Antifungal Activity of Bacillus amyloliquefaciens X30 Against Botrytis fabiopsis 3-3 on Panax notoginseng and Its Mechanism

Abstract

Gray mold disease severely impacts the yield and quality of Panax notoginseng (Burkill) F. H. Chen ex C. Chow & W.G. Huang. In this study, a strain of Botrytis fabiopsis J. Zhang, G.N. Wu & G.Q. Li labeled as 3-3 was isolated from the leaves affected by gray mould disease of P. notoginseng, identified as a novel pathogen for this plant. Targeting the strain 3-3, an antagonistic bacterial strain X30 was isolated from the leaves of P. notoginseng and was preliminarily identified as Bacillus amyloliquefaciens (Fukumoto) Priest et al. through morphological and molecular biological analyses. The in vitro antifungal test showed that strain X30, at a concentration of 1 × 10⁸ CFU mL⁻¹, had an inhibition rate of 84.63% against the B. fabiopsis strain 3-3, and it exhibited broad-spectrum antifungal activity against other major pathogenic fungi of P. notoginseng, including Alternaria alternata (Fr.) Keissl., Rhizoctonia solani J.G. Kühn and others. Additionally, strain X30 was found to produce ammonia, fix nitrogen, secrete plant growth hormones, and release multiple hydrolytic enzymes, thus possessing both plant-growth-promoting and antimicrobial traits. In pot experiments, an X30 suspension at 1 × 10⁸ CFU mL⁻¹ achieved 61.04% control rate against B. fabiopsis. Using non-targeted metabolomics, compounds in the culture filtrate of strain X30 were analyzed, and two organic acid compounds with antimicrobial activity were identified. Among them, phenylpyruvic acid had an EC₅₀ value of 312 µg mL⁻¹ against pathogen 3-3, while 2,6-dihydroxybenzoic acid had an EC₅₀ value of 660 µg mL⁻¹. B. amyloliquefaciens X30 provides a theoretical basis for developing green and efficient biocontrol agents against gray mould in P. notoginseng.

1. Introduction

Panax notoginseng (Burkill) F. H. Chen ex C. Chow & W.G. Huang, a member of the genus Panax L. in the family Araliaceae Juss., is a highly valued traditional Chinese medicine renowned for its rich content of dammarane-type tetracyclic triterpenoid saponins. Owing to its remarkable properties in promoting blood circulation, arresting bleeding, reducing swelling and alleviating pain, it has become a major component in several well-known pharmaceutical preparations, including Yunnan Baiyao, Compound Danshen Tablets and Sanqi Trauma Tablets [1,2]. Panax notoginseng has a long growth cycle and requires specific cultivation conditions. Despite recent advancements in planting and processing technologies, disease pressure remains a significant challenge, affecting not only the yield but also the quality of P. notoginseng [3].

Above-ground parts of P. notoginseng in the field are affected by gray mold, black spot, anthracnose, powdery mildew and other diseases [4]. Among them, gray mould is the most common cause of early seedling death. Gray mould in P. notoginseng is caused by Botrytis cinerea Pers.: Fr. The disease affects leaves from April to June, worsening during the rainy season, leading to plant wilting and death [5]. Effective control of gray mould in P. notoginseng has thus become a crucial issue for the development of the P. notoginseng industry. Botrytis cinerea has a high potential for gene drift and a large population size [6]. The long-term use of chemical fungicides can lead to the development of resistance in B. cinerea, resulting in increased dosage requirements. This not only raises economic costs but also disrupts biodiversity [7,8]. Resistance to commonly used fungicides such as chlorothalonil, carbendazim, and thiophanate-methyl has already been reported in Botrytis spp. [9,10,11]. Therefore, the development of more environmentally friendly and safe methods for controlling gray mould in P. notoginseng is of great importance [12].

Biological control, a green method of plant disease management, holds great potential for increasing agricultural productivity, enhancing crop resilience, and suppressing pathogen growth, with biopesticides exerting long-lasting effects after application [13,14]. Beneficial microorganisms can effectively suppress the growth of pathogenic Botrytis species. For instance, microbial fermentation broth and secondary metabolites have been shown to inhibit spore germination and mycelial growth, thereby preventing disease onset in various crops [15,16,17]. Metabolomics analysis further reveals that Bacillus species produce key biocontrol compounds such as Iturin A, Fengycin, and Surfactin via polyketide synthase (PKS) and nonribosomal peptide synthetase (NRPS) pathways, offering effective protection against fungal pathogens [16,18,19].

Current research on biocontrol bacteria for P. notoginseng primarily focuses on pathogens causing root rot and black spot disease, with limited reports on gray mold pathogens [16,20,21,22]. To address the lack of effective biological control strategies for P. notoginseng gray mould, this study screened phyllosphere bacteria and identified a Bacillus amyloliquefaciens (Fukumoto) Priest et al. strain X30 with potent antagonistic activity against Botrytis fabiopsis J. Zhang, G.N. Wu & G.Q. Li. Through integrated metabolomic and functional analyses, we elucidated the underlying antifungal mechanisms and bioactive compounds, providing a theoretical basis for developing microbial alternatives to chemical fungicides.

2. Materials and Methods

2.1. Identification of Pathogens

The pathogenic strain 3-3 was isolated from gray-mold lesions on P. notoginseng leaves collected in Wenshan Prefecture, Yunnan, China, and its pathogenicity was confirmed according to Koch’s postulates. Genomic DNA was amplified with the fungal universal primers ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′), as well as the specific primers RPB2-F (5′-GATGATCGTGATCATTTCGG-3′) and RPB2-R (5′-CCCATAGCTTGCTTGCTTACCCAT-3′) targeting the RPB2 gene. PCR (Polymerase Chain Reaction) amplification was conducted in a 20 μL reaction mixture containing 10 μL of 2× Taq PCR StarMix, 1 μL each of forward and reverse primers (10 μM), 1 μL of DNA template, and 7 μL of ddH₂O. The thermocycling conditions for the ITS region were as follows: initial denaturation at 94 °C for 2 min; 36 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min; followed by a final extension at 72 °C for 5 min. For the RPB2 gene, the conditions were: initial denaturation at 94 °C for 2 min; 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min; with a final extension at 72 °C for 10 min. The resulting sequences were queried against the NCBI (National Center of Biotechnology Information) nucleotide database using BLAST database version 5 (Basic Local Alignment Search Tool), and a neighbor-joining phylogenetic tree was constructed with MEGA 7.0 to achieve preliminary identification of the isolate.

2.2. The Antifungal Activity of Strain X30 Against Botrytis fabiopsis and Other P. notoginseng Pathogens

The antimicrobial activity of strain X30 was evaluated against a panel of major pathogens associated with P. notoginseng diseases, including B. fabiopsis strain 3-3, B. cinerea strain 6-6, Alternaria alternata (Fr.) Keissl. strain SC-Z7, Rhizoctonia solani J.G. Kühn strain 7236, Fusarium solani (Mart.) Sacc. strain 4023, Fusarium oxysporum Schltdl. strain 4151, and Ilyonectria robusta (Berk. & Broome) Rossman, L. Lombard & Crous strain 6695 using a dual-culture confrontation assay. The strains of B. fabiopsis 3-3, B. cinerea 6-6, and A. alternata SC-Z7 were isolated and identified from diseased P. notoginseng leaves collected in Wenshan, Yunnan Province. The strains of R. solani 7236, F. solani 4023, F. oxysporum 4151, and I. robusta 6695 were provided by the laboratory of Weiwei Gao.

A 5 mm mycelial plug taken from the actively growing margin of B. fabiopsis strain 3-3 was placed in the center of a fresh 90 mm PDA (Potato Dextrose Agar) plate. Four 10 µL droplets of X30 bacterial suspension (adjusted to 1 × 10⁸ CFU mL⁻¹ based on hemocytometer counts) were deposited at equidistant positions 2 cm away from the plug to create a symmetrical confrontation zone. Control plates received the fungal plug only, and each treatment was replicated three times.

All plates were incubated in the dark at 25 °C; when the control mycelium had reached approximately three-quarters of the Petri-dish diameter, images of 3-3 hyphae were captured with an Axiocam 512 color camera mounted on an Axio Imager A2 upright microscope (Carl Zeiss, Oberkochen, Germany), and colony diameter of B. fabiopsis was measured perpendicular to the confrontation line for both treated and control plates. The percentage inhibition of mycelial growth (PGI) was calculated as follows [12]:

PGI (%) = [(Dc − Dt)/Dc] × 100

where Dc is the colony diameter of the control, and Dt is the colony diameter of the treatment.

2.3. Identification of X30 Strain

Morphological, physiological and biochemical traits of strain X30 were characterized on LB (Lysogeny Broth) agar plates incubated at 28 °C; colony topography, margin type, pigmentation and growth kinetics were recorded daily for 7 d. For molecular identification, the 16S rDNA gene was amplified with universal bacterial primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′), and the recombinase A (recA) locus was targeted with designed primers recA-F (5′-GATCGTCAAGCAGCCTTAGAT-3′) and recA-R (5′-TTACCGACCATAACGCCGAC-3′).

PCR amplification was performed in a 20 μL reaction mixture containing 10 μL of 2× Taq PCR StarMix, 1 μL each of forward and reverse primers (10 μM), 1 μL of DNA template, and 7 μL of ddH₂O. The thermal cycling conditions for the 16S rDNA gene were as follows: initial denaturation at 94 °C for 5 min; 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 45 s, and extension at 72 °C for 1 min; followed by a final extension at 72 °C for 5 min. For the recA gene, the conditions were: initial denaturation at 95 °C for 5 min; 30 cycles of denaturation at 95 °C for 30 s, annealing at 56 °C for 30 s, and extension at 72 °C for 1 min; with a final extension at 72 °C for 5 min. The obtained sequences were subjected to BLAST analysis against the NCBI nucleotide database, and a neighbor-joining phylogenetic tree was constructed using MEGA 7.0 software to determine the taxonomic position of the strain.

2.4. Biological Function of Strain X30

To preliminarily elucidate the antimicrobial and plant-growth-promoting capabilities of strain X30, we conducted the following experiments. Nitrogen-fixtion ability was tested on Ashby’s solid medium (26.7 g L⁻¹; Coolaber); Phosphate solubilization was assessed using Menchikova’s P agar: organic phosphorus medium (31.66 g L⁻¹; Coolaber) and inorganic phosphorus medium (26.7 g L⁻¹; Coolaber); Potassium solubilization was evaluated on a potassium-solubilizing medium containing (per liter): 5.0 g potassium feldspar powder, 2.0 g sucrose, 1.5 g Na₂HPO₄, 0.5 g (NH₄)₂SO₄, 0.5 g MgSO₄·7H₂O, and 20.0 g agar; Siderophore secretion was detected on modified Chrome Azurol S (CAS) solid medium (Coolaber); Ammonia production was determined in peptone water (5.0 g L⁻¹ peptone) using Nessler’s reagent [23]. ACC deaminase (ACCD) activity was measured using a commercial enzyme assay kit, while indole-3-acetic acid (IAA) and gibberellin (GA) concentrations were determined using commercial enzyme-linked immunosorbent assay (ELISA) kits [24,25].

Extracellular enzymatic activities of amylase, cellulase, and protease were assessed on specific substrate media: starch agar (5.0 g beef extract, 10.0 g peptone, 5.0 g NaCl, 1.0 g soluble starch, and 15.0 g agar per liter), carboxymethyl-cellulose (CMC) agar (10.0 g sodium carboxymethyl cellulose, 10.0 g peptone, 15.0 g NaCl, 1.0 g KH₂PO₄, and 15.0 g agar per liter), and skim-milk agar (10.0 g skim milk, 10.0 g peptone, 5.0 g NaCl, 0.1 g CaCl₂, and 15.0 g agar per liter), respectively. Additionally, chitinase and β-1,3-glucanase activities were quantified using commercial enzyme assay kits [23].

2.5. Antifungal Activity Test of Metabolites of X30 Strain

2.5.1. Inhibitory Effect of Volatile Organic Compounds (VOCs) on 3-3 Strain

A two-plate assay was used to evaluate the volatile-mediated antifungal activity of strain X30. One hundred microlitres of cell suspension prepared from 24 h LB cultures were evenly spread on LB agar to give final inocula of 1 × 10⁸ or 1 × 10⁴ CFU mL⁻¹. After 24 h incubation at 28 °C, a 5 mm mycelial plug taken from the advancing edge of a 4-day-old PDA culture of strain 3-3 was placed in the center of a fresh PDA plate. The LB plate was inverted and sealed face-to-face with the PDA plate using Parafilm, creating an airtight compartment that allowed only volatile metabolites to diffuse. Control consisted of PDA plates paired with sterile LB agar without bacteria. Each treatment was replicated four times. All paired plates were incubated in darkness at 25 °C for 4 d, after which radial growth of the fungus was measured along two perpendicular axes. The inhibition rate of mycelial growth by volatile organic compounds was calculated as described in Section 2.2.

2.5.2. Inhibitory Effect of Sterile Fermentation Broth on 3-3 Strain

A single colony of X30 was inoculated into 5 mL of LB liquid medium and cultured at 37 °C with shaking at 180 rpm (revolutions per minute) for 24 h. The culture was then resuspended in LB liquid medium to a concentration of 1 × 10⁸ CFU mL⁻¹ and inoculated into a 250 mL Erlenmeyer flask containing 100 mL of LB liquid medium at a 1% (v/v) inoculum volume. The flask was incubated at 28 °C with shaking at 180 rpm for 48 h. The culture broth was centrifuged at 10,000 rpm for 10 min to collect the supernatant, which was then filtered through a 0.22 µm Millipore filter membrane to obtain a cell-free filtrate. The inhibitory effect of the filtrate on the pathogen strain 3-3 was evaluated using the Oxford cup method, with sterile LB liquid medium as the control. Each experiment was repeated four times, and the plates were incubated in darkness at 25 °C for 4 days. Colony diameters were measured using the cross method. The calculation method for the inhibition rate of mycelial growth by the cell-free filtrate is as described in Section 2.2.

2.6. Preventive Effect of Strain X30 Against P. notoginseng Gray Mold

To evaluate the practical biocontrol efficacy of strain X30, a foliar infection assay was performed using the pathogen strain 3-3 on P. notoginseng leaves pretreated with X30 suspension. The experiment was conducted in the light culture chamber at the Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences. One-year-old P. notoginseng seedlings with uniform growth were transplanted into plastic pots containing a sterilized soil mixture of pine needle soil and black soil at a 2:1 ratio, one seedling per pot. Seven days after transplantation, leaves were surface-sterilized with 75% ethanol before treatment. The X30 suspension was adjusted to 1 × 10⁸ CFU mL⁻¹ in sterile water and sprayed evenly onto leaves until runoff. On each plant, seven leaves were inoculated with a 5 mm mycelial plug of strain 3-3 and enclosed in a plastic bag to maintain humidity. The treatments included: (i) spraying with sterile water (negative control, CK0); (ii) inoculating with strain 3-3 alone (positive control, CK1); (iii) a safety control treated solely with X30 suspension to assess phytotoxicity; (iv) Inoculation with strain 3-3 after spraying with the X30 bacterial suspension (treatment group, 3-3 & X13). Each treatment was replicated six times. Plants were maintained at 24 °C, and the experiment was terminated after three days. Photographs were taken of the diseased leaves, and the lesion area was measured using ImageJ software (version 1.54g; National Institutes of Health, Bethesda, MD, USA). Disease severity was expressed as the relative lesion area per plant. The protective efficacy was calculated as follows [26]:

Protective efficacy (%) = [(Ac − At)/Ac] × 100

where Ac is the mean relative lesion area of the control plants, and At is the mean relative lesion area of the X30-treated plants.

2.7. Ultra-High Performance Liquid Chromatography-Tandem Mass Spectrometry (UHPLC-MS/MS) Analysis of the Secondary Metabolites of X30 Strain

The sample preparation method for the experimental group was identical to that described in Section 2.5.2 for the cell-free filtrate of strain X30, with sterile LB liquid medium serving as the control. An accurately measured sample volume was combined with 400 µL of cold methanol solution (methanol:water = 4:1) and homogenized at low temperature using a high-throughput tissue homogeniser. After vortexing, the mixture was subjected to ultrasonic extraction on ice for 30 min, followed by incubation at −20 °C for 30 min. The sample was then centrifuged at 13,000 rpm (4 °C, 15 min), and the supernatant was transferred to an LC-MS vial with an insert for analysis. The LC-MS analysis was performed using an SCIEX UPLC-Triple Quadrupole Time-of-Flight system. The chromatographic separation was achieved on a BEH C18 column (100 mm × 2.1 mm i.d., 1.7 µm; Waters, Milford, CT, USA). The mobile phase A consisted of water with 0.1% formic acid, and mobile phase B was acetonitrile/isopropanol (1:1) with 0.1% formic acid. The gradient elution program was as follows: 0–3 min, 0–20% B; 3–9 min, 20–60% B; 9–11 min, 60–100% B; 100% B for 2.5 min; 13.5–13.6 min, 100% to 0% B; and 0% B for 2.4 min. The flow rate was 0.40 mL min⁻¹, the injection volume was 20 µL, and the column temperature was maintained at 40 °C. Mass spectrometry was performed in both positive and negative ion modes. The electrospray capillary voltage was set at 1.0 kV, with a sampling cone voltage of 40 V and a collision energy of 6 eV. The ion source temperature was 120 °C, and the desolvation temperature was 500 °C. The gas flow rate was 900 L h⁻¹, and the mass spectrometry scan range was from 50 to 1000 m z⁻¹, with a resolution of 30,000. Quality control (QC) samples, prepared by pooling equal volumes of all analyzed samples, were inserted every eight analytical runs to monitor system stability.

Raw data were imported into the metabolomics processing software Progenesis QI (v. 2.3, Waters Corporation, Milford, CT, USA) for baseline filtering, peak detection, integration, retention-time alignment and peak alignment. Subsequent data preprocessing generated the final data matrix used for downstream analyses. Compound annotation was performed by matching accurate mass and MS/MS fragmentation patterns embedded in the same software against public repositories (HMDB, https://hmdb.ca/ (accessed on 1 May 2025); METLIN, https://metlin.scripps.edu/ (accessed on 1 May 2025)) supplemented with an in-house spectral library.

2.8. Analysis of Differential Metabolites

After data pre-treatment, principal-component analysis (PCA) was applied to inspect sample distribution, and Spearman correlation was used to check variable consistency. Partial least squares-discriminant analysis (PLS-DA) was conducted to analyze the differences between groups, and the model performance was evaluated through cross-validation. Student’s t-test (two-tailed, p < 0.05) supplied p-values for each metabolite. Putative compound classifications and pathway information were retrieved from the Kyoto Encyclopaedia of Genes and Genomes (KEGG). Metabolites were considered differential when log₂FC (fold-change) > 1, p-value (probability value) < 0.05 and VIP (variable importance in projection) > 1. Enrichment of KEGG pathways by these markers was evaluated with a hypergeometric test; pathways with adjusted p < 0.05 were regarded as significantly enriched.

2.9. EC₅₀ Values of Antifungal Compounds and Their Effects on Mycelial Morphology

To further identify the organic acid compounds with antimicrobial activity, we assessed their inhibitory effects using the dilution plate method. Compounds were added to PDA agar to achieve final concentrations of 500 and 1000 µg mL⁻¹. Plates containing 1% DMSO (Dimethyl Sulfoxide) without the compound served as positive controls (CK1), while PDA plates without the compound acted as blank controls (CK0). A 5 mm mycelial plug from a 4-day-old culture was inoculated onto each PDA plate. The plates were sealed and incubated in darkness at 25 °C, with results recorded when the fungal colony in the control group extended beyond three-quarters of the plate.

Several compounds with the best inhibitory effects were selected for dose–response assays, with concentrations set at 50, 100, 200, 400, 800, 1200, and 1600 µg mL⁻¹. Each treatment was replicated four times. The calculation method for the inhibition rate of hyphal growth by the compounds is as described in Section 2.2. The half-maximal effective concentration (EC₅₀) was calculated from the dose–response regression equation. The pathogen was cultured in medium containing the EC₅₀ concentration of the compound, and hyphal growth was observed under a microscope after 4 days.

2.10. Statistical Analysis

All statistical analyses were conducted with GraphPad Prism Statistics 10, and results are expressed as means ± standard deviation (SD). An independent-samples t-test was used to compare VOC-mediated inhibition rates with those observed in the pot assay. The data of broad-spectrum antifungal activity and inhibition rates at gradient concentrations of individual compounds were evaluated using one-way ANOVA (Analysis of Variance). The half-maximal effective concentration (EC₅₀) of the compounds was estimated by nonlinear regression (curve fitting). Based on the model fitting results, statistical differences between different treatment groups were determined by comparing the 95% confidence intervals of the EC₅₀ values. The significance level was set at p < 0.05.

3. Results

3.1. Identification of Pathogenic Strains

Pathogenic strain 3-3 was isolated from gray-mold lesions on P. notoginseng leaves collected in Wenshan, Yunnan Province, and its pathogenicity was confirmed by Koch’s postulates; re-inoculated leaflets developed water-soaked, soft-rot symptoms identical to those observed in the field. The fungus grew rapidly on PDA, filling three-quarters of a 90 mm Petri dish within 4 d at 25 °C. Young colonies were white and sparse, with abundant aerial hyphae; maturing mycelium gradually became denser and, after 25 d, produced both superficial and immersed sclerotia. Macro-sclerotia were black, sub-globose to irregular, ≈5 mm × 5 mm, whereas micro-sclerotia were visible from the reverse of the plate (Figure 1c,d). Light microscopy revealed dense, grape-like conidiophores bearing ovoid conidia (Figure 1b). PCR with universal primers ITS1 and ITS4 yielded a 533 bp amplicon (GenBank accession PV608498.1) that shared 99% identity with B. fabiopsis. Further identification was conducted using the conserved RPB2 gene sequence, and a phylogenetic tree was constructed in MEGA 7.0. Strain 3-3 (1127 bp, GenBank accession PX113379.1) in the same clade as B. fabiopsis strain B11 and BroadbeanBC-2 (100% identity), leading to the preliminary identification of the pathogen as B. fabiopsis, a new pathogen of P. notoginseng (Figure 1e).

3.2. The Antifungal Effect of X30 Strain on B. fabiopsis and Other Main Pathogens

Strain X30 exhibited an inhibition rate of 83.42% against B. fabiopsis strain 3-3, significantly suppressing the growth of the gray mold pathogen. The broad-spectrum antimicrobial experiment indicated that X30 has varying degrees of inhibitory effects on multiple pathogens of P. notoginseng. Among them, the most significant antimicrobial effect was observed against A. alternata, with an inhibition rate of 85.45%, demonstrating that X30 possesses extensive antifungal activity (Figure 2a,b). Microscopic observations showed that hyphae of B. fabiopsis strain 3-3 treated with X30 exhibited bending, branching, and deformation, while control hyphae maintained normal morphology (Figure 2c). We therefore conclude that X30 exerts its antifungal effect, at least in part, by disrupting normal mycelial architecture.

3.3. Morphological and Molecular Biological Identification of X30 Strain

Strain X30 is an aerobic, Gram-positive rod, ca. 1–2 µm in length. On LB agar, it forms nearly circular, off-white, opaque colonies with undulate margins and a rough mat surface, rugose surface with conspicuous folds (Figure 3b,c). Nearly full-length 16S rDNA gene sequence (1423 bp; GenBank PV608487.1) shares ≥98% identity with B. amyloliquefaciens. Neighbor-joining phylogeny constructed with MEGA 7.0 places X30 in a well-supported clade with the type strains of B. amyloliquefaciens, clearly separated from other Bacillus spp. (Figure 3d). Analysis of the recA gene (571 bp; GenBank PX113383.1) corroborates this assignment, clustering X30 with B. amyloliquefaciens BCRC14193 at 100% nucleotide identity (Figure 3e). Both loci therefore identify X30 as B. amyloliquefaciens.

3.4. Biological Function of X30 Strain

In plant-growth-promotion assays, strain X30 produced detectable levels of ammonia and nitrogen-fixing, while releasing indole-3-acetic acid (IAA), ACC deaminase (ACCD) and gibberellin (GA). However, it exhibited no phosphate-solubilizing, potassium-mobilizing, or siderophore-producing activity. Enzyme profiling revealed that every cell-wall-degrading hydrolase assayed—amylase, cellulase, protease, chitinase and β-1,3-glucanase was secreted by X30 (

Table 1. Biological functions of strain X30. (Table + represents positive results, − represents negative results).

Biological Functions	Results	Activity/Content
Ammonia Production	+
Nitrogen Fixation	+
Siderophore Production	−	-
Potassium Solubilisation	−	-
Phosphate Solubilisation	−	-
Indole-3-Acetic Acid (IAA)	+	50.37 pmol L−1
ACC Deaminase (ACCD)	+	447.08 U L−1
Gibberellin (GA)	+	117.02 pmol L−1
Amylase	+	4.15 (D2/D1)
Protease	+	3.36 (D2/D1)
Cellulase	+	2.89 (D2/D1)
Chitinase	+	12.79 U mL−1
β-1,3-Glucanase	+	2.48 mg mL−1

), indicating that the bacterium can antagonize phytopathogens through a battery of lytic enzymes.

3.5. Antifungal Activity Test of Metabolites of B. amyloliquefaciens X30

3.5.1. Antifungal Activity of VOCs from X30 Strain

After 4 days of co-culture, VOCs produced by strain X30 significantly inhibited the mycelial growth of pathogen 3-3 (p ≤ 0.05). The inhibitory effect increased with the bacterial inoculum density, reaching 63.49% at an X30 concentration of 1 × 10⁸ CFU mL⁻¹ (Figure 4a,c). This finding indicates that strain X30 can influence the mycelial growth of 3-3 through the release of bioactive VOCs.

3.5.2. Antifungal Activity of Cell-Free Filtrate of X30 Strain

The antifungal activity of the cell-free sterile filtrate of strain X30 against B. fabiopsis was evaluated using the Oxford cup method. Results showed that a distinct inhibition zone was formed around the cup, with a maximum inhibition rate of 66.84% against pathogen strain 3-3 (Figure 4b,c). This indicates that strain X30 can inhibit the growth of the pathogen through the production of secondary metabolites.

3.6. The Protective Effect of X30 Strain on the Leaves of P. notoginseng Infected by B. fabiopsis in Pot Experiment

Compared with CK1 (pathogen-inoculated control), leaves of P. notoginseng treated with X30 bacterial suspension and subsequently inoculated with the pathogen exhibited a significantly reduced area of rot, with a disease prevention rate of 65.77 ± 23.36%. Plants treated solely with X30 suspension showed no significant difference from those treated with sterile water (CK0) (Figure 5). These results indicate that X30 can effectively prevent disease without causing pathogenic effects on the plants.

3.7. UHPLC-MS/MS Analysis of Secondary Metabolites of Strain X30

3.7.1. Multivariate Statistical Analysis of X30-Treated and Control Groups

A total of 1465 metabolites were identified in this study. Principal component analysis (PCA) was employed to assess the quality of the LC-MS data. The results showed that samples from the experimental group were tightly clustered without any outliers, confirming the accuracy of the analysis (Figure 6a). There was no overlap between the experimental group (X30) and the control group (Control), indicating that the principal components of the compounds in the experimental and control groups were distinct. Supervised partial least squares-discriminant analysis (PLS-DA) was used to identify differential metabolites between the Control and X30 groups. The x-axis accounted for 91.6% of the predictive variation (between-group differences), while the y-axis accounted for 1.11% of the orthogonal variation (within-group differences), demonstrating a significant shift in the metabolic profiles between the experimental group X30 and the control group Control (Figure 6b).

3.7.2. Abundance Characteristics of Differential Metabolites in Different Samples

A heatmap analysis of differentially expressed metabolites revealed that samples from the treatment and control groups formed two distinct clusters, indicating that X30 treatment induced a systemic reconfiguration of the metabolic network. The tight clustering of samples within each group demonstrated good biological reproducibility and experimental stability. The heatmap identified several key metabolites that were significantly upregulated in the treatment group, with surfactin and its analogues being particularly enriched, suggesting that metabolites from strain X30 may possess antimicrobial properties (Figure 6c).

3.7.3. KEGG Metabolic Pathway Analysis

Topology analysis confirmed the significant enrichment of core pathways: phenylalanine metabolism and tryptophan metabolism, which are among the pathways with the highest impact values. These aromatic amino acid pathways serve as precursors for the synthesis of diverse bioactive secondary metabolites. The enrichment of the aminobenzoate degradation pathway indicates that strain X30 synthesizes structurally specific aromatic degradation products. These compounds are often closely related to antibiotic synthesis or quorum-sensing signaling molecules and may constitute a key metabolic basis for their antimicrobial activity (Figure 6d). The KEGG enrichment analysis in Figure 6e further revealed the significant enrichment of tryptophan and phenylalanine metabolism, consistent with the key metabolic pathway analysis in Figure 6d.

3.8. Comparative Analysis of Antifungal Activity of Compounds

3.8.1. Preliminary Screening of Antifungal Activity of Compounds

Potential organic acid antimicrobial compounds were identified from metabolites based on the criteria of log₂FC > 1, p-value < 0.05, and VIP > 1. These compounds, including 2-hydroxyisovaleric acid, protocatechuic acid, phenylpyruvic acid and 2,6-dihydroxybenzoic acid, were tested for their antifungal activity against B. fabiopsis. The results showed that these compounds (at 1000 µg mL⁻¹) exhibited varying degrees of inhibitory effects on the growth of the pathogen. Compared with the control group, phenylpyruvic acid demonstrated the strongest inhibitory activity, followed by 2,6-dihydroxybenzoic acid (Figure S1).

3.8.2. Compounds Antifungal Activity Gradient Concentration Test

The two most effective compounds were selected for dose–response assays. Both compounds (50–1600 µg mL⁻¹) exhibited concentration-dependent inhibition of mycelial growth of strain 3-3, with inhibition rates significantly increasing as concentrations rose (Figure 7a). Phenylpyruvic acid showed the highest inhibitory activity, reaching 97.93% at 1200 µg mL⁻¹ (Figure 7b). Logarithmic transformation analysis revealed an EC₅₀ value of 312 µg mL⁻¹ for phenylpyruvic acid against B. fabiopsis 3-3 (Figure 7b). The EC₅₀ value of 2,6-dihydroxybenzoic acid was 660 µg mL⁻¹ (Figure 7b). Figure S2 shows the chemical formulas, metabolite mass spectra, and expression differences in these two compounds. The violin plot demonstrates that the relative abundance of the compounds in the treatment group (X30) is significantly higher than that in the control group, indicating that these two compounds are the key metabolites responsible for the antifungal activity exhibited by strain X30. All treatment groups showed statistically significant differences in inhibition rates compared with the control (p < 0.05). These results suggest that phenylpyruvic acid has potential as a biocontrol agent for the management of plant diseases.

3.8.3. Effects of Compounds on Mycelium

To investigate the effects of the compounds on mycelial ultrastructure, we examined hyphae from treated and control groups under a microscope. In the presence of phenylpyruvic acid, the membrane structure of 3-3 hyphae was severely disrupted, with visible leakage of cytoplasmic contents, and some hyphae appeared fragmented and dissolved. In contrast, treatment with 2,6-dihydroxybenzoic acid caused hyphae to twist, fold, and become deformed. Hyphae in the control group remained intact and plump (Figure 7d). These observations indicate that both compounds can alter the ultrastructure of the pathogen, leading to subsequent growth inhibition.

4. Discussion

Fungi in the genus Botrytis have an extremely broad host range among plants. The increasing resistance of Botrytis spp. to chemical fungicides has made their control increasingly challenging [27,28]. Based on multigene sequence analyses (e.g., RPB2, HSP60, G3PDH), the genus Botrytis currently comprises over 30 species [29]. We isolated a pathogenic strain 3-3 from P. notoginseng leaves affected by gray mold. Morphological and molecular analyses identified this strain as B. fabiopsis. First described on Vicia faba L. in central China, B. fabiopsis has subsequently been isolated from faba beans and chickpeas in Latvia [30]. Phylogenetic analysis shows it is more closely related to Botrytis spp. on monocots than to the more widely host-specific B. cinerea on dicots [29,31]. Our study identifies B. fabiopsis as a new pathogenic fungus of P. notoginseng. This discovery expands the known host range of B. fabiopsis and highlights its potential threat to Panax species.

Strain X30, isolated from P. notoginseng leaves, exhibited antagonistic activity against B. fabiopsis and six other pathogens of P. notoginseng. Molecular identification confirmed it as B. amyloliquefaciens, a species widely recognized for its biocontrol potential. It is the first report of a biocontrol agent targeting gray mold in P. notoginseng. The protective efficacy of X30 on live plants suggests its potential as an alternative to chemical fungicides for controlling this emerging pathogen.

We found that strain X30 exhibited multiple plant-growth-promoting traits, including ammonia production, nitrogen fixation, and phytohormone secretion. Bacillus amyloliquefaciens is a crucial plant-growth-promoting bacterium [32,33]. For example, B. amyloliquefaciens YB1701 produces IAA and exhibits ACC deaminase activity, significantly promoting rice seedling growth [34]. Additionally, the secretion of hydrolytic enzymes by X30 may contribute to the disruption of fungal cell membranes and cell walls. Such enzymatic activities likely explain the morphological deformations observed in treated B. fabiopsis hyphae, similar to previous reports on other Botrytis species [35,36].

The production of secondary metabolites is a key mechanism by which antagonistic bacteria inhibit pathogens [37]. Both non-volatile and volatile metabolites of strain X30 suppressed the mycelial growth of the pathogen, with non-volatile metabolites showing a stronger effect. UHPLC-MS/MS analysis revealed that the secondary metabolites of X30 primarily consist of lipids, organic acids, aromatic compounds, and organic heterocyclic compounds (Figure 6c). Current research on Bacillus spp. has primarily focused on lipid and protein metabolites [38,39]. In contrast, we focused on organic acid compounds. Our metabolomics-guided approach identified phenylpyruvic acid and 2,6-dihydroxybenzoic acid as key antifungal constituents. These compounds exhibited potent inhibitory activity against B. fabiopsis. Phenylpyruvic acid functions as an inhibitor of amino acid metabolism. It acts as a competitive substrate or inhibitor of transaminases, thereby affecting microbial protein synthesis [40,41]. Research on the antimicrobial activity of 2,6-dihydroxybenzoic acid has not been reported. Microscopic observations suggest that these two compounds might affect fungal hyphae through different modes of action, one primarily causing membrane damage and the other inducing morphological deformities, thereby offering a potential dual-mode antifungal strategy. Notably, this study represents the first report of phenylpyruvic acid and 2,6-dihydroxybenzoic acid as antifungal metabolites from Bacillus spp., expanding the known repertoire of biocontrol-active compounds produced by this species and providing new targets for metabolic engineering in microbial pesticide development.

Future research should focus on targeted metabolomics to determine the absolute concentrations of these bioactive compounds. This will help to elucidate their specific contributions to the resistance mechanisms of strain X30. The transition from the laboratory to field application is a complex process that requires further field experiments to verify the practical effectiveness and feasibility of the X30 strain. Furthermore, refine the fermentation conditions for X30 to enhance its effectiveness in managing plant diseases and to determine if its protective effects can be generalized to other crops.

5. Conclusions

A strain of B. amyloliquefaciens X30 was isolated from the leaves of P. notoginseng, and it has shown significant antifungal activity against B. fabiopsis strain 3-3 and six other pathogenic fungi that are common to P. notoginseng. Its antifungal action relies on the combined action of multiple cell-wall-degrading enzymes and secondary metabolites (Figure 8). Two bioactive molecules were purified from the metabolome: phenylpyruvic acid (EC₅₀ 312 µg mL⁻¹), which disrupted membrane integrity and caused cytoplasmic leakage, and 2,6-dihydroxybenzoic acid (EC₅₀ 660 µg mL⁻¹), which induced severe hyphal deformation. Beyond pathogen suppression, X30 exhibits nitrogen fixation, produces ammonia, and secretes indole-3-acetic acid and gibberellins, indicating substantial plant-growth-promoting potential. In pot assays, a single foliar application at a concentration of 1 × 10⁸ CFU mL⁻¹ achieved a disease control rate of 61.04% for gray mold without phytotoxicity, suggesting the promise of strain X30 as a safe and effective biocontrol agent for sustainable management of gray mold in P. notoginseng.