Introduction
Plant pathogenic fungi, bacteria, viruses, and viroids can reduce agricultural productivity and result in yield losses of up to 14% for various crops (Peng et al., 2021). An essential step in the production of agricultural products is the application of pesticides to combat plant diseases. Chemical pesticides have been used extensively to control plant diseases; without them, the production of fruits, vegetables, and grains would have dropped by 78%, 54%, and 32%, respectively. Regardless of all the above, excessive pesticide usage in agriculture pollutes the environment and has a negative impact on human health (Syed Ab Rahman et al., 2018; Tudi et al., 2021). Furthermore, pesticides can alter the composition of plant-associated microbial communities (Meena et al., 2020). Therefore, the development of ecofriendly pesticide alternatives is urgently required. Due to their safe and environmentally friendly effects on crops, antagonistic bacteria have become a potent substitute for conventional pesticides in the management of crop diseases in recent years. Numerous strains of the Bacillus species are being evaluated for use as biopesticides and have gained prominence as a biocontrol agent for plant diseases. Bacillus strains are the most promising group of plant growth-promoting rhizobacteria (PGPR), which play important roles in promoting plant growth, enhancing growth hormones, producing antioxidant enzymes, nitrogen fixation, phosphate solubilization, phytohormones, and volatile organic compounds (VOCs), and triggering induced systemic resistance (ISR) by producing different types of secondary metabolites that can potentially inhibit the growth of plant pathogens and control soil-borne diseases (Ongena and Jacques, 2008; Chowdhury et al., 2015; García-Fraile et al., 2015; Fan et al., 2018; Chandran et al., 2020; Samaras et al., 2021).
Bacillus atrophaeus is a significant member of the plant growth-promoting rhizobacteria (PGPR), which are known for enhancing plant growth development as well as controlling plant pathogenic fungi and bacteria. When applied to seedlings and harvested fruits, they stimulate plant growth and improve the resistance of plants against insect pests and plant diseases such as powdery mildew and tomato gray mold (Reva et al., 2013; Zhang et al., 2013; Huang et al., 2015). Bacterial secondary metabolites are important not only for producer cells but also have a positive impact on their host. These secondary metabolites have significant applications in agriculture and pharmaceuticals as bioactive compounds (Bornscheuer, 2016). Advances in whole-genome sequencing technology have enabled the detection of putative antimicrobial and genome mining tools, enabling researchers to uncover the molecular basis of strain-versatile lifestyles and prioritize industrially important secondary metabolites at the genomic level. These specialized metabolites represent a possible way to improve crop yield and produce antimicrobial activities to control plant diseases (Chun et al., 2017, 2019; Wang et al., 2020; Iqbal et al., 2021). For example, genome mining of B. atrophaeus L193 has revealed a non-ribosomal peptide synthetase gene cluster involved in the production of surfactin, fengycin, bacillomycin, and iturin (Rodríguez et al., 2018).
Genome analysis of B. atrophaeus GQJK17 revealed eight gene clusters that produce antimicrobial secondary metabolites such as surfactin, fengycin, bacillaene, and bacillibactin (Ma et al., 2018). Surfactin, fengycin, bacillomycin, iturin, bacillaene, and bacillibactin have been reported to have antimicrobial properties (Wang et al., 2024). Bacillibactin has been reported to inhibit the growth and invasion of Phytophthora capsici and Fusarium oxysporum (Woo and Kim, 2008; Yu et al., 2011). Surfactin was reported to have a broad spectrum of antibacterial activity to significantly inhibit bacterial diseases such as Arabidopsis root infection caused by Psuedomonas syringae and tomato wilt caused by Ralstonia solanacearum (Bais et al., 2004; Xiong et al., 2015), and fengycin exhibited antifungal activity against a broad spectrum of filamentous fungi (Vanittanakom et al., 1986).
In this study, we selected the B. atrophaeus HAB-5 strain, which was isolated from the cotton rhizosphere in Xinjiang Province, Republic of China. The HAB-5 strain has potential as a biological agent for antifungal and antiviral agents. Previous studies conducted in our laboratory have shown that HAB-5 can provide high control efficacy against 22 plant pathogenic fungi such as Alternaria alternata MfAa-1, Alternaria brassicicola MfAb-1, Colletotrichum gloeosporioides MnCg-2, Colletotrichum musae BnCy-1, Colletotrichum gloeosporioides MnCg-3, Colletotrichum gloeosporioides MnCg-4, Colletotrichum gloeosporioides MnCg-1, Colletotrichum gloeosporioides MnCg-5, Colletotrichum gloeosporioides MnCg-6, Colletotrichum gloeosporioides MnCg-7, Colletotrichum gloeosporioides MnCg-9, Colletotrichum gloeosporioides MnCg-10, Colletotrichum gloeosporioides MnCg-11, Corynespora cassiicola HbCc-1, Corynespora cassiicola HbCc-2, Curvularia geniculata CyCg-1, Fusarium oxysporum f. sp. cubense FOC-4, Fusarium proliferatum PgFp-1, Fusarium oxysporum PgFp-2, Phyllosticta theaefolia, Phytophtora nicotianae, and Trichothecium roseum PyTr-1, and the inhibition range from 21.07 to 67.29% was recorded (Rajaofera et al., 2020). It also prevents disease infection, protects tobacco seedlings from P. nicotianae, and exhibits an antiviral effect against tobacco mosaic virus (TMV) by activating the signaling of regulatory genes (NPR1), defense genes (PR-1a, PR-1b, and chia5), and hypertensive response-related genes (Hsr 203j and Hin1; Rajaofera et al., 2019). Similarly, HAB-5 was found to be effective against C. gloeosporioides through the volatilization of antimicrobial volatile compounds such as octadecane, hexadecanoic acid, methyl ester, and chloroacetic acid, tetradecyl ester, chloroacetic acid, tetradecyl ester, octadecane, hexadecanoic acid, and methyl ester (Rajaofera et al., 2018). In addition, the HAB-5 strain exhibited a remarkable ability to improve the growth of tobacco plants, and the inoculated plant exhibited a significant increase in fresh shoot weight, dry shoot weight, fresh root weight, and dry root weight by 76.47%, 80.58%, 71.71%, and 82.10%, respectively, compared with the non-inoculated control plant (Rajaofera et al., 2019). These interesting features have piqued our interest in B. atrophaeus HAB-5. Analysis of the whole genome of HAB-5 revealed the PGPR and biological control activities. The objective of this study was to perform genome assembly of B. atrophaeus HAB-5 and comparative genome analysis of B. atrophaeus HAB-5 with B. amyloliquefaciens DSM7, B. atrophaeus SRCM101359, B. velezensis FZB42, B. velezensis HAB-2, and B. subtilis 168. In addition to evaluating gene clusters encoding potential secondary metabolites of B. atrophaeus, HAB-5 may contribute to plant growth-promoting and biocontrol activities.
Materials and methods
<italic>Bacillus atrophaeus</italic> HAB-5 strain selection and genomic DNA extraction
HAB-5 was isolated from Xinjiang Province, Republic of China (Rajaofera et al., 2019) and preserved in the Key Laboratory of Green Prevention and Control of Tropical Plant Disease and Pest at Hainan University, Ministry of Education. The HAB-5 was cultivated in Luria–Bertani (LB) agar medium at 37°C with shaking at 180 rpm and was used to extract genome DNA. The bacterial culture was subjected to genomic DNA extraction using a commercial kit according to the manufacturer’s instructions (Sigma Aldrich, St. Louis, MO, United States). NanoDrop and Qubit (Thermo Fisher Scientific United States) were utilized for optical density measurement and quality control.
Genome sequences of <italic>Bacillus atrophaeus</italic> HAB-5
The complete genome of HAB-5 was sequenced by third generation sequencing on the PacBio RS II sequencing platform (Pacific Biosciences). Fragment DNA samples were sheared and treated with Exo VII to remove single-stranded ends, and size selection was performed to retain longer reads (>10 k reads) for sequencing. Blunt reactions were performed, and the SMRT bell template was annealed for sequencing. The large insert libraries were sequenced through single-molecule real-time (SMRT) sequencing, and the cells were run on the Pac Biosciences RS II systems using P6-C6 chemistry. After 180 min of mode data collection, all reads were spliced into contigs and combined into scaffolds.
Genome assembly, gene function annotation, and genome component prediction of <italic>Bacillus atrophaeus</italic> HAB-5
The Pac Bio reads were assembled into contigs using de novo hierarchical genome assembly process (HGAP) software in the single-molecule real-time sequencing (SMRT) portal using default parameters (Chin et al., 2013, 2016). The assembly results were then corrected based on NGS data using Pilon software (Walker et al., 2014). Finally, the gaps between contigs were filled by comparing the contigs assembled from PacBio using MUMmer software (Delcher et al., 1999). The genome sequences were annotated by the National Center for Biotechnology Information (NCBI) by using the Prokaryotic Genomes Automatic Annotation Pipeline (PGAAP). Functional description of putative protein-encoding genes was performed using BLASTx, with an E-value of 1e − 5. We used GenoVi software for circular genome representations.1 The Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology assignment and prediction of KEGG pathways were performed by Kanehisa et al. (2004) to identify the components of cellular processes (CP), environmental, genetic, human disease, metabolism, and organismal system pathways. The COG (Cluster of Orthologous Group) annotated the predicted genes in accordance with Tatusov et al. (2003); the gene ontology (GO) was completed by Bard and Winter (2000). Genome components were predicted using a glimmer2 using Markov models. Scam-SE was used for tRNA, rRNA, and sRNA recognition, and other types of RNAmmers were predicted by comparison with the Rfam database (Lowe and Eddy, 1997).
The estimation of the core and pan genomes
To categorize the core and pan genomes, HAB-5 was analyzed using the Prokaryotic Genome Annotation Pipeline (PGAP) to identify core orthologs from strains (Zhao et al., 2012). The size of the core genome was determined as the number of common genes shared by all analyzed genomes, and the pan-genome size was defined as the sum of all gene families. Species-specific core orthologous genes and strain-specific unique genes were also examined in the HAB-5 genome sequences. The Ortho MC tool was used to identify core-specific genes in the genomes (Li et al., 2003). The gene accumulation curve was produced using the R package gg plot 2 using the results from Roary.
Gene family clustering of <italic>Bacillus atrophaeus</italic> HAB-5 and collinearity analysis
Hclustersg software was used to carry out gene family alignment (Maqbool and Babri, 2007), and muscle software was used to analyze the alignment sequence for the cluster gene family (Edgar, 2004). The parameters were set as follows: a Blastp E-value threshold of 1e-5 to ensure the quality of the comparisons. Genome-wide collinearity among strains HAB-5, DSM7, SRCM101359, FZB42, HAB-2, and 168 was determined using the BLASTp database, with an e-value of ≤1e − 5 and an identity threshold of ≥85% at both nucleic acid and amino acid levels. For the analysis of genome synteny and collinearity, D-GENIES and C-Sibelia software were used. Visualization of the alignment of the synteny blocks was achieved using Circos (Krzywinski et al., 2009; Minkin et al., 2013; Cabanettes and Klopp, 2018).
Phylogenetic trees and heat map synteny
All genomes used in this study were downloaded in the FASTA format from the NCBI database to construct neighbor-joining phylogenetic trees based on 16Sr RNA. Molecular Evolutionary Genetic Analysis (MEGA) was used to construct neighbor-joining phylogenetic trees (Tamura et al., 2011) with the p-distance model and 1,000 bootstraps. An analysis subset of SNPs identified in all single-copy genes. A phylogenetic tree of SNPs was constructed using HAB-5 against the reference genomes by Tree Best (Nandi et al., 2010) with 1,000 bootstraps. The average nucleotide identity analysis was performed using Jsspecies 1.2.1 (Richter et al., 2016). The CIMminer3 was used for heat maps based on the Average Nucleotide Identity values, and pairwise genome alignment for synteny was performed using Mauva Version 2.4.0 (Darling et al., 2004).
Genome mining analysis of secondary metabolite gene clusters
The web-based tool, Antibiotics and Secondary Metabolites Analysis SHell (antiSMASH 7.0) software, was used to predict the biosynthesis gene clusters of secondary metabolites in HAB-5 (Blin et al., 2023). antiSMASH is available at http://antismash.secondarymetabolites.org/. The assembled sequences were uploaded to antiSMASH reports containing both known and unknown clusters to identify similar clusters by genome comparisons with detailed NRP function annotation, and the chemical structure of the gene cluster was generated (Medema et al., 2011; Blin et al., 2013; Weber et al., 2015). The Roary pan genome pipeline was used to identify gene cluster homologies, and gene cluster synteny maps were produced using the R package genoPlotR (Guy et al., 2010; Page et al., 2015).
Discussion
Genome sequences were performed to ascertain the molecular basis of the mechanisms underlying the promotion of plant growth and the biocontrol capabilities of B. atrophaeus HAB-5, and a comparative genome analysis was conducted with other Bacillus strains. The genomic analysis showed that 1,182 (Mb) of clean data were created and that the HAB-5 genome contained a circular chromosome with a size of 4,083,597 bp and a GC content of 43.36%. Additionally, the genome did not contain any plasmids. A comparative analysis showed that the genome size of HAB-5 (4,083,597 bp) was larger than that of DSM7 (3,980,199 bp), FZb42 (3,918,596 bp), and HAB-2 (3,894,648 bp), but it was still smaller than that of 168 (4,215,606 bp) and SRCM101359 (4,180,819 bp). The GC content of HAB-5 (44.21%) was higher than that of 168 (43.1%) and SRCM101359 (43%) but it was lower than that of FZb42 (46.6%), HAB-2 (46.6%), and DSM7 (46.1%). To determine the relationship between HAB-5 and other strains, a phylogenetic tree based on the 16SrRNA gene sequence and a phylogenomic tree were constructed. The results showed that HAB-5 is closely related to SRCM101359 and showed the highest similarity.
In previous studies, many plant growth-promoting bacteria were analyzed at the whole-genome level to gain an in-depth understanding of PGP mechanisms in bacteria such as Pseudomonas aeruginosa B18 (Singh et al., 2021), B. velezensis HAB-2 (Xu et al., 2020), B. megaterium BM89 and B. subtilis BS87 (Chandra et al., 2021), K. variicola UC4115 (Guerrieri et al., 2021), and Streptomyces (Subramaniam et al., 2020). Numerous genes involved in biocontrol and growth-promoting activities in plants have been identified by whole genome sequencing. The presence of antimicrobial genes was also revealed by an analysis of the B. atrophaeus genome for the presence of secondary metabolites. Bacillus atrophaeus strains are remarkably capable of producing secondary metabolites with antimicrobial compounds, such as terpenes, polyketides, and non-ribosomally synthesized peptides (NRPs; Liu et al., 2012; Chan et al., 2013; Ma et al., 2018). In a previous study, genomic research revealed that B. atrophaeus L193 carries a cluster of genes known as non-ribosomal peptide synthetases. These genes include fenC, srfA-A, BmyB, and ituD, which are involved in the production of surfactin, fengycin, bacillomycin, and iturin (Rodríguez et al., 2018). Eight gene clusters that produce antimicrobial secondary metabolites, such as surfactin, bacillaene, fengycin, and bacillibactin, were found in the genome study of B. atrophaeus GQJK17 (Ma et al., 2018).
In the present studies, the whole genome sequencing of HAB-5 identified the genes encoding for novel antimicrobial peptides associated with its biocontrol properties. There were 11 gene clusters predicted in the genome of HAB-5: three gene clusters encoding for NRPS (non-ribosomal peptide synthetases), two gene clusters encoding for Transat-pks, two gene clusters encoding for terpene, one encoding for lanthipeptide, one gene cluster for T3PKS, one encoding RiPP, and one encoding thiopeptide that synthesized bacillaene, fengycin, bacillibactin, surfactin, and rhizocticin A. These secondary metabolites show antifungal and antibacterial activities against plant pathogens. According to Ongena and Jacques (2008), surfactin possesses antibacterial and antifungal properties, and fengycin and rhizocticin exhibit antifungal properties (Kugler et al., 1990). Bacillaene exhibits antimicrobial activity against many types of plant-harmful bacteria and fungi (Patel et al., 1995; Um et al., 2013; Müller et al., 2014), and two of the HAB-5 secondary metabolites gene clusters that may produce terpenes were identified; however, the other four are yet unknown. Terpenes are large, diversified, naturally occurring organic compounds that are present in bacteria, fungi, plants, and animals. They have a variety of medicinal uses and can be added to food and cosmetic products. They also have antifungal and anticarcinogenic characteristics (Zhao et al., 2016). Additionally, it plays a significant role in defending numerous plant, animal, and microbe species from infections and insects, as well as transmitting messages to non-specific and mutuality regarding the existence of food, partners, and adversaries, as well as from abiotic and biotic stressors (Gershenzon and Dudareva, 2007). Furthermore, the HAB-5 genome showed an amazing capacity to create bacillibactin, a type of siderophore that is characterized by short peptide molecules with functional groups and a side chain that can provide a set of ligands to coordinate ferric ions (Crosa and Walsh, 2002). Bacillibactin is a kind of strong siderophore that increases the absorption of ferric ions in soil for plant growth and to secrete volatile compounds (Nithyapriya et al., 2021). Furthermore, the gene cluster of Bacillibactin strain HAB-5 also contains other genes that promote plant growth and codify useful substances such as butanone, protease, phytase, and phosphatase (Ma et al., 2018; Rajaofera et al., 2020). Among the metabolites, VOCs have gained great attention for their potential in the control of plant pathogens. It has been reported that the strain HAB-5 produces a variety of VOCs, which have strong antifungal effects, inhibiting the growth of C. gloeosporioides (Rajaofera et al., 2019). Besides, HAB-5 has detected volatile chemical-producing genes such as acoA, acoB, acoR, acuB, and acuC. Acetion, one of the active bacterial volatile compounds, was released to stimulate the induced systemic resistance (ISR) of plants (Zhang et al., 2015).
Whole genome sequencing has revealed several genes linked to the promotion of plant growth, including chitinase, nitrogen fixation, phosphorus solubilization, auxin synthesis, iron acquisition, potassium, and IAA (Subramaniam et al., 2020; Chandra et al., 2021; Guerrieri et al., 2021; Iqbal et al., 2021; Singh et al., 2021). Most of the genes linked to promoting plant growth were found in HAB-5 after the genome HAB-5 annotation was completed in the current study. IAA is an important phytohormone that controls cell enlargement and tissue differentiation in plants. Genes, such as trpA, trpB, trpC, trpD, trpE, trpS, ywkB, miaA, and nadE, that contribute to the production of IAA, ethylene, and ammonia have been predicted in strain HAB-5. The occurrence of the gene clusters (trpABCDEG, trpBCDES, trpABCDR, trpABD, and trpBE), which are responsible for IAA production, also supports our findings from whole genome analysis of strains P. aeruginosa B18 (Singh et al., 2021), Rhizobacteria (Gupta et al., 2014), B. cereus T4S (Babalola et al., 2021), Sphingomona ssp. LK11 (Asaf et al., 2018), and Enterobacter roggenkampii ED5 (Guo et al., 2020). Similarly, we identified the genes involved in nitrogen fixation (nif3-like, gltP), nitrogen metabolism (gltX, glnR, glnA), and nitrogen regulation dissimulator nitrate (nadR, nirB, nirD, nasD, narl, narH, narJ, and nark) in the genome of HAB-5. Bacteria contain several iron transporters, e.g., Ybt, Feo, Efe, Yfe, Fet, and Fhu in Yersinia pestis (Forman et al., 2007). Among such iron transporters, the Fhu system—a siderophore receptor called FhuD—participates in siderophore (hydroxamate)-dependent iron (III) transport and was initially discovered in Escherichia coli (Kammler et al., 1993). In the present studies, several other genes (fbpA, fetB, feuC, feuB, feuA, and fecD) involved in iron transportation were also identified in HAB-5, according to the current research. Additionally, comparable findings were noted for the following bacteria: E. coli (FeoAB; Kammler et al., 1993), Shigella fexneri (FecIRABCDE; Luck et al., 2001), Staphylococcus (SirABC; Dale et al., 2004), and Yersinia pestis (Efe, Yfe, and Fet; Forman et al., 2007). Additionally, we discovered a few genes related to phosphate uptake and solubilization, some of which have been thoroughly investigated. These genes include pst (specific transporter) and pho (alkaline phosphate; Gebhard et al., 2006; Martín and Liras, 2021). The HAB-5 genome contained the phosphate solubilization-related genes (ispH, pstA, pstC, pstS, pstB, gltP, and phoH). Moreover, the genomes of Streptomyces and Mycobacterium tuberculosis have been revealed to contain pstA, pstC, pstS, and pstB, as well as phoA, phoC, and phoX, respectively (Gebhard et al., 2006; Martín and Liras, 2021). Furthermore, the HAB-5 genome contained genes associated with sulfur metabolisms (cysC, sat, cysK, cysS, and sulP), which have been found and described as sulfur metabolisms in Bacillus.