# METALLIC MICRONUTRIENT HOMEOSTASIS IN PLANTS

EDITED BY : Manuel González-Guerrero, Hannetz Roschzttardtz and Diego Fabian Gomez-Casati PUBLISHED IN : Frontiers in Plant Science

#### Frontiers eBook Copyright Statement

The copyright in the text of individual articles in this eBook is the property of their respective authors or their respective institutions or funders. The copyright in graphics and images within each article may be subject to copyright of other parties. In both cases this is subject to a license granted to Frontiers. The compilation of articles constituting this eBook is the property of Frontiers.

Each article within this eBook, and the eBook itself, are published under the most recent version of the Creative Commons CC-BY licence. The version current at the date of publication of this eBook is CC-BY 4.0. If the CC-BY licence is updated, the licence granted by Frontiers is automatically updated to the new version.

When exercising any right under the CC-BY licence, Frontiers must be attributed as the original publisher of the article or eBook, as applicable.

Authors have the responsibility of ensuring that any graphics or other materials which are the property of others may be included in the CC-BY licence, but this should be checked before relying on the CC-BY licence to reproduce those materials. Any copyright notices relating to those materials must be complied with.

Copyright and source acknowledgement notices may not be removed and must be displayed in any copy, derivative work or partial copy which includes the elements in question.

All copyright, and all rights therein, are protected by national and international copyright laws. The above represents a summary only. For further information please read Frontiers' Conditions for Website Use and Copyright Statement, and the applicable CC-BY licence.

ISSN 1664-8714 ISBN 978-2-88963-235-0 DOI 10.3389/978-2-88963-235-0

#### 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

# METALLIC MICRONUTRIENT HOMEOSTASIS IN PLANTS

Topic Editors:

Manuel González-Guerrero, Polytechnic University of Madrid, Spain Hannetz Roschzttardtz, Pontificia Universidad Católica, Chile Diego Fabian Gomez-Casati, National University of Rosario, Argentina

Citation: González-Guerrero, M., Roschzttardtz, H., Gomez-Casati, D. F., eds. (2019). Metallic Micronutrient Homeostasis in Plants. Lausanne: Frontiers Media SA. doi: 10.3389/978-2-88963-235-0

# Table of Contents


Miguel A. Ibeas, Susana Grant-Grant, Nathalia Navarro, M. F. Perez and Hannetz Roschzttardtz


Kareem A. Mosa, Mohamed El-Naggar, Kalidoss Ramamoorthy, Hussain Alawadhi, Attiat Elnaggar, Sylvie Wartanian, Emy Ibrahim and Hala Hani


María J. García, Francisco J. Corpas, Carlos Lucena, Esteban Alcántara, Rafael Pérez-Vicente, Ángel M. Zamarreño, Eva Bacaicoa, José M. García-Mina, Petra Bauer and Francisco J. Romera


Miguel Angel Ibeas, Susana Grant-Grant, Maria Fernanda Coronas, Joaquín Ignacio Vargas-Pérez, Nathalia Navarro, Isidro Abreu, Hiram Castillo-Michel, Natalia Avalos-Cembrano, Julio Paez Valencia, Fernanda Perez, Manuel González-Guerrero and Hannetz Roschzttardtz

# Editorial: Metallic Micronutrient Homeostasis in Plants

Hannetz Roschzttardtz <sup>1</sup> \*, Manuel González-Guerrero<sup>2</sup> \* and Diego F. Gomez-Casati <sup>3</sup> \*

<sup>1</sup> Facultad de Ciencias Biológicas, Pontificia Universidad Católica de Chile, Santiago, Chile, <sup>2</sup> Centro de Biotecnología y Genómica de Plantas (UPM-INIA), Universidad Politécnica de Madrid, Madrid, Spain, <sup>3</sup> Centro de Estudios Fotosintéticos y Bioquímicos (CEFOBI-CONICET), Universidad Nacional de Rosario, Rosario, Argentina

Keywords: metals, plant nutrition, metal accumulation, frataxin, nodule, medicago

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

#### **Metallic Micronutrient Homeostasis in Plants**

Transition elements, such as copper, iron, or zinc, are essential nutrients for plants. They participate in every biological process. However, intracellular metal levels must be maintained within a narrow physiological concentration. Too little, and not enough cofactors are available to the cell; too much, and Fenton-type reactions and mismetallation events will disrupt many cellular processes. As a result, plants have developed complex systems to control metal uptake and to deliver them to all tissues and cells. In this Research Topic, we have collected some of the most recent work furthering our understanding of Metallic Micronutrient Homeostasis in Plants.

Plant dependence on metals contrasts to the common low bioavailability of these nutrients in many soil types. Growing evidence indicates that plants have optimized the use of metals facilitating their relocation from one cellular compartment to another, depending on their need in different physiological processes. In this Topic, Zhang and Krämer have reanalyzed available bioinformatic data to show that there is a diurnal regulation of the translation of proteins involved in Fe-S metabolism in Arabidopsis thaliana, including that of frataxin, a protein involved in iron transfer in Fe-S protein biogenesis, and also in copper control in mitochondria and chloroplasts, as reviewed in this issue by Gomez-Casati et al. Iron is not the only metal micronutrient that is regulated by day-night cycles. Copper homeostasis is also controlled in this way, the likely consequence of plastocyanin synthesis for photosynthesis, and the toxic effect of copper in Fe-S synthesis. Andrés-Colás et al. showed that copper transporter COPT3 transcription is controlled in this way, in a process that is dependent of copper availability to the plant and on transcription factor TCP16.

The study of metal (re)distribution in plants can be greatly assisted by metal-imaging methods. Using X-ray fluorescence approaches, Vigani et al. showed how iron deficiency affects zinc distribution in leaves. With a similar approach, Ibeas et al. have determined the iron distribution in Chenopodium quinoa seeds and used it to validate the Perls-DAB histochemical method to visualize iron. This approach not only led to illustrate how iron distribution in seeds changed during dicot evolution but showed how subcellular iron distribution changes during embryo development (Ibeas et al.). Metal imaging approaches are also important to evaluate how metal nanoparticles could enter into trophic chains through plants, and the physiological effects that might cause there (Mosa et al.).

The evolving metal distribution in a cell and an organism could also reflect how some metalloenzymes are expressed with different demand for metal cofactors. For this, we would need high throughput metalloproteomics methods to identify the precise nature of the accepting metalloprotein to put into context many of the observed phenotypes. For instance, symbiotic nitrogen fixation requires one or several zinc-proteins that although dependent on zinc transfer into the endoplasmic reticulum of nodule cells, exert their effect in nitrogen-fixing symbiosomes and in nodule development (León-Mediavilla et al.).

#### Edited and reviewed by:

Felipe Klein Ricachenevsky, Universidade Federal de Santa Maria, Brazil

#### \*Correspondence:

Hannetz Roschzttardtz hroschzttardtz@bio.puc.cl Manuel González-Guerrero manuel.gonzalez@upm.es Diego F. Gomez-Casati gomezcasati@cefobi-conicet.gov.ar

#### Specialty section:

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

Received: 13 June 2019 Accepted: 02 July 2019 Published: 17 July 2019

#### Citation:

Roschzttardtz H, González-Guerrero M and Gomez-Casati DF (2019) Editorial: Metallic Micronutrient Homeostasis in Plants. Front. Plant Sci. 10:927. doi: 10.3389/fpls.2019.00927

**4**

Hormones have been known to control metal homeostasis, but the precise mechanisms of how hormone levels are controlled by metals still remains elusive. In this topic, García et al. reported how long-distance iron trafficking mediated by OPT3 could control ethylene metabolism as well as GSNO levels. In addition, micronutrients such as silicon, modulate hormone levels to overcome abiotic stress such as chilling and also improve the nutrition by metals (particularly zinc and manganese) (Moradtalab et al.).

In summary, the work shown in this Research Topic illustrates our advances in defining how cells and organisms control and redistribute essential metal nutrients. This distribution is affected by biotic and abiotic interactions, hormones and responds to different physiological states of plant development. The work also illustrates the need to have new methods to determine and visualize metals, as well-new developments in metalloproteomics and in metal-speciation analyses.

### AUTHOR CONTRIBUTIONS

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

**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 © 2019 Roschzttardtz, González-Guerrero and Gomez-Casati. 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) and the copyright owner(s) 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.

# Dynamic Subcellular Localization of Iron during Embryo Development in Brassicaceae Seeds

Miguel A. Ibeas<sup>1</sup> , Susana Grant-Grant<sup>1</sup> , Nathalia Navarro<sup>1</sup> , M. F. Perez<sup>2</sup> and Hannetz Roschzttardtz<sup>1</sup> \*

<sup>1</sup> Departamento de Genética Molecular y Microbiología, Pontificia Universidad Católica de Chile, Santiago, Chile, <sup>2</sup> Departamento de Ecología, Pontificia Universidad Católica de Chile, Santiago, Chile

Iron is an essential micronutrient for plants. Little is know about how iron is loaded in embryo during seed development. In this article we used Perls/DAB staining in order to reveal iron localization at the cellular and subcellular levels in different Brassicaceae seed species. In dry seeds of Brassica napus, Nasturtium officinale, Lepidium sativum, Camelina sativa, and Brassica oleracea iron localizes in vacuoles of cells surrounding provasculature in cotyledons and hypocotyl. Using B. napus and N. officinale as model plants we determined where iron localizes during seed development. Our results indicate that iron is not detectable by Perls/DAB staining in heart stage embryo cells. Interestingly, at torpedo development stage iron localizes in nuclei of different cells type, including integument, free cell endosperm and almost all embryo cells. Later, iron is detected in cytoplasmic structures in different embryo cell types. Our results indicate that iron accumulates in nuclei in specific stages of embryo maturation before to be localized in vacuoles of cells surrounding provasculature in mature seeds.

Keywords: embryo development, iron, Brassicaceae, nucleus, perls/DAB

### INTRODUCTION

Nutrient reserves in the seed must be sufficient to sustain plant establishment until the root system has developed enough to provide nutrients from the soils. High nutrient content of seeds is particularly important for plants growing in unfavorable nutritional conditions and has been related to higher seed viability and seedling vigor. Besides its impact on plant growth, nutrient levels in seeds are an important consideration for human and/or livestock seed-based nutrition (Roschzttardtz et al., 2017).

Iron is an essential micronutrient for plant growth and development. Despite its importance, the prevalent low iron bioavailability in the soils of main agricultural areas of the world limits plant productivity, fertility, and germination rates (Guerinot and Yi, 1994). As a consequence, iron contents in seeds is diminished which results in negative impacts in human and animal health, since seeds are a main source of food for humans and animals (Fan et al., 2008; DeFries et al., 2015). In humans, iron deficiency in women and children under 2 years is a serious and growing public health problem and a major concern for the World Health Organization. Therefore, understanding seed iron distribution and storage at the physiological and molecular level is key to design biotechnological applications to improve iron content of staple seeds.

Different approaches have been used in order to study where iron localizes in plant organs and at subcellular level. In plant, iron accumulates in different compartments depending of the tissue and

#### Edited by:

Felipe Klein Ricachenevsky, Universidade Federal de Santa Maria, Brazil

#### Reviewed by:

Ricardo Fabiano Hettwer Giehl, Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung (IPK), Germany Sebastien Thomine, Centre National de la Recherche Scientifique (CNRS), France

#### \*Correspondence:

Hannetz Roschzttardtz hroschzttardtz@bio.puc.cl

#### Specialty section:

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

Received: 04 September 2017 Accepted: 12 December 2017 Published: 22 December 2017

#### Citation:

Ibeas MA, Grant-Grant S, Navarro N, Perez MF and Roschzttardtz H (2017) Dynamic Subcellular Localization of Iron during Embryo Development in Brassicaceae Seeds. Front. Plant Sci. 8:2186. doi: 10.3389/fpls.2017.02186

**6**

cell type. Among these are apoplast in root vasculature, plastids in leaves and pollen grain, vacuoles in embryos, and nuclei in different cell types (Roschzttardtz et al., 2009, 2011, 2013; Lanquar et al., 2010).

Regarding seeds, it has been described that iron accumulates in embryo during maturation stage of seed development (Roschzttardtz et al., 2009). So far, VACUOLAR IRON TRANSPORTER1 (VIT1) is the only player described involved in the iron loading into the vacuoles of endodermis cell layer in Arabidopsis embryos. Mutants in VIT1 mislocalize iron, which accumulates in vacuoles of cortex cells in hypocotyl and subepidermal cells of abaxial side of cotyledons (Kim et al., 2006; Roschzttardtz et al., 2009; Eroglu et al., 2017). Interestingly, vit1 mutants do not grow correctly in iron deficient conditions, suggesting that iron localization in seed has an important role in seed physiology. In the absence of functional VIT1 or during germination, the manganese and iron transporter MTP8 is responsible for loading iron in the vacuoles of subepidermal cells of cotyledons (Eroglu et al., 2017).

Little is known about the mechanism involved in iron loading in embryo during seed development. Recently Grillet et al. (2014) described that ascorbate efflux from Pea embryos is required for iron loading in seeds.

In this article we used, as model plants, species belonging to the Brassicaceae family in order to describe the subcellular compartments where iron accumulates in embryo during seed development. One of our models, Brassica napus L., is one of the most important oil crops almost all over the world. The tetraploid B. napus is a hybrid derived from Brassica oleracea and Brassica rapa and it shares more than 86% similarity in protein coding sequence with Arabidopsis thaliana (Cavell et al., 1998; Parkin et al., 2005). Furthermore, rapseed is not only important for vegetal oil, but also a major source for industrial materials such as biofuel (Chen et al., 2017).

Our results show that iron is accumulated first in nuclei of integument cells, endosperm and embryo cells in the early maturation stage of seed development before being located in vacuoles of endodermis cells. Some differences in iron localization were also observed in mature embryos of Brassica napus, Nasturtium officinale, Lepidium sativum, Brassica oleracea and Camelina sativa compared with Arabidopsis thaliana.

### MATERIALS AND METHODS

#### Plant Material and Growth Conditions

Brassica napus, Nasturtium officinale, Camelina sativa, Lepidium sativum, and Brassica oleracea seeds were purchased at a local market. B. napus and N. officinale plants were grown on soil in a greenhouse at 23◦C under long-day conditions (16-h/8-h day/night cycle).

### Histochemical Staining of Iron with the Perls/DAB Procedure

Iron staining was performed according to Roschzttardtz et al. (2009). Brassica napus, Nasturtium officinale, Camelina sativa, Brassica oleracea, and Lepidium sativum dry seeds or seeds from B. Napus and N. officinale at different developmental stage were vacuum infiltrated with fixation solution (2% w/v paraformaldehyde in 1 mM phosphate buffer pH 7.0) for 45 min and incubated for 16 h in the same solution. The fixated seeds were dehydrated with ascendant concentration of ethanol (50, 70, 80, 90, 95, and 100%), later were incubated 12 h with a solution of butanol/ethanol 1:1 (v/v), and finally were incubated 12 h with 100% butanol. Then, the seeds were embedded in the Technovit 7100 resin (Kulzer) according to the manufacturer's instructions and thin sections (3 µm) were cut and were deposited on glass slides. For Perls/DAB staining, slides were incubated with 2% (v/v) HCl and 2% (w/v) K-ferrocyanide (Sigma–Aldrich) for 45 min For the DAB intensification, each glass slide was washed with distilled water, later were incubated in a methanol solution containing 0.01 M NaN<sup>3</sup> (Sigma–Aldrich) and 0.3% (v/v) H2O<sup>2</sup> (Merck) for 1 h, and then washed with 0.1 M phosphate buffer (pH 7.4). For the intensification reaction, the glass slides were incubated between 10 and 30 min in a 0.1 M phosphate buffer (pH 7.4) solution containing 0.025% (v/v) H2O2, and 0.005% (w/v) CoCl<sup>2</sup> ∗ 6H2O (intensification solution). To stop the reaction each slide was rinsed with distilled water.

#### Iron Quantification

A microwave-assisted acid digestion was performed using 6.0 mL of concentrated HNO<sup>3</sup> (Winkler) and 1.0 mL of 30% (v/v) H2O<sup>2</sup> added over 10–20 mg of seeds of each genotype. The colorless digestate were filled up to 15 mL using deionized water. The iron content of each genotype was measured in triplicate using inductively coupled plasma mass spectrometry (ICP-MS).

### Microscopy and Staining with DAPI, Perls/DAB and Toluidine Blue

Thin sections were first stained with DAPI in order to visualize nuclei (Roschzttardtz et al., 2013), the same section was stained by Perls/DAB, as described above. Then were stained with toluidine blue. Each slide was incubated in a 0.5% (w/v) solution of toluidine blue for 2 min and rinsed with distilled water. The samples were observed and photographed with a Nikon Eclipse 80i microscope.

### RESULTS

### Iron Concentration and Distribution in Dry Seeds of Close Related Species to Arabidopsis thaliana

To evaluate iron concentration and distribution in dry seeds from closely related Arabidopsis species, first we determine iron content in dry seeds from Arabidopsis thaliana, Brassica napus and Nasturtium officinale (**Figure 1A**). Dry seeds were isolated, and a microwave-assisted acid digestion was performed. **Figure 1B** show iron concentration of each species determined by inductively coupled plasma mass spectrometry (ICP-MS). Iron concentration in Arabidopsis thaliana and Brassica napus dry seeds (70.4 and 64.6 µg of iron/g of seeds, respectively) was lower compared to Nasturtium officinale (116.6 µg of iron/g of seeds),

suggesting that there is no relation between iron concentration and seed size, as Brassica napus have the biggest seeds (**Figure 1A**) but also have a similar iron content as Arabidopsis thaliana.

In order to determine where iron localizes in dry seed embryos from species closely related to Arabidopsis thaliana, we analyzed seven different plant species: Brassica napus, Nasturtium officinale, Camelina sativa, Arabidopsis thaliana, Lepidium sativum, Brassica oleracea and Brassica oleracea var. capitate (**Figure 2** and Supplementary Figure S2). Embryos from dry seeds were isolated and fixed. Thin sections of different embryo regions were analyzed, in particular, cotyledon and hypocotyl. **Figure 2** and Supplementary Figure S2 show Perls/DAB staining revealing iron distribution in embryos from these species. Cotyledons accumulate iron in vacuoles of cells surrounding provasculature (**Figures 2A–D** and Supplementary Figures S2A,B) as has been described before for Arabidopsis thaliana (Roschzttardtz et al., 2009). Provasculature complexity in cotyledon is clearly higher than Arabidopsis (Roschzttardtz et al., 2014, 2017). Iron accumulation was not observed in other cell types like protodermis and cortex cells in dry seed embryos (**Figures 2A–D** and Supplementary Figures S2A,B).

Interestingly, transversal sections of hypocotyl show differences in the number of cell layer where iron accumulates compared to Arabidopsis hypocotyl, where only one cell layer accumulates iron (Roschzttardtz et al., 2009). We found that at least two-cell layers accumulate iron in the hypocotyl of Brassica napus, Nasturtium officinale, Camelina sativa, Lepidium sativum and Brassica oleracea dry seed embryos (**Figures 2E–G** and Supplementary Figure S2C,D, respectively). Despite the differences in the number of cell layers where iron accumulates in dry seeds of Brassica napus, Nasturtium officinale, Camelina sativa, Lepidium sativum and Brassica oleracea compared to Arabidopsis thaliana, at subcellular level iron is stored in the vacuoles in cells surrounding provasculature (**Figures 2A–D** and Supplementary Figures S2A,B).

Analyses of embryonic root tips indicate that iron accumulates in different cells identified as endodermis-cortex cells (Supplementary Figure S3).

Supplementary Figure S1 shows the phylogenetic tree of the 5 species of Brassicaceae family used in this study. Interestingly, iron localization in vacuoles is a conserved trait in the analyzed plant species. As indicated above, differences were only observed in the number of cell layers that accumulate iron in the hypocotyls.

#### Iron Distribution during Brassica napus Seed Development

As we showed, iron accumulates in vacuoles in embryos of seven species of Brassicaceae (**Figure 2** and Supplementary Figure S2). We used Brassica napus plants as model to study where iron localizes during seed development. B. napus produces bigger seeds compared with Arabidopsis plants allowing to study with more details different seed regions during its development (**Figure 1A**). In a phylogenetic point of view, B. napus is the more distant specie to A. thaliana used in this study. The fact that iron localizes in vacuoles of mature B. napus seeds is a strong indication that it is a conserved trait. Using A. thaliana as plant model it has been described that iron accumulates in embryo during seed maturation stages (Roschzttardtz et al., 2009), for this reason we analyzed three different stages of seed maturation, torpedo, bent cotyledon and mature stage before seed desiccation. In order to describe seed structures where iron accumulates before to be loaded into the embryo, we used in our analysis whole seeds including seed coat, endosperm and embryo. Whole seeds of B. napus containing embryos in different maturation stages as indicated above, were fixed and embedded in Technovit resin according to Section "Material and Methods."

The analysis of seeds with torpedo embryo stages revealed that iron localizes mainly in nuclei of integument cells, free cell endosperm and in different cell layers in the embryo (**Figure 3**). At this embryo stage, Perls/DAB and Toluidine blue staining show that in protodermal, cortex, endodermis and provasculature cells iron is localized in nuclei (**Figures 3A,B**). As vacuoles look empty after both staining, our results suggest that vacuoles of embryo cells are not an important compartment for iron storage at this developmental stage. Interestingly, strong iron staining was observed in nuclei (one to three iron pools were observed) of the free cell endosperm surrounding the embryo (**Figures 3C,D**).

Later, in seeds containing bent cotyledon embryo stage, integuments no longer contain iron inside the nuclei (**Figures 4A,B**). In the embryo, the number of cells containing

FIGURE 2 | Iron distribution in Brassica napus, Nasturtium officinale, Camelina sativa, and Arabidopsis thaliana dry seed embryos. Embryos were dissected from dry seeds and thin sections were obtained and stained with Perls/DAB in order to reveal iron accumulation. Brassica napus sections are shown in (A,E); Nasturtium officinale sections are shown in (B,F); Camelina sativa sections are shown in (C,G) and Arabidopsis thaliana sections are shown in (D,H). Cotyledons and hypocotyl are shown in (A–D) and (E–H) respectively. In panel A to D a zoom is shown in order to indicate iron accumulation into the vacuoles. Bars in panels correspond to 100 µm.

iron in nuclei decreases. In embryonic root tips and cotyledons, iron is localized in structures surrounding the nuclei (**Figures 4A,B,E,F**), while vacuoles are starting to load iron in the hypocotyl endodermis cell layer (**Figures 4C,D**).

In mature seeds before desiccation, iron does not localize in nuclei (**Figure 5**) and vacuoles are the principal compartments for iron storage (**Figures 5B,D**). At this late stage of seed development, cells adjacent to provasculature load iron and protodermis and cortex cells have not detectable iron as the previous seed developmental stages (**Figure 5**).

In order to determine whether iron is accumulated in nuclei before torpedo stages, we use N. officinale developing

embryos. Our results indicate that in the heart stage of embryo development iron is not detectable by Perls/DAB staining in embryo cells. This result strongly supports the conclusion that iron localizes in nuclei of embryo cells in the torpedo stage (Supplementary Figure S5). Similar stages of seed maturation were analyzed using N. officinale. No differences in iron distribution were observed compared to B. napus (Supplementary Figure S4).

#### Iron Distribution at Subcellular Level during Seed Maturation

In order to confirm the subcellular compartments accumulating iron during development we performed consecutively staining with DAPI, Perls/DAB and toluidine blue on the same sections of different developmental stages of B. napus embryos. Iron accumulates in nuclei of different cell types in early maturation stages of seed development (**Figures 6A–E**). In bent cotyledon stage, iron is not longer detected in nuclei and accumulates in cytoplasmic structures (**Figures 6K–O**). Finally in mature stages, vacuoles of cells surrounding provasculature start to accumulate iron (**Figures 6P–T**). Details of unique cells are shown in **Figure 6**, indicating unambiguously were iron localizes in different cases, nuclei, cytoplasmic structures or vacuoles.

### DISCUSSION

During seed development embryo accumulates nutrients that will be used in the transition of heterotrophy to autotrophy metabolism during germination and post germination stage of development. In the case of iron, the principal transporter expressed in root involved in iron acquisition in Arabidopsis IRT1 is expressed 3–4 days after germination (Lanquar et al., 2010). Before roots are able to acquire iron from the soil, iron that was stored during seed development is remobilized from

FIGURE 5 | Iron distribution in Brassica napus seeds in mature embryo stage before seed desiccation. Brassica napus seeds at the mature embryo before seed desiccation stage dissected from siliques were embedded in Technovit resin and sectioned (3 µm) and stained with Perls/DAB (A–F) and then with Toluidine blue (B,D,F). Cotyledons, hypocotyl and endosperm-seed coat are shown in (A–F), respectively. In (B,D) a zoom are shown in order to indicate iron accumulation in the vacuoles. The scale bar represents 100 µm.

vacuoles of endodermis and bundle sheath cells (Roschzttardtz et al., 2009; Lanquar et al., 2010; Mary et al., 2015). When iron is not correctly localized as in vit1 mutants or not remobilized as in nramp3 nramp4 double mutant, germinated seeds have difficulties to growth in iron deficiency conditions (Kim et al., 2006; Lanquar et al., 2010; Mary et al., 2015). This indicates that iron storage and localization in embryos is pivotal for seed physiology (Kim et al., 2006).

In this article, different stages of seed maturation were analyzed by a histochemical approach (Perls/DAB staining) in order to reveal iron localization. Perls/DAB method is specific and very sensitive for iron detection and can detect both Fe2<sup>+</sup> and Fe3<sup>+</sup> forms (Roschzttardtz et al., 2009). Other methods more quantitative have been used to reveal iron localization in seeds, indicating that an Arabidopsis embryo contains a fraction of the total iron of a seed (Schnell Ramos et al., 2013). Despite of not being a quantitative method, Perls/DAB staining allows to obtain information at subcellular level of any plant tissues and stages of plant development (Roschzttardtz et al., 2013).

Our analysis showed that iron was not localized in an unique subcellular compartment during seed development. In early maturation stages (embryo in torpedo stage) iron was localized principally in nuclei of integument, endosperm and embryo cells. This result is very surprising, indicating that vacuoles are not the only relevant subcellular compartment for iron storage during seed development (**Figures 3**, **4**). Interestingly, Perls/DAB did not detect iron in embryo at heart stage, suggesting strongly that iron is accumulated in nuclei not before torpedo stages (Supplementary Figure S5). Later iron seems to be gradually concentrated in structures surrounding nuclei (possibly mitochondria or endoplasmic reticulum) before to be loaded into the vacuoles of endodermis cells (**Figure 5**). The micronutrient manganese has been detected in endoplasmic reticulum during seed development, but more studies may be performed in order to identify the nature of the structures where iron is detected before to be mobilized to the vacuoles (Otegui et al., 2002). A model of iron dynamic during embryo loading is shown in **Figure 7**. The presence of iron in nuclei of pea embryos was

staining. The subsequent staining was performed over the same sections. The scale bar represents 20 µm.

previously reported in Roschzttardtz et al. (2011), however, in this present study we shown that iron localizes in nuclei only during torpedo stages in Brassicaceae seeds. These results open interesting questions about the role of nuclei at this embryo stage of development. Our results suggest that nuclei may be a reservoir of iron during seed development or that iron may have unknown functions in nuclei of embryos in torpedo stage. Very recently, a specific role of iron in promoting meristematic cell division in adventitious root formation has been proposed (Hilo et al., 2017), however, the molecular role of iron in nuclei remains unknown.

Regarding iron distribution, we found that species closely related to Arabidopsis thaliana have some differences in dry seed embryos. For instance, iron is accumulated into the vacuoles of two-cell layer in mature embryos of Lepidium sativum showing a clear difference with Arabidopsis mature embryos, where only one cell layer accumulates iron (**Figure 2**; Roschzttardtz et al., 2009). A study reporting iron localization in dry seeds of 13 genoypes of three Phaseolus species has shown that large concentrations of iron are accumulated in the cytoplasm of cells surrounding the provascular bundles (Cvitanich et al., 2010). In the case of monocots, the highest iron concentration occurs

in scutellum and outer regions of the embryo while a very low Fe is detected in the outermost layers of the endosperm and the single-layered aleurone that surrounds the endosperm

Pt, protodermis. N, nuclei; V, vacuole.

(Johnson et al., 2011). This suggests that different mechanisms of iron acquisition may play important roles in different seed species. In our group, we are interested in elucidating the molecular mechanisms that may explain these differences.

#### AUTHOR CONTRIBUTIONS

fpls-08-02186 December 22, 2017 Time: 9:41 # 8

MI, SG-G, NN, and HR performed experiments and designed the research. MP was involved in the phylogenetic tree analysis. All authors participated in manuscript preparation.

#### FUNDING

This work was funded by FONDECYT 1160334 (Chilean Government) and INTER 6809 (Pontificia Universidad Católica

### REFERENCES


de Chile-VRI) to HR, and Millennium Nucleus Center for Plant Systems and Synthetic Biology (NC130030). Ph.D. students work was supported by Conicyt-Chile grants 21160350 (to MI) and 21170951 (to SG-G).

#### ACKNOWLEDGMENT

The authors are greatly indebted to Xavier Jordana and María Isabel Gómez for their continued support and encouragement.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2017.02186/ full#supplementary-material

manganese deficiency. Plant Physiol. 152, 1986–1999. doi: 10.1104/pp.109. 150946


**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 Ibeas, Grant-Grant, Navarro, Perez and Roschzttardtz. 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.

# Silicon Improves Chilling Tolerance During Early Growth of Maize by Effects on Micronutrient Homeostasis and Hormonal Balances

Narges Moradtalab<sup>1</sup> \*, Markus Weinmann<sup>1</sup> , Frank Walker <sup>2</sup> , Birgit Höglinger <sup>2</sup> , Uwe Ludewig<sup>1</sup> and Guenter Neumann<sup>1</sup>

1 Institute of Crop Science (340h), University of Hohenheim, Stuttgart, Germany, <sup>2</sup> Institute of Phytomedicine (360), University of Hohenheim, Stuttgart, Germany

#### Edited by:

Manuel González-Guerrero, Universidad Politécnica de Madrid (UPM), Spain

#### Reviewed by:

Yongchao Liang, Zhejiang University, China Julia Quintana González, Ruhr University Bochum, Germany

\*Correspondence:

Narges Moradtalab n.moradtalab@uni-hohenheim.de

#### Specialty section:

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

Received: 18 December 2017 Accepted: 16 March 2018 Published: 26 April 2018

#### Citation:

Moradtalab N, Weinmann M, Walker F, Höglinger B, Ludewig U and Neumann G (2018) Silicon Improves Chilling Tolerance During Early Growth of Maize by Effects on Micronutrient Homeostasis and Hormonal Balances. Front. Plant Sci. 9:420. doi: 10.3389/fpls.2018.00420 Low soil temperature in spring is a major constraint for the cultivation of tropical and subtropical crops in temperate climates, associated with inhibition of root growth and activity, affecting early growth and frequently plant performance and final yield. This study was initiated to investigate the physiological base of cold-protective effects induced by supplementation with silicon (Si), widely recommended as a stress-protective mineral nutrient. Maize was used as a cold-sensitive model plant, exposed to chilling stress and low root-zone temperature (RZT) during early growth in a lab to field approach. In a pot experiment, 2–weeks exposure of maize seedlings to low RZT of 12–14◦C, induced leaf chlorosis and necrosis, inhibition of shoot and root growth and micronutrient limitation (particularly Zn and Mn). These phenotypes were mitigated by seed treatments with the respective micronutrients, but surprisingly, also by Si application. Both, silicon and micronutrient treatments were associated with increased activity of superoxide dismutase in shoot and roots (as a key enzyme for detoxification of reactive oxygen species, depending on Zn and Mn as cofactors), increased tissue concentrations of phenolics, proline, and antioxidants, but reduced levels of H2O2. These findings suggest that mitigation of oxidative stress is a major effect of Zn, Mn, and Si applied as cold stress protectants. In a soil–free culture system without external nutrient supply, Si significantly reduced large leaching losses of Zn and Mn from germinating seeds exposed to low-temperature stress. Silicon also increased the translocation of micronutrient seed reserves to the growing seedling, especially the Zn shoot translocation. In later stages of seedling development (10 days after sowing), cold stress reduced the root and shoot contents of important hormonal growth regulators (indole acetic acid, gibberellic acid, zeatin). Silicon restored the hormonal balances to a level comparable with non-stressed plants and stimulated the production of hormones involved in stress adaptation (abscisic, salicylic, and jasmonic acids). Beneficial effects of Si seed treatments on seedling establishment and the nutritional status of Zn and Mn were also measured for a field-grown silage maize, exposed to chilling stress by early sowing. This translated into increased final biomass yield.

Keywords: silicon, micronutrients, germination, chilling stress, maize, oxidative stress

#### INTRODUCTION

In the context of global warming, there is an increasing trend for the cultivation of crops with tropical and subtropical origins, such as maize, soybean, Miscanthus or Sorghum also in temperate climates, e.g., in Central Europe. Under these conditions, early sowing is required for efficient use of the comparatively shorter vegetation periods and to escape from detrimental effects of summer drought (Hund et al., 2004). However, low temperature and cold and wet soils during early spring represent major constraints for the cultivation of tropical crops bearing the risk of poor germination, impaired seedling establishment, and reduced nutrient acquisition due to limited root growth and activity. This is frequently associated with poor vegetative plant development, reduced stress resistance and finally reduced crop yield (Duncan and Hesketh, 1968; Muldoon et al., 1984; Imran et al., 2013), although under favorable conditions short cold periods can be tolerated and later compensated until final harvest (Saeidnejad et al., 2012). In plant species, such as maize, with optimum temperatures of 25–30◦C for germination and plant growth, even moderately low soil temperatures <15◦C are already detrimental (Cutforth et al., 1986; Kasper and Bland, 1992).

Due to root growth limitation and slow diffusion rates, plant availability, and acquisition of sparingly soluble nutrients, such as P, NH<sup>+</sup> 4 , K, Fe, Zn, Mn, and Cu, is particularly affected by low soil temperatures (Duncan and Hesketh, 1968; Kramer and Boyer, 1995; Wan et al., 2001). Therefore, placement of these nutrients close to the seedling roots or as seed treatments are among the most widespread practical measures to counteract detrimental effects of low soil temperatures on seedling establishment and early growth of maize (Imran et al., 2013; Bradácová et al., ˇ 2016; Nkebiwe et al., 2016). Meanwhile, starter applications of ammonium phosphates by shallow subsurface placement below the seeds (Nkebiwe et al., 2016) belong to the standard fertilization strategies employed for maize production systems in temperate climates. Micronutrients with stress-protective functions, such as Zn and Mn, are frequently applied as foliar sprays. However, this is not possible in the seedling stage and usually, the formulations are more expensive than soil fertilizers. More cost-effective placement strategies including seed dressings are increasingly employed to promote stress resistance, early growth, and crop establishment (Farooq et al., 2012; Bradácová et al., ˇ 2016). Similar stress-protective effects have been recorded also for amendments with Si (Liang et al., 2015).

When plants are exposed to environmental stress factors, such as chilling, drought or salinity, an imbalance between production and detoxification of reactive oxygen species (ROS) promotes accumulation of ROS, which induces oxidative damage to cellular components (Gong et al., 2005). Many studies have highlighted a role of Si in the suppression of oxidative damage in various plant species under a wide range of stress conditions. Protective effects have been reported against drought stress in wheat (Gong et al., 2005, 2008; Pei et al., 2010), salinity and boron toxicity in tomato, spinach (Al-aghabary et al., 2005; Gunes et al., 2007), cotton (Gossett et al., 1994), and barley (Liang et al., 2003; Inal et al., 2009), or low temperature stress in wheat (Liang et al., 2008) and cucumber (Liu et al., 2009). Mitigation of oxidative stress by Si treatments has been related to increased expression of enzymatic ROS detoxification systems, such as superoxide dismutases, catalases, peroxidases, ascorbate peroxidase, and increased accumulation of antioxidants (phenolics, proline, ascorbic acid), similar to the effects mediated by stress-protective micronutrients, such as Zn, Mn, Fe, and Cu (Cakmak, 2000; Datnoff et al., 2007). However, the links between Si application and induction of the protective mechanisms against oxidative stress are still largely unknown.

Therefore, this study was designed to investigate more in detail the physiological background of Si effects on cold stress mitigation during early growth of maize. We hypothesized that the protective role of silicon is related to the homeostasis of cold stress-protective micronutrients, such as Zn and Mn (Imran et al., 2013; Bradácová et al., ˇ 2016). In a comparative investigation, Zn, Mn, and Si were applied as seed treatments or by starter fertigation to maize plants, subsequently exposed to low temperatures of 12–14◦C on a silty loam soil taken from a field site of maize cultivation and also in a soilfree culture system. Plant growth, symptoms of oxidative leaf damage and the mineral nutritional status in different plant organs were documented in relation with the expression of various physiological stress indicators (production of H2O2, activities of superoxide dismutase and peroxidase, accumulation of antioxidants, phenolics, and proline) and with changes in hormonal balances. Finally, a preliminary field experiment was conducted on the same soil, where low-temperature stress was provoked due to early sowing at the mid of April. This allowed evaluating the expression of micronutrient and Si effects and their impact on final yield under practical conditions.

**Abbreviations:** Zn, zinc; Si, silicon, Mn, manganese, IAA, indole acetic acid; SA, salicylic acid; JA, jasmonic acid; GA, gibberellin; COR, cold response; RZT, root zone temperature; DAS, days after sowing, ROS, reactive oxygen species; SOD, superoxide dismutase; POD, peroxidase.

### MATERIALS AND METHODS

## Plant Cultivation

#### Soil Culture Experiment

Zea mays L cv. Colisee was used as test plant. Soil material (siltyloam, pH 6.9) was derived from the Ap horizon of a maize cultivation field site at the Hohenheim University experimental station Ihinger Hof, Renningen, Germany (Supplementary Table 1). After sieving with 2 mm mesh size, fertilization was performed with Ca(NO3)2, 100 mg N kg−<sup>1</sup> DM; Ca(H2PO4)2, 80 mg P kg−<sup>1</sup> DM; K2SO4, 150 mg K kg−1DM and MgSO4, 50 mg Mg kg−<sup>1</sup> DM. For improvement of the soil structure, the fertilized soil was mixed with quartz sand (ratio 2:1). Plastic pots were filled with 1,800 ml of this soil substrate and inserted into a cooling system, designed to control the root zone temperature of plants. An immersion water bath circulator (Thermomix 1480/Frigomix 1497, Braun, Melsungen, Germany) was connected to the cooling system, equipped with a closed pipe system which was installed into moist peat substrate to circulate the refrigerating fluid through the moist peat layer surrounding the culture vessels (Bradácˇová et al., 2016). The plants were regularly watered to 70% of substrate water holding capacity (WHC) with deionized water and cultivated for 2 weeks at a root zone temperature of 20–22◦C, 2 weeks at low root zone temperature (12–14◦C), followed by a 2 weeks' recovery phase at 20–22◦C.

Silicon was supplied as silicic acid (H4SiO4) prepared by passing K2SiO<sup>3</sup> through a column filled with a cation–exchange resin (Amberlite IR−120, H<sup>+</sup> form, Sigma Aldrich, Germany) according to Maksimovic et al. (2007) and Pavlovic et al. (2013). Silicon fertigation was performed at a dosage of 40 mg H4SiO<sup>4</sup> kg−<sup>1</sup> soil DM applied with a pipette close to the plant, directly on top of the soil substrate in four weekly intervals, starting at the sowing date. Seed dressing with Zn and Mn was performed with commercial formulations: Lebosol <sup>R</sup> Mn500 SC and Lebosol <sup>R</sup> Zn700 SC (Lebosol <sup>R</sup> Dünger GmbH, Ermstein, Germany) according to the manufacturer instructions. Lebosol <sup>R</sup> Zn700 SC: 2 ml 4,000 seeds−<sup>1</sup> , Lebosol <sup>R</sup> Mn500 SC: 4 ml 4,000 seeds−<sup>1</sup> .

#### Soil–Free Culture Experiment

Germination of maize seeds (Z. mays cv. Colisee), surface– sterilized by 1 min soaking in ethanol (99% v/v), was performed in pre-sterilized Petri dishes (9 cm diameter) on moist filter paper (10 seeds per petri dish). The seeds were soaked with 3 mL deionized water (–Si control) or 3 mL freshly prepared H4SiO<sup>4</sup> [1.0 mM Si] (Maksimovic et al., 2007; Pavlovic et al., 2013), respectively with four replicates per treatment. The optimum level of Si seed application was determined in a pilot experiment with different levels of Si supply (0–3.0 mM Si, Supplementary Figure 1). Covered Petri dishes were placed into a laboratory incubator (AtmoCONTROL, ICP, 750 Memmert GmbH, Schwabach, Germany) for 3 days in the dark at 18◦C in thin unsealed plastic bags to minimize evaporation. For further seedling development, the germinated seeds were placed on the upper edge of rolled filter papers (10<sup>∗</sup> 60 cm; 2 seeds per roll), moistened with 100 ml distilled water or H4SiO<sup>4</sup> solution [1 mM Si] at 3 and 5 days after sowing (DAS). The filter rolls were transferred into the laboratory incubator and incubated for 7 days at 12◦C (16/8 h light/dark period and relative humidity of 60%, with a light intensity of 200 µmol·m−<sup>2</sup> ·s <sup>−</sup><sup>1</sup> or at 18◦C outside the incubator.

#### Field Experiment

In 2016, a field experiment was established for silo maize production (Z. mays cv. Rolandinio) at the experimental station "Ihinger Hof " University of Hohenheim, Renningen, Germany (soil properties are specified in Supplementary Table 1). To increase the probability of low-temperature stress in spring, sowing was conducted on April 22nd. Accordingly, a mean air temperature of 13.6◦C during April and May was recorded associated with heavy rainfall (469 mm) leading to cold and wet soil conditions during germination and emergence. Fertilization was conducted prior to sowing by broadcast application with subsequent soil incorporation of stabilized ammonium sulfate NovaTec <sup>R</sup> solub 21: 161 kg N ha−<sup>1</sup> (Compo–Expert, Münster, Germany) and by underfoot placement of di-ammonium phosphate (29 kg N, 32 kg P ha−<sup>1</sup> ). Seed treatments were performed as seed dressings with Zn and Mn (Lebosol <sup>R</sup> Mn500 SC and Lebosol <sup>R</sup> Zn700 SC (Lebosol <sup>R</sup> Dünger GmbH, Ermstein, Germany) as described in experiment (1), and as seed priming with 1 mM potassium silicate (KSi, PottaSol, BioFa, Münsingen Germany) with a 24 h seed soaking period and re-drying during 24 h at 28 ± 2 ◦C. A seed water priming treatment without Si was included as a control. Sowing was performed on April 22nd with a sowing density of 9 seeds m−<sup>2</sup> , a row distance of 75 cm and a sowing depth of 6–7 cm. Foliar Si application (Vitanica <sup>R</sup> Si, Compo, Münster, Germany 16 L ha−<sup>1</sup> + 100 ml ha−<sup>1</sup> Greemax <sup>R</sup> Stallen Bio Agro AG, Basel, Switzerland) was conducted with a backpack sprayer (Solo <sup>R</sup> , Sindelfingen–Maichingen, Germany) at 69, and 75 DAS. Due to the extremely cold and wet soil conditions by the end of April, seedling emergence was severely biased, particularly in the untreated control variants. Therefore, after recording of emergence rates at 41 days after sowing (DAS), re–sowing was performed in the heavily affected plots to maintain a comparable level of inter-plant competition for light, water, and nutrients within the rows during the rest of the culture period. Final harvest was conducted at 214 DAS.

#### Plant Analysis

Visual scorings of leaf chlorosis, necrosis, and anthocyanin formation and determination of shoot height was performed for all experiments. Reflectometric leaf chlorophyll measurements were performed with a SPAD meter (Konica Minolta INC, Osaka, Japan). Estimates of damaged leaf area were obtained by the equation, leaf area (cm<sup>2</sup> ) = x/y, where x is the weight (g) of the area covered by leaf drawings on a transparent millimeter graph paper, and y is the weight of 1 cm<sup>2</sup> of the same graph paper (Pandey and Singh, 2011). Based on these data, the percentage of the necrotic area was calculated. After final harvest, root and shoot dry matter was determined after 60◦C oven-drying. Root length measurements were performed by digital image analysis using the WinRHIZO root analysis software package (Regent Instruments Inc., Quebec, Canada).

#### Analysis of Mineral Nutrients

One hundred milligrams of dried, milled shoot material were ashed for 5 h in a muffle furnace at 500◦C. After cooling, the samples were digested twice with 1 mL of 3.4 M HNO<sup>3</sup> and evaporated until dryness to precipitate SiO2. The ash was dissolved in 1 mL of 4 M HCl, subsequently diluted 10 times with hot deionized water, and boiled for 2 min to convert meta-, and pyro-phosphates to orthophosphate. After addition of 0.1 mL Cs/La buffer to 4.9 mL ash solution, Mg, Fe, Mn, and Zn concentrations were measured by atomic absorption spectrometry (ATI Unicam Solaar 939, Thermo Electron, Waltham, USA). Spectrophotometrical determination (Hitachi U−3300 spectrophotometer, Hitachi LtD. Corporation Japan) of orthophosphate was conducted after addition of molybdate-vanadate color reagent according to the method of Gericke and Kurmis (1952). K and Ca were measured by flame emission photometry (ELEX 6361, Eppendorf, Hamburg, Germany). Silicon was analyzed by ICP–OES (Vista–PRO, Varian Inc., Palo Alto, USA). For the digestion, 0.250 g of sample DM was suspended in 1 mL of H2O and 2.5 mL of conc. HNO3. The digestion is carried out by means of microwave–heated pressure digestion with HNO<sup>3</sup> and HF at 220◦C in a digestion system Ultra clave II (MLS GmbH, Leutkirch, Germany). The digestion took place over 20 min, the entire digestion program with heating and cooling phases comprised 2 h. After digestion, 0.5 ml HF solution (1% v/v) was added to dissolve sparingly soluble silicates. The solutions were adjusted to 10 mL with distilled H2O (VDLUFA Method book VII, 2011) and used for ICP–OES analysis.

#### Superoxide Dismutase Assay

The superoxide dismutase (SOD, EC 1.15.1.1) assay was optimized for root and shoot tissues of maize according to the method described by Beauchamp and Fridovich (1971) and modifications suggested by Giannopolitis and Ries (1977) and Hajiboland and Hasani (2007). One hundred milligrams of fresh plant material, frozen in liquid nitrogen and stored at −80◦C, were ground with a pre-cooled mortar and pestle, and homogenized in 1.5 ml extraction buffer containing 25 mM HEPES pH 7.8 and 0.1 mM EDTA. After centrifugation at 10,000 × g (4◦C for 10 min), aliquots of supernatant were transferred into 2 ml reaction tubes and kept on ice. For preparation of the reaction mixture, 1 ml cuvettes covered with aluminum foil for light protection, were filled with 300 µl 62.5 mM HEPES, 75 µl 1.0 mM EDTA, 75 µl 120 mM Na2CO3, 75 µl 120 mM Lmethionine, 150 µl 750µM nitro-blue tetrazolium (NBT), and 100 µl of plant extract. Finally, 225 µl of 10µM riboflavin was added. The light reaction was started by removing the aluminum foil, exposing the samples to a light source (8000 Lux) for 25 min. During the light phase, NBT is reduced to a dark blue formazan, measured spectrophotometrically (Spectrophotometer U−3300, Hitachi, Tokyo, Japan) at a wavelength of 650 nm. The final SOD activity, which inhibits the NBT reduction, was calculated as the difference between the absorbance of the sample and a control without plant extract, divided by 50% absorbance of the control. The specific SOD activity was expressed as SOD units per mg total protein. Total, protein content was determined according to Bradford (1976).

### Peroxidase Assay

Peroxidase (POD, EC1.11.1.7) activity was determined using the guaiacol test (Chance and Maehly, 1955; Hajiboland and Hasani, 2007). The tetra-guaiacol formed during the reaction is measured photometrically at 470 nm. The enzyme was extracted from fresh leaf material (100 mg) by 10 mM phosphate buffer (pH 7.0) and centrifuged 1,000 g for 10 min. The test mixture (1 mL) contained 10 mM phosphate buffer (300 µL, pH 7.0), 5 mM H2O<sup>2</sup> (300 µL), and 4 mM guaiacol (300 µL). The reaction was started by addition of the enzyme extract (100 µL) at 25◦C. The formation of tetraguaiacol was recorded over a reaction period of 5 min and the specific enzyme activity was expressed in µmoles tetraguiacol formation mg−<sup>1</sup> total protein.

## Determination of H2O<sup>2</sup>

Hydrogen peroxide levels were determined as described by Harinasut et al. (2003). Leaf tissues (100 mg fresh weight) were homogenized in an ice bath with 5 ml 0.1% (w/v) trichloroacetic acid. The homogenate was centrifuged at 12,000 g for 15 min. 0.5 ml of the supernatant was added to 0.5 ml of 10 mM potassium phosphate buffer (pH 7.0) and 1 ml of 1 M KI. The absorbance of the supernatant was recorded at 390 nm. The concentration of hydrogen peroxide was determined using a standard curve ranging from 0 to 120µM of H2O2.

### Determination of Total Soluble Sugars, Phenolics, and Proline

For determination of soluble sugars, leaf and root samples were homogenized in 100 mM phosphate buffer (pH 7.5) at 4◦C. After centrifugation at 12,000 g for 15 min, the supernatant was used for determination of total soluble sugars (Yemm and Willis, 1954). An aliquot of the supernatant was mixed with anthronesulfuric acid reagent (Yemm and Willis, 1954) and incubated for 10 min at 100◦C. After cooling, the absorbance was recorded at 625 nm. A calibration curve was created using glucose as external standard (Merck, Darmstadt, Germany). Total phenolics concentration was determined spectrophotometrically at 750 nm, using the Folin method (Hajiboland et al., 2017). For determination of proline, samples were homogenized with 3 % (v/v) sulfosalicylic acid and the homogenate was centrifuged at 3,000 g for 20 min. The supernatant was treated with acetic acid and acid ninhydrin, boiled for 1 h, and then the absorbance was determined at 520 nm. Proline (Sigma-Aldrich, Munich, Germany) was used for the production of a standard curve (Bates et al., 1973).

### Determination of Total Soluble Anthocyanins and Flavonoids

Determination of anthocyanins was conducted spectrophotometrically at 510 nm according to Plessi et al. (2007). One hundred milligrams of fresh shoot material was extracted with 2 ml methanol/HCl conc. (98:2 v/v). After centrifugation at 12,000 × g for 10 min, each 0.5 mL of the supernatant was used for spectrophotometric determination by using a pH differential method at pH 1 and pH 4.5 adjusted with 4.5 ml of MES buffer After 5 h incubation at 4◦C the absorbance was read at 510 nm from each group. The results were calculated as cyanidine-3-glycoside equivalents, using the

formula 1A <sup>∗</sup>MW∗DF∗V ∗ 100/ε <sup>∗</sup>Wt, where 1A is Abs (pH 1) – Abs (pH 4.5), MW = molecular weight of cyanidine-3-glycoside (484.83 g mol−<sup>1</sup> ), DF = dilution factor, V = final volume of the supernatant (0.5 ml), ε = molar absorbance factor of cyanidine-3-glycoside (26,900 Mm−<sup>1</sup> cm−<sup>1</sup> ), L = diameter of the light path [cm], and Wt = sample fresh weight (0.1 g). Total leaf flavonoids were determined in methanolic extracts according to Hajiboland et al. (2014). One hundred milligrams of fresh leaf material was extracted in AlCl3-methanol (2%, w/v) and after centrifugation at 12,000 × g for 10 min, the supernatant was used for determination at 415 nm with quercetin (Sigma, Munich, Germany) as an external standard.

### Determination of Total Antioxidants

The 1,1-diphenyl-2-picrylhydrazyl radical (DPPH •) has been used to evaluate the free radical scavenging activity of antioxidants (Panico et al., 2009). The DPPH solution was prepared by adding 2.37 mg DPPH (Sigma-Aldrich, Munich, Germany) in 2 ml 99% ethanol. One hundred milligrams of fresh leaf samples were grinded in 1 ml of extraction solution (1:1 Ethanol:Water). After centrifugation at 12,000 g for 10 min at 4◦C, 50 µl supernatant was used in a reaction cell which contained 50 µl freshly prepared 3 mM DPPH solution and 900 µl ethanol (99%). After incubation for 10 min in a dark room at 25◦C, the absorbance was determined at 515 nm. A reference solution contained 50 µl DPPH solution and 950 µl of ethanol (99%). The decline in absorbance at 515 nm was recorded for each sample and the quenching percentage of the DPPH radical was calculated based on the observed decrease in absorbance using the formula:

% Inhibition = [(A0-A1)/A0] × 100, where A<sup>0</sup> is the absorbance value of the DPPH • blank solution and A<sup>1</sup> is the absorbance value of the sample solution.

### Determination of Phytohormones by UHPLC-MS Analysis

Frozen maize tissue samples (shoot, roots) of 1 g of were ground to a fine powder with liquid nitrogen and extracted twice with 2.5 ml of 80% methanol in falcon tubes. Thereafter, the samples were further homogenized by ultrasonication (Micra D-9 homogenizer, Art, Müllheim Germany) for 1 min and 15 s at 10,000 rpm. Two milliliters of the methanol extracts were transferred to microtubes and centrifuged at 5,645 × g for 5 min. Thereafter, 350 µl of the supernatant was mixed with 700 µl ultra-pure water and centrifuged at 5,645 × g for 5 min. The supernatant was cleaned by membrane filtration (Chromafil <sup>R</sup> O-20/15 MS) and transferred to HPLC vials. UHPLC-MS analysis was carried out on a Velos LTQ System (Thermo Fisher Scientific, Waltham, Massachusetts, USA) fitted with a Synergi Polar column, 4µ, 150 <sup>∗</sup> 3.0 mm, (Phenomenex, Torrance, California, USA).The injection volume was 3 µL and the flow rate was adjusted to 0.5 ml min−<sup>1</sup> for gradient elution with mobile phase (A): water and 5% acetonitrile; mobile phase (B): acetonitrile and a gradient profile of: 0–1 min, 95% A, 5% B, 11–13 min, 10% A, 90% B, 13.1 min, 95% A, 5% B, 16 min 95% A, 5% B). All standards were purchased from Sigma Aldrich, (Sigma Aldrich, St. Louis, Missouri, USA) including (+/–)-jasmonic acid; 3-indoleacetic-acid, gibberellic acid, (+/–)-abscisic acid; trans-zeatin; salicylic acid.

### Statistical Analyses

The study was carried out in a completely randomized design for pot experiments and a randomized block design for the field experiment. Data are presented as means ± SE. For statistical analysis of significant differences between treatment groups, a one-way ANOVA followed by a Tukey–test (p < 0.05 significance level) were performed using the Sigma-Plot software 10.0 (Systat Software GmbH, Erkrath, Germany). For the statical evaluation of yield data from the field experiment, t-grouping instead of the Tukey test was applied as recommended by Mudra (1958).

### RESULTS

### Soil Culture Experiment

In the first experiment, soil-grown maize seedlings were exposed to 14 days reduced root zone temperature of 12–14◦C in a root cooling system. Cooling of the roots started at 2 weeks after sowing, followed by a 14-d recovery period. Seed treatments were performed with a commercial Zn/Mn seed dressing formulation. Silicon was applied as silicic acid (40 mg kg−<sup>1</sup> dry soil) by fertigation in four weekly intervals. The optimal dosage for silicic acid application has been determined in a pilot experiment (**Figure 1**).

#### Plant Growth and Development

Confirming the results of our earlier studies (Bradácˇová et al., 2016), 2 weeks exposure of maize plants to low RZT of 12–14◦C was associated with induction of leaf chlorosis (**Figures 2A,C**) necrosis, formation of stress anthocyanins (**Figure 2A**) limited shoot and root growth (**Figures 2B,E**) and an impaired micronutrient status (particularly Zn and Mn) below the deficiency thresholds (**Figures 2D**, **3A**; Bergmann, 1988). The stress responses and micronutrient deficiencies were mitigated

TABLE 1 | Superoxide dismutase activity (SOD), peroxidase activity (POD), and H2O2 concentrations in root tissue and SOD, total protein, total antioxidant capacity, total soluble phenolics, total soluble flavonoids, proline, anthocyanin, total soluble sugars, and Si concentrations in the shoot tissue of maize plants (cv Colisee) exposed to a 2-weeks period at reduced root zone temperature (12–14◦C) on a silty loam soil, pH 6.9.


Untreated control: (Ctrl); Zn Mn seed dressing: (ZnMnSeed), and silicon (H4SiO4) fertigation: (+Si). Means of three replicates. Significant differences (P < 0.05) are indicated by different characters.

by seed treatments with the respective micronutrients, but surprisingly, also by Si application (**Figure 2**). Shoot biomass and chlorophyll contents determined by SPAD measurements were increased by ∼50% (**Figures 2B,C**) and total root length even by 90% (**Figure 2E**).

#### Mineral Nutritional Status

Macronutrient analysis of the shoot tissue revealed no significant differences for the P, Ca, and K concentrations (**Figure 3A**). While the Ca and K status was in the sufficiency range, low concentrations were recorded for P and Mg. Seed dressing with Zn/Mn or Si fertigation significantly increased particularly the Mg status of the plants (Bergmann, 1988). The shoot micronutrient concentrations of the untreated control were 0.013 mg g−<sup>1</sup> DM for Zn and 0.025 mg g−<sup>1</sup> DM for Mn (**Figure 3B**) which is below the reported deficiency thresholds (Bergmann, 1988). The concentrations increased to the sufficiency range in response to the Zn/Mn seed dressing treatment, but interestingly also after Si fertigation (**Figure 3A**). The Fe and Cu status was low but not critical, without significant differences between the treatments (**Figure 3A**). Shoot Si was significantly increased only by Si fertigation (**Table 1**). In contrast to the nutrient concentrations, as indicators for the

plant nutritional status, total shoot contents of all investigated nutrients were significantly increased by the treatments with the chilling stress protectants Si and Zn/Mn (**Figures 3C,D**). This demonstrates that the Si and Zn/Mn applications generally improved nutrient acquisition and any surplus of mineral nutrients was readily transformed into biomass production (**Figure 2B**).

#### Oxidative Stress Indicators

In the root tissue directly exposed to low-temperature stress, Zn/Mn and Si applications increased the activities of superoxide dismutase (SOD, **Table 1**) and peroxidase (POD, **Table 1**) by 45– 46 and 71–84%, respectively. These enzymes are involved in oxidative stress defense and accordingly, H2O<sup>2</sup> concentrations declined by 54–63% (**Table 1**). In the shoot, SOD activities even increased by 179–183% (**Table 1**). Shoot concentrations of total proteins, antioxidants, total phenolics, total flavonoids, and proline concentrations were also strongly increased (**Table 1**) by both the Zn/Mn and Si treatments, while leaf anthocyanins, significantly declined (**Table 1**). In addition, the concentration of total soluble sugars of shoot tissue was significantly enhanced by Si treatment (**Table 1**).

#### Soil–Free Culture Experiment

Although Si induced a recovery of maize seedling growth after exposure to low RZT, which was associated with improved root development and mitigation of Zn and Mn deficiency (**Figure 3**, **Table 1**) we hypothesized that Si may exert also effects on improvement of the micronutrient status independent of root uptake. Thus, a second experiment was conducted to separate the root–mediated nutrient acquisition from internal redistribution of Zn and Mn from the seed to the establishing seedling. This was achieved by exposing maize seedlings to low temperature stress (7 d, 12◦C) in a soil–free culture system with and without Si application via seed soaking (3 d during germination at 18◦C), where the seedlings were exclusively dependent on their nutrient seed reserves and root uptake was excluded.

During the first 3 days of cultivation at 18◦C, no significant treatment differences were recorded for germination rates (**Figure 4**). However, the subsequent 7 d–cold stress periods at 12◦C induced a retardation in seedling growth, associated with a reduction in shoot and root biomass production by 53 and 60%, respectively, and intense necrotic and chlorotic leaf damage (**Table 2**) similar to the symptoms observed in soil culture (**Figures 2A** and **5**). Silicon treatments reduced the necrotic leaf

TABLE 2 | Effect of Si seed soaking on biomass production, oxidative leaf damage [% necrotic leaf area], tissue concentrations of Zn, Mn, Si, soluble sugars, and total antioxidants in maize seedlings (cv Colisee) exposed to 7 days chilling stress at 12◦C in a soil-free filter roll culture system.


Means of five replicates. Significant differences (P< 0.05) are indicated by different characters.

area by more than 90%. This was associated with increased levels of total antioxidants (+27%) and soluble sugars (+117%) in the leaf tissue. The nutritional status of Zn and Mn was significantly increased by the Si treatment and a trend for increased root and shoot biomass production was detectable after the cold period of 7 d (**Table 2**).

During the 7-days cold stress period, shoot and root concentrations of all important hormonal growth regulators (indole acetic acid—IAA, zeatin, gibberellic acid—GA), as well as stress–related hormones, such as abscisic (ABA), jasmonic and salicylic acids, declined significantly. However, after Si seed soaking, hormonal concentrations were restored comparable to those of unstressed plants or even further increased (shoot concentrations of IAA, gibberellic acid, and ABA, **Table 3**).

Since plant cultivation in this experiment was conducted in a nutrient–free culture system, lacking any additional Zn

or Mn (Si treatment solution pre-purified by cation exchange chromatography) (Maksimovic et al., 2007; Pavlovic et al., 2013), it was possible to monitor the redistribution of seed– stored Zn and Mn during seedling development. Thus, any additional nutrient uptake via the roots was excluded. During the 10 days culture period, 50% of the seed Zn contents and 42% of seed Mn were translocated to the developing control seedlings. Interestingly, the translocation of micronutrient seed reserves was stimulated by the Si treatments, resulting in significantly higher total Zn (+67%) and Mn (+62%) contents at the end of the cold stress period, as compared with the untreated controls (**Figure 6**). By comparison with the remaining nutrient contents in the seeds, it was also possible to calculate the nutrient leaching losses. In the untreated control plants exposed to chilling stress, leaching accounted for 29% of the original seed Zn contents and for 41% of Mn. However, silicon treatment significantly reduced the large leaching losses of Zn and Mn by 70 and 48%, respectively (**Figure 6**).

Apart from the micronutrients, also the re-distribution of K was investigated, because of its well-established role as an indicator for nutrient leaching in response to impairment of tissue and membrane integrity (Cakmak and Marschner, 1987) and its function in cold stress protection (Wang et al., 2013). At the end of the cold stress period, untreated control seedlings had lost 72% of their K seed reserves by leaching and only 22% had been translocated to the developing seedling. Silicon treatments reduced these large leaching losses by 17% and increased the K contents of the seedling by 38% (**Figure 6**). For other tested macronutrients, such as P, Mg, and Ca, no significant treatment effects were recorded (data not shown). However, Si not only reduced the leaching losses and improved the translocation of seed reserves of mineral nutrients to growing tissues. Additional selective effects were detectable for the mineral nutrient ratio between shoots and roots. While Si pre-treatments increased the Mn and K contents in shoots and roots to a similar extent (Mn: +68–77%; K: +38–46%), Si preferentially increased the Zn shoot contents by 64%, but less, only by 34%, in the root tissue (**Figure 6**).

TABLE 3 | Endogenous concentrations of phytohormones in maize seedlings (cv Colisee) grown for 7 days at 18 or 12◦C, in a soil-free filter roll culture system with (+Si) or without (–Si) silicic acid seed-soaking (–Si) during a 3-day pre-germination period at 18◦C.


IAA, indole acetic acid; ABA, abcsisic acid; GA, gibberellic acid; JA, jasmonic acid; SA, salicylic acid. Means of three replicates. Significant differences (P < 0.05) are indicated by different characters. ND, Not Detectable.

## Field Experiment

To investigate the benefit of Zn/Mn and Si starter treatments on maize performance under on-farm conditions, a field experiment was established. Early sowing on April 22nd resulted in cold stress during germination and early growth. Due to low temperatures (13.6◦C) and high precipitation (469 mm) during April and May, emergence and seedling growth were heavily affected by cold stress, but also by oxygen limitation and Aspergillus niger infections on the cold and wet silty-loam soil. By the mid of May, at 41 DAS (BBCH 15, stage 1), an emergence rate of only 44% was recorded in the untreated control variant (**Table 1**). However, emergence was significantly increased by the Zn/Mn seed dressing (+12%) and particularly by seed soaking with KSi (+28%). To account for a potential water priming effect of the seed soaking treatment (Lutts et al., 2016), also a watersoaked control was included, which increased emergence by 8% (**Table 4**).

At 49 DAS (BBCH 17, stage 1), a mineral nutrient analysis was conducted for the youngest fully developed leaves. In the control variant, the Zn–nutritional status was low (**Table 4**) as expected, and K and P were even below the deficiency thresholds. Mg, Cu, and Mn supply, by contrast, was sufficient (Supplementary Table 2). Similar to the pot experiment, both, Zn/Mn and Si treatments significantly increased the Zn concentrations to a sufficient level of 65 and 59 µg g−<sup>1</sup> dry matter (Bergmann, 1988). A trend for an improved status was recorded also for the remaining nutrients (Supplementary Table 2). The protective effects of Zn/Mn and Si starter treatments were finally reflected in substantially higher biomass yield of the heavily cold-affected plots by 56 and 82%, respectively, when calculated according to the plant density determined at 41 DAS. In the Si treatments, 25% of the yield increase could be attributed to a water priming effect. Foliar Si application had no significant impact on final biomass yield (**Table 4**). However, due to the extremely low emergence (44%), particularly in the untreated plots, re-sowing was performed in the most heavily affected parts, by the mid of May after the end of the stress period, to provide comparable inter-plant competition for light, water, and nutrients within the rows during the rest of the culture period. But even including the non-stressed plants after re-sowing into the yield calculation, a TABLE 4 | Emergence, Zn/Mn status (DAS), and final biomass yield of field-grown silo maize (Rolandinio) on a silty loam soil pH 6.9 at the experimental station "Ihinger Hof" University of Hohenheim with underfoot placement of di-ammonium phosphate (29 kg N, 32 kg P ha−<sup>1</sup> ) and stabilized ammonium sulfate fertilization (161 kg N ha−<sup>1</sup> ) with or without Zn/Mn seed dressing (Lebosol Mn500 SC, Zn700 SC), potassium silicate [1 mM] seed priming, water priming of foliar Si application of Si as (16 L Si [1 mM] + 100 ml Greemax® ha−<sup>1</sup> ).


Yield determinations with (in brackets) and without re-sowing by the end of May 2016. Means of five replicates per treatment. Significant differences (P < 0.05) are indicated by different characters.

significant yield increase (+10.6%) was measured only for the Si seed priming variant (**Table 4**).

#### DISCUSSION

#### Si Mimics Cold Stress-Protective Effects of Zn/Mn Starter Applications

Similar to earlier reports (Imran et al., 2013; Bradácová et al., ˇ 2016), starter treatments by seed dressing or seed priming with Zn and Mn significantly increased cold tolerance of maize in the pot experiment with controlled root-zone temperature (**Figure 2**). The 14-d chilling treatments with 12◦C root zone temperature induced Zn and Mn deficiencies in the soil-grown maize seedlings (**Figure 3**). This seems to be not primarily related to Zn/Mn availability in the soil, since Imran et al. (2013) demonstrated that Zn and Mn accumulation in the shoot tissue was suppressed during the cold stress period, even with freely available nutrient supply in a hydroponic culture system. Only Zn and Mn uptake before the onset of the stress period via seed priming, seed dressing or fertigation, could increase shoot micronutrient accumulation above the deficiency thresholds, detectable even after the end of the 2-weeks cold stress period (**Figure 3**). This is in line with findings of Engels and Marschner (1992, 1996), demonstrating that limited Zn and Mn shoot accumulation in maize exposed to low root zone temperatures was particularly dependent on cold-stress effects affecting nutrient uptake and root activity.

In accordance with the low Zn/Mn status, the cold-stressed control plants exhibited various symptoms characteristic for Zn and Mn limitation (Cakmak, 2000), such as chlorosis and oxidative leaf damage, stunted shoot, and root growth (**Figure 2**) reduced activity of enzymes involved in ROS detoxification with micronutrients as co-factors (SODs, PODs), and impaired biosynthesis of phenolic antioxidants (**Table 1**) which depends on Cu and Mn cofactors (Datnoff et al., 2007). Consequently, excessive accumulation of ROS (**Table 1**) resulted in oxidative damage of plant tissues (**Figure 2**) considered as one of the major constraints for cold-stressed plants (Baek and Skinner, 2012; Saeidnejad et al., 2012). Supplementation of Zn and Mn via seed priming (Imran et al., 2013) or seed dressing were able to overcome Zn/Mn deficiency (**Figure 3**) and largely mitigated the related stress symptoms described above. Surprisingly, also starter fertigation or seed soaking with Si completely mimicked all cold-stress-protective effects of Zn/Mn starter applications (**Figures 2**, **Table 1**). Silicon increased the Zn/Mn status, activities of ROS detoxification enzymes, and accumulation of antioxidants to a comparable level as the Zn/Mn treatments (**Figures 3**, **Table 1**) although Si was applied as free silicic acid (H4SiO4), pre-purified via cation exchange chromatography, and no Zn or Mn was detectable in the application solution.

Also in the field experiment, seed treatments with Si and Zn Mn improved emergence and the micronutrient status (particularly Zn) of maize seedlings, exposed to sub-optimal germination temperatures and cold and wet soil conditions by early sowing at the mid of April (**Table 4**). A certain protective effect was recorded also for the control treatment with water-primed seeds but less expressed as compared with Si seed priming. This is in line with the well-documented beneficial effects of water priming on seed germination by metabolic pre-activation (Lutts et al., 2016). In accordance with earlier observations (Imran et al., 2013; Bradácová et al., ˇ 2016), application of cold stress protectants was ineffective after the onset of the stress period, i.e., when Si was supplied by the foliar application (**Table 4**). By contrast, silicon seed priming was the most effective treatment and increased emergence by 64% as compared with the untreated control, which translated into a significant increase in final yield in both scenarios of yield determination (with and without re-sowing after the end of the cold stress period; **Table 4**). Similar effects have been reported also in two additional field experiments conducted under comparable climatic conditions when the micronutrient status of the maize seedlings was increased by direct supplementation of Zn, Mn, and Fe via seed priming. In these field trials, marketable grain yields increased by 13–15% (Imran et al., 2013). The surprisingly long-lasting effects of the starter treatments with cold stress protectants may be related to the intense stimulation of root growth (**Figure 2D**) which has an advantage for plant performance not only during the stress period but also during the recovery phase and under more favorable growth conditions.

#### Silicon Reduces Leaching Losses and Promotes Utilization of Zn/Mn Seed Reserves

The striking similarity of Zn/Mn and Si effects on cold-stressed maize seedlings raises the question whether Si exerts its protective effects via improvement of the plant micronutrient (Zn/Mn) status. Restoration of cold stress-induced root growth inhibition was among the most apparent effects of Si or Zn/Mn applications (**Figure 2E**). Of course, root growth stimulation can contribute to improved nutrient acquisition in general, as demonstrated for increased shoot accumulation of P, K, Mg, Ca, Fe, Zn, Mn, and Cu recorded after the cold stress period in Si-treated maize plants (**Figure 3**). However, Si was able to improve selectively the Zn/Mn status of maize seedlings, already during the first week of the cold stress period, before a marked stimulation of root growth was detectable (**Table 2**). Moreover, this effect was observed in a soil–free culture system excluding the option for further Zn/Mn root uptake from the external medium (**Figure 5**). This implicates that the improved micronutrient status of the Sitreated maize seedlings exposed to low-temperature stress cannot be exclusively attributed to Si-induced root growth stimulation but involves also Si effects counteracting nutrient leaching and promoting internal distribution of Zn and Mn. Imran et al. (2015) demonstrated that seed reserves can cover the Zn and Mn demand of maize seedlings for about 2–3 weeks. However, cold stress is a well-documented stress factor leading to electrolyte leakage via oxidative membrane damage (Bewley and Black, 1994), which can limit the seed reserves of mineral nutrients. Accordingly, our study revealed large leaching losses of 30– 40% for the Zn/Mn seed reserves and even 70% for K, as another mineral nutrient with cold-protective functions (Wang et al., 2013), during the first 10 days of seedling development after a 12◦C cold stress period of 7 days (**Figure 6**). Seed soaking with Si dramatically reduced the leaching losses by 70% (Zn), 50% (Mn), and 15% (K), leading to an improved nutrient supply to the developing seedling. This effect may be attributed to the well-documented protective functions of Si against oxidative membrane damage (He et al., 2010). However, Si seed soaking increased the root and shoot contents of Mn by 77 and 68%, respectively, while the root contents of Zn were increased only by 34%, but by 64% in the shoot tissue (**Figure 6**). This indicates a selective effect of Si on the root/shoot distribution of Zn. A similar improved Zn status by Si treatment has been reported for Zn-deficient soybean plants (Pascual et al., 2016).

Different mechanisms have been proposed for effects of Si, improving internal Zn availability in Zn-deficient plants. Increased production of phenolics with metal chelating properties induced by Si treatments (Pavlovic et al., 2013), as observed also in the present study (**Table 1**) may increase internal mobility and transport of Zn within the plant. This may be related to the improved Mn status of Si treated plants (**Figure 3**) as an important enzymatic cofactor for the biosynthesis of phenolics (Datnoff et al., 2007). In later stages of plant development, the same mechanism may be responsible also for the remobilization of Zn sequestered in the apoplast together with iron plaques (Chen et al., 1980), as similarly demonstrated for apoplastic Fe remobilization in cucumber (Pavlovic et al., 2013). Bityutskii et al. (2014) could not confirm this interaction, but in that study, plants were grown in nutrient solution without Zn supply, which may have prevented the accumulation of apoplastic Zn pools. Other studies suggest direct Si-metal interactions counteracting apoplastic metal immobilization and supporting metal transport in plants (Pavlovic et al., 2013; Hernandez– Apaolaza, 2014; Stevic et al., 2016). Effects of Si on the expression of metal acquisition and transport genes have been reported by Pavlovic et al. (2013, 2016), but the underlying mechanisms are unknown.

### Si Restores the Levels of Hormonal Growth Regulators in Cold-Stress-Affected Maize Seedlings

Due to the impairment of Zn/Mn-dependent ROS detoxification systems, induced by limited Zn/Mn availability in coldstressed plants (**Figures 2**, **3**) excessive ROS accumulation causes oxidative damage, leading to leaf chlorosis and necrosis (**Figure 2**). This is associated with an impairment of photosynthesis, resulting in a reduced root allocation of assimilates required for root development, as previously reported also by Sowinski et al. (1998). Moreover, excessive production of ROS can promote oxidative degradation of indole acetic acid and result in a 50% reduction of IAA contents in Zn-deficient Phaseolus vulgaris, which could be restored by Zn fertilization. Therefore, auxin deficiency was considered as another important factor for growth limitation in Zn deficient plants (Cakmak et al., 1989). Similarly, in our study cold stress induced, both, Zn/Mn limitation (**Table 1**) and a 60% reduction of IAA accumulation in the shoot and root tissue (**Table 3**) associated with inhibition of shoot and root growth and all symptoms were reverted by exogenous Si application (**Figure 2**). More recent studies suggest that cold stress additionally affects root growth via inhibition of PIN2 and PIN3-mediated basipetal auxin transport within the roots (Shibasaki et al., 2009).

Shoot growth in cold-stressed plants seems to be also affected by a reduction of bioactive growth-promoting gibberellic acid (GA) levels, leading to an increased abundance of nuclear DELLA-protein growth repressors via a signaling pathway involving CBF/DREB1 transcription factors (Miura and Furumoto, 2013; Eremina et al., 2016). Accordingly, in our study GA levels in shoot and roots of cold-stressed maize seedlings declined by ∼60–70%. This effect was completely reverted by Si seed soaking (**Table 3**). In line with this observation, in Arabidopsis thaliana it was demonstrated that cold stress stimulates GA degradation by upregulation of the GA 2-oxidase gene and simultaneously impairs GA-biosynthesis by repressing the GA 20-oxidase gene (Eremina et al., 2016). For Zn-deficient plants, reduced GA concentrations have been recorded (Suge et al., 1986; Sekimoto et al., 1997) similar to the reduction in IAA levels reported by Cakmak et al. (1989). More recently, it was shown that various steps of GA biosynthesis depend on the presence of IAA (Ross et al., 2001). Therefore, the observed reduction of GA levels in cold-stressed maize seedlings (may be a consequence of auxin deficiency, resulting from the limited Zn supply caused by low temperature stress. Interestingly, it was shown that exogenous application of IAA and GA could increase the accumulation of flavonoids and other phenolics in buckwheat (Fagopyrum esculentum) seedlings (Park et al., 2017). This observation is in line with the increased production of phenolics and antioxidants, induced by Si application in cold-stressed maize seedlings (**Table 1**) that may be triggered by the increased IAA and GA levels in the respective plants (**Table 3**).

Similar to auxin and GA, also the levels of the metabolically active cytokinin form "zeatin," with functions in stimulation of cell division and cell expansion, declined in cold-stressed maize seedlings (**Table 3**). A similar decline had been previously reported for Arabidopsis,rice and, wheat (Eremina et al., 2016). In rice, this was associated with a significant downregulation of gene expression related to cytokinin biosynthesis (Maruyama et al., 2014). Accordingly, external application of cytokinins increased cold tolerance in Arabidopsis (Jeon et al., 2010; Shi et al., 2012). In our study, cold-stress protection by Si application was also associated with increased zeatin concentrations in the root and shoot tissue (**Table 3**).

Taken together, the results demonstrated that cold stress significantly reduced the internal concentrations of the most important hormonal regulators of plant growth. Furthermore, the cold-protective effect of Si application was related to a restoration of hormonal levels comparable to those of unstressed plants (**Table 3**).

### Si Increases the Levels of Hormonal Stress Regulators in Cold Stress-Affected Maize Seedlings

The Si treatments influenced also the levels of abscisic (ABA) salicylic (SA) and jasmonic acids (JA), known as hormones more directly involved in the regulation of abiotic and biotic stress responses (**Table 3**). These hormones have been also implicated in cold stress signaling (Miura and Furumoto, 2013; Eremina et al., 2016). Improved cold acclimation and increased cold tolerance by exogenous applications of ABA, SA, and JA have been reported for various plant species (Horváth et al., 2007; Kumar et al., 2008; Eremina et al., 2016; Hu et al., 2017). Mutants affected in ABA, SA, and JA metabolism show altered responsiveness to cold stress (Eremina et al., 2016). Accordingly, in our study, exposure of maize, as cold-sensitive plant species to a 7-days cold period, significantly decreased the levels of ABA, SA, and JA. The cold-protective effect of Si starter treatments was associated with an increase of ABA and to a smaller extent also of SA and JA, particularly in the shoot (**Table 3**). Abscisic acid is considered as a central regulator of cold stress responses in plants and seems to regulate the adaptive expression of coldrelated genes with cold-protective functions via CBF-dependent and independent pathways, in cross-talks involving also SA and JA (Szalai et al., 2011; Eremina et al., 2016). Direct links between ABA and induction of oxidative stress defense enzymes, such as SOD in cold-stressed plants similar to our study (**Figure 3**) have been reported by Kumar et al. (2008), Szalai et al. (2011), and Li and Zhang (2012). Moreover, the cold stress-induced accumulation of cryo-protectants, such as proline (**Figure 3**) has been linked with a reduction of leaching losses via ABA-induced protection against oxidative membrane damage (Chen and Li, 2002). Proline biosynthesis at least partially depends on ABA signaling (Szabados and Savouré, 2010). Increased ABA levels in the Si-treated plants may be also related to the improved Znnutritional status induced by the Si starter treatments (**Figure 3**). Accordingly, Cakmak et al. (1989) and more recently Wang et al. (2012) reported that Zn deficiency reduced the ABA levels in Phaseolus vulgaris and in apple rootstocks.

The surprisingly distinct effects of the Si treatments on the hormonal balances (**Table 3**) and the timing of hormonal changes just at the beginning of detectable growth responses (**Table 2**) are clear indicators for the proposed interactions of the Si amendments with hormonal cold stress signaling. Nevertheless, the obvious interactions of Si with hormonal balances require further investigations considering quantitative changes during plant development and a higher spatial resolution, since it is known that local changes in hormone concentrations at the cellular level control adaptive growth and development (Eremina et al., 2016). A conceptual model of the proposed interactions between Si, micronutrients, and hormones mediating chilling tolerance in maize seedlings and its relation to plant growth is presented in **Figure 7** and summarizes our findings.

### CONCLUDING REMARKS

The findings of the present study suggest that induced deficiency of Zn and Mn as a consequence of leaching during early development and limited root growth and activity is a major factor determining the sensitivity of young maize plants to chilling stress, with options for mitigation by supplementing germinating seeds with Zn, Mn, or Si. In this context, the protective effect of Si treatments is related to an improved Zn and Mn status, starting already during germination with a protective effect of Si against leaching losses of micronutrient seed reserves and improved translocation to the developing seedling. The improved micronutrient status can at least partially explain the ability of the plants to maintain a balanced hormonal (IAA, GA, cytokinin) status that restores plant growth and particularly root development, which allows further nutrient uptake. The improved micronutrient status also helps to increase the expression of enzymatic (SOD, POD) and non-enzymatic (phenolic antioxidants) defense systems against cold-induced oxidative stress. Finally, it stimulates the accumulation of stress hormones (ABA, SA, JA), potentially priming various physiological adaptations to chilling stress via expression of

cold response genes. It remains to be established whether this applies also to protective effects of Si against other abiotic stresses, such as drought or salinity, which are also associated with oxidative stress. Remarkably, the improved micronutrient status by Si treatment under low temperature stress is not only restricted to controlled lab conditions. The preliminary field experiment demonstrated that the observed protective effects of Si starter treatments on seedling performance can translate into improved yields in agricultural practice and requires further investigation considering different soil types and maize varieties.

#### AUTHOR CONTRIBUTIONS

NM, GN, and MW: conceived and designed the experiments; NM: conducted the experiments, collected the data, and wrote the manuscript with GN and UL; FW, BH, and NM: developed and performed the UHPLC-MS analysis of phytohormones. All authors approved the final manuscript.

#### REFERENCES


#### ACKNOWLEDGMENTS

This work was supported by the European's Seventh Framework Programme (FP/2007–2013) under Grant Agreement no. 312117.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018. 00420/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 © 2018 Moradtalab, Weinmann, Walker, Höglinger, Ludewig and Neumann. 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) 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.

# Expression of the Intracellular COPT3-Mediated Cu Transport Is Temporally Regulated by the TCP16 Transcription Factor

Nuria Andrés-Colás<sup>1</sup>† , Angela Carrió-Seguí<sup>1</sup> , Salah E. Abdel-Ghany<sup>2</sup> , Marinus Pilon<sup>2</sup> and Lola Peñarrubia<sup>1</sup> \*

<sup>1</sup> Departament de Bioquímica i Biologia Molecular, Estructura de Recerca Interdisciplinar en Biotecnologia i Biomedicina, Universitat de València, Valencia, Spain, <sup>2</sup> Department of Biology, Colorado State University, Fort Collins, CO, United States

#### Edited by:

Manuel González-Guerrero, Universidad Politécnica de Madrid (UPM), Spain

#### Reviewed by:

Crysten Elizabeth Blaby-Haas, Brookhaven National Laboratory (DOE), United States Marc Hanikenne, University of Liège, Belgium

> \*Correspondence: Lola Peñarrubia penarrub@uv.es

#### †Present address:

Nuria Andrés-Colás, Departamento de Biotecnología, Universitat Politècnica de València, Valencia, Spain

#### Specialty section:

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

Received: 28 March 2018 Accepted: 08 June 2018 Published: 03 July 2018

#### Citation:

Andrés-Colás N, Carrió-Seguí A, Abdel-Ghany SE, Pilon M and Peñarrubia L (2018) Expression of the Intracellular COPT3-Mediated Cu Transport Is Temporally Regulated by the TCP16 Transcription Factor. Front. Plant Sci. 9:910. doi: 10.3389/fpls.2018.00910 Copper is an essential element in plants. When scarce, copper is acquired from extracellular environment or remobilized from intracellular sites, through members of the high affinity copper transporters family COPT located at the plasma membrane and internal membrane, respectively. Here, we show that COPT3 is an intracellular copper transporter, located at a compartment of the secretory pathway, that is mainly expressed in pollen grains and vascular bundles. Contrary to the COPT1 plasma membrane member, the expression of the internal COPT3 membrane transporter was higher at 12 h than at 0 h of a neutral photoperiod day under copper deficiency. The screening of a library of conditionally overexpressed transcription factors implicated members of the TCP family in the COPT3 differential temporal expression pattern. Particularly, in vitro, TCP16 was found to bind to the COPT3 promoter and down-regulated its expression. Accordingly, TCP16 was mainly expressed at 0 h under copper deficiency and induced at 12 h by copper excess. Moreover, TCP16 overexpression resulted in increased sensitivity to copper deficiency, whereas the tcp16 mutant was sensitive to copper excess. Both copper content and the expression of particular copper status markers were altered in plants with modified levels of TCP16. Consistent with TCP16 affecting pollen development, the lack of COPT3 function led to altered pollen morphology. Furthermore, analysis of copt3 and COPT3 overexpressing plants revealed that COPT3 function exerted a negative effect on TCP16 expression. Taken together, these results suggest a differential daily regulation of copper uptake depending on the external and internal copper pools, in which TCP16 inhibits copper remobilization at dawn through repression of intracellular transporters.

Keywords: copper transport, COPT3, heavy metals, TCP16, transcriptional regulation

## INTRODUCTION

Copper (Cu) is an essential micronutrient for the growth and development of aerobic organisms. Under metal deficiency, Cu<sup>+</sup> is incorporated through high affinity COPper Transporters, denoted COPTs in plants (Kampfenkel et al., 1995; Sancenón et al., 2003; Puig, 2014; Peñarrubia et al., 2015) and referred to as CTR (SLC31) in other organisms (Kim et al., 2013). In Arabidopsis thaliana, the COPT family can be divided into two subfamilies: the plasma membrane members

**31**

COPT1, COPT2, and COPT6 (pmCOPT), which are transcriptionally induced under Cu deficiency (Sancenón et al., 2004; Garcia-Molina et al., 2013; Perea-García et al., 2013), and the members located in internal membranes COPT3 and COPT5 (imCOPT), which are not clearly induced by Cu deficiency (Sancenón et al., 2003; Garcia-Molina et al., 2011; Klaumann et al., 2011). This subdivision might distinguish at least two Cu<sup>+</sup> sources in the cells (external and internal), differentially mobilized based on the type of COPT activated.

The transcriptional response to Cu deficiency is mainly orchestrated by the transcription factor (TF) SQUAMOSA PROMOTER BINDING PROTEIN-LIKE7 (SPL7) through binding to GTAC motifs in the promoters of target genes, such as pmCOPT (Yamasaki et al., 2009; Bernal et al., 2012). Thus, pmCOPT-mediated cytosolic Cu<sup>+</sup> uptake from the extracellular environment is highly increased by SPL7 under Cu deficiency. It has been suggested that the SPL7-mediated auto-regulatory effect of Cu on pmCOPT expression could establish a feedback loop responsible for cyclic expression, peaking at dawn (Andrés-Colás et al., 2010; Peñarrubia et al., 2010). Deregulated Cu<sup>+</sup> uptake in COPT-overexpressing plants causes abnormal development in the absence of environmental cycles (Andrés-Colás et al., 2010; Perea-García et al., 2016a,b). Furthermore, the observed interaction between SPL7 and ELONGATED HYPOCOTYL5 (HY5) underscores a connection between Cu homeostasis and light (Zhang et al., 2014).

In yeast, it has been shown that, due to the Cu<sup>+</sup> toxicity, practically no free Cu<sup>+</sup> is found in the cytosolic compartment (Rae et al., 1999). Therefore, pmCOPT-mediated Cu<sup>+</sup> uptake is probably coupled to cupro-chaperone-mediated delivery to different protein targets, among them the P-type ATPase RESPONSIVE-TO-ANTAGONIST1 (RAN1) located at the endoplasmic reticulum (ER) where it pumps Cu<sup>+</sup> into the lumen (Hirayama et al., 1999). SPL7 presents an operative transmembrane domain that allow his localization at the endomembrane system, most likely the ER. During ER stress, as a result of Cu deficiency, SPL7 localizes in the nucleus to modulate the Cu deficiency responses, after activation by proteolytic cleavage. In this sense, SPL7 could function as a double Cu sensor in both the nucleo-cytoplasm and the ER lumen (Garcia-Molina et al., 2014).

With regard to the imCOPT-mediated Cu<sup>+</sup> transport from internal sources, COPT5 plays an important role in the plant response to severe environmental Cu scarcity (Garcia-Molina et al., 2011; Klaumann et al., 2011). COPT5 functions in remobilizing Cu from prevacuolar vesicles, which could act as internal stores or recycling vesicles to provide the metal to key Cu-dependent processes such as photosynthesis (Garcia-Molina et al., 2011; Klaumann et al., 2011; Carrió-Seguí et al., 2015). Little is known about the function of COPT3. COPT3 has been classified as a member of the imCOPT subfamily based on its partial complementation of the respiratory and Cu transport defect exhibited by a ctr11ctr31 yeast mutant (Sancenón et al., 2003; Garcia-Molina et al., 2013). Apparently, imCOPT expression is not affected by Cu and temporal factors modulating imCOPT expression remain unexplored.

The TCP (named after TEOSINTE BRANCHED 1, CYCLOIDEA and PROLIFERATING CELL FACTOR 1) protein family precisely orchestrates spatial and temporal plant responses to both environmental and endogenous factors (Martín-Trillo and Cubas, 2010; Kieffer et al., 2011; Li, 2015; Danisman, 2016; Dhaka et al., 2017). The TCP family is constituted by plant-specific TFs that share a conserved non-canonical basic helix–loop–helix (bHLH) DNA binding domain, termed TCP domain (Cubas et al., 1999). The TCP members, are grouped into two subfamilies, based on the TCP domain structure and their roles (Cubas et al., 1999). These subfamilies are denoted as class I or PROLIFERATING CELL FACTOR (PCF) and class II, which is divided in CINCINNATA (CIN) and CYCLOIDEA (CYC)/(TEOSINTE BRANCHED 1) TB1 proteins. TCPs bind cis-acting regulatory elements (CAREs) known as site II, in the promoter regions of various genes. CAREs come in two classes: class I (GTGGGNCC) and class II (GTGGNCCC), which have different but similar binding preferences (Kosugi and Ohashi, 2002; Viola et al., 2012). The peculiar class I member TCP16 is an exception with preference for class II binding site since it contains Asp instead of Gly at a key discriminatory position (Uberti-Manassero et al., 2016).

A key mechanism underlying temporal control is the circadian clock. Among the central Arabidopsis clock components are the TFs CIRCADIAN CLOCK-ASSOCIATED 1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY) (for a review, see Nohales and Kay, 2016). Some TCP members interact with different components of the core circadian clock as shown in both yeast two-hybrid and protein–protein interaction assays (Giraud et al., 2010), which indicates that the TCP family is intricately linked to circadian regulation of gene expression in Arabidopsis. TCP21, termed CHE (for CCA1 Hiking Expedition), binds TOC1 (timing of CAB expression 1), which provides an explanation of how TOC1 can regulate expression of CCA1, as TOC1 lacks a DNA binding domain (Pruneda-Paz et al., 2009). Moreover, the concomitant binding of TCP20/TCP22 and LWD1 (LIGHT-REGULATED WD1) to the CCA1 promoter activates its expression at dawn (Wu et al., 2016). Furthermore, TCPs appear to link the diurnal changes in mitochondrial function, particularly in genes encoding components of the oxidative phosphorylation machinery, with transcriptional changes that are regulated and integrated with the central clock function. This provides a molecular link between cellular and organelle metabolic activity and the circadian clock in plants (Palatnik et al., 2003; Welchen and Gonzalez, 2006; Giraud et al., 2010; Danisman, 2016).

Other developmental plant process that requires a TCP-mediated precise spatial and temporal control is the regulation of floral organ development, including secondary cell wall thickening necessary to release pollen grains. This developmental program is under the precise control of TCP24, which functions as a negative regulator (Wang et al., 2015). Moreover, the inhibition of the TCP16 function results in abortion of early pollen development (Takeda et al., 2006). In rice, class I TCP genes have been mainly implicated in stress adaptation, such as salinity tolerance (Almeida et al., 2017) or cold stress (Wang et al., 2014). OsTCP19 facilitates abiotic stress

tolerance by manipulating the abscisic acid (ABA) signaling network (Mukhopadhyay and Tyagi, 2015). The wide crosstalk between TCP and hormones has been recently summarized (Nicolas and Cubas, 2016).

Environmental signals, such as nutrient availability, also lead to TCP-mediated regulation. In this sense, TCP20 and NIN-like proteins (NLP6 and NPL7) are involved in nitrate availability responses (Guan et al., 2014, 2017). This is only an example of the high range of TCPs interactions with other TFs (Bemer et al., 2017). These multiple interactions highlight the central role of TCPs in plant molecular networks that integrate environmental and endogenous processes in plants (Danisman, 2016; Dhaka et al., 2017). Metal availability is a key environmental factor under precise temporal control intricately linked with the circadian clock (Andrés-Colás et al., 2010; Hermans et al., 2010; Chen et al., 2013; Hong et al., 2013; Salomé et al., 2013). In this regard, it has been shown that TCP20 transcriptionally repress the expression of the subgroup Ib of bHLH TFs, previously implicated in iron homeostasis (Wang et al., 2007). Moreover, these TFs are up-regulated in the transition from cell proliferation to cell expansion during sink-source transitions (Andriankaja et al., 2014).

Although the activation of Cu<sup>+</sup> uptake through the pmCOPT transporters under Cu deficiency by SLP7 is a well-established process (Yamasaki et al., 2009; Bernal et al., 2012), their temporal control, as well as the transcriptional regulation of imCOPTs, COPT3 and COPT5, remain unsolved. In this work, we have identified TCP16 as a TF that, besides SPL7, could participate in Cu homeostasis via temporal modulation of gene expression in Arabidopsis.

#### MATERIALS AND METHODS

#### Plant Growth Conditions and Treatments

A. thaliana plants, ecotype Col 0, and the transgenic lines indicated in Supplementary Table SI were grown as previously described (Andrés-Colás et al., 2010). The half-strength Murashige and Skoog (1/<sup>2</sup> MS) medium was either commercial (Sigma) or prepared in the laboratory as follows: macronutrients 12.5 ml (NH4NO<sup>3</sup> 825 mg/l, KNO<sup>3</sup> 950 mg/l, MgSO4·7H2O 90.35 mg/l, KH2PO<sup>4</sup> 85 mg/l, and CaCl<sup>2</sup> 166.25 mg/l), micronutrients 0.5 ml (H3BO<sup>3</sup> 3.1 mg/l, MnSO4·H2O 8.45 mg/l, ZnSO4·7H2O 4.3 mg/l, NaMoO4·2H2O 0.125 mg/l, and CoCl2·6H2O 0.0125 mg/l), Fe-EDTA 2.5 ml (FeSO4·7H2O 13.9 mg/l and Na2EDTA·2H2O 18.63 mg/l), KI 1.1 ml (0.41 mg/l), in both cases supplemented with MES 0.5 g/l, sucrose 10 g/l, agar 8 g/l, pH 5.7 with KOH. Variable CuSO<sup>4</sup> concentrations were added when indicated. Media were supplemented with 100 µM BCS for the Cu-deprived media.

A COPT3 promoter (COPT3p), covering 1,248 bp upstream from the start codon, was fused to the uidA (GUS) reporter gene (COPT3p:GUS) by substitution of the CaMV35S promoter in the pBI121 vector. At least two independent transgenic Arabidopsis stable homozygous lines harboring the COPT3p:GUS chimeric construct were obtained and analyzed.

The phenotype of two independent TRANSPLANTA (TPT) TPT TCP16 lines (Coego et al., 2014) was analyzed on <sup>1</sup>/<sup>2</sup> MS medium with 100 µM BCS, with or without 2 µM β-estradiol after germinating on <sup>1</sup>/<sup>2</sup> MS medium 2 days or for the indicated period, and under long (16 h light-23◦C/8 h dark-16◦C) or neutral (12 h light-23◦C/12 h dark-16◦C) photoperiod conditions, as indicated. For the gene expression analysis by RT-qPCR of TPT TCP16 lines, plants were grown directly with 2 µM β-estradiol. For this expression analysis, one TPT TCP16 line was used and compared to the wild-type. Samples were collected at 0, 12, or 24 h from plants grown under neutral conditions (0 or 24 h, start of light; 12 h, end of light), as indicated.

For the genotyping of the T-DNA insertion lines, plants were self-pollinated and homozygous lines were obtained. PCR (Supplementary Table SIV) or RT-qPCR (Supplementary Table SV) were performed with specific oligonucleotides to genotype or check the loss of expression in the lines, respectively. For the tcp16 mutant, one line was analyzed and compared to the previously obtained TCP16 RNAi line. For the copt3 mutant, one line was analyzed and compared to the COPT3-HA overexpressing line. For the copt5 mutant, one of the previously characterized lines was used as a control for sensitivity to Cu deficiency.

#### Electrophoretic Mobility Shift Assay (EMSA)

Biotin-labeled and unlabeled oligonucleotides (Supplementary Table SII), including the putative TCP binding motifs of COPT3 and COPT5 promoters, and a fragment of the COPT2 promoter (without TCP binding motifs), were synthetized (VWR) and annealed in TEN buffer (10 mM Tris Base pH 7.8, 1 mM EDTA, 0.1 M NaCl) to generate the probes. The purified TCP16 and TCP23 proteins were obtained from the TRANSPLANTA consortium (Coego et al., 2014). Briefly, full-length TCP16 and TCP23 expression constructs were cloned in the destination vector pER8 and mobilized into pDONR201 using BP clonase reaction. cDNAs were transferred to destination vector pDEST-TH1 using LR clonase, yielding Maltose Binding Proteins (MBP) N-terminal fusions and constructs checked by sequencing. MBP-TCP16 and MBP-TCP23 constructs were transformed into BL-21 strain for expression. Induction of bacterial cultures was routinely at 25◦C for 6 h with 1 mM Isopropyl β-D-1-thiogalactopyranoside. Expression of recombinant proteins was assessed by Western blot with an anti MBP antibody (BioLab) (Franco-Zorrilla et al., 2014). MBP-TCP16 and MBP-TCP23 proteins were bound to an amylose resin and eluted using maltose. EMSA was carried out with 960–1320 ng of purified protein, 0.01 pmol labeled probe and 225x unlabeled probe, as indicated, in binding buffer in 10 µl of total reaction volume. Protein buffer (2x TEN, 1 mM DTT, 1 mM protease inhibitor). Binding buffer (20 mM HEPES-KOH pH 7.8, 100 mM KCl, 1 mM EDTA, 0.1% BSA, 10 ng salmon sperm DNA, 10% glycerol). Binding reaction was performed at room temperature for 30 min. Electrophoresis was performed at 32 V on ice in a pre-run 5% native polyacrylamide gel in TBE buffer. Transfer was performed onto a nylon membrane at 40 V for 2 h on ice. Membrane crosslinking was at 120

mJ/cm<sup>2</sup> 45–60 s at 254 nm in Stratalinker. Streptavidin-HRP conjugate antibody (Pr.Nr.21126, Pierce) was used for detection of the labeled probe. Relative bound DNA was quantified using ImageJ 1.42q software<sup>1</sup> . The experiment was repeated at least two independent times.

### Biochemical Fractionation

Plants overexpressing the COPT3-HA fusion protein (Andrés-Colás et al., 2010) were grown on soil. Chloroplasts were isolated and fractionated into stroma and thylakoids from leaves of 3- to 4-week-old plants as described (Pilon-Smits et al., 2002). Samples were normalized based on the number of chloroplasts and chlorophyll content, as determined by the method of Bruinsma (1961) as described (Pilon-Smits et al., 2002). Proteins were quantified by Bradford (1976) assay and separated by native 15% PAGE and then blotted on a nitrocellulose membrane. Immunodetection of SEC12, CpNifS, and PC was used as control of ER, stroma and thylakoids proteins with specific antiserum (Bar-Peled and Raikhel, 1997; Pilon-Smits et al., 2002; Abdel-Ghany et al., 2005). COPT3-HA was detected with anti-HA 3F10 specific antibody (Roche).

For sucrose density gradient fractionation the leaves of 4-week-old plants were ground with mortar and pestle in membrane isolation buffer [20 mM HEPES-KOH, pH 7; 50 mM C2H3KO2; 5 mM EDTA; 250 mM sorbitol, 1 mM DTT plus CompleteTM protease inhibitor cocktail (Roche)] and centrifuged at 2,000 × g for 10 min at 4◦C. 3 ml of the supernatant were applied to the top of continuous 10 ml 20–60% (w/v) sucrose gradients, either with or without 5 mM MgCl<sup>2</sup> added and centrifuged at 150,000 × g for 3 h at 4◦C. Fractions of 0.5 ml were taken from the top and concentrated with TCA. The proteins in the fractions were electrophoresed in 12.5% SDS-PAGE, blotted and immunodetected using antibodies against the HA epitope (3F10, Roche), the ER SEC12, the plasma membrane α-AHA and the mitochondrial PMO35 markers, as described above.

#### Subcellular Localization in Arabidopsis Protoplasts

The complete COPT3 coding sequence was obtained from Arabidopsis genomic DNA by PCR using the following specific primers, which introduce the adequate restriction sites for cloning: C3-SalI F, 5<sup>0</sup> CCACGCGTCG ACATGAACGGCATGAGTGGATC; C3-NcoI R, 5<sup>0</sup> CCATG CCATGGAACAATGTGATTGAACCTCGG. The C-terminus was fused with the GFP reporter and its expression was controlled by the constitutive CaMV35S promoter through its insertion into the transient expression vector pGFPau with the SpeI and SalI restriction enzymes.

The COPT3-GFP construct was used to transform Arabidopsis protoplasts obtained from the fresh leaf tissue of 3-week-old plants grown on soil, as previously described (Abdel-Ghany et al., 2005). After 16 h under continuous light at 23◦C in the wash solution, confocal images were obtained using a fluorescence confocal microscope TCS SP vertical (DM-R) (Leica) equipped with an argon ion (458 and 488 nm), He-Ne I (543 nm) and He-Ne II (633 nm) excitation laser systems and a 60× objective lens. The fluorescence signals were detected at 500–530 nm for GFP and at 650–750 nm for chlorophyll, after exciting at 488 and 633 nm, respectively.

#### GUS Staining and Pollen Preparations for Scanning Electron Microscopy

Assays were performed as described (Jefferson et al., 1987). Briefly, the organs were embedded with the substrate solution [100 mM NaPO<sup>4</sup> pH 7.2, 0.5 mM K3Fe(CN)6, 0.5 mM K4Fe(CN)6, 0.1% (v/v) Triton X-100, 0.5 mM 5-bromo-4-chloro-3-indolyl-β-D-glucuronide (X-Gluc, AppliChem) and 10 mM EDTA pH 7.2]. Reactions took place at 37◦C.

Pollen was mounted on standard stubs and coated with gold-palladium in a Bio-Rad E5600 ion sputter for 3 min prior to observation on a Hitachi S4100 FE scanning electron microscope. Digital images were acquired with the application EMIP.

### Cu Content Measurements

Cu content was determined by atomic absorption as described (Andrés-Colás et al., 2006; Carrió-Seguí et al., 2015) at the "Servei Central de Suport a la Investigació Experimental SCSIE" (Universitat de València) and the "Servicios Centrales de Investigación" (Universidad de Almería).

### Gene Expression by RT-PCR

Total RNA was isolated from A. thaliana seedlings with trizol reagent (Ambion). RNA was quantified by UV spectrophotometry and its integrity was visually assessed on ethidium bromide-stained agarose gels. Total RNA (1.5 µg) was first converted into cDNA by reverse transcription (RT) using SuperScript II reverse transcriptase (Invitrogen) and anchored oligo(dT)<sup>15</sup> (Roche) and 18S reverse primer. PCR was performed under the following conditions to maintain a linear response in the range of the cDNA concentrations used (see Supplementary Table SIV for the primer-specific sequences): 30 cycles, except 20 cycles for 18S, of three temperature segments of 30 s (Td 94◦C/Th 55◦C/Te 72◦C). The PCR products were visualized in 2% agarose gels. Real-time PCRs (qPCR) were carried out with SYBR-Green qPCR Super-Mix-UDG with ROX (Invitrogen) and specific oligonucleotides (Supplementary Table SV) in a StepOnePlus Real-Time PCR System (Applied Biosystems) under 1 cycle of 95◦C for 2 min and 40 cycles consisting in 95◦C for 30 s and 60◦C for 30 s. The results correspond to the comparative Ct (cycle threshold) method (11Ct). The UBQ10 gene was used as a loading control. Values are relative expression with respect to the first sample in each graph, in arbitrary units.

#### Computer-Assisted Sequence

The theoretical promoter sequences analysis was performed by Patmatch from TAIR<sup>2</sup> .

<sup>1</sup>https://imagej.nih.gov/ij/

<sup>2</sup>www.arabidopsis.org

### RESULTS

#### COPT3 Protein Was Intracellular Localized

The COPT3 gene (At5g59040) is located on chromosome V of the A. thaliana genome, adjacent to COPT1. The two genes are organized head-to-head in opposite orientations separated by 2,266 bp (Supplementary Figure S1A). The COPT3 gene encodes a 151 amino acid protein that displays features, which are conserved in COPT/CTR-type transport proteins. These conserved features include three transmembrane domains (TMDs) with an external amino terminus, containing a conserved Met residue and a cytosolic carboxy terminus, as well as the Mx3M and the Gx3G motifs within TMD2 and TMD3, respectively (Peñarrubia et al., 2010; Puig, 2014).

The 1.4 kb promoter region of the COPT3 gene contains a number of potential cis regulatory elements (Supplementary Figure S1B). One of these is a putative plastid expression box at position −388. Sequence analysis of the coding region suggested that a putative transit sequence for targeting to the chloroplast may be present in COPT3 (PSORT<sup>3</sup> ) (Supplementary Figure S1C). In order to analyze its subcellular localization, chloroplasts were isolated from an Arabidopsis transgenic line expressing the COPT3 coding sequence tagged with the human influenza hemagglutinin (HA) epitope under the control of the 35S cauliflower mosaic virus (CaMV35S) promoter (COPT3-HA) (Andrés-Colás et al., 2010). The analysis of isolated chloroplast fractions clearly indicated that COPT3 was not present in plastids (Supplementary Figures S2A,C).

Next, COPT3 subcellular localization was analyzed by transient expression in Arabidopsis protoplasts of the COPT3 coding region tagged with the green fluorescence protein (GFP) under the control of the CaMV35S promoter (COPT3-GFP) (Supplementary Table SI). The signal obtained confirmed an intracellular localization of COPT3 excluding the plasma membrane and chloroplasts (**Figure 1A**). Moreover, sucrose density gradient fractionation of membranes from leaves of plants expressing the COPT3-HA construct indicated that the COPT3 distribution pattern is more similar to the ER protein marker SEC12 than to the other markers, such as the mitochondrial PMO35 protein or the plasma membrane α-AHA protein (Supplementary Figures S2B,D). These results point to a putative COPT3 localization in the endomembrane system, maybe in the ER.

#### COPT3 Was Mainly Expressed in Pollen and Vascular Bundles

The analysis of the COPT3 promoter (Supplementary Figure S1B) indicated the presence of several GTGA boxes (7), described in the late pollen g10 gene promoter (Rogers et al., 2001). Additionally, one of the two co-dependent elements responsible for the pollen-specific activation of tomato LAT52 gene (Bate and Twell, 1998), the AGAAA element, were present at multiple positions (12). There were also several boxes for expression in

<sup>3</sup>https://psort.hgc.jp/

FIGURE 1 | COPT3 expression pattern and regulation by Cu. (A) COPT3 subcellular localization. Arabidopsis protoplasts isolated from 30-day-old leaves and transiently transformed with the 35S:COPT3-GFP construct or the 35S:GFP as a control, were analyzed by confocal microscopy at 16 h post transformation. Green and red fluorescence are indicative of the localization of the GFP protein and the chlorophyll, respectively. Representative protoplasts from at least two independent experiments are shown on the same scale, including their merged and light fields. (B) Histochemical β-glucuronidase staining of the Arabidopsis lines transformed with the COPT3p:GUS construct. A representative flower (left), mature anthers and filaments (middle) and seedling shoot (right) from at least two independent lines are shown. (C) COPT1 and COPT3 temporal expression pattern and regulation by Cu. Seven-day-old wild-type seedlings grown under Cu deficiency (1/<sup>2</sup> MS with 100 µM BCS; –Cu) and Cu sufficiency conditions (1/<sup>2</sup> MS with 1 µM CuSO4; Cu). Samples were taken at 0 h (dark bars) and 12 h (light bars) of the 12 h light/12 h dark cycle. Total RNA was extracted and analyzed by RT-qPCR with specific oligonucleotides for COPT1 and COPT3. The relative expression in arbitrary units is represented. Values correspond to arithmetic means (2−11Ct) ± standard deviation from three biological replicates with three technical replicates for each (n = 3). Asterisks indicate significant differences for the same group of samples with respect to the 0 h sample (P < 0.05) according to T-test.

embryo (3), endosperm (4) and aleurone (2) (Supplementary Figure S1B).

In order to determine the tissue expression pattern of COPT3 throughout the plant, we first performed RT-PCR analysis from various organs of adult plants. The result indicated that COPT3 expression was detected in flowers and dried seeds, but also in stems and leaves, whereas it was hardly detectable in roots (Supplementary Figure S3A). These results were consistent with the Genevestigator database<sup>4</sup> , which also indicated that overall

<sup>4</sup>https://genevestigator.com/gv/doc/intro\_plant.jsp

levels of COPT3 expression were low throughout the plant life cycle compared to other COPT family members.

To further study the COPT3 spatial expression pattern, stable transgenic Arabidopsis lines harboring the GUS reporter gene driven by the COPT3 promoter (COPT3p:GUS) were obtained (Supplementary Table SI). During the development of reproductive organs, GUS staining was observed in anthers with a strong signal in pollen (**Figure 1B**, left). Accordingly, GUS staining was detected in the stamen filaments only when the styles were elongating (**Figure 1B**, middle). Moreover, GUS staining was also detected in the leaf vascular bundles (**Figure 1B**, right).

#### COPT3 Expression Was Up-Regulated at Dusk and Down-Regulated by Cu

The COPT3 promoter also displayed regulatory elements conserved in light regulated genes, denoted as I-box (GATAA), as well as an element (CAANNNNATC) required for the tomato LHC circadian expression and an Evening Element (AAAATATCT) involved in circadian regulation (Terzaghi and Cashmore, 1995; Harmer et al., 2000; Rawat et al., 2005) (Supplementary Figure S1B). Moreover, based on the DIURNAL DataBase<sup>5</sup> (Mockler et al., 2007), COPT3 expression oscillated with a phase of around 24 h under different circadian and diurnal conditions and peaked at 12 h (end of light period) of the 12 h light/12 h dark neutral photoperiod cycle (Supplementary Figure S3B). Furthermore, it was previously shown that altered Cu transport, through COPT1 and COPT3 overexpression, affected circadian rhythms regulation (Andrés-Colás et al., 2010). Taken together, these data may indicate a COPT3 temporal regulation. As a first approach to address its study, COPT3 expression was tested at 0 and 12 h in 7-day-old seedlings grown on neutral photoperiod conditions. The result confirmed the higher COPT3 expression at 12 h (end of the light period), opposite to COPT1 expression that was higher at 0 h (start of the light period), under Cu deficiency (**Figure 1C**). This result confirmed our previous data on the temporal expression of other pmCOPT members, such as COPT2 and COPT6, both peaking at dawn (Perea-García et al., 2016a), whereas the other imCOPT, COPT5, peaked at dusk (not shown). These data are also in agreement with data in the DIURNAL DataBase for the pmCOPT and imCOPT expression, where both subfamily member types oppositely oscillated during the diurnal cycle.

Moreover, the COPT3 promoter displayed three putative Cu deficiency response elements (GTAC) (Supplementary Figure S1B), previously described in the promoters of Cu-deficiency regulated genes that may be target sites for SPL7 (Yamasaki et al., 2009; Bernal et al., 2012). Although COPT3 expression was previously reported to be independent of the Cu levels (Sancenón et al., 2003), the presence of these elements suggested a possible up-regulation of COPT3 under Cu deficiency. In order to check if COPT3 was differentially regulated by Cu over day and night, samples of seedlings grown under Cu deficiency and excess were checked by RT-qPCR at 0 and 12 h (**Figure 1C**). COPT3 expression was significantly increased under Cu deficiency, specifically at 12 h (end of the light period). Furthermore, a wide range of Cu concentrations were tested and, in general, higher COPT3 expression levels were observed in Cu deficiency media when compared to Cu sufficiency or Cu excess (Supplementary Figure S3C). However, the temporal specific increase of COPT3 expression under Cu deficiency at 12 h, did not correlate to the expression peak of SPL7 at 0 h (Perea-García et al., 2016a).

### Two Independent TPT TCP16 Lines Were Sensitive to Cu Deficiency

In order to find regulatory factors involved in the temporal pattern of COPT3 expression, a screen of a conditional overexpression TF library, denoted TRANSPLANTA (TPT) (Coego et al., 2014), was performed under Cu deficiency conditions. The TRANSPLANTA collection contains 634 Arabidopsis TFs transferred into a vector (pER8) that conferred a β-estradiol-inducible gene overexpression (Zuo et al., 2000). At least, two independent single insertion and homozygous transgenic lines were generated for each TF (Coego et al., 2014). Reporter lines under the same promoter (pER8G:GUS-GFP) were used to optimize the expression conditions in our experimental set up (Supplementary Figure S4A). The treatment with 2 µM β-estradiol induced reporter expression after 12 h and treatment with 100 µM BCS did not modify GUS expression (Supplementary Figure S4B). Based on these results, a screen was performed in 7-day-old seedlings germinated on <sup>1</sup>/<sup>2</sup> MS medium and then grown on 100 µM BCS with 2 µM β-estradiol. Afterward, seedlings were transferred to fresh plates containing 10 µM Cu and 2 µM β-estradiol for checking the recovery of root growth with the aim to discard lines with phenotypes unrelated to Cu deficiency. The copt5 mutant line, which exhibited a defect in root elongation under Cu deficiency (Garcia-Molina et al., 2011) and WT seedlings were used as controls (Supplementary Figure S4C).

One of the TF families in the TPT lines with a higher percentage of the members showing a short root phenotype in the screen under Cu deficiency, which was reverted under high Cu, was the TCP family (not shown). The TRANSPLANTA collection contains TPT lines for 17 TCPs out of a total of 24 members present in the Arabidopsis genome. TPT lines from six of them (TCP14, TCP16, TCP19, TCP20, TCP22, and TCP24) displayed a short root phenotype under Cu deficiency, which reverted under high Cu (**Table 1**). All of them belong to the TCP class I PCF, except TCP24. It is noteworthy that most TCP factors display putative Cu deficiency responsive GTAC boxes in their proximal promoters (500 bp), except TCP4 and TCP17 (**Table 1**). Thus, TPT lines positive in the screening corresponded to TCPs that all contain GTAC boxes. Among them, TCP16 (At3g45150) was chosen for further study since the similar curly leaves phenotype found in plants where TCP16 was fused to a repressor domain (Uberti-Manassero et al., 2016) and in plants overexpressing COPT1 and COPT3 (Andrés-Colás et al., 2010; Garcia-Molina et al., 2013).

Two independent TPT TCP16 lines, TPT 3.45150.1B (TPT TCP16-B) and TPT 3.45150.1I (TPT TCP16-I) (Supplementary Table SI), were sensitive to Cu deficiency, as shown by a reduced root elongation (**Figure 2A**). The reduced root length

<sup>5</sup>http://diurnal.mocklerlab.org/

TABLE 1 | TCPs characteristics and screening results.


TRANSPLANTA, available TPT lines (+) or not (−). Screening, germinated in <sup>1</sup>/<sup>2</sup> MS medium, grown in 100 µM BCS and recovered with 10 µM Cu after 100 µM BCS. Class I/II CAREs, presence (+) of class I/II TCP CAREs as indicated. Other CARE versions, presence of different versions of the TCP motifs as indicated. CuRE, number of GTAC motifs in the first 500 bp upstream (<sup>∗</sup> , 3 GTAC motifs along 65 bp). Motifs were considered in the first 500 bp upstream and in both strains.

was specifically linked to the Cu deficiency since the root length defect was not observed in the presence of excess Cu (**Figure 2**). Moreover, this phenotype was indeed due to the induced overexpression of TCP16 since it was not observed in the TPT TCP16 lines in the absence of β-estradiol (Supplementary Figure S4C).

#### The TCP16 Transcription Factor Bound the COPT3 Promoter

The promoters of genes participating in Cu homeostasis were analyzed for the presence of putative cis CARE elements recognized by TCPs. This indicated that COPT3 and COPT5 were the only members of the COPT family that displayed putative CAREs (Supplementary Table SII). These CARE elements in the COPT3 (TTGAGCCCAT) and COPT5 (GTGAGCCCAC) promoters where identified as a specific version of the previously described TCP16 CARE (Martín-Trillo and Cubas, 2010).

In order to check if TCP16 had a direct effect on imCOPTs regulation, binding to the promoter regions containing the putative CARE elements in COPT3 and COPT5 promoters (Supplementary Table SIII) was analyzed by Electrophoretic Mobility Shift Assay (EMSA) with the purified TCP16 protein (provided by the TRANSPLANTA consortium) (Coego et al., 2014). TCP16 interacted with the COPT3 promoter, as shown by a retarded COPT3 probe band in the presence of the TCP16 protein (**Figure 3**). An excess of COPT3 unlabeled probe reduced the TCP16 binding, as shown by a lower intensity of the retarded COPT3 probe band in the presence of the TCP16 protein. However, with the same amount of the COPT2 unlabeled probe (the COPT2 promoter has no CARE; Supplementary Table SIII), a minor reduction was observed, pointing to the specificity of the TCP16 interaction with the COPT3 promoter. As well as COPT3, TCP16 also bound the COPT5 promoter and specifically compete with an excess of the COPT5 but not of the COPT2 unlabeled probe (**Figure 3**). TCP23 is a class I TCP member involved in plant development (Balsemão-Pires et al., 2013) that resulted negative in the TPT screening (**Table 1**). Under the same experimental conditions, no interactions of TCP23 with the COPT3 (Supplementary Figure S5A) and the COPT5

promoters (Supplementary Figure S5B) were detected. Taken together, the TCP16 TF specifically bound to the COPT3 and COPT5 promoters in vitro.

#### TCP16 Was Involved in Repression of COPT3 Expression

To check the effect of TCP16 on COPT3 expression, TCP16 and COPT3 mRNA levels were determined in 7-day-old TPT TCP16 seedlings at different times after β-estradiol induction. Shortterm kinetics indicated that TCP16 expression levels increased 4–5 times after 24 h induction in both lines (**Figure 4A**). In parallel to the increase in TCP16 expression at 24 h after induction, COPT3 expression levels were reduced to 25–15% (**Figure 4B**). Besides, long-term TCP16 overexpression by β-estradiol was dependent on the diurnal time, being higher at 12 h of the neutral photoperiod cycle (**Figure 4C**) and provoked a repression of COPT3 expression, specifically at this time of the

day, under Cu deficiency (**Figure 4D**). As a control, the GFP-GUS line showed similar levels of GUS overexpression at 0 and 12 h of the neutral photoperiod cycle (Supplementary Figure S4A) and other TPT lines showed a robust β-estradiol-dependent induction of the TF transgene although to a different extent depending on the line (Coego et al., 2014). These data pointed to a repressive TCP16 role on COPT3 expression.

experiments are shown.

### TCP16 Expression Was Up-Regulated at Dawn and by Cu

Since the TCP16 promoter contains a putative SPL7-responsive GTAC box (**Table 1**), we checked its expression in WT seedlings in the same samples that were used for COPT3 expression analysis (**Figure 1C**). The results indicated that TCP16 expression was higher at 0 h than at 12 h of the 12 h light/12 h dark cycle, under Cu deficiency (**Figure 5A**). These results may point to a diurnal oscillation of TCP16 expression opposite to the one shown by COPT3. Unfortunately, the TCP16 expression pattern is unavailable in the DIURNAL DataBase to be compared with the results shown here. Moreover, TCP16 expression was significantly higher under Cu excess at 12 h with respect to Cu deficiency, thus high transcript levels remained along day and night under Cu excess (**Figure 5A**), coinciding with the reduction in COPT3 expression (**Figure 1C**). Taken together, these data pointed to a role for TCP16 as a temporal transcriptional repressor of COPT3, specially at dawn under Cu deficiency and along day and night under Cu excess.

Cu levels were determined in the conditionally overexpressing TPT TCP16-B and TPT TCP16-I lines (**Figure 5B**). TPT TCP16 lines had lower Cu content than WT under Cu deficiency. This result pointed to affected Cu uptake. To test this possibility, the expression of pmCOPTs (COPT1 and COPT2) was analyzed. Both

pmCOPT were selectively reduced at 12 h in TPT TCP16-B under Cu deficiency conditions (**Figures 5C,D**), despite the absence of TCP16 binding boxes in their promoters (Supplementary Table SII).

### The Loss-of-Function of TCP16 Exhibited Copper-Related Phenotypes

In order to have a better understanding of the role of TCP16 in regulating Cu homeostasis, we used both a RNA interference line TCP16RNAi (Takeda et al., 2006) and a T-DNA insertion mutant tcp16 (N462818) (Supplementary Table SI). The tcp16 mutant contains the T-DNA insert at +110 bp of the TCP16 coding sequence (Supplementary Figures S6A,B). A homozygous tcp16 line was selected (Supplementary Figure S6C) and the loss of the TCP16 expression at 0 h corroborated by RT-qPCR (Supplementary Figure S6D). The growth of tcp16 was checked under Cu deficiency and excess in the medium.

FIGURE 5 | TCP16 temporal expression pattern, regulation by Cu, and Cu status in the TCP16 TPT lines. (A) TCP16 expression. Six-day-old wild-type seedlings grown in <sup>1</sup>/<sup>2</sup> MS medium (MS) and the same medium supplemented with 100 µM BCS (–Cu) or 10 µM CuSO<sup>4</sup> (+Cu). Samples were taken at 0 h (dark bars) and 12 h (light bars) of the 12 h light/12 h dark cycle. Total RNA was extracted and analyzed by RT-qPCR with specific oligonucleotides for TCP16. The relative expression in arbitrary units is represented. Values correspond to arithmetic means (2−11Ct) ± standard deviation from at least three biological replicates (n ≥ 3). Asterisk indicates significant differences for the same group of samples with respect to the other conditions (P ≤ 0.002) according to T-test. (B) Cu content in TPT TCP16 plants. Seven-day-old wild-type (white bars), TCP16-B (dark gray bars) and TCP16-I (light gray bars) TPT seedlings grown in <sup>1</sup>/<sup>2</sup> MS medium supplemented with 100 µM BCS (–Cu) or with 10 µM CuSO<sup>4</sup> (+Cu), in the presence of 2 µM β-estradiol, under 12 h light/12 h dark cycles. Values correspond to arithmetic means ± standard deviation from three biological replicates of 120 mg of DW. Asterisks indicate significant differences between WT and mutant lines in the same condition (P ≤ 0.05) according to T-test. (C,D) Cu-status marker expression in TPT TCP16 line. Six-day-old wild-type (WT) and TPT TCP16-B seedlings grown in <sup>1</sup>/<sup>2</sup> MS medium supplemented with 100 µM BCS and 2 µM β-estradiol. Samples were taken at 0 h (dark bars) and 12 h (light bars) of the 12 h light/12 h dark cycle. Total RNA was extracted and analyzed by RT-qPCR with specific oligonucleotides for COPT1 (C) and COPT2 (D). The relative expression in arbitrary units is represented. Values correspond to arithmetic means (2−11Ct) ± standard deviation from four biological replicates (n = 4). Asterisks indicate significant differences for the same group of samples with respect to the WT line (∗P ≤ 0.02; ∗∗P ≤ 0.005; ∗∗∗P ≤ 0.0005) according to T-test.

The tcp16 seedlings showed defects in root elongation and fresh weight mostly under Cu excess (**Figures 6A,B**), coincident with the conditions where TCP16 was mainly expressed (**Figure 5A**). Moreover, Cu content was also determined by atomic absorbance and, whereas conditionally overexpressing TPT TCP16 lines had lower Cu content than WT under Cu deficiency (**Figure 5B**), a decreased level of Cu content was observed in the loss-of-function TCP16RNAi and tcp16 lines under Cu excess (**Figure 6C**).

To check the effect of the TCP16 loss-of-function on COPT3 expression, COPT3 mRNA levels were determined in 7-day-old seedlings of the TCP16RNAi and tcp16 lines under Cu excess. An increase in COPT3 expression levels was observed specifically at

FIGURE 6 | Phenotype of tcp16 seedlings under different Cu status and COPT3 expression in TCP16 RNAi and knockout lines. (A) Photographs of representative 6-day-old wild-type (WT) and tcp16 seedlings grown in <sup>1</sup>/<sup>2</sup> MS medium supplemented with 100 µM BCS (–Cu) or 10 µM CuSO<sup>4</sup> (+Cu) under 12 h light/12 h dark cycles. A representative image of at least three independent experiments is shown. Scale bars, 1 cm. (B) Fresh weight of wild-type (white bars) and tcp16 (dark gray bars) 7-day-old seedlings grown in the same conditions shown in (A). Bars correspond to means ± standard deviation of six biological replicates of five seedlings each. Asterisks indicate significant differences (∗P < 0.05; ∗∗P < 0.01) according to Duncan test. (C) Cu content in WT (white bars), TCP16 RNAi (dark gray bars) and tcp16 (light gray bars) 7-day-old seedlings grown in the same conditions shown in (A). Bars correspond to means ± standard deviation of three biological replicates of 120 mg of DW. Asterisks indicate statistical differences between WT and mutant lines in the same condition (P < 0.05) according to T-test. (D) COPT3 expression. Six-day-old wild-type (WT), TCP16 RNAi and tcp16 seedlings grown in <sup>1</sup>/<sup>2</sup> MS medium supplemented with 10 µM CuSO4. Samples were taken at 0 h (dark bars) and 12 h (light bars) of the 12 h light/12 h dark cycle. Total RNA was extracted and analyzed by RT-qPCR with specific oligonucleotides for COPT3. The relative expression in arbitrary units is represented. Values correspond to arithmetic means (2−11Ct) ± standard deviation from at least three biological replicates (n ≥ 3). Asterisks indicate significant differences for the same group of samples with respect to the WT line (∗P ≤ 0.06; ∗∗P ≤ 0.009) according to T-test.

the end of the light period (12 h) (**Figure 6D**). This effect was the opposite to the one observed in the conditionally overexpressing TPT TCP16-B line (**Figure 4D**), further pointing to the function of TCP16 as a repressor of COPT3.

#### COPT3 Function Was Required for Proper Repression of TCP16

The possibility that COPT3 function could reciprocally affect TCP16 expression was also checked. For that purpose, we analyzed the TCP16 expression in transgenic lines with altered COPT3 expression levels and used the COPT3-HA line (Andrés-Colás et al., 2010) and loss-of-function copt3 mutant (GK633G06) (Supplementary Table SI). The copt3 mutant contains a T-DNA insertion at +109 bp of the COPT3 coding sequence (Supplementary Figure S7A). A homozygous copt3 line was selected (Supplementary Figure S7B) and the loss of the COPT3 expression in flowers corroborated by RT-qPCR (Supplementary Figure S7C). The COPT3-HA line was sensitive to Cu excess, as shown by a reduction in root elongation and alteration of the flower morphology and flowering time (Andrés-Colás et al., 2010). On the contrary, the copt3 mutant was more sensitive to Cu deficiency than WT and accordingly accumulated less Cu than controls (not shown). The levels of the Cu deficiency marker COPT2 were higher at 12 h in the copt3 mutant (**Figure 7A**), whereas SDH1-2, a mitochondrial marker of Cu excess (Andrés-Colás et al., 2013), was significantly reduced at this time in flowers (**Figure 7B**) further pointing to decreased Cu levels in the copt3 mutant.

The expression analysis showed that TCP16 levels were higher at 12 h in the copt3 mutant line, in contrast with a lower expression at 0 h in the COPT3-HA transgenic line (**Figure 7C**).

FIGURE 7 | Phenotype of copt3 knockout seedlings under different Cu status. (A,B) Cu-status marker expression in copt3 lines. Flowers from wild type (WT) and copt3 knockout plants grown in soil. Samples were taken at 0 h (dark bars) and 12 h (light bars) of the 12 h light/12 h dark cycle. Total RNA was extracted and analyzed by RT-qPCR with specific oligonucleotides for COPT2 (A) and SDH1-2 (B). The relative expression in arbitrary units is represented. Values correspond to arithmetic means (2−11Ct) ± standard deviation from at least three biological replicates (n ≥ 3). Asterisk indicates significant differences for the same group of samples with respect to the WT line (∗P ≤ 0.04; ∗∗P ≤ 0.03) according to T-test. (C) TCP16 expression in copt3 and COPT3OE lines. Six-day-old wild-type (WT), copt3 and COPT3OE seedlings grown in <sup>1</sup>/<sup>2</sup> MS medium supplemented with 100 µM BCS. Samples were taken at 0 h (dark bars) and 12 h (light bars) of the 12 h light/12 h dark cycle. Total RNA was extracted and analyzed by RT-qPCR with specific oligonucleotides for TCP16. The relative expression in arbitrary units is represented. Values correspond to arithmetic means (2−11Ct) ± standard deviation from at least two biological replicates (n ≥ 2). Asterisks indicate significant differences for the same group of samples with respect to the WT line (∗P ≤ 0.15; ∗∗P ≤ 0.005) according to T-test.

These data suggested that, in addition to the already mentioned repression of TCP16 over COPT3, a reciprocal repressing effect of the COPT3 function on TCP16 expression was taking also place.

In agreement with the observation that COPT3 expression was higher in pollen, the copt3 mutant showed a significantly higher percentage of pollen ornamentation defects than the WT plants, but only under Cu deficient conditions (**Figure 8**). TCP16 was also highly expressed in pollen and abortion of early pollen development has been shown in the TCP16RNAi plants (Takeda et al., 2006). These results reinforced the relevance of Cu homeostasis for pollen viability and the temporal and reciprocal regulation established between TCP16 and COPT3, necessary to fully accomplish this crucial process in plants.

#### DISCUSSION

COPT1 and COPT3 transporters belong to two subfamilies (pmCOPT and imCOPT, respectively) involved in Cu<sup>+</sup> uptake from different extra- and intra-cellular pools. The COPT1 and COPT3 are two flanking genes organized head-to-head in opposite orientations. In general, bidirectional activity was shown to be an inherent feature of most promoters, being especially relevant in divergent promoters (Seila et al., 2008; Wakano et al., 2012). In Arabidopsis, 5,763 divergent gene pairs were reported (Krom and Ramakrishna, 2008) and among them, 462 are separated by a small distance (<250 bp) sharing a single bidirectional promoter that may regulate the co-expression of the two genes (Dhadi et al., 2009). The intergenic region between transcriptional start sites of COPT1 and COPT3 is 2,266 bp (Supplementary Figure S1A), a similar size to the 2,177 DNA segment between the genes cab1 and cab2 which was shown to function as a bidirectional promoter (Mitra et al., 2009). It is thus possible, that COPT1 and COPT3 spatial expression is co-regulated via the shared bidirectional promoter. Indeed, both were mostly present in pollen, seeds and vascular bundles (**Figures 1B** and Supplementary Figure S3A) (Sancenón et al., 2004; Bock et al., 2006). However, whereas COPT3 was expressed early in pollen development, COPT1 was expressed at later stages (Bock et al., 2006). Indeed, Cu is highly required for pollen development and its regulated delivery through COPT transporters could be an important step. Both SPL7 and a Cu-DEFICIENCY-INDUCED TRANSCRIPTION FACTOR 1 (CITF1) belonging to the bHLH family (bHLH160) were recently shown to participate in the regulation of Cu delivery to the anthers and in jasmonic acid synthesis during Cu deficiency (Yan et al., 2017).

Our data indicated that COPT3 might be located in a compartment of the secretory pathway, which could be the ER, where COPT3 would recover Cu<sup>+</sup> from the ER lumen. Although further experimental approaches are required to localize COPT3 to a precise organelle, it is interesting to note that the ER was also the proposed location of the Cu deficiency sensor SPL7 (Garcia-Molina et al., 2014). SPL7 was proposed to sense both cytosolic and the ER lumen Cu status. Our COPT3 localization data pointed that COPT3 could be interestingly involved in the partitioning of these two differential Cu pools (**Figure 9**).

The fact that the COPT3 transcriptional expression pattern was initially described as not being affected by Cu status in the medium (Sancenón et al., 2003), could be attributed to the particular temporal dependence of the Cu-regulation of COPT3 expression. Both COPT1 and COPT3 present several GTAC elements nearby the translational start sites that could be involved in SPL7-mediated Cu deficiency responses (Yamasaki et al., 2009; Bernal et al., 2012). In regard to this, the expression in phase of both the activator SPL7 and the target COPT1 could drive a robust Cu deficiency response for COPT1 expression, whereas the antiphase expression between the SPL7 and the target COPT3 could temporally affect the intensity of the SPL7-mediated regulation, as already modeled (Peñarrubia et al., 2015). This result suggested that, at the temporal level, the promoter region could be alternatively used in each direction for COPT1 and COPT3 transcription, instead of co-expression. However, the SPL7-mediated antiphase regulation of COPT3 expression cannot explain per se the anti-correlated expression observed at the temporal level between COPT1 and COPT3 driven from a putative bidirectional promoter (**Figure 1C**), but rather

suggested the presence of other temporal transcriptional regulators.

Thanks to the TRANSPLANTA consortium, we could screen for conditionally inducible TFs that might participate in the response to Cu availability (Coego et al., 2014). The TCP family was one of the most overrepresented with regard to the number of positive members in the screen. Moreover, the presence of putative elements denoted CAREs in the COPT3 and COPT5 promoters, as well as in other Cu homeostasis components (Supplementary Table SII), is compatible with a regulatory role of TCPs in Cu homeostasis. We selected the TCP family for further study, because pollen morphology was also affected in plants with altered levels of COPT1, COPT3, and TCP16 (Sancenón et al., 2004; Takeda et al., 2006; this study). Moreover, CHE is a TCP factor that was involved in CCA1 repression by interacting with TOC1, both components of the circadian clock (Pruneda-Paz et al., 2009), which expression was altered in COPT1 overexpressing plants (Andrés-Colás et al., 2010). COPT1 and COPT3 overexpression displayed similar phenotypes that were attributed to the temporal deregulation of the Cu entrance that could affect the circadian rhythms (Andrés-Colás et al., 2010). This fact could explain the phenotype observed for the COPT3 overexpressing and copt3 seedlings under Cu deficiency in this work. Finally, in microarrays analysis performed in transgenic plants with modified levels of COPT2, differential expression of several TCP members was observed (Perea-García et al., 2013).

There are several lines of evidence to support the model that TCP16 represses COPT3 expression. First, TCP16 specifically bound to the COPT3 and COPT5 promoters in vitro as shown by EMSA (**Figure 3**). And second, COPT3 was repressed as TCP16 expression increased (**Figure 4**) and conversely, COPT3 was upregulated in a tcp16 mutant (**Figure 6**). The presence of a putative TCP16 binding site nearby the translational start site of COPT3, but not of COPT1, could account for the particular temporal repression of COPT3. The relative short distance (109 bp) between the GTAC boxes and the putative TCP16 binding site in the COPT3 promoter (Supplementary Figure S1) brings to discussion if there is a competence between SPL7 and TCP16 for binding. Any kind of interaction between the activation and the repression function of SPL7 and TCP16, respectively, is also plausible since TCPs interact with a wide variety of other TF families, including other SPL member in the described TCP4-SPL9 temporal interaction taking place during flower development (Rubio-Somoza et al., 2014; Bemer et al., 2017). Furthermore, as a conclusion of the COPT3 expression analysis, TCP16 could act as a repressor of COPT3-mediated Cu transport in a time specific manner.

Under Cu deficient conditions, we hypothesized that COPT3-mediated Cu recovery from the secretory pathway was repressed by TCP16 at 0 h, while the pmCOPT would participate in the uptake extracellular Cu at this time being activated by SPL7 (**Figure 9**). Under Cu excess, pmCOPTs were not activated by SPL7 and TCP16 will further repressed COPT3 and COPT5 along day and night (**Figure 9**). On the

other hand, the expression of the pmCOPT transporters at 12 h could be subjected to a feedback autoregulatory loop that was proposed to act as a biochemical oscillator (Peñarrubia et al., 2010). Moreover, COPT3 and TCP16 were mutually repressing each other's expression. TCP16 may directly act as a COPT3 repressor, whereas COPT3 probably indirectly affected TCP16 expression. Subsequently, Cu entrance from extracellular pools was prioritized at dawn and Cu mobilization from internal pools was favored at dusk. Whether this affected cytosolic Cu or the destiny of Cu coming from the different pools was being used for separate purposes still deserves further investigation. In addition, HMA5 and RAN1 also display putative CARE elements in their promoters (Supplementary Table SII) that could indicate that Cu transport in both directions (entrance and exit) through internal membranes was under control of TCPs.

The fine regulation exerted by TCP16 as a repressor of Cu entrance from internal stores specifically at dawn suggested a temporal division requirement for incompatible processes. Among the possibilities, the avoidance of a putative excessive increase in oxidative stress in this period that could not be properly counteracted. In this sense, Cu<sup>+</sup> uptake through COPT transporters imposed an increased oxidative stress (Rodrigo-Moreno et al., 2013) that could interfere at multiple cellular processes and damage biological structures (Ravet and Pilon, 2013). The redox state of the cell was shown to influence the DNA binding ability of class I TCP proteins (Viola et al., 2013). The oxidation of a conserved cysteine residue (C-20) leaded to the formation of intermolecular disulfide bonds that cannot bound target promoters (Viola et al., 2016). Although the C-20 residue is not conserved, a single cysteine (C-107) residue is present in TCP16 (Supplementary Figure S6A).

On the other hand, TCP16 repression could be aimed to protect a Cu sensitive process operating at dawn. In this sense, a specially Cu sensitive process that takes place in the mitochondria is the Fe-S cluster assembly, required for multiple processes including the respiratory electron transfer chain (Brancaccio et al., 2017). Since the mitochondrial matrix contains a labile Cu<sup>+</sup> pool and is also the place where the Fe-S cluster assembly machinery resides, a strict temporal regulated Cu uptake might prevent a blocking of mitochondrial Fe-S protein maturation (Brancaccio et al., 2017). This process is in agreement with the regulatory function of mitochondrial proteins by TCPs (Welchen and Gonzalez, 2006) and is also connecting Fe and Cu homeostasis, as already described for the copt2 mutant (Perea-García et al., 2013). Accordingly, tcp16 seedlings were sensitive to Cu excess (**Figure 6**), maybe due to an impaired temporal Cu entrance to the mitochondria. The source of Cu that reaches organelles from an endosymbiotic origin, such as mitochondria and chloroplasts, remains an unsolved question. Under environmental nutrient deprivation, a putative Cu source is the lumen of the endocytic compartments that were recently shown to participate in dynamic intracellular metal homeostasis (Blaby-Haas and Merchant, 2014; Hong-Hermesdorf et al., 2014). Although further work is needed to confirm this hypothesis, the internal membrane COPT3 and COPT5 transporters might participate in Cu delivery from the secretory pathway to organelles under metal deficiency. In agreement, photosynthesis is affected in copt5 mutants (Garcia-Molina et al., 2011). The mitochondrial SDH1-2 promoter displayed 3 putative CARE elements (Welchen and Gonzalez, 2006) and it was shown to be regulated by TCPs (Giraud et al., 2010). Moreover, SDH1-2 was a good marker for mild Cu excess (Andrés-Colás et al., 2013). In accordance, SDH1-2 expression was down-regulated in the copt3 mutant (**Figure 7B**). This result suggested that a Cu-TCP interplay may mediate mitochondrial SDH1-2 expression. This would constitute a new pathway for gene expression regulation under mild Cu excess that might be aimed to protect mitochondria from Cu toxicity.

The fact that COPT1 and COPT3 were mostly expressed in vascular tissues points to a role for the temporal differences in metal long distance transport. In this sense, since the higher Cu affinity for common metal chelators (Álvarez-Fernández et al., 2014), competition with other metals, such as Fe, in the xylem transport could involve a metal interference in long distance transport, especially relevant under metal deficiencies. Although further work will be needed to address this hypothesis, a putative solution could be a metal differential temporal arrangement in vascular transport.

Finally, the in silico analysis of the hormone-responsive cis-elements present in the promoter sequences (1,000 bp upstream of the five prime untranslated region) from the COPT1 and COPT3 genes indicated a differential hormonal response (Peñarrubia et al., 2015). Major differences were observed for those cis-elements involved in ABA and gibberellic acid (GA) signaling. Whereas the total elements for ABA were 11 and 3, those for GA were 7 and 20 in the COPT1 and COPT3 promoters, respectively (Peñarrubia et al., 2015). Since the antagonism between ABA and GA is well-known (Weiss and Ori, 2007), as well as their interplay with the circadian clock (Atamian and Harmer, 2016) and their wide crosstalk with TCPs (Nicolas and Cubas, 2016), the results shown here underscore the role of phytohormones in the temporal orchestration of metal homeostasis that might control plant development depending on the environmental nutrient conditions.

#### AUTHOR CONTRIBUTIONS

LP conceived the idea and wrote the manuscript. MP, NA-C, and SA-G conceived and performed the COPT3 localization experiments. NA-C performed the TF screening. NA-C and AC-S performed the physiological and molecular experiments in mutant plants.

### FUNDING

This work has been supported by grants BIO2017-87828-C2-1- P (LP) and the TRANSPLANTA Consortium (CSD2007-00057) from the Spanish Ministry of Economy and Competitiveness, and by FEDER funds from the European Union. NA-C and AC-S were recipients of a predoctoral FPI fellowship from the Spanish Ministry of Economy and Competitiveness.

#### ACKNOWLEDGMENTS

fpls-09-00910 June 30, 2018 Time: 17:41 # 14

We acknowledge the SCSIE (Universitat de València) for the sequencing and greenhouse services, Dr. Sergi Puig for critical reading of the manuscript and Drs. Pablo Vera and José Luis Carrasco (IBMCP-UPV València) for providing the

#### REFERENCES


TCP16 and TCP23 proteins for EMSA analysis. TCP16 RNAi line was kindly provided by Chiharu Ueguchi (Takeda et al., 2006).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018.00910/ full#supplementary-material

development and function through transcriptomics. Plant Physiol. 140, 1151–1168. doi: 10.1104/pp.105.074708


proteins with the circadian clock in Arabidopsis thaliana. Plant Cell 22, 3921–3934. doi: 10.1105/tpc.110.074518



**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 © 2018 Andrés-Colás, Carrió-Seguí, Abdel-Ghany, Pilon and Peñarrubia. 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) and the copyright owner(s) 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.

# MtMTP2-Facilitated Zinc Transport Into Intracellular Compartments Is Essential for Nodule Development in Medicago truncatula

Javier León-Mediavilla<sup>1</sup> , Marta Senovilla<sup>1</sup> , Jesús Montiel<sup>1</sup> , Patricia Gil-Díez<sup>1</sup> , Ángela Saez<sup>1</sup> , Igor S. Kryvoruchko<sup>2</sup>† , María Reguera<sup>1</sup> , Michael K. Udvardi<sup>2</sup> , Juan Imperial1,3 and Manuel González-Guerrero1,4 \*

#### Edited by:

Michael A. Grusak, Agricultural Research Service (USDA), United States

#### Reviewed by:

Ulrike Mathesius, Australian National University, Australia Erik Limpens, Wageningen University & Research, Netherlands

#### \*Correspondence:

Manuel González-Guerrero manuel.gonzalez@upm.es

#### †Present address:

Igor S. Kryvoruchko, Department of Molecular Biology and Genetics, Bogazici University, Istanbul, Turkey

#### Specialty section:

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

Received: 15 March 2018 Accepted: 19 June 2018 Published: 10 July 2018

#### Citation:

León-Mediavilla J, Senovilla M, Montiel J, Gil-Díez P, Saez Á, Kryvoruchko IS, Reguera M, Udvardi MK, Imperial J and González-Guerrero M (2018) MtMTP2-Facilitated Zinc Transport Into Intracellular Compartments Is Essential for Nodule Development in Medicago truncatula. Front. Plant Sci. 9:990. doi: 10.3389/fpls.2018.00990 <sup>1</sup> Centro de Biotecnología y Genómica de Plantas (UPM-INIA), Universidad Politécnica de Madrid, Madrid, Spain, <sup>2</sup> Noble Research Institute, Ardmore, OK, United States, <sup>3</sup> Instituto de Ciencias Ambientales, Consejo Superior de Investigaciones Científicas, Madrid, Spain, <sup>4</sup> Escuela Técnica Superior de Ingeniería Agronómica, Alimentaría y de Biosistemas, Universidad Politécnica de Madrid (UPM), Madrid, Spain

Zinc (Zn) is an essential nutrient for plants that is involved in almost every biological process. This includes symbiotic nitrogen fixation, a process carried out by endosymbiotic bacteria (rhizobia) living within differentiated plant cells of legume root nodules. Zn transport in nodules involves delivery from the root, via the vasculature, release into the apoplast and uptake into nodule cells. Once in the cytosol, Zn can be used directly by cytosolic proteins or delivered into organelles, including symbiosomes of infected cells, by Zn efflux transporters. Medicago truncatula MtMTP2 (Medtr4g064893) is a nodule-induced Zn-efflux protein that was localized to an intracellular compartment in root epidermal and endodermal cells, as well as in nodule cells. Although the MtMTP2 gene is expressed in roots, shoots, and nodules, mtp2 mutants exhibited growth defects only under symbiotic, nitrogen-fixing conditions. Loss of MtMTP2 function resulted in altered nodule development, defects in bacteroid differentiation, and severe reduction of nitrogenase activity. The results presented here support a role of MtMTP2 in intracellular compartmentation of Zn, which is required for effective symbiotic nitrogen fixation in M. truncatula.

Keywords: zinc, cation diffusion facilitator, metal transport protein, symbiotic nitrogen fixation, metal nutrition, nodulation

### INTRODUCTION

Zinc (Zn) is an essential nutrient for plants as a cofactor of enzymes or as a structural element (Coleman, 1998; Broadley et al., 2007). Consequently, plants grown in soils with low Zn bioavailability, which include some of the main agricultural areas of the world, have severe growth defects (Alloway, 2008; Marschner, 2012). These include interveinal chlorosis, necrotic leaves, and stunted growth, the result of alterations in the plethora of processes mediated by Zn proteins (Coleman, 1998; Broadley et al., 2007). To prevent this and ensure proper Zn allocation, plants have developed a complex network of transcription factors, transporters, and small Zn-chelating molecules that direct this metal to the proper tissue, cell compartments, and apoproteins (Assunção et al., 2010; Sinclair and Krämer, 2012; Olsen and Palmgren, 2014).

Zinc transport is mediated typically by four different transporter families: the Zrt1/Irt1-like (ZIP) and the yellow stripe-like (YSL) families for transport into the cytosol; and the metal tolerance protein (MTP) and Zn2+-ATPase families for efflux out of the cytosol (DiDonato et al., 2004; Eren and Argüello, 2004; Desbrosses-Fonrouge et al., 2005; Ishimaru et al., 2005; Olsen and Palmgren, 2014). Uptake of Zn from soils in dicots is mediated by ZIP proteins (Korshunova et al., 1999), in a process induced by soil acidification (Pedas and Husted, 2009). Zn can move symplastically from cell to cell and is released from endodermal cells into the xylem, via Zn2+-ATPases (Hussain et al., 2004). YSL transporters are likely candidates to mediate Zn loading into the phloem, as a Zn-nicotianamine complex (Waters et al., 2006). Within cells, Zn is transported into organelles by MTP or Zn2+-ATPases, either to be stored when in excess, or to be used to assemble Zn-proteins (Blaudez et al., 2003; Desbrosses-Fonrouge et al., 2005; Morel et al., 2009). Overall, these processes are regulated by a set of transcription factors that orchestrate Zn homeostasis (Assunção et al., 2010).

While leaves are the main Zn sink in most plants during vegetative growth (Broadley et al., 2007), legumes have an additional one: nitrogen-fixing root nodules (González-Guerrero et al., 2014, 2016). Nodules are root- or stem-associated organs that develop as a result of complex chemical exchanges with soil bacteria known as rhizobia (Downie, 2014). After detection of specific nodulation factors synthesized by the colonizing rhizobia (Oldroyd, 2013), cells of the root pericycle and cortex proliferate to originate the nodule primordia (Xiao et al., 2014). As the nodule develops, a root hair curls to surround the rhizobia in the root rhizoplane. The plasma membrane of this hair cell retracts into the cytosol, forming an infection thread that guides the rhizobia from the epidermis into the nodule core (Gage, 2002). There, in an endocytic-like process, rhizobia are released into the cytosol of cortical cells (Limpens et al., 2009). Under the proper physico-chemical conditions, rhizobia differentiate into bacteroids (Kereszt et al., 2011). Surrounded by a specialized plant-derived membrane, the symbiosome membrane, and protected from oxygen, bacteroids are able to synthesize nitrogenase, the iron-molybdenum enzyme complex responsible for converting atmospheric N<sup>2</sup> into NH<sup>4</sup> <sup>+</sup> (Rubio and Ludden, 2005). Fixed nitrogen is transferred to the host plant, in exchange for photosynthate and mineral nutrients from the plant (Udvardi and Poole, 2013). Two morphological types of nodules are known as follows: determinate and indeterminate (Brewin, 1991). Indeterminate nodules, such as those present in the genera Medicago or Pisum, are characterized by the presence of a persistent apical meristem(s) that produce cylindrical or coralloid-shaped organs (Vasse et al., 1990). As a consequence, different developmental zones can be distinguished in such nodules: the meristem or Zone I, the region where rhizobia colonize the nodule and differentiate into bacteroids or Zone II, the nitrogen fixation zone or Zone III, and, in old nodules, the senescent zone or Zone IV (Vasse et al., 1990). To this, some authors add an Interzone between Zones II and III, were oxygen levels transition from atmospheric levels (20%) to microaerobiosis (<1%); and a Zone V, where the rhizobia grow saprophytically (Timmers et al., 2000; Roux et al., 2014). In addition to the zonation pattern, determinate and indeterminate nodules differ in the process of bacteroid differentiation, that is irreversible for the majority of indeterminate nodules, to the extent that they cannot proliferate if released from the nodules. These bacteroids also have a higher ploidy level and larger size, as a consequence of cysteine-rich peptides released by the host cell that participate in bacteroid development (Van de Velde et al., 2010; Kereszt et al., 2011; Kondorosi et al., 2013; Stonoha-Arther and Wang, 2018).

The specific role(s) of Zn in symbiotic nitrogen fixation is not clear. Given the multitude of Zn-proteins in a cell, Zn could act at several different levels. It has been reported that plants growing under Zn deficiency suffer reduced growth and reduced nitrogenase activity (Ibrikci and Moraghan, 1993; O'Hara, 2001). Similar effects result from silencing the expression of MtZIP6, a Medicago truncatula Zn transporter located in the plasma membrane of nitrogen-fixing nodule cells (Abreu et al., 2017). Nodule development in these silenced plants was also affected, resulting in smaller nodules compared to wild-type controls. Understanding how reduced uptake of Zn into cells affects nodule development and symbiotic nitrogen fixation is complicated by the fact that Zn plays a role in numerous intracellular processes. Knowledge of the intracellular fate of Zn would help in this regard. To this end, we are characterizing Zn transporters likely to be involved in organelle loading, especially be members of the MTP and Zn2+-ATPase families.

Exploration of publicly available transcriptome databases (Benedito et al., 2008; Roux et al., 2014) revealed no Zn2+- ATPase to be upregulated during nodule development. In contrast, M. truncatula MtMTP2 (Medtr4g064893) was found to be expressed at higher levels in nodules than in any other organ. Here, we show that MtMTP2 is a Zn efflux protein located in an intracellular compartment, with an important role in nodule development and bacteroid differentiation.

### MATERIALS AND METHODS

### Biological Materials and Growth Conditions

Medicago truncatula R108 ecotype, the Tnt1-insertion mutants mtp2-1 (NF11171) and mtp2-2 (NF18305) were used in this study. Seeds were scarified in concentrated H2SO<sup>4</sup> for 7 min and washed in dH2O. Later, seeds were surface sterilized using 50% bleach for 90 s, washed in sterile dH2O, and left overnight in sterile water to facilitate imbibition. After 48 h at 4◦C, seeds were germinated in water-agar plates at 22◦C for 48 h. Seedlings were then transplanted to sterilized perlite pots and inoculated with Sinorhizobium meliloti 2011, S. meliloti 2011 transformed with the GFP expressing pHC60 vector (Cheng and Walker, 1998), or S. meliloti 1021 expressing pCMB13 DsRED (Gage, 2002), as indicated. Plants were cultivated in a greenhouse in 16 h of light and 22◦C conditions, and watered every 2 days with Jenner's solution or water, alternatively (Brito et al., 1994). This nutrient solution contained 5 mM CaSO4, 1 mM KCl, 1 mM K2HPO4, 1 mM MgSO4, 11.5 µM H3BO3, 7.3 µM Fe-citrate, 3.6 µM MnSO4, 0.38 µM ZnSO4, 0.16 mM

CuSO4, and 4 nM (NH4)6Mo2O24. Nodules were collected 28 dpi. Non-nodulated plants were watered every 2 weeks with solutions supplemented with 20 mM NH4NO3. For hairyroot transformations, M. truncatula seedlings were transformed with Agrobacterium rhizogenes ARqua1 carrying the appropriate binary vector as described (Boisson-Dernier et al., 2001).

Complementation assays were performed using the yeast (Saccharomyces cerevisiae) strains DY1457 (MATa ade6 can1 his3 leu2 trp1 ura3), the yeast double mutant zrc1cot1 (MATa ade6 can1 his3 leu2 trp1 ura3 zrc1::his3 cot1::ura3) (MacDiarmid et al., 2000), DY150 (MATa ade2-1 his3-11 leu2-3,112 trp1-1 ura3-52 can1-100(oc)), the mutant ccc1 (MATa ade2-1 his3-11 leu2-3,112 trp1-1 ura3-52 can1-100(oc) ccc1::his3; Li et al., 2001), BY4741 (his3 leu2 met1 ura3), and the strain 1smf1 (his3 leu2 met1 ura3 1smf1; ThermoFisher). All strains were grown in synthetic dextrose (SD) medium (Sherman et al., 1981) supplemented with necessary auxotrophic requirements, with 2% (w/v) glucose as the carbon source, and supplemented with Zn, iron, or manganese, when required.

#### Sequence Analysis and Protein Structure Prediction

To identify M. truncatula MTP family members, BLASTN and BLASTX searches were carried out in the M. truncatula Genome Project site<sup>1</sup> and include 13 members: MtMTP1, Medtr8g024240; MtMTP2, Medtr4g064893; MtMTP3, Medtr2g036390; MtMTP4, Medtr3g080090; MtMTP5, Medtr7g093290; MtMTP6, Medtr1g088870; MtMTP7, Medtr4g008150; MtMTP8, Medtr3g062610; MtMTP9, Medtr2g064405; MtMTP10, Medtr8g046550; MtMTP11, Medtr7g022890; MtMTP12, Medtr6g463330, MtMTP13, Medtr5g075680. Sequences from model MTP genes were obtained from the Transporter Classification Database<sup>2</sup> (Saier et al., 2014), NCBI<sup>3</sup> and Phytozome<sup>4</sup> , and include A. thaliana MTPs (AtMTP1, At2g46800; AtMTP2, At3g61940, AtMTP3, At3g58810; AtMTP4, At2g29410; AtMTP5, At3g12100; AtMTP6, At2g4783; AtMTP7, At1g51610; AtMTP8, At3g58060; AtMTP9, At1g79520; AtMTP10, At1g16310; AtMTP11, At2g39450; AtMTP12, At2g04620), Oryza sativa MTPs (OsMTP1, LOC\_Os05g03780; OsMTP8, LOC\_Os05g03780) Cucumis sativus MTPs (CsMTP1, Cucsa.362220; CsMTP4, Cucsa.146570; CsMTP8, Medtr3g062610; CsMTP9, Cucsa.118550), Anemone halleri MTPs (AhMTP1-A, FN428855; AhMTP1-B, Fn386317; AhMTP1-C, Fn386316; AhMTP1-D, Fn386315), Hordeum vulgare MTPs (HvMTP1, HORVU1Hr1G015500; HvMTP8.1, HORVU4Hr1G065110.1), and Populus trichocarpa MTPs (PtdMTP1, Potri.014G106200; PtMTP11, POPTR\_0010s21810). All these protein sequences were processed with MEGA7<sup>5</sup> . First, protein sequences were aligned using the Clustal Omega algorithm<sup>6</sup> (Sievers et al., 2011), and the alignment was visually examined to exclude alignment artifacts. Then, phylogenetic reconstruction was performed using the Neighbor-joining method, the Jones–Taylor–Thornton (JTT) substitution model, and assuming uniform rates (Saitou and Nei, 1987; Jones et al., 1992). Deletion sites were excluded from the alignment following the partial deletion method (95% site coverage cutoff). Unrooted tree visualization was carried out using FigTree<sup>7</sup> .

MtMTP2 protein sequence from R108 was obtained from the Medicago Hapmap website<sup>8</sup> . The automated protein homologymodeling server SWISS-model<sup>9</sup> (Biasini et al., 2014) was used to predict the MtMTP2 protein structure based on the template 3h90 from the Escherichia coli Zn transporter YiiP (Lu et al., 2009). Protein structure was visualized using PyMOL (Schörindeger LLC, United States).

#### RNA Extraction and RT-qPCR

RNA was isolated from leaves, roots, or nodules from threepooled plants (from independent experiments each) following the protocol previously described by Abreu et al. (2017). Briefly, RNA was extracted using Tri-Reagent <sup>R</sup> (Life Technologies, Carlsbad, CA) followed by a DNase treatment and later cleaned with RNeasy Minikit (Qiagen, Valencia, CA). Denaturing agarose gel was used to verify RNA quality. One microgram of DNA-free RNA was employed to generate cDNA by using PrimeScript RT Reagent Kit (Takara). Gene expression was determined by quantitative Real time RT-PCR (9700, Applied Biosystems, Carlsbad, CA, United States) using primers listed in Supplementary Table S1. The M. truncatula ubiquitin carboxylterminal hydrolase gene was used to normalize the results. Realtime cycler conditions have been previously described (González-Guerrero et al., 2010). The threshold cycle (Ct) was determined in triplicate. The relative levels of transcription were calculated using the 2−11Ct method (Livak and Schmittgen, 2001). As control, a non-RT sample was used to detect any possible DNA contamination.

#### Yeast Complementation Assays

Yeast complementation was performed by cloning the MtMTP2 cDNA between the XbaI and BamHI sites of the yeast expression vector pAMBV or pDR196. Cloning in pAMBV was carried out by homologous recombination of MtMT2 cDNA using primers 5 MtMTP2 XbaI pMBV and 3 MtMTP2 BamHI pAMBV (Supplementary Table S1). Cloning in pDR196 was carried out by restriction digestion and T4 ligation of the DNA fragment resulting from the XbaI and BamHI digestion of the amplicon resulting from amplifying by PCR MtMTP2 cDNA with primers 5MtMTP2-XbaI and 3MtMTP2-BamHI (Supplementary Table S1). Yeast transformations were performed using a lithium acetate-based method (Schiestl and Gietz, 1989). Cells transformed with pAMBV or pAMBV::MtMTP2 (in Zn phenotypic assays) or pDR196 or pDR196 (in the case of iron or manganese phenotypic assays) were selected in SD medium by leucine or uracil autotrophy, respectively.

<sup>1</sup>http://www.jcvi.org/medicago/index.php

<sup>2</sup>http://www.tcdb.org/

<sup>3</sup>http://www.ncbi.nlm.nih.gov/

<sup>4</sup>https://phytozome.jgi.doe.gov/pz/portal.html

<sup>5</sup>http://www.megasoftware.net

<sup>6</sup>https://www.ebi.ac.uk/Tools/msa/clustalo/

<sup>7</sup>http://tree.bio.ed.ac.uk/software/figtree/

<sup>8</sup>www.medicagohapmap.org

<sup>9</sup>https://swissmodel.expasy.org

For phenotypic tests, DY1457 and zrc1cot1 transformants were plated in SD with or without supplementation of 500 µM ZnSO4, DY150, and ccc1 transformants were plated in SD with or without supplementation with 4 mM FeSO<sup>4</sup> and BY4741 and 1smf1 transformants were plated in SD with or without supplementation with 10 mM MnCl2.

#### GUS Staining

A transcriptional fusion was constructed by amplifying 889 bases upstream of MtMTP2 start codon using primers indicated on Supplementary Table S1, cloned in pDONR207 (Invitrogen), and transferred to pGWB3 (Nakagawa et al., 2007) using Gateway technology <sup>R</sup> (Invitrogen). This led to the fusion of the promoter region of MtMTP2 with the β-glucoronidase (gus) gene in pGWB3. pGWB3::MtMTP2 was transformed in A. rhizogenes ARqua1 and used to obtain M. truncatula composite root plants as indicated (Boisson-Dernier et al., 2001). GUS activity was determined in 28 dpi plants as described (Vernoud et al., 1999). The process was carried out from biological material originated from three independent assays carried out at different times of the year to select a representative image.

#### Immunolocalization of MtMTP2-HA

By using Gateway Technology <sup>R</sup> (Invitrogen), a DNA fragment of the full length MtMTP2 genomic region and the 1,961 bases upstream of its start codon, was cloned into the plasmid pGWB13 (Nakagawa et al., 2007). Hairy-root transformation was performed as previously described (Boisson-Dernier et al., 2001). For confocal microscopy, transformed plants were inoculated with S. meliloti 2011 containing the pHC60 plasmid that constitutively expresses GFP (Cheng and Walker, 1998) or DsRED (Gage, 2002). Roots and nodules collected from 28 dpi plants were fixed by overnight incubation in 4% paraformaldehyde, 2.5% sucrose in phosphate buffer saline (PBS) at 4◦C. After washing in PBS, nodules were cut in 100 µm sections with a Vibratome 1000 plus (Vibratome, St. Louis, MO, United States). Sections were dehydrated in a methanol series (30, 50, 70, 100% in PBS) for 5 min and then rehydrated. Cell walls were treated with 4% cellulase in PBS for 1 h at room temperature and with 0.1% Tween 20 in PBS for an additional 15 min. Sections were blocked with 5% bovine serum albumin (BSA) in PBS before their incubation with an anti-HA mouse monoclonal antibody (Sigma, St. Louis, MO) for 2 h at room temperature. After washing, an Alexa 594-conjugated anti-mouse rabbit monoclonal antibody (Sigma) was added to the sections for 1 h at room temperature. DNA was stained with DAPI after washing. Images were acquired with a confocal laser-scanning microscope (Leica SP8, Wetzlar, Germany). The process was carried out from biological material originated from three independent assays carried out at different times of the year to select a representative image.

Immunolocalization of MtMTP2-HA in an electronmicroscope was carried out with M. truncatula plants transformed with A. rhizogenes ARqua1 pGWB13 carrying MtMTP2 full gene and the 1,961 bases upstream the start codon. Transformed plants were inoculated with S. meliloti 2011 and 28 dpi nodules were collected and fixed in 1% formaldehyde and 0.5% glutaraldehyde in 50 mM potassium phosphate (pH 7.4) for 2 h. After that, the fixation solution was renewed for 1.5 h. Samples were washed in 50 mM potassium phosphate (pH 7.4) 3 × 30 min and 3 × 10 min. Nodules were dehydrated by incubation with ethanol dilution series of 30, 50, 70, 90 (10 min each), 96 (30 min), and 100% (1 h). Nodules were included in a series of ethanol and LR–white resin (London Resin Company Ltd., United Kingdom) dilutions: 1:3 (3 h), 1:1 (overnight), and 3:1 (3 h). Samples were included in resin during 48 h. All the process was performed at 4◦C. Nodules were placed in gelatine capsules and filled with resin and polymerized at 60◦C for 24 h. One-micron thin sections were prepared at Centro Nacional de Microscopía Electrónica (Madrid, Spain) with a Reichert Ultracut S-ultramicrotome fitted with a diamond knife. Thin sections were blocked in 2% bovine serum albumin in PBS for 30 min. Anti-HA rabbit monoclonal antibody (Sigma) was used as primary antibody, a 1:20 dilution in PBS. Samples were washed 10 times in PBS for 2 min. Anti-rabbit goat conjugated to a 15-nm gold particle (BBI solutions) was used as secondary antibody diluted 1:150 in PBS. Incubation was performed for 1 h followed by 10 washes in PBS for 2 min and 15 times in water for 2 min. Sections were stained with 2% uranyl acetate and imaged in a JEM 1400 electron microscope at 80 kV.

## Acetylene Reduction Assay

Nitrogenase activity was measured by the acetylene reduction assay (Hardy et al., 1968). Nitrogen fixation was assayed in 28 dpi wild-type and mutant plants in 30 ml tubes fitted with rubber stoppers. Each tube contained roots from five independently transformed plants. Three milliliters of air inside were replaced with 3 ml of acetylene. Tubes were incubated at room temperature for 30 min. Gas samples (0.5 ml) were analyzed in a Shimadzu GC-8A gas chromatograph fitted with a Porapak N column. The amount of ethylene produced was determined by measuring the height of the ethylene peak relative to background. Each point consists of three tubes each with five pooled plants measured in triplicate.

### Metal Content Determination

Inductively coupled plasma mass spectrometry (ICP-MS) was carried out at the Metal Analysis Unit of the Scientific and Technology Centre, Universidad de Barcelona (Barcelona, Spain). These samples were digested with HNO3, H2O2, HF in a Teflon reactor at 90◦C. The sample was diluted with deionized water. The final volume of the solution was calculated by weight difference with the original sample, and with the measured density of the solution, obtained from weighting a small aliquot of known volume. Samples were digested with three blanks in parallel. Metal determination was carried out in Agilent 7500ce instrument under standard conditions (RF power 1550 W, Nebulizer Burgener AriMist HP, Nebulizer Ar flow 0.75 l/min, sample pump 0.1 rps, QP resolution 0.7 amu at 10% height (7Li, <sup>89</sup>Y, <sup>205</sup>Tl), integration time 0.9 s, reading replicates 3, calibration linear through zero, internal standard online addition <sup>103</sup>Rh, gas cell mode He collision). Calibration was carried out with five measurements using commercial certified solutions analyzed and compared with reference NIST solutions.

## Confocal Imaging of Bacteroids and Colony-Forming-Units Assays

Confocal microscopy images of bacteroids were obtained from 28 dpi nodules. Nodules were ground with a micropestle in TY medium to release the bacteroids. The nodule homogenate was filtered using CellTrics <sup>R</sup> 30 µm columns and stained using propidium iodide (50 µg/ml) to visualize the bacteroids by using confocal microscopy at a 535 nm Ex/617 nm Em. Colonyforming-units (CFU) were obtained from fresh nodules harvested at 28 dpi. Nodules were weighed and surface-sterilized in 70% ethanol for 10 min followed by five washes with distillated water. The tissue was later ground with a micropestle in 200 µl TY medium. Serial dilutions of the homogenate were plated on TY solid media. Plates were incubated 48 h at 30◦C, and the number of colonies was recorded (Montiel et al., 2016).

### Statistical Tests

Results are presented as mean value ± standard deviation. Multiple comparisons were performed by one-way analysis of variance (ANOVA) followed by Tukey HSD post hoc at a probability level of 5% (P < 0.05). Pairwise comparisons were done by using Student's t-test at a probability level of 5% (P < 0.05). The JMP <sup>R</sup> (ver.11.0) statistical package (SAS Institute) was used for statistical analyses.

### RESULTS

#### MtMTP2 Is Up-Regulated During Nodule Development

Out of the thirteen MTP genes in the M. truncatula genome, MtMTP2 was the one with the highest expression levels in nodules, as reported in the Medicago Gene Expression Atlas (Benedito et al., 2008) and in the Symbimics database (Roux et al., 2014; Supplementary Figure S1). MtMTP2 expression analysis was performed to identify organs in which the gene/protein operates. Relatively high levels of MtMTP2 transcripts were found in nodules compared to shoots and roots of inoculated plants (**Figure 1**). Shoots of nitrogen-fertilized, non-nodulated plants exhibited higher transcript levels than those of nodulated plants, although levels were still much lower than in nodules (**Figure 1**).

MTP proteins fall into seven groups, based on sequence similarity, which roughly correspond to putative metal substrates and subcellular localizations (Ricachenevsky et al., 2013). To gain insight into the possible metal substrates of MtMTP2, phylogenetic analysis was performed on 13 M. truncatula MTPs and homologous proteins in Arabidopsis thaliana, Oryza sativa, Crocus sativus, Arabidopsis halleri, Populus trichocarpa, and Hordeum vulgare. MtMTP2 showed strong similarity to CsMTP4 (54% similarity) described as a protein involved in Zn homeostasis in cucumber (Migocka et al., 2015) and AtMTP4 (48% similarity), a predicted Zn transporter of A. thaliana (Waters and Grusak, 2008; **Figure 1B**). To explore further Zn as a candidate substrate, a predicted tertiary structure of MtMTP2 was obtained by homology modeling based on the known crystal structure of the E. coli Zn transporter YiiP (Lu et al., 2009). The generated model revealed a Zn-binding domain made of residues H88, D92, H271, and D275 that corresponded to the YiiP site Z1 (D45, D49, H153, and D157; **Figure 1C**). The substitution of YiiP D45 residue by a histidine, H88, observed in MtMTP2 is also conserved in other plant MTP proteins (**Figure 1D**).

## MtMTP2 Complements a Zinc Detoxification-Deficient Yeast Mutant

In S. cerevisiae, ZRC1 and COT1 are tonoplast transporters responsible for the storage of Zn in vacuoles (MacDiarmid et al., 2000). Yeast zrc1/cot1 double mutants are hyper-sensitive to Zn in the growth medium. Genetic complementation assays using a yeast zrc1/cot1 double mutant showed that expression of MtMTP2 enabled the mutant strain to grow on otherwise toxic levels of Zn (500 µM ZnSO4), consistent with a role of MtMTP2 in Zn efflux out of the cytosol (**Figure 2A**). In contrast, complementation assays using a yeast ccc1 mutant affected in the transport of iron into the vacuole (Li et al., 2001), and an smf1 mutant strain, which is unable to store manganese in the vacuole (Portnoy et al., 2000), showed no recovery of growth on media supplemented with 4 mM FeSO<sup>4</sup> or 10 mM MnCl2, respectively (**Figures 2B,C**), indicating that MtMTP2 does not transport Fe or Mn in yeast.

### MtMTP2 Is Located in an Endomembrane Compartment in Cells of Nodule Zones II to III

To determine MtMTP2 expression distribution in nodules and roots, a segment of 889 bp upstream of the MtMTP2 start codon was fused to the gus reporter gene and subsequently expressed in roots of M. truncatula inoculated with rhizobia. MtMTP2 promoter activity was detected in roots and nodules (**Figure 3A**) and was most active in the segment from late Zone II to early Zone III, while lower GUS signal was detected at the meristematic zone and late zone III (**Figure 3B**). In silico analysis of MtMTP2 expression in nodules using data obtained from the Symbimics database<sup>10</sup> (Roux et al., 2014) was consistent with GUS assays showing an increased expression pattern in the late differentiation zone (proximal Zone II) and in Zone III (Supplementary Figure S2). In roots, MtMTP2-regulated GUS activity was faintly detected at the epidermis, pericycle, and vascular tissue (**Figure 3C**).

Subcellular localization of MtMTP2 was performed by fusing the DNA segment from 1,961 bp upstream of the MtMTP2 start codon to the last codon before the stop codon to either GFP or to 3 hemagglutinin (HA) epitopes, resulting in pMtMTP2::MtMTP2-GFP and pMtMTP2:: MtMTP2-HA, respectively. Plants were transformed with these constructs and inoculated with S. meliloti strains constitutively expressing DsRED or GFP, followed by DAPI staining (**Figure 4**). Subcellular localization using both constructs was consistent with the results obtained with the GUS reporter assays. pMtMTP2::MtMTP2- GFP was localized intracellularly from the infection and

<sup>10</sup>https://iant.toulouse.inra.fr/symbimics/

FIGURE 2 | MtMTP2 yeast complementation assays. (A) Yeast strain DY1457 was transformed with the pYPGE15 empty vector, while the double mutant zrt1/cot1 was transformed with the empty pYPGE15 or with pYPEG15 containing MtMTP2 coding DNA sequence. Serial dilutions (10×) of each transformant were grown for 3 days at 28◦C on SD media with all the required amino acids and 500 µM ZnSO4. Positive control was obtained without supplementing the media with ZnSO4. (B) Yeast strain DY150 was transformed with the pDR196 empty vector, while ccc1 mutant was transformed either with the empty pDR196 or with pDR196 containing MtMTP2 coding DNA sequence. Serial dilutions (10×) of each transformant were grown for 3 days at 28◦C on SD media with all the required amino acids and 4 mM FeSO4. Positive control was obtained without supplementing the media with FeSO4. (C) Yeast strain BY4741 was transformed with the pDR196 empty vector, while a deletion strain in smf1 was transformed either with empty pDR196 or with pDR196 containing MtMTP2 coding DNA sequence. Serial dilutions (10×) of each transformant were grown for 3 days at 28◦C on SD media with all the required amino acids and 10 µM MnCl2. Positive control was obtained without MnCl<sup>2</sup> supplementation.

differentiation zones and into the fixation zone in 28 dpi nodules, while no signal was found at the meristematic zone (**Figure 4A**). Localization using the pMtMTP2::MtMTP2- HA construct by immunostaining with Alexa594-conjugated antibody (Supplementary Figure S3), showed an identical pattern of distribution to that observed with pMtMTP2::MtMTP2- GFP. High magnification imaging of nodules 28 dpi allowed the visualization of MtMTP2 intracellularly in infected cells (**Figures 4B,C**). No autofluorescence signal was detected under the experimental conditions used, as shown when primary anti-HA antibody for MtMTP2-HA detection was removed (Supplementary Figure S4), or when no MtMTP2-GFP protein was present (Supplementary Figure S5). In roots, MtMTP2- HA was detected in the epidermis and around the vasculature (**Figure 4D**), consistent with the promoter-GUS assays.

To further clarify the subcellular localization of MtMTP2, immunolocalization of MtMTP2-HA with a gold-conjugated antibody and transmission electron microscopy was used

(**Figure 4E**). Gold particles were found concentrated in electron dense structures corresponding to intracellular compartments resembling endoplasmic reticulum (or associated domains) of infected cells and non-infected cells. No gold particles were detected associated with symbiosomes. When no primary antibody was used, no gold particles were found to be concentrated in any cell sections (Supplementary Figure S6).

### MtMTP2 Mutants Exhibit Abnormal Accumulation of Zinc in Nodules and Impaired Nitrogen Fixation and Growth

To determine the physiological role of MtMTP2, two homozygous mutant lines of MtMTP2, mtpt2-1 (NF11171), and mtp2-2 (NF18305) were evaluated. Mutant line mtpt2-1 harbors the tnt1 insertion at the promoter region (−69 upstream of the start codon) while mtp2-2 contains a tnt1 insertion at the unique exon of the gene (+623 downstream the start codon; **Figure 5A**). These insertions resulted in a reduced level of MtMTP2 expression in nodules, 80% reduction in the case of mtpt2-1 and more than 99% in mtpt2-2 (**Figure 5B**). Accordingly, they were designated as a knock-down and a knock-out MtMTP2 mutants, respectively. Plant phenotypes were analyzed when nitrogen was provided in the nutrient solution as ammonium nitrate. Under these non-symbiotic conditions, no changes in the plant phenotype were detected in the mtp2 mutants compared to wild-type plants (**Figure 5C**). Plant biomass (determined by the dry weight of shoots and roots) did not show significant differences among the genotypes analyzed (**Figure 5D**). Similarly,

FIGURE 5 | MtMTP2 mutants grown under non-symbiotic conditions did not show an altered phenotype. (A) Position of the transposable element of tobacco (Nicotiana tabacum) cell type 1 (Tnt1) in the MtMTP2 genomic region in mtp2-1 and mtp2-2 mutants. (B) MtMTP2 expression in roots (black) and nodules (white) of WT, mtp2-1, and mtp2-2 plants relative to the internal gene Ubiquitin carboxyl-terminal hydrolase 1. Data are the mean ± SD of two independent experiments with four pooled plants. (C) From the left to the right, representative plants of M. truncatula WT, mtp2-1, and mtp2-2 N-fertilized plants. Bars: 1 cm. (D) Dry weight of shoots (black) and roots (white). Data represent the mean ± SD of three experiments pooling, at least, ten independent plants (n = 10).

no significant differences between wild type and mutant lines were observed when either no Zn or excess Zn (100×) were added to the nutrient solution (Supplementary Figure S7).

In contrast, under symbiotic conditions without mineral-N, mutant plants exhibited reduced growth and an altered nodule development (**Figure 6**). Nodules of the mtp2-2 mutant were small, round, and white, in contrast to the long, cylindrical, pink nodules of the wild-type (**Figures 6A,B**). A time-course analysis of nodule growth was performed, which showed a progressive delay in nodule growth, but no delay in the start of nodulation of the mutants (Supplementary Figure S8). Consistent with the plant phenotypes observed, plant biomass (determined as dry weight) was reduced in both mutants (**Figure 6C**). Shoot biomass suffered a more drastic decrease linked to the mutation than did root biomass (a 36% reduction in the case of mtpt2-1 and 55% for mtpt2-2 shoots). However, root biomass was still significantly diminished in both mutants, and to a similar extent (20%). In order to determine if the phenotypic alterations observed were a consequence of a decline in nitrogenase activity, acetylene reduction assays were performed (Hardy et al., 1968). The knock-down mutant, mtpt2-1 exhibited a reduction of 60% in nitrogenase activity while the knock-out mutant, mtpt2-2 barely showed detectable enzyme activity (**Figure 6D**). Similar activity profile was observed when data were normalized to nodule number per root (Supplementary Figure S9). Zn content in shoots, roots, and nodules were determined in order to evaluate changes in the putative metal substrate of MtMTP2. While no significant changes in Zn content were detected in roots, and a

slight increase was detected in the shoots of mtpt2-2, nodules of knock-out plants showed a substantial accumulation of Zn (∼40% increase; **Figure 6E**). This phenotype was not the result of additional insertions in the tnt1 lines. Both mutant lines only share insertions in MtMTP2. Moreover, segregants containing the two wild-type copies of MtMTP2 did not show any significant differences with wild-type plants (Supplementary Figure S10). A Zn gradient including suboptimal (0 µM added ZnSO4), control (0.38 µM ZnSO4) and supra-optimal Zn conditions (38 µM ZnSO4) was applied in an attempt to complement the mtp2 mutants' defective symbiotic phenotype (Supplementary Figure S11). None of the Zn conditions tested enabled recovery in growth (including biomass; Supplementary Figure S11A,B) nor in nitrogenase activity (Supplementary Figure S11C).

different (Tukey's HSD, p < 0.05).

#### Absence of MTP2 Alters Nodule and Bacteroid Development

The smaller size of mtp2 mutant nodules indicated altered nodule development. To explore this further, nodules were

toluidine blue and visualized by light microscopy. Bars: 100 µm. (B) Detail of infected nodule cells of M. truncatula WT, mtp2-1, and mtp2-2 nodules stained with toluidine blue and visualized by light microscopy. Bars: 10 µm. (C) Confocal image of S. meliloti stained with propidium iodide from 28-dpi M. truncatula WT, mtp2-1, and mtp2-2 nodules. (D) Cell length of S. meliloti bacteroids isolated from WT, mtp2-1, and mtp2-2 nodules. Data are the mean ± SD of at least 100 cells. (E) S. meliloti CFU per mg fresh weight of 28-dpi nodules from WT, mtp2-1, and mtp2-2 plants. Data are the mean of three independent experiments ± SE. Values with different letters are significantly different (Tukey's HSD, p < 0.05). (F) Expression of the senescence marker genes Chit2 (chitinase Medtr5g022560) and CysProt (cysteine protease Medtr6g079630) in 28 dpi M. truncatula nodules of WT, mtp2-1, and mtp2-2 plants relative to the internal standard gene Ubiquitin carboxyl-terminal hydrolase. Data are the mean of three independent experiments ± SE. Values with different letters are significantly different (Tukey's HSD, p < 0.05).

sectioned, stained with toluidine blue, and observed by light microscopy. Nodules of mtpt2-2 a reduced infection zone and the disappearance of the fixation zone compared to WT nodules (**Figures 7A,B**).

The lack of a fixation zone in mtp2-2 suggested that rhizobia may not have differentiated fully into bacteroids in these nodules. To test this idea, bacteroids were isolated from WT and mtp2 mutant nodules and characterized (**Figures 7C,D**). Bacteroids from mtp2-2 nodules were shorter in length than those of wild-type nodules, longer than free-living rhizobia (**Figure 7D**) and formed more colonies on solid growth medium (**Figure 7E**). These results are consistent with lack of full differentiation of rhizobia into nitrogen-fixing bacteroids within mtp2-2 mutant nodules, or with early senescence that would hamper further nodule development. To test the later, gene expression of a chitinase (Medtr5g022560) and cystine protease (Medtr6g079630), two genes induced in senescence (Xi et al., 2013), was determined in 28 dpi nodules from wild-type and mtp2-1 and mtp2-2 plants (**Figure 7F**). The result showed an induction of senescent genes in mtpt2-2 nodules.

#### DISCUSSION

MTP proteins are members of the Cation Diffusion Facilitator (CDF) family and known in plants (Ricachenevsky et al., 2013).These proteins are typically involved in Zn2+, Mn2+, or Fe2<sup>+</sup> efflux from the cytosol, either out of the cell or into organelles (Desbrosses-Fonrouge et al., 2005; Eroglu et al., 2015). From a physiological point of view, their functions can be diverse, including metal detoxification (Desbrosses-Fonrouge et al., 2005), metal storage and allocation to sink organs (Eroglu et al., 2017), and metalation of apometalloproteins (Ellis et al., 2004). From a structural point of view, they seem to function as homodimers (Wei et al., 2004; Lu and Fu, 2007; Lu et al., 2009), in which each of the monomers is able to independently pump cations. This is made possible by transmembrane metal binding sites that, by being arranged in a specific geometry, confer specificity to the transporter (Argüello et al., 2012).

The genome of M. truncatula encodes 13 MTP proteins of which only one, MtMTP1, was previously characterized (Chen et al., 2009). MtMTP1 is a Zn transporter also involved in Zn efflux from the cytosol. It is expressed in roots and shoots. Zn modulates MtMTP1 transcription: in roots, it is downregulated, while it is up-regulated in shoots in response to Zn supply. Although there is no published information on its expression in nodules, its overall expression decreases with S. meliloti inoculation, and the ˆMedicago Gene Expression Atlas and the Symbimics databases indicate that it is downregulated in nodules. In this manuscript, we characterized MtMTP2 as an additional Zn2+-transporting MTP family member that is involved in nodule development. Homology modeling of the structure of MtMTP2 shows that its metal substrate would likely be tetrahedrally coordinated by two histidine and two glutamate residues, since they occupy a similar location to the transmembrane metal binding site of template E. coli YiiP (Lu and Fu, 2007; Lu et al., 2009). Three of the four amino acid residues are conserved between both proteins, and the fourth, a change from glutamate to histidine, is consistent with Zn binding, both by its occurrence in many Zn-coordinating sites, and by its being conserved in other plant Zn-transporting MTPs (Ricachenevsky et al., 2013). Further supporting this ability to transport Zn are two related observations: the capability of MtMTP2 to functionally complement the Zn transport defect of the yeast zrc1/cot1 double mutant (MacDiarmid et al., 2000); as well as the changes in Zn concentration in nodules of the mtp2-2 mutant line. Although MTP proteins have been shown to be able to transport more than one substrate, MtMTP2 did not complement defects in Fe or Mn transport of specific yeast mutants (Podar et al., 2012; Migocka et al., 2015; Eroglu et al., 2017).

MtMTP2 is expressed in different plant organs. In roots, it is located in the epidermal and in vascular and endodermal cells; in nodules, in cells in Zones II and III, as indicated by promoter::gus fusions, fluorescence of a GFP-labeled MtMTP2, and immunolocalization of a HA-tagged protein. The latter two approaches also provide insight into the subcellular localization of MtMTP2, which is associated with an endomembrane compartment. Given the importance of Zn in symbiotic nitrogen fixation (Ibrikci and Moraghan, 1993; O'Hara, 2001), and the direction of transport of MTPs in general and MtMTP2 in particular (Wei and Fu, 2006), it was tempting to speculate that MtMTP2 might deliver Zn across the symbiosome membrane. However, electron microscopy analyses of the localization of HA-tagged MtMTP2 indicated that this is not the case, since no protein was associated with symbiosomes. Instead, signal was located in electron-dense bodies in the cell cytosol, that are distributed all over the cells. This allows us to discard plastids or mitochondria as putative locations based on their unique morphology, as well as the nucleus (based on its uniqueness), Golgi cisternae (very few in a cell), or late endosomal compartment (with a close distribution to the plasma membrane). Putative localization would be the endoplasmic reticulum. Considering that transport into the vacuole would mean a role in Zn storage, rather than a more active role in the cell functioning, and the severe phenotype observed by MtMTP2 mutation, it can be speculated that the intracellular compartment would correspond to the endoplasmic reticulum. Previously, other MTP/CDF proteins have been associated to the endoplasmic reticulum, where they would play a role in metallating metalloproteins. Mutation of yeast Msc2 gene results in the induction of the unfolded protein response, as a consequence of the Zn cofactor not being inserted in the proteins (Ellis et al., 2004, 2005). Similar roles of CDF proteins being involved in metallation of proteins have been attributed to Schizosaccharomyces pombe Zhf1 (Choi et al., 2018), or to mammalian ZnT5 and ZnT6 (Suzuki et al., 2005). However, co-localization with ER-specific markers would be needed to conclusively demonstrate MtMTP2 subcellular localization.

In spite of being expressed in many plant organs, MtMTP2 is primarily involved in nodule development. No aberrant

phenotype was observed for mtp2 mutants under non-symbiotic conditions, indicating that either MtMTP2 is not required for key physiological processes in plants in the vegetative stage of growth when watered with NH4NO3, or that another, yetto-be-determined protein can serve as a substitute for MTP2. However, when nitrogen is provided by endosymbiotic rhizobia in root nodules, mutating MtMTP2 has a dramatic effect. As was the case with silencing the Zn transporter MtZIP6 (Abreu et al., 2017), altering Zn homeostasis in the nodule resulted in reduced nodule development and a substantial decrease in nitrogenase activity. However, the effect of mutating MtMTP2 was more severe because its loss lead to alterations in nodule development, in bacteroid maturation, and nodule senescence. Phenotypic differences between MtZIP6 silenced plants and the MtMTP2 knock-out mutant seem to result from the inability to completely silence gene expression, since an activity of just 40% in the knock-down mtp2-1 line is enough to allow for bacteroid development, a situation similar to what was reported for MtZIP6 RNAi plants. This result is striking since it indicates the existence of one or several Zn proteins that receive Zn in an endomembrane compartment (likely the endoplasmic reticulum) that have an effect on bacteroid differentiation and/or nodule development, leading to early senescence. Alternatively, it could be argued that MtMTP2 might be protecting the nodule against Zn toxicity by sequestering the metal, as has been proposed for other MTP transporters (Blaudez et al., 2003; Desbrosses-Fonrouge et al., 2005). However, when no Zn was provided in the nutrient solution, no improvement of the mutant phenotype was observed, suggesting that no toxicity effect was at play. Moreover, altered nodule development has also been reported when mutating iron transporter SEN1 in Lotus japonicus (Hakoyama et al., 2012). Lowering Zn levels in the nutrient solution did not have any effect on plant growth or nitrogen fixation. This could be due to not being able to diminish enough the Zn levels (traces in perlite) or to already having achieved the bare minimum nitrogenase activity in mtp2-2 plants. Future work will be directed toward characterizing the nodulation Zn-proteome to identify the

#### REFERENCES


Zn-proteins that might be governing nodule development and bacteroid differentiation.

### AUTHOR CONTRIBUTIONS

JL-M and MS carried out most of the experimental work with assistance from ÁS (yeast complementation), PG-D (nodule development time course and effect of added metals on mtp2- 2 phenotype), and JM (bacteroid development). IK and MU obtained the mtp2-1 and mtp2-2 mutants. MR, JI, and MG-G were responsible for experimental design, data analyses, and wrote the manuscript with contributions from all authors.

### FUNDING

This work was supported by the Spanish Ministry of Economy and Competitiveness (grant AGL-2012-32974), a Marie Curie International Reintegration grant (IRG-2010-276771), and a European Research Council Starting grant (ERC-2013-StG-335284) to MG-G.

### ACKNOWLEDGMENTS

The authors would like to thank Dr. David Eide (University of Wisconsin-Madison) for providing the DY1457 and zrc1cot1 yeast strains, Dr. Jack Kaplan (University of Utah) for the DY150 and the ccc1 yeast strains, Dr. Stephan Pollmann for the gift of the pAMBV vector, and members of CBGP laboratory 281 for critical discussions of the manuscript.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018.00990/ full#supplementary-material


the hydrogenase structural subunits. J. Bacteriol. 176, 5297–5303. doi: 10.1128/ jb.176.17.5297-5303.1994



**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 © 2018 León-Mediavilla, Senovilla, Montiel, Gil-Díez, Saez, Kryvoruchko, Reguera, Udvardi, Imperial and González-Guerrero. 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) and the copyright owner(s) 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.

# Copper Nanoparticles Induced Genotoxicty, Oxidative Stress, and Changes in Superoxide Dismutase (SOD) Gene Expression in Cucumber (Cucumis sativus) Plants

Kareem A. Mosa1,2 \*, Mohamed El-Naggar<sup>3</sup> , Kalidoss Ramamoorthy<sup>4</sup> , Hussain Alawadhi<sup>5</sup> , Attiat Elnaggar<sup>4</sup> , Sylvie Wartanian<sup>1</sup> , Emy Ibrahim<sup>1</sup> and Hala Hani<sup>1</sup>

<sup>1</sup> Department of Applied Biology, College of Sciences, University of Sharjah, Sharjah, United Arab Emirates, <sup>2</sup> Department of Biotechnology, Faculty of Agriculture, Al-Azhar University, Cairo, Egypt, <sup>3</sup> Department of Chemistry, College of Sciences, University of Sharjah, Sharjah, United Arab Emirates, <sup>4</sup> Environmental and Chemical Biology Research Group, Research Institute of Science and Engineering, University of Sharjah, Sharjah, United Arab Emirates, <sup>5</sup> Center of Advanced Materials Research, Research Institute of Sciences and Engineering, University of Sharjah, Sharjah, United Arab Emirates

#### Edited by:

Hannetz Roschzttardtz, Pontificia Universidad Católica de Chile, Chile

#### Reviewed by:

Julio Paez Valencia, University of Wisconsin-Madison, United States Anjana Jajoo, Devi Ahilya University, India

> \*Correspondence: Kareem A. Mosa

kmosa@sharjah.ac.ae

#### Specialty section:

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

Received: 17 February 2018 Accepted: 04 June 2018 Published: 16 July 2018

#### Citation:

Mosa KA, El-Naggar M, Ramamoorthy K, Alawadhi H, Elnaggar A, Wartanian S, Ibrahim E and Hani H (2018) Copper Nanoparticles Induced Genotoxicty, Oxidative Stress, and Changes in Superoxide Dismutase (SOD) Gene Expression in Cucumber (Cucumis sativus) Plants. Front. Plant Sci. 9:872. doi: 10.3389/fpls.2018.00872 With the increased use of metal nanoparticles (NPs), their access to the food chain has become a main concern to scientists and holds controversial social implications. This research particularly sheds light on copper nanoparticles (CuNP), as they have been commonly used in several industries nowadays. In this study, we investigated the phytotoxicity of CuNP on cucumber (Cucumis sativus) plants grown hydroponically. Atomic Absorption Spectroscopy (AAS), X-Ray Fluorescence (XRF), and Scanning Electron Microscopy (SEM) analysis confirmed that C. sativus treated with CuNP accumulated CuNP in the plant tissues, with higher levels in roots, with amounts that were concentration dependent. Furthermore, genotoxicity was assessed using Random amplified polymorphic DNA (RAPD) technique, and our results showed that CuNP caused genomic alterations in C. sativus. Phenotypical, physiological, and biochemical changes were assessed by determining the CuNP treated plant's total biomass, chlorophyll, H2O<sup>2</sup> and MDA contents, and electrolyte leakage percentage. The results revealed notable adverse phenotypical changes along with decreased biomass and decreased levels of the photosynthetic pigments (Chlorophyll a and b) in a concentration-dependent manner. Moreover, CuNP induced damage to the root plasma membrane as determined by the increased electrolyte leakage. A significant increase in H2O<sup>2</sup> and MDA contents were detected in C. sativus CuNP treated plants. Additionally, copper-zinc superoxide dismutase (Cu-Zn SOD) gene expression was induced under CuNP treatment. Overall, our results demonstrated that CuNP of 10–30 nm size were toxic to C. sativus plants. This finding will encourage the safe production and disposal NPs. Thus, reducing nano-metallic bioaccumulation into our food chain through crop plants; that possesses a threat to the ecological system.

Keywords: phytotoxicity, copper nanoparticles, hydrophonic system, XRF analysis, genotoxicity, Cucumis sativus

## INTRODUCTION

fpls-09-00872 July 12, 2018 Time: 18:27 # 2

Nanotechnology, being a revolutionary science, has been a top trend field in the scientific community because of the enormous uses and utilization of nanoparticles (NPs) in many industrial sectors and research fields. NPs have been used in various domains including medicine, electronics, and cosmetics (Roco, 2003; Biswas and Wu, 2005; Nowack and Bucheli, 2007). Despite the obvious benefits that NPs offer, there are open questions as to how the NPs used for everyday life may affect the environment. The extensive utilization of these NPs allows them to be unintentionally released into the environment either during production, use or disposal (Gottschalk and Nowack, 2011; Peralta-Videa et al., 2011; Rico et al., 2011). Increased production of NPs leads to toxic effects on living organisms including plants (Bhatt and Tripathi, 2011). Eventually, phytotoxicity in crop plants leads to threats to human health through the food chain (Servin et al., 2013). Although several reports representing phytotoxicity of NPs in plants have been demonstrated (Dreher, 2004; Yang and Watts, 2005; Wiesner et al., 2006; Handy et al., 2008; Wang et al., 2012; Yang et al., 2015), comprehensive knowledge is still lacking.

According to the United States Environmental Protection Agency (USEPA), engineered NPs are divided into four categories; carbon-based materials, metal-based materials, dendrimers, and composites (USEPA, 2007). NPs that have both positive and negative impacts on higher plants have been reported. For example, SiO<sup>2</sup> and TiO<sup>2</sup> enhanced the absorbance of water and fertilizers in soybean (Glycine max) by increasing the level of nitrate reductase enzyme to stimulate antioxidant system (Lu et al., 2002). Pisum sativum plants (green peas) treated with ZnO NP showed induced root growth compared to control plants (Mukherjee et al., 2014). A particular concentration of TiO<sup>2</sup> NP has been shown to increase the growth of spinach by increasing photosynthesis and nitrogen metabolism (Hong et al., 2005a,b; Zheng et al., 2005; Yang et al., 2006). On the other hand, inhibition of root elongation of corn, cucumber, soybean, cabbage, and carrot was demonstrated when plants were treated with nanoscale alumina (nano-Al2O3) powders (Yang and Watts, 2005). To the best of our knowledge, this study by Yang and Watts, 2005 was the first study which highlighted the negative impact of metal NPs in plants and initiated the focus on studying the phytotoxic effect of metal NPs on plants. Later, several reports investigated the phytotoxic impact of several NPs in different plant species. For example, maize seedlings, treated with TiO<sup>2</sup> NP showed reduced leaf growth and transpiration (Asli and Neumann, 2009).

Copper nanoparticles (CuNP) received great interest in the textile industry due to their immediate antimicrobial effects when introduced into synthetic and natural fibers. Also, it is now being used in the microelectronics industry for its highly conductive properties where it is being implied to conductive pastes and heaters (Ravishankar Rai and Jamuna Bai, 2011). Copper oxide nanoparticles (CuONP) have been used as catalysts for a typical C-N cross-coupling reactions (Pan et al., 2011; Poreddy et al., 2015). Copper oxide is a semiconductor with special optical, electrical, and magnetic properties; it has been used in developing super capacitors and sensors (Dagher et al., 2014; Devi et al., 2014; Zhang et al., 2014). Due to its antimicrobial and biocidal activities, CuONP attracted more researchers than other metal NP for medical applications (Perreault et al., 2012; Nations et al., 2015). In the medical field, CuONP are used to detect viruses in the human body (Ahamed et al., 2014). Sankar et al. (2014) highlighted the anticancer activity of CuONP on human lung cancer cells and showed apoptosis due to reactive oxygen species (ROS) dependent disruption of mitochondrial membranes.

Despite their many useful applications, CuNP might also have a negative impact on the environment. Generally, the toxicity of the NPs depends on size, surface charge, and pH of the environment (Chang et al., 2012). CuNP on algae showed higher toxic effect than its bulk material and the existence of the toxicity remains after 72 h of the treatment (Aruoja et al., 2009). Mussels Mytilus galloprovincialis treated with CuONP exhibited higher production of ROS which leads to genotoxicity and cancer (Ruiz et al., 2015). Additionally, potential toxic effects of CuONP have been reported in different tissues in rainbow trout (Oncorhynchus mykiss). (Baek and An, 2011; Isani et al., 2013; Ostaszewska et al., 2016).

Phytotoxicity of CuNP has been studied on wheat and mung bean plants grown on agar media and was found to reduce seedling and shoot growth (Lee et al., 2008). Additionally, Kim et al., 2012 reported the phytotoxic effect of CuONP with a size of 50 nm on cucumber plants grown hydroponically which exhibited a significant increase in ROS enzymes; catalase (CAT), peroxidase (POD), and superoxide dismutase (SOD) (Kim et al., 2012). In another report, CuONP significantly reduced the growth of cucumber plants cultivated on soil microcosm system (Kim et al., 2013). Although, the previous reports suggested the phytotoxic effects of CuNP on plants, the underlying physiological, cellular, and molecular mechanisms need further investigation.

In this study, a comprehensive analysis on cucumber (Cucumis sativus) plants treated with CuNP with an average size of 10–30 nm in a hydroponic system has been performed. The hydroponic system was used to maintain the uniform concentration of NPs in each different treatment compared to soil system grown plants. The phytotoxicity and genotoxicity of CuNP and its accumulation in C. sativus plant tissues were investigated in details.

## MATERIALS AND METHODS

#### Characterization of CuNP

Copper nanoparticles with an average particle size of 20 nm were purchased from Hengqiu Graphene Technology (Suzhou) Co., Ltd., Shanghai, China with a purity of 99.9%. To find out the exact size and nature of the NPs, they were analyzed using SEM. The analysis was carried out using a TESCAN VEGA4 XM SEM (SE Detector, 30 kV, high vacuum). Conductive carbon tape was used to attach the sample to the measurement stub.

### Seed Germination and Seedling Development

Soil mixture was prepared by mixing autoclaved Peat Moss with Perlite in a 3:1 ratio and was then made moist by adding distilled water. The mixture was then added to sterile plastic pots. The overnight soaked seeds of C. sativus were placed randomly onto the surface of the pots and a pinch of soil was added on top to fully cover the seeds. The pots were placed in the growth chamber and were watered at regular intervals for 3 weeks.

### Development of Hydroponic System

Cucumis sativus plants of 3 weeks old were carefully removed from the soil, washed with distilled water, transferred to glass jars containing 20% Hoagland solution prepared from Hoagland's No. 2 Basal Salt Mixture (Sigma-Aldrich), and placed in the growth chamber maintained at 25◦C (12 h with light) and 15◦C (12 h dark) for 3 weeks. After 7 days of growth in the hydroponic system, well-developed lateral root system was observed. The jars were replaced every 3 days with fresh Hoagland solutions to maintain stable health.

### Nanoparticles Treatment and Plant Tissues Harvesting

Various concentrations (50, 100, and 200 mg/L) of CuNP powders were dispersed in distilled water and sonicated for 30 min, and the required Hoagland powder was mixed with it prior treatment. The 6 weeks old plants were treated by replacing the Hoagland solution with fresh Hoagland solution containing CuNP and placed back in the growth chamber for 4 days. There were 8 replicate plants per treatment. After the 4 days treatment, plants were washed to remove excess particles and dried on a tissue paper for 30 min. Fresh biomass was determined for each plant before and after treatment. Shoot and root of four replicates were separated, and kept in a 50 ml polystyrene falcon tube and then stored at −80◦C until further use. The remaining four replicates were kept in a paper bags and dried in the oven (65◦C) till complete dryness for metal analysis.

#### Metal Analysis

Oven dried tissues of shoots and roots were weighed, crushed to powder, and digested with Nitric acid (HNO3) at 95◦C water bath till brown fumes ceased. After cooling; distilled water and 30% H2O<sup>2</sup> were added and samples were heated till no effervescence was observed. HCl was added to the mixture and heated. The samples were allowed to cool and filtered, and then metal accumulation was measured using Atomic Absorption Spectroscopy.

### XRF and SEM Analysis of CuNP in Plant Tissues

The XRF analysis was performed using a Horiba XGT 7200 X-ray analytical microscope. The microscope is equipped with a 50-W Rhodium X-ray source (operating at up to 50 kV) and a silicon drift detector. Two beam sizes of diameter 50 µm and 1.2 mm are available for spot and mapping analysis. XRF microscopy was used for elemental analysis of C. sativus roots treated with 200 mg/L CuNP. Oven dried root samples treated with CuNP were ground using mortar and pestles. The powdered samples were coated with a thin layer of gold for further analysis with SEM. The SEM analysis was carried out similar to the analysis of CuNP powders described previously.

## DNA Extraction and RAPD Analysis

Root samples stored at −80◦C were powdered in liquid nitrogen, and DNA extraction was performed using NORGEN BIOTEK DNA extraction kit for plant/fungi according to the manufacturer instructions. The quality of the extracted DNA was checked by gel electrophoresis on 1% agarose gel and was quantified using NanoDrop. PCR reactions were prepared using NORGEN BIOTEK CROP master mix kit according to the manufacturer's instruction. The PCR tubes were inserted into the PCR machine, and the random amplified polymorphic DNA (RAPD) program was set (5 min initial denaturation at 94◦C, 35 cycles at 95◦C for 3 min, annealing for 1 min at 40◦C, extension at 72◦C for 2 min, and final extension set at 72◦C for 7 min). The PCR products were then run at 2% agarose gel observed under UV to check for the presence or absence of bands. RAPD primers used in this experiment were OPA-07 (GAAACGGGTG), OPA-08 (GTGACGTAGG), and OPA-09 (GGGTAACGCC).

## Chlorophyll Analysis

Shoot samples were ground in liquid nitrogen. Chlorophyll a and b contents were quantified by adding 5 ml of 80% acetone to 300 mg of powder tissue samples and placed in a shaker in the dark for 15 min, then centrifuged at 4◦C for 15 min at a speed of 3,000 RPM. The supernatant was transferred to a new centrifuge tube and diluted with 80% acetone (1:5 ratio). The absorbance was measured using a spectrophotometer at two wavelengths: 663 nm (chlorophyll a) and 645 nm (chlorophyll b), and chlorophyll contents were calculated.

### Electrolyte Leakage Estimation

After CuNP treatment, 0.1 g of control and treated root tips were excised and placed in falcon tubes containing 10 ml deionized water. The initial electrical conductivity (CI) was measured at time zero. After 48 h, C<sup>N</sup> was measured. The samples were then autoclaved at 121◦C, allowed to cool and the final electrical conductivity (CF) was measured. Three readings of the three replicates were measured. Degree of electrical leakage was calculated using the following formula: E<sup>T</sup> = ((CN−CI)/CF) × 100. The experiments were performed in triplicate.

### DAB Staining

After the CuNP treatment, the plant roots were washed well with tap and distilled water to get rid of the surface attached NPs. A small portions of the roots were excised and stained with 3, 3<sup>0</sup> -diaminobenzidine (DAB) (Daudi and O'Brien, 2012). Two ml of the DAB staining solution (1 mg/ml with 10 mM Na2HPO4) were added into the portion of the root samples and kept in desiccator for 5 min to allow the DAB into the tissues. Hence the DAB is light sensitive the samples were covered with

aluminum foil. The samples were kept in a shaker for 4–5 h for incubation. After the incubation, DAB solution was removed and washing solution (ethanol:acetic acid:glycerol = 3:1:1) were added and kept for 15 min at 90◦C for chlorophyll removal. Then the wash solutions were changed and observed under light microscope.

## Lipid Peroxidation and H2O<sup>2</sup> Content

Lipid peroxidation was analyzed by measuring malondialdehyde (MDA) content according to Zhou and Leul (1998). 100 mg of control and treated plant tissues were homogenized with 1 ml of 0.1% (w/v) trichloroacetic acid (TCA) then centrifuged at 10,000 rpm for 15 min. In another 15 ml Falcon tube, 2 ml of TBA (0.5%) in TCA (20%) were added along with 800 µl of the supernatant. This mixture was incubated at 80◦C water bath for 30 min and then immediately cooled into ice for 5 min. Then the mixture was centrifuged for 13,500 rpm, 5 min at 4 ◦C. The supernatant was transferred into cuvettes and measured absorbance at 532 and 600 nm after blanking. The level of MDA was represented as nmol g−<sup>1</sup> FW using extinction coefficient of 155 mM−<sup>1</sup> cm−<sup>1</sup> .

For determination of H2O<sup>2</sup> content, 0.1 g of ground tissues were mixed with 2 ml of 0.1% TCA (w/v) and then the homogenate were centrifuged at 12,000 rpm for 15 min (Velikova et al., 2000). 0.5 ml of the supernatant was mixed with 0.5 ml of potassium phosphate buffer (10 mM, pH 7) and 1 ml of 1 M potassium iodide (KI) was added and checked their absorbance at 390 nm. The H2O<sup>2</sup> content was determined using extinction coefficient of 0.28 µm cm−<sup>1</sup> and it was measured as nmol g−<sup>1</sup> FW.

### RNA Extraction and Quantitative Real-Time PCR (qRT-PCR)

0.1 g of the −80◦C stored control and CuNP treated plant tissues were ground in liquid nitrogen and were used to extract total RNA using Qiagen RNeasy <sup>R</sup> Plant Mini kit protocol. RNA concentrations were checked using NanoDrop, and the final concentrations of 1 µg RNA were used for cDNA synthesis using InvitrogenTM SuperScriptTM II Reverse Transcriptase according to manufacturer's protocol. Real-time PCR was performed with the following primer pairs; copper-zinc SOD (Cu-Zn SOD)\_F (GGTTGCTGGTGATGATGGTACTG), Cu-Zn SOD\_R (TGCATGGACAACAATAGACCTTCC), Actin\_F (GGAAATACAGTGTCTGGATTGGAG), Actin\_R (TGAACTTAGAAGCACTTCCTGTG) with Power SYBRTM Green PCR Master mix (Applied BiosystemsTM) using BioRad CFX96 TouchTM Real-Time PCR detection system. The real time data were analyzed using 2−11<sup>C</sup> <sup>T</sup> method (Livak and Schmittgen, 2001).

## RESULTS

### Characterization of CuNP

SEM analysis confirmed that the purchased CuNP sizes were in the 10–30 nm range (**Figure 1A**). Scanning Electron Microscope-Electron Dispersive X-ray Spectroscopy (SEM-EDS) analysis also showed the presence of oxygen along with copper, indicating that at least part of the CuNP was oxidized by air (**Figure 1B**).

#### Biomass Analysis of CuNP Treated C. sativus Plants

Plants treated with various concentrations of CuNP showed a clear senescence effect (leaf yellowing) at the end of the treatment period (4 days), with brown spots observed compared to control plants (**Figure 2A**). The total biomass of the C. sativus plants was measured before and after treatment (**Figure 2B**). Biomass analysis showed significant biomass reduction of C. sativus plants treated with 100, and 200 mg/ L CuNP for 4 days compared with the biomass of the same plants before treatment. Although, the biomass of the plants treated with 50 mg/L CuNP showed decreased biomass as well, but it was not significant.

Furthermore, as expected, the non-treated control plants (0 mg/ L CuNP) showed an enhanced but not significant increase in the biomass after the 4 days.

#### Copper Detection and Accumulation

To investigate whether C. sativus plants treated with CuNP mediate CuNP accumulation, XRF microscopy was utilized to measure the relative elemental composition represented as mass% of C. sativus shoot and root tissues treated with CuNP (**Figures 3A,B**). Besides copper, other elements of the Hoagland solution, such as Ca, K, Si, P, S, and Cl were also detected. Interestingly, the XRF analysis of shoots and roots exhibited more copper in roots compared to shoots (**Figures 3A,B**). To confirm and to quantify the Cu levels, the concentrations of Cu in shoots and roots of C. sativus plants were measured after 4 days of CuNP treatment using AAS (**Figures 3C,D**). The results showed increased Cu accumulation in these plants in proportion to the CuNP concentrations (100 and 200 mg/L). Again, it was observed that the accumulation of CuNP was more in the roots compared to the shoots. SEM analysis of C. sativus root samples confirmed the presence of CuNP in aggregated form and the sizes ranged between 80–140 nm in the treated plant tissues (**Figure 4**). However, no CuNP were observed in the control roots.

## Genotoxicity Analysis by RAPD

Random amplified polymorphic DNA was employed to assess the CuNP toxicity effect on the genomic level of C. sativus. Genomic DNA was extracted from the root of both control, and CuNP treated samples. Various set of primers were used to amplify all DNA samples. Control sample amplified OPA- 07 primer showed five bands ranging between 300 to 800 bps. For 50 mg/L treatment of CuNP sample DNA amplified same bands as the control DNA sample as well as two additional bands at 200 and much above 1000 bp marker band. Similar band patterns were observed for DNA samples treated with 100 mg/L CuNP. In the case of the 200 mg/L CuNP treated DNA sample amplified same bands of control DNA with the exception of 800 bp and the additional 200 bp band (**Figure 5A**). Control sample DNA for OPA- 08 primer amplified three high intensity bands between 200 to 800 bps. For 50 mg/L treatment of CuNP sample DNA amplified several bands ranging from 300 to above 1000 bps marker band. In case of 100 and 200 mg/L treated with CuNP sample amplified almost similar band patterns of same intensity (**Figure 5B**).

Control sample DNA amplified with OPA- 09 primer exhibited six bands ranging from 100 to 1000 bps. Samples treated with 50 and 100 mg/L CuNP showed two additional bands at 400 and 500 bps. However, 200 mg/L treated sample DNA amplified same band pattern as 50 and 100 mg/L treated sample DNA along with an additional band at 600 bps (**Figure 5C**).

### Effect of CuNP on Chlorophyll Content

When the plants were treated with CuNP for 4 days, induced senescence between control and treated plants was observed. To quantify the levels of senescence, the contents of chlorophyll a and b were measured. Compared to the untreated control plants, shoots of CuNP treated plants showed a significant decrease in chlorophyll a and b contents as the concentration of CuNP increased (**Figure 6**).

#### Electrolyte Leakage Analysis

To measure the plasma membrane integrity of the C. sativus plants treated with CuNP, electrolyte leakage analysis was performed. The results showed increased electrolyte leakage in CuNP treated plants compared to control plants (**Figure 7**). Hence, CuNP induced damage to root plasma membrane of C. sativus as indicated by the significant increase in electrolyte leakage in 50 and 200 mg/L CuNP treatment.

### Effect of CuNP on H2O<sup>2</sup> and MDA Contents

To examine whether CuNP induced oxidative stress on C. sativus plants, H2O<sup>2</sup> and lipid peroxidation contents were measured (**Figure 8**). H2O<sup>2</sup> levels in both shoots and roots of all the concentrations we checked were higher compared to control plants. The levels were almost four times higher than control plants of shoots and roots (**Figures 8A,B**).

In terms of lipid peroxidation, we measured MDA levels in both control and CuNP treated C. sativus plants. The root and shoot samples of C. sativus treated plants showed significant increase in all the concentrations of CuNP as compared to the control plants (**Figures 8C,D**).

## H2O<sup>2</sup> Detection by DAB Staining

To visualize the H2O<sup>2</sup> accumulation in root tips due to CuNP treatment, we performed the histochemical staining with DAB reagent. The roots observed under the light microscope showed a clear difference between control and treated plants. CuNP

FIGURE 4 | Scanning Electron Microscope (SEM) image showing (A) no CuNP were observed in the control roots, and (B) CuNP were observed in aggregates in 200 mg/L CuNP treated root samples of C. sativus.

treated C. sativus roots stained more than control plants in a dose-dependent manner as the CuNP concentration increases it induces higher H2O<sup>2</sup> accumulation (**Figure 9**).

#### Impact of CuNP on Superoxide Dismutase (SOD) Gene Expression

Superoxide dismutase is one of the major enzymatic antioxidants. In order to evaluate the regulation of C. sativus Cu-Zn SOD gene in response to CuNP treatment, quantitative RT-PCR (qRT-PCR) was performed to analyze the changes in transcripts of Cu-Zn SOD in response to 50, 100, and 200 mg/L of CuNP in C. sativus shoots and roots (**Figure 10**). Expression levels of Cu-Zn SOD were induced in shoots under 50, 100, and 200 mg/L of CuNP compared to control untreated plants (0 mg/L of CuNP) (**Figure 10A**). Similarly, in 50 mg/L CuNP treated roots, the expression level of Cu-Zn SOD was increased by six-fold compared to control untreated roots. However, a clear decrease but still significantly higher than control untreated roots was detected in Cu-Zn SOD expression level in response to 100, and 200 mg/L of CuNP (**Figure 10B**).

## DISCUSSION

Generally, metal NPs have really strong bonds that allow them to reside within the environment for a long time as they are hardly degraded or destroyed (Svintradze and Pidaparti, 2010). This puts the ecological system in great danger as metal NPs possess highly reactive or catalytic properties compared to their bulk material, and thus can be potentially toxic. The current research

of CuNP].

bars represent standard errors of mean values of four replicates. Statistically significant difference was calculated at <sup>∗</sup>P ≤ 0.05, ∗∗P ≤ 0.01.

highlighted the phytotoxicity effect of CuNP on cucumber (C. sativus) plants under hydroponic growth conditions, due to the limited studies available concerning the implications of CuNP and its phytotoxic effect on C. sativus plants. We have chosen the hydroponic system for this study to obtain a uniform concentration of NPs compared to soil systems. If we perform such experiments in soil grown plants, it's difficult to attain the uniformity in the concentration of NPs as well as the clear impact of toxicity on plants. Furthermore, for maintaining the uniformity of the NPs by growing plants on agar media, though its well suited for seedlings stage (Taylor and Foy, 1985; Munzuroglu and Geckil, 2002; Lee et al., 2008) rather than mature plants (6–7 week old in our study). C. sativus was selected for our study as one of the plant species recommended by McLean and Bledsoe (1996) for phytotoxicity studies. There are high chances of Cu oxidation by air to become copper oxide (Miyagawa et al., 2016). The purchased CuNP got oxidized to form copper oxide by air exposure as indicated by SEM-EDS analysis.

In the present study, we observed that plants exposed to CuNP exhibited higher toxicity based on the concentration of the CuNP with regard to its biomass reduction, which was similar to the reported earlier in hydroponically and soil grown cucumber plants treated with CuONP of 50 nm size (Kim et al., 2013). X-Ray Fluorescence (XRF) is one of the advanced techniques to estimate the relative quantity of elements present in the sample as semi-quantitatively (mass%). Earlier, Attaelmanan and Kawam, 2012 evaluated the elemental composition of Calotropis procera using such XRF microscope (Attaelmanan and Kawam, 2012). We used the same technique and the same instrument to show the elemental composition of C. sativus shoot and root samples treated with CuNP. There were several reports regarding micro XRF analysis of NPs treated plant tissues. micro XRF analysis of TiO<sup>2</sup> of hydroponically grown cucumber plants showed titanium accumulation in their leaf trichomes (Servin et al., 2012). The translocation of TiO<sup>2</sup> NP and multi-walled carbon nanotubes in red clover and wheat plant tissues were studied using XRF techniques (Gogos et al., 2016). Wang et al. (2017) demonstrated that phytotoxicity of silver sulfide NP in dicot and monocot plants and used XRF analysis to show the NP accumulation. Our elemental analysis of CuNP treated samples revealed 64 and 1.38% of Cu in root and shoot tissues, respectively.

To characterize the CuNP in cucumber treated plant tissues we performed Scanning Electron Microscopy (SEM) analysis. SEM analysis of CuNP treated root samples showed the aggregates of CuNP. Several reports already demonstrated the detection of NPs in aggregated form within the plant tissues regardless the small size of the NPs used for the treatment. González-Melendi et al. (2007) reported that charged NPs can be targeted into specific locations of plant tissues for delivery purposes and found that they tend to form aggregates in their targeted positions. TEM analysis of carbon-coated iron NP into Cucurbita pepo L plants showed the aggregates of NPs in various parts of the plants (Corredor et al., 2009). Another study revealed that green synthesized gold NP were found to increase rice germination, where the TEM analysis revealed the aggregated form of gold NP in root tissues of rice (Nji Tsi et al., 2017). Similarly, in our study SEM-EDS analysis showed the aggregation of CuNP in treated C. sativus.

Random amplified polymorphic DNA technique was employed in this study to check for any genetic alteration caused by CuNP. Unlike traditional PCR, RAPD technology is rapid, reproducible and does not require specific knowledge of the DNA sequence, which makes it ideal for examining and estimating the genomic variation in genotoxic studies (Atienzar et al., 1999). RAPD has been used to confirm the genotoxic effect of ZnO and CuO NP on buckwheat (Fagopyrum esculentum) seedlings (Comrey and Lee, 2013). Additionally, RAPD analysis of zucchini (C. pepo) plants treated with TiO<sup>2</sup> NP exhibited DNA changes compared to the control plants (Moreno-Olivas et al., 2014). Genotoxicity of CeO<sup>2</sup> and ZnO NP on soybean seedlings was demonstrated by the appearance of new DNA bands through RAPD analysis by López-Moreno et al. (2010).

Furthermore, the genotoxicity of nCeO<sup>2</sup> and nTiO<sup>2</sup> NP were reported in seedlings of Hordeum vulgare L. using RAPD approach (Marchiol et al., 2016). In our investigation, three standard primers (OPA7, OPA8, and OPA9) were employed, and different bands appeared/disappeared when comparing the treated plants to the untreated control plants, rendering CuNP toxic to C. sativus at high exposure concentrations. The presence of these new bands may exhibit alterations in the priming sites leading to new annealing conditions in addition to homologous recombination which lead to the appearance of new bands (Atienzar and Jha, 2006). Our plants were more sensitive to the effects of CuNP as more band alterations were observed when compared to the control. The only difference between the treated and control plants was the presence or absence of CuNP which support that the changes in the DNA were caused by the effect of the CuNP.

The assessments of chlorophyll content of plants exposed to NPs were included as new parameters for phytotoxicity analysis (Ruttkay-Nedecky et al., 2017). When NPs are adsorbed to the root surface, it interrupts the absorption of macro and micronutrients required for plant's development reducing the nutrient uptake and thus resulting in lower chlorophyll content. Martinez-Fernandez et al. (2016) investigated the uptake of water contaminated with iron oxide NP by the roots of Helianthus annuus. They found that macronutrients such as Ca, K, Mg, and S reduced subsequently in the plant's shoot, and accordingly the chlorophyll pigmentation contents were decreased as well (Martinez-Fernandez et al., 2016). Our study has also demonstrated the negative effect of CuNP on chlorophyll a and chlorophyll b contents of C. sativus. The decrease in chlorophyll contents was obvious to the eyes after assessing treated and untreated plants.

The plasma membranes permeability was analyzed by testing the electrolyte leakage of the control and CuNP treated root tissues. It has been reported that 500 mg/L cerium oxide (nCeO2) NP increased electrolyte leakage in the roots of rice seedlings (Hernandez-Viezcas et al., 2013). Additionally, a significant increase in electrolyte leakage was reported in asparagus lettuce root cultured in an agar medium treated with cerium oxide NP (Cui et al., 2014). In our study, C. sativus was sensitive to CuNP and significantly increased electrolyte leakage which caused the high reactivity of CuNP. CuNP changed the permeability of the plant membrane and as a result, damaged the plant's cell membranes which increased the probability of NPs entering into the cells. C. sativus showed increased permeability when plants were treated with CuNP which could be then accumulated into the roots and could also be translocated to the shoots, thus affecting the cell metabolic pathway. It has been already

reported that CuONP is able to be translocated from roots to shoots (Shi et al., 2014). Our results demonstrated that CuNP were highly accumulated in roots with slight accumulations in the shoot system, indicating that only a small quantity of CuNP was translocated to the shoot of C. sativus during the treatment time period of our experiment (4 days). Kim et al. (2012) reported the increased accumulation of Cu and Zn ions in roots of C. sativus when treated with CuO and ZnO NP. It's already reported that plant systems are tend to absorb CuONP according to their concentration in the nutrient solution (Shi et al., 2014). Plants with more thin roots accumulate more metals than those with less thick roots as the surface area is decreased.

It was apparent that our C. sativus plants had numerous thin and long root systems resulting in higher root accumulation of Cu.

It has been already highlighted earlier that metal toxicity in plants lead to enhanced production of ROS which is causing oxidative stress by increased electrolyte leakage, protein oxidation, lipid peroxidation, DNA damage and finally cell death (Meriga et al., 2004; Sharma et al., 2012; Wani et al., 2018). Cu is catalyzing the overproduction of ROS, such as H2O<sup>2</sup> by Haber-Weiss and Fenton reactions leading to lipid peroxidation, damaging nucleic acids, and oxidizing proteins (Gill and Tuteja, 2010; Fidalgo et al., 2013). Hence we measured

the major molecules associated with oxidative stress such as MDA and H2O<sup>2</sup> in C. sativus plants upon CuNP treatment. The highest levels of MDA were observed in C. sativus shoots and roots treated with 100 and 200, and 50 and 100 mg/L CuNP, respectively. The increase in MDA levels is directly proportional to the concentration of the CuNP used for the treatment. Similarly, such effect has been already shown in rice seedlings treated with CuONP which induced MDA levels compared to control plants (Wang et al., 2014). Likewise, the major reactive oxygen molecule, H2O2, levels were measured in control and CuNP treated plants. Our results demonstrated that all the treatment conditions showed increased H2O<sup>2</sup> levels in both roots and shoots samples. Such induction of oxidative stress mediated H2O<sup>2</sup> levels were also confirmed by DAB staining of root tips where dark brown stains were observed in all the treated root tip samples. Similarly, TiO2NP treated Hydrilla verticillata showed increased H2O<sup>2</sup> which was due to oxidative stress caused by the NPs (Spengler et al., 2017).

In order to reduce the oxidative stress, plants develop a defense mechanism involving the antioxidant enzymes such as SOD to scavenge excessive ROS. SODs are classified into three isoforms; Cu-Zn SOD, manganese SOD (Mn SOD), and iron SOD (Fe SOD) based on their metal cofactors (Alscher et al., 2002). In general, SOD is catalyzing the disproportionating superoxide anion free radicals into hydrogen peroxide and molecular oxygen (Gill and Tuteja, 2010). Our results showed a significant upregulation of Cu-Zn SOD gene expression in roots and shoots of CuNP treated C. sativus plants, demonstrating the increased ROS production and the conversion of superoxide anion free radicals into hydrogen peroxide which is displayed in our study by the increased H2O<sup>2</sup> levels in both root and shoot samples of CuNP treated C. sativus. Similarly, a significant increase in SOD gene expression was reported in P. sativum L roots treated with 100, 200, 400, and 500 mg dm−<sup>3</sup> CuONP (Nair and Chung, 2015).

#### CONCLUSION

This study was performed to assess the phytotoxicity of CuNP on C. sativus. We investigated the toxicity on the

#### REFERENCES


physiological, phenotypical, biochemical, and genomic levels. It was found that tested CuNP of the size 10–30 nm were toxic to C. sativus. CuNP showed a decrease in the total biomass of the C. sativus treated plants. XRF, Scanning Electron Microscope (SEM), and Atomic Absorption Spectroscopy (AAS) analysis demonstrated that CuNP were accumulated in the C. sativus plant tissues, with higher accumulation level in root tissues. RAPD PCR analysis confirmed the genotoxic effect of CuNP which induced genomic DNA modifications in C. sativus. Additionally, CuNP showed a significant decrease in chlorophyll a and b contents, increase in H2O<sup>2</sup> and MDA contents, as well as an increase in electrolyte leakage which induced damage to cucumber root plasma membrane. Collectively, this demonstrated that CuNP induced oxidative stress in C. sativus. Finally, Cu-Zn SOD gene expression was induced under CuNP treatment. The availability of C. sativus genome information could enhance our understanding of CuNP phytotoxicity through deep molecular and gene expression analysis studies.

### AUTHOR CONTRIBUTIONS

KM conceived and designed the experiments. KM, KR, AE, SW, EI, and HH performed the experiments. KM, ME-N, and KR analyzed the data. ME-N performed the AAS metal analysis. HA performed the XRF and SEM analysis. KM wrote the manuscript. KM, ME-N, KR, and HA edited the manuscript.

#### FUNDING

This research was financially supported by the Office of VC for Research and Graduate Studies at the University of Sharjah.

### ACKNOWLEDGMENTS

We would like to thank Mr. Mohamed Shameer, Ms. Mona Mahfood, and Ms. Fatima Shaheen for their assistance.



C-N cross-coupling catalysis. Crystal Res. Technol. 46, 1167–1174. doi: 10.1002/ crat.201100258


**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 © 2018 Mosa, El-Naggar, Ramamoorthy, Alawadhi, Elnaggar, Wartanian, Ibrahim and Hani. 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) and the copyright owner(s) 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.

# Cellular Fractionation and Nanoscopic X-Ray Fluorescence Imaging Analyses Reveal Changes of Zinc Distribution in Leaf Cells of Iron-Deficient Plants

Gianpiero Vigani1,2 \*, Sylvain Bohic<sup>3</sup> , Franco Faoro<sup>2</sup> , Bart Vekemans<sup>4</sup> , Lazlo Vincze<sup>4</sup> and Roberto Terzano<sup>5</sup> \*

<sup>1</sup> Plant Physiology Unit, Department of Life Sciences and Systems Biology, University of Turin, Turin, Italy, <sup>2</sup> Department of Agricultural and Environmental Sciences, Production, Landscape, Agroenergy, University of Milano, Milan, Italy, <sup>3</sup> European Synchrotron Radiation Facility, NINA Beamline, Grenoble, France, <sup>4</sup> Department of Analytical Chemistry, Ghent University, Ghent, Belgium, <sup>5</sup> Department of Soil, Plant and Food Sciences, University of Bari, Bari, Italy

#### Edited by:

Hannetz Roschzttardtz, Pontificia Universidad Católica de Chile, Chile

#### Reviewed by:

Ferenc Fodor, Eötvös Loránd University, Hungary Isidro Abreu, Centre for Plant Biotechnology and Genomics, Spain

#### \*Correspondence:

Gianpiero Vigani gianpiero.vigani@unito.it Roberto Terzano roberto.terzano@uniba.it

#### Specialty section:

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

Received: 26 March 2018 Accepted: 10 July 2018 Published: 03 August 2018

#### Citation:

Vigani G, Bohic S, Faoro F, Vekemans B, Vincze L and Terzano R (2018) Cellular Fractionation and Nanoscopic X-Ray Fluorescence Imaging Analyses Reveal Changes of Zinc Distribution in Leaf Cells of Iron-Deficient Plants. Front. Plant Sci. 9:1112. doi: 10.3389/fpls.2018.01112

Multilevel interactions among nutrients occur in the soil-plant system. Among them, Fe and Zn homeostasis in plants are of great relevance because of their importance for plant and human nutrition. However, the mechanisms underlying the interplay between Fe and Zn in plants are still poorly understood. In order to elucidate how Zn interacts with Fe homeostasis, it is crucial to assess Zn distribution either in the plant tissues or within the cells. In this study, we investigated the subcellular Zn distribution in Fe-deficient leaf cells of cucumber plants by using two different approaches: cellular fractionation coupled with inductively coupled plasma mass spectrometry (ICP/MS) and nanoscopic synchrotron X-ray fluorescence imaging. Fe-deficient leaves showed a strong accumulation of Zn as well as a strong alteration of the organelles' ultrastructure at the cellular level. The cellular fractionation-ICP/MS approach revealed that Zn accumulates in both chloroplasts and mitochondria of Fe deficient leaves. Nano-XRF imaging revealed Zn accumulation in chloroplast and mitochondrial compartments, with a higher concentration in chloroplasts. Such results show that (i) both approaches are suitable to investigate Zn distribution at the subcellular level and (ii) cellular Fe and Zn interactions take place mainly in the organelles, especially in the chloroplasts.

Keywords: cellular fractionation, iron, nanoscopic X-ray fluorescence imaging, zinc, plant

### INTRODUCTION

Among nutrients, iron (Fe) represents an essential element for the life cycle of plants, because it is a key cation to ensure electron flow in photosynthetic and respiratory pathways. For this reason, Fe can be a limiting factor for biomass production as well as for the quality of plant products. On the other hand, Fe is also potentially toxic because of its reactivity with oxygen, which catalyzes the formation of reactive oxygen species (Briat et al., 2015a).

**74**

The chlorosis observed in younger leaves represents the most typical visual symptom of Fe deficiency in plants. Indeed, Fe deficiency impairs chlorophyll synthesis, leading to interveinal chlorosis in developing leaves (Rodriguez-Celma et al., 2013) and decreased photosynthesis rates (Terry, 1980). However, the excess of several heavy metals can induce such chlorosis as well (Vert et al., 2002; Leskova et al., 2017), probably because Fe shares a number of similarities (i.e., electronic configuration, availability in the soil, or uptake mechanisms by plant) with other transition metals.

Accordingly, the changes in the Fe nutritional status of a plant are usually associated with changes in a given subset of metals, such as Mn, Zn, Cu, Mo, Co (Baxter et al., 2008a,b; Murgia and Vigani, 2015), pinpointing to the existence of a multi-level interaction among nutrients in plant. For example, a strong interaction between Fe and Mo has been recently characterized in cucumber plants (Vigani et al., 2017). Also, Tomasi et al. (2014) observed that zinc (Zn) accumulates in the leaves of Fe-deficient cucumber plants as compared to Fe-sufficient and Fe-resupplied plants.

Iron deficiency is often determined by the physico-chemical characteristics of the soil. Indeed, in alkaline and calcareous soils Fe is mainly present as insoluble oxi-hydroxide compounds leading to a very low bioavailability of Fe for plant uptake. To survive in these stress conditions, plants have developed adaptive mechanisms to cope with the low bioavailability of Fe in order to increase its acquisition (Tsai and Schmidt, 2017 and references therein). Dicots and non-graminaceous monocots have evolved a reduction-based mechanism known as Strategy I. The main responses induced by Fe deficiency in Strategy I species are the increase in (i) a Fe(III)-chelate reductase activity (FC-R), the function of which is the NAD(P)H-dependent reduction of the ferric to the ferrous form; (ii) the activity of a ferrous iron transporter (IRT1), and (iii) a H+-ATPase activity whose task is to extrude protons useful both to acidify the rhizosphere and to generate and maintain a transmembrane electrochemical gradient which facilitates the Fe2<sup>+</sup> uptake (Tsai and Schmidt, 2017 and references therein).

IRT1 has a weak substrate specificity and contributes therefore to the accumulation of a broad range of divalent transition metals, including Zn (Vert et al., 2002; Arrivault et al., 2006; Haydon et al., 2012). Similar to Fe, Zn is still an essential microelement for cell life, being a crucial cofactor for all six classes of enzymes: oxidoreductases, hydrolases, transferases, lyases, isomerases, and ligases (Coleman, 1998). Moreover, Zn plays an important structural role in regulatory proteins (Berga and Shi, 1996). However, high concentrations of Zn can be toxic for the cell causing oxidative stress (Sresty and Madhava Rao, 1999).

Zn is a highly effective cofactor since its coordination geometry is highly flexible (Fraústo da Silva and Williams, 1991). For this reason, Zn is an important structural component of small protein motifs, named Zinc fingers, characterized by the coordination of one or more Zn ions in order to stabilize protein structures. Zinc finger motifs are known to bind several target ligands, for example, DNA.

Thanks to these properties, Zn can play different roles when interacting with proteins: (i) catalytic, where Zn ions directly participate in the reaction (e.g., carbonic anhydrase); (ii) cocatalytic, where Zn plays a catalytic role together with several metal ions that interact with each other in a co-catalytic Zn site (e.g., alkaline phosphatase); (iii) structural, where Zn ions stabilize the tertiary structure of the enzyme in a similar way as the disulphide bonds (e.g., DNA-binding proteins; alcohol dehydrogenase) (Vallee and Auld, 1990; Escudero-Almanza et al., 2012).

Despite plant symptoms under Zn excess resemble symptoms of Fe-deficient plants, the mechanism underlying the interplay between Fe and Zn in plants is poorly understood. Indeed, it is known that Zn excess causes physiological Fe deficiency, probably due to an alteration of Fe transport form soil. The ability to take up iron and zinc by IRT1 is inhibited by the addition of excess zinc and iron, respectively (Rogers et al., 2000).

In order to understand how Zn interacts with Fe homeostasis, it is crucial to know how Zn is distributed in the plant tissues and in the cells of Fe-deficient plants.

The distribution of metals in plants can be studied by various imaging techniques, using histochemical methods or analytical techniques employing electrons, charged particles, or X-ray beams (Terzano et al., 2013). Nevertheless, scanty information is still available concerning the Zn distribution within plant cells (Wu et al., 2005; Li et al., 2006). Some authors stated, indeed, that one hurdle to the study of subcellular zinc homeostasis is the lack of techniques applicable to the study of this spectroscopically silent ion (Blaby-Haas and Merchant, 2013). X-ray fluorescence (XRF) methods are powerful analytical techniques for nondestructive elemental analysis with the capability to detect trace level concentrations of multiple elements. At third-generation synchrotron facilities, submicrometer resolutions and parts-permillion (ppm) to parts-per-billion (ppb) minimum detection limits can be obtained within a feasible experimental time frame.

Micro X-ray florescence (µ-XRF) analyses on cucumber plants showed an amount of Zn more than double in leaf tissues of Fe deficient plants as compared to plants resupplied with natural Fe-sources (Tomasi et al., 2014). Subcellular elemental distributions in plant samples have been imaged by µ-XRF for As in Ceratophyllum demersum (Mishra et al., 2016); Cd and Zn in arbuscular mycorrhizas (Nayuki et al., 2014); and As, Fe, Zn, Mn, and macronutrients in Oryza sativa (Moore et al., 2014). However, in all these studies, the X-rays were focused to a beam size ranging from about 500 nm to 20 µm, thus not allowing a clear visualization of the cell's internal structures and organelles. A new generation of synchrotron nanoprobes is nowadays available with ever-smaller spot sizes, down to a few tens of nanometers (Laforce et al., 2014). An example of such synchrotron X-ray nanoprobe is the newly developed ID16 NINA (nano imaging nano analysis) beamline at the European Synchrotron Radiation Facility (ESRF) of Grenoble (France) (Laforce et al., 2014).

In this work, we provide evidence about the localization and homeostasis of Zn at sub-cellular level in Fe-deficient cucumber leaves by mapping for the first time a plant cell with XRF at nanometers resolution, and by comparing XRF data with data from cell fractionation coupled to ICP/MS analysis.

## MATERIALS AND METHODS

fpls-09-01112 August 1, 2018 Time: 8:2 # 3

#### Plant Material and Growth Conditions

Seeds of cucumber (Cucumis sativus L. cv. Marketer) were surface-sterilized and sown in Agriperlite, watered with 0.1 mm CaSO4, allowed to germinate in the dark at 26◦C for 3 days, and then 70 seedlings were transferred to a box containing 20 L of the following nutrient solution: 2 mM Ca(NO)3, 0.75 mM K2SO4, 0.65 mM MgSO4, 0.5 mM KH2PO4, 10 µM H3BO3, 1 µM MnSO4, 0.5 µM CuSO4, 0.5 µM ZnSO4, 0.05 µM (NH4)Mo7O24; 0.1 mM Fe(III)-EDTA was added to control (+Fe) plants but omitted in Fe-deficient (–Fe) plants. The pH was adjusted to 6.0– 6.2 with NaOH. Aerated hydroponic cultures were maintained in a growth chamber with a day:night regime of 16:8 h and a photosynthetic photon flux density (PPFD) of 200 µmol photons m−<sup>2</sup> s −1 . The temperature was 18◦C in the dark and 24◦C in the light. Expanded leaves of 10 days-old cucumber plants, grown in the presence (+Fe) and in the absence of Fe (−Fe) in the medium were harvested. As previously described, 10-days-old cucumber leaves of plants grown under Fe deficiency displayed evident chlorosis symptoms and the chlorophyll content decreased by about 80% with respect to +Fe leaves (Vigani et al., 2013a).

#### Purification of Mitochondria and Chloroplasts

Mitochondria and chloroplasts were purified according to Rödiger et al. (2010) from expanded leaves of 10-days-old cucumber plants grown in the presence (+Fe) and in the absence of Fe (−Fe) in the medium. During purification steps, the following cellular fraction were harvested: total extract (TE, collected after homogenized tissues filtration), soluble fraction (SF, collected after mitochondrial precipitation and representing the cytosolic plus vacuolar lumen compartments), purified chloroplasts (Ch, collected at the interface between 35/80% percoll gradient), and purified mitochondria (Mit, collected at the interface between 23/40% percoll gradient) (**Supplementary Figure S1**). The purity degree was tested by Western blot analysis according to Vigani et al. (2015). Zinc and Fe content was determined on TE, SF, Ch, and Mit fractions.

#### Ionomics

Sampled tissues were dried and then mineralized in HNO<sup>3</sup> by using a Microwave Digestion System (Multiwave ECO). Chloroplast and mitochondrial fractions were mineralized in HNO<sup>3</sup> at 100–120◦C according to Vigani et al. (2017). Metal content was determined by inductively coupled plasma-mass spectrometry (ICP/MS, aurora M90 BRUKER) according to Vigani et al. (2013a, 2017). Data collected are from three independent experiments.

#### Transmission Electron Microscopy

Samples of +Fe and −Fe expanded leaf tissues were fixed in a mixture of 3% (v/v) glutaraldehyde and 2% (w/v) paraformaldehyde in 0.1 mM phosphate buffer, pH 7, overnight at room temperature. Samples were subsequently post fixed with 1% (w/v) osmium tetroxide in the same buffer for 1 h at 4◦C and dehydrated in a graded ethanol series before being embedded in SPURR resin (Electron Microscopy Sciences, Washington, PA, United States) according to Vigani et al. (2009). Ultrathin sections (100 nm thickness) were cut from at least three leaf samples from different plants and contrasted with uranyl acetate and lead citrate and examined with a Jeol JEM-100 SX TEM at 80 kV.

#### Nano X-Ray Fluorescence Imaging

Ultrathin sections were prepared as described for TEM analyses with a few modifications: osmium post-fixation was omitted and thin sections (around 200 nm thickness) were stained with toluidine blue before mounting them on SiN support meshes for X-ray analysis (DuraSiN, distributed by Electron Microscopy Sciences). Staining was necessary to visualize the portion of the sample to analyze at the synchrotron facility through the use of an optical microscope.

The nano-XRF analyses were performed at the nano-imaging beamline ID 16A of the ESRF in Grenoble (France). The synchrotron X-rays were focused at an energy of 17 keV to a beam size varying from 100 to 20 nm (depending on the sample) by means of an X-ray focusing optics based on a Kirkpatrick–Baez mirror system with a multilayer surface coating. A six elements silicon drift diode detector was used to collect the fluorescence signal.

XRF data were analyzed by using PyMca software (version 5.1.1). Zinc distribution maps are displayed by using the same grayscale for all the samples so that they can be directly visually compared. The relative amount of Zn in chloroplasts and mitochondria was obtained from the XRF sum spectrum of each organelle by dividing the peak area of the Kα line of the element multiplied per 10<sup>3</sup> for the area of the scatter signal, used to normalize the data for synchrotron current variations. The sum spectra were extrapolated from XRF maps by using PyMca and selecting the pixels forming the image of each organelle. All the spectra from each pixel were then summed to have the final sum spectrum. Eight chloroplasts and four mitochondria were considered for each sample. The number of mitochondria was lower because it was more difficult to distinguish them from other organelles (like peroxisomes) due to the low resolution of the XRF images, which did not allow to clearly distinguish the internal structures for all the organelles.

### RESULTS

### Transmission Electron Microscopy

Ultrathin sections obtained from mesophyll of +Fe and – Fe cucumber leaves were observed by TEM at different magnifications (**Figure 1**). From these observations, it is evident that the most striking morphological alteration in −Fe leaf tissues is related to chloroplast ultrastructure (**Figures 1b,d**). The organelles appeared swollen and some of them almost roundish with a dense stroma and a very low number of thylakoids, usually not organized in grana (**Figure 1d**). In some instance, also mitochondria showed altered organization with reduced number of cristae (**Figure 1d**), while no other apparent cytopathic effect

from cucumber plants grown in control condition (+Fe) or under Fe deficiency (–Fe). Control leaf tissues (a,c) and –Fe leaf tissues (b,d). All scale bars = 1 µm (a,b) 0.1 µm (c,d). M, mitochondrion; Ch, chloroplast; V, vacuole.

was detectable in cell ultrastructure, in comparisons with +Fe tissue (**Figures 1a,c**).

#### Leaf Ionome Profile

To evaluate the metabolic consequences of induced Fe deficiency, ionome of +Fe and −Fe plants was profiled by quantifying the content of the macronutrients sodium (Na), magnesium (Mg), potassium (K), calcium (Ca) as well as of the micronutrients manganese (Mn), zinc (Zn), copper (Cu), Fe and Mo in leaves (**Table 1**).

As expected, the growth of –Fe plants modulated Fe concentration, as seen by the reduced Fe content (by about 59%) in −Fe leaves when compared with +Fe ones.

The increased levels of K, Mg, Mn, Zn, and Cu, under Fe deficiency, indicate that these elements are dependent on Fe availability in leaves, while Na and Ca are not affected (**Table 1**). Moreover, Mo levels decreased in −Fe leaves relative to control leaves. The ionome profile of the leaves revealed that under Fedeficiency Zn was the element which displayed the highest overall increase respect to the +Fe plants.



Value significantly different (p < 0.05) in Student's t-test are indicated with<sup>∗</sup> . Percentage changes of Fe and Zn content between +Fe and −Fe treatment are highlighted in bold.

#### Cellular Fractionation-ICP/MS Analyses

In order to further investigate the Zn distribution in −Fe cells in relation to Fe, the subcellular distribution of Fe and Zn was assessed by ICP/MS analysis on four different cellular fractions collected from leaf tissues of +Fe and – Fe plants: TE, SF, Ch, and Mit (see Materials and Methods section and **Supplementary Figure S1**). Significant differences in protein yield for each fraction between +Fe and –Fe samples were not observed, as the variability among biological replicates was high. Forty grams of fresh weight of leaf material (regardless of treatments) yielded proteins amount of 0.9– 1.8 g TE, 0.8–1.5 g SF, 1–3 mg Ch, and 70–100 µg Mit [i.e., 23–45 mg TE protein per gram of fresh weight (mg/g), 20– 38 mg/g of SF proteins, 250–300 µg/g of Ch proteins, and 1.7–2.5 µg/g of Mit proteins]. In the TE fraction, a significant decrease of Fe concentration (expressed as ng/mg of protein) as well as a significant increase of Zn was detected in – Fe samples, while in the SF only a significant increase of Zn concentration was observed (**Figure 2A**). Interestingly, Fe concentration did not change in –Fe respect to +Fe in SF samples. Under Fe deficiency, Zn accumulated in both chloroplast and mitochondrial-enriched fractions (**Figures 2B,C**). However, in – Fe samples, Fe content in the chloroplasts and mitochondria dramatically decreased as compared to +Fe plants, being below the detection limit in chloroplasts, and reduced by 90% in the mitochondria (**Figures 2B,C**). Although possible contamination among fractions might occur under cellular fractionation, such analysis revealed that under Fe deficiency: (i) Fe content was greater (at least detectable) in mitochondria than in chloroplasts; and (ii) Zn accumulated more in chloroplasts than in mitochondria.

#### Nano XRF Imaging of Leaf Cells and Zn Distribution in Chloroplasts and Mitochondria

Leaf mesophyll cells were imaged by nano-XRF and elemental distribution maps were obtained as reported in **Figure 3** for a Fe-deficient plant. Clear distribution maps were obtained

only for few elements: Zn (**Figure 3a**), S (**Figure 3c**), and Ca (**Figure 3d**). An RGB map showing all the three elements together is also reported in **Figure 3b**. For other elements like Fe (**Figure 3e**) and Cu (**Figure 3f**), no subcellular distribution was visible except for contaminating particles (also visible in the Zn, S, and Ca maps), likely deriving from the staining process or atmospheric particulate contamination. Despite this contamination, while for certain elements like Fe and Cu, the concentration within the cell was too low to allow distribution mapping by this technique, Zn distribution appears to be not affected by staining and/or particulate except for the presence of highly concentrated spots, corresponding to the same hotspots of Fe, Cu, S, and Ca (**Figure 3**). Sum spectra of single chloroplasts and mitochondria from −Fe and +Fe plants are presented in **Figure 4**. Beside Zn, S, Ca, Fe, and Cu, other elements were detected: Si, Cl, K, Mn, Ni, and Pb. The Si signal derives from the SiN mesh used as support for the ultrathin sections. Pb signal is due to the staining solution used to allow the visualization of the cell structures within the thin section. Sum spectra showing the elemental composition of cytoplasm as compared to the full investigated areas (also containing organelles and vacuoles) are presented as **Supplementary Figure S2**. **Figure 5** shows Zn distribution maps within leaf organelles for +Fe (**Figures 5a,b**) and −Fe (**Figures 5c,d**) samples. As can be seen, Zn concentration in all the organelles of −Fe plants is higher than that in +Fe plants (darker pixels correspond to higher concentrations). The distribution maps of S, Ca, Fe, and Cu in the same areas are presented as **Supplementary Figure S3**. Chloroplasts are clearly discernible because of their large size and typical internal structures (thylakoid membranes are clearly visible). Peroxisomes appear as small rounded objects while mitochondria show a more elongated shape (**Figure 5**).

On the basis of their characteristic size and shape, various chloroplasts and mitochondria were identified in different cells and the XRF sum spectra of each organelle were used

FIGURE 3 | Nano X-ray fluorescence (K<sup>α</sup> lines) metal distribution maps of Fe-deficient leaf cells for Zn (a), S (c), Ca (d), Fe (e), and Cu (f). Darker pixels correspond to higher elemental concentrations. Scale bars based on counts are reported for each element. An RGB image of the distribution of Zn (blue), S (green), and Ca (red) is also presented (b). N, nucleus; V, vacuole.

to extrapolate Zn relative concentration. As can be seen from **Figure 6**, both in chloroplasts and mitochondria Zn concentration is higher in −Fe plants compared to +Fe plants. In particular, Zn concentration in −Fe chloroplasts was about 85% higher than in control +Fe chloroplasts and about 55% higher in mitochondria. These data thus confirm what was observed with cell fractionation followed by ICP/MS analyses, i.e., a higher increase of Zn concentration in chloroplasts compared to mitochondria in −Fe plants.

#### DISCUSSION

In this work, we aimed to provide new evidences about the interplay between Fe and Zn in plants by characterizing the subcellular localization of accumulated Zn in leaf tissues of Fedeficient cucumber plants. The investigation of subcellular Zn homeostasis in plants is quite difficult as specific techniques applicable for such ion are still lacking (Blaby-Haas and Merchant, 2013).

Therefore, we compared the results from two different analytical approaches in order to assess Zn distribution at sub-cellular level under Fe deficiency: ion determination by ICP/MS technique on different cellular fractions separated by ultracentrifugation, and nanoscopic XRF imaging method.

ICP/MS has been established as the most reliable technique for quantifying metals, metalloids, and some non-metals in a wide range of samples with a wide working range of concentrations, high sensitivity, and low interferences. The use of ICP/MS, by allowing simultaneous measurement of multiple elements in the same sample tissue, has significant advantages for biological applications. However, few reports analyzed the metal content at subcellular level in plants (Tan et al., 2010; Vigani et al., 2017).

According to the data obtained by cellular fractionation and organelles purification followed by ICP/MS analysis, under Fe deficiency the Fe content decreased in TE, Ch, and Mit cellular fractions, while Zn content increased in all the cellular fractions collected. In particular, Fe content in chloroplasts and mitochondria was strongly reduced, with a decrease of 90% in mitochondria and close to 100% in chloroplasts, not being detected anymore by ICP/MS. It is well known that chloroplasts represent the major sink of Fe in leaf cells (Nouet et al., 2011; Vigani et al., 2013b). It has been estimated that 68% of the Fe present in the vegetative shoot of Arabidopsis plant was found to be in the chloroplast (Shikanai et al., 2003). Therefore, our results suggest that Fe deficiency weakens Fe content more in

chloroplasts than in mitochondria or, alternatively, that under Fe starvation conditions mitochondria might have a priority for Fe allocation with respect to chloroplasts. Accordingly, in the green microalgae, mitochondria seem to be more resistant than chloroplast to Fe deficiency, suggesting a preference for Fe delivery to mitochondria when Fe is low (Nouet et al., 2011). Additionally, it has been stated that chloroplast play a central role in the regulation of Fe economy strategy of the cell under

Fe deficiency in Chlomydomonas reinhardtii (Page et al., 2012). Recently, Hantzis et al. (2018) observed that in Arabidopsis Fe deficiency first affect photosynthesis than respiration, suggesting that photosynthesis is more dispensable under Fe deficiency compared to respiration. Accordingly, our results along with those provided by Hantzis et al. (2018) demonstrated that chloroplast is the main target of Fe deficiency highlighting the hypothesis that respiration might be preserved compared to photosynthesis under such nutritional stress.

Additionally, ICP/MS analysis revealed that Zn accumulation under Fe deficiency was significant both in chloroplast and in mitochondrial compartments. Therefore, the opposite behavior between Fe and Zn was observed in both organelles: when the Fe content decreased, the content of Zn increased. Interestingly, this did not occur in the SF, suggesting that both organelles may represent important cellular compartments where Fe and Zn interplay take place.

Nanoscopic XRF Imaging analyses of cucumber leaf confirmed that Zn accumulate mainly in chloroplasts and mitochondria of −Fe plants. Unfortunately, Fe distribution was not detectable because of the low concentration and the lower sensibility of the technique for this element, as compared to Zn. Beside the higher fluorescence yield of Zn (0.435) compared to Fe (0.32), also the energy of the X-rays used in the experiment (17 keV) favored the excitation of Zn atoms (9.657 keV) rather than Fe ones (7.109 keV). Overall, our results showed that Fe deficiency significantly affects the distribution of Zn in leaf cellular compartments: in particular chloroplast might represent a Zn sink under Fe deficiency.

An important issue to consider when dealing with synchrotron XRF analyses is the possible creation of artifacts. In this type of experiments, artifacts may derive mainly from sample preparation and photon damage (Lombi et al., 2011; Zhao et al., 2014; Kopittke et al., 2017). To image subcellular structures, ultrathin sectioning is needed as well as high flux synchrotron X-ray sources (to focus a sufficient number of photons on nm-sized spots). Therefore, we cannot completely exclude that artifacts were created during sample preparation and/or analysis. However, differently from Fe and Cu, it seems that Zn distribution was not biased by these issues, as it can be argued from the clean distribution maps and the relative Zn concentrations in −Fe and +Fe samples, which confirm the ICP/MS data.

The interplay between Fe and Zn has been already observed in different plants, but the specific molecular mechanism underlining such interaction is less investigated (Leskova et al., 2017). Indeed, it is known that at the root level Zn competes with Fe for the uptake via IRT1 (Korshunova et al., 1999). Additionally, it has been observed that foliar Fe supply strongly suppressed the development of chlorosis upon excess Zn and reduced shoot Zn accumulation (Leskova et al., 2017). Such results were not observed when foliar Fe was supplied to irt1 mutant grown under Zn excess, suggesting a probably shootderived interaction between Fe and Zn, which might be relevant for the induction of Fe deficiency symptoms instead of the IRT1-mediated Zn uptake. Recently, it has been stated that the activation of Fe deficiency responses of plants growing under excess Zn were caused by an antagonist interaction between Zn and Fe (Leskova et al., 2017). In plants, Zn-dependent processes are located in all cellular compartments, including mitochondria (Heazlewood et al., 2004), and chloroplasts (Blaby-Haas and Merchant, 2013). To date, the requirement for Zn inside plant mitochondria has been linked to its role in important processes such as the degradation of mitochondrial presequences (Moberg et al., 2003; Tan et al., 2010). In this contest, several Zn-dependent metalloproteases have been identified in plant mitochondria (Moberg et al., 2003). Additionally, Zn is involved in the mitochondrial RNA editing and in in the small TIM (Translocase of the Inner Membrane) folding, and it is also requests by the COX4 subunit of cytochrome c oxidase in plant mitochondria (Vigani and Hanikenne, 2018 and references therein).

The role of Zn in chloroplasts has been overlooked because of its inability to perform redox chemistry. However, the interest in understanding Zn homeostasis in plant is gaining prominence (Sinclair and Kramer, 2012). Blaby-Haas and Merchant (2013) summarized the metal-dependent metabolic pathways in chloroplasts by considering different publicly available data. Chloroplast Zn-requiring enzymes are Superoxide dismutase (Cu/Zn), Alkenal/one oxidoreductase, Carboxy-transferase beta subunit of the acetyl-CoA carboxylase, proteases (FtsH proteins, peptidase M50 family protein, stromal processing peptidase), Zinc-finger proteins (involved in chloroplast and palisade cell development, in the ribosomal subunits, in the cyclic electron flow and chloroplast splicing factor). Therefore, several Znrequiring proteins are localized in chloroplasts. Considering that Fe deficiency negatively impacts on morphology and functions of both chloroplasts and mitochondria, it is plausible to think that an increased requirement of Zn might sustain ROS detoxification and proteins degradation processes. However, further studies are needed in order to understand the specific role of Zn under Fe deficiency in such cellular compartments.

#### CONCLUSION

fpls-09-01112 August 1, 2018 Time: 8:2 # 9

Understanding nutrient interactions is a challenging and relevant issue in the plant nutrition field as it might reveal useful insights to develop new plant genotypes with improved mineral use efficiency and for biofortification programs (Briat et al., 2015b).

Iron and Zn represent important targets for such programs and several evidences reveal a strong interaction among these nutrients in plants. An impaired balance between Fe and Zn is perceived as Fe deficiency by plant. In order to identify the cellular site where such interaction take place we determined the subcellular distribution of Zn in −Fe leaf cells and we observed that Zn strongly accumulates in the cells of the leaves of −Fe cucumber plants. By comparing cellular fractionation-ICP/MS and nanoscopic XRF imaging results, we showed that under Fe-deficiency Zn accumulates both in chloroplasts and in mitochondria, with a higher level in chloroplasts. Our findings highlight the importance of focusing future investigations on Fe and Zn interaction at the subcellular level, especially on chloroplasts and mitochondria.

However, in order to reduce the possibility of artifacts creation during sample preparation and/or analysis, two possible approaches should be considered in future nanoscopic XRF experiments: (i) frozen-hydrated samples can be prepared by a fast-freezing process, cut as thin sections by using a cryomicrotome, and kept at low temperature until the analysis. Cryopreparations are usually less disturbing and limit the redistribution of metals and changes in speciation. However, also beamlines equipped with cryostages or cryochambers are then needed to preserve the frozen-hydrated state during the analysis. (ii) In order to avoid chemical staining, fast preliminary maps of larger areas in the sample can be acquired to locate the areas of interest prior to longer scans of the subcellular structures. For this purpose multi-crystal detector arrays, like the Maia detector (Kirkham et al., 2010), could dramatically reduce the sample exposition time thus reducing the risks of beam damage.

#### REFERENCES


### AUTHOR CONTRIBUTIONS

GV and RT conceived and designed the study. GV performed the cellular fraction analysis. FF carried out TEM analysis and revised the manuscript. SB, BV, LV, GV, and RT carried out nano XRF imaging analysis. GV and RT drafted the article. FF, SB, BV, and LV revised it critically. All the authors approved the final version of the article.

### FUNDING

This research was supported by a FIRB-Programme Futuro in Ricerca grant from the Italian Ministry of University and Research – MIUR (Grant No. RBFR127WJ9). This work was also supported by the local research funds of Department of Life Science and Systems Biology, University of Turin.

#### ACKNOWLEDGMENTS

The authors thank Dr. Dario Maffi for the assistance in the ultrathin sections for both TEM and Nano X ray fluorescent imaging analysis and Giorgio Lucchini for assistance in the ICP-MS analysis.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018.01112/ full#supplementary-material

FIGURE S1 | Workflow of the isolation of chloroplast and mitochondria from leaf tissues. Tube after percoll gradient for chloroplast (left) and mitochondria (right) purification are reported. Cellular fractions collected and analyzed are highlighted in yellow.

FIGURE S2 | XRF sum spectra of the whole area imaged in Figure 5b (+Fe Full) and 5C (−Fe Full) in comparison with sum spectra of the cytosol extracted from pixel analysis of Figure 5b (+Fe Cytoplasm) and Figure 5c (−Fe Cytoplasm).

FIGURE S3 | Subcellular distribution maps (XRF K<sup>α</sup> lines) of S, Ca, Fe, Cu within leaf cells for +Fe (A,B) and −Fe (C,D) samples, as in Figure 5 of the main text. Darker pixels correspond to higher elemental concentrations. A scale bar based on counts is also reported.

plant's physiological status. Proc. Natl. Acad. Sci. U.S.A. 105, 12081–12086. doi: 10.1073/pnas.0804175105


involvement of PHOSPHATE STARVATION RESPONSE 1 (PHR1). Front. Plant Sci. 6:290. doi: 10.3389/fpls.2015.00290



**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 © 2018 Vigani, Bohic, Faoro, Vekemans, Vincze and Terzano. 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) and the copyright owner(s) 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 Shoot Fe Signaling Pathway Requiring the OPT3 Transporter Controls GSNO Reductase and Ethylene in Arabidopsis thaliana Roots

María J. García<sup>1</sup> , Francisco J. Corpas<sup>2</sup> , Carlos Lucena<sup>3</sup> , Esteban Alcántara<sup>3</sup> , Rafael Pérez-Vicente<sup>1</sup> , Ángel M. Zamarreño<sup>4</sup> , Eva Bacaicoa<sup>4</sup> , José M. García-Mina<sup>4</sup> , Petra Bauer<sup>5</sup> and Francisco J. Romera<sup>3</sup> \*

<sup>1</sup> Department of Botany, Ecology and Plant Physiology, Campus de Excelencia Internacional Agroalimentario, Universidad de Córdoba, Córdoba, Spain, <sup>2</sup> Department of Biochemistry, Cell and Molecular Biology of Plants, Estación Experimental del Zaidín, Spanish National Research Council, Granada, Spain, <sup>3</sup> Department of Agronomy, Campus de Excelencia Internacional Agroalimentario, Universidad de Córdoba, Córdoba, Spain, <sup>4</sup> Department of Environmental Biology, Faculty of Sciences, University of Navarra, Pamplona, Spain, <sup>5</sup> Institute of Botany, University of Düsseldorf, Düsseldorf, Germany

#### Edited by:

Hannetz Roschzttardtz, Pontificia Universidad Católica de Chile, Chile

#### Reviewed by:

Ping Lan, Institute of Soil Science (CAS), China Louis Grillet, Academia Sinica, Taiwan

> \*Correspondence: Francisco J. Romera ag1roruf@uco.es

#### Specialty section:

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

Received: 19 March 2018 Accepted: 23 August 2018 Published: 11 September 2018

#### Citation:

García MJ, Corpas FJ, Lucena C, Alcántara E, Pérez-Vicente R, Zamarreño ÁM, Bacaicoa E, García-Mina JM, Bauer P and Romera FJ (2018) A Shoot Fe Signaling Pathway Requiring the OPT3 Transporter Controls GSNO Reductase and Ethylene in Arabidopsis thaliana Roots. Front. Plant Sci. 9:1325. doi: 10.3389/fpls.2018.01325 Ethylene, nitric oxide (NO) and glutathione (GSH) increase in Fe-deficient roots of Strategy I species where they participate in the up-regulation of Fe acquisition genes. However, S-nitrosoglutathione (GSNO), derived from NO and GSH, decreases in Fe-deficient roots. GSNO content is regulated by the GSNO-degrading enzyme S-nitrosoglutathione reductase (GSNOR). On the other hand, there are several results showing that the regulation of Fe acquisition genes does not solely depend on hormones and signaling molecules (such as ethylene or NO), which would act as activators, but also on the internal Fe content of plants, which would act as a repressor. Moreover, different results suggest that total Fe in roots is not the repressor of Fe acquisition genes, but rather the repressor is a Fe signal that moves from shoots to roots through the phloem [hereafter named LOng Distance Iron Signal (LODIS)]. To look further in the possible interactions between LODIS, ethylene and GSNOR, we compared Arabidopsis WT Columbia and LODIS-deficient mutant opt3-2 plants subjected to different Fe treatments that alter LODIS content. The opt3-2 mutant is impaired in the loading of shoot Fe into the phloem and presents constitutive expression of Fe acquisition genes. In roots of both Columbia and opt3-2 plants we determined 1-aminocyclopropane-1-carboxylic acid (ACC, ethylene precursor), expression of ethylene synthesis and signaling genes, and GSNOR expression and activity. The results obtained showed that both 'ethylene' (ACC and the expression of ethylene synthesis and signaling genes) and 'GSNOR' (expression and activity) increased in Fe-deficient WT Columbia roots. Additionally, Fe-sufficient opt3-2 roots had higher 'ethylene' and 'GSNOR' than Fesufficient WT Columbia roots. The increase of both 'ethylene' and 'GSNOR' was not related to the total root Fe content but to the absence of a Fe shoot signal (LODIS), and was associated with the up-regulation of Fe acquisition genes. The possible relationship between GSNOR(GSNO) and ethylene is discussed.

Keywords: ethylene, glutathione (GSH), iron, long distance iron signal (LODIS), nitric oxide (NO), phloem, S-nitrosoglutathione (GSNO), S-nitrosoglutathione reductase (GSNOR)

#### García et al. Shoot Fe and Fe Responses

### INTRODUCTION

fpls-09-01325 September 8, 2018 Time: 14:26 # 2

Iron (Fe) is abundant in most soils, mainly as Fe3+, although its availability to plants is low, especially in calcareous soils (Römheld and Marschner, 1986). Based on the mechanisms used by plant roots to facilitate mobilization and uptake of Fe, plants are classified into Strategy I species and Strategy II species (Römheld and Marschner, 1986; Ivanov et al., 2012; Kobayashi and Nishizawa, 2012). Dicots, such as Arabidopsis and pea (Pisum sativum), are Strategy I species and reduce Fe3<sup>+</sup> to Fe2+, by means of a ferric reductase (encoded by FRO2 in Arabidopsis) at the root surface, prior to its subsequent uptake through a Fe2<sup>+</sup> transporter (encoded by IRT1 in Arabidopsis; Ivanov et al., 2012; Kobayashi and Nishizawa, 2012). When grown under Fe deficiency, Strategy I species develop several physiological and morphological responses, mainly in roots, which favor Fe acquisition and are generally known as Fe deficiency responses. Among the physiological responses are the up-regulation of the ferric reductase and the Fe2<sup>+</sup> transporter genes, as well as many other Fe-related genes (Ivanov et al., 2012; Kobayashi and Nishizawa, 2012; Brumbarova et al., 2015; Lucena et al., 2015). In the last years, several transcription factors (TFs) that participate in the activation of these genes have been found (Ivanov et al., 2012; Kobayashi and Nishizawa, 2012; Brumbarova et al., 2015; Zhang et al., 2015; Li et al., 2016; Liang et al., 2017). In Arabidopsis, the master regulator of most of the Ferelated genes is FIT (bHLH29), homolog of the tomato FER gene (Bauer et al., 2007 and references therein). The FIT regulatory network comprises other bHLH TFs of the Ib subgroup, such as bHLH38, bHLH39, bHLH100, and bHLH101. All of them have redundant functions and can interact with FIT to form heterodimers that activate the expression of the Fe acquisition genes FRO2 and IRT1 (Yuan et al., 2008; Wang et al., 2013; Brumbarova et al., 2015). FIT is induced in roots in response to Fe deficiency while the other Ib bHLH genes cited above are induced in both roots and leaves in response to Fe deficiency (Brumbarova et al., 2015 and references therein). Lately, it has been found that, under Fe-deficiency conditions, IVc subgroup bHLH TFs, like bHLH105 and bHLH115, activate the expression of FIT/bHLH38/39/100/101 (Zhang et al., 2015; Li et al., 2016; Liang et al., 2017).

The regulation of the above TFs and genes is not totally understood, but several works support a role for hormones and signaling molecules in the activation of FIT (and other Ib bHLH TFs) and, consequently, in the up-regulation of the ferric reductase and the Fe2<sup>+</sup> transporter genes. Among them are auxin, ethylene, salicylic acid, nitric oxide (NO), sucrose, and glutathione (GSH). All of them increase in Fe-deficient roots although their precise roles and interactions are not totally known (Zaharieva and Abadía, 2003; Zaharieva et al., 2004; Lucena et al., 2006, 2015; Graziano and Lamattina, 2007; Waters et al., 2007; Bacaicoa et al., 2009, 2011; Chen et al., 2010; García et al., 2010, 2011; Lingam et al., 2011; Meiser et al., 2011; Romera et al., 2011, 2017; Koen et al., 2012; Yang et al., 2014; Shanmugam et al., 2015; Lin et al., 2016; Shen et al., 2016; Li and Lan, 2017; Kailasam et al., 2018). There are also some hormones, such as cytokinins and jasmonates, that have been involved in the suppression of Fe deficiency responses (Séguéla et al., 2008; Maurer et al., 2011). Besides the activation of Fe-related genes, auxin, ethylene, and NO have also been involved in the regulation of morphological responses to Fe deficiency, such as the development of subapical root hairs (Romera et al., 2011, 2017; Lucena et al., 2015; Li and Lan, 2017). Additionally, ethylene is involved in restricting the suberization of the endodermis under Fe deficiency (Barberon et al., 2016), and NO and GSH are involved in improving Fe availability inside plants (Graziano et al., 2002; Ramírez et al., 2013).

In a previous work, García et al. (2011) showed that ethylene can induce NO accumulation in the subapical region of the roots, where most Fe responses are located. On the other hand, they found that NO can upregulate many ethylene synthesis and signaling genes in roots. This mutual and positive influence between ethylene and NO has also been described in the development of root hairs under Mg deficiency (Liu et al., 2017) and in other physiological processes (García et al., 2011 and references therein; Lin et al., 2013). Both ethylene and NO greatly activate the expression of Fe acquisition genes in plants grown with low levels of Fe (or without Fe), but have much less effect in plants grown with high levels of Fe (Lucena et al., 2006; Graziano and Lamattina, 2007; Chen et al., 2010; García et al., 2011). Similar results have also been found when other activators of Fe acquisition genes, like auxin (Chen et al., 2010) or sucrose (Lin et al., 2016), have been applied to plants grown with different Fe levels. All these results suggest that the up-regulation of Fe acquisition genes does not solely depend on hormones and signaling molecules (such as ethylene, auxin, sucrose, or NO), that would act as activators, but also on the internal Fe content of plants, that would act as a repressor (Lucena et al., 2006; García et al., 2011, 2013; Romera et al., 2011). However, different results suggest that total Fe in roots is not the repressor of Fe acquisition genes. As examples, Arabidopsis opt3, frd3, nas4x-1 and bts3 mutants, Arabidopsis AtHSCB overexpressing lines, tomato chln mutant, and pea dgl and brz mutants, all of them present constitutive activation of Fe acquisition genes when grown under Fe sufficiency despite the high accumulation of Fe inside their roots (García et al., 2013 and references therein; Romera et al., 2015; Leaden et al., 2016; Hindt et al., 2017; Khan et al., 2018). Several of the above genotypes are affected, either directly or indirectly, in the transport of Fe through the phloem, which suggests that phloem Fe could be a key factor in the repression of Fe acquisition genes (García et al., 2013; Khan et al., 2018). In any case, there is no doubt that shoots play a very important role in the regulation of Fe acquisition genes in roots (García et al., 2013 and references therein; Mendoza-Cózatl et al., 2014; Zhai et al., 2014; Gayomba et al., 2015). In accordance with this, Khan et al. (2018) have recently demonstrated that the leaf vasculature respond to Fe deficiency considerably faster than roots, and Kumar et al. (2017) have found, by using reciprocal shoot/root grafts between Arabidopsis ysl1ysl3 double mutant and WT plants, that the ability to express the Fe deficiency responses depends on the genotype of the shoot.

OPT3 (OligoPeptide Transporter3) is the main transporter implicated in the loading of shoot Fe into the phloem (Mendoza-Cózatl et al., 2014; Zhai et al., 2014; Kumar et al., 2017;

Khan et al., 2018). In supporting this view, it should be noted that the foliar application of Fe did not inhibit the expression of Fe acquisition genes in Fe-deficient Arabidopsis opt3-2 roots while it did in Fe-deficient Arabidopsis WT roots (García et al., 2013). Furthermore, phloem sap measurements in opt3 mutants show a 50% reduction in Fe content (Khan et al., 2018). OPT3 is expressed in the plasma membrane of phloem cells, mainly in shoots, and its shoot-specific expression is sufficient to complement the Fe-deficiency response in opt3-2 roots (Mendoza-Cózatl et al., 2014; Zhai et al., 2014; Gayomba et al., 2015). These results suggest that OPT3 plays an important role in the iron-signaling network between leaves and roots (García et al., 2013; Mendoza-Cózatl et al., 2014; Zhai et al., 2014; Gayomba et al., 2015). The precise substrate of the OPT3 transporter is not yet clear. Zhai et al. (2014) found that OPT3 can transport Fe2<sup>+</sup> ions when expressed in Xenopus oocytes. However, while Wintz et al. (2003) found that OPT3 was able to rescue the fet3fet4 strain of yeast, impaired in Fe uptake, Mendoza-Cózatl et al. (2014) found that not. Even if OPT3 transports Fe ions, these ions should be chelated in the phloem sap due to the poor solubility of Fe at the alkaline pH of this fluid (Gutiérrez-Carbonell et al., 2015). Although OPT3 does not mediate GSH transport in S. cerevisiae (Zhai et al., 2014), other OPT transporters do (Lubkowitz, 2011 and references therein; Zhang et al., 2016 and references therein). Consequently, some proposed chelating agents of Fe are GSH-derived compounds, like S-nitrosoglutathione (GSNO) or nitrosyl-iron complexes (NICs, formed from the interaction of Fe, NO and thiols); also peptides and proteins (Krüger et al., 2002; Ramírez et al., 2011; Rahmanto et al., 2012; Darbani et al., 2013; García et al., 2013; Buet and Simontacchi, 2015). Based on these findings, we suggest the name LODIS (LOng Distance Iron Signal) for the shoot Fe signal related to OPT3 and moving through the phloem, which causes the repression of Fe acquisition genes in roots. Although the LODIS nature is still unknown, it is possible to investigate its consequences by comparing the LODIS-deficient mutant opt3-2 and the WT cultivar Columbia, and by comparing plants treated or not with foliar application of Fe.

Another important question, which remains unsolved, is how the Fe status is perceived by the roots. At this point, it has been proposed that Fe (probably, LODIS or a LODIS-derived signal) could be sensed by the BRUTUS (BTS) protein in Strategy I plants (Kobayashi and Nishizawa, 2014). BTS is a homolog of HRZs, known Fe sensors in rice (Kobayashi and Nishizawa, 2014). The current data suggests that BTS acts as a negative regulator of Fe responses (Zhang et al., 2015; Hindt et al., 2017). This would explain why Fe-sufficient bts-2 and bts-3 roots present higher expression of Fe acquisition genes than Fe-sufficient WT roots (Zhang et al., 2015; Hindt et al., 2017).

To integrate both positive and negative signals in the regulation of Fe acquisition genes in roots, Lucena et al. (2006) proposed a model that implicated both LODIS and ethylene in such a regulation. Subsequently, this model has been extended to other positive signals besides ethylene, like NO and auxin (García et al., 2011; Romera et al., 2011, 2017). According to this model, auxin/ethylene/NO would act as activators of the expression of Fe acquisition genes in roots, while LODIS would act to repress their expression (García et al., 2011; Romera et al., 2011, 2017). This model does not exclude the role of other positive and negative signals (most of them reviewed by García et al., 2015; Romera et al., 2017). The way LODIS represses the expression of Fe acquisition genes is not totally known, but several results suggest that it could negatively affect ethylene action. García et al. (2013) found that ACC up-regulated the expression of Fe acquisition genes when applied to roots of Fe-deficient plants but not when applied simultaneously with foliar Fe. This result suggests that LODIS could block ethylene action, but does not preclude an additional negative effect of LODIS on ethylene production.

One of the objectives of this work was to study the role of LODIS on ethylene synthesis in roots. For this, we determined 1 aminocyclopropane-1-carboxylic acid (ACC) concentration and the expression of the key ethylene synthesis genes SAM1, ACS6, and ACO2 in roots of Arabidopsis WT Columbia plants subjected to different Fe treatments that alter LODIS content. It should be noted that ACC content and the expression of the above ethylene synthesis genes increase under Fe deficiency (García et al., 2010; Gutiérrez-Carbonell et al., 2015; Lucena et al., 2015 and references therein; Ye et al., 2015; Romera et al., 2017 and references therein). In a parallel study, similar determinations were done in the Arabidopsis LODIS-deficient mutant opt3- 2, and in other Arabidopsis mutants that behave like LODISdeficient, such as frd3-3 and nas4x-1. All of these mutants have constitutive activation of Fe acquisition genes when grown under Fe sufficient conditions (see section "Materials and Methods"; Rogers and Guerinot, 2002; Klatte et al., 2009; García et al., 2013). A second objective of this work was to study the possible relationship of S-nitrosoglutathione reductase (GSNOR) with Fe deficiency and, more specifically, with LODIS. Both GSH and NO increase in roots under Fe deficiency and have been implicated in the activation of Fe responses (Zaharieva and Abadía, 2003; Zaharieva et al., 2004; Graziano and Lamattina, 2007; Bacaicoa et al., 2009; Chen et al., 2010; García et al., 2011; Koen et al., 2012; Shanmugam et al., 2015; Kailasam et al., 2018). However, S-nitrosoglutathione (GSNO), the most abundant lowmolecular-weight S-nitrosothiol in plants (Liu et al., 2018), and which is derived from GSH and NO (Corpas et al., 2013), decreases under Fe deficiency in Arabidopsis roots (Shanmugam et al., 2015; Kailasam et al., 2018). Since GSNOR regulates GSNO content by decomposing it to oxidized glutathione (GSSG) and H3N (Leterrier et al., 2011; Corpas et al., 2013), we wanted to know whether GSNOR increases in roots under Fe deficiency. For this, we determined GSNOR expression and activity under Fe deficiency and also under different Fe treatments and genetic backgrounds that could influence LODIS levels.

The results show that ethylene (ACC and the expression of ethylene synthesis genes) and GSNOR (expression and activity) increase under conditions that restrict LODIS accumulation in roots, such as Fe deficiency in WT or Fe sufficiency in the LODISdeficient mutant opt3-2, while they decrease under conditions that favor its accumulation, such as Fe sufficiency or foliar application of Fe in WT. The up-regulation of GSNOR under Fe deficiency in the WT suggests that lower GSNO levels in roots could be a prerequisite for the up-regulation of ethylene synthesis and, consequently, for that of Fe acquisition genes.

#### MATERIALS AND METHODS

fpls-09-01325 September 8, 2018 Time: 14:26 # 4

### Plant Materials, Growth Conditions, and Treatments

To study the role of LODIS on ethylene and GSNOR in roots, wild-type Arabidopsis [Arabidopsis thaliana (L.) Heynh ecotype Columbia] and pea [Pisum sativum L. cv. Sparkle] plants were used. Additionally, we used some mutants that show constitutive up-regulation of Fe acquisition genes when grown under Fesufficient conditions and that are, or behave like, LODIS-deficient (García et al., 2013 and references therein). These mutants were Arabidopsis opt3-2, frd3-3 and nas4x-1, and pea dgl (Sparkle [dgl,dgl]). OPT3 is expressed mainly in shoots, where it could act as a transporter involved in the loading of Fe2<sup>+</sup> ions into the phloem (Stacey et al., 2008; Zhai et al., 2014; see **Figure 12**). The Arabidopsis frd3-3 mutant is impaired in xylem Fe transport (Rogers and Guerinot, 2002; Roschzttardtz et al., 2011; Gayomba et al., 2015) but, as a consequence, it is also defective in the transport of Fe from shoots to roots through the phloem (it behaves like LODIS-deficient), since less Fe gets into leaves to enter the phloem (Lucena et al., 2006; see **Figure 12**). The Arabidopsis nas4x-1 mutant is defective in the synthesis of NA, which is involved in Fe loading and unloading of the phloem (Klatte et al., 2009; Schuler et al., 2012). Both frd3-3 and nas4x-1 mutants show chlorosis when grown under Fe-sufficient conditions (Lucena et al., 2006; Klatte et al., 2009). In relation to the pea dgl mutant, the specific gene related to this mutation has not been identified yet. However, several results suggest that the dgl mutant phenotype may be related to defects in phloem Fe transport since it behaves like a LODIS-deficient mutant (García et al., 2013; Romera et al., 2015). Although the dgl mutant originated from the DGV (Dippes Gelbe Viktoria) cultivar, we used a near isogenic line homozygous for the dgl mutation that was introgressed into the Sparkle cultivar and described here as Sparkle [dgl, dgl] (Marentes and Grusak, 1998).

Arabidopsis and pea plants were grown on aerated nutrient solution in a growth chamber, as previously described (Lucena et al., 2006, 2007). When appropriate, plants were transferred to the different treatments. The treatments imposed were: +**Fe**: nutrient solution with 40 µM Fe-EDDHA, except for pea WT Sparkle, that was 20 µM Fe-EDDHA, and for pea dgl, that was 3 µM Fe-EDDHA (this mutant grows adequately with 2–3 µM Fe-EDDHA concentration but presents symptoms of Fe toxicity when grown with higher levels of Fe; Romera et al., 2015); −**Fe**: nutrient solution without Fe during different times (from 6 to 72 h, depending on experiments); −**Fe**+**foliarFe**: −Fe treatment during 2 or 3 days and FeSO<sup>4</sup> application to leaves during the last 24 h; −**P**: nutrient solution without P during 2 days; −**S**: nutrient solution without S during 2 days. It should be noted that, in our experimental conditions, the expression of Fe acquisition genes and the ferric reductase activity was weak after 24 h of Fe deficiency and reached their maximum after 48 h of Fe deficiency. FeSO<sup>4</sup> was dissolved in deionized water (1.8 mM) and Tween 20 was added as surfactant. Leaves were sprayed once until total moistening. After treatments, root ferric reductase activity was determined as described previously (Lucena et al., 2006). Finally, the roots were collected and kept at −80◦C for subsequent analysis of ACC and mRNA levels. In some experiments, GSNOR activity, GSNO and GSH were determined in fresh roots.

#### ACC Determination

The extraction, purification and quantification of ACC was carried out using the method described by Mora et al. (2012). Briefly, ACC of roots was extracted with 20 µl of d4ACC [3 µg/mL in acetonitrile/acetic acid 0.2% (90/10)] and 3 ml of MeOH/H2O/HCOOH (15/4/1, v/v/v) at −20◦C. Purification was carried out using a Strata C18-E cartridge (Ref 8B-S001-FBJ, Phenomenex, Torrance, CA, United States) preconditioned with 4 ml of methanol and 2 ml of MeOH/H2O/HCOOH (15/4/1, v/v/v). Finally, the eluted fraction was centrifuged (10,000 rpm, 8 min) and injected in the LC/MS/MS systems, and ACC was quantified by HPLC linked to a 3200 QTRAP LC/MS/MS system (Applied Biosystems/MDS Sciex, Ontario, Canada), equipped with a turbo ion spray interface.

#### GSNOR Activity Determination

GSNOR activity was assayed spectrophotometrically at 25◦C by monitoring the oxidation of NADH at 340 nm as described by Sakamoto et al. (2002). The root extracts were incubated in an assay mixture containing 20 mM TRIS-HCl (pH 8.0), 0.2 mM NADH, and 0.5 mM EDTA, and the reaction was started by adding GSNO (Calbiochem) to the mixture at a final concentration of 400 µM. The activity was expressed as nmol NADH consumed min−<sup>1</sup> mg−<sup>1</sup> protein (ε<sup>340</sup> = 6.22 mM−<sup>1</sup> cm−<sup>1</sup> ).

#### Quantification of GSNO and GSH by Liquid Chromatography Electrospray Ionization Mass Spectrometry (LC–ES/MS)

Arabidopsis samples (300 mg) were ground using a mortar and pestle in the presence of 1 ml of 0.1 M HCl. Homogenates were centrifuged at 15,000 g for 20 min at 4◦C. The supernatants were collected and filtered through 0.22-mm polyvinylidene fluoride filters and immediately analyzed. All procedures were carried out at 4◦C and were protected from light to avoid potential degradation of the analytes (GSNO and GSH). The LC–ES/MS system consisted of a Waters Allience 2695 HPLC system connected to a Micromass Quattro micro API triple quadrupole mass spectrometer, both obtained from the Waters Corporation. HPLC was carried out using an Atlantis <sup>R</sup> T3 3 µm 2.1 mm × 100 mm Column obtained from the Waters Corporation. The Micromass Quattro Micro API mass spectrometer was used in positive electrospray ionization mode for detection and quantification of GSNO and GSH (Airaki et al., 2011; Leterrier et al., 2012).

#### qRT-PCR Analysis

Roots were ground to a fine powder with a mortar and pestle in liquid nitrogen. Total RNA was extracted using the Tri Reagent solution (Molecular Research Center, Inc., Cincinnati, OH, United States) according to the manufacturer's

instructions. M-MLV reverse transcriptase (Promega, Madison, WI, United States) was used to generate cDNA from 3 µg of DNase-treated root RNA as the template and random hexamers as the primers.

The study of gene expression by qRT-PCR was performed by using a qRT-PCR Bio-Rad CFX connect thermal cycler and the following amplification profile: initial denaturation and polymerase activation (95◦C for 3 min), amplification and quantification repeated 40 times (90◦C for 10 s, 57◦C for 15 s and 72◦C for 30 s), and a final melting curve stage of 65 to 95◦C with increment of 0.5◦C for 5 s, to ensure the absence of primer dimer or non-specific amplification products. PCR reactions were set up in 20 µl of SYBR Green Bio-RAD PCR Master Mix, following the manufacturer's instructions. Controls containing water instead of cDNA were included to check for contamination in the reaction components. Gene-specific primers (**Table 1**) were designed by using the Primer-BLAST software from the NCBI site. Standard dilution curves were performed for each primer pair to confirm appropriate efficiency of amplification (E = 100 ± 10%). Constitutively expressed SAND1 and YLS8 genes, which do not respond to changes in the Fe conditions (Han et al., 2013), were used as reference genes to normalize qRT-PCR results. The relative expression levels were calculated from the threshold cycles (Ct) values and the primer efficiencies by the Pfaffl method (Pfaffl, 2001). Each PCR analysis was conducted on three biological replicates and each PCR reaction repeated twice.

#### Statistical Analysis

All experiments were repeated at least twice and representative results are presented. The values of qRT-PCR represent the mean ± SE of three independent biological replicates. The values of other determinations (ACC, GSNO, GSH, ferric reductase



FIGURE 1 | Effect of Fe deficiency and foliar application of Fe on the expression of (A) Fe acquisition genes FRO2, IRT1, and FIT; and (B) ethylene synthesis genes SAM1, ACS6 and ACO2, in roots of Arabidopsis WT Columbia plants. Plants were grown in nutrient solution with Fe (+Fe). Some of them were transferred to nutrient solution without Fe 3 days before harvest (–Fe). Half of the –Fe plants were sprayed with Fe 24 h before harvest (–Fe+fol Fe). Relative expression was calculated in relation to +Fe. Data represent the mean ± SE of three independent biological replicates. Within each gene, bars with different letters indicate significant differences (P < 0.05).

activity, GSNOR activity) represent the mean ± SE of six replicates. Within each gene or genotype, different letters indicate significant differences (P < 0.05) among treatments using oneway analysis of variance (ANOVA) followed by a Duncan's multiple range test (i.e., **Figures 1**, **2**, **6**). Dunnett's test was also used when one or several mutants were compared with the WT for the +Fe treatment (i.e., **Figures 2**, **3**) and when different treatments were compared with a control (i.e., **Figures 3**, **9**). In this latter case, ∗∗ indicate significant differences (P < 0.05).

### RESULTS

#### Effect of LODIS on Ethylene Synthesis in Roots

To observe whether LODIS affects ethylene synthesis in roots, we determined ACC (ethylene precursor) concentration and the expression of several genes implicated in ethylene synthesis (SAM1, ACS6, and ACO2; García et al., 2010) in roots of Arabidopsis WT Columbia plants grown under different Fe treatments that could affect LODIS accumulation: with Fe (LODIS sufficient), without Fe (LODIS deficient), and without Fe but with Fe sprayed on leaves (LODIS sufficient). The

expression of all the ethylene synthesis genes studied and ACC production greatly increased (sometimes, more than 2-fold) in Fe-deficient Columbia roots, in parallel to the expression of the key Fe acquisition genes FRO2, IRT1, and FIT (**Figures 1**, **2**). On the other hand, the foliar application of Fe to Fedeficient Columbia plants greatly repressed the expression of Fe acquisition (**Figure 1A**) and ethylene synthesis genes (**Figure 1B**), and ACC production (**Figure 2**), which suggests that LODIS inhibits ethylene synthesis. This inhibitory effect of LODIS was further supported by the fact that the foliar application of Fe did not significantly inhibit ACC production in the Arabidopsis LODIS-deficient mutant opt3-2 (**Figure 2**; García et al., 2013; Zhai et al., 2014). This mutant also showed higher ACC concentration in Fe-sufficient roots compared to the WT Columbia (**Figure 2**), which suggests that total Fe in roots does not control ethylene synthesis. Similar higher ACC levels were found in Fe-sufficient roots of other Arabidopsis mutants that behave like LODIS-deficient (frd3-3 and nas4x-1) and that present constitutive activation of Fe acquisition genes under Fesufficiency (**Figure 3A**; García et al., 2013; see section "Materials and Methods").

It should be noted that, although the frd3-3 and opt3-2 mutants usually accumulate more metals than the WT cultivar (Rogers and Guerinot, 2002; Stacey et al., 2008), neither of them showed any symptom of metal toxicity in our experimental conditions (**Figure 3B**). The frd3-3 and nas4x-1 mutants showed chlorosis, as expected (**Figure 3B**; see section "Materials and Methods").

The higher ACC content in Fe-sufficient opt3-2 roots than in Columbia roots (**Figure 2**) was correlated with higher expression of the ethylene synthesis genes SAM1, ACS6 and ACO2, and of the ethylene signaling genes EIN2, EIN3 and EIL1, in the mutant (from 5- to 30-fold; **Figure 4**). Collectively, these results suggest that ethylene (synthesis and signaling) drastically increases under low LODIS accumulation in roots (i.e., Fe deficiency in WT or Fe sufficiency in the LODIS-deficient mutant opt3-2).

appearance (B) of Arabidopsis WT Columbia and opt3-2, frd3-3, and nas4x-1 mutant plants. Note the chlorosis in the frd3-3 and nas4x-1 mutants. Plants were grown in nutrient solution with Fe. Data represent the mean ± SE of six replicates. Significant differences from Col are indicated: ∗∗P < 0.05.

### Effect of LODIS on GSNOR Expression and Activity in Roots

To study the possible relationship of GSNOR with Fe deficiency, we analyzed GSNOR expression and activity under different Fe treatments and genetic backgrounds that could influence LODIS levels. The results showed that GSNOR1 expression increased (more than 10-fold) under Fe deficiency in roots of the WT cultivar Columbia (**Figure 5A**). This up-regulation of GSNOR1 occurred very quickly, after few hours of Fe depletion, and slightly before the up-regulation of the main regulator of the Fe deficiency responses, FIT (**Figure 5B**), other Fe acquisition genes (bHLH38, IRT1, FRO2; **Figures 5C–E**), and the enhancement of ferric reductase activity (**Figure 5F**). In agreement with the GSNOR1 up-regulation, the GSNOR activity significantly increased (approximately 50%) in Fe-deficient Columbia roots (**Figure 6**). This activity was restored to normal level upon foliar application of Fe (**Figure 6**), which suggests that it could be inhibited by the accumulation of LODIS in roots. To further confirm this possibility, we compared GSNOR1 expression in Fe-sufficient WT and opt3-2 roots. As shown in **Figure 7**, GSNOR1 expression was much higher (more than 50-fold) in Fe-sufficient opt3-2 roots (LODIS-deficient)

than in Fe-sufficient Columbia roots, which suggests that its expression depends on LODIS and not on total root Fe content. The up-regulation of GSNOR1 expression also occurred under other nutrient deficiencies, such as P deficiency and S deficiency (**Figure 8**). Furthermore, GSNOR1 expression was up-regulated by ACC treatment in Fe-sufficient Columbia roots (**Figure 9**).

In addition to GSNOR, we also determined GSNO concentration in roots of Arabidopsis WT Columbia and opt3-2 plants subjected to different Fe treatments. The results

mean ± SE of three independent biological replicates. Relative expression was calculated in relation to +Fe. For ferric reductase activity, data represent the

showed a decrease of approximately 30% of GSNO content in Fe-deficient WT roots and, by contrast, a great increase upon foliar application of Fe (**Figure 10**). GSNO also decreased in Fe-deficient opt3-2 roots but was less affected by the foliar application of Fe (**Figure 10**). The lower GSNO content in Fe-deficient roots could be explained by the higher GSNOR expression and activity in these roots (**Figures 5**, **6**). In supporting this view, Fe-sufficient opt3-2 roots, which had higher GSNOR expression than Fe-sufficient WT roots (**Figure 7**), had lower GSNO content (approximately 50%; **Figure 10**). Similar results were also found with the pea LODIS-deficient mutant dgl and its WT cultivar Sparkle. As shown in **Figure 11**, GSNO content also decreased in Fe-deficient Sparkle roots (approximately 25%) and, similarly to opt3-2, GSNO content was lower in Fe-sufficient dgl roots (approximately 40%) than in Fe-sufficient Sparkle roots.

Collectively, all these results suggest that GSNOR (expression and activity) increases, and consequently GSNO decreases, under low LODIS accumulation in roots (i.e., Fe deficiency in WT or Fe sufficiency in LODIS-deficient mutants).

#### Effect of Fe Deficiency on GSH Content in Roots and Leaves of WT and opt3-2 Mutant Plants

In our experimental conditions, GSH greatly increased in Fedeficient WT roots but hardly in Fe-deficient opt3-2 roots (**Table 2**). In leaves, the most noticeable result was the higher GSH content in the opt3-2 mutant in relation to Columbia under Fe sufficiency (**Table 2**).

### DISCUSSION

Ethylene increases in Fe-deficient roots, where it activates the expression of key Fe acquisition genes (reviewed in Lucena et al., 2015; Li and Lan, 2017; Romera et al., 2017). A signal

mean ± SE of six replicates.

FIGURE 6 | Effect of Fe deficiency and foliar application of Fe on GSNOR activity in roots of Arabidopsis WT Columbia plants. Plants were grown in nutrient solution with Fe (+Fe) and some of them were transferred to nutrient solution without Fe 2 days before harvest (–Fe). Half of the –Fe plants were sprayed with Fe 24 h before harvest (–Fe+fol Fe). Data represent the mean ± SE of six replicates. Bars with different letters indicate significant differences (P < 0.05).

related to phloem Fe, LODIS, blocks this activating effect (García et al., 2013). However, it is not known whether LODIS could affect ethylene synthesis in roots. Results presented in this work (**Figures 1**, **2**, **3A**, **4**) support that LODIS does affect ethylene synthesis. The foliar application of Fe greatly decreased the expression of Fe acquisition and ethylene synthesis genes (**Figure 1**), and ACC production (**Figure 2**), in Fe-deficient WT Columbia roots. However, in Fe-deficient roots of the LODISdeficient mutant opt3-2 (García et al., 2013; Mendoza-Cózatl et al., 2014; Zhai et al., 2014), foliar application of Fe had almost no effect on ACC content of roots (**Figure 2**). Moreover, the expression of several ethylene synthesis and signaling genes was much higher in Fe-sufficient opt3-2 roots than in Fe-sufficient Columbia roots (**Figure 4**). In the same way, Fe-sufficient roots of opt3-2 and other Arabidopsis mutants that behave as LODISdeficient, such as frd3-3 and nas4x-1 (see section "Materials and Methods"), had greater ACC content than Fe-sufficient Columbia roots (**Figures 2**, **3A**). All these results suggest that

FIGURE 8 | Effect of P or S deficiency on GSNOR1 expression in roots of Arabidopsis WT Columbia plants. Plants were grown in complete nutrient solution (Control). Some of them were transferred to nutrient solution without P (–P) or without S (–S) during the last 2 days. Relative expression was calculated in relation to Control. Data represent the mean ± SE of three independent biological replicates. Significant differences from the Control treatment are indicated: ∗∗P < 0.05.

ethylene synthesis (and signaling) in roots does not depend on total root Fe content (these mutants accumulate high levels of Fe inside roots under Fe sufficiency; García et al., 2013 and references therein) but on LODIS levels, that depend on the OPT3 transporter.

Since heavy metals can enhance ethylene production and the expression of ethylene signaling genes (Keunen et al., 2016), it can be argued that the higher ACC content and the higher expression of ethylene-related genes in Fe-sufficient opt3-2 roots (and in Fe-sufficient frd3-3 and nas4x-1 roots) could be caused by the high accumulation of metals described in some of these mutants. Against this argument, we have to say that, in our experiments, plants were grown with low levels of metals and they did not present leaf toxicity symptoms (**Figure 3B**). Moreover, after the work of Vert et al. (2002) it is known that the accumulation of metals in mutants that present constitutive activation of IRT1 expression, such as opt3-2 (Stacey et al., 2008) and frd3 (Rogers and Guerinot, 2002), is mainly due to the broad substrate range

FIGURE 10 | Effect of Fe deficiency and foliar application of Fe on GSNO content in roots of Arabidopsis WT Columbia and opt3-2 mutant plants. Plants were grown in nutrient solution with Fe (+Fe). Some of them were transferred to nutrient solution without Fe 2 days before harvest (–Fe). Half of the –Fe plants were sprayed with Fe 24 h before harvest (–Fe+fol Fe). Data represent the mean ± SE of six replicates. Within each genotype, bars with different letters indicate significant differences (P < 0.05). Significant difference between the +Fe treatment from opt3-2 and Col is also indicated: ∗∗P < 0.05.

of this transporter. IRT1 can transport several divalent metals,

besides Fe2+, such as Mn2+, Zn2+, or Cd2<sup>+</sup> (Korshunova et al.,

(P < 0.05). Significant difference between the +Fe treatment from dgl and

1999; Lucena et al., 2006). Since IRT1 expression is activated by ethylene (García et al., 2010; Lingam et al., 2011; Blum et al., 2014; Marín-de la Rosa et al., 2014), the most probable sequence is: (1) enhanced ethylene production → (2) activation of IRT1 expression → (3) accumulation of metals, and not: (1) accumulation of metals → (2) enhanced ethylene production → (3) activation of IRT1 expression. No accumulation of metals would occur without IRT1 activity. In supporting this view, it should be noted that the inhibition of ethylene synthesis with cobalt, a potent ethylene inhibitor, drastically inhibited IRT1 expression and the accumulation of metals in tomato plants (Lucena et al., 2006) and in pea brz mutant plants (also named E107, has constitutive activation of Fe responses; Romera et al., 1996, 2015). Taken together, all these results suggest that ethylene is necessary for IRT1 up-regulation and, consequently, for the acquisition and accumulation of metals. This does not discard that, if plants accumulate high levels of metals, they could cause a further increase of ethylene production, but as a later side effect.

Similarly to ethylene, GSH (**Table 2**) and NO also increase in Fe-deficient WT roots, where they have been implicated in the activation of Fe deficiency responses (Zaharieva and Abadía, 2003; Zaharieva et al., 2004; Graziano and Lamattina, 2007; Bacaicoa et al., 2009; Chen et al., 2010; García et al., 2011; Koen et al., 2012; Shanmugam et al., 2015; Kailasam et al., 2018). However, results in this work show that GSNO content in WT roots, in contrast to its precursors GSH and NO, decreases upon Fe deficiency (**Figures 10**, **11**), which agrees with previous results (Shanmugam et al., 2015; Kailasam et al., 2018). Since GSNOR reduces GSNO to GSSG and other byproducts, like hydroxylamine and H3N (Leterrier et al., 2011; Corpas et al., 2013; Kubienová et al., 2014), we wanted to know whether GSNOR increases in Fe deficient roots. In fact, GSNOR1 expression and activity in roots increased early after Fe deficiency and slightly before the up-regulation of Fe acquisition genes (**Figures 5**, **6**). This suggests that GSNOR could be responsible for the lower GSNO contents found in Fe-deficient WT roots (**Figures 10**, **11**; Shanmugam et al., 2015; Kailasam et al., 2018). GSNOR1 is highly responsive to the Fe status of the plant but, as for ethylene (see above), the results indicate that its up-regulation does not depend on the total root Fe content but on LODIS levels. First, the foliar application of Fe greatly increased GSNO content in Fe-deficient Columbia roots (probably by decreasing GSNOR activity, as seen in **Figure 6**) but hardly in Fedeficient roots of the LODIS-deficient mutant opt3-2 (**Figure 10**), which is impaired in the movement of Fe from shoots to

TABLE 2 | Effect of Fe deficiency on GSH (glutathione) concentration in roots and leaves of Arabidopsis WT Columbia and opt3-2 mutant plants.


Plants were grown in nutrient solution with Fe (+Fe) and half of them transferred to nutrient solution without Fe (−Fe) during the last 2 days. Data represent the mean ± SE of six replicates. Significant differences from the +Fe treatment are indicated: ∗∗P < 0.01.

Sparkle is also indicated: ∗∗P < 0.05.

roots (García et al., 2013; Mendoza-Cózatl et al., 2014; Zhai et al., 2014). Second, Fe-sufficient opt3-2 roots presented much higher GSNOR1 expression (and less GSNO content) than Fesufficient Columbia roots (**Figures 7**, **10**). Similarly, a lower GSNO content was found in Fe-sufficient roots of the pea LODIS-deficient mutant dgl than in Fe-sufficient roots of its WT Sparkle (**Figure 11**). All these results clearly suggest that GSNOR expression (and consequently GSNO content)

does not depend on total root Fe content but on LODIS levels.

The results described above clearly show that the higher NO accumulation found in Fe-deficient WT roots (Graziano and Lamattina, 2007; Chen et al., 2010; García et al., 2011) is associated with lower GSNO contents (**Figures 10**, **11**). Similar results have been found by others in roots of Arabidopsis plants subjected to Fe deficiency (Shanmugam et al., 2015; Kailasam et al., 2018), to arsenic stress (Leterrier et al., 2012) or to Cd stress (Liu et al., 2018). In the two latter cases, the lower GSNO levels have been associated with a higher GSNOR activity (Leterrier et al., 2012; Liu et al., 2018), which also occurs under other abiotic stress conditions (Corpas et al., 2008; Airaki et al., 2012; Kubienová et al., 2014; Cheng et al., 2015). In relation to Cd stress, it is important to mention that Cd can cause Fe deficiency (Alcántara et al., 1994 and references therein) and induce Fe deficiency responses (Yoshihara et al., 2006). In our work, besides Fe deficiency, GSNOR1 expression also increased under P or S deficiency (**Figure 8**), which indicates that GSNOR could be implicated in the responses to different nutritional deficiencies. The higher NO accumulation associated with a lower GSNO content in Fe-deficient roots suggests that NO and GSNO do not have the same exact roles, probably because they have different lifetime and location. NO lifetime is relatively short (less than 10 s) while GSNO is a relatively stable store for NO and might constitute a vehicle for its long-distance transport (Corpas et al., 2011; Malik et al., 2011), mediating a signaling function throughout a process of S-transnitrosation between GSNO and proteins (Corpas et al., 2013). In relation to their location, both NO and GSNO can present different cellular and intracellular location (Malik et al., 2011; Shi et al., 2015). NO production is associated with phloem cells (Corpas et al., 2004; Gaupels et al., 2008) and, since the GSNOR protein is also located in the phloem (Rustérucci et al., 2007), it is possible that GSNO could be a vehicle to transmit the NOsignal out of the phloem. In agreement with the distinct role of NO and GSNO, Yun et al. (2016) have found that both of them act differently in the responses associated with plant immunity. In relation to Fe deficiency responses in dicot plants, Kailasam et al. (2018) have recently shown, by using chemical screening, that GSNO probably affects the regulation of Fe acquisition genes through FIT while NO could affect them through other TFs, such as bHLH38 and bHLH39. It should be noted that the SKB1 protein, implicated in bHLH38 and bHLH39 expression (Fan et al., 2014), can be nitrosylated (Hu et al., 2015). In this way, it is possible that GSNO, through the nitrosylation/de-nitrosylation of SKB1, could affect the expression of these Ib bHLH TFs (Fan et al., 2014; Lucena et al., 2015). In supporting this view, bHLH38, bHLH39, and bHLH100 are up-regulated in leaves of the Arabidopsis gsnor1-3 mutant, which has altered GSNO content (Xu et al., 2013).

Results in this work, along with previous results (García et al., 2013), suggest that LODIS can inhibit both ethylene synthesis and signaling, and GSNOR expression and activity in roots (see Working Model in **Figure 12**). The inhibition of GSNOR would increase GSNO content in roots, which agrees with the higher GSNO contents found in Fe-sufficient WT roots or in roots of Fe-deficient WT plants treated with foliar Fe (**Figures 10**, **11**; Shanmugam et al., 2015; Kailasam et al., 2018). After all these results, several questions arise: what is the nature of LODIS?; is there any relationship between GSNO(NO), GSH and ethylene?; is there any relationship between GSNO(NO)/ethylene and the putative Fe sensor BTS? In relation to the nature of LODIS, it is unknown at present. Although Zhai et al. (2014) found that OPT3 can transport Fe2<sup>+</sup> ions when expressed in Xenopus oocytes, the substrate of OPT3 remains controversial because the Fe transport function of OPT3 could not be confirmed in other experiments (Mendoza-Cózatl et al., 2014). Even if Fe enters the phloem through OPT3 as free ions, it would be difficult for Fe to move in this way, and probably the Fe ions would be chelated (**Figure 12**; see section "Introduction"). OPT3 does not mediate GSH transport in S. cerevisiae (Zhai et al., 2014), as other OPT transporters did (Zhang et al., 2016 and references therein). However, the higher leaf GSH content found in the opt3-2 mutant in relation to Columbia, under Fe sufficiency (**Table 2**) or under Cd treatment (Akmakjian, 2011), supports the idea of LODIS being a compound related, directly or indirectly, to GSH, such as GS-Fe-NO or NICs (Ramírez et al., 2011; Rahmanto et al., 2012; Darbani et al., 2013; Buet and Simontacchi, 2015). Nonetheless, this idea deserves further research.

In relation to the possible interactions between GSNO(NO), GSH and ethylene, it should be said that, besides the mutual and positive influence between NO and ethylene already described in previous works (**Figure 13**; García et al., 2011; Romera et al., 2011; Liu et al., 2017), GSNO(NO), GSH and ethylene could also interact at other levels. GSH can enhance ethylene production by affecting some ethylene synthesis enzymes (Datta et al., 2015). On the other hand, the relationship between GSNO and ethylene can be feasible because both GSNOR (Rustérucci et al., 2007; see above) and several ethylene synthesis enzymes, like 5-methilthioribose kinase (induced in Fe-deficient roots; García et al., 2010; Romera et al., 2017 and references therein), are located in the phloem (Pommerrenig et al., 2011). Moreover, enzymes implicated in ethylene synthesis, such as SAM synthetases (also named methionine adenosyltransferases; Sauter et al., 2013), can be inhibited by S-nitrosylation (Lindermayr et al., 2006; Freschi, 2013). Since GSNO has been implicated in the reversible S-nitrosylation of proteins (Malik et al., 2011; Begara-Morales et al., 2015; Zaffagnini et al., 2016), it is possible that higher GSNO levels (such as those found in Fe-sufficient WT roots: **Figures 10**, **11**) could contribute to S-nitrosylation of ethylene synthesis enzymes and, consequently, to the inhibition of ethylene synthesis (**Figure 13**). By contrast, lower GSNO levels (such as those found in Fe-deficient WT roots or in Fe-sufficient opt3-2 and dgl roots: **Figures 10**, **11**) could contribute to de-nitrosylation of ethylene synthesis enzymes and, consequently, to promotion of ethylene synthesis (**Figures 1B**, **2**). In supporting this view, silencing GSNOR in Nicotiana attenuata, which leads to higher GSNO content, decreased the herbivore-induced accumulation of ethylene (Wünsche et al., 2011). Reciprocally to the possible

influence of GSNO(NO) on ethylene synthesis, ethylene (ACC) could decrease GSNO content by activating GSNOR1 expression (**Figures 9**, **13**). GSNOR1 expression is also influenced by other hormones, such as salicylic acid which activates it or jasmonic acid which down-regulates it (Díaz et al., 2003). Very recently, it has been found that ethylene (ACC) can increase NO content by activating enzymes involved in its synthesis, such as nitrate reductase and nitric oxide synthaselike (Liu et al., 2017). This would imply that ethylene could simultaneously increase NO accumulation and decrease GSNO content (**Figure 13**).

Finally, in the relationship between GSNO(NO)/ethylene and the putative Fe sensor BTS, some results suggest that the BTS protein could act upstream of ethylene. The SAM1 gene, encoding a SAM synthetase (involved in ethylene synthesis; Sauter et al., 2013), is up-regulated in the bts-3 mutant (Hindt et al., 2017). In the same way, several genes whose expression is activated by ethylene, such as FIT, bHLH38, bHLH39, MYB72, BGLU42, FRO2, IRT1, FRD3, and NAS2 (García et al., 2010), are up-regulated in bts-3 roots under Fesufficient conditions (Hindt et al., 2017). In rice, roots from HRZ-knockdown plants have higher jasmonate content than the WT ones when grown under Fe sufficiency (Kobayashi et al., 2016). All these results suggest that BTS/HRZs may act as sensors of LODIS, as suggested by Kobayashi and Nishizawa (2014), and depending on its binding state could then affect the synthesis of hormones, like ethylene or jasmonate.

### CONCLUSION

The results presented in this work, along with already published results, suggest that shoots, through LODIS, play an important role in the regulation of Fe acquisition genes in roots, which is depicted in the Working Model of **Figure 12**. According to this model, under Fe sufficiency, sufficient Fe goes to leaves and enters the phloem to form LODIS, which travels to roots and represses ethylene synthesis and signaling, and GSNOR expression and activity, which leads to increased GSNO content. GSNO(NO), GSH, and ethylene can interact in different ways (see above paragraph and **Figure 13**). Moreover, a high GSNO content can compromise auxin signaling and transport (Shi et al., 2015), which is important because auxin has also been involved in the regulation of Fe responses, where is intimately interrelated with ethylene and NO (Chen et al., 2010; Romera et al., 2011, 2017). Under Fe deficiency, the lack of LODIS derepresses ethylene synthesis and signaling, and GSNOR expression and activity, which leads to decreased GSNO content. All this would cause the up-regulation of Fe acquisition genes.

This working model could open the way to better understand the role of shoots and roots in the regulation of Fe deficiency responses. According to this model, LODIS would have a double role in the regulation of Fe acquisition genes, by controlling both ethylene synthesis and signaling (**Figure 12**). This double role of LODIS would explain why the ethylene overproducer mutant eto and the treatments with ethylene did not activate the expression of Fe acquisition genes when plants are grown with high levels of Fe (Lucena et al., 2015): despite high ethylene levels, LODIS would block ethylene action. All the results discussed in this work imply that the relationship of GSH, NO and GSNO with ethylene, with GSNOR, with LODIS and with the expression of Fe acquisition genes is complex and deserves further research.

### AUTHOR'S NOTE

A very recent work by Grillet et al. (2018) shows that some peptides (IRON MAN or IMA peptides) associated with the phloem, and preferentially expressed in leaves, could be implicated in the regulation of Fe responses in roots. The upregulation of Fe responses in lines overexpressing genes encoding these peptides and their ability to bind Fe2<sup>+</sup> suggest that these peptides could act either as activators (unbound IMA peptides) or repressors (IMA peptides bound to Fe) of the responses. Perhaps LODIS is related to the ratio between IMA peptides bound to Fe and unbound IMA peptides.

#### AUTHOR CONTRIBUTIONS

fpls-09-01325 September 8, 2018 Time: 14:26 # 14

FR and MG designed the experiments after discussions with PB, RP-V, and EA. MG and CL conducted the laboratory work. FC and MG determined GSNOR, GSNO, and GSH. ÁZ, EB, and JG-M determined ACC. FR, MG, RP-V, and EA wrote the manuscript that was improved by the other authors.

#### FUNDING

This work was supported by the European Regional Development Fund from the European Union, the 'Ministerio de Economía

#### REFERENCES


y Competitividad' (Projects AGL2013-40822-R and AGL2015- 65104-P) and the 'Junta de Andalucía' (Research Groups AGR115, BIO159, and BIO192).

#### ACKNOWLEDGMENTS

We thank Dr. Brian M. Waters, of the University of Nebraska (Lincoln, NE, United States), for the English correction and valuable suggestions in the editing of the manuscript. We thank Dr. Michael Grusak for providing seeds of the Sparkle [dgl,dgl] mutant, the Arabidopsis Biological Resource Center for providing seeds of the frd3-3 mutant and Dr. Stacey for providing seeds of the opt3-2 mutant. We also thank Inmaculada Montilla and Carmelo Ruíz-Torres for their excellent technical support.



with a broad substrate range. Plant Mol. Biol. 40, 37–44. doi: 10.1023/A: 1026438615520


H+-ATPase gene expression by affecting FER (or FER-like) gene activity. J. Exp. Bot. 57, 4145–4154. doi: 10.1093/jxb/erl189


disconnected tissues throughout Arabidopsis development. Plant Cell 23, 2725–2737. doi: 10.1105/tpc.111.088088


molecular views in tobacco plants. Plant Cell Rep. 25, 365–373. doi: 10.1007/ s00299-005-0092-3


iron signaling and redistribution of iron and cadmium in Arabidopsis. Plant Cell 26, 2249–2264. doi: 10.1105/tpc.114.123737


**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 © 2018 García, Corpas, Lucena, Alcántara, Pérez-Vicente, Zamarreño, Bacaicoa, García-Mina, Bauer and Romera. 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) and the copyright owner(s) 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.

# Differential Diel Translation of Transcripts With Roles in the Transfer and Utilization of Iron-Sulfur Clusters in Arabidopsis

#### Hongliang Zhang and Ute Krämer\*

Molecular Genetics and Physiology of Plants, Faculty of Biology and Biotechnology, Ruhr University Bochum, Bochum, Germany

#### Edited by:

Manuel González-Guerrero, Universidad Politécnica de Madrid (UPM), Spain

#### Reviewed by:

Ping Lan, Institute of Soil Science (CAS), China Marinus Pilon, Colorado State University, United States Louis Grillet, Academia Sinica, Taiwan

#### \*Correspondence:

Ute Krämer Ute.Kraemer@ruhr-uni-bochum.de; Ute.Kraemer@rub.de

#### Specialty section:

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

Received: 07 August 2018 Accepted: 22 October 2018 Published: 13 November 2018

#### Citation:

Zhang H and Krämer U (2018) Differential Diel Translation of Transcripts With Roles in the Transfer and Utilization of Iron-Sulfur Clusters in Arabidopsis. Front. Plant Sci. 9:1641. doi: 10.3389/fpls.2018.01641 Iron-sulfur (Fe-S) clusters are evolutionarily ancient ubiquitous protein cofactors which have mostly catalytic functions but can also have structural roles. In Arabidopsis thaliana, we presently know a total of 124 Fe-S metalloproteins that are encoded in the genome. Fe-S clusters are highly sensitive to oxidation. Therefore, we hypothesized that Fe-S cluster protein biogenesis is adjusted following the daily rhythms in metabolism driven by photosynthesis at the whole-plant, organ, cellular and sub-cellular levels. It had been concluded previously that little such regulation occurs at the transcript level among the genes functioning in Fe-S cluster assembly. As an initial step toward testing our hypothesis, we thus addressed the diel time course of the translation state of relevant transcripts based on publicly available genome-wide microarray data. This analysis can answer whether the translation of the pool of transcripts of a given gene is temporarily either enhanced or suppressed, and when during the day. Thirty-three percent of the transcripts with functions in Fe-S cluster assembly exhibited significant changes in translation state over a diurnal time course, compared to 26% of all detected transcripts. These transcripts comprised functions in all three steps of cluster assembly including persulfide formation, Fe-S cluster formation and Fe-S cluster transfer to target apoproteins. The number of Fe-S cluster carrier/transfer functions contributed more than half of these transcripts, which reached maxima in translation state either during the night or the end of the night. Similarly, translation state of mitochondrial frataxin and ferredoxin, which are thought to contribute Fe and electrons during cluster formation, peaked during the night. By contrast, translation state of chloroplast SUFE2 in persulfide formation and cytosolic Fe-S cluster formation scaffold protein NBP35 reached maxima in translation state during the day. Among the transcripts encoding target Fe-S clusterutilizing proteins, 19% exhibited diurnal variation in translation state. Day-time maxima of translation state were most common among these transcripts, with none of the maxima during the night (ZT18). We conclude that diurnal regulation of translation state is important in metalloprotein biogenesis. Future models of Fe-S protein biogenesis require more comprehensive data and will have to accommodate diurnal dynamics.

Keywords: translatome, microarray, diurnal, metalloprotein, Fe, Zn, Mn, Cu

### INTRODUCTION

fpls-09-01641 November 10, 2018 Time: 13:42 # 2

Across all groups of biological organisms, proteins that can bind one or several iron-sulfur (Fe-S) cluster cofactors in vivo (Fe-S proteins) fulfill central cellular biochemical functions. Several types exist of protein-bound Fe-S clusters, based on chemical structures and oxidation states. Among these, cubane [4Fe–4S] and rhombic [2Fe–2S] are the most widespread Fe– S clusters, containing ferrous or ferric Fe(+II) or (+III) and sulfide S(−II) (Lill, 2009; Balk and Schaedler, 2014). Disruption of Fe-S protein biogenesis has been associated with serious diseases in human and animals (Lill, 2009; Wachnowsky et al., 2017). For example, frataxin is thought to function as the iron donor for the mitochondrial Isu1 scaffold protein complex in Fe-S cluster assembly in human. Depletion of frataxin causes the neurodegenerative disease Friedreich's ataxia, attributed to defective Fe–S protein activity and iron accumulation (Campuzano et al., 1996; Rötig et al., 1997).

In plants, Fe-S proteins are abundant in mitochondria and chloroplasts, for example in respiratory and photosynthetic electron transfer chains (Blaby-Haas and Merchant, 2013; Couturier et al., 2013), and Fe-S cluster assembly occurs in both of these organelles (Balk and Schaedler, 2014). Notably, the metabolism, repair and epigenetic modification of DNA all depend on Fe-S protein functions (Buzas et al., 2014; Zhang, 2015). The catalytic subunits of eukaryotic DNA polymerases α, δ, ε, and ζ contain [4Fe-4S] cofactors that are required for protein-protein interactions in the active protein complex (Netz et al., 2012). In Arabidopsis, Fe-S cluster-binding residues are conserved in these proteins, namely POLA1, POLD1, POLE1A/POL2A, POLE1B/POL2B, and REV3, the catalytic subunit of DNA polymerase ζ(Shultz et al., 2007; **Supplementary Dataset 1**). Moreover, proteins of the Demeter (DME) family contain a [4Fe-4S] cofactor and function in the demethylation and base excision repair of DNA (Choi et al., 2002; Wöhrmann et al., 2012; Buzas, 2016). Expanding from the list of Fe-S proteins updated in 2010 (Balk and Pilon, 2011), a total of 124 genes encoding proteins that contain Fe-S clusters are presently known in Arabidopsis (**Supplementary Dataset 1**).

The biosynthesis of Fe-S proteins requires cellular Fe-S cluster assembly. The first knowledge on Fe-S assembly pathways was obtained in Escherichia coli and Azotobacter vinelandii (Lill, 2009). Subsequent work revealed that genes encoding Fe-S assembly proteins are well conserved from prokaryotes to eukaryotes (Balk and Lobréaux, 2005; Lill, 2009; Py and Barras, 2010). In recent years, an increasing number of proteins of the Fe-S cluster assembly pathways were characterized employing mutants of A. thaliana (Couturier et al., 2013; Balk and Schaedler, 2014), and yet, our present knowledge in plants still relies to a large degree on inference from other model organisms. In plants, there are three cellular pathways of Fe-S cluster biogenesis. These are the Sulfur mobilization (SUF) pathway in plastids, the Iron Sulfur Cluster (ISC) pathway in mitochondria, and Cytosolic Iron-sulfur protein Assembly (CIA) pathway in the cytosol. Although each of these Fe-S cluster assembly pathways involves different proteins, all three of them proceed according to common biosynthetic principles (**Figure 1**). First, cysteine is metabolized by cysteine desulfurase to generate a persulfide (protein-S-S-H) at the active site of the enzyme. Subsequently, a terminal sulfane S<sup>0</sup> of this persulfide is transferred to a scaffold protein complex, followed by the reduction to sulfide (S−II). The integration of iron on the scaffold protein results in the formation of the Fe-S cluster. Finally, the assembled Fe-S cluster is transferred to target apoproteins via carrier/transfer proteins. Based on three excellent reviews (Balk and Pilon, 2011; Couturier et al., 2013; Balk and Schaedler, 2014), we have assembled a list of 49 A. thaliana genes encoding proteins involved in Fe-S cluster assembly (**Supplementary Dataset 2**). Note that some of these Fe-S cluster assembly proteins bind Fe-S clusters themselves, mostly Fe-S scaffold and carrier/transfer proteins.

Although we have good knowledge of the Fe-S protein biogenesis machinery, we know comparably little about the regulation of Fe-S protein biogenesis. The Fe-S protein biogenesis machinery of bacteria involves transcriptional regulators, such as Fur, OxyR and IscR. For example, under oxidative stress, OxyR activates the expression of the suf operon in E. coli (Roche et al., 2013). However, none of the known bacterial regulators of Fe-S protein biogenesis has any orthologs in a plant or an algal genome (Couturier et al., 2013), suggesting that plants utilize different regulatory mechanisms. Protein-bound Fe-S clusters can be destroyed by oxygen or reactive oxygen species (ROS), such as hydrogen peroxide or nitric oxide (Beinert et al., 1997). Free Fe2<sup>+</sup> released from Fe-S clusters can catalyze the Fenton reaction, which generates even more ROS (Halliwell and Gutteridge, 1992; Ravet et al., 2009). Transcript levels of few genes encoding Fe–S cluster assembly proteins respond to oxidative stress, nutritional iron deficiency, and heavy metal excess, in plants (Liang et al., 2014) (see also Discussion). Overall there is only little environment-dependent regulation in plants of Fe-S cluster assembly at the transcript level (Balk and Schaedler, 2014).

In plants, contrasting photosynthetic activity between day and night confers a diel rhythm in plant metabolism as well as in the generation of ROS, primarily locally in chloroplasts but also across cells, tissues and organs of plants (Lai et al., 2012; Sierla et al., 2013; Trotta et al., 2014). Plants are capable of regulatory adjustment to these fluctuations, which can be controlled by light, stress, metabolic state or the circadian clock (Lai et al., 2012**;** Sierla et al., 2013; Trotta et al., 2014). Surprisingly, it was noted that Fe-S cluster assembly – a potentially highly ROSsensitive process – exhibits no diurnal profile of regulation at the transcript level (Balk and Schaedler, 2014). Consequently, Fe-S protein biogenesis of plants is either sufficiently ROS-insensitive, or alternatively it could be diurnally controlled at a different level of regulation.

To examine whether Fe-S protein biogenesis is diurnally regulated at the level of translational regulation, we re-analyzed published microarray-based diurnal translatome data (Missra et al., 2015). Accordingly, subsets of Fe-S cluster assembly and Fe-S utilizing proteins are under pronounced translational control. Among these, the translation of transcripts encoding Fe-cluster assembly proteins tended to peak at night while the translation of the majority of transcripts encoding Fe cluster-utilizing proteins reached a minimum. We conclude that the diurnal regulation of metalloprotein biogenesis deserves to be addressed in a

more quantitative and comprehensive translatome sequencing approach.

in the Arabidopsis genome (see Supplementary Dataset 2).

#### MATERIALS AND METHODS

Diel translatome time course data from ATH1 GeneChip microarray hybridization was taken from Missra et al. (2015) (**Figure 2**). Briefly, RNA was fractionated into non-polysomal (NP), small polysomal (SP), and large polysomal (LP) fractions using sucrose density centrifugation. In these fractions, the mRNA molecules were estimated to be bound by two and seven ribosomes on average so that SP and LP fractions were weighted by 2 and 7, respectively (Missra et al., 2015). The translation state (TL) of a given mRNA was approximated as follows: TL = (2 × SP + 7 × LP)/(NP + SP + LP), resulting in a global maximum TL of 6.48 and a minimum TL of 0.46. We additionally estimated relative rates of protein biosynthesis (PBR) for each protein across the diel time course by multiplying translation state with transcript levels (PBR = TL× TX, whereby TX = NP + SP + LP) as given in Missra et al. (2015). This calculation is based on the assumption that there is a correlation between the amount of ribosome-associated mRNA and its rate of translation, which is a simplification. The authors made no adjustment for global oscillation of TL across all expressed genes, which peaked at noon (ZT6; ∼ 0.8) and was minimal at the end of the night (ZT0; ∼ 0.68). Following the strategy employed by Missra et al. (2015), transcripts concluded to exhibit diel variation of TL fulfilled one of three criteria: 1TL (difference between maximum and minimum average TL of three replicates) > 0.7, or ANOVA (Analysis Of Variance) P < 0.05, or Significance Analysis of Microarrays (SAM) with a collective FDR < 0.10 (Tusher et al., 2001).

For analyzing the TL of Fe-S proteins and Fe-S assembly proteins in this study, we first assembled a list of 124 Fe-S cluster-containing proteins (**Supplementary Dataset 1**) and 49

Fe-S assembly proteins (**Supplementary Dataset 2**) identified in A. thaliana so far, expanding from previous work (Balk and Pilon, 2011; Couturier et al., 2013; Balk and Schaedler, 2014). Then, the translatome data of the genes in the two lists of Fe-S proteins and Fe-S assembly proteins were extracted from **Supplementary Dataset 1** of the original publication (Missra et al., 2015) as shown in **Supplementary Datasets 3**, **4** of this publication. Note that we consider as Fe-S cluster-utilizing proteins (**Supplementary Dataset 4**) only those Fe-S clustercontaining proteins that are not involved in Fe-S cluster assembly (**Supplementary Dataset 3**). Translatome data were further analyzed by using the software Genesis 1.7.6 (Sturn et al., 2002). First, translatome data were scaled by the mean center method. That is, for each transcript at each time point, Z = TLZT0,6,12,or18 – average (TLZT0 + TLZT6 + TLZT12 + TLZT18). The purpose of this was to aid visualization of diel cycles of a number of genes together. Subsequently, the transcripts were clustered into different groups by the average linkage clustering method (Sokal and Michener, 1958). Analysis of TL of metalloproteins dependent on cofactors other than Fe-S clusters was done based on manually assembled lists (**Supplementary Dataset 5**). Only the transcripts showing significant diurnal dynamics at the translatome level were used when calculating the proportion of transcripts reaching maximum TL at each time point.

#### RESULTS

#### Global Translation Status of Fe-S Proteins and Fe-S Assembly Proteins Over a Diurnal Cycle

According to the previous analysis, out of 12,342 detected transcripts, 3,218 transcripts (26%) varied significantly in their translation state (TL) over a diurnal cycle (**Supplementary Dataset 6**). In this dataset, 39 (out of a total of 49) transcripts encoding Fe-S assembly proteins and 68 (out of 97) transcripts encoding Fe-S cluster-utilizing proteins were detected as expressed (**Supplementary Datasets 3**, **4**). Among these, 13 transcripts encoding Fe-S assembly proteins (33%) and 13 transcripts encoding Fe-S cluster-utilizing proteins (19%) showed diurnal fluctuations in TL (**Table 1**). The apparent overrepresentation of Fe-S cluster assembly genes among genes with diurnal changes in TL was not statistically significant according to a Fisher's exact test.

#### Transcripts Encoding Fe-S Assembly Proteins With Diurnal Changes in Translation State

Thirteen transcripts encoding Fe-S assembly proteins varied significantly in their TL over the diurnal cycle. Among these 13 transcripts, 7 correspond to mitochondrial, 5 to plastidic and one to cytosolic proteins. These transcripts clustered into two groups of similar TL profiles over the diurnal cycle (**Figure 3A**). TL of transcripts in the first cluster peaked during and at the end of the night (ZT18 or ZT0). Interestingly, 7 out of 9 transcripts in this group encode proteins associated with Fe-S cluster carriage/transfer, including mitochondrial ISCA2, ISCA4, NFU4, and GRXS15 and chloroplast SUFA, NFU1 and GRXS14. The remaining two transcripts in this cluster encode the candidate Fe donor protein mitochondrial frataxin (FH) and the alpha-helical mitochondrial ferredoxin, respectively. This ferredoxin protein is thought to provide an electron for Fe-S cluster synthesis on ISU1, the major scaffold protein in mitochondria. TL of transcripts in the second cluster peaked during the day (ZT6 or ZT12). Among the four transcripts that reached peak TL during the day, two encode plastidic cysteine desulfurase activators and interactors of CpNifS, namely SUFE2 and SUFE3 (Narayana Murthy et al., 2007). The other two transcripts encode the cytosolic Fe-S assembly scaffold protein NBP35 (Bych et al., 2008; Kohbushi et al., 2009; Bastow et al., 2017), and HSCA2, one of two abundant mitochondrial HSP70-type chaperones which was proposed to function in Fe-S cluster transfer (Balk and Pilon, 2011). The estimation of protein biosynthesis rates (PBR) (**Supplementary Figure 1A** and **Supplementary Dataset 7**) resulted in diel profiles that were similar to TL for most Fe-S assembly proteins, but considerably

TABLE 1 | Extent of diurnal regulation of translation state of Fe-S cluster-related proteins.


<sup>a</sup>Transcript levels detected as present in all ribosome-associated fractions at all four time points (Missra et al., 2015). <sup>b</sup>Number counts for transcripts exhibiting significant diurnal variation in TL according to Missra et al. (2015) (see Materials and Methods).

more diverse than diel TL profiles for Fe-S carrier/transfer proteins (see **Figure 3A**).

occupancy per transcript. #Putative assembly protein. Fe-S: indicates the ability to bind an Fe-S cluster in (A).

### Transcripts Encoding Fe-S Proteins With Diurnal Changes in Translation State

Twenty-three transcripts encoding Fe-S proteins varied significantly in TL over the diurnal time course. Among these, 13 transcripts encode Fe-S cluster-utilizing proteins (**Figure 3B**) and 10 transcripts encode Fe-S proteins that are involved in cluster assembly and Fe-S protein biogenesis (**Figure 3A**). The 13 transcripts encoding Fe-S cluster-utilizing proteins comprised 6 plastidic, 4 nuclear, 2 cytosolic, and 1 mitochondrial localized proteins. Proteins detected to exhibit diurnal regulation of TL have various functions, in particular DNA and RNA metabolism as well as isoprenoid and leucine biosynthesis. All except two of these proteins are predicted to bind the most abundant [4Fe-4S] cluster, which is less stable than the [2Fe-2S] cluster. Only DWARF27B, a homolog of a rice β-carotene isomerase protein acting in strigolactone biosynthesis, and Xanthine the dehydrogenase XDH1 bind [2Fe-2S] clusters (Zhang et al., 2012; Balk and Schaedler, 2014). Fe-S cluster-utilizing proteins fell into three groups of differing diel TL profiles (**Figure 3B**). TL of 8 transcripts peaked during the day (ZT6 or ZT12), TL of two transcripts peaked at ZT0 and ZT6, and TL of three transcripts peaked at the end of the night (ZT0) (**Figure 3B**). TL of none

of the Fe-S cluster-utilizing proteins peaked at ZT18 during the night, in contrast to the Fe-S carrier/transfer proteins that exhibited diel TL differences (see **Figure 3A**). When, instead of TL, we considered estimates of PBR, we mostly observed no changes or only minor shifts for few proteins (**Supplementary Figure 1** and **Supplementary Dataset 8**).

Among the transcripts showing diel changes in TL, ten transcripts encode Fe-S proteins that are also involved in Fe-S protein biogenesis (see **Figure 3A**). The TL of 7 of these 10 transcripts peaked during the night (ZT18) and one peaked at the end of the night (ISCA4). Only the cytosolic Fe-S cluster scaffold protein NBP35 and SUFE3 peaked during the daytime.

### Transcripts Encoding Proteins Binding Heme and Cationic Fe, Zn, Cu, or Mn With Diurnal Changes in Translation State

To compare Fe-S cluster-containing proteins with other metalloproteins, we determined the proportion showing diurnal variation in TL among proteins binding heme/cytochrome cofactors (29%) and proteins binding cationic Fe (28%), Zn (22%), Cu (30%), or Mn (39%) (**Table 2**). Among these, in comparison to all functions (26%) only Zn-dependent proteins are significantly differently represented, with an underrepresentation in contrast to the trend for overrepresentation of heme/cytochrome-dependent, Cu-dependent and Mn-dependent proteins. In order to compare the timing of biogenesis of metalloproteins among those exhibiting significant variation in TL over the diurnal cycle, we determined the proportion of transcripts peaking in TL at each time point (**Figure 4**). Whereas TL of 62% of transcripts encoding Fe-S cluster assembly proteins peaked at ZT18, TL of none of the transcripts encoding Fe-S cluster-utilizing proteins peaked at this time point (**Figure 4A**; statistically significantly different from all detected transcripts at P < 0.01, Fisher's exact test), in accordance with observations described earlier (see **Figure 3**). The diel TL profile of transcripts encoding heme/cytochromeas well as Fe, Zn, and Cu cation-binding proteins was similar to that of all detected transcripts, with a maximum of 30–50%

TABLE 2 | Extent of diurnal regulation of translation state of heme/cytochrome and Fe, Zn, Cu, and Mn cation-binding metalloproteins.


<sup>a</sup>Transcript levels detected as present in all ribosome-associated fractions at all four time points (Missra et al., 2015). <sup>b</sup>Number counts for transcripts exhibiting significant diurnal variation in TL according to Missra et al. (2015) (see Materials and Methods). <sup>∗</sup>P < 0.05 (Pearson Chi-Square test).

peaking at ZT18, a less pronounced secondary maximum of 25–42% peaking at ZT6 and a minimum of 0–17% peaking at ZT12 (**Figures 4A,B**). Notably, none of the transcripts encoding Cu-dependent proteins showed an afternoon peak in TL at ZT12 (statistically significantly different from all detected transcripts, P < 0.05, Fisher's exact text), and as many as 50% peaked during the night at ZT18. The diel course of TL of transcripts encoding Mn-dependent proteins differed from that of all others, with a maximum of 39% of proteins peaking at the end of the night (ZT0).

#### DISCUSSION

Here we addressed the translation state of transcripts encoding Fe-S cluster-binding proteins and other metalloproteins over a diurnal cycle. While this has not been addressed to date, previous studies have addressed the responses of various components in Fe-S cluster assembly to changed environmental conditions.

Besides oxygen (see Introduction), iron and sulfur supply were reported to affect Fe-S cluster biosynthesis. Under iron deficiency conditions, SUFB expression is down-regulated in both A. thaliana (Xu et al., 2005; Sivitz et al., 2012; Rodríguez-Celma et al., 2013) and in rice (Liang et al., 2014). Corresponding to the decreased transcript level, SUFB protein abundance is also decreased (Hantzis et al., 2018). SUFB is a component of the SUFBC2D complex which functions as a scaffold protein in the plastidic SUF pathway (Hu et al., 2017). Besides SUFB, the protein abundance of a Fe-S carrier/transfer protein SUFA was decreased under Fe deficiency, whereas SUFA transcript abundance was unaffected, suggesting the possibility of regulation at the translational or protein level. Recently, it was reported that S and Fe uptake are closely coordinated (Astolfi et al., 2010; Astolfi et al., 2012; Zuchi et al., 2015). This is not surprising, given that Fe-S clusters are the most abundant among all the Fecontaining cofactors (Forieri et al., 2013; Balk and Schaedler, 2014; Vigani and Briat, 2016). Fe deficiency can result in the down-regulation of transcript levels of sulfite reductase (SiR). SiR is a protein involved in sulfate assimilation (Forieri et al., 2017), which contains an Fe4S<sup>4</sup> cluster itself. This response might contribute to a decreased S flux into cysteine, the only S donor for Fe-S cluster biosynthesis (Balk and Schaedler, 2014).

Light is another environmental factor known to enhance the incorporation of Fe-S clusters into Fe-S cluster-utilizing proteins of isolated chloroplasts (Takahashi et al., 1986) as well as the expression of some Fe-S cluster assembly genes. SUFE1 transcript levels were reported to increase 2.3-fold in the light (Ye et al., 2006), and light-grown plants contained 2.5 times higher SUFE1 protein levels than dark-grown plants. Another study showed that prolonged darkness leads to significant decreases in the protein abundance of several key components in SUF pathway, including GRXS14, NFU2, NFU3, SUFA, and SUFB (Rey et al., 2017).

It was reported that the expression of major Fe homeostasis genes is under circadian regulation, such as IRT1, bHLH39, and FER1 (Hong et al., 2013). Regulation of ferritin gene (AtFER1, AtFER3, and AtFER4) expression by the clock was

significant diurnal variation in TL were used in this analysis (see Tables 1, 2). Asterisks mark significant differences from all detected transcripts at the same time point (Fisher's exact test; <sup>∗</sup>P < 0.05, ∗∗P < 0.01).

confirmed by showing that they are direct transcriptional targets of PRR7, a central clock component (Liu et al., 2013). In turn, Fe nutrition status can modulate the plant circadian clock, with period lengthening under Fe-deficient conditions (Chen et al., 2013; Hong et al., 2013; Salomé et al., 2013). We observed no significant diel regulation of translation state of these transcripts (**Supplementary Dataset 9**).

#### TL of Transcripts Encoding Proteins in Early Steps of Fe-S Cluster Assembly and in Fe-S Cluster Transfer Show Contrary Diurnal Dynamics in Translation State

In our analysis, we separated Fe-S cluster-associated functions into Fe-S cluster assembly and Fe-S cluster utilization. By comparison to all genes detected as expressed, those with functions in Fe-S cluster assembly were slightly, ca. 1.3-fold overrepresented among the genes exhibiting diurnal variation in TL. By contrast, genes associated with Fe-S cluster utilization were slightly, ca. 0.7-fold, underrepresented (**Table 1**). Expanding this analysis, genes encoding Mn-dependent proteins were the most overrepresented, ca. 1.5-fold, among genes exhibiting diurnal variation in TL, followed by genes encoding Cudependent proteins (1.2-fold) (**Table 2**). The proportion of genes exhibiting diurnal variation in TL among heme/cytochromedependent and Fe cation-dependent proteins was similar to that of all expressed genes. It is important to note that none of these deviations were statistically significant, possibly because of a lack of sensitivity of statistical testing given the small number of genes in each of the functional classes. The only significant difference was an underrepresentation, approximately 0.85-fold, of genes encoding Zn-dependent proteins among those exhibiting significant diurnal variation in TL. The considerably larger number of genes in this class is likely to have facilitated the statistical detection of this effect. This result may reflect the high proportion of transcription factors of overall low abundance among Zn-dependent proteins, by comparison to genes encoding highly abundant proteins functioning in metabolism. Taking these results together, diel dynamics in TL are relevant for all classes of metalloproteins.

However, based on this preliminary analysis, the pervasiveness of such diel dynamics in TL among metalloproteins is similar or only slightly higher than for non-metalloproteins. Moreover, our results suggest differences between different types of metalloproteins and metalloprotein functions, with overall more pronounced diel dynamics in TL among Mn-dependent proteins and Fe-S cluster assembly proteins and overall less diel dynamics in TL among Zn-dependent proteins. Evidently, TL analyzed in this study cannot be seen as a proxy of the overall rate of protein biogenesis (PBR), which is better described by an estimate of the total amount of ribosomal occupancy on transcripts derived from a given gene in the future (e.g., in this study: PBRest = 2 × SP + 7 × LP). A re-analysis of data shown in **Figure 3** suggested little difference between the diel dynamics in TL and the diel dynamics in PBR (**Supplementary Figure 1**) except for Fe-S carrier/transfer proteins, for which diel PBR profiles were rather variable.

Among the metalloprotein-encoding genes exhibiting diurnal variation in TL, we observed some striking differences in the diel profiles (**Figure 4**). The deviating profiles for Mn and, very pronouncedly, Cu-dependent metalloproteins are consistent with an overall metal-specific differentiation of TL over the diel cycle for these metals. In addition, there was a clear difference between genes encoding Fe-S cluster assembly proteins and those encoding Fe-S cluster-utilizing proteins. More than 60% of the former peaked during the night at ZT18, whereas none of the latter peaked at this time point. Instead, TL of genes encoding Fe-S cluster-utilizing proteins peaked mostly ZT6 (42%) and ZT12 (33%). A possible interpretation of this observation is the occurrence of a temporal gap between the biogenesis of Fe-S clusters and their incorporation into freshly synthesized polypeptide chains of Fe-S cluster-utilizing apo-metalloproteins.

A more detailed examination of the diel TL profiles of Fe-S cluster assembly proteins identified two distinct clusters (**Figure 3A**). The Fe-S assembly proteins in the first cluster, for which TL peaked during the night (ZT18 and to a lesser degree ZT0) mainly participate in the third step of Fe-S cluster protein biogenesis, which is Fe-S cluster transfer to target apoproteins via carrier/transfer proteins (**Figure 1**). By contrast, the three Fe-S cluster assembly proteins SUFE2, SUFE3, and NBP35 in the second cluster, which peaked during the day (ZT12 and ZT6), contribute to the first two steps of Fe-S protein biogenesis, persulfide formation and Fe-S cluster formation on the scaffold protein complex (**Figure 1**). These data suggest that some functions in Fe-S cluster synthesis and Fe-S cluster transfer to apoproteins have opposing diel dynamics of association with ribosomes. Taking all our observations together, the biosynthesis of apometalloproteins that require Fe-S clusters appears to occur preferentially during the day, as well as just before dawn for some cytosolic/nuclear Fe-S proteins (**Figure 3** and **Supplementary Figure 1**).

In contrast to cytoplasmic NBP35, the main scaffold proteins SUFB, C and D of plastids and ISU1 of mitochondria did not show diurnal dynamics at the level of translation state. In the context of the model described above, it is possible that their basal transcript levels and translation states are sufficiently high throughout the day, which can satisfy the need of Fe-S biogenesis even during the peak phase of Fe-S protein synthesis. It is worth noting that besides stimulating the cysteine desulfurase activity of NFS2 alike SUFE2 (SufE activity), SUFE3 possesses quinolinate synthase activity which is required for NAD cofactor synthesis (Narayana Murthy et al., 2007). The quinolinate synthase activity of SUFE3 is dependent on the presence of a highly oxygensensitive [4Fe-4S] cluster (Cicchillo et al., 2005; Ollagnier-de Choudens et al., 2005; Narayana Murthy et al., 2007). SUFE3 cannot complement the embryo lethal phenotype of a sufe1 knock-out mutant, suggesting the SufE activity of SUFE3 is likely dedicated to Fe-S cluster formation in its own quinolinate synthase domain (Narayana Murthy et al., 2007; Balk and Pilon, 2011).

Three mitochondrial proteins are less easily integrated into the general patterns observed here, frataxin (FH), ferredoxin and HSCA2. FH and ferredoxin are thought to have a role in the second step of Fe-S cluster assembly by analogy to human and yeast (Balk and Schaedler, 2014), but their TLs peak at night alongside the TL of several proteins with Fe-S carrier/transfer functions (see **Figure 3A**, in agreement with **Supplementary Figure 1**). FH was proposed to deliver Fe to ISU1 by direct interaction with the NFS1/ISD11/ISU1 complex in mitochondria (Balk and Schaedler, 2014). An alternative model has also been discussed in which FH promotes the interaction of the cysteine desulfurase NFS1 with the scaffold ISU proteins to favor the sulfur transfer reaction (Couturier et al., 2013). FH was reported to localize to both mitochondria and plastids (Turowski et al., 2015). HSCA1 can functionally complement the yeast ssq1 knockout mutant and its ATPase activity is enhanced by HSCB and ISU1 (Xu et al., 2009). The role of HSCA2 in Fe-S protein biogenesis remains to be investigated. Additional and more detailed information will be required to test the possible models arising from this study. In particular also, diel variation in TL was only detected for a subset of proteins in each functional class.

#### Extensive Translational Cycling of Fe-S Transfer/Carrier Proteins May Be Determined by Their Specific Roles

Among the 13 Fe-S assembly proteins cycling in translation state, 7 are Fe-S carrier/transfer proteins. ISCA4, NFU4, GRXS15, and ISCA2 localize in mitochondria, whereas NFU1, SUFA, and GRXS14 are localized in plastids. So far, a number of Fe-S transfer/carrier proteins have been identified in A. thaliana and they are thought to form a flexible network for Fe-S cluster delivery (Couturier et al., 2013). The function of some Fe-S transfer/carrier proteins may be redundant, because knock-out mutants for most genes encoding Fe-S transfer/carrier proteins are viable (Couturier et al., 2013). For instance, plants deficient in GRXS14 or GRXS16 did not display any growth defect, whereas the growth rate of plants lacking both was reduced (Rey et al., 2017). However, some reports showed that Fe-S carrier/transfer proteins also have specificity for a particular Fe-S target protein or group of targets (Balk and Schaedler, 2014). For instance, NFU2 and HCF101 are involved in the maturation of one or several proteins belonging to PSI and some other stromal proteins (Lezhneva et al., 2004; Stöckel and Oelmüller, 2004; Touraine et al., 2004; Yabe et al., 2004). Thus, we presently cannot decide whether these 7 translationally cycling Fe-S carrier/transfer proteins may preferentially function in the biogenesis of specific Fe-S proteins.

Another feature of Fe-S carrier/transfer proteins is that they act in different manners depending on their redox environment. Mutants for some genes encoding Fe-S carrier/transfer proteins only showed lethal phenotypes specifically under oxygenic conditions. The maturation of two Fe-S proteins in E. coli, IspG and IspH, requires a different combination of Fe-S carrier/transfer protein isoforms under different oxygenic conditions. Under aerobic conditions, it uses IscU, IscA and then ErpA for cluster insertion into apo-IspG and apo-IspH, whereas it uses IscU and either ErpA or IscA under anaerobic conditions (Vinella et al., 2009). Another example is cytoplasmic NAR1 in eukaryotes, an Fe-S carrier/transfer protein in CIA

pathway. NAR1 forms a complex with CIA1, AE7, and MET18 to transfer Fe-S clusters from the scaffold protein NBP35 to target apoproteins. Interestingly, nar1 mutants in plants, yeast, and nematodes display phenotypes only under normal oxygen pressure but not under low oxygen pressure (Fujii et al., 2009; Mondy et al., 2014). Therefore, the function of some Fe-S carrier/transfer proteins has a close relationship with the oxygen levels in vivo and in vitro, which might explain the existence of extensive cycling of TL of Fe-S transfer/carrier proteins over the diurnal cycle as observed here.

#### Open Questions and Prospects

Here we report that the TL of subsets of Fe-S assembly and Fe-S utilizing proteins, as well as of other metalloproteins, vary over a diurnal cycle. It was concluded earlier that Fe-S assembly proteins are not diurnally regulated at the transcript level (Balk and Schaedler, 2014). But there is no investigation so far into whether the rates of Fe-S protein biosynthesis are under circadian or diurnal regulation. Knowledge on the regulation of Fe-S protein biogenesis in plants grown under various regimes of temperature and photoperiod would also help us better understand the mechanisms of de novo synthesis or repair of Fe-S clusters (Couturier et al., 2013).

Although much progress was made in elucidating the Fe-S protein biogenesis machinery in plants in the past decades, we are still lacking plant-specific knowledge on the function of a number of predicted Fe-S cluster assembly proteins. Another issue is that the environment-dependent regulation of Fe-S cluster protein biogenesis is largely unknown. Initially, it is important to understand whether and how Fe-S cluster assembly proteins and Fe-S cluster-utilizing proteins are regulated at the transcript, translational and protein levels under various environmental conditions, such as iron and sulfur starvation and other environmental stresses known to enhance internal levels of ROS.

#### CONCLUSION

We assessed the translation state of Fe-S and other metaldependent proteins genome-wide over a diurnal cycle through the re-analysis of a published microarray dataset. Our analysis

#### REFERENCES


suggested that there are notable and complex diel dynamics. Overall, translation state of Fe-S cluster assembly proteins peaked in the middle and at the end of the night (ZT0 and ZT18). Notably, translation state of none of the Fe-S cluster-utilizing proteins peaked during the night, and instead peaked during the day (ZT6 and ZT12). Finally, we observed a trend for peaks in translation state of transcripts encoding proteins requiring different metals at different times of the day. These observations were made only for subsets of proteins of each group. While the biological implications of these observations deserve further study, the diurnal regulation of metalloprotein biogenesis at the level of translational regulation is substantial and should be addressed in a more quantitative and comprehensive study. RNA sequencing-based methods are now available to address this biological question with far higher accuracy (Juntawong et al., 2014).

### AUTHOR CONTRIBUTIONS

HZ performed all data analysis, wrote first draft of manuscript and edited manuscript. UK conceived study and edited manuscript.

### FUNDING

This work was funded by the Deutsche Forshcungsgemeinschaft in SPP1927 Iron-Sulfur for Life.

### ACKNOWLEDGMENTS

We thank Dr. Janneke Balk, JIC Norwich, for repeated sharing of information.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018.01641/ full#supplementary-material



and mitochondrion and is involved in chloroplast Fe-S protein function in Arabidopsis. PLoS One 10:e0141443. doi: 10.1371/journal.pone.0141443


iron-sulfur cluster scaffold protein, is required for biogenesis of ferredoxin and photosystem I. Plant Cell 16, 993–1007. doi: 10.1105/tpc.020511


**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 © 2018 Zhang and Krämer. 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) and the copyright owner(s) 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 Frataxin in Metal Metabolism

#### Diego F. Gomez-Casati\*, Maria V. Busi and Maria A. Pagani

Centro de Estudios Fotosintéticos y Bioquímicos (CEFOBI-CONICET), Universidad Nacional de Rosario, Rosario, Argentina

Frataxin is a highly conserved protein from prokaryotes to eukaryotes. Several functions related to iron metabolism have been postulated for this protein, including Fe-S cluster and heme synthesis, response to oxidative damage and oxidative phosphorylation. In plants, the presence of one or two isoforms of this protein with dual localization in mitochondria and chloroplasts has been reported. Frataxin deficiency affects iron metabolism in both organelles, leading to an impairment of mitochondrial respiration, and chlorophyll and photosynthetic electron transport deficiency in chloroplasts. In addition, plant frataxins can react with Cu2<sup>+</sup> ions and dimerize, which causes the reduction of free Cu ions. This could provide an additional defense mechanism against the oxidation of Fe-S groups by Cu ions. While there is a consensus on the involvement of frataxin in iron homeostasis in most organisms, the interaction of plant frataxins with Cu ions, the presence of different isoforms, and/or the localization in two plant organelles suggest that this protein might have additional functions in vegetal tissues.

#### Edited by:

Felipe Klein Ricachenevsky, Universidade Federal de Santa Maria, Brazil

#### Reviewed by:

Khurram Bashir, RIKEN, Japan Daniel H. Gonzalez, National University of the Littoral, Argentina Joaquin Medina, Centro de Biotecnología y Genómica de Plantas (CBGP), Spain

#### \*Correspondence:

Diego F. Gomez-Casati gomezcasati@cefobi-conicet.gov.ar

#### Specialty section:

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

Received: 30 May 2018 Accepted: 02 November 2018 Published: 21 November 2018

#### Citation:

Gomez-Casati DF, Busi MV and Pagani MA (2018) Plant Frataxin in Metal Metabolism. Front. Plant Sci. 9:1706. doi: 10.3389/fpls.2018.01706 Keywords: frataxin, iron, copper, Fe-S clusters, metal homeostasis

#### IRON FUNCTIONS IN PLANTS, UPTAKE AND DISTRIBUTION

Iron is an essential element for almost all life forms. It is part of cofactors that carry out electron transfer functions, and is involved in chemical transitions (e.g., hydroxylations), hydration and dehydration reactions and radical-mediated rearrangements. Iron also participates in oxygen sensing and transport, and regulation of protein stability (Connorton et al., 2017). Iron is essential for plant growth but, at the same time, is highly reactive and toxic via the Fenton reaction. Thus, plants tightly control iron homeostasis and react to both deficiency and overload of iron. Photosynthetic organisms are distinguished by a high iron requirement for the function of both mitochondria and chloroplasts. These organelles are thought to play a major role in the iron metabolism of the plant cell because this metal/ion serves as an essential cofactor for many enzymes involved in the mitochondrial respiratory chain and electron transfer in the chloroplastic photosynthetic complexes (Morrissey and Guerinot, 2009).

Plants mainly acquire iron from the rhizosphere. Although iron is one of the most abundant metals in the land surface, its availability for plant roots is very low, dependent on the soil reduction potential and pH. In soils that are aerobic or at higher pH, Fe is readily oxidized and is thus predominately in the form of insoluble ferric Fe(III) oxides. At lower pH, the ferric ion is freed from the oxide and becomes more available for uptake. Thirty percent of the world's land for cultivation is too alkaline for optimal plant growth – the most common problem is iron deficiency. Many plant foods like rice, maize, and wheat constitute poor sources of dietary iron (Takahashi et al., 2001; Walker and Connolly, 2008; Morrissey and Guerinot, 2009).

Iron uptake in plants has classically been divided into Strategy I (reducing, dicotyledonous and non-graminaceous monocots) and Strategy II (chelating, graminaceous monocots), the main

Gomez-Casati et al. Frataxin and Metal Metabolism

difference being the oxidation state of the iron (Romheld and Marschner, 1986). In the rhizosphere, iron is mostly found as ferric oxyhydrates of very low solubility. The participation of FRO2 (Ferric Reduction Oxidase 2), IRT1 (Iron-Regulated Transporter 1) and AHA2 (a proton pump located in plasma membrane) has been reported in Strategy I in tomato and Arabidopsis thaliana. Phytosiderophores (plant-derived small organic molecules with a high affinity for iron) and oligopeptide transporters participate in Strategy II in rice, maize and barley (including the iron-siderophore transporters YS1, first characterized in maize, and YSL15 in rice) (Higuchi et al., 1999; Curie et al., 2001; Inoue et al., 2009; Nozoye et al., 2011). Rice is particular among monocots because it uses Strategy II to acquire Fe from rhizosphere but also has the Strategy I-like system (Ishimaru et al., 2006). Under flooded conditions, when Fe(II) is more stable and abundant, rice also absorbs Fe(II) directly via OsIRT1 (Oryza sativa IRT1) and OsNRAMPs (Natural Resistance Associated Macrophage Proteins) (Takahashi et al., 2011). It has been reported that IRT1 has a major role in the regulation of plant iron homeostasis and it is essential for plant growth under iron-limited conditions (Vert et al., 2002). IRT1 undergoes ubiquitin-dependent endocytosis to prevent the uptake of other divalent metals such as Mn, Zn, and Co (Barberon et al., 2014; Dubeaux et al., 2015). In addition, other transporters such as several NRAMPs also participate in the uptake of iron, but as a low-affinity systems (Curie et al., 2000; Castaings et al., 2016).

Several phenylpropanoid-pathway enzymes are upregulated under iron deficiency in Arabidopsis such as PAL1, PAL2 (two Phe ammonia-lyases), 4CL1 and 4CL2 (two 4-coumarate:CoA ligases), the ABC transporter PDR9 (responsible for coumarin secretion into the rhizosphere) and MAT3 (an enzyme that produces S-adenosyl methionine which is involved in coumarin biosynthesis) (Lan et al., 2011; Rodriguez-Celma et al., 2013; Mai et al., 2016). It has been demonstrated that the phenolic compounds secreted facilitate Fe(III) availability for the FRO2 reductase to generate Fe(II) which is transported by IRT1 (Fourcroy et al., 2016). This suggests that these phenolic compounds are important in/for Strategy I iron uptake and are also involved in iron mobilization from insoluble pools (soil) or root apoplast (Fourcroy et al., 2014; Schmid et al., 2014). The catecholic coumarins found at the highest levels in the exudates of iron-deficient wild-type Arabidopsis are esculetin, fraxetin, scopoletin, and sideretin (Fourcroy et al., 2014; Schmid et al., 2014). Fraxetin is the major coumarin exuded into the rhizosphere in response to iron deficiency in alkaline conditions, while sideretin is exuded in acidic conditions. Fraxetin and sideretin are synthesized from scopoletin by hydroxylases [2-ODD (S8H)], which generates fraxetin, and a cytochrome P450 (CYP82C4), which oxidizes fraxetin to generate sideretin) and both compounds efficiently mobilize and reduce insoluble Fe(III), rescuing the chlorotic phenotypes (Rajniak et al., 2018). To date, most efforts in understanding soil iron-uptake limitations have focused on the role of soil pH and have ignored other potentially relevant factors such as interactions with soil organic matter or other metals such as Zn(II) or Mn(II) (Rajniak et al., 2018).

Because of its low solubility and high toxicity, iron must form complexes with chelators to be translocated without causing damage by redox reactions (Dubeaux et al., 2015). One of the complexes necessary for transport by the symplast occurs between the reduced form of iron (Fe2+) and nicotianamine, a non-protein amino acid produced by nicotianamine synthase (Inoue et al., 2003; Klatte et al., 2009; Rellan-Alvarez et al., 2010), which also chelates other divalent cations (e.g., Zn2+). Once iron has passed the endodermis, it is taken into the xylem to reach the shoot. The dominant form of iron in the xylem is Fe3<sup>+</sup> bound to citrate and, therefore, the flow of citrate is essential for iron translocation (Rellan-Alvarez et al., 2010). This translocation is mediated by the efflux transporter FRD3 in Arabidopsis (Green and Rogers, 2004) and its ortholog FRDL1 in rice (Yokosho et al., 2016).

The leaves are the main receptor organ for iron because they are where it is needed for photosynthesis. Here, iron reenters the symplast by the action of FRO proteins (Finazzi et al., 2015; Bashir et al., 2016). Subsequently, it can be remobilized and arrives at other sink organs through the phloem, where it is mainly transported as Fe(II)- nicotianamine complexes. In Arabidopsis, OPT3 (of the oligopeptide transporter family protein) is involved in this process (Mendoza-Cozatl et al., 2014; Zhai et al., 2014). The seed is considered the final destination of iron, where iron reserves are fundamental for germination. The YSL transporters are involved in the iron loading of the seeds (Le Jean et al., 2005).

Mitochondria and chloroplasts represent major iron sinks within cells, as iron is required for the proper functioning of the respiratory chain and photosynthetic protein complexes. The mechanisms by which Fe is obtained by chloroplasts and mitochondria are not as well-known as iron uptake at the root epidermis. There is evidence of reduction-based mechanisms for chloroplast and mitochondria Fe-acquisition, since the presence of FRO7 and FRO3/FRO8 has been reported in each organelle (Jeong et al., 2008). A mitochondrial iron transporter, which belongs to the mitochondrial carrier family (MCF) also present in yeast, zebrafish, humans, and Drosophila, was identified in rice and named MIT (Bashir et al., 2011). The protein PIC1 (Permease In Chloroplasts 1) was the first molecular component involved in plastid Fe-transport identified in Arabidopsis and tobacco (Duy et al., 2007; Gong et al., 2015). In addition, other proteins are supposed to participate in Fe transport across the chloroplast envelope, including two transporters from the yellow stripe 1-like family, YSL4 and YSL6, which have been characterized as potential plastid Fe-efflux transporters in Arabidopsis (Divol et al., 2013). However, in summary, mitochondrial and chloroplast iron transport is till far from being completely deciphered, nor it is understood their cross talk regarding iron homeostasis.

### BIOSYNTHESIS OF IRON COFACTORS: THE PARTICIPATION OF FRATAXIN

Iron is an important component of Fe-S clusters found in many ferrosulfoproteins. Three different systems capable of mediating

Fe-S cluster assembly have been identified: NIF (nitrogen fixation) found in azeotropic bacteria, SUF (mobilization of sulfur) found in archaea, bacteria and plastids, and ISC (ironsulfur cluster) present in bacteria and mitochondria (Romheld and Marschner, 1983; Patzer and Hantke, 1999; Dos Santos et al., 2004; Lill and Muhlenhoff, 2008; Morrissey and Guerinot, 2009; Balk and Pilon, 2011). These systems have in common three steps: (i) the production of sulfur from cysteine, catalyzed by a cysteine desulfurase; (ii) the assembly of Fe-S clusters onto scaffold proteins; and (iii) the transfer of the mature Fe-S cluster into an apoprotein. It has been reported that a small mitochondrial protein, frataxin, could be involved in several steps of this process (Balk and Lobreaux, 2005; Lill and Muhlenhoff, 2008; Lillig and Lill, 2009; Balk and Pilon, 2011; Turowski et al., 2012).

Frataxin is a nuclear-encoded mitochondrial protein whose deficiency is the cause of Friedreich's ataxia, a hereditary cardioand neurodegenerative disease in humans. This protein plays a role in Fe-S cluster biosynthesis, protection against oxidative stress and iron metabolism. Frataxin is highly conserved throughout evolution, being present in humans, plants, flies, worms, and bacteria (Gibson et al., 1996; Babcock et al., 1997; Adinolfi et al., 2002; Busi and Gomez-Casati, 2012; Han et al., 2017). Some hints about frataxin function can be gleaned from the evolutionary record (Morrissey and Guerinot, 2009). The appearance of frataxin in eukaryotes occurred about the time of the endosymbiotic event creating mitochondria from the purple bacterial ancestor, and it was probably acquired by mitochondria together with other components of the ISC operon (Waters et al., 2007; Morrissey and Guerinot, 2009). Different organisms developed mechanisms to avoid the toxicity of free metal ions that allow the control of their uptake, storage and release. It was postulated that frataxin potentially fulfills some of these functions (Busi and Gomez-Casati, 2012).

In 2004, we identified the first frataxin protein in a photosynthetic organism, A. thaliana (Busi et al., 2004, 2006). The functionality of AtFH was assessed by complementation of a yeast frataxin-null mutant, suggesting that AtFH was involved in plant mitochondrial respiration and stress responses (Busi et al., 2004). In agreement with this hypothesis, AtFH-deficient plants presented retarded growth, showed an increment of reactive oxygen species (ROS) and an induction of oxidative stress markers. Interestingly, we also found an increment of aconitase and succinate dehydrogenase-2 (SDH2-1) transcripts, both coding for mitochondrial Fe-S-containing proteins. The reduction of the activities of both enzymes indicates that AtFH also participates in Fe-S cluster assembly or the insertion of Fe-S clusters into apoproteins, possibly in cooperation with other proteins such as Nfs1, HscB, Isd11, and Isu (among others), by the formation of a multiprotein complex (Gonzalez-Cabo et al., 2005; Busi et al., 2006; Maliandi et al., 2007; Shan et al., 2007; Shan and Cortopassi, 2012; Turowski et al., 2012; Leaden et al., 2014; **Figure 1**). Consistent with the essential role of AtFH in cellular function is the observation that homozygous null mutants result in a lethal phenotype (Busi et al., 2004, 2006; Vazzola et al., 2007; Martin et al., 2009). Our data also substantiate the hypothesis that AtFH, apart from its role in protecting bioavailable iron within mitochondria and the biogenesis of Fe-S groups, plays a role in the biosynthesis of heme in plants (Maliandi et al., 2011). We have shown, by in vitro experiments, that AtFH catalyzes the formation of heme when it is incubated with Fe(II) and protoporphyrin IX (Armas et al., 2019). Despite plant ferrochelatases are only present in chloroplasts (Lister et al., 2001; Masuda et al., 2003), a modest but detectable ferrochelatase activity has been observed in plant mitochondria (Cornah et al., 2002), pointing toward frataxin as the enzyme responsible for the reaction in this organelle.

We have reported that in Arabidopsis, AtFH is dualtargeted to mitochondria and chloroplasts (Turowski et al., 2015) and its deficiency alters the normal functioning of chloroplasts by affecting the levels of Fe, chlorophyll, and Fe-S proteins, suggesting that AtFH plays a role as a modulator of both the mitochondrial ISC and chloroplast SUF systems (Turowski et al., 2015; **Figure 1**). In AtFH deficient plants, we also found a reduction of about 40% in ferredoxin levels and about 30% in nitrite reductase (a chloroplastic Fe-Scontaining protein) activity (Turowski et al., 2015). Thus, it is possible that plant frataxins play similar roles in the two organelles, as iron donors, regulating the activity of Fe-S proteins, and, possibly, modulating the activity of the ISC and SUF systems.

Although we identified only one frataxin gene in A. thaliana, other plants could have at least two isoforms (Murgia et al., 2009). Recently, we reported the presence of two functional frataxin isoforms in Zea mays, ZmFH-1 and ZmFH-2, located in both, mitochondria and chloroplasts (Buchensky et al., 2017). The biochemical, biophysical and physiological studies showed some differences between the two isoforms in protection against oxidants, aggregation state and expression patterns. ZmFH-1 showed to be more efficient against oxidative damage and it is expressed in a higher extent in almost all tissues respect to ZmFH-2. Furthermore, ZmFH-2 undergoes some conformational changes when exposed to air, possibly due to a C-terminal extension that could give high thermodynamic stability to the protein in comparison to ZmFH-1 (Buchensky et al., 2017; Sanchez et al., 2018). These results suggest that the two proteins play similar but not identical roles in plant cell metabolism.

In some cases, frataxin assembly seems to be a consequence of iron incorporation into the protein. The assembly of yeast frataxin, for instance, seems to be driven by iron oxidation and accumulation by iron core formation, whereas iron core degradation results in protein disassembly (Adamec et al., 2000; Aloria et al., 2004). Human frataxin assembly has been proposed to be a means of detoxifying redox-active iron. However, in the case of human frataxin, iron does not seem to be the main factor for assembly and it was postulated that the assembly is a physiological property of the protein that allows it to perform diverse cellular functions (O'Neill et al., 2005a,b).

#### FRATAXIN AND COPPER

Cu is a trace element necessary for many different processes in plants. More than 50% of Cu present in plants is found in chloroplasts, which underscores the need for Cu in photosynthesis (Raldugina et al., 2016). The major Cu protein is plastocyanin, which is an essential component of the electron transport chain of photosystem I (Schubert et al., 2002; Kramer and Clemens, 2006). Plant mitochondria also require Cu for the assembly and function of cytochrome c oxidase, the terminal enzyme of the respiratory chain, among other Cu proteins (Garcia et al., 2014). Nevertheless, free copper ions are dangerous inside cells since they can directly attack functional sites in proteins or induce ROS production through Fenton and Haver–Weiss reactions (Halliwell, 2006). As a result, many protective mechanisms exist, such as Cu metallochaperones (Robinson and Winge, 2010), low molecular weight thiol ligands such as glutathione, and the mitochondrial anionic compound known as copper ligand (or CuL) described in yeast (Cobine et al., 2004). Still, Cu is toxic when present in excess, probably because these housekeeping defense mechanisms are overwhelmed.

Copper toxicity has classically been associated with ROS production and oxidative damage. However, more recent evidence shows that damage and inhibition of Fe-S metabolism is the more likely initial manifestation of Cu toxicity (Dupont et al., 2011). Dehydratase family enzymes are rapidly inactivated

upon exposure of Escherichia coli cells to low micromolar copper levels due to the displacement of iron atoms from their solventexposed Fe-S clusters (Macomber et al., 2007). Copper stress in Bacillus subtilis leads to enhanced expression of Fe-S cluster scaffold (SufU) and many Fe-S proteins, as well as iron and sulfur uptake pathways (Chillappagari et al., 2010). Additional work performed in vitro confirms that mammalian ISCA1/2 and GLRX5 Fe-S clusters are destabilized by the presence of Cu(I) (Brancaccio et al., 2017).

De novo Fe-S cluster biogenesis in plants occurs in chloroplasts and mitochondria and as we mentioned above, plant frataxins are located in both organelles (Turowski et al., 2015; Buchensky et al., 2017). We have shown that frataxins can react in vitro with the free cupric ion to give the cuprous form (Sanchez et al., 2018). In this reaction, the conserved cysteine residue of frataxin is oxidized forming a disulfide bridge with another frataxin unit (Buchensky et al., 2017). Moreover, plant frataxins oxidized and in their dimeric form (AtFH and ZmFH2) can bind free Cu(I) ions (Sanchez et al., 2018). Frataxin oxidation and the resulting dimerization might be necessary to bring together enough ligands for the cuprous ion. Cu(I) binding sites are dominated by amino acids with sulfur ligands like cysteine and methionine, but histidine can also bind Cu(I), although less tightly. The cuprous ion is coordinated by 2, 3, or 4 ligands, and proteins involved in copper resistance tend to have a coordination environment of low affinity and high coordination number (usually 4) (Rubino and Franz, 2012). AtFH and ZmFH-2 contain three histidine and two methionine residues (the cysteine residue is oxidized and unable to bind copper). Among the potential plant frataxin Cu(I) ligands, two histidine residues are relatively close (seven and eight amino acids apart in ZmFH-2 and AtFH, respectively). In this manner, a plant frataxin dimer might fulfill the requirements of four lowaffinity ligands for a Cu(I) ion. The reaction observed in vitro might reasonably occur in vivo, with concomitant reduction of free copper ions, hence protecting the Fe-S clusters in plants. In this way, frataxins could comprise an additional defense for Fe-S clusters in the presence of excess copper (**Figure 1**).

#### REFERENCES


There is additional evidence from different organisms that frataxin might be involved in copper metabolism. Friedreich's ataxia patients show altered Cu distribution in the dentate nucleus of the central nervous system (Koeppen et al., 2012), and in the Drosophila model of the disease there is a generalized increase in copper content (Soriano et al., 2016), as there is in mitochondria of yeast frataxin-null mutants (Han et al., 2017). Moreover, in frataxin knockdown flies, treatment with copper chelators improved their impaired motor performance without altering the iron accumulation phenotype, implying a direct role of Cu in the pathophysiology of the disease (Soriano et al., 2016).

Saccharomyces cerevisiae frataxin-null mutants are more sensitive to Cu than wild-type cells (Foury and Cazzalini, 1997; Han et al., 2017) and in anaerobic growth conditions, in which little oxygen is available for ROS generation, copper accumulation and its toxicity increases (Strain and Culotta, 1996; Outten et al., 2001). Ha-Duong and coworkers have shown that yeast frataxin can bind Cu(II) and Cu(I) ions, with higher affinities than iron (Han et al., 2017). Thus, we might assume that yeast frataxin protects Fe-S clusters from copper toxicity by preventing free Cu ions from interacting with them in the same way we have suggested for plant frataxins. Yeast complementation assays with plant frataxins support this hypothesis. Arabidopsis and maize frataxins can completely restore the ability of a S. cerevisiae frataxin-null strain to grow in high copper medium (Sanchez et al., 2018), which led us to infer that plant and yeast frataxins might perform similar and conserved molecular functions in copper metabolism.

### AUTHOR CONTRIBUTIONS

DG-C, MB, and MP collaborated in the writing of the manuscript.

### FUNDING

This work was supported by grants from ANPCyT (PICT 2014- 2184/2016-350/2016-0264). DG-C, MB, and MP are research members of CONICET.

by Yfh1p, a putative homolog of frataxin. Science 276, 1709–1712. doi: 10.1126/ science.276.5319.1709




gene and of the new isolated ferric reductase CsFRO1 and iron transporter CsIRT1 genes in cucumber plants. Plant Physiol. Biochem. 45, 293–301. doi: 10.1016/j.plaphy.2007.03.011


**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 © 2018 Gomez-Casati, Busi and Pagani. 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) and the copyright owner(s) 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 Diverse Iron Distribution in Eudicotyledoneae Seeds: From Arabidopsis to Quinoa

Miguel Angel Ibeas<sup>1</sup> , Susana Grant-Grant<sup>1</sup> , Maria Fernanda Coronas<sup>1</sup> , Joaquín Ignacio Vargas-Pérez<sup>1</sup> , Nathalia Navarro<sup>1</sup> , Isidro Abreu<sup>2</sup> , Hiram Castillo-Michel<sup>3</sup> , Natalia Avalos-Cembrano<sup>1</sup> , Julio Paez Valencia<sup>4</sup> , Fernanda Perez<sup>5</sup> , Manuel González-Guerrero<sup>2</sup> and Hannetz Roschzttardtz<sup>1</sup> \*

<sup>1</sup> Facultad de Ciencias Biológicas, Pontificia Universidad Católica de Chile, Santiago, Chile, <sup>2</sup> Centro de Biotecnología y Genómica de Plantas (UPM-INIA), Universidad Politécnica de Madrid, Madrid, Spain, <sup>3</sup> ID21, European Synchrotron Radiation Facility (ESRF), Grenoble, France, <sup>4</sup> Department of Botany, University of Wisconsin–Madison, Madison, WI, United States, <sup>5</sup> Departamento de Ecología, Pontificia Universidad Católica de Chile, Santiago, Chile

#### Edited by:

Felipe Klein Ricachenevsky, Universidade Federal de Santa Maria, Brazil

#### Reviewed by:

Louis Grillet, Academia Sinica, Taiwan Paloma Koprovski Menguer, John Innes Centre (JIC), United Kingdom

#### \*Correspondence:

Hannetz Roschzttardtz hroschzttardtz@bio.puc.cl

#### Specialty section:

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

Received: 30 August 2018 Accepted: 20 December 2018 Published: 15 January 2019

#### Citation:

Ibeas MA, Grant-Grant S, Coronas MF, Vargas-Pérez JI, Navarro N, Abreu I, Castillo-Michel H, Avalos-Cembrano N, Paez Valencia J, Perez F, González-Guerrero M and Roschzttardtz H (2019) The Diverse Iron Distribution in Eudicotyledoneae Seeds: From Arabidopsis to Quinoa. Front. Plant Sci. 9:1985. doi: 10.3389/fpls.2018.01985 Seeds accumulate iron during embryo maturation stages of embryogenesis. Using Arabidopsis thaliana as model plant, it has been described that mature embryos accumulate iron within a specific cell layer, the endodermis. This distribution pattern was conserved in most of the analyzed members from Brassicales, with the exception of the basal Vasconcellea pubescens that also showed elevated amounts of iron in cortex cells. To determine whether the V. pubescens iron distribution was indicative of a wider pattern in non-Brassicales Eudicotyledoneae, we studied iron distribution pattern in different embryos belonging to plant species from different Orders from Eudicotyledoneae and one basal from Magnoliidae. The results obtained indicate that iron distribution in A. thaliana embryo is an extreme case of apomorphic character found in Brassicales, not-extensive to the rest of Eudicotyledoneae.

Keywords: apomorphy, Arabidopsis, quinoa, embryo, iron, phylogeny

### INTRODUCTION

Increased iron content in seeds is an important agronomic trait. This is due to the relevance of this element in seed production (Marschner, 2005; Roschzttardtz et al., 2011), embryo development, and seedling germination and growth (Lanquar et al., 2005; Kim et al., 2006), as well as in human nutrition (Murgia et al., 2012). In spite of this essential role, the prevalent low iron bioavailability in the soil of most of the main agricultural areas of the world, limits plant productivity, fertility, and even germination rates (Guerinot and Yi, 1994). Consequently, a substantial effort has been dedicated to unraveling the molecular bases controlling iron homeostasis, how plants incorporate and distribute iron throughout their organs, and how iron is stored in the seeds (Walker and Waters, 2011; Kobayashi and Nishizawa, 2012; Grillet et al., 2014; Flis et al., 2016; Tsai and Schmidt, 2017).

Iron distribution in seeds has been studied in monocots and Eudicotyledoneae. Using rice and wheat as models it was concluded that most of the iron in monocot plants is stored in the aleurone layer (Iwai et al., 2012; De Brier et al., 2016). In Eudicotyledoneae, the majority of work has been carried out in Arabidopsis. It has been estimated that approximately 50% of the seed iron content

is stored in endodermal cells (Roschzttardtz et al., 2009; Schnell Ramos et al., 2013). Results from experiments using multiple iron-imaging methods showed that the vast majority of the seed embryo iron is located in vacuoles (Kim et al., 2006; Roschzttardtz et al., 2009; Schnell Ramos et al., 2013). This is in contrast to other plant tissues in which iron has also been detected in plastids, associated with ferritin, and in nucleoli (Roschzttardtz et al., 2011, 2013).

Iron accumulates steadily in the endodermal vacuole during the maturation stage of embryo development, and is subsequently used during post-germinative growth (Lanquar et al., 2005; Roschzttardtz et al., 2009). The defects shown by the vit1 and nramp3 nramp4 mutants under iron deficiency conditions suggest that proper iron storage seems essential for seed germination and post-germinative growth (Lanquar et al., 2005; Kim et al., 2006; Mary et al., 2015). These mutant plants are defective in iron loading into the vacuoles of endodermal cells during embryo maturation, a process mediated by the transporter AtVIT1 (VACUOLAR IRON TRANSPORTER1) (Kim et al., 2006); or in iron recovery from vacuoles in these cells via the AtNRAMP3 and AtNRAMP4 (NATURAL RESISTANCE ASSOCIATED MACROPHAGE PROTEIN3 and 4) transporters during germination (Lanquar et al., 2005).

Closely related Brassicaceae embryos accumulate iron very similarly to Arabidopsis, i.e., in vacuoles of cells surrounding the embryo provasculature (Ibeas et al., 2017). However, reports on another Eudicotyledoneae, the legume Phaseolus vulgaris, showed broader iron localization. While iron hotspots were detected in cotyledons in a pattern resembling the provasculature, the metal was also evenly distributed in the rest of the organ (Cvitanich et al., 2010). This alternative way of accumulating iron in the seed could indicate that the storage of this nutrient in Eudicotyledoneae is more diverse than what has been reported to date. Considering the importance of Eudicotyledoneae as crops, and the need to improve the iron content in the edible parts (DeFries et al., 2015), we set to study iron distribution in embryos from a wide range of orders belonging to dicotyledonous plants. Our results show that the iron distribution pattern in the Arabidopsis embryo is not extended to all Eudicotyledoneae, it rather seems to be a derived character only observed in the Brassicaceae family. We also found that Caryophyllales has a distinct iron distribution pattern and that iron loading during Chenopodium quinoa seed development is different than what we had previously observed in Brassicaceae (Ibeas et al., 2017), suggesting there might be different strategies for the storage of iron in these seeds.

### MATERIALS AND METHODS

#### Plant Material and Growth Conditions

Most of the seeds used in this study were purchased at a local market or harvested in public gardens. Some seeds were collected in the field and identified before being deposited in the herbarium at the Departamento de Ecología, Pontificia Universidad Católica de Chile. The list of species used in this study is in **Supplementary** **Table S1**. C. quinoa plants were grown on soil in a greenhouse at 23◦C under long-day condition (16-h/8-h day/night cycle).

### Histochemical Staining of Iron With Perls/DAB Method

Seed embryos from the species indicated in **Supplementary Table S1** or C. quinoa seeds at different developmental stages were isolated and fixed with 2% w/v paraformaldehyde in 1 mM phosphate buffer pH 7.0 for 45 min. The following steps were performed according to Roschzttardtz et al. (2009) and Ibeas et al. (2017).

#### Synchrotron µ-XRF

Quinoa seeds were embedded in OCT (Optimal Cutting Temperature) resin and plunge-frozen in isopentane chilled in liquid nitrogen. Longitudinal sections (30 µm) were prepared using a cryomicrotome (Leica, RM2265-LN22) at –50◦C and immediately imaged at beamline ID21 at the ESRF (European Synchrotron Radiation Facility, Grenoble, France). Elemental distribution was mapped by micro X-ray fluorescence (µXRF) under cryogenic conditions (Cotte et al., 2017). The beam was focused with the use of KB mirrors to a size of 0.5 × 0.9 µm<sup>2</sup> (V × H). The fluorescence signal was detected using an 80 mm<sup>2</sup> active area SGX Si drift detector with a Be window. Two photodiodes were used to measure the incident and transmitted beam intensities. Scans were acquired with an energy of 7.2 keV (Fe k-edge) with a dwell time of 100 ms per pixel, and a pixel size of 2 µm × 2 µm. RGB color maps and Fe, Mn, and P elemental maps were created using PyMca software (Soléa et al., 2007).

### Phylogenetic Analysis

A phylogenetic tree of the species in the data set was assembled using the phylogeny of angiosperms at family level, from the Angiosperm Phylogeny Group (The Angiosperm Phylogeny Group et al., 2016). To resolve relationships within Brassicaceae we followed Guo et al. (2017), which uses 77 protein coding regions. Branch lengths were set to 1.0. Species were assigned to four categories according to Fe location in seed embryos: endodermis; inner layers of cortex, external cortex or protodermis. Ancestral states of iron location were reconstructed using Mesquite (Maddison and Maddison, 2007) based on one−parameter model.

### RESULTS

### Differences in Iron Distribution in Embryos of the Order Brassicales

Analyses on species from the Brassicaceae family showed that their embryos have similar subcellular iron distribution in cotyledons and hypocotyl (Ibeas et al., 2017). In Arabidopsis thaliana, Camelina sativa, Nasturtium officinale, Lepidium sativum, and Brassica napus dry seed embryos, iron accumulates in the vacuoles of cells surrounding the provasculature. Some differences in the number of cells that store iron were observed in the hypocotyl: while Arabidopsis only presented a one cell layer,

the other species have two or more (Roschzttardtz et al., 2009; Ibeas et al., 2017).

We extended this study to members of other families within the Brassicales order. We performed Perls/DAB staining on Cleome hassleriana (Cleomaceae Family) and Capparis spinosa (Capparaceae Family). Embryo sections revealed that iron accumulates in cells that surround the provasculature in a similar manner as embryos from Brassicaceae species (**Figure 1**). Some differences were observed in C. spinosa, where a group of cells between provasculature regions in cotyledons showed iron accumulation in vacuoles (**Supplementary Figure S1**). Surprisingly, when analyzing the member of the Caricaceae family V. pubescens (a more ancestral Brassicales), we observed that several cortex cells accumulate iron. This pattern sets it apart from the rest of Brassicales embryos studied to date and suggests that Brassicaceae embryo cortex cells may also accumulate iron (**Figure 1**).

### Iron Distribution in Embryos From Different Orders of Eudicotyledoneae

To inquire whether the observation that V. pubescencs has a differential iron distribution is an exception or an ancestral trait lost at some point in the Brassicales order, Perls/DAB analyses were carried out in embryos from six representative Eudicotyledoneae Orders (Sapindales, Rosales, Zygophyllales, Solanales, Asterales, and Caryophyllales) and one basal specie from Magnoliidae (Magnoliales). Interestingly, the analysis of embryos from those orders showed that iron accumulates in several cell layers including endodermis and cortex cells. In some cases, iron was also detected in the protodermis (Solanales, Asterales, and Magnoliales; **Figure 2**). These results indicate that the Arabidopsis embryo has an uncommon iron distribution compared to embryos of seed plants from other Orders of the Eudicotyledoneae class. Moreover, a more detailed analysis of Porlieria chilensis Perls/DAB-stained section shows that iron also seems to accumulate extracellularly, in what might be described as the cortex apoplast (**Figure 2**).

### Iron Distribution in Caryophyllales Embryos

In order to evaluate iron distribution in species from other orders, seed embryos of several species belonging to Caryophyllales were used to determine whether iron distribution might differ within the same order. **Figure 3A** shows embryos from Fagopyrum esculentum, C. quinoa and Phytolacca dioica. In all cases, embryos were isolated and fixed, and then thin sections were analyzed, in particular hypocotyl and cotyledon. Using Perls/DAB staining we found that in all the species analyzed, iron was present in several embryo cell types, including endodermis and cortex cells. This result indicates that this pattern of iron distribution

is conserved within this order (**Figure 3A**). We also analyzed embryos from Spinacia oleracea, Beta vulgaris, and Rheum rhabarbarum (**Supplementary Table S1**).

Phylogeny Group et al. (2016). Bars in panels correspond to 100 µm.

Next, we used synchrotron-based micro X-ray fluorescence (SµXRF) to independently confirm iron distribution using Perls/DAB staining in C. quinoa seeds. Dry seeds were sectioned and analyzed by S-µXRF, showing iron accumulation in several cell layers in hypocotyl but excluding provasculature (**Figure 3B**). S-µXRF also showed Mn accumulation in several cell layers of C. quinoa hypocotyl (probably cortex) in contrast with the restricted localization of Mn in the subepidermal layer of plant model A. thaliana (Kim et al., 2006; Punshon et al., 2012; Schnell Ramos et al., 2013; Eroglu et al., 2017).

### Iron Distribution During Chenopodium quinoa Seed Development

Previously, our group reported for the first time that some Brassicaceae species accumulate iron in two embryo cell types (endodermis and one cell layer of cortex). We also showed that in early stages of B. napus seed development, iron accumulates in nuclei of the free cell endosperm and in all embryo cell types. Later, iron is relocated to cytoplasmic structures, and finally, at the mature stage, iron is accumulated in vacuoles of the endodermis and cortex (Ibeas et al., 2017).

Because the iron distribution pattern in Caryophyllales is different from the pattern described above, we used C. quinoa as a model to study iron localization during seed development. C. quinoa is an emerging crop that has potential health benefits and an exceptional nutritional value. We analyzed four different stages of seed maturation in C. quinoa: early stages (between 3 and 7 days postanthesis), an intermediate stage (14 days postanthesis/cotyledon stage) and a late stage (21 days postanthesis/mature stage). In order to describe the seed structures where iron accumulates before being loaded into the embryo, we used whole seeds including seed coat, perisperm and embryo in our analyses.

Analysis of seed sections containing early embryo developmental stages revealed that there are no detectable iron pools in seed coat, perisperm or embryos (**Figure 4**). In seeds containing embryos at the cotyledon stage iron is detected inside the nuclei and in structures surrounding the nuclei (**Figures 5A,B**). Interestingly, longitudinal sections of the entire seed showed detectable iron pools in the integuments. Strong iron staining was observed in cytoplasmic structures (**Figures 5C,D**). In order to confirm the subcellular compartments that accumulate iron we performed Perls/DAB and Toluidine blue staining on the same sections. The zoom of specific cells in **Figures 5B,D**, show that iron localizes in nuclei and cytoplasmic structures in the embryo. Moreover, there is strong iron staining in cytoplasmic structures in the integuments.

In mature seeds, iron is no longer localized in nuclei (**Figure 6**). At this stage, iron is detected in several embryonic cell types, including endodermis and cortex cells of cotyledons and hypocotyl (**Figures 6B,D**). Perls/DAB and Toluidine blue staining show that iron pools do not localize in nuclei in cotyledon and hypocotyl cells (Zoom in **Figures 6B,D**).

#### DISCUSSION

Dietary iron deficiency is a major issue for human health, affecting more than 2 billion people around the world (Murgia et al., 2012). Seeds are a pivotal source of iron for humans and animals (Roschzttardtz et al., 2017). Therefore, understanding the molecular mechanisms involved in iron loading and distribution in seeds is critical for the development of biotechnological approaches to improve seed iron content. Generation of biofortified crops could be a financially sound strategy to deliver nutrients to the population, as an alternative to the continuous governmental investment in fortification and supplementation programs (DeFries et al., 2015). Thus, genetic biofortification of staple crops could result in an environmentally friendly and cost-effective strategy for the improvement of nutritional health (Guerinot and Yi, 1994).

Pioneer studies, using SµXRF and a histochemical method for iron detection in plant tissues (Perls/DAB stain), showed that iron accumulates in the vacuoles of the endodermis cell layer of A. thaliana mature embryos (Kim et al., 2006; Roschzttardtz et al., 2009; Eroglu et al., 2017; Ibeas et al., 2017). While this distribution pattern is conserved in the rest of the Brassicaceae family (Ibeas et al., 2017), it is not present in legumes (Cvitanich et al., 2010). These observations suggest that embryos may have multiple ways to store iron. Consequently, by exploring this diversity, we should be able to propose suitable models to dissect the molecular mechanisms of seed iron storage and eventually translate these findings to crops of commercial and nutritional value. Toward this end, we used Perls/DAB staining to determine iron distribution in several plant species along the Eudicotyledoneae phylogeny. As a result, we obtained the first phylogenetic study of iron distribution in plant seed embryos. The results indicate that seed embryos belonging to different Eudicotyledoneae orders accumulate iron in several cell layers, with more diversity that what can be

inferred by simply using Arabidopsis as model. In particular, cortex cells accumulate iron in all the embryo species analyzed except for Brassicales families Capparaceae, Cleomaceae and Brassicaceae. These results strongly suggest that iron distribution in cortex cells in embryos from Eudicotyledoneae embryo plants is a conserved characteristic from a phylogenetic point of view. We used S-µXRF as a second and independent method to confirm that in quinoa embryos, iron accumulates in several cell layers including the cortex and endodermis cells (**Figure 3B**).

Quinoa seed has high nutritional value and has recently started to be used as a novel functional food (Abugoch, 2009). This plant contains 15 mg of iron per 100 g of seeds, covering the daily iron needs of infants and adults (National Academy Press, 2001; Konishi et al., 2004). We evaluated iron accumulation during C. quinoa seed development and found that it does not accumulate iron in any seed tissue during the early developmental stages (**Figure 4**). This is in contrast with what was previously described for B. napus seeds, in which strong iron staining was observed in the nuclei of the free cell endosperm surrounding the embryo, and also in several cell layers of the embryo at the torpedo stage (Ibeas et al., 2017). This discovery opens new questions on the mechanisms of iron distribution and accumulation in these seeds, and will be used as the basis for the development of biotechnological strategies to increase total iron content in seeds for human consumption.

A side finding of this work is the broad distribution of Mn in the embryo of C. quinoa hypocotyl (**Figure 3B**), compared to the restricted location of Mn in the hypocotyl subepidermal tissue of A. thaliana embryos (Kim et al., 2006; Punshon et al., 2012; Schnell Ramos et al., 2013; Eroglu et al., 2017). It has recently been shown that mutations in the genes cax1cax3 (Punshon et al., 2012), mtp8 and vit1 (Eroglu et al., 2017) cause a broad distribution of Mn indicating that they are responsible for this localization. As proposed for iron, this finding will open new strategies for the development of Mn fortified seeds.

In addition to Brassicales, our analyses also showed that at least some Zygophyllales, such as P. chilensis, also have an unusual iron distribution. In this case, iron is distributed inside and outside the cells (apoplast). This iron pool would necessarily be accumulated in a different manner from the vacuolar-located. For instance, there would be no need for particular metal transporters as is the case for Arabidopsis (Lanquar et al., 2005; Kim et al., 2006). It would also be interesting to determine the

FIGURE 5 | Iron distribution in C. quinoa seeds at cotyledon stage. C. quinoa seeds at the cotyledon stage (14 days postanthesis) dissected from fruits were embedded in Technovit resin, sectioned (3 µm) and then stained with Perls/DAB (B,D). In B the zoom shows iron accumulation into and around the nuclei. In B and D toluidine blue was used to show different cell structures. Unstained sections were used as control (A,C). The scale bar represents 100 µm.

postanthesis), were collected from fruits, embedded in Technovit resin and sectioned (3 µm). The sections were then stained with Perls/DAB (B,D). Cotyledons and hypocotyl are shown in A,B and C,D, respectively. The zoom in B and D show iron accumulation in cell structures different from nuclei. Perls/DAB and Toluidine blue were used in order to reveal different cell structures. Unstained sections were used as control (A,C). The scale bar represents 100 µm.

FIGURE 7 | A global view of iron distribution in embryos. Phylogenetic tree showing all the species analyzed in this study. Colors of circles indicate type of cells that accumulate iron in the hypocotyl. White, only endodermis; light gray, endodermis + internal cortex (C<sup>I</sup> ); dark gray, endodermis + internal cortex + external cortex (CE); black, endodermis + internal cortex + external cortex + protodermis (PD). Cortex cell are defined as the cell layers that are between the endodermis and the protodermis, we refer as internal cortex the cell layers than are near to the endodermis. Ratio of number of cells where iron is accumulated versus total number of cell layers from the endodermis to the protodermis for embryos from different orders are indicated in bars. Colors of bars correspond to the same code described above.

iron ligands, and how assimilable it is.Further analyses of this and other Zygophyllales species will be carried out to ascertain the extension of this phenotype and to determine its molecular bases.

Regarding the widespread use of Arabidopsis as a Eudicotyledoneae model plant for iron nutrition, our results show that iron distribution in Arabidopsis embryos is not a widely conserved character in Eudicotyledoneae seed embryos and it corresponds to an apomorphic character. Ancestral reconstruction of the phylogeny of angiosperms with Maximum Parsimony or Maximum Likelihood indicated that the embryo of the ancestor of Brassicales, and probably of all Eudicotyledoneae, accumulates iron in several layers of the cortex (**Figure 7**). The loss of iron in the outer layers of the cortex evolved in the core of Brassicales, whereas the ability to accumulate iron only in endodermis is a derived trait of Arabidopsis. According to Beilstein et al. (2010), the apomorphy described in our study emerged between 120 and 70 Mya. In our opinion, Brassicales seeds could be used to evaluate molecular mechanisms involved in iron accumulation in the cortex cells, using information obtained from other species described in this article.

These results also open new questions on the mechanisms involved in iron accumulation and remobilization in cortex and protodermis cells, during embryogenesis and seed germination, respectively. From an evolutionary point of view, it will be interesting to study which competitive advantage determined the selection of this iron distribution in Brassicales species.

#### AUTHOR CONTRIBUTIONS

fpls-09-01985 January 14, 2019 Time: 16:25 # 9

MI, SG-G, JV-P, NN, and HR performed the embryo dissections and Perls/DAB staining. IA, HC-M, and MG-G performed the SµXRF analysis. FP performed the phylogenetic analysis. All authors participated in the writing of the manuscript.

#### FUNDING

This work was funded by FONDECYT 1160334 (Chilean Government) to HR. Ph.D. students' work was supported by

### REFERENCES


Conicyt-Chile grants 21160350 (to MI), 21170951 (to SG-G), and 21151344 (to JV-P).

#### ACKNOWLEDGMENTS

The authors are greatly indebted to Dr. Xavier Jordana, María Isabel Gómez and Dr. Juan Keymer for its continued support and encouragement. Vasconcellea pubescens seeds were kindly provided by Dr. Maria José Letelier and Dr. Sebastián Godoy.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018.01985/ full#supplementary-material



Walker, E., and Waters, B. (2011). The role of transition metal homeostasis in plant seed development. Curr. Opin. Plant Biol. 14, 318–324. doi: 10.1016/j.pbi.2011. 03.025

**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 © 2019 Ibeas, Grant-Grant, Coronas, Vargas-Pérez, Navarro, Abreu, Castillo-Michel, Avalos-Cembrano, Paez Valencia, Perez, González-Guerrero and Roschzttardtz. 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) and the copyright owner(s) 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.