Antimicrobial biosynthetic potential and diversity of culturable soil actinobacteria from forest ecosystems of Northeast India

Actinobacteria is a goldmine for the discovery of abundant secondary metabolites with diverse biological activities. This study explores antimicrobial biosynthetic potential and diversity of actinobacteria from Pobitora Wildlife Sanctuary and Kaziranga National Park of Assam, India, lying in the Indo-Burma mega-biodiversity hotspot. A total of 107 actinobacteria were isolated, of which 77 exhibited significant antagonistic activity. 24 isolates tested positive for at least one of the polyketide synthase type I, polyketide synthase type II or non-ribosomal peptide synthase genes within their genome. Their secondary metabolite pathway products were predicted to be involved in the production of ansamycin, benzoisochromanequinone, streptogramin using DoBISCUIT database. Molecular identification indicated that these actinobacteria predominantly belonged to genus Streptomyces, followed by Nocardia and Kribbella. 4 strains, viz. Streptomyces sp. PB-79 (GenBank accession no. KU901725; 1313 bp), Streptomyces sp. Kz-28 (GenBank accession no. KY000534; 1378 bp), Streptomyces sp. Kz-32 (GenBank accession no. KY000536; 1377 bp) and Streptomyces sp. Kz-67 (GenBank accession no. KY000540; 1383 bp) showed ~89.5% similarity to the nearest type strain in EzTaxon database and may be considered novel. Streptomyces sp. Kz-24 (GenBank accession no. KY000533; 1367 bp) showed only 96.2% sequence similarity to S. malaysiensis and exhibited minimum inhibitory concentration of 0.024 µg/mL against methicilin resistant Staphylococcus aureus ATCC 43300 and Candida albicans MTCC 227. This study establishes that actinobacteria isolated from the poorly explored Indo-Burma mega-biodiversity hotspot may be an extremely rich reservoir for production of biologically active compounds for human welfare.

compounds have to be found. This necessity has led to the search for new bioactive compounds among poorly explored habitats which can efficiently target these life-threatening pathogens 12 . Hence, actinobacteria may be a potential solution to these problems. In order to find new microbial products derived from newly identified activities of actinobacteria, the search has to be shifted from routinely explored ecological niches to unexplored ones 13,14 .
Northeast India, a part of the Indo-Burma mega-biodiversity hotspot, is well known for its rich biodiversity. The region has diverse climatic, edaphic and altitudinal variations resulting in wide ecological habitats. Northeast India is the connection between the Indian, Indo-Malayan and Indo-Chinese bio-geographic regions. It is home to a wide spectrum of India's flora and fauna 15 . Unlike the floral and faunal diversity, the microbial diversity is also relatively unexplored in this part of the world. The local environment may influence the evolution of novel secondary metabolic pathways in organisms found in the biodiversity hotspots 16 . Forests are considered to be the most bio-diversified terrestrial ecosystems on earth. They are the largest possible resource available to obtain novel microorganisms and their valuable natural products 17,18 . Quite a few researchers have reported actinobacteria from the forest ecosystems of Northeast India for the search of natural products endowed with antimicrobial activity [19][20][21][22][23][24][25][26] . However, the literature review reveals that Pobitora Wildlife Sanctuary, located in Assam, India is yet an unexplored forest ecosystem and we had previously reported regarding the bioactivity prospective of its microflora 24 . Thus, this forest ecosystem can be considered as an unexplored source of actinobacteria producing bioactive metabolites. Regarding the study of actinobacteria from Kaziranga National Park of Assam, this forest has been previously explored by few researchers and as such, holds a lot of promise for isolation of actinobacteria having pharmaceutical potential 19,20 .
In the light of the above studies, the present investigation was undertaken with an aim to isolate actinobacteria from Pobitora Wildlife Sanctuary (26 °12′ to 26 °16′N and 91 °58′ to 92 °05′E) and Kaziranga National Park (26 °30′ to 26 °45′N; 93 °08′ to 93 °36′E) of Assam, India and screening them against an array of microbial pathogens responsible for human pathogenesis. Emphasis was given for screening of the isolates for the presence of antibiotic biosynthetic genes and the different chemical classes of antimicrobial compounds they produce were also predicted using DoBISCUIT. Analysis of genetic diversity of the actinobacteria isolates was carried out by 16S rDNA-ARDRA. An attempt was made to study a promising actinobacteria exhibiting potent antimicrobial potential against a wide range of microorganisms.

Results
isolation of actinobacteria. A total of 107 presumptive actinobacteria of different phenotypes were isolated from different environmental sites of Pobitora Wildlife Sanctuary (n = 54) and Kaziranga National Park (n = 53) of Assam, India. The details of the nature of soil samples, its pH and the number of actinobacteria isolated are given in Table 1. They were associated with scanty to profuse sporulating capacity and showed the presence of distinctive colonial morphology, mycelia colour and pigment production (See Supplementary Table S1). The white colour series was found to be the most dominant one (n = 48; 44.8%). The light microscopy results showed the spiral chain morphology of the aerial mycelia (See Supplementary Fig. S1).
Detection and analysis of PKS-I, PKS-II and NRPS genes for prediction of chemical classes. All the 77 antagonistic actinobacteria were evaluated for their biosynthetic potential in terms of natural product drug discovery. 24 isolates indicated the presence of at least one of the PKS-I, PKS-II or NRPS genes. PKS-I genes were detected in 6 isolates, PKS-II in 20 isolates and NRPS genes were detected in 2 isolates. The partial gene sequences of PKS-I, PKS-II and NRPS were deposited in GenBank under the following accession numbers KY073865-KY073869, KY235144-KY235162, KU721842, KU721843, KY271082 and KY274457 (Table 3).
These genes were translated to amino acid sequences and the secondary metabolite pathway products were identified using DoBISCUIT database. The genes of all the isolates showed similarities to the phylum actinobacteria at the amino acid level. PKS-I sequences shared 56-68% similarity with their closest matches at the amino acid level in BLASTP search. The amino acid sequences of PB-10, PB-32, PB-47, PB-52, PB-64 and Kz-24 matched with the KS domain and are closely related to the gene product involved in concanamycin A, tautomycin, chlorothricin, nanchangmycin, oligomycin and rifamycin production respectively. PKS-II sequences of 20 actinobacteria shared 68-95% similarity with their closest matches at the amino acid level. Their closest match pathway products were predicted as actinorhodin, naphthocyclinone, jadomycin B, landomycin. These compounds belonged to diverse polyketide groups such as glycosides, anthracycline, naphthoquinone, angucycline, benzoisochromanequinone. Highest sequence identity (95%) at amino acid level was observed in PB-66 which was most closely involved in the synthesis of urdamycin. Adenylation domain of NRPS genes was screened to be positive in PB-52 and PB-64. At the amino acid level, NRPS fragments derived from PB-52 and PB-64 showed 40-53% similarity to their closest relatives. The closest match pathway products of PB-52 and PB-64 isolates were analyzed as virginiamycin and oxazolomycin belonging to streptogramin and polyene-type alkaloid group of compounds respectively (Table 3). Interestingly, PB-64 harbors all the three antibiotic biosynthetic genes, i.e. PKS-I, PKS-II and NRPS. This isolate had the ability to inhibit all the test microorganisms during in vitro antimicrobial screening.
Amplified Ribosomal DNA Restriction Analysis (ARDRA). The restriction digestion profile of the 77 antagonistic actinobacteria were analysed by ARDRA fingerprinting analysis using HinfI restriction endonuclease. Digestion with Hinf1 showed different restriction banding patterns and the dendrogram was constructed. Fragments which were smaller than 100 bp in size could not be reproducibly visualized and so not considered. Critical analysis of the dendrogram obtained from Unweighted Pair Group Method with Arithmetic Mean (UPGMA) revealed three major clusters consisting of Streptomyces sp. Nocardia sp. and Kribbella sp. Overall analysis of the soil samples (Sample 1-4 and A-E) from the two forest ecosystems, indicated Streptomyces as the dominant genus in these habitats (Fig. 1).
Kz-24 was selected based on its promising antimicrobial potential against test microorganisms and the partial 16S rDNA sequence (1367 bp) was deposited to NCBI GenBank with accession number KY000533. The strain indicated maximum 16S rDNA similarity (96.2%) with S. malaysiensis NBRC 16446 (AB249918). The phylogenetic tree also showed the closest similarity to S. malaysiensis based on maximum likelihood method (Fig. 2).

Minimum inhibitory concentration (MIC) of EA-Kz-24. Broth dilution method was used to determine
MIC of EA-Kz-24 ranging from 100-0.024 μg/mL against all test microorganisms (Table 4). EA-Kz-24 exhibited lowest MIC against MRSA ATCC 43300 and C. albicans MTCC 227 (0.024 μg/mL) whereas highest was recorded against S. marcescens MTCC 97 (50 μg/mL). According to Clinical and Laboratory Standards Institute (CLSI) recommendations for MIC, S. marcescens MTCC 97 were found to be resistant to EA-Kz-24 (MIC: 50 μg/mL), as ≤8 μg/mL was taken as susceptible, ≤16 μg/mL as intermediate and ≥32 μg/mL as resistant. SEM analysis. SEM was performed for the assessment of antibacterial and anti-candidal activity of EA-Kz-24 against MRSA ATCC 43300 and C. albicans MTCC 227. SEM indicated significant morphological changes in the cells of the test microorganisms including cell deformity and shrinkage leading to prominent loss and integrity of cell shape after treatment with 1 × MIC EA-Kz-24. 10% DMSO-treated control cells appeared smooth with intact cell surface (Fig. 3).

GC-MS analysis.
Chemical composition of EA-Kz-24 was characterized with GC-MS. Based on the retention time, molecular weight and molecular formula, thirteen chemical compounds were identified by comparing their mass spectra with the NIST library as shown in Table 5. The peak area of the compound is directly proportional to its antimicrobial metabolite quantity (See Supplementary Fig. S4).

Discussion
Actinobacteria have been recognized as the most proliferant producers of natural bioactive compounds like antimicrobials, wide range of enzymes and valuable secondary metabolites with incredible diversity of biological activities [27][28][29] . Our primary goal was to study the functionality of culturable actinobacteria from the Indo-Burma mega-biodiversity hotspot with key emphasis towards understanding their biosynthetic potential.
In total, 107 presumptive actinobacteria were isolated from nine varied soil samples collected from Pobitora Wildlife Sanctuary and Kaziranga National Park of Assam, India. Forest is considered as a wealthy biological diversity with millions of animals, plants and microorganisms 30,31 . Soil samples are known to be rich in organic  www.nature.com/scientificreports www.nature.com/scientificreports/ matter and microorganisms where most of the biological activities occur. A total of 34 actinobacteria were isolated from leaf litter soil and tree rhizosphere soil each, 26 from sediment soil and only 13 isolates from grass rhizosphere soil. The pH range of the soil samples was found to be in the range 4.5-6.0 implying to acidic soil pH of the forest ecosystems. Forest soils are previously reported to be of low pH 32 . Actinomycetes isolation agar, Streptomyces agar and GLM agar were used for actinobacteria isolation which was reported previously 12,19 . Maximum numbers of actinobacteria (n = 52; 49%) were recovered from Actinomycetes isolation agar as this media contains glycerol and asparagine which most actinobacteria use as a source of carbon and nitrogen respectively. Additionally, it also contains sodium propionate which makes it most suitable for the isolation of actinobacteria because it acts as an antifungal agent 19 . The use of amending the isolation media with amphotericin B and rifampicin has been confirmed to be a good strategy for promotion of growth of slow growing actinobacteria in the absence of fast growing contaminants 19 Table 3. Amino acid sequence similarities of the PKS-I, PKS-II and NRPS genes of the actinobacteria and predicted chemical classes for functional genes. Morphological differentiation of all the 107 actinobacteria was done based on their colony morphology, aerial and substrate mycelium colour and diffusible pigment production. The chromogenicity of aerial mycelium is a significant character for grouping of actinobacteria 35 . Furthermore, 38 actinobacteria produced diffusible pigment which is considered as a characteristic feature for identification and classification of Streptomyces 36 .
Actinobacteria have been screened from diverse habitats for the search of novel bioactive compounds from last few years [37][38][39] . Yet there are very few reports regarding the exploration of actinobacteria from forest ecosystems for production of bioactive compounds. In the preliminary screening by spot inoculation method, 77 actinobacteria (72%) were found to be potential antagonists against at least one of the test microorganisms. A total of 51   www.nature.com/scientificreports www.nature.com/scientificreports/ antibiotic compounds display antimicrobial activity against Gram-positive bacteria, only about 1.5% of those are effective against Gram-negative bacteria. There are limited effective antifungal agents for treating life-threatening fungal infections, which is a major challenge to the pharmaceutical industry 40 . In the present study, the result of significant antimicrobial activity against Gram-negative bacteria and fungus strongly suggests that forest ecosystems of Assam are a good source of antagonistic actinobacteria exhibiting promising antimicrobial activity.
Out of 77 antagonistic actinobacteria, 63, 56, 53, 59 and 58 number of isolates produced amylase, cellulase, protease, lipase and esterase enzymes respectively. Interestingly, 24 isolates produced all the five enzymes. Ramesh and Mathivanan 41 and Meena et al. 42 isolated actinobacteria from marine sources with multi-enzyme activity. As per the findings of Ramesh and Mathivanan 41 , they isolated actinobacteria from marine sources and found that majority of the isolates produced lipase which plays an important role in the degradation of polymers in oceans for adaptation in the extreme environment. However, in this study, 56 actinobacteria isolated from the forest ecosystems produced cellulase. The population of cellulase producing actinobacteria is reasonably high in forest floors and soil for the purpose of soil cycling and decomposition of tough plant materials and woody stems. These actinobacteria are largely responsible for the breakdown of large biopolymers like cellulose, hemicellulose, lignin and chitin 32,43,44 .
Biosynthetic gene clusters PKS-I, PKS-II and NRPS play a fundamental role in the biosynthesis of microbial natural products 45 . Out of the 77 actinobacteria exhibiting antimicrobial activity, 24 isolates tested positive for at least one of these biosynthetic genes. This result indicated that 31% of the isolates with antimicrobial potential from the forest ecosystems possessed one of the biosynthetic genes. A total of 6 isolates (8%) were found  www.nature.com/scientificreports www.nature.com/scientificreports/ positive for the presence of PKS-I genes while 20 isolates (26%) were positive for PKS-II genes and NRPS genes were detected in only 2 isolates (3%). These findings were in concurrence with the reports of Lee et al. 46 , who isolated actinobacteria from mangrove forest soil in Malaysia and reported that PKS-II genes were found to be most frequent among the actinobacteria endowed with antimicrobial activity in forest ecosystems, followed by PKS-I and NRPS genes. Lack of amplification of biosynthetic gene sequences in some of the isolates may be due to absence of these genes in their genome 47,48 . The absence of PKS and NRPS genes does not detriment the antagonistic activity of the isolates, signifying that additional biosynthetic mechanisms or types of bioactive agents may be involved in the production of antimicrobial activity 49 . Genus Streptomyces and Nocardia are previously reported as recognized producers of polyketides and nonribosomal peptides 24,38 . Earlier reports confirm that Streptomyces strains possess multiple copies of PKS and NRPS genes the functions of which are still not explored fully [50][51][52] . In the antimicrobial screening, Streptomyces sp. PB-64 exhibited bioactivity against all the test microorganisms and interestingly it indicated the presence of all the three biosynthetic genes i.e. PKS-I, PKS-II and NRPS. Challis 53 reported that microbial genome sequences harvest many orphan or cryptic biosynthetic gene clusters which have the capability to direct the synthesis of novel, structurally complex natural products. The secondary metabolite pathway products of these biosynthetic gene clusters were predicted using DoBISCUIT database 54 . In recent years, prediction of chemical classes have been applied effectively for the discovery of type I and type II polyketides and nonribosomal peptides 38,55,56   www.nature.com/scientificreports www.nature.com/scientificreports/ similarity to PB-64 isolate and interestingly, this isolate was positive for the presence of all the three biosynthetic genes. These compounds are reported to be antibacterial in nature 61,62 .
16S rDNA-ARDRA is an excellent molecular genome typing method for classification of actinobacteria at the genus level 63 . Comparative diversity analysis of the 77 antagonistic actinobacteria through ARDRA using Hinf1 revealed significant difference among the isolates indicating its true diversity study. The genetic variation among the isolates might be due to mutation or recombination 64,65 . Natural populations of soil bacterium may exhibit genetic diversity due to variable habitat conditions and soil properties 66 . Also, the genetic composition and diversity of actinobacteria is influenced by variety in frequency and intensity of competition among locations of isolation 67 . ARDRA has proved its use in differentiating bacterial species within same genus and bacterial strains within same species 68,69 . Genetic diversity of the forests soil consisted of Streptomyces, Nocardia and Kribbella sp. which was determined by sequencing of 16S rDNA. Based on ARDRA fingerprinting analysis, strong antimicrobial activity and presence of antibiotic biosynthetic genes (PKS-I, PKS-II and NRPS), 41 isolates were selected as representatives and partially identified by 16S rDNA sequencing. BLAST search results implied that these 41 isolates belonged to the genus Streptomyces (33 isolates), Nocardia (7 isolates) and Kribbella (1 isolate). From the sequencing result, it was clear that Streptomyces was found to be the dominant genus in the soil of protected forest ecosystems which was also reported previously 12,19,70,71 .
Kz-24 was selected based on its promising antimicrobial potential against the test microorganisms. MIC of EA-Kz-24 support the popular notion that antimicrobial metabolites extracted from Streptomyces sp. at very low concentrations can be one of the finest sources of potent antimicrobials for treatment of infectious diseases especially those caused by clinically resistant pathogens, such as P. aeruginosa, MRSA, C. albicans 72 . Our finding is similar with the result of Kumar et al. 73 where the crude ethyl acetate extract of S. lavendulae SCA5 showed potent antimicrobial action with an MIC of 125 μg/mL against bacteria and MIC against fungi was 31.25 μg/mL. Kz-24 showed 96.2% 16S rDNA sequence similarity with S. malaysiensis NBRC 16446 (AB249918). The aerial mycelium of Kz-24 was brown in colour in GLM agar containing yeast extract/malt extract while S. malaysiensis NBRC 16446 was repoted to be dark grey in a medium with same composition 74 . S. malaysiensis was previously reported to exhibit strong antifungal activity and low antibacterial activity against B. subtilis NCIB 3610 [74][75][76] . This is the first report on its antibacterial activity against E. coli MTCC 40.
SEM experiments further confirmed that the strong antimicrobial activity of EA-Kz-24 led to significant morphological changes in the selected test microorganisms leading to cell shrinkage and cytosolic loss. These results are in symmetry with the findings of Sharma et al. 24 , Supaphon et al. 77 , Nurkanto and Julistiono 78 .
Actinobacteria are significant producers of bioactive secondary metabolites with different biological activity. There are many reports available for the use of GC-MS to analyze microbial metabolites chemically [79][80][81] . In this study, GC-MS analysis was performed on EA-Kz-24 and thirteen chemical compounds with different retention time and abundance were detected. The compounds identified were esters, alkenes, a phenolic compound, diketopiperazine and pyrrolidinopiperazine. Phenolic compounds are known to be powerful antimicrobials and antioxidants because they can reduce free radicals by hydrogen-donating ability 82 . Studies led by Balachandran et al. 83 and Kumar et al. 84 exhibited maximum antimicrobial action with highest phenolic compounds in GC-MS fractions. The antimicrobial activity of 3,5-bis(1,1-dimethylethyl)-phenol by Nocardia sp. PB-52 was reported for the first time by our group 24 . The pyrrolizidine compounds present in EA-Kz-24 included hexahydro-pyrrolo[1,2-a] pyrazine-1,4-dione (22.91%), hexahydro-3-(2-methylpropyl)-pyrrolo[1,2-a]pyrazine-1,4-dione (37.27%) and hexahydro-3-(phenylmethyl)-pyrrolo[1,2-a]pyrazine-1,4-dione (6%). Previous findings conclude that these compounds possess promising antimicrobial activity 85,86 . These compounds could be the key contributors for potent antimicrobial action of EA-Kz-24. A subset of researchers reported the antagonistic potential of alkenes  www.nature.com/scientificreports www.nature.com/scientificreports/ such as (Z)-3-tetradecene and (E)-9-octadecene against an array of test pathogens 87-89 but (Z)-3-tridecene has not been reported for any antimicrobial activity till date. Manilal et al. 90 90 reported strong antimicrobial activity of ester compounds. Niku-Paavola et al. 92 reported that a piperazinedione compound, 3-(2-methylpropyl)-2,5-piperazinedione, from Lactobacillus plantarum strongly inhibited the growth of Pantoea agglomerans and Fusarium avenaceum. Musthafa et al. 93 reported the effect of 2,5-piperazinedione in reducing the production of quorum sensing dependent factors in Pseudomonas aeruginosa PAO1 both in vivo and in vitro. According to the reports by Jain et al. 94 , 1-(4-chlorophenyl)-1-propyl piperazine and 1-(4-methylphenyl)-1-propyl piperazine exhibited excellent inhibitory activity against S. aureus and P. aeruginosa respectively. Jadhav et al. 95 reported strong antibacterial and antifungal activity by different piperazine group of compounds. Thus, it can be concluded that propanoic acid decyl ester, propanoic acid,3-mercapto-dodecyl ester, 3-(phenylmethyl)-2,5-piperazinedione and N-acetyl-3-methyl-1,4-diazabicyclo[4.3.0]nonan-2,5-dione present in EA-Kz-24 might have a significant role to play in its inhibitory effect against a wide range of test microorganisms which is reported for the first time. The combinatorial effect of bioactive compounds found in GC-MS analysis was demonstrated previously [79][80][81] . We, therefore, suggest that these compounds might be the key contributing factor for the antimicrobial action of EA-Kz-24. The forest ecosystems of biodiversity hotspots represent diverse and largely underexplored ecosystem for the isolation of actinobacteria producing effective antimicrobial compounds. It can be inferred from our findings that actinobacteria can be the fundamental sources for the discovery of valuable antibiotic compounds of high industrial and commercial values for human welfare.

Materials and Methods
Sample collection and measurement of soil pH. Soil samples were collected from Pobitora Wildlife Sanctuary and Kaziranga National Park of Assam, India (Fig. 4). Soil samples each weighing ∼50 g were collected randomly in sterile zip-lock plastic bags within 50 m 2 area from a depth of 5 to 20 cm after removing the top soil. Four samples from each site were bulked and homogenized to prepare composite samples. pH of the soil samples was also measured 41 .
Selective isolation and preliminary identification of actinobacteria. The composite soil samples were suspended in physiological saline (NaCl 9 g/L), thoroughly homogenized by stirring and serial dilutions up to 10 −4 were plated on three isolation media: Actinomycetes isolation agar, Streptomyces agar and GLM agar (Yeast extract, 3 g; malt extract, 3 g; peptone Type I, 5 g; starch, 10 g; agar, 20 g; distilled water, 1000 mL) supplemented with amphotericin B (75 µg/mL) and rifampicin (2.5 µg/mL) 19,24 . The inoculated plates were incubated at 28 °C and examined regularly for the appearance of actinobacteria colonies until 4 weeks.
The pure isolates were grouped according to their colony morphology, colour of aerial and substrate mycelium, colour of diffusible pigments and spore chain morphology 96,97 . The spore chain morphological features were observed by light microscopy.
Screening for antimicrobial potential. The 99 . The experiment was repeated for three times.
The actinobacteria found to be promising in the preliminary antimicrobial screening were subjected to secondary screening by disc diffusion method (Bauer et al., 1966) Screening for the production of extracellular enzymes. All the 77 antagonistic actinobacteria were screened qualitatively for the production of 5 important extracellular enzymes, i.e. amylase, cellulase, protease, lipase and esterase. Each actinobacteria was spot inoculated on agar plates amended with the respective substrates such as starch, carboxyl methyl cellulose, casein and Tween 80 and Tween 20 and was incubated for up to 10 days at 28 °C 100-103 .

PCR amplification of biosynthetic genes PKS-I, PKS-II and NRPS and analysis of biosynthetic
genes for prediction of chemical classes. Genomic DNA isolation of actinobacteria and PCR amplification of PKS-I and NRPS was performed as described previously 24,104 . Degenerate primers KSαF (5′-TSG CST GCT TCG AYG CSA TCA-3′) and KSαR (5′-TGG AAN CCG CCG AAB CCG CT-3′) targeting ketosynthase gene of the minimal PKS cluster were used for amplification of PKS-II 105 . PCR reactions were performed in a final volume of 50 μl in Proflex PCR System (Applied Biosystems, USA). The reaction mixture comprised of template DNA (50 ng), each dNTP (0.2 mM), 1X Taq DNA polymerase buffer, MgCl 2 (1.5 mM), each primer (0.2 µM) and 1 U Taq DNA polymerase. The profile used for amplification of PKS-II genes were programmed as: initial denaturation at 94 °C for 5 mins; followed by 35 cycles at 95 °C for I min, 65 °C for 1 min, 72 °C for 2 mins and a final extension at 72 °C for 10 mins. The size of the amplicons was 613 bp (KSαF/ KSαR). The amplified products were determined by 1.8% (w/v) agarose gel electrophoresis and partially sequenced by automated DNA sequencer at Scigenome Labs Pvt. Ltd. (India).
The nucleotide sequences of PKS-I, PKS-II and NRPS were translated into amino acid sequences using the web tool ORF FINDER (http://www.ncbi.nlm.nih.gov/ projects/gorf/). The deduced amino acid sequences were used as queries to search related gene products in the NCBI and DoBISCUIT (Database of BIoSynthesis clusters CUrated and InTegrated, http://www. bio.nite.go.jp/pks/) 54 using the BLASTP algorithm 106 with default parameters. The secondary metabolites pathway products of these biosynthetic genes were identified using DoBISCUIT.
ARDRA. 16S rDNA PCR amplification was carried out as previously described 24,107 . ARDRA was performed for the identification of number of polymorphic groups and then select the representative actinobacteria isolates among these groups 108 . 50 ng of purified 16S rDNA PCR products were digested with 1.5 U of Hinf1 109,110 following the manufacturer's protocol (New England Biolabs, UK). The mixture was incubated at 37 °C for 4 hours. Fully digested restriction fragments together with 100 bp and 1 kb markers were resolved by 2% (w/v) agarose gel electrophoresis at 100 volts for 90 min containing 10 μg/mL ethidium bromide. Different ARDRA banding patterns were observed and the isolates were grouped accordingly. ARDRA banding pattern in binary data was graded visually where "1" indicated the presence of band and "0" indicated for absence of band. This binary data was useful for construction of the dendrogram. Similarities among these actinobacteria were calculated by Jaccard's coefficient in the SimQual program. Similarity index matrix was used to cluster the actinobacteria with SAHN tool based on UPGMA method and the TreePlot program of NTSYS-pc 2.02e software package 111 . Identification of actinobacteria by 16S rDNA sequence analysis. Based on the antimicrobial potential, presence of biosynthetic genes and ARDRA, actinobacteria were identified based on the 16S rDNA sequencing using the facility at Scigenom Labs Pvt. Ltd. (India) and Molbiogen (India). Identification of nearest phylogenetic neighbours of sequenced 16S rDNA was carried out using EzTaxon database (http://www. eztaxon. org/) 112 . The 16S rDNA gene sequences used in the phylogenetic analysis was retrieved from NCBI GeneBank. The aligned sequences were used to reconstruct the phylogeny using maximum likelihood method algorithm by MEGA version 6 113 . Bootstrap analysis carried out with 1000 replications determined the support of each clade 114 . Determination of MIC of Kz-24. MIC of Kz-24 was performed according to CLSI 115 and Andrews 116 using broth dilution method with slight modifications. To 5 mL of Mueller Hinton broth (for bacterial test organisms) and Sabouraud dextrose broth (for yeasts), 1×10 5 cfu/mL inoculum of test microorganisms (log phase culture) was added in different test tubes and incubated at 37 °C for bacteria (24 hours) and 25 °C for yeasts (48 hours). 10% DMSO was used to dissolve EA-Kz-24 (1 mg/mL) and the extract was prepared for MIC screening with two fold serial dilution (100-0.024 μg/mL). MIC was determined in presence of EA-Kz-24 after 24-48 hours. 10 μL