Mechanisms of Surfactin from Bacillus subtilis SF1 against Fusarium foetens: A Novel Pathogen Inducing Potato Wilt

Fusarium wilt is a severe and worldwide disease in potato cultivation. In this study, Fusarium foetens was first identified as the pathogen of potato wilt. Bacillus subtilis SF1 has the potential for controlling potato wilt induced by F. foetens, resulting in a mycelium growth inhibition of 52.50 ± 2.59% in vitro and a significant decrease in incidence rate by 45.56% in vivo. This research highlighted the antifungal activity of surfactin from B. subtilis SF1 and attempted to reveal the unknown antifungal mechanisms. Surfactin inhibited F. foetens mycelium growth beyond the concentration of 20 μg/μL. Surfactin-treated mycelium appeared to have morphological malformation. Surfactin enhanced reduced glutathione production and caused the increase in values of the extracellular fluids in OD260 and OD280. Surfactin induced differential protein expression and changed the genes’ transcription levels. Surfactin binds to fungal DNA via groove-binding mode, with a binding constant of Kb 2.97 × 104 M−1. Moreover, B. subtilis SF1 harbored genes encoding plant-promoting determinants, making potato seedlings grow vigorously. The results will help provide a comprehensive understanding of the mechanisms of surfactin against filamentous fungi and the application of surfactin-producing microbial in the biocontrol of plant pathogenic fungi.


Introduction
Potato (Solanum tuberosum L.) is the fourth largest crop, after rice, maize, and wheat. It ranks first as a non-cereal food crop for human consumption and has great potential for ensuring food security in developing countries [1]. About 1.3 billion people in China and India consume fresh potatoes as a staple food [2]. Diseases caused by fungi have always been a severe problem in potato cultivation [3]. Fusarium spp. are the most common soilborne pathogens responsible for yield and commercial losses during potato production [4]. These pathogens affect potatoes at any growth stage by inducing Fusarium wilt (FW) on plants and Fusarium dry rot on tubers. FW is prevalent worldwide, leading to 30-78% yield losses in some parts of China [5]. FW is a vascular disease mainly caused by F. oxysporum. Fusarium spp., such as F. solani, F. graminearum, and F. sambucinum, have also been reported as FW pathogens [6][7][8].
Management strategies for potato Fusarium diseases have mainly focused on chemical fungicides, crop rotation, and breeding of resistant cultivars [4,9,10]. However, due to the rapid differentiation of pathogenicity, the appearance of resistant pathogenic strains, and the lack of superior cultivars with high levels of resistance, Fusarium diseases are difficult to control. A recent approach that has gained attention is the use of bacterial antagonists, such as members of the genus Bacillus, particularly Bacillus subtilis, which

Pathogen Isolation and Pathogenicity Tests
The symptomatic potato plants were collected from Laiyang, Shandong Province of China, in April 2021. The pathogen was isolated according to our previous publication [26]. After cleaning and sterilization, the stem base of the symptomatic potatoes was cut into pieces 1 cm long and 0.5 cm wide, placed on potato dextrose agar (PDA) Petri dishes, and cultured at 25 • C for 3-5 days. The obtained colonies were transferred onto a new plate repeatedly.
Three F. foetens isolates were selected randomly and conducted infection on 90 potato seedlings grown under greenhouse conditions for pathogenicity tests. Each isolate infected ten potato seedlings each time. The experiments were repeated three times. The seed tubers were planted in pots of peat and perlite (3:1 v/v). Two weeks later, The seedlings were transplanted into new pots with F. foetens spores (1 × 10 7 CFU/mL) at a volume of 100 mL.
The seedlings in pots with sterilized water were the negative control. All the seedlings were maintained in the greenhouse with 12 h light and 12 h dark per day at 25 • C for 30 days. The disease symptoms were investigated and the symptomatic to re-isolate the pathogen was collected.

Pathogen Identification
Multiple gene loci, internal transcribed spacer rDNA (ITS), EF-1α, and β-tubulin, were applied to the molecular identification of the pathogen causing potato wilt [21]. The ITS, EF-1α, and β-tubulin genes were amplified from the genomic DNA with PCR primers ITS1/ITS4, EF-1/EF-2, and T1/T22 [27,28], respectively. The amplified products were purified and sequenced. Then, the gene sequences were analyzed using BLAST in the NCBI database. The gene sequences of EF-1α and β-tubulin were combined to build a phylogenetic tree with Neighbor-Joining (NJ) in MEGA 10.0. The branch supports of NJ analysis were evaluated using a bootstrapping (BS) method of 1000 replicates.
The fungus grown on PDA was incubated at 25 • C in the dark for 10-14 days. The colony morphology of the fungus was observed using stereo microscope (M205C, Leica, Wetzlar, Germany). The morphology of conidia, phialide, and chlamydospore was observed with positive fluorescence microscope (CX21, Olympus, Tokyo, Japan) after staining the fungal structures with cotton blue.
2.3. The Antifungal Activity of B. subtilis SF1 against F. foetens In Vitro B. subtilis SF1 (ACCC 19742) was purchased from the Agricultural Culture Collection of China. Illumina was used for the whole-genome sequencing of B. subtilis SF1 by oeBiotech (Shanghai, China). The whole genome shotgun project has been deposited at DDBJ/ENA/GenBank under the accession JANYMH000000000. The antifungal activity of B. subtilis SF1 against F. foetens was evaluated with a dual culture plate assay, as described in the previous study [29]. To investigate the antifungal activity of the metabolites produced by B. subtilis SF1, the culture filtrate of B. subtilis SF1 was prepared and mixed with water in different volumes to prepare the PDA medium. F. foetens plug (5 mm) was inoculated onto the plates containing different proportions of culture filtrate. The colony diameters of F. foetens were measured when the mycelium of control completely overspread the Petri dish surface. The growth inhibition for F. foetens was estimated using the formula: Growth inhibition% = ((D1 − D2)/D1) × 100%, where D1: Mean diameter of the colony for the control and D2: Mean diameter of the colony in presence of culture filtrate or cells of B. subtilis SF1. All the cultures were incubated at 25 • C.
The culture filtrate of B. subtilis SF1 was prepared following the procedures [30]: B. subtilis SF1 cells were inoculated into LB and cultured for 24 h. Then, the culture was transferred into the fermentation media consisting of glucose 40 g/L, beef extract 15 g/L, (NH 4 ) 2 SO 4 4 g/L, KH 2 PO 4 14 g/L, 14 g Na 2 HPO 4 g/L and MgSO 4 1.5 g/L. After being cultured for 40 h, the fermentation broth was centrifuged (4 • C, 6000× g), and the supernatant was collected and sterilized twice with a 0.22 µm millipore filter.

Effect of B. subtilis SF1 on Fusarium Wilt Incidence
The tests on the effect of B. subtilis SF1 on potato FW incidence were performed by the root-dip method previously described [31]. Wash roots of potato seedings, then dip into conidia suspension (1 × 10 7 CFU/mL) of F. foetens. Transplant the infected seedlings into the pots supplemented with 100 mL of B. subtilils SF1 cell suspension (1 × 10 7 cells/mL). Use sterilized water in place of B. subtilis SF1 culture in positive control. The seedlings without infection and B. subtilis SF1 inoculation were negative control. The disease incidence was assessed after 60 days using the following formulas: Incidence rate (%) = (No. of diseased plants/30) × 100%. To investigate the effect of B. subtilis SF1 on potato plant growth, the pots of healthy potato seedlings were inoculated with B. subtilis SF1 cells (100 mL, 1 × 10 7 cells/mL). After culturing for 20 days, compare the seedlings treated with SF1 with the seedlings in negative control by monitoring the number of leaves and plant height. Ten 4 of 19 potato seedlings were used in each experiment group each time. The experiments were repeated three times, and 120 potato seedlings were used. All the pots were put in the field under natural lighting. The average outdoor temperature was 15 and 22 • C at night and daytime, respectively.

Hydrolytic Enzymes Production
The protein-coding genes in the B. subtilis SF1 genome were annotated in the Carbohydrate-Active enZYmes database. The activities of cellulase, protease, β-glucanase, and chitinase were assessed by measuring the halo formation on agar plates supplemented with carboxymethyl cellulose, skim milk, barley flour, and colloidal chitin, respectively, in the basal medium LB [32,33]. The concentration of skim milk was 10%, and the concentration of other additives was 1%. The agar plates were cultured at 30 • C for 5 days. A clear halo zone around the colony indicated B. subtilis SF1 was positive for enzyme production.

Plant Growth Promoting-Determinants Produced by B. subtilis SF1
Based on the published gene sequences, the genes associated with plant promotion were blasted in the B. subtilis SF1 genome. The phosphate solubilization was confirmed by the appearance of a halo zone around the B. subtilis SF1 colony on agar after bromophenol blue staining [34]. Use the National Botanical Research Institute's Phosphate growth medium (NBRIP) to examine inorganic phosphate solubilization. Replace the phosphorus source in the NBRIP with calcium phytate to assess organophosphates solubilization. B. subtilis SF1 was inoculated onto chromium azure S (CAS) medium, incubating for 5 days at 30 • C, to determine siderophores production by the appearance of a halo zone around the colony on CAS [35]. Indole-3-acetic acid (IAA) production was evaluated by the method of Bric et al. [36].

Extraction, Purification, and Identification of Surfactin Produced by B. subtilis SF1
The surfactin produced by B. subtilis SF1 was extracted using the acid precipitation method, as described previously [37]. The extracted surfactin was analyzed with the highperformance liquid chromatography (HPLC, 1100, Agilent, Santa Clara, CA, USA) equipped with a YMC C18 column (4.6 × 2.50 mm, 5 µm). The mobile phase consisted of 90% acetonitrile and 10% water (0.05% trifluoroacetic acid) with a flow rate of 0.8 mL per minute at 30 • C. The standard surfactin was purchased from Abison Company (Beijing, China).
The purification of surfactin was performed by preparative chromatography using the HPLC system (SCL-10A, Shimadzu, Tokyo, Japan) with a YMC C18 column (10.0 × 250 mm, 5 µm). The molecular weight of the surfactin was identified with LC-MS system (Orbitrap Fusion Lumos, Thermo Fisher, Waltham, MA, USA) equipped with a Kromasil C18 column (4.6 × 150 mm, 5 µm) and monitored at 210 nm as well as in positive-ion mode. The solvent gradient used acetonitrile and water at a flow rate of 3 mL/min with sample elution starting with 70% acetonitrile, followed by a linear gradient to 100% acetonitrile over 30 min.

Effect of Surfactin on the Growth and Morphology of F. foetens
The effect of surfactin on F. foetens growth was tested by the filter paper disk method [38]. The filter paper was dipped into 10 µL of surfactin at the concentration of 5, 10, 20, and 50 µg/µL, respectively, dried, and placed on a plate coated with 100 µL of 1 × 10 6 CFU/mL of F. foetens spores. A filter paper sheet dipped into the sterilized water was the control. The plates were incubated at 25 • C for 40 days and observed for mycelium growth.
To investigate the effect of surfactin on F. foetens morphology, the F. foetens mycelium cultured for 5-7 days on PDA was picked onto a glass slide and added surfactin to a final concentration of 1 µg/µL to immerse the mycelium. The mycelium treated with sterile PBS was the control. The glass slides were placed on a moisturizing gauze in Petri dishes, incubating at 25 • C for 4 h. Then, the mycelium morphology was observed under positive fluorescence microscopy (Axio Scope A1, Leica, Wetzlar, Germany) and scanning electron microscopy (SEM, JSM-7500F, JEOL, Tokyo, Japan). The specimen preparation for SEM was according to a previous publication [17].
2.9. Effect of Surfactin on the Nucleic Acids and Proteins of F. foetens F. Foetens spores (1 × 10 6 CFU/mL) were added to PDB supplemented with 0.5 and 1 µg/µL of surfactin, respectively, and incubated at 25 • C for 1 to 4 h. Culture in PDB without surfactin was the control. At the end of the incubation, centrifuged at 5000× g for 5 min, then the supernatant was collected and measured at OD 260 and OD 280 levels to assess the leakage of nucleic acids and proteins from F. foetens.
To study the effect of surfactin on F. foetens protein expression, the PDB medium inoculated with 1 × 10 6 CFU/mL spores was added to 0.5 µg/µL of surfactin and then incubated at 25 • C for 120 h. Culture in PDB without surfactin treatment was the control. After incubation, the mycelium was collected by centrifugation, and the total intracellular protein was extracted and separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Effect of Surfactin on Transcription Level of F. foetens
The effect of surfactin on the mRNA levels was determined using quantitative PCR (qPCR). The F. foetens cells were cultured in PDB with/without surfactin (0.5 µg/µL) for 120 h as the above method in 2.9. After culturing, the total RNA of the collected mycelium was extracted with QIAGEN RNeasy Mini Kit (QIAGEN, Hilden, Germany). cDNA was synthesized from purified total RNA using the QuantiTect reverse transcription kit (QIAGEN, Hilden, Germany). qPCR was performed using PrimeScriptTM RT reagent Kit with gDNA Eraser (Takara, Osaka, Japan). β-actin was selected as the reference gene to adjust the veracity of the qPCR assay. Primers were designed by Tsingke Biotechnology Co., Ltd., Beijing, China. Three parallel measurements were performed for every sample in a LightCycler 96 real-time PCR instrument (CFX96 touch, Bio, Washington, DC, USA). Relative transcription levels were presented using the 2 −∆∆CT method.

Effect of Surfactin on GSH Production of F. foetens
To investigate the effect of surfactin on GSH production in F. foetens, the F. foetens spores (1 × 10 6 CFU/mL) were inoculated in PDB without and with surfactin (0.5, 1, and 2 µg/µL, respectively). After incubation at 25 • C for 120 h, the mycelium was collected and mixed with 10% trifluoroacetic acid at a volume ratio of 1:1. The mixture was centrifuged at 5000× g for 5 min, then add 0.1 M potassium phosphate buffer into the supernatant. Ellman's reagent was used to measure the absorbance value at 412 nm using a spectrophotometer after 5 min.

DNA Binding Assay
The interaction between surfactin and the genomic DNA of F. foetens was detected by gel retardation experiments, as described previously [39]. The genomic DNA was extracted, and its purity was tested by OD 260 /OD 280 > 1.8. F. foetens DNA (ff-DNA) was treated with surfactin at concentrations of 0.05, 0.1, 0.5, and 1 µg/µL, respectively, and incubated for 1 h at room temperature, then subjected to electrophoresis on a 1% agarose gel and photographed under UV observation.
The binding mode of surfactin with ff-DNA was confirmed using UV-visible and fluorescence spectroscopy. The UV-visible spectroscopy analysis referred to a variation of the previous method [40]. The concentration of ff-DNA (30 µg/mL) was kept constant in 0.01 M PBS (pH 7.2) while gradually increasing the concentrations of surfactin by dropping the surfactin solution (1 µg/µL). It was equilibrated for 10 min after each drop, then the absorption spectrum of ff-DNA was measured in the range of 200 to 320 nm using an UV-visible spectrophotometer (Cary60, Agilent, Santa Clara, CA, USA). The fluorescence spectroscopy analysis was performed, as previously described [41]. Surfactin is a non-fluorescent compound, and ethidium bromide (EB) was used as a fluorescent probe to examine the interaction between surfactin and ff-DNA. The concentration of ff-DNA (30 µg/mL) was kept constant in 0.01 M PBS (pH 7.2), equilibrated for 10 min after EB (1.59 × 10 −5 M) addition, and then equilibrated for 20 min after each successive drop of surfactin solution (1 µg/µL). The fluorescence emission spectra in the range of 520 to 620 nm were scanned at room temperature using a fluorescence spectrophotometer (F4600, Hitachi, Tokyo, Japan) with a fixed excitation wavelength of 485 nm, an incident slit of 5 nm and an exit slit of 5 nm.

Statistical Analysis
The diameters of the colonies were expressed as mean values ± SD. All the assays were repeated in at least three separate experiments. Data were subjected to analysis of variance using SPSS version 20.0 for windows. Mean values were compared using Duncan's multiple range test at the 5% (p < 0.05) level of significance.

Isolation of the Pathogen and Molecular Identification
Twenty potato plants with typical wilt symptoms were collected from Laiyang city, China ( Figure 1A). A total of 99 fungi were isolated from the stem base of these diseased plants. The ITS fragments of these fungi were amplified and sequenced. The ITS sequence alignment results in the NCBI database showed that these fungi belonged to the genera Fusarium, Alternaria, and Colletotrichum, accounting for 75.76% (75), 18.19% (18) and 6.06% (6), respectively. Subsequently, partial sequences of EF-1α and β-tubulin were used to further identify 75 Fusarium strains at the species level. Fifty-three Fusarium isolates shared identical ITS (OM370930, NCBI), EF-1α (OM370932, NCBI), and β-tubulin (OM370931, NCBI) sequences, which showed 100% identity to the corresponding gene sequences of F. foetens CBS 110,286 (ITS for NR_159865.1, EF-1α for MT011001.1, β-tubulin for MT011049.1 in NCBI). A phylogenetic tree was constructed using the neighbor-joining method with the combination of the EF-1α and β-tubulin gene sequences from 10 Fusarium strains (Supplementary Figure S1). The F. foetens isolate was grouped with F. foetens CBS 110286 and was closely related to F. oxysporum.

Morphological Characteristics of F. foetens
The colony morphology of F. foetens is shown in Figure 1E,F. The aerial mycelia in the front colony formed thick white tufts evenly covering the entire Petri dish, whereas the reverse colony appeared purplish-red in the center. The colony diameter was 33.8 ± 1.6 mm on PDA after 4 days, and 57.0 ± 1.9 mm after 7 days. The average radial growth rate was 8.2 ± 0.3 mm/day. Conspicuous aerial mycelia gave older colonies a dotted appearance ( Figure 1G), which appeared to be stromata-supporting sporodochia [21]. However, no sporodochia occurred until 28 days of culture. Aerial mycelia bore solitary monophialides, occasionally polyphialides ( Figure 1H-J). The phialides were either cylindrical or slightly tapered toward the tip or narrowly flask-shaped, with the widest point in the middle ( Figure 1J). The chlamydospores were globose to subglobose, rare, mostly terminal, and smooth ( Figure 1K,L). Microconidia that formed laterally from hyphae of the aerial mycelium were 0 to 3 septate and fusiform to slightly curved ( Figure 1M-S). It is worth noting that F. foetens produced a pungent distinct odor.

Pathogenicity Test
Three representative F. foetens isolates were selected randomly and inoculated into potato seedlings pots cultivated in the greenhouse. After 20 days of cultivation in the greenhouse, the seedlings in the negative control group did not exhibit any wilting symptoms ( Figure 1B). The tested fungi developed wilting symptoms and were re-isolated from the infected plants, indicating that they were the pathogen causing the potato wilt. After inoculating with F. foetens on days 4-6, the potato plant stems near the roots appeared purplish red, this color gradually spread to the top of the stem, and the lower leaves began to wilt ( Figure 1C). On day 9, the plant tilted showing a damping-off trend ( Figure 1D). As the disease worsened, the plant damped and died after 2-3 weeks. Therefore, F. foetens was identified as the pathogen of potato FW, and this fungus was discovered to be responsible for potato wilt for the first time.

Morphological Characteristics of F. foetens
The colony morphology of F. foetens is shown in Figure 1E,F. The aerial mycelia in the front colony formed thick white tufts evenly covering the entire Petri dish, whereas the reverse colony appeared purplish-red in the center. The colony diameter was 33.8 ± 1.6 mm on PDA after 4 days, and 57.0 ± 1.9 mm after 7 days. The average radial growth rate was 8.2 ± 0.3 mm/day. Conspicuous aerial mycelia gave older colonies a dotted appearance ( Figure 1G), which appeared to be stromata-supporting sporodochia [21]. However, no sporodochia occurred until 28 days of culture. Aerial mycelia bore solitary monophialides, occasionally polyphialides ( Figure 1H-J). The phialides were either cylindrical or slightly tapered toward the tip or narrowly flask-shaped, with the widest point in the middle (Figure 1J). The chlamydospores were globose to subglobose, rare, mostly terminal, and smooth ( Figure 1K,L). Microconidia that formed laterally from hyphae of the aerial mycelium were 0 to 3 septate and fusiform to slightly curved ( Figure 1M-S). It is worth noting that F. foetens produced a pungent distinct odor.

Pathogenicity Test
Three representative F. foetens isolates were selected randomly and inoculated into potato seedlings pots cultivated in the greenhouse. After 20 days of cultivation in the greenhouse, the seedlings in the negative control group did not exhibit any wilting symptoms ( Figure 1B). The tested fungi developed wilting symptoms and were re-isolated from the infected plants, indicating that they were the pathogen causing the potato wilt. After The antifungal activity of B. subtilis SF1 against F. foetens was evaluated in vitro. As shown in Figure 2A, F. foetens was significantly inhibited by B. subtilis SF1 in the confrontation assay, revealing 52.50 ± 2.59% inhibited growth. To investigate the antifungal activity of the metabolites produced by B. subtilis SF1, the F. foetens spores were inoculated onto PDA mixed with different volumes of the SF1 culture filtrate. Figure 2C-E shows that the antifungal activity against F. foetens increased with an increase in the proportion of supplemented culture filtrate. When the volume ratios of the culture filtrate to water in the PDA medium were 3:2 and 5:0, growth rates were inhibited by 25.32 ± 1.96% and 55.76 ± 2.46%, respectively. Based on the F. foetens colony diameters in Figure 2B, the average mycelial growth rates were 3.6 ± 0.3 and 8.3 ± 0.3 mm/day on PDA with and without the B. subtilis SF1 culture filtrate, respectively. These results suggest that B. subtilis SF1 suppressed F. foetens with antifungal metabolites. antifungal activity against F. foetens increased with an increase in the proportion of supplemented culture filtrate. When the volume ratios of the culture filtrate to water in the PDA medium were 3:2 and 5:0, growth rates were inhibited by 25.32 ± 1.96% and 55.76 ± 2.46%, respectively. Based on the F. foetens colony diameters in Figure 2B, the average mycelial growth rates were 3.6 ± 0.3 and 8.3 ± 0.3 mm/day on PDA with and without the B. subtilis SF1 culture filtrate, respectively. These results suggest that B. subtilis SF1 suppressed F. foetens with antifungal metabolites.

Effect of B. subtilis SF1 on FW Incidence In Vivo
The efficacy of B. subtilis SF1 against potato FW induced by F. foetens was evaluated in vivo. The negative control seedlings did not exhibit any symptoms. The positive control seedlings (treated with F. foetens) presented 100% morbidity. Inoculating the B. subtilis SF1 spore suspension after infection resulted in a lower incidence rate of 54.44 ± 2.31%, which decreased by 45.56% compared to the positive control (Supplementary Figure S2). The healthy potato plants treated with B. subtilis SF1 spores grew better than the potato plants in the negative control, as they were taller and had more leaves (Supplementary Figure

Effect of B. subtilis SF1 on FW Incidence In Vivo
The efficacy of B. subtilis SF1 against potato FW induced by F. foetens was evaluated in vivo. The negative control seedlings did not exhibit any symptoms. The positive control seedlings (treated with F. foetens) presented 100% morbidity. Inoculating the B. subtilis SF1 spore suspension after infection resulted in a lower incidence rate of 54.44 ± 2.31%, which decreased by 45.56% compared to the positive control (Supplementary Figure S2). The healthy potato plants treated with B. subtilis SF1 spores grew better than the potato plants in the negative control, as they were taller and had more leaves (Supplementary Figure S2). These results indicate that B. subtilis SF1 alleviated the incidence of potato FW caused by F. foetens in vivo and promoted potato growth.

Production of Hydrolytic Enzymes
In total, 145 genes were annotated in the B. subtilis SF1 genome in the Carbohydrate-Active enZYmes database. Among the annotated genes, 54 genes encoded glycoside hydrolases, and most of these (61.12%) were involved in sugar metabolism ( Figure 3A,B). Cellulase, β-glucanase, chitinase, and protease are fungal cell wall degradation-associated enzymes. Hydrolase assays on agar plates indicated that B. subtilis SF1 secreted protease, glucanase, and cellulase ( Figure 3C-E). The undetected activity of chitinase on agar (data not shown) agrees with the result that no chitinase-encoding genes exist in the genome. The genes encoding beta-glucanase, endo-1,4-beta-glucanase, and extracellular protease in the genome of B. subtilis SF1 are shown in the Supplementary Table S1. To inactivate the heat-sensitive antifungal components in the B. subtilis SF1 metabolites, the B. subtilis SF1 culture filtrate was heated at 100 • C for 10 min, mixed with PDA to prepare the medium, and inoculated with F. foetens. Compared to the unheated sample (57.24 ± 2.37%), the inhibition of growth against F. foetens decreased (45.37 ± 2.07%) by heat treatment (Supplementary Figure S3). The remaining inhibited growth indicated that the vital antifungal components in the B. subtilis SF1 culture filtrate were heat-insensitive. Therefore, hydrolytic enzymes were not the antifungal ingredients that affected F. foetens in B. subtilis SF1. the heat-sensitive antifungal components in the B. subtilis SF1 metabolites, the B. subtilis SF1 culture filtrate was heated at 100 °C for 10 min, mixed with PDA to prepare the medium, and inoculated with F. foetens. Compared to the unheated sample (57.24 ± 2.37%), the inhibition of growth against F. foetens decreased (45.37 ± 2.07%) by heat treatment (Supplementary Figure S3). The remaining inhibited growth indicated that the vital antifungal components in the B. subtilis SF1 culture filtrate were heat-insensitive. Therefore, hydrolytic enzymes were not the antifungal ingredients that affected F. foetens in B. subtilis SF1.

Plant Growth-Promoting Determinants Produced by B. subtilis SF1
The genes involved in plant growth-promoting were summarized based on the genome analysis of B. subtlis SF1. As shown in Supplementary Table S2, B. subtilis SF1 contained a series of genes associated with root colonization, biofilm formation, swarming motility, and elicitation of plant basal defense. A clear halo zone was observed around the colony on CAS agar, indicating that B. subtilis SF1 produced siderophores ( Figure 4A). A hydrolysis halo appeared around the B. subtilis SF1 colony on the organophosphate agar plate ( Figure 4B), indicating that B. subtilis SF1 degrades organophosphates. No solubilization of inorganic phosphate was detected on the NBRIP plate (data not shown). No IAA was produced by B. subtilis SF1 as no color change occurred in the samples with or without tryptophan ( Figure 4C), which was consistent with the absence of IAA synthesis-related genes in the B. subtilis SF1 genome. B. subtilis SF1 synthesized metabolites to enhance plant

Plant Growth-Promoting Determinants Produced by B. subtilis SF1
The genes involved in plant growth-promoting were summarized based on the genome analysis of B. subtlis SF1. As shown in Supplementary Table S2, B. subtilis SF1 contained a series of genes associated with root colonization, biofilm formation, swarming motility, and elicitation of plant basal defense. A clear halo zone was observed around the colony on CAS agar, indicating that B. subtilis SF1 produced siderophores ( Figure 4A). A hydrolysis halo appeared around the B. subtilis SF1 colony on the organophosphate agar plate ( Figure 4B), indicating that B. subtilis SF1 degrades organophosphates. No solubilization of inorganic phosphate was detected on the NBRIP plate (data not shown). No IAA was produced by B. subtilis SF1 as no color change occurred in the samples with or without tryptophan ( Figure 4C), which was consistent with the absence of IAA synthesis-related genes in the B. subtilis SF1 genome. B. subtilis SF1 synthesized metabolites to enhance plant growth, illustrating why the potato seedlings treated with B. subtilis SF1 grew more vigorously than the negative control seedlings.  Figure 5A). Crude surfactin was purified by preparative HPLC. The molecular mass of the purified surfactin was measured by liquid chromatography-mass spectroscopy (LC-MS). The four tested surfactin homologs produced main peaks at m/z 1030 (S1), 1044 (S2), 1044 (S3), and 1058 (S4), corresponding to molecular masses of 1007 (S1; C13), 1021 (S2 and S3; C14), and 1035 (S4; C15) Da, respectively (Supplementary Figure S4). S2 and S3 were isomers with the same molecular weight of 1021 Da. The relative contents of S1, S2, S3, and S4 in purified surfactin were 16.67, 19.51, 20.23, and 42.89%, respectively. growth, illustrating why the potato seedlings treated with B. subtilis SF1 grew more vigorously than the negative control seedlings.

Effect of Surfactin on the Growth and Morphology of F. foetens
The effect of surfactin on the growth of F. foetens was evaluated using the filter disk method ( Figure 5B). The F. foetens mycelia initially spread over the filter paper as if surfactin had no inhibitory effect on the growth of F. foetens. After the 30-day incubation, the mycelia on/around the filter paper appeared to have significantly collapsed. The mycelia around the filter paper gradually began to dissolve and almost disappeared, indicating a clear circle of inhibition. Surfactin effectively inhibited the growth of F. foetens at a concentration of 20 μg/μL. Higher concentrations resulted in larger zones of inhibition. The an-

Effect of Surfactin on the Growth and Morphology of F. foetens
The effect of surfactin on the growth of F. foetens was evaluated using the filter disk method ( Figure 5B). The F. foetens mycelia initially spread over the filter paper as if surfactin had no inhibitory effect on the growth of F. foetens. After the 30-day incubation, the mycelia on/around the filter paper appeared to have significantly collapsed. The mycelia around the filter paper gradually began to dissolve and almost disappeared, indicating a clear circle of inhibition. Surfactin effectively inhibited the growth of F. foetens at a concentration of 20 µg/µL. Higher concentrations resulted in larger zones of inhibition. The antifungal effect of surfactin was significant but was slow and persistent.
To investigate the effect of surfactin on the F. foetens mycelial morphology, 1 µg/µL surfactin was used to treat the mycelia. The results of SEM and fluorescence microscopy are shown in Figure 5C-H. The mycelia of the control had smooth edges and uniform thickness ( Figure 5C-F), while the surfactin-treated mycelia appeared to have an uneven, broken, and wrinkled surface (Figure 5D,E,G,H). The most significant morphological changes to the F. foetens mycelium caused by surfactin were abnormal vacuolation and swelling.

Effect of Surfactin on F. foetens Membrane Integrity
The effect of surfactin on F. foetens membrane integrity was reflected by the release of nucleic acids and proteins. F. foetens was cultured in PDB supplemented with different concentrations of surfactin, and the OD 260 and OD 280 values of the extracellular fluid were assessed. As shown in Figure 6A,B, the OD 260 values increased with increasing surfactin concentration, and a similar profile was observed for OD 280 . These results demonstrated that nucleic acids and proteins were rapidly released into the external fungal body after the surfactin treatment. The leakage of nucleic acids and proteins suggested that the surfactin produced by B. subtilis SF1 damaged F. foetens membrane integrity.

Effect of Surfactin on F. foetens Protein Expression
To determine the effect of surfactin on protein expression by F. foetens, the total intracellular proteins of F. foetens were collected and analyzed by SDS-PAGE after exposure to surfactin (0.5 μg/μL) for 120 h. As shown in Figure 6C, surfactin decreased the densities of some of the protein bands, and many of the bands were almost undetectable. The presence of surfactin also resulted in the appearance of several protein bands with significantly different molecular weights. These results suggest that surfactin produced by B.

Effect of Surfactin on F. foetens Protein Expression
To determine the effect of surfactin on protein expression by F. foetens, the total intracellular proteins of F. foetens were collected and analyzed by SDS-PAGE after exposure to surfactin (0.5 µg/µL) for 120 h. As shown in Figure 6C, surfactin decreased the densities of some of the protein bands, and many of the bands were almost undetectable. The presence of surfactin also resulted in the appearance of several protein bands with significantly different molecular weights. These results suggest that surfactin produced by B. subtilis SF1 induced differential protein expression in F. foetens.

Effect of Surfactin on GSH Production by F. foetens
GSH is an important intracellular non-enzymatic antioxidant. To determine the effect of surfactin on F. foetens intracellular GSH content, an F. foetens culture was treated with different concentrations of surfactin for 120 h. As shown in Figure 6D, the surfactin treatment resulted in increased GSH production, and GSH content increased with increasing concentrations of surfactin.

Effect of Surfactin on F. foetens Transcription Levels
The transcription levels of some genes related to basal metabolism and antioxidation were determined to investigate the effect of surfactin on various metabolic pathways. As shown in Figure 6E, the transcription levels of genes, such as FBA, PFK, and PGK in glycolysis, IDH in the TCA pathway, GSY1 in glycogen metabolism, PCK1 and GPM in gluconeogenesis, and FAS1 in fatty acid metabolic processes, were downregulated. The gene expression levels of CPX, GSH-PX, and GR were upregulated.

Interaction between Surfactin and ff-DNA by Electrophoresis, UV-Visible
Spectroscopy and Fluorescence Spectroscopy ff-DNA was treated with different concentrations of surfactin and analyzed by electrophoresis. As shown in Figure 7A, the bands dimmed as the surfactin concentration was increased, the bands dimmed. DNA bound by surfactin was not displayed on the agarose gel, the results indicate that surfactin bound to ff-DNA.
Intercalations, grooves, and electrostatic interactions are non-covalent binding modes by which small molecules interact with DNA [42]. The UV-visible spectra of ff-DNA showed a slight decrease at 260 nm and no redshift ( Figure 7B). The lack of a redshift indicates that the binding mode was not intercalation [43]. The small hypochromic effect indicates that the binding mode of surfactin to DNA might be groove or electrostatic interaction [44].
Because surfactin does not fluoresce, the binding mode of ff-DNA with surfactin was further confirmed by fluorescent EB-competitive binding. EB is a typical DNA-intercalation marker [45]. As shown in Figure 7C, the intensity of the emission band at 600 nm of the EB-DNA system decreased with increasing concentrations of surfactin when the EB-DNA solution was excited at 485 nm. Fluorescence quenching was due to the replacement of EB from the complex (EB-DNA) or when a new complex (EB-DNA-surfactin) formed. The fluorescence quenching constant was determined by the Stern-Volmer equation: F 0 /F = 1 + K q τ 0 [Surfactin] = 1 + K SV [Surfactin], where F 0 and F indicate the fluorescence intensities without and with the quencher surfactin, respectively, [Surfactin] is the concentration of surfactin, τ 0 is the average fluorescence lifetime of the molecule and its value is 10 −8 s. K SV is the Stern-Volmer quenching constant, which was obtained from the slope of F 0 /F versus [surfactin] plot ( Figure 7D). Based on the calculation, the quenching rate constant, K q , was 1.07 × 10 11 L/mol·s, which is greater than the maximum diffusion collision quenching rate constant of quenching agents on biomolecules (2 × 10 10 L/mol·s). These results confirm that the quenching of EB-DNA fluorescence was initiated by the formation of a complex between surfactin and EB-DNA through a static mechanism rather than dynamic collision.
3.3.7. Interaction between Surfactin and ff-DNA by Electrophoresis, UV-Visible Spectroscopy and Fluorescence Spectroscopy ff-DNA was treated with different concentrations of surfactin and analyzed by electrophoresis. As shown in Figure 7A, the bands dimmed as the surfactin concentration was increased, the bands dimmed. DNA bound by surfactin was not displayed on the agarose gel, the results indicate that surfactin bound to ff-DNA. Intercalations, grooves, and electrostatic interactions are non−covalent binding modes by which small molecules interact with DNA [42]. The UV−visible spectra of ff-DNA showed a slight decrease at 260 nm and no redshift ( Figure 7B). The lack of a redshift indicates that the binding mode was not intercalation [43]. The small hypochromic effect indicates that the binding mode of surfactin to DNA might be groove or electrostatic interaction [44].
Because surfactin does not fluoresce, the binding mode of ff-DNA with surfactin was further confirmed by fluorescent EB−competitive binding. EB is a typical DNA-intercalation marker [45]. As shown in Figure 7C Figure 7D). Based on the calculation, the quenching rate constant, Kq, was 1.07 × 10 11 L/mol·s, which is greater than the maximum diffusion collision quenching rate constant of quenching agents on biomolecules (2 × 10 10 L/mol·s). These  [43]. In our experiment, the K b value of the surfactin was 2.97 × 10 4 M −1 , which was lower than that expected for an intercalating compound but was consistent with one that undergoes groove binding [40]. Moreover, only molecules with planar aromatic rings are involved in intercalation between DNA base pairs. The non-planar nature of the compounds also shows the non-intercalative binding with DNA [46,47]. The binding mode of ff-DNA and surfactin was electrostatic interaction or groove. Electrostatic binding occurs due to the interaction between the negatively charged phosphate backbone of DNA and the positively charged ends of small molecules [42]. Surfactin has a net negative charge [48]. Thus, all of these results support that surfactin undergoes groove binding to ff-DNA.

Discussion
FW induced by Fusarium is a severe disease that occurs in potato cultivation worldwide. It is difficult to prevent and control FW by selecting resistant varieties or using chemical agents [49]. B. subtilis spp. have been used to control FW on crops, such as cucumber, pepper, tomato, and muskmelon [50][51][52][53]. In this study, F. foetens was identified as the cause of potato FW. However, systematic studies on preventing F. foetens are unavailable. Surfactin produced by B. subtilis SF1 exhibited significant inhibitory activity against F. foetens, and the underlying antifungal mechanisms of surfactin were revealed.
F. foetens closely resembles F. oxysporum [54]. The colony and spore morphology of F. foetens are very similar to those of F. oxysporum. A pungent colony odor distinguished the fungus from F. oxysporum. The phylogenetic analysis indicated that F. foetens was closely related to F. oxysporum. F. oxysporum is well-known to cause FW during potato production.
The closest relative provided some evidence for their similar morphology and the ability to infect potatoes to induce FW.
B. subtilis SF1 exhibited significant inhibitory activity against F. foetens in the in vitro assay. Growth inhibition of 55.76 ± 2.46% was observed when adding the B. subtilis SF1 culture filtrate to PDA, indicating that the antifungal activity was attributed to metabolites produced by B. subtilis SF1. Extracellular enzymes produced by Bacillus, such as chitinase, protease, β-glucanase, and cellulose, affect the growth and morphology of fungi [55,56]. These enzymes degrade the hyphal cell wall, change its permeability, and cause the hyphae to swell [57][58][59]. The hydrolytic enzymes perhaps were not the key antifungal ingredients produced by B. subtilis SF1, due to their lower activities on agar as well as the absence of chitinase-coding genes in the SF1 genome. The growth inhibition was still significant (45.37%) even after the B. subtilis SF1 culture filtrate was heated at 100 • C for 10 min, indicating that key antifungal components in B. subtilis SF1 were heat-resistant. These results suggest that hydrolytic enzymes made a limited contribution to the antifungal activity of B. subtilis SF1.
The secondary metabolites of B. subtilis SF1 were predicted with AntiSMASH. The presence and antifungal activity of surfactin was confirmed by HPLC and the filter paper disk method, respectively. Interestingly, the inhibiting effect of surfactin was slow and persistent. The F. foetens mycelia initially spread and grew on the filter paper. However, a clear inhibition zone became apparent around the filter paper disk after 30 days of exposure to surfactin. The antifungal characteristics of surfactin were significantly different from fengycin (another lipopeptide produced by Bacillus) and carbendazol, which directly inhibit the growth of fungus and form a clear circle of inhibition [39,60]. Various lipopeptide antibiotics produced by B. subtilis are highly stable at high temperatures, with more than 90% of its activity being retained even after samples had been held at 100 • C for 2 h [61,62]. The thermal resistance of surfactin is consistent with the result that heating did not destroy the activity of the antifungal substances in B. subtilis SF1. Thus, surfactin may play a key role in the activities of B. subtilis SF1 against F. foetens.
Based on the in vitro antifungal assays with the filter paper disk, surfactin produced by B. subtilis SF1 acted on the growing mycelia and eventually led to the death and lysis of the F. foetens mycelia. What are the underlying mechanisms used by surfactin to inhibit F. foetens? Surfactin induces ion-conducting pores in artificial lipid membranes and alters membrane permeabilization [63,64]. The leakage of nucleic acids and proteins of F. foetens after exposure to surfactin confirmed that surfactin damaged the membrane integrity of F. foetens. In a previous study, surfactin [∆Leu6] (lacking amino acid Leu-6) possessed significant antifungal activity against F. moniliforme and induced leakage of nucleic acids and proteins [17]. On the other hand, surfactin changes mycelial morphology. The surfactin-treated mycelium appeared to have an uneven, broken, and wrinkled surface, with swelling and vacuolations. Similar morphological abnormalities were observed when B. subtilis IBFCBF-4 was co-cultured with F. oxysporum [65]. Hu et al. reported that fengycin changes the permeability of the cell membrane and causes F. moniliforme hyphae to appear vacuolated [66]. Surfactin may dehydrate and interact with the phospholipid acyl chains, resulting in considerable membrane fluidization [64]. Surfactin produced by B. subtilis SF1 changed the membrane permeabilization of F. foetens cells, inducing abnormal mycelial morphology and lysis of intracellular substances.
Can surfactin inhibit the fungus via other mechanisms? Surfactin is involved in the interaction between B. subtilis and Aspergillus niger, resulting in altered metabolism in the bacterium and the fungus [67]. Several new protein bands and a significant fading of the original protein bands on SDS-PAGE demonstrated that surfactin caused differential protein expression in F. foetens. Surfactin extracted from Brevibacillus brevis KN8(2) alters the protein expression of F. moniliforme on SDS-PAGE [19]. The results of qRNA revealed that the transcription level of some key genes during glycolysis (FBA, PFK, PGK), TCA pathway (IDH), glycogen metabolism (GSY1), gluconeogenesis (PCK1, GPM), and fatty acid metabolic processes (FAS1) were downregulated. Surfactin causes a significant change in the transcriptional profiling of Candida albicans, and there are 773 and 617 genes with at least a 1.5-fold increase or decrease in transcription, respectively [68]. Therefore, the growth inhibition of F. foetens was related to reduced expression of the enzymes involved in the basal metabolism of F. foetens. Surfactin-induced differential protein expression may have affected various fungal metabolic pathways.
Surfactin produced by Bacillus spp. induced an increase in reactive oxygen species (ROS), causing the death of Magnaporthe grisea [69]. Krishnan et al. presumed that surfactin induces ROS production in F. moniliforme causing protein and DNA damage [19]. However, no significant ROS changes were detected in our results (data not shown), and the CPX and GSH-PX genes involved in antioxidative defense were upregulated, indicating that ROS were not involved in inhibiting F. foetens. Surfactin exposure increased GSH production by F. foetens, which agreed with the upregulation of glutathione reductase-encoding genes and agrees with the results of Ágnes et al. [68]. GSH has free radical scavenging, anti-oxidation, and electrophile elimination activities [70]. The increase in the level of GSH in F. foetens was a response to surfactin.
DNA carries the genetic information of the entire organism and plays a significant role in the transcription, expression, growth, and development of organisms. DNA is also a target for exogenous substances to attack and cause changes in the body, leading to disease [71]. The lipopeptides, fengycin and iturin, bind with DNA, and the binding modes have been confirmed [39]. In our study, the interaction between surfactin and ff-DNA in groove binding mode was confirmed by UV-visible and fluorescence spectroscopy. Groove binders affect gene expression in vitro by inhibiting sequence selective binding of various transcription factors to DNA [72]. These effects may result in the expression or repression of downstream genes. Groove binding also results in structural distortion or damage to DNA that impedes replication and transcription by blocking the movement of helicases, topoisomerases, and polymerases [73]. We presumed that the interaction between surfactin and ff-DNA potentially altered replication or transcription, which was part of the mechanism by which B. subtilis SF1 inhibited F. foetens. However, the changes induced by the interaction between surfactin and ff-DNA and its importance in the antifungal effect need further study.
Surfactin mounts innate immunity in plants and thus decreases the invasive growth caused by fungal infection [12]. Inoculating B. subtilis SF1 after infection led to a significant decrease in the FW incidence rate (45.56%), and healthy potato seedlings treated with B. subtilis SF1 appeared to grow vigorously, compared to untreated potato seedlings. Surfactin is essential for biofilm formation and surface motility of plant growth-promoting organisms [74,75]. Siderophore production and phosphate solubilization, both important for plant growth, were detected in B. subtilis SF1. B. subtilis SF1 also harbors many genes associated with root colonization, biofilm formation, swarming motility, and elicitation of plant basal defense. Thus, the alleviation of morbidity by B. subtilis SF1 was a combined effect of antifungal activity by surfactin and the secretion of various plant growth-promoting determinants.

Conclusions
The present study is the first to report that F. foetens infects potatoes and causes a large amount of FW disease. Surfactin was the main antifungal component produced by B. subtilis SF1. Surfactin inhibited mycelial growth, causing malformed mycelial morphology, cell leakage, differential protein expression, and GSH production. The underlying mechanisms of surfactin against F. foetens were presumed to be: (i) Surfactin altered cell membrane permeability, leading to lysis of cell contents and morphological malformation of the mycelia; (ii) surfactin caused differential protein expression and changes in cell metabolic pathways; (iii) surfactin bound to DNA in groove binding mode, impeding replication or transcription. Surfactin was a mixture of four homologs in our study. The presence of different types of surfactin in different proportions resulted in diverse antimicrobial activities. The antifungal activity of a single fraction or a mixture of different fractions of the surfactin produced by B. subtilis SF1 needs to be determined in further studies.
Supplementary Materials: The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/jof9030367/s1, Figure S1: Phylogenic tree inferred from the combined dataset of EF-1 α and β-tubulin based on the Neighbor-Joining method. Fusarium foetens isolate is the pathogen of potato wilt described in the study. The scale bar indicates genetic distance. Bootstrap values in 1000 replicates for major lineages at the nodes were shown as percentages; Figure S2 Table S1: The genes encoding beta-glucanase, endo-1,4-beta-glucanase, and extracellular protease in the genome of B. subtilis SF1; Table S2: Representative genes involved in plant growth promotion in the genome of B. subtilis SF1; Table S3: The predicted secondary metabolites and the corresponding synthesis gene clusters in the genome of B. subtilis SF1.
Author Contributions: H.Y. and L.L. designed and supervised this project; sampling, fungal isolation, description, sequencing and phylogenetic analysis were finished by X.L. and P.L.; X.J. and B.L. carried out the analysis of antifungal mechanisms; L.L. and X.J. wrote the manuscript; L.G. assisted during manuscript preparation. All authors have read and agreed to the published version of the manuscript. Data Availability Statement: The whole genome shotgun project for Bacillus subtilis SF1 has been deposited at DDBJ/ENA/Genbank under the accession JANYMH000000000. The sequences of ITS, EF-1α, and β-tubulin for Fusarium foetens have been submitted to NCBI with the accession number OM370930, OM370932, and OM370931, respectively. The authors state that the current study existing the necessary data for the conclusion and Supplementary Material will be fully exposed.