Genophenotypic insights into biocontrol and growth-promoting mechanisms of Bacillus velezensis BV25 against tomato Fusarium wilt

Abstract

Background

Fusarium wilt of tomato (Solanum lycopersicum L.), caused by Fusarium oxysporum f. sp. lycopersici (FOL), is a devastating soil-borne disease. Due to the limitations of conventional methods, sustainable biocontrol strategies are required. Bacillus velezensis is considered a promising biocontrol agent; however, the strain-specific nature of its antifungal and plant-growth-promoting mechanisms demands further investigation.

Results

We demonstrated the efficacy of B. velezensis BV25 against tomato Fusarium wilt. In vitro assays showed that FOL growth was significantly inhibited by both BV25 and its crude extracts; further, the inhibitory effect of the extracts was dose-dependent. In greenhouse trials, application of B. velezensis BV25 fermentation broth not only significantly suppressed the severity of Fusarium wilt but also promoted plant growth, even in the absence of the pathogen. The suppression of Fusarium wilt by B. velezensis BV25 treatment was correlated with the priming of plant defenses, as evidenced by the elevated activities of defense enzymes (superoxide dismutase, SOD; peroxidase, POD; catalase, CAT; phenylalanine ammonia-lyase, PAL). The complete genome sequencing of B. velezensis BV25, coupled with ANI analysis, determined its species designation. Our focus then turned to its functional genetic repertoire. We identified a diverse complement of Secondary metabolite (SM) biosynthetic gene clusters (BGCs) for known antimicrobials such as fengycin and surfactin, a suite of genes encoding fungal cell wall-degrading Carbohydrate-Active Enzymes (CAZymes), and multiple genes implicated in plant growth promotion including indole-3-acetic acid (IAA) synthesis, siderophore production, and nutrient uptake.

Conclusion

Our findings provide an extensive understanding of the biocontrol mechanisms of B. velezensis BV25, which include direct antagonism, induction of systemic resistance, and direct plant growth promotion. The integration of phenotypic evidence with genomic data positions B. velezensis BV25 as a robust candidate for developing sustainable biocontrol inoculants to manage tomato Fusarium wilt.

Introduction

Tomato (Solanum lycopersicum L.) is essential to global vegetable production, yet its cultivation is severely constrained by soil-borne diseases [1]. Among these, Fusarium wilt, caused by Fusarium oxysporum f. sp. lycopersici (FOL), is a major threat to tomato cultivation [2, 3]. Upon infection, FOL colonizes the xylem vessels, disrupting the transport of water and nutrients, which leads to wilting, yellowing, and ultimately, plant death, resulting in substantial economic losses [4, 5]. The reliance on chemical fungicides for managing Fusarium wilt faces significant challenges from the emergence of resistant pathogen strains and growing environmental concerns [6]. While the development of resistant tomato varieties offers a viable strategy, the rapid evolution of FOL races often renders this approach ineffective [7, 8]. Consequently, these limitations have driven the search for sustainable and eco-friendly alternatives [9], with biological control using beneficial microorganisms presenting a promising solution [10].

Bacillus velezensis is a well-characterized biocontrol agent that suppresses plant pathogens [11]. The synthesis of diverse bioactive secondary metabolites (SMs) [12], encoded by various biosynthetic gene clusters (BGCs) [13], leads to the production of compounds including lipopeptides such as surfactin, fengycin, and iturin, as well as polyketides, which collectively exhibit direct antifungal activity against pathogenic fungi [14, 15]. Furthermore, B. velezensis can compete for nutrients and space [16], produce cell wall-degrading enzymes [17], and induce systemic resistance (ISR) in plants [18], providing a multi-layered defense system. B. velezensis also functions as a plant growth-promoting rhizobacterium (PGPR), directly enhancing plant development through mechanisms such as phosphate solubilization, siderophore production for iron acquisition, and the synthesis of phytohormones [19], which are underpinned by characterized genetic modules including phosphate solubilization regulons, the bacillibactin siderophore cluster, and the trp operon for auxin biosynthesis [15, 20, 21]. The biocontrol and plant growth-promoting activities of B. velezensis are well-documented [15]; however, the genomic factors underlying its strain-specific biocontrol efficacy against particular pathogens such as FOL remain largely underexplored, and the specific genetic determinants of B. velezensis BV25 responsible for its antagonism against FOL, as well as its plant growth-promoting traits, also remain uncharacterized due to the high strain diversity. Therefore, elucidating the genomic basis of these traits is fundamental to deciphering its mode of action and advancing its commercial application [22].

This study aimed to evaluate the biocontrol efficacy of B. velezensis BV25 against tomato Fusarium wilt and to elucidate the underlying mechanisms. Previous studies have established that B. velezensis BV25 is an effective biocontrol agent against the southern root-knot nematode [23]. However, its potential for controlling tomato Fusarium wilt remains unexplored. To fill this research gap, we intended to verify the direct antagonistic activity against FOL through in vitro dual-culture assays and in vivo pot experiments, and to determine the activity of defense enzymes (superoxide dismutase, SOD; peroxidase, POD; catalase, CAT; phenylalanine ammonia-lyase, PAL) associated with systemic resistance. We planned to sequence and analyze the complete genome of B. velezensis BV25 to decipher the genetic basis of its multifaceted biocontrol activity, including the identification of biosynthetic gene clusters (BGCs) for SMs, Carbohydrate-Active Enzymes (CAZymes), and genes associated with plant growth promotion, as well as Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses.

Materials and methods

Microbial strains and plant materials

The biocontrol bacterial strain B. velezensis BV25, utilized in this study, was originally isolated from cucumber (Cucumis sativus L.) root systems sampled from a greenhouse at the experimental base of the Chinese Academy of Agricultural Sciences (CAAS) in Langfang, Hebei Province, China. This strain is maintained at the Institute of Vegetables and Flowers, CAAS (Beijing, China) and has been deposited in the China General Microbiological Culture Collection Center (CGMCC Accession No.16523), with routine storage at -80 °C for long-term preservation. Seeds of the tomato cultivar ‘Money Maker’ provided by the Institute of Vegetables and Flowers, CAAS (Beijing, China) were employed in all plant experiments.

Methods

Morphological characterization of B. velezensis BV25

For strain revival and morphological observation, the frozen B. velezensis BV25 stock was inoculated into Luria-Bertani (LB) broth (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl, pH 7.0) and incubated at 36 °C with agitation at 200 rpm for 24 h. The revived culture was then streaked for isolation onto LB agar plates and incubated at 36 °C for 48 h to observe colonial morphology.

To obtain sufficient biomass for scanning electron microscopy (SEM) analysis, the revived B. velezensis BV25 culture was inoculated into Yeast–Molasses–Corn flour (YMC) broth (20 g/L yeast powder, 20 g/L molasses, 20 g/L corn flour, pH 7.0) and cultured on a shaker at 36 °C for 5 days [24]. After incubation, microbial cells were harvested at 6,000 rpm for 15 min at 4 °C. The biomass was washed twice with sterile distilled water. Samples were fixed in 2.5% glutaraldehyde (dissolved in 0.1 M sodium phosphate buffer, pH 7.0) and incubated at 4 °C for 12 h. A graded ethanol dehydration series (30%, 50%, 70%, 90%, and 100%) was applied, followed by critical point drying to preserve delicate surface structures. Dried specimens were mounted onto aluminum stubs using conductive adhesive tape and sputter-coated with a thin layer of gold to enhance conductivity. SEM imaging was performed with a Hitachi S-800 electron microscope under an accelerating voltage of 3 kV, beam current of 10 µA, and a working distance of 8 mm. The entire procedure was performed in three independent biological replicates, and representative images were selected for analysis.

Culture conditions and fermentation

For subsequent greenhouse and antagonism assays, B. velezensis BV25 was initially cultivated on LB agar plates at 36 °C for 48 h. For seed culture preparation, a single colony was transferred to a 500 mL Erlenmeyer flask containing 200 mL of LB broth and incubated at 36 °C with shaking at 200 rpm for 24 h. This seed culture was subsequently used to inoculate the fermentation medium, which consisted of YMC broth in 500 mL Erlenmeyer flasks (200 mL working volume) [24]. The fermentation was carried out at 36 °C with agitation at 180 rpm for 5 days.

Inhibition of FOL by B. velezensis BV25

The direct antagonistic activity of B. velezensis BV25 against FOL was evaluated using a dual-culture assay. A mycelial plug of FOL (8 mm in diameter) was inoculated on one side of a Potato Dextrose Agar (PDA) (39 g/L, pH 5.6) plate, 2.5 cm from the center. Simultaneously, B. velezensis BV25 was streaked on the opposite side of the same plate, at an equal distance from the center. Control plates were inoculated with a FOL plug alone [25]. Each treatment was replicated three times. All plates were incubated at 36 °C for 7 days. Following incubation, the area of fungal colonies was measured using ImageJ software. The inhibition percentage of mycelial growth was calculated according to the formula:   [26].

Where: a - colony diameter of control group; b - colony diameter of treatment group.

Antifungal activity of bacterial crude extracts

Crude extracts from B. velezensis BV25 were obtained by ethyl acetate extraction. This involved extracting 1000 mL of fermentation broth two times with an equal volume of ethyl acetate. The combined organic phases were then concentrated under reduced pressure at 40 °C and 90 mmHg using a rotary evaporator, yielding approximately 60 mg of crude extract. To evaluate the antifungal activity, the crude extract was dissolved in methanol, and serial dilutions were prepared and incorporated into PDA medium to achieve final concentrations of 50, 150, 300, and 500 µg/mL. A mycelial plug of FOL was inoculated at the center of each treated PDA plate. Control plates contained PDA amended with an equivalent volume of methanol without the crude extract [27]. All plates were incubated at 28 °C for 7 days, with each concentration tested in three replicate plates. The inhibitory activity was quantified by measuring fungal colony areas using ImageJ software, and the percentage inhibition of mycelial growth was calculated as previously described in "Inhibition of FOL by B. velezensis BV25" section.

Greenhouse experiment

Tomato seedlings of the ‘Money Maker’ cultivar, a standard experimental variety with moderate yield potential that is susceptible to tomato Fusarium wilt, were grown in sterilized potting mix under controlled greenhouse conditions maintained at 26–28 °C, 65–75% relative humidity, and a 16 h light/8 h dark photoperiod [28]. Plants were divided into four treatment groups receiving different root drenching treatments (20 mL per plant): control check (CK) (sterile water), FOL (FOL conidial suspension + sterile water), FOL+BV25 (FOL conidial suspension + BV25 fermentation broth), and BV25 (BV25 fermentation broth). All biocontrol treatments were applied twice with a 3-day interval between applications. At the 2–3 true leaf stage, the FOL and FOL+BV25 groups were inoculated with FOL by applying 20 mL of a conidial suspension (1 × 10⁶ conidia/mL) to the soil of each pot [29]. The experiment was independently repeated three times. Within each run, each treatment included twenty tomato plants, resulting in a total sample size of sixty plants per treatment. Plant growth parameters were also measured at the end of the experimental period. Disease severity was evaluated using a 0–5 rating scale based on typical FOL-induced symptoms including leaf wilting, chlorosis, and tissue necrosis 21 days after pathogen inoculation [30], and the disease index was calculated as: [31].

Where: d - disease grade; n - number of plants at each grade; N - total number of plants; D_Max - highest disease grade.

Assay of defense-related enzymes

Assessment of the induced resistance response in tomato plants following treatment with B. velezensis BV25 fermentation broth involved quantifying the activities of defense-related enzymes in leaf tissues [32]. Leaf samples were collected from all treatment groups at 0, 12, 24, and 48 h after the final root drenching application. The fifth to eighth leaves from the base of the stem were selected, pooled randomly, and immediately frozen in liquid nitrogen. Frozen leaf tissues were ground to a fine powder using a pre-chilled mortar and pestle. The resulting powder was transferred to centrifuge tubes and homogenized in 0.02 M phosphate buffer (pH 7.4) at a ratio of 1:9 (w/v, tissue: buffer). The homogenate was vortexed for 1 min and centrifuged at 3000 × g for 4 min at 4 °C. The supernatant was collected for enzymatic assays. Enzyme activities were determined using commercial assay kits (BoxBio Science & Technology Co., Ltd., Beijing, China) according to the manufacturer’s protocols. Absorbance readings were obtained using a Synergy LX multi-mode microplate reader (BioTek Instruments, USA). All measurements were performed in triplicate, and enzyme activities were calculated based on the standard curves and formulas provided with the respective kits.

Genome sequencing assembly and annotation

The genomic DNA of B. velezensis BV25 was isolated and purified using the Invitrogen PureLink^® Genomic DNA kit. DNA concentration and quality were verified using the NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific), followed by additional purification with the Quick-DNA Miniprep Plus Kit. Whole-genome sequencing was performed by Shanghai BIOZERON Co., Ltd. using a combination of next-generation sequencing and third-generation sequencing technologies. For Illumina sequencing, paired-end libraries with insert sizes of approximately 400 bp were prepared following standard protocols. DNA fragmentation was achieved using Covaris, followed by end repair, adapter ligation, and PCR amplification. The completed library was sequenced to generate 150 bp paired-end reads. For PacBio sequencing, SMRTbell libraries were constructed using the Express Template Prep Kit 2.0 (Pacific Biosciences). Size selection was performed using Sage Sciences’ BluePippin with a cutoff of 8000 bp. The library was sequenced on the PacBio Sequel II platform.

Genome assembly was conducted using Unicycler with default parameters. The raw Illumina reads were quality-controlled using Trimmomatic with parameters “SLIDINGWINDOW:4:15 MINLEN:75”. PacBio reads were converted to FASTA format using Samtools. The assembly was circularized using Circlator, and genome characteristics including GC content and depth were analyzed using custom Perl scripts. Genome annotation was performed using GeneMark for gene prediction, which was selected due to its high accuracy for bacterial genomes and compatibility with downstream functional annotation pipelines. Functional annotation was carried out through BLASTP analysis against multiple databases including Non-Redundant Protein Database (NR), SwissProt, KEGG, GO, Clusters of Orthologous Groups (COG), as well as Antibiotic and Secondary Metabolite Analysis Shell (antiSMASH) v8.0.4. BGC similarity was classified as “High” at a homology threshold of ≥ 70%, a universal criterion for prokaryotic secondary metabolite analysis. Additional analyses included signal peptide prediction with SignalP and transmembrane helix identification with TMHMM. tRNA and rRNA genes were identified using tRNAscan-SE and RNAmmer, respectively. The circular genome map (Fig. 5) was generated using Circos software (implemented via the circlize package v0.4.4 in R v3.5.1) [33] based on the genome annotation described above.

Fig. 5.

Genome circular map of B. velezensis BV25. This is a circular map of the B. velezensis BV25 genome, where the central area shows the strain name and its genome size (3974396 bp). The outer regions of the map present two core pieces of information: (1) Classification and distribution of non-coding RNAs (including tRNA, 5 S rRNA, 16 S rRNA, 23 S rRNA, etc.) of BV25; (2) Functional annotation distribution of genes based on COG categories, with different colors corresponding to distinct COG functional classifications (specific correspondences can be referred to the legend), which intuitively reflects the functional distribution characteristics of genes in the BV25 genome

Average nucleotide identity (ANI) analysis

Average nucleotide identity (ANI) analysis was used to determine the definitive taxonomic placement of B. velezensis BV25. The complete genome sequence of B. velezensis BV25 was compared with those of representative Bacillus strains, with their corresponding GenBank accession numbers: B. velezensis FZB42 (CP000560), B. paralicheniformis BL-09 (CP010524), B. velezensis DH8043 (CP047268), B. subtilis 168 (NC_000964), B. halotolerans LN2 (CP126100), and B. licheniformis ATCC 14580 (AE017333). All reference genomes were obtained from the NCBI database. Pairwise ANI values were calculated using the FastANI software [34].

Statistical analysis

Statistical analyses were conducted in GraphPad Prism 10.31. Data meeting the assumption of homogeneity of variances (Brown-Forsythe and Bartlett’s tests, p > 0.05) were subjected to parametric tests. Comparisons among multiple groups were performed using one-way ANOVA followed by Tukey’s post hoc test; comparisons between two groups were made using unpaired Student’s t-test. Statistical significance was defined at p < 0.05. All graphical data are presented as mean ± SD.

Results

Morphological characterization of B. velezensis BV25

The colonial and cellular morphology of B. velezensis BV25 was characterized by colony observation on LB agar plates and SEM. When grown on LB agar plates, B. velezensis BV25 formed circular, off-white colonies with smooth margins and a uniform, creamy texture (Fig. 1A) [35]. Further ultrastructural observation via SEM revealed that the bacterial cells displayed a typical rod-shaped morphology, consistent with members of B. velezensis. The cells occurred primarily in aggregates, maintained structural integrity, and displayed an intact cell surface (Fig. 1B) [36]. This morphological characterization provides foundational insight into the cellular architecture of the biocontrol agent.

Fig. 1.

Morphological observation of B. velezensis BV25. A: Colony morphology of B. velezensis BV25 cultured on LB (Front and Back); B: SEM images of B. velezensis BV25, displaying the cellular morphology under different magnifications

In vitro antagonistic activity of B. velezensis BV25 Against FOL

Antagonism of bacterial culture

The direct antagonistic capability of B. velezensis BV25 against FOL was assessed using a dual-culture assay. The bacterial strain demonstrated strong antifungal activity, effectively restricting the mycelial growth of the pathogen (Fig. 2A). Treatment with B. velezensis BV25 resulted in a significantly reduced fungal colony area compared to the FOL control (p < 0.05), corresponding to a mycelial growth inhibition rate of approximately 31.4% (Fig. 2B).

Fig. 2.

Antifungal activity of B. velezensis BV25 against FOL. A: Dual culture assay of B. velezensis BV25 and FOL on PDA; B: Statistical analysis of the fungal colony area in the dual culture assay, where *** indicates an extremely significant difference between groups (p < 0.001). Data are presented as mean ± SD (n = 3 technical replicates); C: Antifungal effects of B. velezensis BV25 crude extract (at different concentrations) on FOL, observed on PDA; D: Statistical analysis of the inhibition rate of B. velezensis BV25 crude extract (at different concentrations) against FOL; different letters (A-D) above the columns indicate significant differences between groups (p < 0.05). Data are presented as mean ± SD (n = 3 technical replicates)

Inhibition by crude extracts

The antifungal effect of the crude extracts from B. velezensis BV25 was further evaluated by incorporating the extracts into PDA medium at different concentrations. The crude extracts exhibited a clear, dose-dependent suppression of FOL mycelial growth (Fig. 2C). The inhibition rates, quantified from the colony areas, were determined to be 36.8%, 61.8%, 77.4%, and 91.9% at concentrations of 50, 150, 300, and 500 µg/mL, respectively (Fig. 2D). These results indicate that the SMs produced by B. velezensis BV25 are major contributors to its antagonistic activity.

Greenhouse evaluation of biocontrol efficacy and plant growth promotion

The protective effect of B. velezensis BV25 against FOL was evaluated under greenhouse conditions. As illustrated in Fig. 3A, tomato plants inoculated with FOL alone exhibited severe disease symptoms, including stunted growth, leaf chlorosis, and a substantial reduction in root development. In contrast, plants treated with B. velezensis BV25 fermentation broth prior to FOL challenge (FOL+BV25) showed markedly reduced disease severity, maintaining healthier shoot and root systems. Consistent with the phenotypic observations, the disease index reached approximately 99% in the FOL-treated group, whereas the FOL+BV25 combination significantly suppressed disease development, achieving an index of only 17.6% (Fig. 3B). Assessment of plant physiological parameters confirmed a clear growth-promoting effect of BV25 (Fig. 3C; Table 1). Treatment with BV25 alone produced the highest values, significantly outperforming the CK with a plant height, root length, fresh weight, and dry weight of 8.31 cm, 10.37 cm, 13.28 g, and 4.03 g, representing increases of 10.8%, 39.2%, 22.9%, and 15.5% over CK (7.50 cm, 7.45 cm, 10.80 g, 3.49 g). When compared to the severely stunted FOL-inoculated plants (1.12 cm, 1.72 cm, 4.58 g, 1.53 g), this superiority was even more pronounced, with metrics reaching 7.4-, 6.0-, 2.9-, and 2.6-fold higher levels, respectively. Crucially, FOL+BV25 treatment effectively mitigated disease damage, restoring growth to 7.20 cm, 7.28 cm, 10.23 g, and 3.36 g—levels that were 6.4-, 4.2-, 2.2-, and 2.2-fold greater than those in the FOL-only group. These data robustly demonstrate the dual capacity of B. velezensis BV25 to promote plant growth and alleviate disease.

Fig. 3.

Greenhouse experiments of different treatment groups on tomato plants. A: Phenotypic performance of tomato plants in the pot experiment under different treatments; B: Statistical analysis of the disease index of tomato plants under different treatments; different letters (A-C) above the columns indicate significant differences between groups (p < 0.05). Data are presented as mean ± SD (n = 3 independent experiments); C: Statistical analysis of growth indicators (plant height, root length, dry weight, fresh weight) of tomato plants under different treatments; different letters (A-C) above the columns indicate significant differences between groups (p < 0.05). Data are presented as mean ± SD (n = 3 independent experiments)

Table 1.

Growth parameters of tomato plants under different treatments

Treatment	Plant height (cm)	Root length (cm)	Dry weight (g)	Fresh weight (g)
CK	7.50 ± 0.50 (0.11)	7.45 ± 0.47 (0.11)	3.49 ± 0.28 (0.06)	10.80 ± 0.86 (0.19)
FOL	1.12 ± 1.15 (0.26)	1.72 ± 1.04 (0.23)	1.53 ± 0.30 (0.07)	4.58 ± 0.91 (0.20)
BV25 + FOL	7.20 ± 0.30 (0.07)	7.28 ± 0.27 (0.06)	3.36 ± 0.29 (0.06)	10.23 ± 0.88 (0.20)
BV25	8.31 ± 0.38 (0.08)	10.37 ± 0.92 (0.21)	4.03 ± 0.27 (0.06)	13.28 ± 0.88 (0.20)

Values are expressed as mean ± standard deviation (SD), with standard error (SE) in parentheses

Induction of defense-related enzymes in tomato by B. velezensis BV25

SOD activity was most prominently induced in the FOL+BV25 treatment, reaching its peak of 394.6 U/g fresh weight (FW) at 24 h, a level significantly higher than the 309.3 U/g FW recorded in the FOL treatment at the same time point. The FOL-treated group itself reached its maximum SOD activity of 309.3 U/g FW at 12 h and subsequently plateaued, whereas the BV25 group exhibited a mild but steady increase from 0 to 48 h, attaining a final activity of 257.3 U/g FW (Fig. 4A).

Fig. 4.

Enzyme activity assays at 0, 12, 24 and 48 h post treatment. A: Activity of SOD in tomato plants under different treatments; B: Activity of POD in tomato plants under different treatments; C: Activity of CAT in tomato plants under different treatments; D: Activity of PAL in tomato plants under different treatments; different letters above columns indicate significant differences between groups (p < 0.05). Data for all enzyme activities are presented as mean ± SD (n = 3 independent experiments)

For POD activity, the FOL+BV25 treatment triggered a sharp increase, peaking at 24 h before stabilizing. When applied alone, BV25 induced a moderate rise that peaked at 24 h with an activity of 204.6 U/g FW, followed by a gradual decline. In contrast, the FOL-inoculated group showed an earlier peak, reaching 215.6 U/g FW at 12 h, after which activity slowly decreased (Fig. 4B). CAT activity demonstrated a progressive upward trend from 0 to 48 h in both the FOL+BV25 and BV25 groups, reaching final levels of 48.6 and 45.3 U/g FW, respectively. The FOL treatment, however, led to a transient peak in CAT activity of 41.7 U/g FW at 12 h, after which it declined (Fig. 4C).Similarly, PAL activity increased gradually over the 48-hour period in the FOL+BV25 and BV25 treatments, reaching 44.3 and 33.4 U/g FW, respectively. In the FOL-inoculated group, PAL activity peaked at 24 h and then began to decline (Fig. 4D).

Genome assembly and annotation of B. velezensis BV25

The genome of B. velezensis BV25 was sequenced using a hybrid approach combining PacBio and Illumina technologies. The final assembly has been deposited in the Genome Warehouse of the National Genomics Data Center (https://ngdc.cncb.ac.cn/gwh) under the accession number GWHHAIH00000000.1. The complete genome comprises a single circular chromosome of 3,974,398 bp and a circular plasmid of 7,276 bp, totaling 3,981,674 bp with a GC content of 46.54% (Fig. 5). Genome annotation predicted 4,164 protein-coding genes, 86 tRNA genes, and 27 rRNA genes (9 of 5 S, 9 of 16 S, and 9 of 23 S). A total of 2.46 Gb of Illumina reads and 170,549 PacBio reads (614.9 Mb) with a mean length of 3,605 bp (N50: 6,680 bp) were combined to produce this high-quality assembly (Fig. 5).

Genomic identification and taxonomic placement of strain BV25

Genomic identification confirmed strain BV25 as Bacillus velezensis, supported by ANI values exceeding the prokaryotic species threshold of 95–96% when compared to established type strains. The highest similarities were observed with B. velezensis FZB42 (97.47%) and B. velezensis DH8043 (97.24%). Overall, the ANI values between BV25 and other closely related Bacillus strains, including B. paralicheniformis, B. subtilis, B. licheniformis, and relatives, ranged from 78.77% to 97.47% (Fig. 6).

Fig. 6.

Heatmap of ANI values based on whole-genome sequences of B. velezensis BV25 and other Bacillus strains. This heatmap displays the ANI values between B. velezensis BV25 and multiple Bacillus strains (including B. velezensis FZB42, B. paralicheniformis BL09, etc.). The color gradient on the right corresponds to the ANI percentage, and the specific ANI values are labeled in each cell, which intuitively reflects the genomic similarity between B. velezensis BV25 and the tested Bacillus strains

Functional annotation and metabolic pathway profiling

GO classification assigned putative functions to the protein-coding genes of B. velezensis BV25 across three primary categories (Fig. 7A). Within Biological Processes, metabolic processes constituted the most predominant group. Molecular Functions were predominantly associated with catalytic activity and transporter activity, while Cellular Components represented the smallest proportion of annotations. This functional profile highlights the genome’s strong bias toward metabolic versatility, evidenced by the concurrent enrichment of genes involved in metabolic processes and those encoding catalytic and transport functions.

Fig. 7.

Functional enrichment analysis of B. velezensis BV25 genes. A: GO functional enrichment analysis of B. velezensis BV25 genes, covering three core categories: biological process, cellular component, and molecular function. The upper bar chart reflects the number of genes corresponding to each GO term, and the lower part lists the specific GO functional terms under each category; B: KEGG pathway enrichment analysis of B. velezensis BV25 genes; the bar chart displays the number of genes enriched in different metabolic/functional pathways, with the horizontal axis indicating the names of the enriched KEGG pathways

Complementary KEGG pathway analysis further elucidated the strain’s metabolic capacity, with metabolic pathways being the most populated category (Fig. 7B). The genomic identification of numerous BGCs was reflected in the functional analysis, where the biosynthesis of secondary metabolites ranked as the second most abundant pathway. These annotation results collectively underscore the genetic foundation for the strain’s prolific metabolic activity, particularly its pronounced capability for synthesizing diverse SMs.

Predicted biosynthetic genes of secondary metabolites in B. velezensis BV25

AntiSMASH analysis uncovered an impressive arsenal of 14 secondary metabolite BGCs within the B. velezensis BV25 genome (Table 2). Crucially, 8 of these clusters showed high similarity to gene clusters for known antimicrobial compounds. These included the potent antifungal lipopeptides fengycin and surfactin; the polyketide antibiotics bacillaene, difficidin, and macrolactin H; the siderophore bacillibactin; the dipeptide antibiotic bacilysin; and the bacteriocin plantazolicin. The presence of these BGCs provides a comprehensive genetic basis for the strain’s broad-spectrum antimicrobial activity and its efficacy in biocontrol.

Table 2.

Predicted BGCs in the genome of B.velezensis BV25, identified by antiSMASH analysis

Clusters	Types	Cluster Range	Most Similar Known Cluster	Similarity
Cluster 1.1	NRPS	304,747–370,154	surfactin	High
Cluster 1.2	azole-containing-RiPP	584,120–613,284
Cluster 1.3	RRE-containing, azole-containing-RiPP	696,389–719,566	plantazolicin	High
Cluster 1.4	PKS-like	929,618–970,862
Cluster 1.5	terpene	1,052,897–1,073,637
Cluster 1.6	transAT-PKS	1,377,018–1,465,323	macrolactin H	High
Cluster 1.7	transAT-PKS, T3PKS, NRPS	1,686,732–1,796,864	bacillaene
Cluster 1.8	NRPS, transAT-PKS, betalactone	1,854,327–1,992,156	fengycin	High
Cluster 1.9	terpene	2,055,098–2,076,981
Cluster 1.10	T3PKS	2,145,612–2,186,718
Cluster 1.11	transAT-PKS	2,302,201–2,408,367	difficidin	High
Cluster 1.12	terpene-precursor	2,431,679–2,452,569
Cluster 1.13	terpene-precursor, NRP-metallophore, NRPS, RiPP-like	3,060,899–3,126,312	bacillibactin	High
Cluster 1.14	other	3,643,321–3,684,739	bacilysin	High

Genomic mining of CAZymes reveals the antifungal potential of B. velezensis BV25

We conducted a comprehensive analysis of the B. velezensis BV25 genome, revealing a rich repertoire of CAZymes, with a total of 144 genes identified. These included 47 glycoside hydrolases (GHs), 39 glycosyltransferases (GTs), 31 carbohydrate esterases (CEs), 22 carbohydrate-binding modules (CBMs), 4 polysaccharide lyases (PLs), and 8 auxiliary activities (AAs). Significantly, several genes were predicted to encode enzymes targeting key structural components of fungal cell walls, such as chitin (GH18) and β-1,3-glucan (GH16), underpinning the strain’s observed strong antagonistic activity against FOL.

Genomic mining reveals the plant growth-promoting potential of B. velezensis BV25

Genomic analysis of B. velezensis BV25 uncovered a diverse repertoire of genes with putative plant growth-promoting functions, pivotal for root colonization, nutrient acquisition, and phytohormone modulation. The initial step of root colonization is supported by the presence of a core set of chemotaxis genes, including cheA, cheB, cheC, cheD, cheR, cheW, and cheY. Subsequently, the strain possesses genetic determinants for enhancing nutrient availability, featuring genes for nitrate/nitrite assimilation (nasA, nasB, nasD, nasE, nasF), potassium uptake (ktrA, ktrB, ktrC, ktrD), and phosphate metabolism, including the pstABCS transporter and associated regulators (phoA, phoB, phoD, phoE, phoL, phoR, phoP). Furthermore, the genomic inventory revealed pathways for the synthesis of various phytohormones and growth-stimulating metabolites, such as indole-3-acetic acid (IAA) involving genes from the trp pathway (trpA, trpB, trpD, trpE, trpS) and yclC, cytokinins (miaA, miaB), polyamines (speA, speB, speE, speH), and acetoin/2,3-butanediol (alsD, alsS, alsR, ilvA, ilvB, ilvC, ilvD, ilvE, ilvG, ilvH, ilvI, ilvK, ilvN, ilvX, bdhA).

Discussion

FOL severely limits sustainable tomato production worldwide [37]. Here, we employed an integrated phenotype-to-genotype strategy to define the biocontrol efficacy of B. velezensis BV25 against FOL and to decipher this strain’s functional basis [38]. Our results established that B. velezensis BV25 suppresses Fusarium wilt by concurrently deploying direct antagonism, inducing systemic resistance, and promoting plant growth. The complete genome sequence of B. velezensis BV25 elucidated the genetic underpinnings of these traits, encoding the specific antimicrobial compounds and growth-promoting pathways responsible for its observed efficacy [39, 40].

To ensure accurate genomic annotation, we first resolved the taxonomy of strain BV25. Genomic analysis revealed an ANI of > 97% with B. velezensis FZB42, which definitively classified it within this species and provided the solid foundation for comparing its functional genomics with other strains of this species [41].

Initial in vitro assays established that both B. velezensis BV25 cells and their crude extract significantly suppressed the mycelial growth of FOL, with the crude extract exhibiting a clear dose-dependent inhibition. This direct antagonism is strongly supported by the genomic evidence. The B. velezensis BV25 genome encodes numerous genes for antimicrobial activity, among which is a substantial set of CAZymes, including genes from the GH18 and GH16 families, known to target key fungal cell wall components such as chitin and β-1,3-glucan [42, 43]. The presence of a suite of cell wall-degrading enzymes is a common trait among many PGPR strains [43–45]. Furthermore, genome mining via antiSMASH revealed an impressive capacity for SMs, predicting 14 BGCs. Crucially, eight of these BGCs show high similarity to known antimicrobial compounds, including the potent antifungal lipopeptides surfactin and fengycin [46, 47], as well as polyketide antibiotics like bacillaene and difficidin [48–50]. The genome of B. velezensis BV25 is distinguished by a notably expanded and diverse repertoire of biosynthetic BGCs and CAZymes, which constitutes a key strain-specific feature. While the model strain FZB42 possesses 9 BGCs and 91 CAZymes [13, 51], B. velezensis SQR9 harbors 12 BGCs, B. velezensis Lzh-5 contains 13 BGCs and 131 CAZymes [52], B. velezensis CRN 2 and CRN 23 encode 74 and 69 CAZymes respectively (with BGCs encoding bacilysin and non-ribosomal peptides) [53], B. velezensis Jin12 features 168 CAZymes [54], and B. velezensis BV25 encodes 14 BGCs and 144 CAZymes. Notably, this reflects a substantial genomic expansion: BV25 boasts a leading BGC count among closely related strains—surpassing FZB42 (9), SQR9 (11), and Lzh-5 (13)—while its CAZyme complement is over 58% more abundant than FZB42’s, outperforms CRN 2, CRN 23, and Lzh-5, and remains highly competitive even against the Jin12 (168 vs. 144). This distinctive genetic endowment underpins B. velezensis BV25’s robust biocontrol capabilities [55, 56]. The additional BGCs likely contribute to its strong antagonistic activity, potentially through novel or enhanced production of antimicrobial compounds such as surfactin, fengycin. Concurrently, the vastly enlarged CAZyme toolkit equips BV25 with superior capacity for environmental interaction and pathogen inhibition, including the degradation of fungal cell walls. Supporting this, KEGG pathway analysis revealed biosynthesis of secondary metabolites as the second most abundant metabolic category, while GO classification showed predominant enrichment in catalytic activity and metabolic processes, which is consistent with the strain’s observed capacity for antibiotic production [57].

In addition to direct pathogen suppression, our greenhouse trials and defense enzyme analyses demonstrated that B. velezensis BV25 also induces systemic resistance in the host plant through the production of specific elicitors. Tomato plants treated with B. velezensis BV25 prior to FOL challenge exhibited a marked reduction in disease severity, correlating with a sharp increase in the activities of key defense enzymes—SOD, POD, CAT, and PAL, in the leaves. Notably, the elevated PAL activity is of particular significance, as PAL is the gateway enzyme of phenylpropanoid metabolism leading to lignin biosynthesis. Enhanced lignin deposition fortifies the plant cell wall, creating a more robust physical barrier against pathogen invasion [58]. This defense enzyme profile is consistent with ISR [59]. We interpret this ISR as a microbe-driven phenotype, wherein bacterial SMs activate the plant’s immune system [60]. Thus, B. velezensis BV25 elicits and primes the plant’s defensive capacity. The induction of this suite of defense enzymes is consistent with the pattern of Bacillus-mediated induced systemic resistance reported in previous studies [61, 62]. Future work employing signaling pathway mutants could definitively test this hypothesis and delineate the specific nodes in the defense network that are primed by B. velezensis BV25. In the absence of pathogen pressure, B. velezensis BV25 still enhanced tomato plant growth, significantly increasing biomass. This growth-promoting capacity is consistent with the well-documented performance of other B. velezensis strains: BS1 enhances root length, plant height, and root fresh weight in pepper [63]; CE100 improves chlorophyll content, nitrogen/phosphorus uptake, and biomass accumulation in walnut while also increasing strawberry fruit yield [64]; RCDL12 boosts shoot length and seed germination rate in rice and mung bean [65]; CH1 promotes oat growth through the production of indole-3-acetic acid (IAA) and biofilm formation [66]; and HNH9 upregulates growth-related genes, thereby increasing the photosynthetic rate and biomass of upland cotton [64]. Genomic analysis revealed a complete pathway for the synthesis of IAA, the gene cluster for the siderophore bacillibactin, and efficient systems for phosphate and potassium uptake and assimilation [67–71]. Accordingly, B. velezensis BV25 likely stimulates plant growth through multiple direct mechanisms: phytohormone production, improved iron acquisition, and enhanced nutrient availability [72]. This direct growth promotion, by generating a more robust plant, potentially creates a positive feedback loop with its biocontrol activities. Such synergistic interactions between different plant-beneficial traits are characteristic of highly effective PGPR, as the combination of multiple mechanisms typically provides more robust and consistent plant growth promotion than single mechanisms alone [73]. It is noteworthy that the plant growth-promoting capacity of Bacillus spp. is highly strain-dependent, and inconsistent results from certain strains have been reported in the literature [56]. The fact that B. velezensis BV25 consistently and significantly enhanced tomato biomass, and that this phenotype is underpinned by a complete set of core beneficial genes, underscores its distinct value and genetic stability as a promising PGPR candidate. This ultimately led to improved overall plant health [74].

While this study established a robust correlation between the genome of B. velezensis BV25 and its phenotypic effects, a deeper, systems-level understanding of its biocontrol activity warrants further investigation. The precise contribution of specific antimicrobial gene clusters to the overall suppression of FOL could be further unraveled by profiling their expression dynamics and the corresponding metabolite production in planta. Furthermore, translating these promising greenhouse results into field efficacy necessitates evaluating B. velezensis BV25’s performance and survival within complex soil microbiomes and under fluctuating environmental conditions. Addressing these aspects will be key to optimizing B. velezensis BV25’s application and solidifying its role in sustainable agriculture [75, 76].

Conclusion

In conclusion, this study presents the first integrated genomic and phenotypic analysis of B. velezensis BV25, systematically characterizing its role as a multi-talented biocontrol agent against tomato Fusarium wilt. Through definitive ANI analysis, we have firmly established its taxonomic position within the B. velezensis species. Its efficacy is mediated by a synergistic combination of direct antagonism via antimicrobials and hydrolytic enzymes, induction of systemic resistance, and direct plant growth promotion. The functional annotation through KEGG and GO analyses further corroborates the strain’s metabolic versatility and biocontrol potential. The co-existence of these functional genetic modules in the B. velezensis BV25 genome underscores its potential as a promising microbial inoculant. Future work will focus on developing stable formulations based on B. velezensis BV25 and validating its efficacy under field conditions, with the goal of translating this strain into a practical tool for sustainable tomato production.

Authors’ contributions

T.X. conceived and designed the study, prepared the experimental materials, conducted bioinformatics analysis, prepared the experimental materials, performed the experiments and data analysis, wrote the manuscript draft, and prepared the figures and tables. X.D. participated in experimental discussions and revised the manuscript. J.Z., J.L., H.Y., and Z.M. participated in the study design. B.X., L.Z., and Y.L. provided support for manuscript review and offered financial support. All authors read and approved the final version of the manuscript.

Funding

This research was funded by grants from the National Key R&D Program of China (2022YFD1400700), the Agricultural Science and Technology Innovation Program of CAAS, as well as the China Agriculture Research System (CARS-25).

Data availability

The genome sequence data generated during the current study are available in the Genome Warehouse of the National Genomics Data Center (https://ngdc.cncb.ac.cn/gwh) under accession number GWHHAIH00000000.1.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.