Identiﬁcation and Characterization of an Antifungal Gene Mt1 from Bacillus subtilis by Affecting Amino Acid Metabolism in Fusarium graminearum

: Fusarium head blight is a devastating disease that causes signiﬁcant economic losses world-wide. Fusarium graminearum is a crucial pathogen that requires close attention when controlling wheat diseases. Here, we aimed to identify genes and proteins that could confer resistance to F. graminearum . By extensively screening recombinants, we identiﬁed an antifungal gene, Mt 1 (240 bp), from Bacillus subtilis 330-2. We recombinantly expressed Mt 1 in F. graminearum and observed a substantial reduction in the production of aerial mycelium, mycelial growth rate, biomass, and pathogenicity. However, recombinant mycelium and spore morphology remained unchanged. Transcriptome analysis of the recombinants revealed signiﬁcant down-regulation of genes related to amino acid metabolism and degradation pathways. This ﬁnding indicated that Mt 1 inhibited amino acid metabolism, leading to limited mycelial growth and, thus, reduced pathogenicity. Based on the results of recombinant phenotypes and transcriptome analysis, we hypothesize that the effect of Mt 1 on F. graminearum could be related to the metabolism of branched-chain amino acids (BCAAs), the most affected metabolic pathway with signiﬁcant down-regulation of several genes. Our ﬁndings provide new insights into antifungal gene research and offer promising targets for developing novel strategies to control Fusarium head blight in wheat.


Introduction
Fusarium graminearum is a fungal pathogen that infects a variety of cereal crops, including wheat, barley, and maize [1,2]. The disease caused by Fg is commonly known as Fusarium head blight (FHB) or scab, and it is a significant economic problem for farmers worldwide [3,4]. The FHB reduces crop yield and quality and contaminates grains with mycotoxins including deoxynivalenol (DON) and zearalenone (ZEA), which pose significant health risks to animals and humans [5][6][7]. Antifungal genes are genes that encode proteins that can inhibit the growth of fungi. These genes have been identified in various organisms, including plants, animals, and bacteria. B. subtilis is a well-studied model strain known for its antifungal properties. It produces various enzymes, including chitinases, lipases, and proteases, that target fungi. For instance, the SG6 strain of B. subtilis secretes lipopeptides that can inhibit the growth of Fg mycelium [8], and the ATCC6633 strain produces mycosubtilin, which has been shown to inhibit both the growth and virulence of Fg [9]. Additionally, the B3 strain of B. subtilis produces iturins and fengycin, both of which are effective inhibitors of Fg [10]. In a study by Aktuganov et al., among 70 Bacillus spp. strains tested, 19 strains exhibited chitinolytic activity [11]. Until now, research on antifungal genes has been primarily focused on identifying active substances and synthesizing their pathways from isolated antagonistic strains. However, we have employed an alternative achieve this goal, we first fragmented the genomic DNA of B. subtilis ( Figure S1a) using ultrasonic cleavage, which ensured random fragmentation of the genomic DNA and allowed for control of fragmentation size and integrity through sonication time, number of sonications, and sonication interval (Figure 1a). After several optimizations, we determined that sonication conditions of 2 s sonication time, 8 s sonication interval, and 30 sonications were optimal for producing DNA fragments ranging in size from 100 to 2000 base pairs, with a concentration between 750 and 1500 base pairs ( Figure S1b). We constructed these fragments into a psxsh-Neo expression vector to create a recombinant expression plasmid library. To assess the quality of the library, we randomly selected transformants to verify the recombination rates and fragment size distribution. Our results showed a recombination rate of 89.6% ( Figure S1c) and a majority of recombinant fragments ranging from 100 to 1000 base pairs, which is advantageous for screening antifungal genes ( Figure S1d). A library of approximately 20,000 recombinant plasmids for subsequent protoplast transformation was obtained.  FgMt1 strain growth rate was significantly reduced on PDA medium and did not differ significantly from the growth rate of the duplicate recombinants (FgMt1-R). A t-test was used for this experimental analysis, "*" indicates p < 0.05, and "ns" indicates no significant difference. The experiment was repeated three times independently, and three single colonies per group were used for the experiment each time.

Effect of Recombinant Gene on Recombinant Strain
The recombinant FgMt1 stain obtained by phenotypic screening exhibited abnormal growth on PDA medium with a slow mycelial growth rate and did not produce aerial mycelium. To assess its impact on FgMt1 hyphae and spore morphology, we stained the septum and nucleus with CFW and DAPI, respectively, and observed them using fluorescence microscopy. Our results showed that the morphology of FgMt1 mycelium and spores was the same for both the wild-type and negative control Fg-GFP (Figure 2a). We The FgMt1 strain growth rate was significantly reduced on PDA medium and did not differ significantly from the growth rate of the duplicate recombinants (FgMt1-R). A t-test was used for this experimental analysis, "*" indicates p < 0.05, and "ns" indicates no significant difference. The experiment was repeated three times independently, and three single colonies per group were used for the experiment each time.
The gene expression library was introduced into the pls1 locus of Fg wild strain 5035 through protoplast transformation and homologous recombination (Figure 1b). Following screening for phenotypic abnormalities, a strain containing the same GFP gene at the pls1 locus was generated as a negative control. The observed phenotypic abnormalities, including altered growth rates, abnormal pigment production, and changes in aerial mycelial volume, suggest that the expression of the recombinant genes may impact certain metabolic and regulatory pathways in Fg and that the recombinant genes have potential as candidates for antifungal genes. Figure 1c displays some examples screened from the 13,000 recombinants. Notably, a recombinant designated FgMt1 was obtained during the screening process, which exhibited a significantly reduced growth rate and no aerial mycelium production on PDA medium (Figure 1d). To verify the accuracy and stability of this recombinant, we examined the recombinant site, the insertion of the pls site ( Figure S2), and the recombinant gene was amplified, reconstructed on the psxsh-Neo plasmid ( Figure S3), and retransferred into Fg-5035 with the new recombinant named FgMt1-R, which yielded recombinants with identical phenotypes. The recombinant gene was then utilized to investigate its effect on the transformed strains.

Effect of Recombinant Gene on Recombinant Strain
The recombinant FgMt1 stain obtained by phenotypic screening exhibited abnormal growth on PDA medium with a slow mycelial growth rate and did not produce aerial mycelium. To assess its impact on FgMt1 hyphae and spore morphology, we stained the septum and nucleus with CFW and DAPI, respectively, and observed them using fluorescence microscopy. Our results showed that the morphology of FgMt1 mycelium and spores was the same for both the wild-type and negative control Fg-GFP (Figure 2a). We further harvested spores from strains Fg-5035, Fg-GFP, and FgMt1 after 5 days of growth in CMC spore-producing medium and counted the number of spore septa and spores. No significant difference was observed in the number of spore septa, the number of spores corresponding to the number of septa (Figure 2b), or spore production (Figure 2c), indicating that the recombinant gene did not affect the morphology of mycelium or the morphology and production of spores. However, FgMt1 exhibited a significant difference in mycelial dry weight compared to the wild type and negative control (Figure 2d). To determine the effect of FgMt1 on pathogenicity, we used wheat coleoptile inoculation to study its impact. Our results showed that wheat coleoptile treated with Fg-5035 and Fg-GFP spore suspensions exhibited severe stem disease and browning necrosis, whereas those treated with FgMt1 spore suspension grew healthy without any disease symptoms ( Figure 2e). Statistical analysis revealed that the length of the disease spots in Fg-5035 and Fg-GFP spore suspension-inoculated wheat coleoptile was significantly higher than that of FgMt1 spore suspension-treated wheat coleoptile (Figure 2e). Confocal microscopy of wheat stem tissues from the inoculated sites showed that the intercellular spaces of negative control seedling tissues were filled with mycelium, whereas no mycelium was observed in FgMt1-treated wheat seedling tissues, and the plant tissues remained intact ( Figure 2f). Overall, our findings indicate that the recombinant gene had no significant effect on spore and mycelial morphology but had a significant impact on mycelial production and pathogenicity.

Exogenous Additive Partially Restores Recombinant Growth Defects
In order to investigate the effect of medium composition on the growth of FgMt1, we selected seven media with different nutrient compositions (Figure 3a and medium ingredients in Supplementary Materials). We compared the growth rates of Fg-5035, Fg-GFP, and FgMt1 on these seven media ( Figure 3b). After 72 h of incubation, we found that the growth rate of FgMt1 on CM medium was significantly higher than that on PDA medium. The growth rate of FgMt1 on YEPD medium was not significantly different from that of the control, but the aerial mycelium volume was significantly increased. However, FgMt1 growth remained slow on the other media, and comparison of the growth rates of PDA and PSA showed that sugars were not the main cause of the growth rate difference in FgMt1. The amount of aerial mycelial biomass was determined by the height of the colony's longitudinal section (Figure 3c). Only on the YEPD medium did FgMt1 produce aerial mycelium, which was significantly lower than the control but had a significant increase in mycelial production compared to other media, where no aerial mycelium was produced (Figure 3d). This might suggest that the exogenous addition of certain nutrients plays a role in the phenotypic recovery of the recombinants and might also indicate that these substances have critical roles in the recovery of the recombinant growth defects. To verify that the components in YEPD could restore the slow mycelial growth and the absence of aerial mycelium production of FgMt1 on PDA medium, although FgMt1 can produce mycelium, it can't produce aerial mycelium on the PDA medium compared to the wild type (as shown in Figure 3c), so we added yeast extract and peptone, respectively, to the PDA media ( Figure 3e). It can be clearly seen that the growth rate of FgMt1 mycelium increased in the PDA medium with the addition of both substances (Figure 3f), and although there is still a gap compared to the control, some aerial mycelium could be produced. Such results suggest that exogenous addition of nutrients can restore to some extent the mutant phenotype of FgMt1 to wild-type, which is most likely due to its own inability to synthesize certain nutrients, and that this phenotype can be compensated for and restored by exogenous addition of some substances. cating that the recombinant gene did not affect the morphology of mycelium or the morphology and production of spores. However, FgMt1 exhibited a significant difference in mycelial dry weight compared to the wild type and negative control (Figure 2d). To determine the effect of FgMt1 on pathogenicity, we used wheat coleoptile inoculation to study its impact. Our results showed that wheat coleoptile treated with Fg-5035 and Fg-GFP spore suspensions exhibited severe stem disease and browning necrosis, whereas those treated with FgMt1 spore suspension grew healthy without any disease symptoms (Figure 2e). Statistical analysis revealed that the length of the disease spots in Fg-5035 and Fg-GFP spore suspension-inoculated wheat coleoptile was significantly higher than that of FgMt1 spore suspension-treated wheat coleoptile (Figure 2e). Confocal microscopy of wheat stem tissues from the inoculated sites showed that the intercellular spaces of negative control seedling tissues were filled with mycelium, whereas no mycelium was observed in FgMt1-treated wheat seedling tissues, and the plant tissues remained intact (Figure 2f). Overall, our findings indicate that the recombinant gene had no significant effect on spore and mycelial morphology but had a significant impact on mycelial production and pathogenicity.

Transcriptome Sequencing to Analyze the Impact Caused by Recombinant Genes
To identify how the recombinant gene affects the recombinant at the transcriptio level, we determined the transcriptome of FgMt1 and analyzed the pathways that the combinant gene may affect. The wild-type 5035 strain and FgMt1 were shake-cultured

Transcriptome Sequencing to Analyze the Impact Caused by Recombinant Genes
To identify how the recombinant gene affects the recombinant at the transcriptional level, we determined the transcriptome of FgMt1 and analyzed the pathways that the recombinant gene may affect. The wild-type 5035 strain and FgMt1 were shake-cultured in PDB medium for 3 days, and the mycelia were collected and used to determine the transcriptome. Principal component analysis of the transcriptome data of strains FgMt1 and Fg5035 showed very strong correlation between sample replicates ( Figure S6). The correlation heatmap analysis showed high correlation among biological replicates of samples with R-values of 0.995 or greater ( Figure S7). Analysis of FgMt1 recombinant gene expression revealed 1996 genes up-regulated, 1977 genes down-regulated, and no significant difference in the expressions of 5483 genes ( Figure S5). The down-regulated genes were found to be enriched in amino acid metabolism pathways, which correlates with the observed phenotypic effects of the recombinant gene on FgMt1 (Figure 4). The results of transcriptome analysis and phenotypic effects suggest that the recombinant gene may have impacted the amino acid metabolic pathway in Fg, leading to down-regulation of related genes, thereby resulting in restricted mycelial growth, limited aerial mycelial production, and reduced pathogenicity. Notably, in the KEGG metabolic pathway analysis, we found that many of the down-regulated genes were enriched in the valine, leucine, and isoleucine degradation and biosynthesis pathways, and they are components of the branched-chain amino acids (BCAAs). The BCAA metabolic pathway is significantly down-regulated in FgMt1, therefore the growth and development of Fg are affected.

EER REVIEW 8 of 15
and isoleucine degradation and biosynthesis pathways, and they are components of the branched-chain amino acids (BCAAs). The BCAA metabolic pathway is significantly down-regulated in FgMt1, therefore the growth and development of Fg are affected.

Recombinant Gene Sequence Alignment and Bioinformatics Analysis
The novel antifungal gene sequence Mt1 (240 bp) was identified from B. subtilis strain

Recombinant Gene Sequence Alignment and Bioinformatics Analysis
The novel antifungal gene sequence Mt1 (240 bp) was identified from B. subtilis strain 330-2 isolated from rapeseed, which has not yet been fully genomically sequenced. Through comparison with the Ensembl Bacteria database, we found that Mt1 is homologous to a 234 bp sequence (KS08-04975) in B. subtilis str. ATCC 13952 (GCA_000772125) ( Figure S11; Table S1). However, the polypeptide encoded by Mt1 is distinct from the homologous sequence (AIW33019), making it a novel protein sequence. The 80-amino acid polypeptide Mt1 exhibits a multi-segmented helix structure and a segmented strand structure, as predicted using computational methods (Figure 5a,c,d). The structure of the polypeptide is critical for its biological function, as we predicted using Materials Approach 4.6. The secondary structure of Mt1 is possibly related to its function and is likely to be the key point of its influence on the metabolic pathway of BCAAs.

Discussion
The expression of the FgMt1 gene has a significant impact on the amino acid me bolic pathway in wild-type Fg-5035. This impact leads to the down-regulation of ma genes related to the pathway. Supplementation of semi-hydrolysis products, such as try tone or yeast extract, can restore some growth defects, including mycelial growth rate a aerial mycelial production. It appears that the FgMt1 gene affects the ability of F. gramin arum to metabolize essential amino acids, which in turn prevents it from synthesizing utilizing nutrients on its own, forcing the fungus to restore its growth capacity on through exogenous additions. While the mechanism of the effect of the FgMt1 gene is n yet clear, it appears that Mt1 affects the pathway of branched-chain amino acid (BCA

Discussion
The expression of the FgMt1 gene has a significant impact on the amino acid metabolic pathway in wild-type Fg-5035. This impact leads to the down-regulation of many genes related to the pathway. Supplementation of semi-hydrolysis products, such as tryptone or yeast extract, can restore some growth defects, including mycelial growth rate and aerial mycelial production. It appears that the FgMt1 gene affects the ability of F. graminearum to metabolize essential amino acids, which in turn prevents it from synthesizing or utilizing nutrients on its own, forcing the fungus to restore its growth capacity only through exoge-nous additions. While the mechanism of the effect of the FgMt1 gene is not yet clear, it appears that Mt1 affects the pathway of branched-chain amino acid (BCAA) metabolism in F. graminearum. Key genes in this pathway, including FgILv1, FgILv2, FgILv3, FgILv5, and FgILv6, are involved in the synthesis and degradation of BCAAs and significantly affect the mycelial growth and virulence of Fg [34,35,39,40]. Knockout mutants of FgILv2 and FgILv6 showed some recovery of the nutritional defects on YEPD medium, consistent with experimental results (Figure 3), and four of five BCAA-related genes were significantly down-regulated in expression in the transcriptome data ( Figure S8). These findings suggest that the FgMt1 gene impacts the BCAA metabolic pathway and affects the growth and virulence of F. graminearum.
BCAAs, a group of amino acids comprising leucine, isoleucine, and valine, are essential for protein synthesis and serve as a source of energy and biosynthetic precursors for cellular processes. In fungi, BCAA metabolism plays a critical role in the synthesis of proteins, lipids, and nucleotides, as well as the maintenance of cellular homeostasis [41][42][43]. Fungal growth, development, and stress responses rely heavily on BCAA metabolism, with BCAA catabolism necessary for the utilization of alternative nitrogen sources and for proper mitochondrial function and cellular redox balance. In particular, the leucine biosynthesis pathway ( Figure S10) and the ILV and LEU genes are crucial to BCAA metabolism-related pathways in fungi [44], with FgLEU1 playing a critical role in leucine biosynthesis and full virulence in Fg [32]. Rajagopal et al. reported that several key enzymes in the leucine synthesis pathway were encoded by FGSG_12952, FGSG_09589, FGSG_10671, FGSG_06675, and FGSG_09512 in Fg [33]. In the FgMt1 comparison, several genes related to the BCAA pathway were significantly down-regulated in expression ( Figure S9), supporting the hypothesis that FgMt1 has a significant impact on the BCAA pathway in F. graminearum.
In F. graminearum, the BCAA metabolic pathway is important for the production of virulence factors and secondary metabolites such as mycotoxins, and its breakdown products are substrates for many primary/secondary metabolites [45]. Undoubtedly, the importance of the BCAAs metabolic pathway for the growth and development of F. graminearum leads to its use as a target pathway for the study of antifungal proteins and some small proteins that exhibit antifungal activity which were isolated from various organisms, including plants and animals [46][47][48]. Antifungal proteins target various cellular processes in fungi, including cell wall synthesis, protein synthesis, cell membrane integrity, spore production, and production of secondary metabolites [49][50][51][52]. Targeting the metabolic pathway of BCAAs could provide a new approach to the study of antifungal proteins, as it is essential for the growth and pathogenicity of F. graminearum. The advantages of Mt1 as an antifungal protein for research are: (1) the gene sequence is derived from the B. subtilis genome and is extremely easy to obtain; (2) its amino acid sequence is shorter and less difficult to manipulate, whether for secretory expression or optimal modification; (3) it already has candidate target pathways for researches; (4) it is more friendly to the environment as a protein that is easy to degrade. Currently, there are many studies on the active regions of antifungal peptides, including the contribution of core sites to the overall antifungal peptides reported, for example, the β-sheet motif at the C-terminus of defensins, which enhances their antifungal activity [53]; the α-helical structure of Coprisin and Mastoparan B, which enhances their antimicrobial activity [54,55]; and amphiphilic peptides likewise have their unique sites of action to fight against fungi, such as His(2aryl)-Trp-Arg [56]. It is well known that the structure of a peptide determines its function, and the possible point of action for Mt1 to be an antimicrobial peptide is that it possesses two helix structures (as shown in Figure 5a,d), which are predicted with a high degree of confidence (as shown in Figure 5c). Then we have good reasons to believe that the core site of action of Mt1 is its helix structure, which determines its function. In subsequent research and drug development, these two structures can also be used as separate objects of study to explore whether they possess the antimicrobial activity and ability of antimicrobial peptides, which will be a very exciting research direction.
Our results suggest that FgMt1 may be defective in nutrient acquisition, specifically nitrogen acquisition. Overall, our findings indicate that the recombinant gene inserted in FgMt1 affects its growth and pathogenicity, likely due to nutrient acquisition defects. These findings contribute to our understanding of the molecular mechanisms involved in the growth and pathogenicity of Fg and provide insights into potential targets for future control strategies.

Medium and Growth Conditions
All media used in this paper are included here (see Medium ingredients in Supplementary Materials). Lysogeny broth (LB), carboxymethylcellulose sodium medium (CMC), potato dextrose agar medium (PDA), potato dextrose medium (PDB), potato sucrose agar medium (PSA), Czapek-Dox medium (CDM), minimal medium (MM), complete medium (CM), cornmeal agar medium (CMA), carrot dextrose medium (CMA), and yeast peptone medium (YEPD). The B. subtilis strain was shaken for 12 h in LB medium, and the F. graminearum strains were sporulated by shaking in CMC medium under light at 220 rpm for 5 days and in all the other 7 media for 3 days. The model of shaker used in all the above culture processes is the Ruihua HZ150L constant temperature culture shaker (Wuhan).

Genomic Library Construction
The B. subtilis 330-2 strain was grown in LB medium at 37 • C with shaking at 170 rpm/ for 12 h, followed by centrifugation at 9600× g for 5 min to collect the cells. Genomic DNA was extracted using the CTAB method after crushing and breaking the mycelium with liquid nitrogen. Ultrasonication was employed to fragment the genomic DNA randomly (2 s ultrasonic treatment and 8 s interval, repeated 30 times) instead of the traditional digestion and ligation reaction method [60]. The resulting DNA fragments were ligated to the psxsh-Neo vector using T4 ligase. The recombinant vectors were transformed into E. coli competent cells (TSC 01) and then transferred into Fg protoplasts [61,62] to create a mutant library. The pls1 gene was used as the target site for the integration of the genomic library [59]. To ensure a sufficient number of transformants, the positive ratio and the number of transformants for each E. coli transformation were counted, and the number of transformants was confirmed to exceed 20,000.

Coleoptile Inoculation
Wheat seeds were sterilized and prepared for inoculation as follows: Seeds were rinsed with distilled water and then soaked in 75% ethanol for 30 s, followed by three rinses with distilled water. Next, seeds were disinfected with a 2% sodium hypochlorite solution for 10 min and rinsed thoroughly with sterile water. After that, seeds were soaked in sterile water for 2 h at 20 • C. A tray was prepared by placing two layers of wet filter paper on it, and seeds were placed on the filter paper with the embryo facing downward, maintaining a distance between them. The tray was then incubated in a constant-temperature incubator at 20 • C for 3 days until the coleoptiles started to grow. Subsequently, the coleoptiles were inoculated with a spore solution by cutting off the tips of the coleoptiles by 2-3 mm with sterile scissors and then adding 2 µL of the spore solution (5 × 10 5 spores/mL) to the incision (the control was Fg-5035 and the treated group was FgMt1). After inoculation, the tray was placed in an artificial climate incubator at 25 • C with 12-h light/dark cycles. Symptoms were checked every day, and the tray was moisturized as needed [63].

RNA-Sequencing
Fg-5035 and FgMt1 mycelia were inoculated in PDB medium and incubated for 3 days in the dark at 170 rpm shaking, then the mycelia were filtered, harvested, and frozen in a −80 • C refrigerator for RNA extraction. Total RNA was extracted from F. graminearum (wt/treatment) using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. PolyA-enriched mRNA was isolated from the total RNA using oligo (dT) magnetic beads, and the RNA was fragmented to an average length of 300 bp by ion interruption. After RNA extraction, purification, and library construction, paired-end (PE) sequencing of the libraries was performed using the Illumina sequencing platform. Sequencing services were provided by Personal Biotechnology Company (Shanghai, China). The resulting data were analyzed using the free online platform Personalbio GenesCloud (https://www.genescloud.cn, accessed on 26 May 2022). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted using the databases established by the Gene Ontology Consortium (http://geneontology.org/, accessed on 14 December 2022) and Kyoto Encyclopedia of Genes and Genomes (http: //www.kegg.jp/, accessed on 14 December 2022), respectively.

Statistical Analysis
GraphPad Prism version 8.00 (GraphPad Software, San Diego, CA, USA, http://www. graphpad.com/, accessed on 9 March 2019) was used in the statistical analysis. The mean and standard deviation (SD) were used as descriptive statistics. A t-test was used for normally distributed variables. All experimental statistics were performed in triplicate, and p < 0.05 was considered a statistically significant difference.

Data Availability Statement:
There is no new data available.

Conflicts of Interest:
The authors declare no conflict of interest.