Microbial Flora Associated with the Halophyte–Salsola imbricate and Its Biotechnical Potential

Halophytes are associated with the intertidal forest ecosystem of Saudi Arabia and seemingly have an immense potential for yielding useful and important natural products. In this study we have aimed to isolate and characterize the endophytic and rhizospheric bacterial communities from the halophyte, Salsola imbricata, In addition these bacterial strains were identified and selected strains were further studied for bioactive secondary metabolites. At least 168 rhizspheric and endophytic bacteria were isolated and of these 22 were active antagonists against the oomycetous fungal plant pathogens, Phytophthora capsici and Pythium ultimum. Active cultures were mainly identified with molecular techniques (16S r DNA) and this revealed 95.7–100% sequence similarities with relevant type strains. These microorgansims were grouped into four major classes: Actinobacteria, Firmicutes, β-Proteobacteria, and γ-Proteobacteria. Production of fungal cell wall lytic enzymes was detected mostly in members of Actinobacteria and Firmicutes. PCR screening for type I polyketide synthases (PKS-I), type II polyketide synthases (PKS-II) and nonribosomal peptide synthetases (NRPS) revealed 13 of the 22 strains (59%) were positive for at least one of these important biosynthetic genes that are known to be involved in the synthesis of important antibiotics. Four bacterial strains of Actinobacteria with potential antagonistic activity including two rhizobacteria, EA52 (Nocardiopsis sp.), EA58 (Pseudonocardia sp.) and two endophytic bacteria Streptomyces sp. (EA65) and Streptomyces sp. (EA67) were selected for secondary metabolite analyses using LC-MS. As a result, the presence of different bioactive compounds in the culture extracts was detected some of which are already reported for their diverse biological activities including antibiotics such as Sulfamethoxypyridazine, Sulfamerazine, and Dimetridazole. In conclusion, this study provides an insight into antagonistic bacterial population especially the Actinobacteria from S. imbricata, producing antifungal metabolites of medical significance and characterized taxonomically in future.

Halophytes are associated with the intertidal forest ecosystem of Saudi Arabia and seemingly have an immense potential for yielding useful and important natural products. In this study we have aimed to isolate and characterize the endophytic and rhizospheric bacterial communities from the halophyte, Salsola imbricata, In addition these bacterial strains were identified and selected strains were further studied for bioactive secondary metabolites. At least 168 rhizspheric and endophytic bacteria were isolated and of these 22 were active antagonists against the oomycetous fungal plant pathogens, Phytophthora capsici and Pythium ultimum. Active cultures were mainly identified with molecular techniques (16S r DNA) and this revealed 95.7-100% sequence similarities with relevant type strains. These microorgansims were grouped into four major classes: Actinobacteria, Firmicutes, β-Proteobacteria, and γ -Proteobacteria. Production of fungal cell wall lytic enzymes was detected mostly in members of Actinobacteria and Firmicutes. PCR screening for type I polyketide synthases (PKS-I), type II polyketide synthases (PKS-II) and nonribosomal peptide synthetases (NRPS) revealed 13 of the 22 strains (59%) were positive for at least one of these important biosynthetic genes that are known to be involved in the synthesis of important antibiotics. Four bacterial strains of Actinobacteria with potential antagonistic activity including two rhizobacteria, EA52 (Nocardiopsis sp.), EA58 (Pseudonocardia sp.) and two endophytic bacteria Streptomyces sp. (EA65) and Streptomyces sp. (EA67) were selected for secondary metabolite analyses using LC-MS. As a result, the presence of different bioactive compounds in the culture extracts was detected some of which are already reported for their diverse biological activities including antibiotics such as Sulfamethoxypyridazine, Sulfamerazine, and Dimetridazole. In conclusion, this study provides an insight into antagonistic bacterial population especially the Actinobacteria from S. imbricata, producing antifungal metabolites of medical significance and characterized taxonomically in future.
Keywords: halophyte, antifungal activity, enzyme production, 16S rRNA gene sequence, LC/MS analyses INTRODUCTION Coastal dune plants consist of shrubs that are associated with tropical and subtropical coastal areas. This ecosystem is considered as a low nutrient environment due to the sparseness of nitrogen and phosphorus (Frosini et al., 2012). Coastal dune areas of the world usually represent a unique flora as a result of variations in salinity, anaerobicity, and temperature fluctuations during different seasons. All of these conditions make this habitat unusual and thus may have been selective for the presence of truly unique microorganisms having a rich biodiversity. Under these conditions microorganisms associated with these costal dune plants play a role in their survival and perform diverse biological functions in plants (Del Vecchio et al., 2015).
Bacteria associated with Coastal dune plants are responsible for various activities including resistance against different pathogens by producing different bioactive metabolites and plant growth promotion compounds. Bacteria are associated with these plants either as rhizospheric or endophytic forms and use nutrients from plants for their growth (de Melo Pereira et al., 2012). Endophytic bacteria get access to internal parts of plants by producing enzymes to hydrolyze cell wall of plants (Cho et al., 2007). These hydrolytic enzymes are also involved in defense of plants against various bacterial and fungal pathogens (Luo et al., 2013).
Endophytes produce metabolites that possess antimicrobial activities (Strobel, 2003). Due to their beneficial effects and broad range, endophytes have been isolated from plants widely ranging from the poles of the earth to the steaming tropics. Plants provide these microorganisms, particularly endophytic bacteria and fungi, a nutrient-rich environment for their growth and development (Strobel et al., 2004;de Melo Pereira et al., 2012). Endophytic Actinobacteria associated with plant roots or leaves are a source of novel potential bioactive substances. Actinobacteria produce various secondary metabolites including antibiotics, antitumor, immunosuppressive, and enzymes widely used in industrial, agriculture, and pharmaceutical industry (Manivasagan et al., 2014). Recently, various antimicrobial compounds and enzymes have been isolated from marine Actinobacteria (Hong et al., 2009;He et al., 2012). Secondary metabolite salinosporamide A, isolated from a marine Actinobacteria is currently in clinical trials as a cancer treatment (Feling et al., 2003). Several other bioactive compounds have been isolated from different marine Actinobacteria (Fu et al., 2012).
The Genus Salsola is comprised of different halophyte species including Salsola imbricata that inhabit the coastal area of the Thuwal region in Jeddah, Saudi Arabia. Recent studies on S. imbricata reported the presence of triterpenesaponins and a new isorhamnetin from the methanolic extract of this plant (Osman et al., 2016). In a recent study from Saudi Arabia, phytochemical analysis of S. imbricata showed presence of bioactive biphenylpropanoids i.e., biphenylsalsonoid A and B. These two bioactive compounds exhibited antioxidant and antibacterial activity against human pathogenic bacteria (Oueslati et al., 2017). Although the coastal dune environment is an unexplored source of natural compounds with diverse activities there has been little or no work on the bacterial flora of this plant. Based on previous studies, it would appear that certain microbial flora components may be present in this plant that has biotechnical potential. Thus, this study was undertaken to explore potential antifungal microflora of halophyte that may produce products of medicinal, agricultural, and industrial significance.

Collection of Plant Samples and Isolation of Rhizobacteria and Endophytic Bacteria
The halophyte, S. imbricata was collected from the coastal area of Jeddah, Saudi Arabia (22 • 15 ′ 54 ′′ North, 39 • 6 ′ 44 ′′ East). We used soil, roots and leaves of plant sample for the isolation of both rhizopheric and endophytic bacteria and the sample plant were placed in sterile plastic bag at the time of collection. For isolation of bacteria from soil associated with roots they were dipped in sterile distilled water and serial dilutions were made (10 −3 , 10 −4 , and 10 −5 ) in filtered autoclaved sea water (FAS) Laboratories, Detroit, MI) in 1000 ml filtered Sea water for bacterial culturing. For isolation of endophytic bacteria, roots, and leaves were surface sterilized. Both roots and leaves were washed several times with tap water to remove adhering soil particles and then sterilized by washing with disinfectants as described previously (Bibi et al., 2012). After sterilization, small pieces of sterilized root and leaf segments were ground in FAS using a sterile mortar with pestle. Aliquots were further serially diluted and plated in triplicate on four different media mentioned above. To suppress fungal growth, 50 µg/ml cycloheximide was mixed to the medium before pouring. The plates were incubated at 28 • C for 2 weeks for bacterial growth.

Isolation of Rhizopheric and Endophytic Actinobacteria
For the isolation of rhizosphere and endophytic Actinobacteria, the same dilutions from soil, roots, and leaves were inoculated onto starch-casein agar (Himedia) in sea water supplemented with cycloheximide and nystatin (50 µg/ml). Further, plates were incubated at 28 • C for up to 10-12 days. By checking morphological characteristics, Actinobacteria were selected and subcultured using the same isolation medium. Individual colonies were streaked to check purity of the strains and stored in 15% (v/v) glycerol. All bacterial strains were labeled according to acquisition numbers of the King Abdulaziz University special infectious agents unit and stored at −70 C.

Screening of Bacteria for Antifungal Activity
All isolated bacteria were screened for their antifungal activity. We used five different fungal pathogens i.e., Phytophthora capsici (P. capsici), Pythium ultimum (Py. ultimum), (obtained in this laboratory from Capsicum annuum and Cucumis sativus respectively), Magnaporthe grisea (KACC 40415), Altenaria malli (KCTC 6972) and Fusarium moniliforme (KCTC 6149) were obtained from Korean Agricultural Culture Collection (KACC) and Korean type culture collection center (KCTC). Antifungal activity was determined by using the cross streak method (Bibi et al., 2012). All isolates were streaked on PDA medium supplemented with ½ R2A in sea water. Each 6 mm mycelial disc of freshly grown fungal pathogen was placed in the center of plate perpendicular to streak of isolates at 4 cm distance from edges of plate and incubated for 4-6 days at 28 • C. Antagonistic strains were checked in duplicate for antagonistic activity evaluated by measuring the inhibition zone of fungal pathogens around each bacterial streak.

Bacterial DNA Extraction and 16S rRNA Gene Sequencing
Genomic DNA was extracted from the selected antagonistic bacterial isolates using a DNA extraction kit (Thermo Scientific, Waltham, USA). To identify antagonistic bacteria, 16S rRNA gene sequencing was performed. Approximately 1500 bp fragment of the 16S rRNA gene was amplified using bacterial universal primers 27F (5 ′ -AGAGTTTGATCCTGGCTCAG-3 ′ ) and 1492R (5 ′ -GGTTACCTTGTTACGACTT-3 ′ ), the 16S rRNA gene fragment was amplified under the following PCR conditions: one cycle of 95 • C for 5 min followed by 28 cycles of 95 • C for 1 min, and annealing at 58 • C for 50 s with extension at 72 • C for 50s, and a final extension step at 72 • C for 10 min (Bibi et al., 2012). PCR products were purified using PCR purification kit (Thermo Scientific, Waltham, USA), and sequenced commercially (Macrogen, South Korea). Antagonistic bacteria were identified by blast searches of their 16S rRNA gene sequences obtained using the EzTaxon server (https:// www.ezbiocloud.net) (Kim et al., 2012) to identify antagonistic bacteria. Phylogenetic positions of the antagonistic bacteria were confirmed using CLUSTALX (Thompson et al., 1997) multiple alignments of the bacterial sequences and BioEdit software (Hall, 1999) was used to edit the gaps. The neighbor-joining method in the MEGA6 Programme with bootstrap values based on 1,000 replications was used for construction of the phylogenetic tree based on the 16S rRNA gene sequences (Tamura et al., 2013).

Evaluation of Hydrolytic Enzymatic Activity
Antagonistic bacterial isolates were further evaluated for their lytic enzyme production. Protease activity was checked using skim milk ½ R2A agar plates. Bacteria exhibiting protease production made clear zone on skim milk agar plates. Amylase production was checked on starch medium. Starch hydrolysis was seen as a clear zone on agar plates by positive strains (Kumar et al., 2012). For lipase activity, tributyrin ½ R2A agar medium was used. After 48 h of incubation at 28 • C a clear zone was detected around bacterial colonies after hydrolysis of tributyrin. To check cellulase activity, CMC agar (carboxy methyl cellulose agar) medium was used. Bacteria were streaked on plates with the respective enzyme medium and incubated at 28 • C for 2 days. Congo red (0.1%) solution was added to plates and put on orbital shaker for 30 min and then rinsed with 1MNaCl (Hendricks et al., 1995). Cellulase activity was observed as a clear zone on CMC agar plates around bacterial colonies.

Optimization of Bacterial Culture Conditions for Production Bioactive Compounds
To optimize culture conditions of selected bacterial strains four different media i.e., ½ R2A broth, ½ TSB, ½NB in sea water and Marine broth in distilled were used for culturing. Bacterial cultures were grown for 72 h and after every 24 h optical density (OD) was checked and antifungal activity was assessed against P. capsici, Py. ultimum using the disc diffusion method. The effect of temperature was optimized within the range of temperatures (20-40 • C) in ½ R2A broth. For pH optimization, different ranges of pH values (5-12) were used for the growth and antifungal compound production in ½ R2A broth.

LC-MS Analysis of Bacterial Culture
In order to assess the actual presence of antimicrobial products in bacterial cultures that had been previously screened and shown to be biologically active we performed LC-MS analyses of the bacterial supernatants. A Bacterial culture (5 ml) was placed on −80 • C for 5 min, and then transfer to 37 • C water bath for 5 min and this procedure was repeated 5 times. It was centrifuged at 15,000 g for 10 min and transfer 3 ml supernatant fluid was transferred to a tube. To it was added 12 ml of acetonitrile and vortexed for 30 s. It was centrifuged again at 15,000 g for 10 min and 300 µl supernatant was taken for LC-MS metabolomic analysis. The injection volume was 3 µl and samples were analyzed on an Agilent 6540B TOF/Q-TOF Mass Spectrometer coupled with an Agilent 1290 UPLC and Dual AJS ESI ion source. An ACQUITY UPLC HSS T3 (100 × 2.1mm, 1.8 um) column and pre-column (Phenomenex Security Guard TM ) was used to separate the sample. The column temperature was set to 45 • C and flow rate was 0.5 ml/min. The acquisition range was from 50 m/z to 1500 m/z and the scan rate was 1.

Nucleotide Sequence Numbers
Nucleotide sequences of the antagonistic bacteria isolated from the halophyte have been deposited in the GenBank database under accession numbers KY436424-KY436445.

Isolation of Rhizospheric and Endophytic Bacteria
In this study, rhizospheric and endophytic bacteria were isolated from the halophyte, S. imbricata growing in a selected site on the Saudi Arabian peninsula. A total of 168 rhizospheric and endophytic bacteria were isolated from soil, roots, and leaves samples. Various methods were used to acquire these bacteria including the use of four different culturing media. It turns out that both high and low nutrient medium content favored the growth of different groups of bacteria. It has been noted that the ½ R2A in SW is a favorable medium for the isolation of bacteria from the marine environment. Thus, all isolated bacteria culture on different media were sub-cultured on ½ R2A in SW. In total 168 bacteria were isolated from different parts of halophyte and soil as well (

Screening of Bacteria against Pathogenic Fungi
All isolated bacteria were tested for their antagonism against plant pathogenic oomycetes. Of the 168 bacteria isolated, 22 (13%) displayed inhibition to both oomycete plant pathogens that was the established bioassay test. The percentage of antagonistic bacteria varied in different parts of halophyte, being highest in roots (n = 9; 5.3%), followed by soil (n = 8; 4.7%), and leaves (n = 5; 2.9%). These antagonistic bacterial isolates were tested further against three other fungal pathogens. Different groups of antagonistic bacteria were identified from S. imbricata, and Actinobacteria was the dominant phylum (  Table 2). Some of the antagonistic isolates (EA56, EA62, EA68-EA70) showed inhibition against oomycetes but did not show inhibition against other pathogenic fungi used ( Table 2).
Frontiers in Microbiology | www.frontiersin.org the antagonistic isolates belonged to the genus Streptomyces (Table 1). Using 16S rRNA gene data, a neighbor Joining (NJ) phylogenetic tree was constructed (Figure 1). High bootstrap values were recovered in phylogenetic tree. A separate cluster has been generated for isolates of class Actinobacteria. Antagonistic strains in this cluster were identified with high bootstraps values (51-100%). Representatives of this class belong to four different genera i.e., Nocardioides, Nocardiopsis, Arthrobacter, and Streptomyces. One strain of β-Proteobacteria was identified with separate cluster also showing high bootstrap value (100%). One strain of γ -Proteobacteria made a distinct cluster with closely related type strain Pseudomonas sp. (EA54) with high a bootstrap value of 100 %. Antagonistic bacterial strains of class Firmicutes were placed in separate cluster belonging to the genus Bacillus. In this study, two antagonistic endophytic bacterial strains were also identified showing low 16S rRNA gene sequence similarity (<98%). One strain of Actinobacteria i.e., Streptomyces sp. (EA64) and one strain belonging to Firmicutes i.e., Bacillus sp. (EA68) were identified with similarities of <98% to the closest type strains ( Table 3). Four antagonistic Actinobacteria were selected for secondary metabolites identification depending on their antagonistic activity against oomycetes.

Enzymatic Activities of Antagonistic Bacteria
Antagonistic bacteria were evaluated for their ability to produce fungal cell wall lytic enzymes. Protease, amylase, lipase, and cellulase activities of these bacteria were examined ( Table 2). The number of bacteria exhibiting protease activity (n = 9; 41%) was high as compared to other enzymatic activities. Most of protease producing bacteria were endophytes (n = 6) and recovered from leaves and roots of halophyte. Amylase activity was observed for 7 (32%) antagonistic bacteria. Mostly strains of Actinobacteria showed amylase activity. Endophytic strains of Actinobacteria exhibited high production of amylase. Lipase activity was observed in 7 (32%) antagonistic bacteria. Strains of Actinobacteria and Firmicutes showed high production of protease. Cellulase activity was observed in only 1 (4.5%) bacterial strain. A strain of N. dassonvillei (EA52) exhibited cellulase activity belonged to Actinobacteria.

Detection of PKS and NRPS Genes
Bacteria possess NRPS/PKS gene clusters especially common in the phyla Proteobacteria, Actinobacteria and Firmicutes. Antagonistic bacterial isolates from this study showed that of FIGURE 1 | Phylogenetic distribution of antagonistic bacteria isolated from S.imbricata on the basis of 16S rRNA gene sequences obtained from bacteria and closely related sequences of the type strains of other species. The phylogenetic relationships were inferred from the 16S rRNA gene sequences (1 kb) by using the neighbor-joining method from distances computed with the Jukes-Cantor algorithm. Bootstrap values (1,000 replicates) are shown next to the branches. GenBank accession numbers for each sequence are shown in parentheses. Bar, 0.01 accumulated changes per nucleotide. Isolates selected for bioactive metabolites identification are highlighted.

Identification of Strain EA52, EA58 (Rhizobacteria), and EA65, EA67 (Endobacteria) Metabolites by LC-MS
On the basis of high inhibition activity against fungal pathogens we have selected four antagonistic bacterial strains from our study to identify secondary metabolites. These four antagonistic isolates belong to class Actinobacteria i.e., Nocardiopsis sp. (EA52) and Pseudonocardia sp. (EA58) were rhizobacteria while Streptomyces sp. (EA65) and Streptomyces sp. (EA67) are endophytes. Production of antifungal compounds was influenced by culturing in R2A as culture medium using optimum culture conditions. The maximum antifungal activity was observed after 48 h of growth at 28 • C with pH 7.5. Metabolites in cultures extract of selected bacteria were analyzed using the LC-MS. LC-MS technique confirms presence of various bioactive compounds although not novel but already known for their bioactivity. Identification of metabolites was determined by LC-MS analysis and comparing results from NIST database. LC-MS analysis of culture extract of these selected strains identified different bioactive metabolites ( Table 4). Strain of Nocardiopsis sp. (EA52) produced 7 identified peaks of different bioactive metabolites in both the positive-and negative-ion mode (Figures 2A,B). These compounds include Thiabendazole, Sulfamethoxypyridazine, N-Nitrosodiethylamine, Ibuprofen, Laberalol, Isoxsuprine, and Oxibendazole. For Pseudonocardia carboxydivorans (EA58), 2 bioactive compounds have been detected among hundreds of peaks known for other different metabolites. These three compounds include Sulfamethoxypyridazine and Dimetridazole (Figures 2C,D). Strain EA65 genetically closely associated as a Streptomyces diastatochromogenes showed presence of nine different bioactive compounds in the culture extract i.e., Sulfamethoxypyridazine, Dimetridazole, Salbutero, Moxidectin, Atenolol, Timolol, Benzydamine, Acebutolol, and Bambuterol (Figures 2E,F). These bioactive secondary metabolites are already known for their biological activities. LC-MS analysis of strain EA67, Streptomyces europaeiscabiei showed the presence of six bioactive compounds among several compounds detected. These bioactive compounds include, Sulfamethoxypyridazine, Benzamide, Sulfamerazine, Dimetridazole, Metronidazole-oh, and Carazolol (Figures 2G,H). These compounds are already known for antimicrobial, antiphytopathogenic and biocontrol properties.

DISCUSSION
There is a need for the discovery of new drugs due to the emergence of resistant bacteria to different antibiotics and cancer cells to anticancer drugs. Currently, marine habitats have been identified as a source of discovering many important novel therapeutic agents (Malve, 2016). S.imbricata, is a halophyte containing various chemical compounds of medicinal use and it is widely spread in the sea shore areas in Saudi Arabia. Generally halophytes with traditional health benefits in folk medicine are a source of attraction for modern comprehensive microbial and chemical investigations. In our previous study (Bibi et al., 2017) we have isolated two rhizobacteria from Salsola exhibiting cellulase and amylase activity on culturing media. One strain EA151 with cellulase activity belonged to Firmicutes while other strain EA152 exhibited amylase activity closely related to type strain of Actinobacteria.
In this study, rhizopheric and endophytic bacterial populations with potent antagonistic activity against phytopathogens were studied. One hundred and sixty eight rhizopheric and endophytic bacteria were isolated and only 22 were found to be potent antagonists of oomycetes. Different cultivation media were used to culture bacteria to isolate novel isolates with unique bioactivities. A study has reported that high nutrient medium favors growth of diverse groups of bacteria, while low nutrient media increase number of bacteria (Vishnivetskaya et al., 2000). To our knowledge, this work is the first study on the isolation, screening, characterization and metabolites identification of both rhizospheric and endophytic bacteria from S. imbricata. In this study, culture-dependent approach was used to study the diversity of antagonistic bacteria from halophyte against oomycetes plant pathogens. Screening of 168 rhizospheric and endophytic bacteria and their 16S rRNA gene sequencing resulted in 22 different bacteria belong to four major classes. Bioactive compounds production by rhizo and endophytic bacteria is one of the phenomenon used by bacteria to defend the host against different pathogens. Antagonistic bacteria belong to four major classes, Firmicutes, γ -Proteobacteria, and β-Proteobacteria. Our results show that Actinobacteria is recognized as a dominant phylum among these antagonitistic bacteria isolated from halophyte ( Table 1). Actinobacteria are distributed in various ecological habitats especially in terrestrial and aquatic ecosystems where they play a potential role in decomposition of plants and animals remains. They are biotechnologically important due to their ability to produce secondary metabolites (Berdy, 2005). Therefore, we have selected four different strains of Actinobacteria for further study of bioactive metabolites produced by these antagonistic rhizopheric and endophytic bacteria. Results of 16S rRNA sequence analyses showed a diversity of different genera of Actinobacteria and the genus Streptomyces remain which are dominant in our study. Results of this study are also compatible with previous studies where many novel Actinobacteria have been recovered from different unexplored marine sources (Fiedler et al., 2005;Bull and Stach, 2007). Halophytes growing near the sea shore in tropical and subtropical areas have become valuable sources for the isolation of microbial flora especially in the group Actinobacteria because of the chemodiversity of environmental factors prevailing there. The presence of major numbers of Streptomyces strains amongst antagonists from the halophyte in this study are in accordance with the results reported previously (Eccleston et al., 2008;Ravikumar et al., 2012). In Actinibacteria, the most active bacterial strains in addition to Streptomyces are, Nocardiopsis, Pseudonocardia, and Arthrobacter which have been reported for their inhibitory activities and production of antimicrobial metabolites (Anandan et al., 2016).
The metabolic diversity of Streptomycetes is remarkable and novel strains with new bioactive compounds including antibiotics have been isolated from extreme habitats (Fiedler et al., 2005). Until now, several studies have reported diverse groups of antimicrobial compounds by bacteria isolated from halophytic endophytes, such as steroids, peptides, alkaloids, terpenoids, quinines, flavonoids, and phenols (Newman and Cragg, 2007). Our results showed that most of the endophytes in this study belonging to Streptomyces and were recovered from root samples. This finding is supported by previous results where maximum number of endophytes were recovered from roots specimens (Taechowisan et al., 2003;Passari et al., 2015) This may be due to the reason that the rhizosphere has a high concentration of nutrients and various bacteria enter different root tissues of the host plant through the rhizosphere and become endophytic in their mode of life (Rosenblueth and Martinez-Romero, 2006).
The second dominant class of bacteria in this study was Firmicutes. The group of bacteria belonging to the class Firmicutes are commonly found as they are easily culturable. Also, from marine sources, species of Bacillus are dominant for the production of antibacterial, antifungal antibiotics, surfactants, and lytic enzymes (Mondol et al., 2013). Endophytic bacterial strains of Bacillus from halophytes are already reported to show inhibition against bacterial and fungal plant pathogens (Menpara and Chanda, 2013). In this study, only two antagonistic rhizopheric and endophytic bacteria were related to the class Proteobacteria belonging to two different classes γ -Proteobacteria and β-Proteobacteria. Members of class γ -Protobacteria from marine source produce the highest number of secondary metabolites (Long and Azam, 2001).
Hydrolytic enzymes from marine bacteria are produced and used by them in catalyzing different biochemical processes in the marine environment (Thatoi et al., 2013). In addition to production of useful secondary metabolites and antibiotics Actinobacteria from marine sources are also known for the production of lytic enzymes of industrial use (Anandan et al., 2016). In this study, mostly strains of Actinobacteria and Firmicutes were active candidates for production of different enzymes. Protease, lipase and amylase activity was mostly observed ( Table 2). Only one strain EA59 exhibited protease, amylase and lipase activity. This strain is endophytic and belongs to Actinobacteria. It seems as if the productions of these hydrolytic enzymes is necessary for antagonism as well as for intracellular colonization of bacteria in the host plant. Only four strain of rhizobacteria belong to Actinobacteria were able to produce at least one type of enzyme tested in this study. While most of the endophytic bacteria were able to show enzymatic activities belong to Actinobacteria. Marine Actinobacteria have the potential to produce extracellular enzymes of industrials use. These extracellular enzymes are usually used by endophytic bacteria for their antagonism against different pathogens invading host plants (Zhao et al., 2016).
Polyketide synthetase (PKS) and nonribosomal polyketide synthetase (NRPS) genes are considered as hallmarks for the production of secondary metabolites. These biosynthetic genes of bacterial secondary metabolites are involved in the biosynthesis of many bioactive products. The presence of the biosynthetic genes PKS and NRPS in rhizospheric and endophytic bacteria associated with this halophyte suggests that these beneficial bacteria are dominant producers of bioactive natural products. Amplification of these three biosynthetic gene clusters domains is an indirect approach to test the knowledge that bioactive products are being produced by the organism. Of 22 isolates screened for detection of PKS-I, PKS-II, and NRPS gene, only 13 strains were positive for at least one of the biosynthetic genes tested. Several species of Streptomyces possess two types of PKSs. One is responsible for antibiotic synthesis, while the other is involved in spore pigments production in various Streptomyces species (Bergh and Uhlén, 1992). In this study, PKS-II and NRPS were detected in most of the Streptomyces. Among them, strains EA52, EA61, EA63, and EA64 were positive for the presence of NRPS and PKS-II, and also exhibited antifungal activity against tested pathogens ( Table 2). While strains EA53, EA56, EA59, EA60, EA62, EA68-EA71 were negative for the presence of PKS-I, PKS-II, and NRPS genes and positive for antifungal activity. Therefore, there is no correlation between presence of PKS and NRPS gene and antimicrobial properties (Qin et al., 2009). It is evident that NRPS PKS-I and PKS-II gene screening using PCR in different bacterial strain help to find out potential strain of bacteria producing important secondary metabolites.
Strains of Streptomycetes are active producers of antimicrobial compounds and more than 70% of antibiotics in the world are produced by them. Endophytic Streptomycetes occupy a biologically important niche in tissues of the host plant, taking nutrition from host and in turn providing protection to the plant. These endophytes produce important metabolites that are not toxic to their host therefore, can be considered as important bioactive metabolites in drug discovery (Castillo et al., 2007). Selection of Actinobacteria for identification of active secondary metabolites resulted in a number of known synthetic bioactive compounds not of microbial origin (Table 4). Based upon antifungal activity and low 16S rRNA gene similarity four out of twenty-two rhizopheric and endophytic strains were ultimately selected for LC-MS analyses. Production of antifungal compounds was influenced by culturing the organism in R2A as a culture medium using optimum culture conditions. The maximum antifungal activity was observed after 48 h of growth with pH 7.5 at 28 • C. Metabolites in cultures extracts of selected bacteria were analyzed using LC-MS. Two rhizopheric strains namely EA52 closely related to Nocardiopsis sp. and EA58 showed similarity with Pseudonocardia sp. in producing diverse metabolites including Thiabendazole, Sulfamethoxypyridazine, N-Nitrosodiethylamine, Ibuprofen, Laberalol, Isoxsuprine, Oxibendazole, and Dimetridazole. Strain EA52 showed close similarity to Nocardiopsis sp. This marine non-Streptomycete genus of Actinobacteria, contained a variety of bioactive compounds and showed a wide range of activities including cytotoxicity, antibacterial, antifungal, and anti-angiogenesis. Nocardiopsis sp. contained α-pyrone (nocapyrone S1) and (4-aminophenyl) acetic acid, N-(2-hydroxyphenyl)-acetamide, cyclo-(L-Pro-L-Val), cyclo-(L-Pro-L-Leu), and cyclo-(L-Pro-L-Ile) when evaluated for secondary metabolite production (Zou et al., 2017) but different secondary metabolites have been detected in closely related strain EA58 in the present study (Figures 2A,B). Antimicrobial activity of strains of Pseudonocardia is already evident from previous studies but no such antifungal activity and metabolites detected in our study were produced from the closely related strain EA58 (Figures 2C,D; Tanvir et al., 2016).
Two endophytic Streptomycetes EA65 and EA67 were analyzed by LC-MS and they exhibited spectra for 15 bioactive compounds including antibacterial, antifungal, antiprotozoal and anthelmintic compounds. Strain EA65 showed close similarity at the 97% level to a rhizobacterium S. diastatochromogenes, a promising source of antifungal metabolites and novel natural antibiotics and used as biocontrol agents (Shentu et al., 2016). S. diastatochromogenes reportedly produces polyketomycin, momofulvenoneA, azdimycin, toyocamycin, concanamycin, and oligomycins. Strain EA65 exhibited the presence of bioactive compounds not reported in S. diastatochromogenes including Sulfamethoxypyridazine, Dimetridazole, Salbutero, Moxidectin, Atenolol, Timolol, Benzydamine, Acebutolol, and Bambuterol (Figures 2E,F). Another endophytic strain EA67 related to S. europaeiscabiei showed the presence of bioactive compounds such as Sulfamethoxypyridazine, Benzamide, Sulfamerazine, Dimetridazole, Metronidazole-oh, and Carazolol (Figures 2G,H). These compounds are already known for antimicrobial, antiphytopathogenic and biocontrol properties. Sulfamethoxypyridazine, Nitroimidazole including Dimetridazole and Metronidazole-oh and Moxidectin are major antimicrobial compounds detected in all four strains in this study. Sulfamethoxypyridazine was detected in all four strains in our study. These antibacterial sulfonamides are synthetic antimicrobial agents used as antibiotics in different inflammatory diseases (Vicente and Pérez-Trallero, 2010). This is first study showing the isolation and screening of antagonistic bacteria from S. imbricata and the identification of active metabolites from crude extracts. Compounds produced by these strains although not new, have been not reported in literature from other organisms and most of them are synthetic so this is the first reported of these Actinobacteria as a natural source for production of these antimicrobial compounds.
In conclusions, present study of antagonistic bacteria from S. imbricata exhibited wide range antifungal and enzymatic activities indicating their industrial, biotechnological and agricultural potential. Further analyses of bioactive metabolites from selected Actinobacteria exhibited production of known antibiotics and bioactive compounds of synthetic nature not reported from bacteria before. These results suggest that halophytes are potential source of bioactive metabolites via their associated bacterial microflora. This study points to an enormous potential of discovering novel bioactive compounds using the methodology described. Currently we are working on some selected novel strains for deep bioactive analysis and genome sequencing that will be presented in our future work.

AUTHOR CONTRIBUTIONS
FB and GS: designed the study and write manuscript; MN and EA: identified the plant samples and help in collection of sample and experimental work; MY and AK collect and analyzed the data. All authors read and approved the final manuscript.