The Extracellular Lipopeptides and Volatile Organic Compounds of Bacillus subtilis DHA41 Display Broad-Spectrum Antifungal Activity against Soil-Borne Phytopathogenic Fungi

Fusarium oxysporum f. sp. niveum (Fon) is a devastating soil-borne fungus causing Fusarium wilt in watermelon. The present study investigated the biochemical mechanism underlying the antifungal activity exhibited by the antagonistic bacterial strain DHA41, particularly against Fon. Molecular characterization based on the 16S rRNA gene confirmed that DHA41 is a strain of Bacillus subtilis, capable of synthesizing antifungal lipopeptides, such as iturins and fengycins, which was further confirmed by detecting corresponding lipopeptide biosynthesis genes, namely ItuB, ItuD, and FenD. The cell-free culture filtrate and extracellular lipopeptide extract of B. subtilis DHA41 demonstrated significant inhibitory effects on the mycelial growth of Fon, Didymella bryoniae, Sclerotinia sclerotiorum, Fusarium graminearum, and Rhizoctonia solani. The lipopeptide extract showed emulsification activity and inhibited Fon mycelial growth by 86.4% at 100 µg/mL. Transmission electron microscope observations confirmed that the lipopeptide extract disrupted Fon cellular integrity. Furthermore, B. subtilis DHA41 emitted volatile organic compounds (VOCs) that exhibited antifungal activity against Fon, D. bryoniae, S. sclerotiorum, and F. graminearum. These findings provide evidence that B. subtilis DHA41 possesses broad-spectrum antifungal activity against different fungi pathogens, including Fon, through the production of extracellular lipopeptides and VOCs.


Introduction
Soil-borne fungi belonging to the Fusarium, Rhizoctonia, Sclerotinia, and Verticillium genera can cause root rot, vascular wilt, damping-off, and sclerotinia stem rot diseases in economically important crops, including vegetables, rice, wheat, cotton, and fruits [1]. Among them, the Fusarium oxysporum species complex can infect more than 150 crop species, including watermelon, tomato, melon, and cotton, and cause severe vascular Fusarium wilt diseases, leading to serious yield loss worldwide [2]. For example, F. oxysporum f. sp. niveum (Fon) causes the most destructive vascular wilt in watermelon, resulting in 30-50% yield losses [3]. Chemical pesticides have become ineffective due to their narrow spectrum activity and non-target environmental impacts, rendering them less effective in disease control. Furthermore, soil-borne fungi can thrive and persist by forming resting

Growth Conditions for Bacterial Strain DHA41 and Fungal Pathogens
Bacterial strain DHA41 was isolated from watermelon rhizosphere soil using the serial dilution method and cultured on a Luria-Bertani (LB) plate at 28 ± 2 • C, as described previously [16]. Fungal pathogens, including F. oxysporum f. sp. Niveum (Fon), F. graminearum (Fg), D. bryoniae (Db), R. solani (Rs), and S. sclerotiorum (Ss), were collected from the Crop Diseases and Insect Pests Laboratory of MARA at Zhejiang University and maintained on potato dextrose agar (PDA) at 28 ± 2 • C [16]. For sporulation, Fon was cultured in a mung bean liquid medium at 28 ± 2 • C with shaking (150 rpm) for 2 d, and spores were collected and adjusted to a final concentration of 4 × 10 6 spores/mL [16].

Amplification of 16S rRNA and Lipopeptide Biosynthesis Genes
Bacterial strain DHA41 was cultured in liquid LB medium under shaking (180 rpm) at 28 ± 2 • C for 12 h. The bacterial cells were collected by centrifugation, and genomic DNA was extracted using the Takara MiniBEST Bacteria Genomic DNA Extraction Kit Ver. 3.0 (Takara, Dalian, China), following the given instructions. Fragments of the 16S rRNA and the lipopeptide biosynthesis genes, namely ItuB, ItuD, and FenD, were amplified using gene-specific primers ( Table 1). The 16S rRNA gene amplicon was sequenced commercially (Zhejiang YouKang Biotech, Hangzhou, China), aligned using the ClustalX program [46], and subjected to phylogenetic analysis using MEGA 11.0 software, following the neighbor joining (NJ) method [47].

Characterization of Cellular Fatty Acids
Bacterial strain DHA41 was cultivated on Tryptic soy agar (Difco Laboratories, Sparks, MD, USA) at 28 ± 2 • C for 24 h. Fatty acid methyl esters (FAMEs) were prepared and analyzed following the protocol of the Sherlock microbial identification system [50]. Briefly, fatty acids were released from the bacterial cells through saponification with NaOH and esterified with 6 N HCl to generate FAMEs. The FAME-containing upper layer was then collected using a methyl tert-butyl ether and hexane solution (1:1, v/v). The FAME profiling was conducted using a Hewlett Packard 5890 Series gas chromatography machine (Ramsey, MN, USA). The fatty acids were identified and quantified by comparing the retention time and peak area with an authentic standard fatty acid mixture (Sigma-Aldrich, St. Louis, MO, USA) as well as with the RTSBA6 6.10 library of bacterial fatty acids [50].

Extraction, Purification, and Characterization of Extracellular Lipopeptides from DHA41
Bacterial strain DHA41 was grown in 100 mL liquid LB medium at 28 ± 2 • C with shaking (200 rpm) for 72 h. The culture was centrifuged (12,000 rpm) at 4 • C for 20 min, and the resultant supernatant was collected. The supernatant was acidified by adding 2 M HCl (pH 2.0) and incubated overnight at 4 • C. Lipopeptide precipitates were collected through centrifugation (15,000 rpm) and resuspended in a methanol and water solution (2:1, v/v). The lipopeptide extract was dried in a rotary vacuum at 40 • C, re-dissolved in dimethyl sulfoxide (DMSO), and stored at −20 • C for further investigation.
Matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry (MALD-TOF-MS) analysis was used to identify the lipopeptides of strain DHA41 [51]. A single colony of bacterial strain DHA41 was picked and homogenized in a matrix solution, as described previously [16]. After centrifugation (12,000 rpm) for 20 min at 4 • C, 1 µL of the resulting supernatant was spotted onto a MALDI-TOF-MS target plate (Bruker Daltonik, Bremen, Germany) and air-dried. The samples were analyzed using an Ultraflex MALDI-TOF-MS spectrometer equipped with a smartbeam laser (Daltonics, Bremen, Germany) for desorption and ionization with a nitrogen laser at 337 nm. The obtained spectra were analyzed to identify different lipopeptides in the extract, with the molecular weight ranging from 800 to 3000 Daltons (Da).

Evaluation of Emulsification Index
The emulsification index was evaluated following a previous protocol [52]. Briefly, bacterial strain DHA41 was grown in 100 mL liquid LB medium at 28 ± 2 • C with shaking (200 rpm) for 72 h. A cell-free supernatant (2 mL) was obtained after centrifugation (12,000 rpm) and mixed with 3 mL of hydrophobic compounds (sunflower oil, mineral oil, and toluene) or a sodium dodecyl sulfate (0.1 g/L; Sigma-Aldrich, St. Louis, MO, USA) and Triton X-100 solutions (0.1%, v/v; Sigma-Aldrich, St. Louis, MO, USA) as controls. The mixture was thoroughly vortexed for 2 min and left at room temperature for 24 h. The height of the stable emulsion layer was measured, and the emulsification index (%) was calculated by comparing the height of the emulsified layer to the total height of the liquid.

Antifungal Activity of Cell-Free Supernatant, Lipopeptide Extract, and VOCs of DHA41
The antifungal activity of the cell-free supernatant of strain DHA41 and its lipopeptide extract was examined using the well culture method [53,54], with minor modifications. The filter-sterilized cell-free supernatant (30 µL), obtained from a 2-day-old culture of strain DHA41, and 30 µg/mL of lipopeptide extract (60 µL) were poured into the wells (5 mm) of the PDA plates, which were then inoculated with fungal mycelial plugs (5 mm in diameter). The wells supplemented with a similar volume of sterile LB medium or DMSO were used as controls, followed by incubations at 28 ± 2 • C for 5 d. To estimate growth inhibition rate, the colony diameters grown in the treated plates were compared to those grown in the control plates.
The effect of VOCs produced by DHA41 on the tested fungi was also examined using the two-sealed-base-plates method as previously described [55]. Briefly, bacterial strain DHA41 was streaked onto the LB medium on one base plate, while fungal mycelial discs (5 mm in diameter) were placed on PDA on another base plate. The two plates were then tightly assembled face-to-face and sealed with three layers of Parafilm and incubated at 28 ± 2 • C for 3-5 d. Two-plate assemblies without inoculation of strain DHA41 served as controls. The growth inhibition rate was calculated by comparing the colony diameters grown in the presence of strain DHA41 to those grown on the control plates.

Minimal Inhibitory Concentration Assay
The minimal inhibitory concentration (MIC) of the lipopeptide extract against Fon was determined as previously described [16]. Briefly, Fon inoculum (50 µL) was added to the wells of a 96-well microtiter plate, followed by supplementation with varying levels of lipopeptide extract (50,75,100,200, and 300 µg/mL). Control wells were supplied with DMSO. After incubation at 28 ± 2 • C for 24 h, the optical density at 600 nm (OD 600 ) was measured. To calculate the growth inhibition rate, the OD 600 values in the lipopeptidesupplemented wells were compared to that in the control wells.

Microscopic Observation
The effect of lipopeptide extract on Fon cell viability was examined by staining using fluorescein diacetate (FDA) and propidium iodide (PI), as previously described [56,57]. In these experiments, live fungal cells exhibited green fluorescence, while dead cells showed red fluorescence. After treating the Fon inoculum with lipopeptide extract (30 µg/mL) for 12 h, fungal mycelia were collected and resuspended in DMSO, followed by staining with fluorescent dyes (FDA and PI) at room temperature for 15 min under dark conditions. The stained mycelia were observed with a Zeiss LSM 880 laser confocal microscope (Jena, Germany).
The effect of the lipopeptide extract on Fon mycelial and conidial cellular structure was examined using transmission electron microscopy (TEM, model H-7650, Hitachi, Tokyo, Japan), as previously described [57,58]. The Fon inoculum (1 mL) was treated with lipopeptide extract (30 µg/mL) and incubated with shaking (150 rpm) at 28 ± 2 • C for 12 h. The samples were then immersed in 2.5% glutaraldehyde overnight, rinsed three times for 15 min with 0.1 M phosphate buffer (pH 7.0), and post-fixed in OsO 4 in the phosphate buffer (1%, w/v) for 2 h. The samples were dehydrated in an ethanol grade (30-70%) for 15 min at each concentration and finally dehydrated twice in absolute acetone for 20 min. The dehydrated samples were embedded in Spurr resin and polymerized for 12 h at 70 • C. Ultrathin sections of the samples were stained with uranyl acetate and lead citrate and observed using the H-7650 TEM.

Statistical Analysis
The experiments in this study were independently conducted three times, and at least three replicates were included for each treatment in an independent experiment. Data from three independent experiments were subjected to statistical analysis using the Tukey's least significant difference (LSD) method to determine the statistical significance among the treatments at a 95% confidence level in SPSS 14.0 software.

Identification of Extracellular Lipopeptides Produced by B. subtilis DHA41
The extracellular lipopeptides produced by B. subtilis DHA41 were characterized by MALDI-TOF-MS analysis, showing multiple spectral peaks that typically corresponded to three families of non-ribosomal lipopeptides, including iturins (1079.  Figure 2). Furthermore, specific bands for genes involved in iturin-(ItuB and ItuD) and fengycin (FenB) biosynthesis, two well-known antifungal lipopeptide families, were PCR amplified using the selected primer pairs from  (Figure 2). Furthermore, specific bands for genes involved in iturin-(ItuB and ItuD) and fengycin (FenB) biosynthesis, two wellknown antifungal lipopeptide families, were PCR amplified using the selected primer pairs from the genomic DNA of B. subtilis DHA41 ( Figure 3A). These data indicate that B. subtilis DHA41 possesses the inherent genetic potential for producing different extracellular lipopeptides.
The emulsification index was analyzed to further confirm the presence of biosurfactants in the lipopeptide extract from B. subtilis DHA41. The lipopeptide extract showed emulsifying activity against two hydrophobic substrates, including mineral oil (31.25 ± 0.03%) and sunflower oil (51 ± 0.50%) ( Figure 3B and Table 3). Interestingly, the lipopeptide extract did not show emulsifying activity against toluene ( Figure 3B and Table 3). These results indicate that B. subtilis DHA41 is capable of producing extracellular biosurfactants in response to specific organic compounds. The emulsification index was analyzed to further confirm the presence of biosurfactants in the lipopeptide extract from B. subtilis DHA41. The lipopeptide extract showed emulsifying activity against two hydrophobic substrates, including mineral oil (31.25 ± 0.03%) and sunflower oil (51 ± 0.50%) ( Figure 3B and Table 3). Interestingly, the lipopeptide extract did not show emulsifying activity against toluene ( Figure 3B and Table 3). These results indicate that B. subtilis DHA41 is capable of producing extracellular biosurfactants in response to specific organic compounds.   lipopeptides.
The emulsification index was analyzed to further confirm the presence of biosurfactants in the lipopeptide extract from B. subtilis DHA41. The lipopeptide extract showed emulsifying activity against two hydrophobic substrates, including mineral oil (31.25 ± 0.03%) and sunflower oil (51 ± 0.50%) ( Figure 3B and Table 3). Interestingly, the lipopeptide extract did not show emulsifying activity against toluene ( Figure 3B and Table 3). These results indicate that B. subtilis DHA41 is capable of producing extracellular biosurfactants in response to specific organic compounds.

Antifungal Effect of Cell-Free Filtrate and Extracellular Lipopeptides of B. subtilis DHA41
The results of the antifungal activity of the cell-free supernatant of B. subtilis DHA41 against different pathogenic fungi showed significant inhibition of Fon, Ss, and Db mycelial growth ( Figure 4A), with inhibition rates of 89.81%, 85.92%, and 86.06%, respectively, compared to the controls ( Figure 4B). These results indicate the presence of potent antimicrobial compounds in the cell-free supernatant of B. subtilis DHA41 that effectively inhibit the mycelial growth of pathogenic fungi.
Next, inhibition zones were observed around the lipopeptide-supplemented wells for Fon, Ss, Db, Rs, and Fg colonies compared to the DMSO-supplemented control wells ( Figure 5A). The largest inhibition zone was observed for Rs (11.50 ± 0.70 mm), followed by Fon (10.33 ± 0.47 mm), Ss (10.33 ± 0.47 mm), Db (8.66 ± 0.47 mm), and Fg (6.83 ± 0.62 mm) ( Figure 5B). Furthermore, the MIC of the lipopeptide extract against Fon was determined. The extracellular lipopeptide extract at 100 µg/mL exhibited the highest inhibition rate (86.4%) against Fon ( Figure 5C). However, Fon growth inhibition rate was decreased at higher concentrations of the lipopeptide extract (200 and 300 µg/mL) compared to 100 µg/mL ( Figure 5C). The observed non-linear antifungal activity of the lipopeptide extract at higher concentrations might be attributed to the saturation effect and hormesis or a dose-dependent response. These results collectively indicate that the extracellular lipopeptide extract of B. subtilis DHA41 possesses significant antifungal activity against multiple phytopathogenic fungi, including Fon. compared to the controls ( Figure 4B). These results indicate the presence of potent antimicrobial compounds in the cell-free supernatant of B. subtilis DHA41 that effectively inhibit the mycelial growth of pathogenic fungi.
Next, inhibition zones were observed around the lipopeptide-supplemented wells for Fon, Ss, Db, Rs, and Fg colonies compared to the DMSO-supplemented control wells ( Figure 5A). The largest inhibition zone was observed for Rs (11.50 ± 0.70 mm), followed by Fon (10.33 ± 0.47 mm), Ss (10.33 ± 0.47 mm), Db (8.66 ± 0.47 mm), and Fg (6.83 ± 0.62 mm) ( Figure 5B). Furthermore, the MIC of the lipopeptide extract against Fon was determined. The extracellular lipopeptide extract at 100 µg/mL exhibited the highest inhibition rate (86.4%) against Fon ( Figure 5C). However, Fon growth inhibition rate was decreased at higher concentrations of the lipopeptide extract (200 and 300 µg/mL) compared to 100 µg/mL ( Figure 5C). The observed non-linear antifungal activity of the lipopeptide extract at higher concentrations might be attributed to the saturation effect and hormesis or a dose-dependent response. These results collectively indicate that the extracellular lipopeptide extract of B. subtilis DHA41 possesses significant antifungal activity against multiple phytopathogenic fungi, including Fon.

Extracellular Lipopeptides Disrupt Fon Cellular Integrity
The cell viability assays using FDA and PI staining [61,62] revealed that the untreated Fon mycelia and conidia exhibited normal morphology and intact structures, as evidenced from strong FDA-generated green fluorescent signals ( Figure 6A,B). However, extracellular lipopeptide extract (30 µg/mL)-treated Fon mycelia and conidia showed PI-generated red fluorescence, revealing damaged morphology and collapsed structures ( Figure 6).  Further, the ultrastructure studies demonstrated that the untreated mycelia and conidia showed intact cellular morphology and structures, such as cellular membranes and cytoplasm, while mycelia and conidia treated with the lipopeptide extract displayed abnormal morphology and structure, as revealed by shrinking of the cytoplasm, plasma membrane damage, and cell wall disintegration (Figure 7). These data indicate that the extracellular lipopeptide extract of B. amyloliquefaciens DHA41 can disrupt Fon integrity, leading to cellular damage and decreased viability.
It has previously been discovered that the non-ribosomally synthesized extracellular lipopeptides are associated with the antimicrobial activity and biocontrol of plant diseases by Bacillus spp. [29,31]. Generally, Bacillus spp. produce three families of lipopeptides, including surfactins, fengycins, and iturins [29]. In the present study, MALDI-TOF-MS analysis identified three isoforms of iturins, three isoforms of surfactins, and two isoforms of fengycins (Figure 2), indicating the genetic diversity of B. subtilis DHA41 to produce extracellular lipopeptides. This result is further confirmed by detecting ItuB, ItuD, and FenB genes in B. subtilis DHA41 ( Figure 3A). Comparative genomic studies have shown that B. subtilis strains harbor 11 putative large biosynthetic gene clusters, some of which are responsible for lipopeptide production [75]. Recent studies have also highlighted variations in the production of non-ribosomally synthesized lipopeptides among different B. subtilis strains isolated from the same soil sample [76]. For example, analysis of 330 biosynthetic clusters from B. subtilis and their lipopeptides revealed a species-specific pattern of lipopeptide production [77]. Furthermore, B. subtilis DHA41 exhibited significant emulsification activity against some organic oils, such as mineral oil and sunflower oil ( Figure 3B and Table 2), which coincides with previous studies where lipopeptide biosurfactants from Bacillus thuringiensis pak2310 showed emulsification and antifungal activity against F. oxysporum [78]. The emulsification activity is known to contribute to adhesion, bioavailability, desorption, and antimicrobial activity in natural environments [79]. Collectively, the diverse extracellular lipopeptides and emulsification activity play an important role in the antifungal activity and disease-suppressing ability of B. subtilis DHA41.
The cell-free supernatant and extracellular lipopeptide extract of B. subtilis DHA41 exhibited significant broad-spectrum antifungal activity against Fon, Db, Ss, Rs, and Fg ( Figure 4A,B). This finding is consistent with previous observations showing that lipopeptide extracts from B. subtilis SCB-1 and B. amyloliquefaciens CNU114001 showed antifungal activity against diverse fungal pathogens, including F. oxysporum and Ss [80,81]. The lipopeptide extract from B. subtilis DHA41 had a notable impact on the viability of mycelia and conidia of Fon ( Figure 6), leading to disruptions in cellular structure and integrity ( Figure 7). These observations align with previous results demonstrating that lipopeptides from B. velezensis and B. amyloliquefaciens caused morphological changes in Fon, such as cytoplasmic shrinkage, aggregation of organelles, and damage to plasma membranes and cell walls [36,38,53,71]. Furthermore, the Bacillus spp.-produced lipopeptides enter Fon cells through endocytosis [82], subsequently targeting intracellular molecules and triggering metabolic alterations, thus exerting the antifungal effects against Fon [72,83].
In addition to extracellular lipopeptides, antagonistic bacteria also emit a wide range of VOCs [32,35]. These VOCs are low molecular weight compounds that readily emit under normal environmental conditions and exhibit significant antimicrobial activity [33]. In this study, co-incubation of Fon, Db, Ss, and Fg with B. subtilis DHA41 significantly inhibited the mycelial growth of fungal pathogens, implying the production of VOCs by B. subtilis DHA41 (Figure 8). Notably, B. subtilis DHA41-emitted VOCs exhibited varying levels of inhibition against the tested fungal pathogens. B. subtilis, B. amyloliquefaciens, and B. mycoides have been reported to release antifungal VOCs, inhibiting the mycelial growth and spore germination rate of Fon, F. oxysporum f. sp. lycopersici, F. oxysporum f. sp. cubense, F. oxysporum f. sp. radicis-lycopersici, and Ss [40][41][42][43]84]. For example, B. amyloliquefaciens L3-produced VOCs, including 2-heptanone, 2-ethyl-1-hexano, and 2-nonanone, completely inhibited Fon mycelial growth [40]. VOCs of antagonistic bacteria have been shown to improve plant growth and trigger induced systemic resistance in plants [33]; for example, albuterol and 1,3-propanediole, two volatile organic compounds produced by B. subtilis SYST2, and acetoin and 2,3-butanediol, produced by B. amyloliquefaciens L3, promoted the growth of tomato and Arabidopsis plants [73,85]. However, the chemical nature of the B. subtilis DHA41-produced VOCs and their involvement in promoting plant growth and suppressing Fusarium wilt in watermelon [16] need further investigation.

Conclusions
In this study, we found that the antagonistic bacterium B. subtilis DHA41 produces three families of extracellular lipopeptides, including iturins, surfactins, and fengycins, which exhibited significant antifungal activity against five soil-borne phytopathogenic fungi, including Fon, Db, Ss, Rs, and Fg. The extracellular lipopeptide extract of B. subtilis DHA41 effectively inhibited the mycelial growth and spore germination of Fon by disrupting cellular structure and integrity. Furthermore, B. subtilis DHA41 emitted VOCs that displayed inhibitory effects on the mycelial growth of Fon, Db, Ss, and Fg. Our findings highlight the biocontrol capacity of B. subtilis DHA41 in combating Fusarium wilt in watermelon through the production of diverse extracellular lipopeptides and VOCs. The ability of B. subtilis DHA41 to promote plant growth and suppress Fusarium wilt in watermelon, along with the antifungal activity of its active secondary metabolites, suggests