Simultaneous Production of Multiple Antimicrobial Compounds by Bacillus velezensis ML122-2 Isolated From Assam Tea Leaf [Camellia sinensis var. assamica (J.W.Mast.) Kitam.]

Bacillus velezensis ML122-2 is an antimicrobial-producing strain isolated from the leaf of Assam tea or Miang [Camellia sinensis var. assamica (J.W.Mast.) Kitam.]. The cell-free supernatant (CFS) of strain ML122-2 exhibits a broad-spectrum antimicrobial activity against various Gram-positive and Gram-negative bacteria as well as the mold Penicillium expansum. The genome of B. velezensis ML122-2 was sequenced and in silico analysis identified three potential bacteriocin-associated gene clusters, that is, those involved in the production of mersacidin, amylocyclicin, and LCI. Furthermore, six gene clusters exhibiting homology (75–100% DNA sequence identity) to those associated with the secondary metabolites bacilysin, bacillibactin, surfactin, macrolactin H, bacillaene, and plipastatin were identified. Individual antimicrobial activities produced by B. velezensis ML122-2 were purified and characterized by Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry analysis, revealing three antimicrobial peptides with molecular masses corresponding to surfactin, plipastatin, and amylocyclicin. Transcriptional analysis of specific genes associated with mersacidin (mrsA), amylocyclicin (acnA), plipastatin (ppsA), and surfactin (srfAA) production by B. velezensis ML122-2 showed that the first was not transcribed under the conditions tested, while the latter three were consistent with the presence of the associated peptides as determined by mass spectrometry analysis. These findings demonstrate that B. velezensis ML122-2 has the genetic capacity to produce a wide range of antimicrobial activities that may support a specific community structure and highlight the biotechnological properties of Assam tea.


INTRODUCTION
Members of the Bacillus genus, which represent Gram-positive and endospore-forming bacteria, are widespread in a variety of environments including air, soil, aquatic ecosystems, foods, skin, and the gastrointestinal tract of animals (Abriouel et al., 2011). Some species of Bacillus are believed to play a key role in biological control through the production of antimicrobial compounds (e.g., bacteriocins, non-ribosomal polypeptides, and polyketides) and in plant growth promotion, such as Bacillus amyloliquefaciens, Bacillus subtilis, and Bacillus tequilensis (Chen et al., 2007;Gao et al., 2017;Li et al., 2018). The antimicrobial metabolites produced by Bacillus spp. are used in clinical settings to achieve inhibition of pathogens, such as Bacillus cereus, Clostridium difficile, Listeria monocytogenes, and methicillin-resistant Staphylococcus aureus (MRSA; Sabaté and Audisio, 2013;Lv et al., 2020;. Additionally, certain species of Bacillus have been reported to elicit probiotic potential, in particular Bacillus amyloliquefaciens, Bacillus licheniformis, Bacillus pumilus, Bacillus siamensis, and Bacillus subtilis (Du et al., 2018;Jeżewska-Frąckowiak et al., 2019;. Bacteriocins are ribosomally synthesized antimicrobial peptides which exhibit antimicrobial activity mostly against closely related bacterial species (Klaenhammer, 1993). Bacteriocins have been classified into three major classes: Class I bacteriocins are small peptides which undergo post-translational modifications, while Classes II and III are small (0.77-10 kDa) and large (>10 kDa) unmodified linear antimicrobial proteins, respectively (Abriouel et al., 2011;Cotter et al., 2013;Alvarez-Sieiro et al., 2016). In addition to bacteriocins, several species of Bacillus have been described to produce non-ribosomally synthesized peptides (NRPs) and polyketides (PKs) with antimicrobial properties (Patel et al., 1995;Pathak and Keharia, 2013). NRPs and PKs are synthesized by large multi-modular synthetases, non-ribosomal peptide synthetases (NRPSs), polyketide synthetases (PKSs), or hybrid NRPS/PKS enzymes. NRPSs typically consist of one or more modules, each responsible for the enzymatic incorporation of a specific amino acid in a growing peptide. An individual NRPS module typically consists of three core domains, that is, domains responsible for adenylation, thiolation, and condensation. Similarly, a given PKS enzyme comprises acyl transferase, acyl carrier, and ketosynthase domains (Mootz et al., 2002;Wang et al., 2014;Aleti et al., 2015). Prediction of gene clusters responsible for the biosynthesis of antimicrobial compounds using genomemining tools has been applied for the identification and subsequent characterization of genes associated with antimicrobial compound production (Medema et al., 2011;Sekurova et al., 2019). BAGEL is a powerful prediction tool aimed at the identification of bacteriocin-associated genes (De Jong et al., 2006). AntiSMASH is a genome database for gene cluster analysis responsible for the synthesis of secondary metabolite compounds, such as NRPs, PKs, and other antimicrobials (Medema et al., 2011;Weber et al., 2015).
Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) is an analytical technique used for evaluating chemical components which are ionized into charged molecules. It has been applied for identification and analysis of biological molecules, especially proteins and peptides (Singhal et al., 2015). MALDI-TOF MS has also been used to identify and analyze antimicrobial peptides, such as amylocyclicin (Scholz et al., 2014), iturin, fengycin, surfactin (Yang et al., 2015;Théatre et al., 2021), and mersacidin (Viel et al., 2021).
Antimicrobial peptides (AMPs) have received substantial attention as an effective treatment of bacterial infections and as an alternative to antibiotics (Cotter et al., 2005), in many cases supported by their low toxicity to human cells (Yang et al., 2014). Furthermore, specific AMPs have not only been used in the food industry as preservatives but also in agricultural applications as antimicrobial compounds (Dischinger et al., 2014). Several studies have identified antimicrobial-producing Bacillus strains associated with soils and plants and are therefore believed to contribute to the biocontrol of plant pathogens (Shafi et al., 2017;Andrić et al., 2020). The mode of action of bacteriocins may be through interaction with specific membrane receptors causing bacterial membrane disruption and associated electrolyte leakage from bacterial cells, ultimately leading to cell death (Tymoszewska et al., 2017;Perez et al., 2018). In contrast, antibiotics typically act as enzyme inhibitors in DNA replication, protein, and fatty acid synthesis, or cell wall biosynthesis O'Rourke et al., 2020). Previous studies by  and  revealed the antimicrobial activity of B. velezensis ML122-2 isolated Assam tea [Camellia sinensis var. assamica (J.W.Mast.) Kitam.] leaf surface against S. aureus, including MRSA. Moreover, this strain was also shown to exhibit tannin tolerance and probiotic properties. In the current study, we describe the identification, purification, and characterization of antimicrobial compounds produced by B. velezensis ML122-2 revealing co-production of several distinct antimicrobial compounds. Genome and transcriptional analysis of B. velezensis ML122-2 revealed expression of the corresponding gene clusters for these antimicrobial activities.

Antimicrobial Activity Assay
Antibacterial activity was assayed using an agar well diffusion method according to the modified protocol of Sewify et al. (2017). The indicator bacteria (listed in (2 × 250 bp) were used as input for genome assemblies through the MEGAnnotator pipeline in default mode (Lugli et al., 2016). Open reading frames prediction was performed by Prodigal v2.6.3 (Strepis et al., 2020). Protein-encoding genes were automatically annotated using a BlastP v2.2.26 (cut-off value of E 0.0001) sequence alignments against the non-redundant protein (nr) database curated by NCBI. 1 The bacteriocin/ antimicrobial gene clusters were predicted with BAGEL4 software. 2 Meanwhile, gene clusters involved in the biosynthesis of secondary metabolites, such as those involved in the production of NRPs, and PKs, were predicted by antiSMASH software. 3 The genome sequence was deposited in GenBank under accession number JAGTWM000000000.

Purification and Identification of Antimicrobial Compounds in Cell Fractions
Strain ML122-2 was cultivated in 800 ml clarified TSB, which had been passed through a column containing Amberlite XAD-2 resin beads (  Following this, the beads were washed with 250 ml 40% ethanol (Fisher Scientific, UK), and bound peptides were eluted from the column with 250 ml 70% (v/v) isopropanol-containing 0.1% (v/v) trifluoroacetic acid (IPA). In parallel, cells from the corresponding cell pellet were mixed with 250 ml IPA and stirred at room temperature for 3-4 h. Subsequently, the mixture was centrifuged at 8,000 × g at 4°C for 20 min. Both IPA eluent and IPA supernatant obtained from CFS and cell pellets, respectively (20 ml each), were applied to a 1 g Strata-E C18 SPE column (Phenomenex, Cheshire, UK) which was pre-equilibrated with 40% methanol and water. Each column was subsequently washed with 20 ml of 40% ethanol and then eluted using 20 ml IPA. The C18 SPE IPA eluents were assessed by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (Axima TOF 2 MALDI-TOF mass spectrometer, Shimadzu Biotech, Manchester, UK) and the molecular mass of bacteriocins determined in positive ion linear mode according to the protocol described by Hill et al. (2020; Figure 1A).

RP-HPLC Purification of Antimicrobial Peptides
Antimicrobial peptides (except surfactin, see below) were purified from CFS and cell pellets using C18 SPE and a reversed phase HPLC (RP-HPLC). The C18 SPE IPA eluents obtained as described here ( Figure 1A) were applied to a semi prep Proteo Jupiter C12 (250 × 10 mm, 4 μ, 90 Å) followed by running a 40 to 85% isopropanol 0.1% trifluoroacetic acid (TFA) gradient. Eluent B was 99.9% isopropanol-containing 0.1% TFA at a flow rate of 2.5 ml/min. Peptide-containing fractions were detected by measuring the absorbance at 214 nm. Fractions that exhibited antimicrobial activity were collected and pooled, subjected to rotary evaporation, and then lyophilized. Each purified antimicrobial peptide was resuspended in 600 μl 50% isopropanol ( Figure 1A).

Partial Purification of Surfactin From Cell-Free Supernatant
Partial purification of surfactin was achieved by organic solvent extraction according to the protocol described by Lei et al. (2020) with modifications as follows. Strain ML122-2 was cultured in TSB and incubated at 37°C, on a rotating platform at 150 rpm for 48 h before centrifugation at 5,000 × g at 4°C for 15 min. The supernatant was subsequently filtered through a 0.20 μm nylon membrane filter. 25 ml ethyl acetate (Sigma-Aldrich™, St. Louis, MO, United States) was mixed with 25 ml filtered cell-free supernatant using a vortex mixer for 10 min prior to centrifugation at 5,000 × g at 4°C for 60 min. Subsequently, the top phase (organic phase), approximately 25 ml, was transferred into a glass bottle. Solvent evaporation was performed using Genevac™ miVac centrifugal concentrator (Genevac Limited, Suffolk, UK) at room temperature for 80 min. The evaporated solvent extract was resuspended in 0.01 M PBS (5 ml; Figure 1B). Antibacterial activity was investigated using the agar well diffusion method described above.

RT-qPCR Analysis
Transcriptional activity of genes associated with gene clusters predicted to be responsible for mersacidin, amylocyclicin, plipastatin, and surfactin production was investigated using RT-qPCR analysis, whereby the mrsA, ancA, ppsA, and srfAA genes served as target genes, respectively. The housekeeping gene rpsE was used as reference for this analysis. Primers were designed using Primer3Plus 4 and listed in Table 2. B. velezensis ML122-2 was cultivated in TSB and incubated overnight at 37°C on an orbital platform shaker (150 rpm) prior to centrifugation at 5000 × g at 4°C for 10 min. The resulting cell pellet was washed twice with 0.85% (w/v) NaCl and adjusted to an OD 600nm of 0.  . However, it has been suggested that rpoB represents a more robust marker (than the 16S rRNA gene) to determine the phylogeny of bacilli that belong to the so-called "operational group Bacillus amyloliquefaciens, " the latter constituting the closely related species B. amyloliquefaciens, B. velezensis, and B. siamensis (Fan et al., 2017). BlastN analysis of the rpoB gene of strain ML122-2 revealed 100% sequence identity with that of Bacillus velezensis strains and with reduced sequence identity to rpoB of B. siamensis (<98.8%) and B. amyloliquefaciens (<99.8%). This finding confirms that strain ML122-2 belongs to the B. velezensis species rather than B. siamensis. To validate this, the average nucleotide identity (ANI) of ML122-2 was analyzed in comparison with those of strains of the B. amyloliquefaciens, B. siamensis, and B. velezensis species. ML122-2 exhibits ANI values of 97.86, 97.85, and 94.65% with B. velezensis ATR2, B. amyloliquefaciens FBZ42, and B. siamensis SCSIO 05746, respectively. The ML122-2 genome was shown to lack identifiable CRISPR-Cas systems, while it is predicted to contain three prophage-associated regions (13.6, 31.8, and 28.7 kb in length, respectively). Two of these appear to represent incomplete prophage regions, while one is predicted to be intact and located within positions 1,213,485-1,245,310 on the genome. This prophage region contains genes predicted to encode DNA replication enzymes, capsid and tail structural components, and lysis functions. BlastN analysis of this putative prophage region highlights that it is highly conserved among the sequenced genomes of B. velezensis strains.
Further in silico analysis was performed using BAGEL4 and antiSMASH to identify genes involved in the production of antimicrobial or bioactive compounds. A total of four putative bacteriocin or bacteriocin-like gene clusters were predicted by BAGEL4 software including those encoding the biosynthetic and immunity genes for mersacidin, amylocyclicin, ComX, and LCI ( Table 3). The predicted ML122-2 mersacidin gene cluster was shown to comprise of mrsK2, mrsR2, mrsF, mrsG, mrsE, mrsA, mrsR1, mrsD, mrsM, and mrsT and is similar to that of Bacillus sp. HIL-Y85/54728 (Genbank accession no. AJ250862; 98%) which was previously described by Altena et al. (2000; Figure 2A). Therefore, it appears that a complete mersacidin gene cluster is present in the B. velezensis ML122-2 genome. The ML122-2 genome also contains a gene cluster with high identity (98%) to the amylocyclicin cluster of B. velezensis FZB42 (Scholz et al., 2014; Figure 2B). The comX gene cluster of B. velezensis ML122-2 elicits 35% identity with that of B. velezensis FZB42 which encodes the competence pheromone ComX peptide, while the lci gene encodes a putative antimicrobial peptide, and exhibits 89% identity with the corresponding lci gene of B. velezensis FZB42 (Supplementary Figure 1). Gene clusters with nucleotide sequence similarity values below 30% were deemed insignificant.
In addition to ribosomally synthesized antimicrobial peptides, Bacillus spp. have been reported to produce non-ribosomally synthesized antimicrobial compounds. Sequence analysis using antiSMASH identified nine gene clusters predicted to be involved in the production of secondary metabolites including NRPs and PKs, of which six were shown to exhibit 75-100% nucleotide identity to known NRP/PK clusters from strains of Bacillus spp. (Table 3). Of these latter six clusters, five are predicted to encode NRPs (bacilysin, bacillibactin, surfactin, macrolactin H, and plipastatin), while the remaining one is associated with the biosynthesis of a PK (bacillaene). Genes associated with macrolactin biosynthesis are typically identified on the genomes of B. velenzensis strains, while they have not been observed among the genomes of B. siamensis or B. amyloliquefaciens strains (Fan et al., 2017). This finding supports the reassignment of this strain as a B. velezensis strain. The bacilysin-and bacillibactin-associated clusters display 100% sequence identity with equivalent clusters in B. velezensis FZB42 which includes seven (bacABCDEFG) and five (dhbACEBF) subunit genes (Figures 3A,B), respectively. The bacillibactin biosynthesis cluster exhibits 75% nucleotide identity with its counterpart in B. subtilis subsp. subtilis str. 168, while the surfactin gene cluster (srfAA, srfAB, srfAC, and srfAD) exhibits 98 and 79% nucleotide identity with B. velezensis FZB42 and B. subtilis JH642, respectively. The genome of B. velezensis ML122-2 was shown to lack the ycxBCD genes located downstream of the sfp gene ( Figure 3C). Furthermore, the macrolactin H biosynthesis gene cluster, mlnABCDEFGHI, exhibits 100% sequence identity to those of B. velezensis FZB42, whereas the plipastatin biosynthesis gene cluster, ppsABCDE, displays 97% identity with that of B. velezensis FZB42 (Figures 3D,E). The bacillaene-associated gene cluster shows 97% identity to that of B. velezensis FZB42, which consists of eight subunit genes, baeEDLMNJRS ( Figure 3F).

Antimicrobial Purification and Mass Spectrometry Analysis
Based on genome analysis, B. velezensis ML122-2 has the genetic capacity to produce a considerable number of distinct antimicrobial compounds. Accordingly, in order to assess which of the predicted antimicrobial compounds are responsible for the observed antimicrobial activity of B. velezensis ML122-2, we characterized the antimicrobial peptides produced in cell pellets and CFS extracts and analyzed the active fractions by MALDI-TOF MS (see "Materials and Methods"; Figure 1) Figure 4A). Further purification by RP-HPLC allowed separation of antimicrobial activities in two active fractions, one of which corresponded to ion peaks with m/z values of 1,449.9, 1,463.9, 1,471.9, 1,487.9, 1,485.9, and 1,501.9 (Figure 4B), and one which corresponded to the peaks at m/z 6,381.4 and 3,190.3 ( Figure 4C). Since surfactin could not purified by RP-HPLC, possibly due to its inherent hydrophobic nature, (partial) purification of this compound was achieved by solvent extraction with ethyl acetate. Ethyl acetate possesses a lower polarity than isopropanol, which was used in the RP-HPLC purification and may explain its (near) absence in the original purification. Moreover, a previous study revealed that surfactin extraction by ethyl acetate is associated with high purity and yield of the compound (Chen and Juang, 2008). The MALDI-TOF mass spectra of unpurified extract obtained from solvent extraction represented the peaks at m/z 1,032.38, 1,046.25, 1,060.24, and 1,103.99 ( Figure 4D). Different molecular weights for purified plipastatin and purified surfactin have previously been described regarding the production of surfactin and fengycins/plipastatin with fatty acid side chains of 15 to 17 carbon atoms in a Bacillus strain (Koumoutsi et al., 2004), resulting in incremental molecular mass increases of 14 Da for purified plipastatin and surfactin. A previous study by Pathak et al. (2014) reported the mass spectrum [M + H] + of surfactin from Bacillus strain at m/z 994.7, 1,008.7, 1,022.7, 1,036.7, 1,064.7, 1,078.7, and 1,092.7 consistent with unsaturated C 12 -C 17 β-hydroxy fatty acids. Similarly, [M + H] + ions at m/z 1, 433.8, 1,447.8, 1,461.8, 1,475.8, and 1,489.9 were assigned to plipastatin isoforms that correspond to unsaturated C 14 -C 18 β-hydroxy fatty acids (Gao et al., 2017). Here, we successfully purified surfactin via solvent extraction. Furthermore, based on peak height, it appeared that cells represented a better source of amylocyclicin and lipopeptides than CFS (Supplementary Figure 2). Genome analysis identified intact gene clusters associated with the biosynthesis of additional ribosomally (mersacidin) and non-ribosomally (macrolactin, bacillaene, bacilysin, and bacillibactin) synthesized compounds. However, the molecular masses associated with mersacidin, bacilysin, bacillibactin, macrolactin H, and bacillaene (Table 3) were not detected through MALDI-TOF MS in either the crude or purified extracts suggesting that these compounds are not produced under the applied laboratory conditions.
The antibacterial activity of individually purified amylocyclicin, plipastatin, and surfactin against B. cereus TISTR 687, B. subtilis NCDO 10073, E. coli DH5α, L. innocua UCC3, Leu. paramesenteroides NCDO 869, MRSA DMST 20625, and S. aureus ATCC 25923 as determined by the agar well diffusion method is presented in Table 4 and Supplementary Figure 3. The purified amylocyclicin was shown to inhibit growth of all test indicator strains with the inhibitory value ranging between 7.8 and 24.0 mm, while purified plipastatin represented antimicrobial activity against B. cereus TISTR 687, Leu. paramesenteroides NCDO 869, MRSA DMST 20625, and S. aureus ATCC 25923, with an associated zone of inhibition ranging from 7.5 to 7.8 mm, and with no inhibition observed for B. subtilis NCDO 10073 and L. innocua UCC3. The purified surfactin obtained via solvent extraction was demonstrated to elicit antimicrobial activity against L. innocua UCC 3 and Leu. paramesenteroides NCDO 869, producing a zone of inhibition of 11.8 and 11.7 mm, respectively, whereas no inhibition was observed when B. cereus TISTR 687, B. subtilis NCDO 10073, E. coli DH5α, and MRSA DMST 20625 and S. aureus ATCC 25923 were used as indicator bacteria (Table 4

Transcriptional Activity of Genes Associated With the Gene Clusters
To validate the mass spectrometry-based identification of the (partially) purified compounds, transcriptional analysis of genes in the analysis, it was selected as a representative negative control for the transcriptional analysis. At 24-h cultivation, ancA, ppsA, and srfAA genes were upregulated 1.57-, 2.80-, and 1.16-fold, while after 48-h incubation, the transcription levels were upregulated 1.23-, 1.75-, and 2.53-fold, respectively. The relative expression level of gene mrsA at 48-h incubation (0.18-fold) was not significant when compared with 24-h incubation (0.15-fold; Figure 5). The low transcription levels measured for the mrsA gene suggest lack of expression of this gene cluster, being consistent with a failure to detect mersacidin.

DISCUSSION
A previous study by    Herzner et al. (2011). The failure to detect mersacidin may be due to low expression levels of mrsA and associated genes. Possibly, the expression of the mersacidin-associated genes may be induced under stress conditions or specific media and deserves further investigation. It has previously been shown that inactivation of mrsR1 causes loss of mersacidin production in synthetic medium by inhibiting MrsA synthesis (Guder et al., 2002), while knockouts of mrsR2K2 genes explicitly have been shown to prevent induction of mersacidin transcription (Schmitz et al., 2006). Several species of Bacillus can produce secondary metabolites with antimicrobial activity against plant pathogenic bacteria and fungi (Lv et al., 2020), and plant growth-promoting activity with plant hormone production ability, such as indole-3-acetic acid (IAA) and 2,3-butanediol (Chen et al., 2007). Various studies have reported the production of multiple antimicrobial compounds by Bacillus strains. A previous study by Han et al. (2018) revealed a broad-spectrum of antimicrobial activity produced by B. amyloliquefaciens WY047 as a result of the simultaneous production of six antimicrobial substances. Gao et al. (2017) presented an engineered B. subtilis pB2-L with the ability to co-produce surfactin and plipastatin. In the current study, B. velezensis ML122-2 was observed to co-produce amylocyclicin, plipastatin, and surfactin. Several reports highlight the role of amylocyclicin (Chen et al., 2009), plipastatin/fengycin, and surfactin (Ongena et al., 2007) in the biological control of plant pathogens. The study by Scholz et al. (2014) revealed that amylocyclicin produced by B. velezensis FZB42 exhibits high antibacterial activity against Gram-positive bacteria (e.g., B. subtilis, B. cereus, Micrococcus luteus, and Paenibacillus granivorans). Surfactin exhibits antimicrobial and emulsification activities and inhibits biofilm formation (Chen et al., 2015). Plipastatin, also known as fengycin, has been reported to demonstrate antibacterial (e.g., L. monocytogenes, S. aureus, and Salmonella Typhimurium) and antifungal (e.g., Fusarium oxysporum and Pythium ultimum) activities causing cellular membrane distortion and cell membrane pore formation and  Frontiers in Microbiology | www.frontiersin.org ultimately death of cells (Gao et al., 2017;Jeong et al., 2018;Lin et al., 2020). The ability to produce multiple antimicrobial compounds has been described to increase the potential for biological control (Han et al., 2018). Our findings support the notion that B. velezensis ML122-2 found on Assam tea leaf plays a role in microbiological control in Assam tea or Miang cultivation via the production of antimicrobial peptides (Rungsirivanich et al., 2019). Consequently, B. velezensis that can be found both on Assam tea leaves  and in fermented Assam tea products (Unban et al., 2020) may exert a powerful biocontrol function in environments, preventing food spoilage through the production of antimicrobial compounds, such as amylocyclicin, plipastatin, and surfactin.

CONCLUSION
B. velezensis ML122-2 exhibits strong and broad-spectrum antimicrobial activity. Three antimicrobial peptides produced by B. velezensis ML122-2, that is, amylocyclicin, plipastatin, and surfactin, were purified from CFS and cell pellets, and their masses confirmed by MALDI-TOF mass spectrometry, this being consistent with transcriptional activity of specific marker genes for the corresponding gene clusters. Each purified peptide was shown to be antimicrobial, with amylocyclicin, in particular, eliciting substantial antimicrobial activity. These findings show that B. velezensis has the potential to play an important role in microbial biocontrol in Assam tea cultivation and Assam tea fermentation.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm. nih.gov/genbank/, JAGTWM000000000, MH796212, CP000560, AL009126, AJ250862, and X70356.

AUTHOR CONTRIBUTIONS
PR, EP, JM, NT, and DS designed the experiments. PR, EP, PO'C, and DF analyzed the data. PR, JM, and DS investigated the data. NT and DS acquired the funding. PR and NT prepared the original draft. PR, PO'C, JM, NT, and DS reviewed and edited the manuscript. All authors contributed to the article and approved the submitted version.