels of JA and SA in BSM-treated and control plants. JA showed strong up-regulation (F = 4.649,

df = 4, p = 0.029) 10 days after the BSM treatment of the soil (**Figure 4C**), consistent with the RT-qPCR results described above. In addition, BSM treatment also up-regulated the SA level in these plants (F = 0.493, df = 4, p = 0.014) also showed the same result as JA level in those plants (**Figure 4D**), indicating that both of the phytohormones were strongly induced by the herbicide-treatment.

### TMV Infection Induces Gene Expression Involved in SA Response and Biosynthesis

It was remarkable that the BSM-treated plants increased TMV replication during the later viral infection. In this regard, SA as well as related genes was significantly up-regulated in BSMtreated plants compared to the controls. Both PR1a and PAL play a key role in endogenous SA accumulation (Shah, 2003; Chen et al., 2009). Thus, we determined the expression of NtPR1a and NtPAL at 10- and 20-day post-inoculation, and found that both NtPR1a (F = 4.278, df = 4, p < 0.001) and NtPAL (F = 5.540, df = 4, p < 0.001) was down regulated in BSMtreated tobaccos 10-day post-TMV inoculation (**Figure 5A**), while NtPR1a (F = 13.854, df = 4, p = 0.007) expression recovered better in BSM-treated plants at 20-day post-TMV inoculation than in control plants (**Figure 5B**). The expression of NtPAL was not significantly different between treated and control plants at 20-day post-inoculation. Our result suggested that the SA level was suppressed by the TMV accumulation during early infection in BSM treated plants, which could contribute to the higher levels of TMV replication during late infection.

Since, RDR family of RNA polymerase (RDR) can also be regulated by endogenous SA (Lee et al., 2016). RDR1-6 contributes to antiviral silencing and symptom-limiting in plants (Lee et al., 2016), we determined the expression of NtRDR1 and NtRDR6 before the inoculation of TMV. However, the expression of both NtRDR1 (F = 4.408, df = 4, p = 0.163) and NtRDR6 (F = 1.194, df = 4, p = 0.379) didn't show significant change (F = 4.408, df = 4, p = 0.163) upon BSM treatment (**Figures 5C,D**). Though both of RDR1 and RDR6 play substantial roles to assist plants to fight against virus, our data suggests that they are not involved the BSM-mediated response to TMV infection. We also determined the expression of NtRDR1 and NtRDR6 after the inoculation in both BSM-treated and control plants. At 20-day post-inoculation when there was high levels of TMV replication, NtRDR1 showed remarkable down regulation (F = 4.478, df = 4, p = 0.003) while the expression of NtRDR6 was significantly induced (F = 12.282, df = 4, p = 0.036).

### DISCUSSION

Herbicide residue draws much more attention today than ever before because pose threaten to the crop yield and the safety of not only the environment but also the humanity (Nicosia et al., 1991). And great effort was made to solve the problem such as the detection of the residue and the biodegradation in the soil (Lin et al., 2012; Yang et al., 2012; Feng et al., 2013), or the target site in the target weeds in order to reduce the usage (Wei et al., 2015). An attempt was made to discover the interaction between crops and important pests/viruses under the residue of herbicide in the soil in our investigation.

We found that the residue of the herbicide could only cause the avoidance to B. tabaci of plants during the later infestation. The performance of the wingless insect M. persicae, as well as the fecundity of both insects were not affected by the herbicide. However, the residue of BSM can strongly affect the activity of virus in plants. Thus we discover that JA and SA, which play significant roles in biotic interactions in plants, were both induced under the residue of the herbicide in the soil. We also found that SA pathway could be inhibited under the simultaneous existence of TMV and herbicide. We also focus on the factors such as RDRs related to the virus-resistance in plants and found

significant difference from the control type (n = 3, <sup>∗</sup>p < 0.05).

that they could only affected by the virus, but not the herbicide. In this study, the results we found needed to be discussed.

There has been few investigations focusing on the interaction between plants and insects under the application of other herbicides such as atrazine or other kinds of sulfonylurea herbicides (Dewey, 1986; Kjaer and Heimbach, 2001). In the study of atrazine in 1986, the researchers focused on the community of the insects and suggested that herbicides affected the community of the insects community indirectly by reducing the food and habitat (Dewey, 1986). In the study of sulfonylurea herbicide application on different host plants which were not sensitive, the activity of insects feeding on them were not severely affected except the death rate of Gastrophysa polygoni larvae (Kjaer and Heimbach, 2001). In our investigation, first of all, tobaccos are sensitive to BSM residues and show remarkable symptoms. In the second place, we study the interactions between the plants and insects from the aspect of physiology including the changing of JA and SA production. According to the fecundity determination assay, in some investigations, the activity of the pests can increase the production of SA (Shi et al., 2014b; Cao et al., 2016), while SA in some cases can also inhibit the activity of pests (Rodriguez-Alvarez et al., 2015). Although in our investigation, the fecundity of both M. persicae and B. tabaci was not significantly affected, this was in correlation with the study performed by Kjaer and Heimbach (2001). This also indicates the limitation of herbicide on the activity of insects, because the target are in the weeds and not in the insects.

Similarly, few studies have paid attention to the interaction between plants and pathogens as a consequence of herbicidetreatment of. We have found no study focussing on the effect of sulfonylurea herbicide for this interaction. The speculative effects of herbicide application on the invasion of viruses can be divided into the following: plants were not affected by herbicide; plants were affected by the herbicide, the virus concentration was suppressed, enhanced or not affected (Kazinczi et al., 2003). Previous investigations has generally showed that the activity of viruses are inhibited by the application of herbicides (Kazinczi et al., 2006; Hooks et al., 2009). Krcatovic et al. (2008) point out that the treatment of some metabolites such as flavonoids, quercetin, and vitexin on plants also can only suppress the activity of TMV during early infection accompanied with the induce of SA, but the phenomenon cannot last for a long time. This was only consistent with the early infection of TMV in tobaccos treated with BSM in our study. Furthermore, the susceptibility of BSM-treated tobaccos to TMV during the later infection was first shown in the present investigation.

In many studies performed before, JA and SA pathways are presented as antagonists (Turner et al., 2002; Lyons et al., 2013). The suppression of target genes related to JA can lead to the accumulation of SA (Turner et al., 2002). In some studies focused on the plant-insect interactions, the infestation of pests can lead to the suppression of JA level and the enhancement of SA

production (Shi et al., 2014b). Some investigations focusing on the relationship between herbicides and SA pointed out that SA and H2O<sup>2</sup> shared a common signaling pathway, this indicated that SA in rapeseed could be boosted by the treatment of other herbicide such as napropamide (Cui et al., 2010). According to the results of the TMV assay above, the concentration of virus is lower in BSM-treated plants during early infection, this correlates with the result that TMV can be inhibited by the exogenous treatment of SA (Conti et al., 2012). However, the fact that not only increased SA level in sensitive plants exposed to the treatment of sulfonylurea herbicide, but the JA production was affected by the application of herbicide was first discovered in our study, and the result that both of them showed up-regulation are different to previous studies.

In the further investigation, higher expression of genes in SA signal pathway was inhibited by lower concentration of TMV in BSM-treated plants. Thus, we speculate the inhibited SA signal pathway can lead to the higher viral replication. And in this regard, SA production in BSM-treated tobaccos can be more easily blocked in the presence of both virus-inoculation and herbicide-treatment. In this assay, we didn't take JA into consideration because the role of JA involved in the anti-virus is still unclear. Oka et al. (2013) discovered the negative impact of JA on virus-resistance in plants because they found that the application JA in tobaccos (Shannon variety, which is resistant to TMV) can lead to the lesion on leaves. Zhu et al. (2014) performed the exogenous JA and JA-defective mutants assay and discovered that without JA, the plants could be more sensitive to viruses. But the mechanism that SA biosynthesis can be suppressed in plants exposed to the presence of both virus and BSM needs to be further investigated.

RNA-dependent RNA polymerase family though plays a key role in resistance to the infection by the virus. It can mediate the conversion of viral single-strand RNA to double-strand RNA, then lead to further degradation (Harmoko et al., 2013). The function of RDR1 and RDR6 was studied most. But those studies mainly focused on virus-resistance (Di Serio et al., 2010; Jiang et al., 2012). For example, Rodriguez et al. (2014) found that RDR1 located in the downstream of SA signal could be inhibited by the infection of TMV. In addition, RDR6 in rice plays key roles in fighting against not only viruses, but also fungus and bacteria (Wagh et al., 2016). The interaction between RDR family and herbicide was not reported in any study. In our investigation, the activity of RDR family cannot be affected by the residue of herbicide in the soil, though the signal pathway of SA, as well as the infection of virus were affected by the herbicide.

### CONCLUSION

The study investigated the performance of the pests and viruses in susceptibility plants N. tabacum treated with BSM soil residues. The results pointed out that plants can be more resistant to

B. tabaci only during the laer time post-infestation in treatment when BSM residues are present in the soil. However, the longevity of both development stage and life of M. persicae was not affected, nor was the fecundity of both insects. Although plants treated with BSM could be more resistant to TMV accumulation during the early infection, the viral RNA replication was enhanced during the late infection in these plants, and this phenomenon was weaker at lower concentration of BSM. The expression of marker genes related to JA and SA was up-regulated, as well as the contents of both phytohormones. Although, during the early infection of TMV in BSM-treated plants, the expression of SA targeted genes was strongly inhibited, thus we speculate that this was beneficial for the multiplication of TMV. In addition, both of NtRDR1 and NtRDR6 were not involved in the anti-virus which was affected by the application of BSM in our study. This study revealed the effect brought by the residue of herbicide in the soil on fighting against pests and viruses of plants. The correlated physiological phenomenon was also displayed. The study can contribute to the study on the ecosystem of cropland suffering the pollution of the herbicide.

### AUTHOR CONTRIBUTIONS

RL completed the whole research work, summarized the correlated data, and composed the draft. SUI contributed to the

### REFERENCES


assistance of research work and the English correction of the writing. ZW and XY provided the guidance of the work and made some suggestions about the work.

### FUNDING

This program is supported by the Key Project of the Fujian provincial department of science and technology (2012N4001) and the Key Project of the National Research Program of China (2012BAD19B03).

### ACKNOWLEDGMENTS

We thank the expertise of Ms. Shuang Fang and Dr. Jinfang Chu (National Centre for Plant Gene Research, Beijing, China), The Plant Hormone Facility of the Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China) for determining the JA and SA contents of N. tabacum. We also would like to thank Professors Guang-Pu Li and Stefan Olsson at Fujian Agriculture and Forestry University for their critical editing of the manuscript.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer JV and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review. The reviewer ZW declared a shared affiliation, though no other collaboration, with the authors to the handling Editor, who ensured that the process nevertheless met the standards of a fair and objective review.

Copyright © 2016 Li, Islam, Wu and Ye. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpls-07-01970 December 23, 2016 Time: 9:59 # 10

# Isolation and Identification of Plant Growth Promoting Rhizobacteria from Cucumber Rhizosphere and Their Effect on Plant Growth Promotion and Disease Suppression

*Shaikhul Islam1, Abdul M. Akanda1, Ananya Prova1, Md. T. Islam2 and Md. M. Hossain3\**

*<sup>1</sup> Department of Plant Pathology, EXIM Bank Agricultural University, Chapainawabganj, Bangladesh, <sup>2</sup> Department of Biotechnology, Bangabandhu Sheikh Mujibur Rahman Agricultural University, Gazipur, Bangladesh, <sup>3</sup> Department of Plant Pathology, Bangabandhu Sheikh Mujibur Rahman Agricultural University, Gazipur, Bangladesh*

#### *Edited by:*

*Anton Hartmann, Helmholtz Zentrum München – German Research Center for Environmental Health, Germany*

#### *Reviewed by:*

*Abu Hena Mostafa Kamal, University of Texas at Arlington, USA Christian Staehelin, Sun Yat-sen University, China*

> *\*Correspondence: Md. M. Hossain hossainmm@bsmrau.edu.bd*

#### *Specialty section:*

*This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology*

*Received: 16 August 2015 Accepted: 16 November 2015 Published: 02 February 2016*

#### *Citation:*

*Islam S, Akanda AM, Prova A, Islam MT and Hossain MM (2016) Isolation and Identification of Plant Growth Promoting Rhizobacteria from Cucumber Rhizosphere and Their Effect on Plant Growth Promotion and Disease Suppression. Front. Microbiol. 6:1360. doi: 10.3389/fmicb.2015.01360*

Plant growth promoting rhizobacteria (PGPR) are the rhizosphere bacteria that may be utilized to augment plant growth and suppress plant diseases. The objectives of this study were to identify and characterize PGPR indigenous to cucumber rhizosphere in Bangladesh, and to evaluate their ability to suppress Phytophthora crown rot in cucumber. A total of 66 isolates were isolated, out of which 10 (PPB1, PPB2, PPB3, PPB4, PPB5, PPB8, PPB9, PPB10, PPB11, and PPB12) were selected based on their *in vitro* plant growth promoting attributes and antagonism of phytopathogens. Phylogenetic analysis of 16S rRNA sequences identified these isolates as new strains of *Pseudomonas stutzeri*, *Bacillus subtilis*, *Stenotrophomonas maltophilia,* and *Bacillus amyloliquefaciens*. The selected isolates produced high levels (26.78–51.28 μg mL−1) of indole-3-acetic acid, while significant acetylene reduction activities (1.79–4.9 μmole C2H4 mg−<sup>1</sup> protein h−1) were observed in eight isolates. Cucumber plants grown from seeds that were treated with these PGPR strains displayed significantly higher levels of germination, seedling vigour, growth, and N content in root and shoot tissue compared to non-treated control plants. All selected isolates were able to successfully colonize the cucumber roots. Moreover, treating cucumber seeds with these isolates significantly suppressed Phytophthora crown rot caused by *Phytophthora capsici*, and characteristic morphological alterations in *P. capsici* hyphae that grew toward PGPR colonies were observed. Since these PGPR inoculants exhibited multiple traits beneficial to the host plants, they may be applied in the development of new, safe, and effective seed treatments as an alternative to chemical fungicides.

Keywords: PGPR, plant growth promotion, IAA production, biological nitrogen fixation, antagonism, *Phytophthora capsici*, disease suppression

## INTRODUCTION

The cucumber (*Cucumis sativus*) is one of the most widely grown vegetable crops in the world, and is particularly prevalent on the Indian sub-continent. This crop is prone to massive attacks by *Phytophthora capsici* that causes crown rot and blight (Kim et al., 2008; Maleki et al., 2011). *P. capsici* infects susceptible hosts throughout the growing season at any growth stage, and causes yield losses as high as 100% (Lee et al., 2001). This pathogen has a wide host range with more than 50 plant species including Cucurbitaceae, Leguminosae, and Solanaceae (Hausbeck and Lamour, 2004). Although fungicides can control the disease, their use is detrimental to the surrounding environment and to the viability and survival of beneficial rhizosphere microbes (Carson et al., 1962; Hussain et al., 2009; Heckel, 2012). Furthermore, the growing cost of pesticides and the consumer demand for pesticide-free food have led to a search for substitutes for these products. Thus, there has been a need to find effective alternatives to costly and environmentally degrading synthetic pesticides.

Rhizobacteria that benefit plants by stimulating growth and suppressing disease are referred to as plant growth promoting rhizobacteria (PGPR; Kloepper et al., 1980). PGPR have been tested as biocontrol agents for suppression of plant diseases (Gerhardson, 2002), and also as inducers of disease resistance in plants (Cattelan et al., 1999; Bargabus et al., 2002; Bais et al., 2004). In particular, strains of *Pseudomonas*, *Stenotrophomonas,* and *Bacillus* have been successfully used in attempts to control plant pathogens and increase plant growth (Bais et al., 2004; Idris et al., 2007; Liu et al., 2007; Messiha et al., 2007; Chen et al., 2009; El-Sayed et al., 2014). The widely recognized mechanisms of plant growth promotion by PGPR are production of phytohormones, diazotrophic fixation of nitrogen, and solubilization of phosphate. Mechanisms of biocontrol action include competition with phytopathogens for an ecological niche or substrate, as well as production of inhibitory compounds and hydrolytic enzymes that are often active against a broad spectrum of phytopathogens (Zhang and Yuen, 2000; Manjula et al., 2004; Haas and Défago, 2005; Stein, 2005; Detry et al., 2006; Konsoula and Liakopoulou-Kyriakides, 2006; Cazorla et al., 2007).

Many PGPR have been shown to reduce Phytophthora crown rot occurrence on various plants. Ahmed et al. (2003) demonstrated *in vitro* suppression of *P. capsici* by bacterial isolates from the aerial part and rhizosphere of sweet pepper. An endophytic bacterium isolated from black pepper stem and roots, *B. megaterium* IISRBP17 suppressed *P. capsici* on black pepper in greenhouse assays (Aravind et al., 2009). Zhang et al. (2010) demonstrated that PGPR strains used separately or in combinations had the potential to suppress Phytophthora blight on squash in the greenhouse. Shirzad et al. (2012) reported that some fluorescent pseudomonads isolated from different fields of East and West Azarbaijan and Ardebil provinces of Iran exhibited strong antifungal activity against *P. drechsleri* and controlled crown and root rot of cucumber caused by the pathogen. However, little is known about PGPRs with the potential to suppress crown rot caused by *P. capsici* on cucumber. Furthermore, the plant-growthpromoting and biocontrol efficacy of PGPR often depend upon the rhizosphere competence of the microbial inoculants (Lugtenberg and Kamilova, 2009). Rhizosphere competence refers to the survival and colonization potential of PGPR (Bulgarelli et al., 2013), and is thought to be highest for each PGPR strain when associated with its preferred host plant. This to some extent explains why some PGPR strains exhibiting promise as biocontrol agents *in vitro* have variable biocontrol efficacy in the rhizosphere of a given crop under a given set of conditions. The identification and characterization of PGPR populations indigenous to cucumber rhizospheres is therefore critical to discovery of strains that can be utilized to improve growth and Phytophthora crown rot suppression in cucumber. The objectives of the present study were to isolate bacterial strains from the cucumber rhizosphere, to characterize these isolates on the basis of morphological and physiological attributes as well as by 16S rRNA sequence analysis, and to assess the plant growth promoting effects of these isolates *in vivo* and their ability to suppress Phytophthora crown rot in cucumber plants. To our knowledge, this is the first report of PGPR reducing *P. capsici* infection on cucumber.

### MATERIALS AND METHODS

### The Study Site

The experimental site was located at the Field Laboratory of the Department of Plant Pathology, Bangabandhu Sheikh Mujibur Rahman Agricultural University (BSMRAU), Gazipur, Bangladesh. The location of the site is at 24.09◦ N latitude and 90.25◦ E longitude with an elevation of 2–8 m. The study area is within the Madhupur Tract agro-ecological zone (AEZ 28). The soil used for pot experiments belongs to the Salna series and has been classified as "swallow red brown terrace soil" in the Bangladesh soil classification system, which falls under the order Inceptisol (Brammer, 1978). This soil is characterized by clay within 50 cm of the surface and is slightly acidic in nature. The pH value, cation exchange capacity (CEC) and electrical conductivity (EC) of bulk soil samples collected from the study site were 6.38, 6.78 meq 100 g−<sup>1</sup> soil and 0.6 dS m<sup>−</sup>1, respectively. This soil contained 1.08% organic carbon (OC), 1.87% organic matter (OM), 0.10% nitrogen (N), 9 ppm phosphorus (P) and 0.20 meq 100 g−<sup>1</sup> soil exchangeable potassium (K).

### Plant Material, Bacterial Isolation, and Pathogenic Organism

Cucumber (*Cucumis sativus* L.) variety Baromashi (Lal Teer Seed Company, Dhaka, Bangladesh) root samples were collected from the study site along with rhizosphere soil. For isolation of bacteria, 2–5 *g* of fresh roots were washed under running tap water and surface sterilized in 5% NaOCl for 1 min. After washing three times with sterilized distilled water (SDW), the root samples were ground with a sterilized mortar and pestle. Serial dilutions were prepared from the ground roots, and <sup>100</sup> <sup>μ</sup>l aliquots from each dilution of 1 <sup>×</sup> <sup>10</sup><sup>−</sup>6, 1 <sup>×</sup> <sup>10</sup>−7, and <sup>1</sup> <sup>×</sup> <sup>10</sup>−<sup>8</sup> CFU mL−<sup>1</sup> were spread on potato dextrose agar (PDA) plates and incubated for 2 days at 28 ± 2◦C. Morphologically distinct bacterial colonies were selected for further purifications. The purified isolates were preserved temporarily in 20% glycerol solution at −20◦C. The pathogen *P. capsici* was provided by Prof. W. Yuanchao, Nanjing Agricultural University, China.

### Morphological and Biochemical Characterization of Bacterial Isolates

Colony morphology, size, shape, color, and growth pattern were recorded after 24 h of growth on PDA plates at 28 ± 2◦C as described by Somasegaran and Hoben (1994). Cell size was observed by light microscopy. The Gram reaction was performed as described by Vincent and Humphrey (1970). A series of biochemical tests were conducted to characterize the isolated bacteria using the criteria of Bergey's Manual of Systematic Bacteriology (Bergey et al., 1994). For the KOH solubility test, bacteria were aseptically removed from Petri plates with an inoculating wire loop, mixed with 3% KOH solution on a clean slide for 1 min and observed for formation of a thread-like mass. The motility of each isolate was tested in sulfide indole motility (SIM) medium. Using a needle, strains were introduced into test tubes containing SIM, and were incubated at room temperature until the growth was evident (Kirsop and Doyle, 1991). Turbidity away from the line of inoculation was a positive indicator of motility. Catalase and oxidase tests were performed as described in Hayward (1960) and Rajat et al. (2012), respectively. To determine whether the rhizobacterial isolates are better suited to aerobic or anaerobic environments, the citrate test was conducted according to Simmons (1926) using Simmons citrate agar medium. All experiments were done following complete randomized design (CRD) with three replications for each isolate and repeated once.

### Molecular Characterization of Bacterial Isolates

Culture DNA was obtained using the lysozyme-SDSphenol/chloroform method (Maniatis et al., 1982). DNA was extracted with phenol-chloroform-isoamyl alcohol (25:24:1) and precipitated with isopropanol. The extracted DNA was treated with DNase-free RNase (Sigma Chemical Co., St. Louis, MO, USA) at a final concentration of 0.2 mg/ml at 37◦C for 15 min, followed by a second phenol-chloroform-isoamyl alcohol extraction and isopropanol precipitation. Finally, the DNA pellet was re-suspended in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0), stored at −20◦C, and used as template DNA in PCR to amplify the 16S rRNA for phylogenetic analysis.

16S rRNA gene amplification was performed by using the bacterial-specific primers, 27F (5 -AGAGTTTGATCCTGG CTCAG-3 ) and 1492R (5 -GGTTACCTTGTTACGACTT-3 ) (Reysenbach et al., 1992). PCR amplifications were performed with 1 × Ex Taq buffer (Takara Bio Inc, Japan), 0.8 mM dNTP, 0.02 units μl <sup>−</sup><sup>1</sup> Ex Taq polymerase, 0.4 mg ml−<sup>1</sup> BSA, and 1.0 μM of each primer. Three independent PCR amplifications were performed at an annealing temperature of 55◦C (40 s), an initial denaturation temperature of 94◦C (5 min), 30 amplification cycles with denaturation at 94◦C (60 s), annealing (30 s), and extension at 72◦C (60 s), followed by a final extension at 72◦C (10 min). The PCR product was purified using Wizard<sup>R</sup> PCR Preps DNA Purification System (Promega, Madison, WI, USA). Purified double-stranded PCR fragments were directly sequenced with Big Dye Terminator Cycle sequencing kits (Applied Biosystems, Forster City, CA,

USA) using the manufacturer's instructions. Sequences for each region were edited using Chromas Lite 2.01 . The 16S rRNA sequence of the isolate has been deposited in the GenBank database. The BLAST search program2 was used to search for nucleotide sequence homology for the 16S region for bacteria. Highly homologous sequences were aligned, and neighborjoining trees were generated using ClustalX version 2.0.11 and MEGA version 6.06. Bootstrap replication (1000 replications) was used as a statistical support for the nodes in the phylogenetic trees.

### Bioassays for Plant Growth Promoting Traits

### Biological Nitrogen Fixation

Nitrogenase activity of isolates was determined via the acetylene reduction assay/ethylene production assay as described in Hardy et al. (1968). Pure bacterial colonies were inoculated to an airtight 30 ml vial containing 10 ml nitrogen-free basal semisolid medium, and were grown for 48 h at 28 ± 2◦C. Following pellicle formation, the bottles were injected with 10% (v/v) acetylene gas and incubated at 28 ± 2◦C for 24 h. Ethylene production was measured using a G-300 Gas Chromatograph (Model HP 6890, USA) fitted with a Flame Ionization Detector and a Porapak-N column. Hydrogen and oxygen were used as a carrier gas, with a flow rate of 4 kg/cm2, and the column temperature was maintained at 165◦C. The ethylene concentration calibration curve was plotted for each trial, and the viable cell numbers (cfu) of the isolate were determined. The rate of N2 fixation was expressed as the quantity of ethylene accumulated (μmol C2H4 mg−<sup>1</sup> protein h−1) based on the standard curve and peak-area percentage.

### Indole-3-Acetic Acid Production

For detection and quantification of indole-3-acetic acid (IAA) production by bacterial isolates, isolated colonies were inoculated into Jensen's broth (Sucrose 20 g, K2HPO4 1gL−1, MgSO4 7H2O 0.5 g L−1, NaCl 0.5 g L−1, FeSO4 0.1 g L−1, NaMoO4 0.005 g L−1, CaCO3 2gL−1) (Bric et al., 1991) containing 2 mg mL−<sup>1</sup> L-tryptophan. The culture was incubated at 28 <sup>±</sup> <sup>2</sup>◦<sup>C</sup> with continuous shaking at 125 rpm for 48 h (Rahman et al., 2010). Approximately 2 mL of culture solution was centrifuged at 15000 rpm for 1 min, and a 1 mL aliquot of the supernatant was mixed with 2 mL of Salkowski's reagent and incubated 20 min in darkness at room temperature (150 ml concentrated H2SO4, 250 ml distilled water, 7.5 ml 0.5 M FeCl3.6H2O) as described by Gordon and Weber (1951). IAA production was observed as the development of a pinkred color, and the absorbance was measured at 530 nm using a spectrophotometer. The concentration of IAA was determined using a standard curve prepared from pure IAA solutions (0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, and 65 μg ml<sup>−</sup>1).

<sup>1</sup>http://www.technelysium.com.au/chromas.html

<sup>2</sup>http://blast.ncbi.nlm.nih.gov/Blast.cgi

### Preparation of Bacterial Inocula for Cucumber Seed Treatment

Bacterial strains were cultured in 250 ml conical flasks containing 200 ml yeast peptone broth on an orbital shaker at 120 rpm for 72 h at 28 ± 2◦C. Bacterial cells were collected via centrifugation at 15000 rpm for 1 min at 4◦C, and each pellet was washed twice with SDW. The bacterial pellets were suspended in 0.6 ml SDW, vortexed and used for seed treatment. Approximately 30–31 cucumber seeds were surface sterilized in 5% NaOCl for 1 min and washed three times in SDW. Dry seeds were immersed in each bacterial suspension, and the preparation was stirred frequently for 5 min. The treated seeds were spread on a petri dish and air dried overnight at room temperature. The number of bacterial cells per seed, determined via serial dilutions, was approximately 108 CFU/seed.

### Effect of Bacterial Seed Treatment on Germination and Vigour Index in Cucumber

In order to determine the effect of the isolates on germination and seedling vigour, 100 seeds inoculated with each isolate were incubated in ten 9-cm petri dishes on two layers of moistened filter paper. As a control treatment, seeds treated with water instead of bacterial suspensions were also established. In order to maintain sufficient moisture for germination, 5 ml distilled water was added to the petri dishes every other day, and seeds were incubated at 28 ± 2◦C in a light incubator. Germination was considered to have occurred when the radicles were half of the seed length. The germination percentage was recorded every 24 h for 7 days. Root and shoot length were measured after the seventh day. The experiment was planned as a completely randomized design with 10 replications for each isolate.

$$\text{Germination rate (\%)} = \left(\frac{\text{number of seconds generated}}{\text{total number of seconds}}\right) \times 100$$

Vigour index = % germination × total plant length

### Effect of Bacterial Seed Treatment on Growth and Nitrogen Content in Cucumber Plants

In order to test the ability of isolates to promote growth in cucumber plants, surface-sterilized cucumber seeds were inoculated with each isolate as described above. The soil from the study site described above was used as potting medium. After autoclaving twice at 24 h intervals at 121◦C and 15 psi for 20 min, 180 g of the sterilized soil was placed in each sterilized plastic pot (9.5 cm × 7.0 cm size). One pre-germinated cucumber seed was sown in each pot, and plants were grown 3 weeks in a net house with watering on alternate days. After harvest, the fresh weight, dry weight, and root and shoot lengths of the plants were measured. The shoots and roots were separated and dried in an oven at 68 ± 2◦C for 48 h, then ground for determination of tissue-N concentrations (Kjeldahl, 1883).

### Root Colonization

Root colonization by bacterial isolates was determined according to the protocol of Hossain et al. (2008). Roots were harvested from plants at 7, 14, and 21 days of growth. Root systems were thoroughly washed with running tap water to remove adhering soil particles, then were rinsed three times with SDW and blotted to dryness. Roots were divided into top, middle, and bottom regions, and were weighed and homogenized in SDW. Serial dilutions were prepared on PDA plates, and the number of colony-forming units (cfu) per gram root was determined after 24 h of incubation at 28 ± 2◦C.

### Pathogenicity of *P. capsici* in Cucumber

For preparation of zoospore inoculum, *P. capsici* was cultured on PDA plates at 18 ± 2◦C for 7 days. Five-mm blocks were then cut from the culture plates and placed in petri dishes containing SDW. The petridishes were incubated in darkness at room temperature for 72 h, followed by a 1-h cold treatment at 4◦C. Zoospore production was confirmed via light microscopy. In order to test the pathogenicity of *P. capsici* zoospores, cucumber seedlings were planted in pots containing 0, 500, 1000, or 1500 μl zoospores/pot. As 100% mortality was found in case 1000 μl zoospores/pot, two concentrations (500 and 1000 μl) of zoospores suspension per pot were fixed.

### *In Vitro* Screening for Antagonism

To test antagonism of *P. capsici* by each isolate, the pathogen and bacteria were inoculated 3 cm apart on the same agar plate. Fungal growth on each plate was observed, and the zone of inhibition, if present, was determined as described in Riungu et al. (2008):

$$\% \text{ Inhibition of mycellal growth} = \frac{\text{X} - \text{Y}}{\text{X}} \times 100\%$$

Where,

X = Mycelial growth of pathogen in absence of antagonists Y = Mycelial growth of pathogen in presence of antagonists

Morphologies of hyphae in the vicinity of bacterial colonies were observed under a light microscope (Meiji Techno: ML2600), and images were recorded with a digital camera (Model: Canon Digital IXUS 900 Ti). Each experiment was carried out following CRD with three replications for each isolate and repeated once.

### Testing the Effect of Rhizobacterial Seed Treatment on Phytophthora Crown Rot of Cucumber

Cucumber seeds inoculated with each isolate were sown and grown for 7 days in sterilized plastic pots as described above. Seven-day-old seedlings were inoculated with 500 or 1000 μl zoospore suspension/pot as described in Deora et al. (2005). Inoculated plants were kept inside humid chambers for 48 h. Each experiment included 12 plants per treatment with three replications. Surviving plants were counted 7 days after inoculation. Percent disease incidences (PDI) were calculated using the following formula:

$$\text{PDI} = \frac{\text{Number of infected plants}}{\text{Total number of inoculated plants}} \times 100\%$$

Percent protection by PGPR was calculated using following formula:

$$\% \text{ Projection} = \left[ \frac{\text{A} - \text{B}}{\text{A}} \right] \times 100\%$$

Where,

A = PDI in non-inoculated control plants B = PDI in PGPR-treated plants.

### Statistical Analysis

Statistical analyses were performed using SPSS (Version 17) and Microsoft Office Excel 2007. A completely randomized design was used for all experiments, with 3–12 replications for each treatment. The data presented are from representative experiments that were repeated at least twice with similar results. Treatments were compared via ANOVA using the least significant differences test (LSD) at 5% (*P* ≤ 0.05) probability level.

### RESULTS

### Strain Isolation and Biochemical and Molecular Characterization

We obtained a total of 66 rhizobacterial strains from the interior of cucumber roots. Ten isolates – PPB1, PPB2, PPB3, PPB4, PPB5, PPB8, PPB9, PPB10, PPB11, and PPB12 – were selected based on their ability to produce IAA, fix N2, and show *in vitro* antagonism against various pathogens in a preliminary screening. All isolates were rods producing fast-growing, round to irregular colonies with raised elevations and smooth surfaces. Reddish pigmentation was produced by PPB5, while no pigmentation was produced by other isolates (Supplementary Table S1). All 10 isolates were motile and reacted positively to the Gram staining, citrate, catalase and oxidase tests, but reacted negatively to the KOH solubility test (**Table 1**).

Phylogenetic trees constructed from 16S rRNA sequences showed that the selected isolates were mainly members of genus *Bacillus, Pseudomonas,* and *Stenotrophomonas* (**Supplementary Figure S1**). The sequences of the isolates PPB2, PPB5, PPB8, PPB9, and PPB11 showed 99% similarity with *Bacillus subtilis* and were submitted to GenBank under accession numbers KJ690255, KM008605, KM008606, KM092525 and KM092527, respectively (**Table 1**). Isolate PPB1 had 99% homology with *Pseudomonas stutzeri* and was submitted to GenBank under accession number KJ959616. Isolate PPB3 was identified as *Stenotrophomonas maltophilia* with GenBank accession number KJ959617. Isolates PPB4, PPB10, and PPB12 showed 99% sequence homology with

*B. amyloliquefaciens* and were submitted to GenBank under accession number KM008604, KM092526 and KM092528, respectively (**Table 1**).

### Characterization for Plant Growth Promoting Traits

The plant growth promoting characteristics viz., IAA production and ARA were examined with the ten selected PGPR isolates. The results of the assays are presented in **Table 2**. In the presence of tryptophan, the isolated bacteria produced IAA in concentrations between 26.78 μg mL−<sup>1</sup> and 51.28 μg mL−1. The highest and lowest amounts of IAA were produced by strain PPB5 and PPB3, respectively (**Table 2**). Nitrogenese activity, as determined by ARA, was not detected in PPB1 and PPB12 under the conditions tested. However, the ARA values ranged from 1.79 to 4.9 μmole C2H4/mg protein/h for remaining isolates. PPB2 showed the highest activity, while the lowest was recorded for PPB11 (**Table 1**). The other isolates also reduced acetylene in significant amounts. Collectively, these results suggest that the isolates possess a number of traits associated with plant growth promotion.

### Germination and Vigour Index Improvement in Cucumber

The effect of rhizobacterial treatment upon seed germination and vigour index of cucumber varied with different isolates. All treatments had a significant effect on the germination rate and vigour index compared to the control. The PGPR treatments increased the germination rate of cucumber seeds by 8.07–15.32% compared with the control, while the vigour index was increased by 98.62–148.05% (**Figure 1**). In both germination rate and vigour index, the maximum increase was obtained with the PPB9 treatment. These results suggest that rhizobacterial treatment could improve the germination and vigour of cucumber seeds.

### Plant Growth Promotion in Cucumber

All isolates significantly increased the growth of cucumber compared to non-inoculated controls. Treatment with isolate PPB12 produced the maximum shoot and root lengths of 18.23 and 20.47 cm, corresponding to increases of 66.02 and 65.63% above control treatments (**Figure 1**). However, the maximum shoot and root weight enhancement was observed in PPB8 treated plants. Treatment with isolate PPB8 produced shoot fresh and dry weights of 5.29 and 0.60 g plant<sup>−</sup>1, which were 79.32 and 100.00% higher than those of control plants. Similarly, treatment with isolate PPB8 produced root fresh and dry weights of 3.03 and 0.32 g plant<sup>−</sup>1, corresponding to increase of 91.77 and 128.57% above control treatments.

### N Concentration in Cucumber Plants

The N content in plant roots and shoots significantly increased due to inoculation treatments with rhizobacterial isolates (**Figure 2**). The shoot and root N content showed similar trends in response to different treatments; hence, the N content is reported as the total combined shoot and root N. The total N content in PGPR-treated plants ranged from 3.66 mg g−<sup>1</sup> to


*'*−*' corresponds to negative response; '*+*ve' and '*+*' correspond to positive responses.* <sup>∗</sup>*Values are the Mean* ± *SE. The experiment was repeated twice with three replicates for each isolate.*

TABLE 2 | Comparative performance of PGPR in mycelia growth inhibition of *P. capsici* and Phytophthora crown rot disease suppression in cucumber plants.


<sup>∗</sup>*Values are the means* <sup>±</sup> *SE (n* <sup>=</sup> *12). Data within the same column followed by different letters are significantly different.* <sup>a</sup>*Pathogen suppression was measured as percent inhibition of radial growth of P. capsici by antagonistic PGPR in dual plate assay.*

<sup>b</sup>*Disease suppression was expressed as percent protection due to treatment with PGPR relative to control (non-inoculated).*

<sup>c</sup>*Protection (%)* <sup>=</sup> *[(A* <sup>−</sup> *B)/A]* <sup>×</sup> *100 in which, A* <sup>=</sup> *PDI in non-inoculated control plants and B* <sup>=</sup> *PDI in PGPR-treated plants.* <sup>d</sup>*Seven-day-old seedlings were inoculated with 500 or 1000* <sup>μ</sup>*l zoospore suspension/pot.*

*The data presented are from representative experiments that were repeated twice with similar results.*

8.25 mg g−<sup>1</sup> N compared with 2.57 mg g−<sup>1</sup> N for non-inoculated control plants, a 42–221% increase in PGPR-treated plants over control plants (**Figure 2**). The highest N content was recorded in plants grown under PPB2 followed by PPB8, PPB3, PPB9 and other treatments.

### Root Colonization

The ability to colonize the root system is essential for rhizobacteria to be effective plant growth promoters. The root colonization assays showed that all the tested isolates successfully colonized the roots of cucumber plants as tested after only 7 days of seedling growth. The inoculated populations were even higher on 21-day-old roots. Nevertheless, the root population densities varied widely among the isolates (**Figure 3**). The largest root populations were observed for strain PPB2, followed by PPB5 and PPB9 (**Figure 3**). These results demonstrate specific interactions between cucumber plants and the rhizobacterial isolates.

## *In vitro* Antagonism of *Phytophthora capsici*

All rhizobacterial isolates exhibited significant antagonistic activity against *P. capcisi* on PDA. The largest inhibition zone was observed with PPB9 (90.08%) followed by PPB8 (82.05%) (**Table 2**). Distinct morphological alterations in *P. capcisi* hyphae were also detected during dual cultures with the rhizobacterial isolates. Hyphal features observed in the vicinity of bacterial colonies included irregular and excessive branching, abnormal swelling of hyphal diameters, unusually long and pointed hyphal tips, and vacuolization leading to hyphal lysis (**Figure 4**).

FIGURE 1 | Effect of plant growth promoting rhizobacteria (PGPR) treatments on seed germination, vigour and growth characteristics of cucumber seedlings grown in pots under axenic conditions. Error bars are SE from three replicates per same treatment. Data are presented as % increase in germination, vigour index, shoot length, root length, shoot fresh weight, root fresh weight, shoot dry weight, and root dry weight of PGPR-treated cucumber seedlings relative to non-treated control seedlings. The experiment was repeated twice.

FIGURE 2 | Effect of inoculation with PGPR strains on shoot and root N contents of cucumber plants. Error bars are SE from three replicates that received the same treatment. Data represents total shoot and root N concentration (mg g−1), each from 3 sets of 8–10 shoots and roots sampled following harvesting the cucumber plants. Within each frame different letters indicate statistically significant difference between treatments (LSD test, *P* ≤ *0.05*). The experiment was repeated twice.

### Suppression of Phytophthora Crown Rot in Cucumber

All the selected PGPR strains showed consistent suppression of Phytophthora crown rot in the greenhouse experiments. Compared with the control, the average disease protection at 500 μl zoospore suspension ranged from 50 to 88.83% after treatment with rhizobacterial isolates, while protection at 1000 μl zoospore suspension ranged from

FIGURE 3 | Population density (cfu) of different PGPR strains from roots of 7-, 14-, and 21-day-old cucumber seedlings. Error bars are SE from three replicates per treatment. Data are presented as numbers of c.f.u. g−<sup>1</sup> fresh weight, each from three sets of 5–8 whole roots. The data presented are from representative experiments that were repeated twice with similar results.

33.33 to 86.08% (**Table 2**). At both inoculum rates, isolate PPB11 showed the highest disease reduction, and the lowest disease reduction was obtained with PPB5.

### DISCUSSIONS

PGPR colonizing the surface or inner part of roots play important beneficial roles that directly or indirectly influence plant growth and development (Glick et al., 1999; Gerhardt et al., 2009). In this study, 10 PGPR classified as *Pseudomonas stutzeri* (PPB1), *B. subtilis* (PPB2, 5, 8, 9, and 11), *S. maltophilia* (PPB3), and *B. amyloliquefaciens* (PPB4, 10, and 12) were isolated from the rhizosphere of cucumber plants. All the isolated PGPR were gram positive and motile rods, and tested positive for the ability to utilize citrate as a carbon source. Flagellar motility and citrate utilization are both thought to play a significant role in competitive root colonization and maintenance of bacteria in roots (Turnbull et al., 2001; Weisskopf et al., 2011). These strains also tested positive for oxidase and catalase activity. Standard microbiology references suggest that *S. maltophilia* is an oxidase-negative bacterium (Ryan et al., 2009). Recent data, however, suggest that some *S. maltophilia* are oxidase-positive (Carmody et al., 2011), and this was also the case for isolate PPB3 in this study. Our catalase test results corroborate prior studies showing that *B. subtilis*, *Pseudomonas* stutzeri, and *B. amyloliquefaciens* are catalase-positive (Merchant and Packer, 1999; Kamboh et al., 2009). *Bacillus* and *Pseudomonas* are the most frequently reported genera of PGPR (Laguerre et al., 1994; Hallmann and Berg, 2006; Zahid et al., 2015). Similarly, most isolates in this study belong to genus *Bacillus*.

Treatment of cucumber seeds with the selected isolates significantly improved seedling emergence and growth. Several different mechanisms have been suggested for similar observations using other PGPR strains: PGPR might indirectly enhance seed germination and vigour index by reducing the incidence of seed mycoflora, which can be detrimental to plant growth (Begum et al., 2003). Duarah et al. (2011) found that amylase activity during germination was increased in rice and legume after inoculation with PGPR. The amylase hydrolyzes the starch into metabolizable sugars, which provide the energy for growth of roots and shoots in germinating seedlings (Beck and Ziegler, 1989; Akazawa and Nishimura, 2011). One of the most commonly reported mechanisms is the production of phytohormones such as IAA (Patten and Glick, 2002). All the selected isolates in this study produced IAA. Similar studies have shown that IAA production is very common among PGPR (Yasmin et al., 2004; Ng et al., 2012; Zahid et al., 2015). In fact, many isolates in this study produced higher IAA than previously reported strains (Yasmin et al., 2004; Banerjee et al., 2010; Ng et al., 2012). This is an important mechanism of plant growth promotion because IAA promotes root development and uptake of nutrients (Carrillo et al., 2002). It has long been proposed that phytohormones act to coordinate demand and acquisition of nitrogen (Kiba et al., 2011). Therefore, enhanced N-content found in inoculated plants could be due to increased N-uptake by the roots caused by hormonal effects on root morphology and activity. Nitrogen fixation may also play a role in plant growth promotion. All the selected isolates in this study except PPB1 and PPB12 showed acetylene reduction activity, which is a widely accepted surrogate for nitrogenase activity and N2-fixing potential (Andrade et al., 1997). However, defensible proof of N2-fixation needs the application of 15N as tracer of soil N or as 15N2-gas and the

demonstration of significantly changed N-isotope-labeling in the plant biomass. Transfer of N between diazotrophic N-fixing rhizobacteria and the roots of several crops has been demonstrated (Islam et al., 2009; Abbasi et al., 2011; Tajini et al., 2012; Verma et al., 2013). It is interesting to note that in this study all isolates, including the two that demonstrated no acetylene reduction activity, enhanced the N content of cucumber. This suggests that while N2 fixation may be an important mechanism of plant growth promotion, there may also be alternate mechanisms, like hormonal interactions and nutrient uptake or pathogen suppression, which might be more pronounced than the contribution of nitrogen fixation.

Results from our study indicate that PGPR strains applied as a seed treatment significantly reduced disease severity of Phytophthora crown rot on cucumber plants. The fungal antagonists *Pseudomonas stutzeri*, *B. subtilis*, *B. amyloliquifaciens,* and *S. maltophilia* were have been shown to be effective biocontrol agents in prior studies (Dunne et al., 2000; Zhang and Yuen, 2000; Dal Bello et al., 2002; Berg et al., 2005; Islam and Hossain, 2013; Erlacher et al., 2014). Competitive root tip colonization by PGPR strains might play an important role in the efficient control of soil-borne diseases. The crucial colonization level that must be reached has been estimated at 105–106 cfu g−<sup>1</sup> of root in the case of *Pseudomonas* sp., which protect plants from *Gaeumannomyces tritici* or *Pythium* sp. (Haas and Défago, 2005). Most of our selected strains were efficient colonizers of roots, since CFU counts for tested strains were more than 10<sup>7</sup> cfu g−<sup>1</sup> root. However, the former study examined the root colonization by introduced bacteria under natural field soil conditions, while our study did under axenic conditions. In view of that, comparison between root colonization data obtained under these two conditions may not be accurate. Biological control of *P. capsici* can also result from antibiosis by the bacteria (Nakayama et al., 1999; Kawulka et al., 2004; Chung et al., 2008; Lim and Kim, 2010; Mousivand et al., 2012; Islam and Hossain, 2013). All the selected isolates exhibited moderate to high antagonistic activity against *P. capsici in vitro*, and caused clear morphological distortions such as abnormal branching, curling, swelling and lysis of the hyphae at the interaction zone in dual cultures. Excessive branching and curling accompanied by marked ultrastructural alterations including invagination of the hyphal membrane, disintegration or necrosis of hyphal cell walls, accumulation of excessive lipid bodies, enlarged and electron-dense vacuoles, and degradation of cytoplasm were potentially due to bacterial production of antibiotics and lytic enzymes (Deora et al., 2005; Islam and von Tiedemann, 2011). These antibiotics and lytic enzymes cause membrane damage and are particularly inhibitory to zoospores of Oomycete (de Souza et al., 2003; Beneduzi et al., 2012). Induced systemic resistance is most likely another mechanism by which bacteria suppress *P. capsici* (Zhang et al., 2010).

In the present study, we have isolated 10 new strains of PGPR from indigenous cucumber plants. These strains possessed several plant growth promoting traits as well as antifungal activity, and were found to be efficient in controlling Phytophthora crown rot of cucumber seedlings. *In vitro* and *in vivo* evidence suggest that the selected isolates benefit cucumber plants via multiple modes of action including antibiosis against phytopathogens, competitive colonization, and plant growth promotion. This reveals the potential of these strains for biofertilizer applications and commercial use as biocontrol agents in the field. However, from the estimation of a PGPR-potential to a biofertilizer application, it requires a long way of greenhouse experiments with pot filled with different type of soils and finally, field experiments to find out the optimum formulations for the inoculums. Thus, the inoculants can perform close to its optimum potential. Future studies concerning commercialization and field applications of integrated stable bio-formulations as effective biocontrol strategies are in progress.

### AUTHOR CONTRIBUTION

SI was involved in the planning and execution of the research work; collection, analysis and interpretation of the data; manuscript writing etc. following the suggestions and directions of the Major Professor. AMA served as the Member of the Dissertation Committee of SI and was involved in the planning of the work and editing of the manuscript. AP was actively involved in the original research work, data collection, analysis as well

### REFERENCES


as manuscript preparation. TI supplied the *Phytophthora capsici* inocula and oversaw the sequence work of the bacterial isolates and related description in the manuscript. MMH served as the Major Professor of SI and was involved in the research design and planning; analysis and interpretation of data; drafting as well as critical revision of the work for intellectual content.

All authors approve the final version of the manuscript for publication and agrees to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

### ACKNOWLEDGMENTS

The authors would like to acknowledge the financial assistance from University Grant Commissions through Research Management Committee of Bangabandhu Sheikh Mujibur Rahman Agricultural University, Bangladesh.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fmicb. 2015.01360

FIGURE S1 | Phylogenetic tree of 16S rRNA gene sequences showing the relationships among the isolates isolated from cucumber rhizosphere. The data of type strains of related species were from GenBank database.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2016 Islam, Akanda, Prova, Islam and Hossain. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# MATI, a Novel Protein Involved in the Regulation of Herbivore-Associated Signaling Pathways

M. Estrella Santamaría1,2, Manuel Martinez<sup>1</sup> , Ana Arnaiz<sup>1</sup> , Félix Ortego<sup>3</sup> , Vojislava Grbic<sup>2</sup> and Isabel Diaz<sup>1</sup> \*

<sup>1</sup> Centro de Biotecnología y Genómica de Plantas, Universidad Politécnica de Madrid – Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria, Madrid, Spain, <sup>2</sup> Department of Biology, The University of Western Ontario, London, ON, Canada, <sup>3</sup> Departamento de Biología Medioambiental, Centro de Investigaciones Biologicas, Consejo Superior de Investigaciones Científicas, Madrid, Spain

The defense response of the plants against herbivores relies on a complex network of interconnected signaling pathways. In this work, we characterized a new key player in the response of Arabidopsis against the two-spotted spider mite Tetranychus urticae, the MATI (Mite Attack Triggered Immunity) gene. This gene was differentially induced in resistant Bla-2 strain relative to susceptible Kon Arabidopsis accessions after mite attack, suggesting a potential role in the control of spider mites. To study the MATI gene function, it has been performed a deep molecular characterization of the gene combined with feeding bioassays using modified Arabidopsis lines and phytophagous arthropods. The MATI gene belongs to a new gene family that had not been previously characterized. Biotic assays showed that it confers a high tolerance not only to T. urticae, but also to the chewing lepidopteran Spodoptera exigua. Biochemical analyses suggest that MATI encodes a protein involved in the accumulation of reducing agents upon herbivore attack to control plant redox homeostasis avoiding oxidative damage and cell death. Besides, molecular analyses demonstrated that MATI is involved in the modulation of different hormonal signaling pathways, affecting the expression of genes involved in biosynthesis and signaling of the jasmonic acid and salicylic acid hormones. The fact that MATI is also involved in defense through the modulation of the levels of photosynthetic pigments highlights the potential of MATI proteins to be exploited as biotechnological tools for pest control.

Keywords: plant–herbivore interaction, Tetranychus urticae, Arabidopsis thaliana, Spodoptera exigua, hormonal signaling pathways, plant redox status

## INTRODUCTION

Plants are sessile organisms that rely on a battery of mechanisms to detect pathogens and pests in order to mount appropriate defense responses. Particularly, plants have evolved constitutive and inducible defenses to deter phytophagous arthropods, as well as indirect defenses, using volatiles and nectars to attract natural enemies of phytophagous insects and acari (Wu and Baldwin, 2010; Santamaria et al., 2013). As part of these defenses, herbivore-challenged plants can also emit volatiles to warn neighboring plants of an imminent threat (Muroi et al., 2011; Schuman and Baldwin, 2016). In parallel, herbivores respond to plant defenses by developing multiple strategies

#### Edited by:

Gero Benckiser, Justus-Liebig-Universität Gießen, Germany

### Reviewed by:

Andrea Chini, CNB (Centro Nacional de Biotecnología), Consejo Superior de Investigaciones Científicas, Spain Raquel Campos-Herrera, University of Algarve, Portugal

> \*Correspondence: Isabel Diaz i.diaz@upm.es

#### Specialty section:

This article was submitted to Plant Microbe Interactions, a section of the journal Frontiers in Plant Science

Received: 20 January 2017 Accepted: 23 May 2017 Published: 09 June 2017

#### Citation:

Santamaría ME, Martinez M, Arnaiz A, Ortego F, Grbic V and Diaz I (2017) MATI, a Novel Protein Involved in the Regulation of Herbivore-Associated Signaling Pathways. Front. Plant Sci. 8:975. doi: 10.3389/fpls.2017.00975

to avoid them (Elzinga and Jander, 2013; Suchan and Alvarez, 2015), and plants counterattack to implement emergency responses (Hao et al., 2008; Schmelz et al., 2012).

The induction of plant defenses is initiated when specific receptors recognize either the presence of herbivore (through the recognition of Herbivore-Associated Molecular Patterns; HAMPs), the damage incurred by plant tissues as a consequence of herbivore feeding (Damage-Associated Molecular Pattern; DAMPs) or the presence of volatiles emitted as plant–plant cues (Frost et al., 2007; Schuman and Baldwin, 2016). Recognition of these molecular patterns triggers transduction pathways that activate the expression of defense genes. Physiologically, early events in plant–herbivore interactions start with membrane potential depolarization at the feeding site, alteration in cell membrane and ion imbalance followed by changes in the intracellular Ca2<sup>+</sup> and generation of reactive oxygen species (ROS) (Fürstenberg-Hägg et al., 2013). A cascade of protein kinases (CDPKs) as calcium-sensor proteins lead to the synthesis of phytohormones and activation of transcription factors that regulate the gene expression of a wide range of species-specific compounds with anti-nutritional, deterrent, repellent and toxic properties that function to entrap, inhibit, block or modify metabolism, development and fecundity of phytophagous arthropods (Wu and Baldwin, 2010; War et al., 2012; Santamaria et al., 2013). Host transcriptomic and proteomic profiles after arthropod feeding, oviposition or application of insect secretions have demonstrated that plants may discriminate between herbivores and activate specific plant responses (Giri et al., 2006; Kempema et al., 2007; Sotelo et al., 2014; Bandoly et al., 2015). Metabolomic approaches have corroborated plant's ability to differentiate herbivore species and determine the onset of indirect defense responses to complement the direct defenses (Hettenhausen et al., 2013). Although the understanding of plant–arthropod interactions is still rudimentary, it is known that plant defenses locally or systemically induced by herbivores are regulated by a complex hormonal cross-talk (Erb et al., 2012; Wang and Wu, 2013; Gimenez-Ibanez et al., 2016). The central phytohormones that mediate between signal recognition and activation of defenses are Jasmonic Acid (JA), Salicylic Acid (SA), and Ethylene (Et). In general, JA regulates the induced defenses against chewing insects (Schmiesing et al., 2016), mesophyll sucking mites (Zhurov et al., 2014; Alba et al., 2015; Martel et al., 2015) and necrotrophic pathogens (Goossens et al., 2016), as well as mechanical wounding (Pieterse et al., 2009). SA-regulated responses are induced by phloem-feeding insects (Kawazu et al., 2012; Thaler et al., 2012), mesophyll sucking mites (Kant et al., 2004) and biotrophic pathogens (Gimenez-Ibanez et al., 2016), while Et most probably modulates the action of both hormones (Campos et al., 2014). Herbivory also leads to oxidative stress, changes in intracellular pH and desiccation, which modulate the JA pathway either directly or indirectly through the action of other hormones such as abscisic acid (ABA) (Hillwig et al., 2016). Other findings have also suggested important roles for auxins, cytokinins, and brassinosteroids (Diezel et al., 2009; Erb et al., 2012; Mur et al., 2013). However, a better understanding of phytohormone crosstalk transforming the initial perception events into appropriate responses is needed.

Phytophagous mites pierce parenchymatic plant cells using stylets to suck their nutrients, and cause severe chlorosis leading to a reduction in crop yield (Park and Lee, 2002; Farouk and Osman, 2011; Bensoussan et al., 2016). Among phytophagous mites, the two-spotted spider mite, Tetranychus urticae, is a species that feeds on more than 1,100 host plants including a wide range of ornamentals, greenhouse crops and annual and perennial field cultivars (Migeon and Dorkeld, 2006–2015). T. urticae is a model within chelicerate herbivores with its genome sequenced and with a broad range of tools and protocols developed (Grbic et al., 2011; Santamaria et al., 2012b; Cazaux et al., 2014). Mite ability to feed on Arabidopsis thaliana and the wide available toolkits for this plant species have provided an outstanding opportunity for functional studies of plant–mite interaction (Santamaria et al., 2012a, 2015; Zhurov et al., 2014). In addition, due to the spider mite impact in agriculture, research on plant responses to mite infestation has been also performed on crops (Maserti et al., 2011; Agut et al., 2015; Diaz-Riquelme et al., 2016). In tomato, the defense effect of glandular trichomes enriched in acylsugars or terpenoids against spider mites has been highlighted (Resende et al., 2002; Alba et al., 2009; Bleeker et al., 2012) as well as the repellence and oviposition deterrence induced by leaf extracts with high concentrations of methyl ketones (Antonious and Snyder, 2015). Tomato defenses against T. urticae also include the emission of different volatile compounds associated with the attraction of the spider mite predator Phytoseiulus persimilis (Ament et al., 2004). JA is essential but not the unique hormone for establishing the spider mite-induced defense responses (Kant et al., 2004; Ament et al., 2006; Zhurov et al., 2014; Martel et al., 2015). Spider mites feeding also triggers SA plant defense pathways (Kant et al., 2004; Zhurov et al., 2014). As a consequence, plants probably use this JA/SA balance to fine tune defense responses. However, spider mites may manipulate plant defenses downstream of hormonal crosstalk via an unknown mechanism generating considerable ecological costs (Kant et al., 2008; Glas et al., 2014; Alba et al., 2015; Wybouw et al., 2015). Recently, Villarroel et al. (2016) have identified effector/elicitor-like proteins in the mite saliva able to suppress plant defenses downstream of SA in tobacco and, as consequence, to promote mite's reproductive performance. This salivary protein expression is dependent of the host plant (Jonckheere et al., 2016).

In this work, it is characterized the Arabidopsis AT3G14395 gene of unknown function, termed Mite Attack Triggered Immunity (MATI). The MATI gene was selected for further characterization because it showed consistently higher expression levels in the resistant Bla-2 strain relative to susceptible Kon Arabidopsis accessions after mite feeding. This previous result prompt us to study its potential defense properties as defense gene to control phytophagous arthropods. Here, we demonstrate that this gene encodes a novel small protein that contributes to confer plant protection against herbivores through modulation of plant redox status and hormonal signaling pathways.

### MATERIALS AND METHODS

### Plant Material and Growth Conditions

Arabidopsis thaliana Col-0, Kondara (Kon), and Bla-2 (Bla-2) accessions (Nottingham Arabidopsis Seed Collection) were used as wild-types (WT). A. thaliana T-DNA mutants (SALK\_092139C, N672907, referred as mati in this article) were obtained from the Arabidopsis Biological Resource Centre (ABRC<sup>1</sup> ) through the European Arabidopsis Stock Centre (NASC<sup>2</sup> ). T-DNA insertion and homozygous status and gene expression levels of the Salk lines were validated by conventional PCR and quantitative RT-qPCR assays (described below) (Supplementary Figure 1A). For growth in the soil, seeds were planted and incubated 5 days at 4◦C. For in vitro growth, seeds were surface-sterilized with 75% (V/V) ethanol, dried, and plated onto Petri dishes containing 0.53 g of Murashige and Skoog salts (Sigma–Aldrich), 1% (W/V) sucrose, 0.5 g/L MES, and 0.4% (W/V) Phytagel (Sigma–Aldrich), adjusted to pH 5.7 with KOH. Plants and plates were then grown in growth chambers (Sanyo MLR-350-H) under control conditions (23◦C ± 1 ◦C, >70% relative humidity and a 16 h/8 h day/night photoperiod).

To generate overexpression lines, MATI cDNA from Col-0 plants was cloned into pGWB2 (CaMV35S, no tag) and pGWB5 (CaMV35S, C-sGFP) Gateway binary vectors (Nakagawa et al., 2007) using the specific primers included in Supplementary Table 1. The recombinant plasmids were introduced into A. thaliana Col-0 plants using Agrobacterium floral dip transformation (Clough and Bent, 1998), Col-MATI and Col-MATI-GFP plants in this article. pGWB2 plasmids were also introduced into A. thaliana Kon plants, termed Kon-MATI plants in this article. Shoots were regenerated on selective medium containing hygromycin (100 mg/L), and primary transformants (T0) were allowed to self-fertilize. Plants were then selected and self-fertilized twice more to generate the third generation lines (T3). Homozygous plants with one single copy insertion and the highest transgene expression levels coming from different transformation events were selected for our experiments (Supplementary Figure 1A).

### Gene Expression Analyses by Real Time PCR (RT-qPCR)

RT-qPCR assays have been used for different purposes: (i) to validate data from transcriptomic analysis; (ii) to determine the homozygous status of the Salk mutant lines; (iii) to study MATI gene expression in major Arabidopsis tissues. A. thaliana rosettes from Col-0, Bla-2, and Kon accessions were sampled after different time of mite infestation (1, 3, 6, 12, and 24 h) to validate microarray results. Total RNA was extracted following Oñate-Sánchez and Vicente-Carbajosa (2008) and reverse transcribed using Revert AidTM H Minus First Strand cDNA Synthesis Kit (Fermentas). cDNAs from A. thaliana Col-0 flowers, roots, siliques, leaves from stem rosettes at 1- to 3-week-old were also prepared. RT-qPCR was performed for three samples coming from three independent experiments as previously described (Santamaria et al., 2012a) using a SYBR Green Detection System (Roche) and the CFX Manager Software 2.0 (Bio-Rad). Gene expression was referred as relative expression levels (2−1Ct) or fold change (2−11Ct) (Livak and Schmittgen, 2001). Similarly, gene expression levels involved in hormonal signaling pathways were analyzed in rosettes from Col-MATI\_4.1, -mati and WT plants in three samples coming from three independent experiments of non-infested 1 and 24 h infested plants (20 mites/plant). mRNA quantification was expressed as relative expression levels (2−1Ct) normalized to ubiquitin (Livak and Schmittgen, 2001). Specific primers were designed through the Salk Institute T-DNA primer design link<sup>3</sup> or through the PRIMER 3 program<sup>4</sup> . Primer sequences are indicated in Supplementary Table 1.

### Structural and Evolutionary MATI Gene Analyses

Mite Attack Triggered Immunity sequences were downloaded from the TAIR website<sup>5</sup> . Amino acid sequence was subjected to a sequence search in the Pfam database v28.0<sup>6</sup> and SignalP 4.1 program<sup>7</sup> to identify possible domains and a signal peptide within the protein, respectively. MATI proteins from different plant species were compiled using GreenphylDB v4<sup>8</sup> and aligned by MUSCLE v3.8<sup>9</sup> . A phylogenetic tree was constructed by the maximum likelihood PhyML v3.0 method<sup>10</sup>, using a BIONJ starting tree and applying the approximate likelihood-ratio test (aLRT) as statistical test for non-parametric branch support. Displayed trees were visualized in the program MEGA 6.0<sup>11</sup> . Conserved sites were analyzed by drawing sequence logos representing profile hidden Markov models using the Skytign tool<sup>12</sup>. To analyze MATI gene product divergences among Arabidopsis accessions, its cDNA from Col-0, Bla-2, and Kon rosettes was cloned and sequenced using pGEM <sup>R</sup> -T Easy Vector Systems. The three sequences were aligned using Clustal W 1.83<sup>13</sup> .

### Subcellular Localization of MATI-GFP in Onion Epidermal Cells and A. thaliana Overexpression Lines

Mite Attack Triggered Immunity protein location within cell compartments were analyzed by two experimental approaches, using MATI gene fusions to the green fluorescent protein (GFP). First, transient expression assays were performed by particle bombardment of onion epidermal layers and second, stable transformation of Arabidopsis plants. The open reading frame

<sup>9</sup>http://www.ebi.ac.uk/Tools/msa/muscle

<sup>1</sup>https://abrc.osu.edu/

<sup>2</sup>http://arabidopsis.info/

<sup>3</sup>http://signal.salk.edu/tdnaprimers.2.html

<sup>4</sup>http://bioinfo.ut.ee/primer3-0.4.0/

<sup>5</sup>http://www.arabidopsis.org/

<sup>6</sup>http://pfam.xfam.org/

<sup>7</sup>http://www.cbs.dtu.dk/Services/SignalP/

<sup>8</sup>http://www.greenphyl.org

<sup>10</sup>http://www.phylogeny.fr

<sup>11</sup>http://www.megasoftware.net/

<sup>12</sup>http://skylign.org/

<sup>13</sup>http://www.ch.embnet.org/software/ClustalW.html

(ORF) of the MATI gene translationally fused to the N-terminus of the whole reporter gene was cloned into the pGWB5 binary vector following Gateway technology instructions. As controls, the psmRS-GFP plasmid containing the cauliflower mosaic virus 35S promoter (Davis and Vierstra, 1998) and the pRTL21NS/ss-RFP-HDEL plasmid containing the Arabidopsis chitinase signal sequence and the C-terminal HDEL ER retrieval signal, whose protein specifically localizes in the endoplasmic reticulum (ER) (Shockey et al., 2006), were used. Transient transformation of onion (Allium cepa) epidermal cells was performed by particle co-bombardment with a biolistic helium gun device (DuPont PDS-1000/Bio-Rad) as previously described (Diaz et al., 2005). Fluorescent images were acquired after 24 h of incubation at 22◦C in the dark, under the LEICA SP8 confocal microscope (Leica). For the subcellular localization of the MATI-GFP fusion protein in A. thaliana, seedlings from Col-MATI\_4.1-GFP line were observed under the LEICA SP8 confocal microscope (Leica) using the appropriated filters.

### Spider Mite and Beet Armyworm Maintenance and Fitness Analyses

To test the involvement of MATI in plant responses to herbivory, T. urticae and Spodoptera exigua feeding bioassays were performed in MATI overexpressing, mati-mutant ans Col-0 Arabidopsis lines. A colony of T. urticae, London strain (Acari: Tetranychidae) provided by Dr. Miodrag Grbic (UWO, Canada), was reared on beans (Phaseolus vulgaris) and maintained in growth chambers (Sanyo MLR-350-H) at 23◦C ± 1 ◦C, >70% relative humidity and a 16 h/8 h day/night photoperiod. Mites were synchronized by inoculating 100 adult females (random age) on one leaf of bean confined in a closed system under water-soaked cotton. After 1 day, adult females were removed, and 10 days after, population on the leaf was synchronized. Spider mite development and behavior was studied on Col-MATI\_4.1, \_5.2, mati and Col-0 WT plants. T. urticae fecundity assay was performed on detached leaves from 3-week-old plants. The newest emerged leaf (about 1 cm long) from each plant was fit in special dishes and infested with 12 adult synchronized females. After 36 h of infestation, the number of eggs was counted. Eight replicates were used for plant genotype.

Spodoptera exigua eggs were kindly provided by Drs. F. Budia and E. Viñuela (Department of Producción Agraria, ETSI Agronomos-UPM, Spain). Larvae were maintained in boxes feeding on artificial diet (Poitout and Bues, 1970), in growth chambers (Sanyo MLR-350-H) at 25◦C ± 1 ◦C, > 70% relative humidity and a 16 h/8 h day/night photoperiod. The boxes were lined with filter paper to reduce humidity. Vermiculite was provided for pupation when larvae complete its development cycle. Adults were allowed to emerge in cylindrical containers supplied with 10% (V/V) honey solution in water. Eggs were deposited on strips of filter paper and neonate (24 h) and L1 larvae (48 h) were synchronized for the plant and herbivore behavior assays, respectively. Larval performance was tested by placing three freshly hatched (neonate) beet armyworm larvae were placed on 3-week-old plants from the different genotypes (Col-MATI\_4.1, -mati and WT lines) and larval weight was measured in a precision balance Mettler-Toledo MT5 (Mettler-Toledo) after 4 days of feeding. Twelve replicates were used for each genotype.

### Plant Damage Determination

Quantification of plant damage after arthropod feeding was done on Arabidopsis T2 entire plants from selected homozygous transgenic lines (Col-MATI\_4.1, Col-MATI\_5.2; Col-mati, Kon-MATI\_6.1 and Kon-MATI\_7.2) and from the non-transformed Col-0 and Kon controls. Three-week-old plants were infected with 20 T. urticae adults per plant. After 4 days of feeding, leaf damage was assessed by scanning the entire rosette using a hp scanjet (HP Scanjet 5590 Digital Flatbed Scanner series), according to Cazaux et al. (2014). Leaf damage was calculated in mm<sup>2</sup> , using Adobe Photoshop CS software. Six replicates were used for each genotype.

Leaf disks (7 mm diameter) of 3-week-old plants from the 3 Col-0 genotypes (WT, Col-MATI\_4.1 and Col-mati) were infected with two neonate larvae of S. exigua. After 12 h, leaf disks were scanned using the hp scanjet (HP Scanjet 5590 Digital Flatbed Scanner series). Leaf damage was calculated in mm<sup>2</sup> using split channel tool from imageJ software. Thirty two replicates were used for each genotype.

### Co-immunoprecipitation Assays

To identify MATI interaction partners, protein extracts from control and Col-MATI-GFP plants were immuno-purified using GFP-Trap system and analyzed in LTQ-Orbitrap Velos. Five gram of fresh weight of Col-GFP WT and Col-MATI-GFP (Col-MATI\_4.1 line) plants grown for 3 weeks on half-strength MS medium were frozen in liquid nitrogen, extracted by grinding with a mortar and pestle, and added to 15 mL of cold TAP extraction buffer (50 mM TRIS-HCl, 150 mM ClNa, 10% (V/V) glycerol, 0.8% (V/V) Triton and 0.4% (W/V) Chaps (Sigma) and complete protease inhibitor cocktail (Roche). This lysate was centrifuged at 10,000 rpm for 10 min at 4◦C and the protein content of the supernatant was quantified by Nanodrop <sup>R</sup> ND1000. Co-immunoprecipitation (Co-IP) experiments from Arabidopsis plants expressing GFP-tagged proteins were performed using GFP-Trap <sup>R</sup> beads (ChromoTek GmbH), following the recommendations of the manufacturer. The immune-precipitated proteins were detected by immunoblot analysis on nitrocellulose membranes (GE Healthcare) using a monoclonal anti-GFP antibody (Miltenyi Biotec) at 1:5000 dilution in BSA blocking buffer containing 3% (WV) PBS and 0.05% (V/V) Tween-20. Acrylamide gels were staining with Coomassie and Oriole (Bio-Rad) to identify putative interactions.

### Proteomic Analysis

Co-immunoprecipitated proteins from Col WT (Col-0) and Col-MATI\_4.1 samples were digested and analyzed in LTQ-Orbitrap Velos (Gil-Bona et al., 2015). Peptide identification from raw data was carried out using licensed version of search engine MASCOT 2.3.0 and Proteome Discoverer 1.4 (Thermo Scientific). Database search was

performed against UniProt-Swissprot and NCBInr database with taxonomic restriction to A. thaliana. The following constraints were used for the searches: tryptic cleavage after Arg and Lys, up to two missed cleavage sites allowed, and tolerances of 20 ppm for precursor ions and 0.6 Da for MS/MS fragment ions. Besides, searches were performed allowing optional Methionine oxidation and fixed carbamidomethylation of Cysteine. Search against decoy database (integrated decoy approach) was used to FDR calculate and MASCOT percolator filter was applied to MASCOT results. The acceptance criteria for protein identification were a FDR < 1% and at least one peptide identified with high confidence (CI > 95%). Proteins identified as putative MATI interactors were functionally categorized using AgriGO v1.2<sup>14</sup> and STRING v10<sup>15</sup>. In addition, the interaction among MATI putative partners was analyzed in STRING v10. The proteomic analysis was carried out at the Proteomics Facility UCM-PCM, a member of ProteoRed network.

### Electrolyte Leakage

Membrane depolarization is rapidly produced after plant– herbivore interaction followed by alterations in cell membrane and ion imbalance that can be monitored by measuring the electrolyte leakage (EL). Arabidopsis leaf disks (1 cm diameter) from Col-MATI\_4.1, -mati and WT plants were infested with 10 mites and incubated for 24 h. The EL was determined as described by Masood et al. (2006). EL measurements were performed after 2 h (C1) of incubation at 32◦C using a conductometer (EC-Metro BASIC 30, CRISON). Total electrolyte content was determined in the same way after boiling for 10 min (C2). Results were expressed as percentage of EL = (C1/C2) × 100. Six replicates were used for each genotype and treatment.

### Photosynthetic Pigment Measurements

To check whether plants reconfigure their primary metabolism during herbivore feeding to cope with the increased metabolic demands, chlorophyll and carotenoids were analyzed as standard parameters of photosynthetic activities in the three different Col-0 genotypes. Chlorophyll a and b, and total carotenoids were extracted from Col-MATI\_4.1, -mati, WT lines. Arabidopsis rosettes were infested with 10 mites and incubated for 24 h. One hundred milligram of leaves were ground in a mortar with liquid nitrogen and suspended in 10 ml of 80% (V/V) acetone, using photo-protected tubes. After centrifugation at 3000 rpm for 15 min (L344 Eppendorf 5810R centrifuge), the absorbance of the supernatant was measured at 470, 663, and 646 nm, for carotenoids, chlorophyll a and b, respectively, using a UV-vis spectrophotometer (UltroSpecTM 3300pro, Amersham Bioscience). Six replicates were used for each genotype and treatment. Pigment content was calculated using the extinction coefficients and equations determined by Lichtenthaler (1987).

### ROS Production and DAB Staining Quantification

The accumulation of H2O<sup>2</sup> was determined using the 3,3-diaminobenzidine tetrachloridehydrate (DAB) substrate (Sigma–Aldrich) which produces a brown precipitate after oxidation in the presence of H2O<sup>2</sup> (Martinez de Ilarduya et al., 2003). Col-0 Arabidopsis leaf disks (1 cm diameter) from Col-MATI\_4.1, -mati and WT genotypes were infested with 10 mites and incubated for 24 h. Infested and non-infested control disks were stained with DAB following Rodríguez-Herva et al. (2012) and observed under a Zeiss Axiophot microscope. DAB staining specificity was confirmed in presence of the H2O<sup>2</sup> scavenger, ascorbic acid (10 mM). ImageJ was used for the image quantification analysis<sup>16</sup> using Methyl Green DAB vector. Six replicates were done by genotype and treatment.

### Thiol Quantification

The abundance of thiol groups was quantified in Col-MATI\_4.1, -mati and WT a lines. Arabidopsis rosettes were infested with 10 mites and incubated for 24 h. Two hundred milligram of 3-week-old rosettes was homogenized with phosphate buffer (pH 6), centrifuged at 2500 rpm at 4◦C for 10 min. Supernatant was used for the fluorometric thiol group assay following manufacturer instructions (Sigma–Aldrich). Six replicates were used for each genotype and treatment.

### Hormonal Analyses

Since hormones are important regulatory components of defense signaling, and particularly JA and SA are known to play major roles in regulating plant defense responses against spider mites, the accumulation of JA and SA as well as ABA in infested and non-infested Col-0 genotypes was also measured. Plant hormones (OPDA: 12-oxo-phytodienoic acid, JA, JA-Ile, SA, and ABA) were quantified by isotopic dilution mass spectrometry from 3-week-old rosettes of Col-MATI\_4.1, -mati and WT plants after 24 h of spider mite feeding. Six rosettes were pooled per experiment and three independent experiments were performed. Isotope-labeled standards were added to plant samples (∼0.1 g) before extraction as described previously (Durgbanshi et al., 2005). Ultra-performance liquid chromatography (UPLC)-electrospray ionization-tandem mass spectrometry analyses were carried out on an Acquity SDS system (Waters) coupled to a triple quadrupole mass spectrometer (Micromass). Quantification was accomplished with an external calibration.

### Statistical Analysis

Statistical analyses were performed using one-way ANOVA for gene expression validation, expression in different tissues and damage analysis and T. urticae and S. exigua bioassays. Two-way ANOVA was used for gene expression validation, EL, H2O2, antioxidants, photosynthesis, metabolites and hormone gene expression studies. Student-Newman-Keuls multiple

<sup>14</sup>http://bioinfo.cau.edu.cn/agriGO/

<sup>15</sup>http://string-db.org/

<sup>16</sup>https://imagej.nih.gov/ij/

comparison test was applied to all the studies. In figures, significant differences (P < 0.05) among lines for different evaluated parameters, were reported with different letters. The F-value, P-value, and df (degree of fredom) obtained from each statistical analysis are compiled in Supplementary Table 2.

### RESULTS

### MATI, a Gene Putatively Involved in Arabidopsis Defense against Spider Mites

Zhurov et al. (2014) highlighted the natural genetic variation of resistance between Arabidopsis accessions to T. urticae, identifying Bla-2 and Kon as the accessions at the opposing ends of the spectrum. Genome-wide analysis of transcriptional responses of these accessions demonstrated that they are very similar, which led to the conclusion that constitutive differences in levels of JA metabolites may underlay their differential resistance to mite feeding. Among otherwise similar responses, the AT3G14395 gene showed consistently higher expression levels in the resistant Bla-2 strain relative to Kon (**Figure 1A**). These data were validated by RT-qPCR assays (**Figure 1B**). The AT3G14395 gene, named MATI, encodes a specific plant protein of unknown function. The predicted protein, of 8.35 kDa, contains 75 amino acids with no known domains and no signal peptide. Searches in the GreenPhylDB discovered that homologs of the MATI protein were widely and uniquely present in angiosperms, including both monocots and dicots. A phylogenetic tree was constructed using protein sequences of MATI homologs from the 27 different species present in GreenPhylDB (**Figure 2A**). The phylogenetic analysis clearly identified three different groups, one formed by proteins from monocot species and two groups containing proteins from dicotyledonous plants (**Figure 2**). The MATI sequence was classified in the dicot group B together with other additional five protein sequences. Multiple sequence alignment revealed that MATI gene product was identical in Col-0, Bla-2, and Kon accessions (data not shown). A distant paralog, the AT1G30260 from A. thaliana was present in the dicot group A (**Figure 2A**), but was not up-regulated after spider mite feeding in Bla-2 (Supplementary Figure 2).

Multiple sequence alignments of MATI homologs revealed the presence of a conserved N-terminal domain found in all protein sequences. A threonine-isoleucine pair followed by several negatively charged amino acid residues at the C-end of the conserved domain was invariably detected in all proteins. In addition, two proline residues toward the middle of the domain and several charged amino acids scattered along the conserved N-terminal domain were also identified in most cases (**Figure 2B**). Thus, MATI is a gene of an unknown function, but conserved among both mono- and di-cotyledon plants, potentially associated with plant responses to spider mite feeding.

### MATI Gene Expression and MATI Protein Subcellular Location

To determine the expression of MATI gene, RT-qPCR studies were carried out in the major Arabidopsis tissues. MATI mRNA expression levels were mainly detected in flowers, seeds, siliques, and leaves from stem or rosette at different developmental stages, while it was scarcely detectable in roots (**Figure 3A**). Transient expression assays in onion epidermal cell layers were performed to determine MATI subcellular localization. The ORF of MATI translationally fused to the GFP was detected throughout the entire ER network, which was continuous with the nuclear membrane. This subcellular location was revealed by its co-location with the fluorescence emitted by 35S-RFP-HDEL plasmid, used as a control of endomembrane system location (**Figure 3Bb,f,j,n**). The MATI protein was also found in the plasma membrane, cytoplasm and within the nucleus (**Figure 3Ba,e**). Plasmolysis, induced by 1 M mannitol treatment confirmed the MATI location pattern in the cell-detached plasma membrane, with no fluorescence observed either in the cell wall or in the apoplast (**Figure 3Bm,o**). As expected, the 35S-GFP control showed fluorescence throughout the whole cell (**Figure 3Bi,k**). Furthermore, Col-MATI-GFP overexpressing lines were created to corroborate MATI subcellular location in the whole plant. The stable expression of 35S-MATI-GFP plasmid in transgenic Arabidopsis Col-0 plants confirmed the protein location in the endomembrane system, cytoplasm and nucleus as shown transgenic in roots of Arabidopsis transgenic plants (**Figure 3C**).

### Effects of MATI on Plant Resistance and Pest Performance

To investigate the role of MATI protein in plant defense, silenced lines (mati Salk) and overexpressing Arabidopsis Col-0 and Kon lines (Col-MAT and Kon-MATI, respectively) were generated. The characterization of the homozygous mati Salk line revealed a loss-of-function allele generated by the insertion of the T-DNA at the first part of the exon (Supplementary Figure 1A). On the other hand, gene expression analysis of transgenic plants expressing MATI gene constitutively in Col-0 and Kon backgrounds allowed the selection of lines that overexpressed the MATI ORF for further studies (Supplementary Figure 1Ba,b).

Homozygous knock-out (Col-mati), overexpressing lines (Col-MATI\_4.1, \_5.2 and Kon-MATI\_6.1, \_7.2) as well as the corresponding WT plants were infested with spider mites and plant damage was quantified 4 days after mite feeding (**Figures 4A,C**). Col-MATI overexpressing lines showed approximately three times less damage than non-transformed Col WT infested plants (exactly 2.83- and 2.68-fold for Col-MATI\_4.1 and \_5.2, respectively). Similarly, overexpression of MATI in Kon (Kon-MATI\_6.1 and \_7.2) resulted in a significant decrease in damage (1.25- and 2.7-fold for Kon-MATI\_6.1 and \_7.2, respectively) in comparison to infested Kon WT. In contrast, damaged area in the knockout Col-mati lines was about 2.5-fold greater than the damage quantified in Col-0 WT lines after mite feeding. The damage intensity measured as the chlorotic area on

infested leaves was proportional to the MATI expression levels. To ensure that chlorotic area correlates with mite enhanced feeding and is not a consequence of greater cell death, mite performance were determined after feeding on plants with different levels of MATI expression. Fecundity assays were carried out on leaves from different Col-0 genotypes and showed that synchronized mites feeding on WT and mati plants had higher fecundity rates than the ones feeding on overexpressing MATI lines (**Figure 4D**). Thus, greater leaf damage reflects a mite's ability to feed more intensely and to be more fecund, indicating that MATI overexpression correlates with plant's capability to defend against herbivory.

To evaluate whether the plant defense mediated by MATI gene was specific to acari infestation, leaf disks from different Col-0 genotypes were infested with a generalist lepidopteran, the beet armyworm S. exigua, and the leaf damage was quantified. As shown in **Figures 4B,C**, the damaged area was significantly higher in Col-mati and Col-0 WT lines than in the overexpressing Col-MATI\_4.1 line. In addition, the larval weight after 12 h of feeding on Col-mati and WT plants was twice and 1.5-fold greater, respectively, relative to the weight of the beet armyworm larvae fed on Col-MATI\_4.1 (**Figure 4D**). Therefore, MATI overexpression confers plant protection against multiple herbivores.

### MATI Interactome Reveals Its Participation in a Complex Network

Using NCBInr database after Co-IP assays, 94 proteins were specifically found in MATI sample but not in the control WT. Most of these proteins were recognized by the identification of one peptide (Supplementary Table 3). The identified proteins were classified in five different over-represented categories based on their GO biological function related to sulfur compound metabolism, generation of precursor metabolites and energy, photosynthesis and Serine metabolism (Supplementary Figure 3A). When the identification was performed with a singular enrichment analysis in AgriGO, the Go biological process categories overrepresented included oxidation-reduction and metabolic processes, photosynthesis, and generation of

PhyML method using MATI homologs from 27 plant species. Numbers are aLRT values for statistical support. ARATH, Arabidopsis thaliana; BRADI, Brachypodium distachyon; CAJCA, Cajanus cajan; CARPA, Carica papaya; CICAR, Cicer arietinum; CITSI, Citrus sinensis; COFCA, Coffea canephora; CUCSA, Cucumis sativus; ELAGV, Elaeis guineensis; GLYMA, Glycine max; GOSRA, Gossypium raimondii; HORVU, Hordeum vulgare; MALDO, Malus domestica; MANES, Manihot esculenta; MEDTR, Medicago truncatula; MUSAC, Musa acuminata; MUSBA, Musa balbisiana; ORYSA, Oryza sativa; PHAVU, Phaseolus vulgaris; POPTR, Populus trichocarpa; RICCO, Ricinus communis; SETIT, Setaria italica; SOLLY, Solanum lycopersicum; SOLTU, Solanum tuberosum; SORBI, Sorghum bicolor; THECC, Theobroma cacao; ZEAMA, Zea mays. (B) Logos representing hidden Markov models of the N-terminal parts of the MATI domain for the three main groups identified.

Student-Newman-Keuls test). (B) Confocal stacks spanning epidermal onion cells co-transformed with 35S-MATI-GFP, 35S-GFP control and/or 35S-RFP-HDEL control. Confocal projections from green fluorescent protein (GFP) (a,e,i), RFP (b,f,j), merged (c,g,k), and the corresponding Nomarski snapshots (d,h,l). Confocal images (m,n) and projections (o,p) of the MATI localization after plasmolysis with 1 M mannitol are shown. (C) Subcellular location of MATI protein in roots from transgenic Col-MATI-GFP lines. Observations with a Nomarski bright field are also shown (right panel). Bars are as indicated in all images.

precursor metabolites and energy (Supplementary Figure 3B). Additionally, the interactome network for the MATI putative partners revealed an enrichment of interactions between MATI putative interactors, with 247 observed interactions over the 100 expected for a total of 94 proteins. Among those partners putatively involved in sulfur compound metabolic process, 58 interactions were detected over the 6 expected (Supplementary Figure 3C). Since the two main overrepresented functional categories were related with sulfur metabolism and photosynthesis, different parameters involved in these physiological processes were further analyzed.

### MATI Modulates Plant Redox State and Photosynthesis

Since MATI expression was induced early upon mite feeding on Arabidopsis leaves, it was important to determine how MATI might modulate plant responses to herbivore feeding. Thus, the EL was determined in Col-0 genotypes after mite feeding. Results showed a significant higher leakage either constitutively or mite-induced leakage in Col-mati plants than in Col-MATI overexpressing or in WT plants (**Figure 5A**). As these changes in conductance are usually accompanied by ROS generation, H2O<sup>2</sup> concentrations were determined in the three Col-0 genotypes. The quantification of H2O<sup>2</sup> in infested plants, expressed as DAB units, demonstrated that Col-0 WT and Col-mati plants accumulated about twice and ten-times more DAB, respectively, than their corresponding non-infested genotypes. However, the H2O<sup>2</sup> levels in infested Col-MATI\_4.1 line changed only about 1.5-fold the DAB values relative to the non-infested line (**Figure 5B**). As thiol-type compounds are an important class of antioxidants that quench reactive oxidant species, variations in the total thiol group's content among different Col-0 genotypes were also quantified. T. urticae feeding caused a significant decrease in the thiol's group concentration in WT and Col-mati plants, while the thiol accumulation pattern was not significantly altered in Col-MATI samples (**Figure 5C**).

Three-week-old mati-mutant, overexpressing and control Col lines were used to determine chlorophyll and carotenoid content in the presence/absence of spider mites. While reduced levels of both pigments were observed in infested WT and Col-mati lines, high chlorophyll and carotenoid accumulation was detected in overexpressing Col-MATI\_4.1 lines after mite feeding in comparison to non-infested genotypes (**Figures 5D,E**). Consequently, MATI helps to maintain moderate levels of ROS necessary for defense signaling to control redox homeostasis, and to retain high levels of pigments to avoid photosynthetic cost.

### MATI Affects Plant Defense against Mites through Modulation of Hormonal Signaling

The hormonal analyses demonstrated that OPDA, JA, and its bioactive form JA-Ile, accumulated in leaf tissue during the initial 24 h in response to spider mite feeding, independently of the genotype (**Figures 6A–C**). The free OPDA content increased mainly in the spider mite-treated Col-MATI and mati lines which contained about twice the JA precursor in infested WT (**Figure 6A**). The JA and JA-Ile accumulation also increased in Col-mati and WT genotypes, but the highest hormone levels were detected in the Col-MATI overexpressing lines after mite infestation. It was detected as an increase of

FIGURE 4 | Plant damage of Col-0 and Kon genotypes infested with T. urticae and Spodoptera exigua and effects on pest perfomance. (A) Leaf damage on Arabidopsis Col-0 (WT, Col-MATI\_4.1, \_5.2, and Col-mati lines) and Kon (WT, Kon-MATI\_6.1 and \_7.2) genotypes 4 days after T. urticae feeding. (B) Leaf damage on Arabidopsis Col-0 (WT, Col-MATI\_4.1, and Col-mati lines) 12 h after S. exigua feeding. Data are means ± SE of 6 (T. urticae) and 32 (S. exigua) replicates. Different letters indicate significant differences (P < 0.05, One-way ANOVA followed by Student-Newman-Keuls test). (C) Leaf phenotypes of Col-0 genotypes 4 days and 12 h after T. urticae and S. exigua feeding, respectively. (D) T. urticae and S. exigua performance, referred as number mite eggs and armyworm larval weight, after feeding for 4 days and 24 h, respectively, on Col-0 genotypes. Data are means ± SE 6 (T. urticae) and 32 (S. exigua) replicates. Different letters indicate significant differences (P < 0.05, One-way ANOVA followed by Student-Newman-Keuls test). Blue bars (Col background) and gray bars (Kon background).

FIGURE 5 | Redox status and photosynthetic pigment quantification in Arabidopsis Col-0 genotypes 24 h after T. urticae feeding. (A) Electrolyte leakage measurements. (B) Accumulation of hydrogen peroxide (DAB). (C) Total thiol groups. (D) Total carotenoids. (E) Total chlorophyll. WT, Col-mati, and Col-MATI\_4.1 overexpressing lines were used. Data are means ± SE of six replicates. Different letters indicate significant differences (P < 0.05, Two-way ANOVA followed by Student-Newman-Keuls test).

approximately 45- and 16-fold in the JA and JA-Ile content, respectively (**Figures 6B,C**). Simultaneously, SA and ABA concentrations were also determined in the same rosette samples. Interestingly, high levels of SA were accumulated in noninfested Col-MATI lines but they were significantly reduced after mite feeding. In contrast, SA increased in response to infestation in Col-mati and WT lines (**Figure 6D**). Regarding ABA levels, they exclusively increased in infested Col-mati lines (**Figure 6E**). Thus, MATI is involved in the modulation of different hormonal signaling pathways in response to spider mite attack.

To further analyze the molecular basis of the MATI gene function, the expression pattern of some JA-, SA- and ABA-related genes located at different levels in the hormonal signaling pathways was analyzed 24 h after mite feeding. Lipoxygenase 3 (LOX3), Allene Oxide Synthase (AOS), MYC2 transcription factor (MYC2), vegetative storage protein 2 (VSP2), and plant defensin (PDF1.2) were the selected genes to dissect JA pathway. Non-infected Col WT plants showed low levels of constitutive expression of all JA-related genes but their expression increased after spider mite infestation. Interestingly, most of the selected defense genes presented higher mRNA levels in the infested overexpressing MATI than in WT or mati lines. Only the expression values of the PDF1.2 gene related to the JA/Et defense branch were lower in Col-MATI\_4.1 overexpressing line than in Col-mati or WT plants (**Figures 7A–E**). Similarly, Isochorismate Synthase 1 (ICS1), Non-expresser of PR gene1 (NPR1) and Pathogenesis Related (PR1) genes involved in SA-biosynthesis, regulation and final defense product, respectively, were selected to study SA implications. The three genes showed lower expression in infested than in non-infested Col-MATI lines while in Col-mati and WT lines the expression of some genes were increased, some decreased and other did not present differences in response to mites (**Figures 7F–H**). Additionally, the expression of the same JA- and SA-regulated genes was analyzed in early mite infestation assays, just 1 h after mite attack. Regarding the genes involved in the JA pathway, LOX3, AOS, and MYC2 were up-regulated after mite feeding in WT, MATI, and mati lines while the VSP2 gene was exclusively and highly induced in infested overexpressing MATI lines. Curiously, the PDF1.2 gene was only induced in infested WT lines since its basal expression levels in non-infested MATI lines was as higher as the induced levels detected in WT plants after mite feeding. This gene was not induced in mati lines either infested or not infested (Supplementary Figures 4A–E). The expression of the genes involved in SA-biosynthesis, ICS1, NPR1, and PR1, was similar or lower in infested and non-infested Col-MATI lines but their expression levels in mati lines was not altered in response to mites. In contrast, ICS1 and NPR1 genes were up-regulated in infested WT lines and the PR1 gene did not vary (Supplementary Figures 4F–H).

Besides, three genes related to ABA pathway, ABA deficient 1 (ABA1), Regulatory Component of ABA Receptor 3 (RCAR3), and Responsive to Dessication 22 (RD22), were also checked. The ABA1 gene decreased after mite infestation following the same expression pattern in all genotypes, but RCAR3 and RD22 genes exclusively increased in mati knock-down lines after mite infestation (Supplementary Figure 5). All these results are in agreement with the hormonal content measured in the MATI overexpressing and knock-down genotypes and reinforce the importance of MATI in the regulation of different hormonal signaling pathways.

### DISCUSSION

In the last years, many studies have been conducted to disclose the plant molecular events in response to phytophagous arthropods, which have revealed a complex combination of metabolic, developmental and signaling networks (Wu and Baldwin, 2010; War et al., 2012; Johnson et al., 2016; Schuman and Baldwin, 2016). In this scenario, the identification and further characterization of new molecules, compounds and pathways involved in the defense and the understanding of the mechanisms of plant protection is essential for pest control and crop improvement.

The MATI gene was selected as a potential candidate involved in defense against T. urticae based on its high expression levels in the resistant Bla-2 Arabidopsis accession after mite infestation in comparison to levels observed in the susceptible Kon accession (Zhurov et al., 2014). Our results confirm and support the wide role of the MATI protein in plant defense against herbivory based on: (i) the spider mite T. urticae inflicts more leaf damage in Col-mati knockdown than in Col-MATI overexpressing plants, and silencing of MATI improves mite performance; (ii) the Kon-MATI plants are able to partially revert the susceptible phenotype of Kon accession against spider mite attack and; (iii) larvae of the lepidopteran S. exigua grow more poorly on Col-MATI overexpressing lines and consumes less leaf tissue than do in knock-out mati or WT lines. Accordingly, previous transcriptomic studies showed the induction of MATI upon Pieris rapae herbivory (Coolen et al., 2016). MATI belongs to a protein family of unknown function highly conserved among angiosperm species, suggesting a potential role of this protein against herbivore attack across many species. Unfortunately, the absence of any domain of known function in its sequence made it difficult to decipher the molecular function of this small protein. To deal with its biological function, Co-IP assays were performed to identify putative protein interactors. It is assumed that proteins belonging to a biological complex should have related molecular functions, should be located in the same cellular compartment and should participate in a common physiological process (Zhang et al., 2008). Results revealed that MATI protein could be participating in a complex network since the classification of the putative interactors into biological categories highlighted related physiological processes, such as photosynthesis, oxidation-reduction pathways, redox homeostasis, sulfur metabolism and generation of precursor metabolites. Thus, MATI protective role seems to be involved in any aspect of the redox metabolism, and probably associated to buffering protein thiol groups against excessive oxidation. This putative physiological role can be easily integrated in the response of the plant to T. urticae.

In general, the initial plant–herbivore interaction triggers signal transduction pathways that commonly include Ca2+ signaling, production of ROS, phosphorylation cascades and transcriptional regulatory events leading to specific defense responses (Wu and Baldwin, 2010; War et al., 2012). Defensive responses begin when herbivores interact with the plant by introducing elicitors and triggering plant-derived signaling molecules (Zebelo and Maffei, 2015). The first consequence in the plant cell is the appearance of changes in ion fluxes, mainly Ca2+, associated to variations in the plasma membrane polarity leading to EL. Electrolyte leakage is almost instantaneously detected after stress application and it is usually accompanied by ROS accumulation (Demidchik et al., 2014). ROS are considered essential components in signaling toward pests, including spider mites (Leitner et al., 2005; Santamaria et al., 2012a). Excess of oxidative stress caused by H2O<sup>2</sup> and strong plasma membrane depolarization finally result to programmed cell death. Conversely, moderate H2O<sup>2</sup> concentrations and plasma membrane depolarization may differentially sense defense signaling (Foyer and Noctor, 2005; Bailey-Serres and Mittler, 2006; Baxter et al., 2014). Thiol groups are involved in reducing oxidative stress controlling the adverse effects on the cell survival exerted by ROS. Our findings indicate that after mite infestation leaves from Col-mati and WT plants showed a strong increase in the levels of H2O<sup>2</sup> and electrolytic leakage together with a sharp decrease in the accumulation of thiol groups, probably associated to cell death promoted by acari feeding. In contrast, the infestation of MATI overexpressing plants do not lead to an increase of electrolytic leakage and causes a moderate accumulation of H2O2. Thus, the manipulation of plant antioxidant status should evidence a clear impact on herbivore performance and the subsequent plant defense phenotypes. According to this hypothesis, Arabidopsis pad2-1 mutants (glutathione deficient) or vtc1-1 mutants (ascorbic acid deficient) resulted more susceptible to Spodoptera littoralis than control lines (Schlaeppi et al., 2008). Thus, we can affirm that MATI gene is anyway involved in the regulation of the cell redox status mediated by H2O<sup>2</sup> and thiol groups to protect the cell from an excessive oxidation and to facilitate the activation of defense signaling pathways.

Extensive characterization of plant–pest interactions has also demonstrated the integration between ROS and hormonal signaling in plant defense (Kerchev et al., 2012b; Mur et al., 2013). The feeding mode of the herbivore and the plant host determine the participation of phytohormones which appear to modulate the fine-tuning of defenses in response to each herbivore species. Generally, jasmonates are associated with the activation of defense responses toward a wide range of herbivorous insects (Howe and Jander, 2008; Wu and Baldwin, 2010). Likewise, JA-induction of plant defenses against spider mites has been widely described (Li et al., 2002, 2004; Ament et al., 2004; Schweighofer et al., 2007; Zheng et al., 2007; Zhang et al., 2009; Zhurov et al., 2014; Martel et al., 2015; Diaz-Riquelme et al., 2016). In Arabidopsis, Zhurov et al. (2014) reported the pivotal role of JA in establishing effective defenses against mites using mutants defective in JA biosynthesis (aos) and JA-regulated transcription (myc), which displayed increased susceptibility to T. urticae attack. As expected, we found that the biologically active jasmonate molecules JA and JA-Ile accumulated after 24 h mite infestation in all genotypes tested, but with a

transcription factor (MYC2). (D) Vegetative Storage Protein 2 (VSP2). (E) Plant Defensin (PDF1.2). (F) Isochorismate Synthase 1 (ICS1). (G) Non-expresser of PR gene1 (NPR1). (H) Pathogenesis Related (PR1). Genes were analyzed in Arabidopsis Col-MATI\_4.1, Col-mati, and WT plants 24 h after mite feeding. Gene expression levels were normalized to the ubiquitin gene expression. Data are means ± SE of six replicates. Different letters indicate significant differences (P < 0.05, Two-way ANOVA followed by Student-Newman-Keuls test).

remarkably strong accumulation in Col-MATI overexpressing lines. These results link the redox balance to the presence of JA-regulated compounds and involve the MATI gene in the final modulation of both features. The question now is how MATI affects JA signaling pathway. In Arabidopsis, plant defenses against herbivores are differentially regulated by two different branches of the JA signaling pathway that are antagonistically controlled by the transcription factors MYC2 and ORA59. The specialist insect herbivore P. rapae induces the expression of MYC2 and the MYC2-branch marker gene VSP2, and suppresses the transcription of ORA59 and the JA/ethylene branch marker gene PDF1.2 (Verhage et al., 2011; Vos et al., 2013). Our results show an overexpression of genes involved in the production of the JA precursor OPDA and in the JA-regulated responsive gene VSP2, concomitant with a minor induction of PDF1.2 in the Col-MATI plants either at 1 h or at 24 h of infestation. These

findings could be interpreted as a MATI-regulated preference of MYC2-branch against ORA59-branch. On the other hand, JA and JA-Ile production by Arabidopsis Col-0 plants also increased after the attack of the chewing insect S. exigua (Rehrig et al., 2014). Our findings with S. exigua suggest that the JA accumulation mediated by MATI gene also conferred resistance to this chewing insect extending its potential role as key defense modulator. These results correlate with the observed higher levels of JA and JA-Ile in overexpressing MATI plants, corroborating that JA is a functional output of the MATI-regulated Arabidopsis defense and suggesting that MATI exerts its function upstream in the MYC2-branch of the JA signaling pathway.

However, herbivore response does not imply the only activation of the JA signaling pathway. Other hormones are induced upon feeding in a specific plant–insect interaction (Nguyen et al., 2016) prompted by herbivore-specific associated molecular patterns (Acevedo et al., 2015; Xu et al., 2015). The extreme case is presented by many phloem-feeding insects, which mainly induce SA accumulation in the plant (Foyer et al., 2016). The fine-tuning of plant defense responses to specific attackers would be achieved by the cross-talk between JA and other phytohormones (Erb et al., 2012; Pieterse et al., 2012).

The antagonistic cross-talk effect between JA and SA signal transduction pathways have been explored at molecular level in the context of plant resistance to herbivores (Kawazu et al., 2012; Thaler et al., 2012; Vos et al., 2015). This antagonism has been exploited by herbivores to suppress JA-mediated defenses. For instance, the whitefly Bemisia tabaci induces the accumulation of SA in Arabidopsis, which plays a key role in the suppression of the effectual JA defenses of the host plant downstream JA signaling enhancing whitefly performance (Zhang et al., 2013). Despite this functional antagonism, there are several examples of plant–herbivore interactions leading to simultaneous JA and SA accumulation (Diezel et al., 2009; Chung et al., 2013; Zhang et al., 2013). SA content also increases in tomato and Arabidopsis in response to spider mite infestation (Kant et al., 2004; Glas et al., 2014; Zhurov et al., 2014). The significance of SA accumulation could depend on the specific plant–herbivore interaction, since profound divergences have been found in the transcriptional response of tomato and Arabidopsis plants to T. urticae (Zhurov et al., 2014; Martel et al., 2015) and in the tomato response to different T. urticae strains (Alba et al., 2015). In tomato, SA accumulation has been associated with a higher protective role against T. urticae (Kant et al., 2004, 2008; Villarroel et al., 2016). On the contrary, spider mite infested Arabidopsis plants deficient in SA biosynthesis or downstream signaling did not significantly alter plant damage or mite performance (Zhurov et al., 2014). Since SA induction could hamper JA-induced responses, the SA accumulation induced by spider mites in Arabidopsis could be a strategy to try to minimize accurate plant defense responses. Signaling effects caused by MATI are also involved in the modulation of SA levels. MATI overexpressing plants constitutively accumulate SA, which could be related to a higher basal resistance of these plants against pest/pathogens vulnerable to SA-induced defenses. After spider mite feeding, whereas in WT and Col-mati plants the SA accumulation strongly

increase probably as a mechanism triggered by the mite to try to impede a correct defense response, the SA concentration in overexpressing MATI plants decreased, as a response leading to optimize the functionality of the JA-signaling pathway.

ABA signaling is associated to plant resistance to herbivores by interacting JA-signaling (Dinh et al., 2013; Vos et al., 2013). Commonly, this interaction promotes a higher protection but, Hillwig et al. (2016) demonstrated that ABA deficiency increases defense responses toward the aphid Myzus persicae which induces the SA signaling pathway. The ABA content increased in mati knock-down lines after mite infestation, but did not change in Col-MATI or WT plants. These results suggest a residual role of this hormone in the protective response against T. urticae. The effect on ABA accumulation in Col-mati plants could be due to a wide reprogramming of hormonal signaling pathways associated to the absence of the MATI protein.

Phytophagous arthropods have a clear impact on metabolism of the plant, limiting resources for growth and development in favor to plant defense (Huot et al., 2014). The classical plant response to herbivores involves photosynthesis reduction, stomata closure, enhanced respiration and changes in the cellular redox state, among other alterations (Kerchev et al., 2012a,b; Baxter et al., 2014). As expected, a decrease in pigment levels was found in infested Col-mati and WT lines associated to the damage observed in Arabidopsis leaves. In contrast, although leaves of Col-MATI lines were also damaged after mite feeding, they tended to accumulate chlorophyll and carotenoids. Interestingly, analysis of expression data for MATI gene in GENEVESTIGATOR<sup>17</sup> evidences a strong silencing of this gene when plants are subjected to darkness treatments, linking MATI with the capacity of the cell to accumulate photosynthetic pigments. These results suggest that the cell reprogramming exerted by MATI is associated to a minimization of the trade-offs between plant defense and growth.

Taken all data together, a working model of MATI protein in defense against spider mites is proposed (**Figure 8**). Our findings suggest that MATI encodes a protein involved in the accumulation of reducing agents upon spider mite attack to control plant redox homeostasis avoiding excessive oxidative damage. Besides, MATI causes signaling effects that modulate different hormonal signaling pathways, affecting the expression of genes involved in biosynthesis and signaling of the JA and SA hormones. In consequence, MATI leads high plant tolerance to

<sup>17</sup>https://genevestigator.com/gv/

### REFERENCES


spider mites reducing leaf damage and mite fecundity. Future experiments will be performed to explain the exact role of MATI protein in this extremely coordinated modulation of signaling pathways and to analyze if MATI is also implicated in defense against other pests and pathogens. Furthermore, how MATI takes part in the modulation of the levels of photosynthetic pigments is a key question to address the potential of MATI proteins to be exploited as biotechnological tools.

### AUTHOR CONTRIBUTIONS

ID, VG, and MM conceived the research. MS performed most of the experimental research. FO and AA participated in the insect feeding assays. ID, MM, VG, and MS participated in the design, the acquisition, analysis, or interpretation of data for the work. All authors contributed to final version of the manuscript.

### FUNDING

This work was supported by projects from Ministerio de Economía y Competitividad of Spain (projects BIO2014-53508- R and 618105-FACCE-Era Net Plus) and the Government of Canada through Genome Canada and the Ontario Genomics Institute (OGI–046), the Ontario Research Fund (RE08-067) and the Natural Sciences and Engineering Research Council of Canada. MS was recipient of a post-doctoral grant from the Ministerio de Economia y Competitividad of Spain (subprogram Juan de la Cierva 2012).

### ACKNOWLEDGMENTS

We thank Pablo González-Melendi (CBGP-UPM-INIA) for his support in the confocal microscope and Dr. Pierre Rouze (VIB-Ghent University) for his advice on the bioinformatics analysis.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpls.2017.00975/ full#supplementary-material

line population. Exp. Appl. Acarol. 47, 35–47. doi: 10.1007/s10493-008- 9192-4


and emission of (E,E)-4,8,12-trimethyltrideca-1,3,7,11-tetraene in tomato are dependent on both jasmonic acid and salicylic acid signaling pathways. Planta 224, 1197–1208. doi: 10.1007/s00425-006-0301-5


methyl salicylate emission in tomato. Plant Physiol. 135, 483–495. doi: 10.1104/ pp.103.038315


efficient construction of fusion genes for plant transformation. J. Biosci. Bioeng. 104, 34–41. doi: 10.1263/jbb.104.34


and MPK6, modulates innate immunity, jasmonic acid, and ethylene levels in Arabidopsis. Plant Cell 19, 2213–2224. doi: 10.1105/tpc.106.049585


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2017 Santamaría, Martinez, Arnaiz, Ortego, Grbic and Diaz. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Effect of Endophyte Infection and Clipping Treatment on Resistance and Tolerance of Achnatherum sibiricum

### Junhua Qin, Yuan Gao, Hui Liu, Yong Zhou, Anzhi Ren\* and Yubao Gao

College of Life Sciences, Nankai University, Tianjin, China

It is well-documented that endophytes can enhance the resistance of agronomical grasses, such as tall fescue and perennial ryegrass to herbivory. For native grasses, however, the related reports are limited, and the conclusions are variable. Achnatherum sibiricum is a grass native to the Inner Mongolian steppe. This grass is highly infected by endophytes but does not produce detectable endophyte-related alkaloids known under normal conditions. In this study, the contributions of endophytes to the resistance of A. sibiricum to Locusta migratoria were studied. We found that locusts preferred EF (endophyte-free) plants to EI (endophyte-infected) plants, and the weight of locusts fed on EI plants was significantly lower than those fed on EF plants. Hence, endophyte infection significantly enhanced the resistance of the host to L. migratoria. Endophyte infection significantly decreased the concentration of soluble sugar and amino acids while significantly increased the concentration of total phenolic content, and these metabolites may contribute to herbivore resistance of the host. The clipping treatment further strengthened the locust resistance advantage of EI over EF plants. After clipping, the weight of the locusts fed on EI plants significantly decreased compared with those fed on unclipped plants, whereas the weight of the locusts fed on EF plants increased significantly. The results suggested that endophyte infection could increase herbivore resistance while decreasing the tolerance of the host grass by mechanisms apart from endophyte-conferred alkaloid defense.

Keywords: Achnatherum sibiricum, clipping, endophyte, locust, resistance, tolerance

### INTRODUCTION

Plants can coexist with many types of microbial symbiosis, such as rhizosphere bacteria, mycorrhizal fungi and endophytes. Endophytes live asymptomatically within many cool season grasses for at least a portion of their life cycle (Carroll, 1988). Thus far, the beststudied endophytes are Neotyphodium lolii and Neotyphodium coenophialum, which colonize perennial ryegrass (Lolium perenne L.) and tall fescue (Lolium arundinaceum Darbyshire ex. Schreb.), respectively. Pioneering studies by Prestidge et al. (1982) reported that endophyte infection could increase the resistance of tall fescue to Listronotus bonariensis (Argentine stem weevil); since then, more studies on the increased herbivore resistance to endophyte infection have been reported in cultivated grasses, including species of Festuca, Lolium, and Poa (Siegel et al., 1990; Bultman and Conard, 1998; Kunkel et al., 2004). The anti-herbivore

#### Edited by:

Anton Hartmann, Helmholtz Zentrum München, Germany

#### Reviewed by:

Daolong Dou, Nanjing Agricultural University, China Andrea Manzotti, University of Copenhagen, Denmark

> \*Correspondence: Anzhi Ren renanzhi@hotmail.com

#### Specialty section:

This article was submitted to Plant Biotic Interactions, a section of the journal Frontiers in Microbiology

Received: 01 September 2016 Accepted: 28 November 2016 Published: 15 December 2016

#### Citation:

Qin J, Gao Y, Liu H, Zhou Y, Ren A and Gao Y (2016) Effect of Endophyte Infection and Clipping Treatment on Resistance and Tolerance of Achnatherum sibiricum. Front. Microbiol. 7:1988. doi: 10.3389/fmicb.2016.01988

properties of endophyte infection are largely attributable to the production of a variety of alkaloids (Schardl et al., 2004; Bush and Fannin, 2009). Thus far, four classes of alkaloids associated with infected host grasses have been detected, including saturated aminopyrrolizidines (lolines), pyrrolopyrazines (peramines), ergot alkaloids and indolditerpenes (lolitrems) (Bush et al., 1997; Schardl et al., 2012). Lolines have a broad spectrum of activity against insects (Bultman et al., 1997). Peramine is known to act as a feeding deterrent (Prestidge and Gallagher, 1988). Ergot alkaloids are primarily active against vertebrates, whereas lolitrems are responsible for neurotoxic disorders of mammals (Prestidge, 1993).

Endophytes not only exist in cultivated grasses such as tall fescue and perennial ryegrass but are also widely distributed in native grasses (Leuchtmann, 1993). There are a variety of native grass hosts, and the circumstances of endophyte infection are more complicated in native grasses (Afkhami and Rudgers, 2009; Faeth and Saari, 2012). In contrast with cultivated grasses, endophytes in native grasses usually produce fewer classes and a lower concentration of alkaloids (Faeth et al., 2002). Some endophytes only produce peramine, and some do not even produce alkaloids (Leuchtmann et al., 2000). Can endophyte infection in native grasses, which do not produce alkaloids or produce a low concentration of alkaloids, enhance the herbivore resistance of host plants? The limited interactions reported are highly variable. In a feeding assay with the leaf material of Brachypodium sylvaticum, Brem and Leuchtmann (2001) found that insect larvae performed significantly better on a diet of uninfected leaves. Lopez et al. (1995) found that Melanoplus femurrubrum consumed similar amounts of Arizona fescue (Festuca arizonica) regardless of endophyte status. Faeth and Shochat (2010) found that endophyte-infected (EI) Arizona fescue harbored more herbivorous insects than uninfected plants, suggesting that endophyte infection decreases rather than increases resistance to herbivores.

Achnatherum sibiricum (L.) Keng is a caespitose perennial grass that is widely distributed in northern China and commonly infected by Epichloë endophyte (Wei et al., 2006). Within the genus Achnatherum there are five sections, and A. sibiricum belongs to section Achnatheropsis (Tzvel.) N. S. Probatova. In this section there are nine species, including seven Asian and two American species (Wu and Lu, 1995). Before A. sibiricum, two other species, Achnatherum inebrians and Achnatherum robustum, have been reported for their narcotic effects on livestock resulting from endophyte infection, and hence are known as 'drunken horse grass' and 'sleepy grass,' respectively (Petroski et al., 1992; Bruehl et al., 1994). In contrast, A. sibiricum exhibits no obvious herbivore deterrence ability according to local records and our own observations. A. sibiricum is infected by two species of endophytes, Epichloë gansuensis and Epichloë sibirica, in its native populations (Zhang et al., 2009; Li et al., 2015). In the Inner Mongolian steppe, locusts are the primary consumer, and they affect grassland productivity and compete with domestic animals for food resources (Kang et al., 2007). The biogeography of 150 species of locust fauna on the Inner Mongolian Plateau has been studied, of which 10–15 species are considered as grassland pests (Kang et al., 2007). Locusta migratoria (Orthoptera: Acrididae) is oligophagous, feeding mainly on grasses of Gramineae and Cyperaceae (Bernays et al., 1976). It is famous for its wide breeding range, strong stress resistance and fecundity, and it can also have a long-distance migration.

In our previous investigation, we found that alkaloids associated with endophyte infection were detected in neither infected nor uninfected plants when grown under normal conditions. After clipping, only peramine was detected, but its concentration in the sheath of infected plants ranged from 0 to 0.6 ppm. In this study, EI A. sibiricum was adopted as plant material. Here, endophyte infection instead of endophyte species was considered. L. migratoria, a common herbivore of grasses (Yue et al., 2009), was adopted as the feeding herbivore. Clipping is a common practice in our sampling grassland; thus, we use clipping as the interference. We wondered whether endophyte infection has a positive effect on the insect resistance of A. sibiricum and whether clipping can influence insect resistance of infected A. sibiricum. Furthermore, how do the strategies of infected and uninfected A. sibiricum respond to clipping?

### MATERIALS AND METHODS

### Seed Source

Seeds of A. sibiricum were collected from the National Hulunber Grassland Ecosystem Observation and Research Station (49.06◦ N, 119.40◦ E) in 2012. After detection, we found that the endophyte infection rate was 100%. To obtain endophyte-free (EF) seeds, EI seeds were placed in a 60◦C oven for 30 days. The previous study in our lab showed that a high temperature treatment for 30 days could completely destroy endophytes in the seeds and that it had no significant influence on seed germination rate, germination potential, and germination index (Li et al., 2010); a similar method to obtain EF seeds was also reported by Kannadan and Rudgers (2008) in grove bluegrass.

### Locusta migratoria

Locusta migratoria is not a dominant species in the Inner Mongolia steppe. It was chosen as an herbivore because it is known to cause significant damage to grasses and because it is readily available and easily cultured. Eggs were purchased from a local pet shop. After hatching unearth eggs in an oven containing moist vermiculite for approximately 2 weeks in the dark at 25◦C, the nymphs were placed in a special device, which was conducted in a constant temperature room at 25◦C with a 12 h light /12 h dark photoregime, and fed with wheat seedlings. After growing into the requisite instars (second instar or fourth instar), the nymphs were used in the experiments.

### Plant Treatment

The plants used in this experiment were obtained either from EI seeds or EF seeds. The seeds were planted in white plastic pots (20 cm in diameter and 20 cm in depth) filled with vermiculite. After 10 days' growth, 20 plants of approximately equal size were maintained in each pot. After 5 weeks of cultivation, the plants in half pots were clipped with scissors 5 cm above the soil surface, and the other half was retained as a control (CK, unclipping). All seedlings were separated into four groups, i.e., EI-CK, EI-Clipping, EF-CK, and EF-Clipping. The plants were cultivated for a further 3 weeks before the locust feeding experiment was performed. Each group comprised 10 replicates, with five replicates for the locust feeding experiment and the other five replicates for the sampling and measurement of physiological characteristics. The experimental plants were watered and fertilized with Hoagland nutrient solution as needed. The experiment was conducted in the greenhouse at Nankai University, Tianjin, China.

### Choice Feeding Experiment

fmicb-07-01988 December 14, 2016 Time: 11:33 # 3

After 24 h' starvation, fourth instar nymphs of the locusts were transferred to transparent plastic containers of a 19 cm height and a 8 cm diameter. As food, 0.5 g leaf blades per pot per treatment were cut and offered in small glass dishes, separately. 0.5 g plant material was sufficient for the locusts feeding in 1 h to avoid being depleted completely during the trial. There were 10 plastic containers in total. Eight locusts were added to each of five containers. Another five containers were used as a control to calculate the reduced leaf quality due to evaporation. After 1 h of exposure to herbivory, all plants from all containers were harvested, and the fresh biomass was recorded.

### Nymphs' Growth Experiment

We equipped every pot with a steel frame of 45 cm height and 20 cm diameter which was coated with a nylon stocking. To launch the experiment, three weighed second instar nymphs of the locust hatchlings from one population were introduced into each of these resulting cages. There were 20 pots in total, including EI-CK, EI-Clipping, EF-CK and EF-Clipping, and each treatment was performed for five replicates. The experiment lasted for 5 days, and the plant material was adequate for the second instar locusts. In the end, the biomass of the nymphs was weighed and recorded.

### Quantification of Amino Acid Concentrations

The amino acids in the shoot were analyzed by reverse-phase high-performance liquid chromatography (HPLC) with precolumn derivatization using 2, 4-dinitro-fluorobenzene (DNFB) according to Li and Sun (2004). The standard solutions were AA-S-18 (Sigma-Aldrich) and stored at 2–8◦C. The analysis was performed using a Wasters HPLC system (Waters 1500 series). A reverse-phase C18 column (5 µm, 150 mm × 4.6 mm) and fluorescence detector were used for the chromatographic separation. The column was maintained at 25◦C with a gradient (1 ml min−<sup>1</sup> flow) programmed as follows: 84/16 (6 min),76/24 (6 min), 70/30 (6 min), 60/40 (12 min), 50/50 (7 min), 2/98 (5 min), 20/80 (8 min), 84/16 (5 min), and 84/16 (5 min holding) of eluent A/eluent B. Eluent A comprised 50% acetonitrile and 50% water. Eluent B comprised 37.5 mmol L−<sup>1</sup> NaAc, 10% acetonitrile, and 1% N,N-Dimethyl-formamide (with a pH of 6.4 adjusted with glacial acetic acid).

### Other Variables Measured

The soluble sugar content was analyzed according to Buysse and Merckx (1993) and Yu et al. (2011). The total phenolic concentration was determined according to Malinowski et al. (1998).

### Root Morphology and Plant Regrowth Rate

Fresh roots were washed and then scanned with an EPSON 1680 scanner (Epson, Long Beach, CA, USA) in 400 dpi resolution. After scanning, the root image was analyzed with WinRHIZO 2012 software to obtain parameters such as root length, root surface area and root average diameter. To assess the effect of endophyte infection on the regrowth ability of A. sibiricum after clipping, we grew additional EI (n = 5) and EF (n = 5) plants in white plastic pots, treated as before, and the plants were clipped 5 cm above the soil surface after 5 weeks of growth. The removed shoot tissue was oven dried and weighed. The plants were allowed to regrow for 3 weeks, at which time they were clipped at the soil surface. The shoot material was oven dried and weighed. The rate of regrowth of the plants was calculated as (ln[final mass]-ln[initial mass])/21 days (Bultman et al., 2004).

### Statistical Analysis

For the amino acids, we performed a principal components analysis (PCA) on the correlations among the 17 response variables and then performed factor rotation using the varimax method (Rasmussen et al., 2008; Guo et al., 2014). A varimax rotation is a change of coordinates that maximizes the sum of the variances of the squared loadings. This method increases the distinction between the large and small loading variables and so makes the biological interpretation of the axes simpler (Rasmussen et al., 2008). After varimax rotation, we retained four rotated factors (RFs). The RF variables were subjected to a two-way ANOVA with endophyte and clipping treatment as the factors. Other indexes were analyzed using multivariate and univariate analyses of variance. All analyses were performed using SPSS 21.0 software. The effects were considered significant if P < 0.05.

## RESULTS

### Bioassay

In the choice experiment, the performance of L. migratoria was significantly influenced by endophyte infection and the interaction of endophyte infection and clipping (**Table 1**). Locusts preferred EF plants to EI plants. After clipping, EF plants were fed on more by locusts than the unclipped control plants. For EI plants, however, the biomass fed by locusts tended to be less with clipping treatment, but the difference was not significant. Thus, the difference of the biomass fed on by locusts between EI and EF plants were more obvious after clipping (**Figure 1**).

In the growth experiment on the nymphs, the weight of the second instar locusts was significantly influenced by both endophyte infection and the interaction of endophyte infection


TABLE 1 | Two-way ANOVA for leaf consumption and physiological indexes of endophyte-infected (EI) and endophyte-free (EF) Achnatherum sibiricum under the clipping treatment.

Significant P-values (P < 0.05) are in bold.

fmicb-07-01988 December 14, 2016 Time: 11:33 # 4

and clipping (**Table 1**). Endophyte infection negatively affected the weight of L. migratoria nymphs. In the control group, the weight of the locusts fed by EI plants was significantly less than those fed by EF plants. After clipping, the weight of the locusts fed on EI plants decreased more than those fed on the unclipped control while the weight of the locusts fed on EF plants increased. Thus, the difference between the weights of the nymphs fed on the EI and EF plants was more obvious after clipping (**Figure 1**).

### Amino Acid Concentration in Plants

The pre-column derivatization of the DNFB method and the HPLC technique were used to measure the relative composition of amino acids in the shoot of A. sibiricum. Because the responses of the 17 amino acids that were measured were not independent, we then used a PCA to reduce the number of amino acid response variables to a new set of composite variables. To facilitate interpretation of the principal components, we subjected the first four principal components to factor rotation with the most common form of factor rotation, varimax rotation, and we retained four rotated factors (RF1, RF2, RF3, and RF4, which accounted for 80.5% of the total variance) (**Figure 2**). As the values of the RF increased, the variables that load heavily and positively (loading ≥ + 0.5) also increased, while the variables that load heavily but negatively (loading ≤ −0.5) decreased. The standardized univariate responses of these variables are shown in **Figure 3** to facilitate the interpretation of the multivariate responses and to allow a closer inspection of the variables loading heavily onto RF1, RF2, RF3, and RF4.

Five amino acids, Thr, Asp, Lys, His, and Val loaded heavily and positively onto RF1, and Pro loaded heavily but negatively onto RF1 (**Figure 2**). Endophyte infection, the clipping treatment and the interaction between endophyte infection and clipping significantly affected RF1 (**Table 2**). In the unclipped control group, RF1 in the EI plants was much lower than that in the EF plants. After clipping, there was no significant change in RF1 in the EI plants, whereas RF1 significantly increased in the EF plants. Thus, after clipping, the difference between the EI and EF plants was more obvious (**Figures 3A,B**). Leu, Phe, Tyr, Met, Gly, Glu, and Ala loaded heavily and positively onto RF2 (**Figure 2**). Only the interaction between endophyte infection and clipping (P = 0.088) affected RF2 under the 0.1 significance level (**Table 2**). In the unclipped group, RF2 in the EI plants was significantly lower than that in the

EF plants; after clipping, however, there was no significant difference between the EI and EF plants (**Figure 3C,D**). Ile loaded heavily and positively onto RF3 while Cys and Ser loaded heavily but negatively onto RF3 (**Figure 2**). Clipping and the interaction between endophyte infection and clipping significantly affected RF3 (**Table 2**). There was no significant difference in RF3 between the EI and EF plants in the control group. After clipping, RF3 in both the EI and EF plants decreased, but it decreased more in the EI plants. Thus, after clipping, RF3 in the EI plants was much lower than that in the EF

plants (**Figures 3E,F**). Arg, Glu, and Ala loaded heavily and positively onto RF4 (**Figure 2**). Neither endophyte infection nor the clipping treatment had an effect on RF4, and RF4 showed no variation before and after the clipping treatment (**Figures 3G,H**).

### Soluble Sugar Content in Plants

The soluble sugar content in the shoot of A. sibiricum was not only influenced by the presence of endophytes but also by the interaction of endophyte infection and clipping (**Table 1**). Endophyte infection negatively influenced the soluble sugar content. In the control group, the soluble sugar content in the EI plants was much lower than that in the EF plants. The clipping treatment increased the soluble sugar content in the EF plants but decreased that in the EI plants (**Figure 4A**).

### Total Phenolic Content in Plants

The total phenolic content of A. sibiricum was influenced by endophyte presence, clipping and the interaction of endophyte and clipping (**Table 1**). The total phenolic content in the shoot did not differ greatly between the EI and EF plants in the control group. After clipping, the total phenolic content increased both in the EI and EF shoots, but the EI shoots showed a greater increase (**Figure 4B**). Endophyte infection had a positive effect on the total phenolic content in the root. Clipping enhanced the total phenolic content in the root both in the EI and EF plants; similar to the total phenolic content in the shoots, the EI plants showed a greater increase than the EF plants after clipping (**Figure 4C**).

## Root Morphology and Regrowth Rate

The root morphology, such as total length, total surface area and average diameter, was influenced by clipping. The total length and average diameter of the roots were also influenced by the interaction of endophyte infection and clipping (**Table 1**). The total length and average diameter did not differ greatly in the unclipped control groups. Clipping decreased the total length and average diameter of the roots in both the EI and EF plants, but those of the EI plants decreased much more than EF plants (**Figure 5**).

To assess the effect of endophyte infection on the regrowth ability of A. sibiricum after clipping, we clipped the plants and allowed them to regrow for 3 weeks. We found that the regrowth rate of the EI plants was significantly slower than that of the EF plants (**Figure 6**).

## DISCUSSION

Endophyte-conferred herbivore resistance is often attributed to the alkaloids produced by the endophyte (Clay and Schardl, 2002; Sullivan et al., 2007; Schardl et al., 2012; Saikkonen et al., 2013). For example, EI drunken horse grass can produce a certain concentration of ergonovine and ergine (Li et al., 2007), and endophyte infection could increase the resistance of the host plants to Rhopalosiphum padi and Messor aciculatus. In Arizona fescue, only a low concentration of peramine was produced, and infected plants did not show significant resistance to Xanthippus corallipes, Melanoplus femurrubrum, and Acromyrmex versicolor (Saikkonen et al., 1999; Tibbets and

TABLE 2 | Two-way ANOVA for rotated factors of individual amino acids of Achnatherum sibiricum under endophyte infection and the clipping treatment.


Significant P-values (P < 0.05) are in bold.

fmicb-07-01988 December 14, 2016 Time: 11:33 # 7

Faeth, 1999). Shymanovich et al. (2015) observed different aphid resistance of sleepy grass in different populations, with very low survival rates for the aphids feeding on plants infected with the Cloudcroft endophyte, whereas aphid survival was better on EI plants in the Weed population. In addition, they suggested that the alkaloid ergonovine was responsible for aphid mortality. In our experiments, alkaloids in neither infected nor uninfected plants were detected when grown under normal conditions. After clipping, only peramine was detected in a small number of samples in the leaf sheath of infected plants', but the concentration was extremely low. However, we found that infected A. sibiricum had significant resistance to L. migratoria, which suggests that apart from endophyte-conferred alkaloid defense, additional mechanisms such as endophyte-mediated changes in host defense are likely to be implicated in endophytehost-insect interactions. Similar results have been reported recently by Ueno et al. (2016), who found that ozone level did not affect alkaloid concentration but rather significantly affected aphid resistance of infected Lolium multiflorum.

Primary metabolites such as carbohydrates and amino acids are two important macronutrients that influence animal survival, growth and reproduction (Karasov and Martiìnez del Rio, 2007; Simpson and Raubenheimer, 2012; Roeder and Behmer, 2014). The mouthparts of locusts contain large numbers of sensilla groups with neurons sensitive to a range of chemicals, including amino acids and carbohydrates (Chapman, 2003). In the present study, we found that endophyte infection significantly decreased the soluble sugar content and amino acid content of the host plants, suggesting that endophyte infection may increase locust resistance by lowering the palatability of host plants.

In addition to alkaloids, other secondary metabolites such as phenols have been proposed as antifeedants or digestibility reducers (Ballare et al., 1996; Izaguirre et al., 2007). In the present study, we found that endophyte infection significantly increased the total phenolic content in both the shoots and roots of the host plants. Similar results have been reported in perennial ryegrass (Rasmussen et al., 2008; Panka et al., 2013 ´ ) and tall fescue (Malinowski et al., 1998). Here, the higher phenol concentration of the EI leaves may have contributed to the higher locust resistance of the host plants. Combined with the reduction of primary metabolites such as carbohydrates and amino acids, our results suggest that endophyte infection triggers reprogramming of the host metabolism, favoring secondary metabolism at a cost to primary metabolism (Dupont et al., 2015).

Plants can alter their metabolism in response to environmental conditions, such as soil nutrients, water levels and feeding by insects or mechanical damage (Bultman et al., 2004; Behmer, 2009; Machado et al., 2015). Herbivore attacks, for instance, alter nitrogen and carbon dynamics (Arnold and Schultz, 2002; Appel et al., 2012), which often results in dramatic changes in primary and secondary metabolite pools (Skrzypek et al., 2005; Steinbrenner et al., 2011; Machado et al., 2013). In particular, endophyte infection can affect the metabolism response of the host plants. For example, Bultman and Bell (2003) found that the total protein N in uninfected tall fescue significantly increased after clipping, but this was not true for infected plants. Sullivan et al. (2007) found that after mechanical and herbivore damage, uninfected tall fescue also had a significantly higher foliar N% and lower C: N ratio compared with infected hosts. In our study, we found that after clipping, the concentration of soluble sugar and Ile significantly decreased

in infected plants, whereas the concentration of total phenolic significantly increased. For uninfected plants, however, clipping caused an elevation of soluble sugar and amino acid such as Thr, Lys, His, and Val. Clipping also resulted in an increase of the total phenolic content in uninfected plants, but the degree was lower than in infected plants. The clipping treatment made uninfected plants more sensitive, whereas infected plants became more resistant to the locust. The weight of the second instar locusts fed on uninfected plants increased significantly after clipping, whereas it decreased when fed on infected plants. Thus, advantage of endophyte infection in resistance to locusts was more obvious after clipping.

In addition to resistance, endophyte infection might influence the tolerance of host plants (Belesky and Fedders, 1996; Cheplick, 1998; Bultman et al., 2004). Tolerance is the degree to which a plant can regrow and reproduce following damage (Strauss and Agrawal, 1999). Hence, it is of interest to ask how endophytes influence the ability of a plant to tolerate herbivory. In tall fescue, Bultman et al. (2004) found that clipping induces resistance in infected plants at the cost of tolerance. In our study, we also found that the regrowth rate of infected plants was significantly lower than that of uninfected plants after clipping.

Endophyte infection significantly affects both primary and secondary metabolism of its host plant (Rasmussen et al., 2008). It is possible that some of the herbivore resistance effects observed in infected plants are due to the metabolites measured in this study or other, still completely unknown, endophyte-specific compounds. Therefore, wider metabolic studies are needed to understand herbivore resistance of this association.

### CONCLUSION

Under normal conditions, endophyte infection did not induce detectable alkaloid production, but endophyte infection did significantly enhance the resistance of the host to L. migratoria. In this study, the lower content of soluble sugars and amino acids and higher total phenolic content may contribute to higher locust resistance of the host plants. Endophyte infection can mediate the strategies of A. sibiricum response to clipping. After clipping, the infected plants exhibited decreased nutrient content, increased defense substances, and thus increased resistance to L. migratoria at the cost of regrowth. For uninfected plants, however, clipping caused an increase in nutrient substances and regrowth rate, but uninfected plants were more susceptible to L. migratoria.

### AUTHOR CONTRIBUTIONS

Conceived and designed the experiments: AR; performed the experiments: JQ, YG, HL, YZ; analyzed the data: JQ, AR; contributed reagents/materials/analysis tools: YG. Wrote the paper: AR.

### ACKNOWLEDGMENTS

This research was funded by National Key Research and Development Program (2016YFC0500702), National Natural Science Foundation (31270463 and 31570433) and Doctoral Program Foundation of Institutions of Higher Education (20130031110023) of China. We greatly appreciate the support of National Hulunber Grassland Ecosystem Observation and Research Station for their invaluable assistance on this experiment.

### REFERENCES

fmicb-07-01988 December 14, 2016 Time: 11:33 # 9


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2016 Qin, Gao, Liu, Zhou, Ren and Gao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fmicb-07-01988 December 14, 2016 Time: 11:33 # 10