Response of Fusarium pseudograminearum to Biocontrol Agent Bacillus velezensis YB-185 by Phenotypic and Transcriptome Analysis

The use of biological control agents (BCAs) is a promising alternative control measure for Fusarium crown rot (FCR) of wheat caused by Fusarium pseudograminearum. A bacterial strain, YB-185, was isolated from the soil of wheat plants with FCR and identified as Bacillus velezensis. YB-185 exhibited strong inhibition of F. pseudograminearum mycelial growth and conidial germination in culture. Seed treatment with YB-185 in greenhouse and field resulted in reductions in disease by 66.1% and 57.6%, respectively, along with increased grain yield. Microscopy of infected root tissues confirmed that YB-185 reduced root invasion by F. pseudograminearum. RNA-seq of F. pseudograminearum during co-cultivation with B. velezensis YB-185 revealed 5086 differentially expressed genes (DEGs) compared to the control. Down-regulated DEGs included genes for glucan synthesis, fatty acid synthesis, mechanosensitive ion channels, superoxide dismutase, peroxiredoxin, thioredoxin, and plant-cell-wall-degrading enzymes, whereas up-regulated DEGs included genes for chitin synthesis, ergosterol synthesis, glutathione S-transferase, catalase, and ABC transporters. In addition, fungal cell apoptosis increased significantly, as indicated by TUNEL staining, and the scavenging rate of 2,2′-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt radical cation (ABTS·+) in the fungus significantly decreased. Thus, F. pseudograminearum may be trying to maintain normal cell functions by increasing cell wall and membrane synthesis, antioxidant and anti-stress responses, detoxification of bacterial antimicrobial compounds, and transportation of damaging compounds from its cells. However, cell death and free radical accumulation still occurred, indicating that the responses were insufficient to prevent cell damage. Bacillus velezensis YB-185 is a promising BCA against FCR that acts by directly damaging F. pseudograminearum, thus reducing its ability to colonize roots and produce symptoms.


Introduction
Wheat (Triticum aestivum L.) is a major food crop, susceptible to a variety of diseases [1]. Fusarium crown rot (FCR) is one of the most destructive soil-borne wheat diseases in many arid and semi-arid cropping regions of the world [2]. Symptoms are typically a crown rot (http://bacteria.ensembl.org/index.html, accessed on 12 March 2022) and aligned with MAFFT (https://mafft.cbrc.jp/alignment/server/, accessed on 12 March 2022). Perl was used to concatenate the aligned conservative regions of the two genes, and RAxML analysis was run (https://github.com/stamatak/standard-RAxML, accessed on 12 March 2022) to construct a maximum likelihood phylogenetic tree for the combined datasets through the CIPRES web portal (http://www.phylo.org, accessed on 12 March 2022), with 1000 bootstrap iterations.

Antagonistic Activity of Strain YB-185 against F. pseudograminearum
For the conidial germination assay, a 7 mm plug from a 5-day-old F. pseudograminearum colony was inoculated into 50 mL carboxymethyl cellulose (CMC) broth [18] and cultured at 25 • C with shaking at 150 rpm. After 5 days, conidia were harvested by filtering through a layer of Miracloth (Merck Millipore, Billerica, MA, USA) and adjusted to 10 7 conidia/mL using a hemocytometer. Strain YB-185 was grown in NB at 30 • C shaking at 180 rpm. After 10 h, the cells were collected by centrifugation at 8000 rpm for 5 min and resuspended in sterile water. The number of colony forming units (CFU) of YB-185 was determined by dilution plating and was adjusted to 10 8 CFU/mL with sterile water. A 1 mL conidial suspension was transferred into 20 mL PDB containing 1 mL bacterial suspension. PDB containing 1 mL conidial suspension with only sterile water added was used as control. The dual cultures were grown at 25 • C in the dark for 12 h, and the percentage of conidial germination was measured by examining 100 conidia with a microscope (×40 magnification). A conidium was considered to have germinated if the germ tube exceeded one-half the length of the conidium.
To assay the effect of YB-185 culture filtrate on mycelial growth of F. pseudograminearum, YB-185 was grown in 50 mL NB for 48 h at 30 • C, shaking at 180 rpm. The culture was centrifuged at 10,000 rpm for 10 min, and the supernatant passed through a 0.22 um filter. The filtrate was added to PDA at 1:2, 1:5, and 1:10 (v/v). Then, a 5 mm F. pseudograminearum plug was inoculated on the medium and incubated at 25 • C in the dark, with a plug grown on PDA alone as control. At 5 days, colony diameters were measured, and the inhibition percent of mycelial growth was calculated [19]. The morphology of F. pseudograminearum grown on PDA with 1:5 YB-185 filtrate was observed by SEM and transmission electron microscopy (TEM). For SEM and TEM, mycelia from the margin of an F. pseudograminearum colony on PDA containing 1:5 YB-185 filtrate were fixed in 2% glutaraldehyde for 4 h at 4 • C and rinsed three times with 0.1 M sodium phosphate buffer (pH 7.4). The samples were then fixed with 1% osmic acid for 2 h at 25 • C and dehydrated through an ethanol gradient. For SEM, the samples were then directly examined with a Hitachi SU8100 SEM microscope (Hitachi, Tokyo, Japan) using an acceleration voltage of 3.0 kV. For TEM, the samples were embedded in Epon 812 (Nisshin EM, Tokyo, Japan) and examined with a Hitachi HT7800 TEM microscope (Hitachi, Tokyo, Japan) using an acceleration voltage of 80 kV.
To observe apoptosis of F. pseudograminearum, mycelium was collected from the PDA with 1:5 YB-185 filtrate and rinsed twice with PBS. The mycelium was then stained according to the protocol of the one-step TUNEL Apoptosis Assay Kit (Beyotime, Shanghai, China). Stained mycelium was examined with an A2 Axio microscope (Carl Zeiss, Oberkochen, Germany) with fluorescence detected at 450-490 nm. In addition, nuclear morphology in fungal mycelium was observed by staining with 1 µg mL −1 4-6-diamidino-2-phenylindole (DAPI) as per Domachowske et al. [20].

YB-185 Suppression of FCR in Greenhouse
For preparation of F. pseudograminearum inoculum, millet seeds (cultivar: Yugu 31) were sterilized at 121 • C for 30 min in flasks and then inoculated with four 7-day-old fungal mycelial plugs (5 mm in diameter) The flasks were incubated at 25 • C for 10 days, during which the flasks were shaken once a day. Soil (sandy loam) collected from the field in Jiaozuo (Henan, China) was used for the greenhouse experiment, with the content of total N 221.6 mg/kg, available P 16.8 mg/kg, and available K 137.5 mg/kg. Then, the fungal inoculum was mixed into sterile field soil at 0.5 % (W/W). Seeds were surface-sterilized with 1% NaClO and washed with sterile water three times, and then soaked in different concentrations of YB-185 (10 6 to 10 9 CFU/mL) for 2 h. For control plants, seeds were soaked in sterile water. The seeds were dried overnight and sown the next day. Four seeds were planted into 200 g of F. pseudograminearum-inoculated soil in 250 mL plastic pots and arranged in a completely randomized design in a greenhouse maintained at 28 • C with a 16:8 h L:D photoperiod supplied by LED light and 80% RH. Each treatment had ten replicates, and the test was conducted twice.
At 12 days post planting (dpp), segments of the root elongation zone treated or not treated with 10 9 CFU/mL YB-185 were embedded in paraffin and cut into thin slices in cross and longitudinal sections. The samples were then co-stained with wheat-germ agglutinin-Alexa Fluor 488 conjugate (WGA-AF488) to observe hyphae of F. pseudograminearum and propidium iodide (PI) to observe plant cell walls [21]. The images were scanned and digitized as previous described [22].
At 35 dpp, plants from six pots were removed from the soil, and disease severity was graded on the washed roots using a 0-to-7 scale according to Smiley et al. [23]. Disease index and control efficiency was calculated using the formulas: Disease index = 100 × ∑ (grade × the number of infected plants)/(highest grade × the total number of investigated plants); Control efficiency = (disease index of control − disease index of treatment)/disease index of control × 100%.

YB-185 Suppression of FCR in the Field
In 2019, wheat seeds of cultivar Zhengmai 366 were soaked in 10 9 CFU/mL YB-185 cell suspensions as described above. The seeds soaked in sterile water for 2 h were used as a non-treated control. Wheat seeds coated with 2 mL/kg difenoconazole-fludioxonil (4.8%; Syngenta, Beijing, China) were applied as a fungicide treatment control. Seeds were planted October 19 in field plots in Jiaozuo (Henan, China) that had been fertilized with 225 kg/ha carbamide and 120 kg/ha diammonium phosphate 2 days before sowing. The plots were 1.5 × 12 m and arranged in randomized complete block design with three replicates each of YB-185 treatment, fungicide treatment, and non-treated control. Plots were weeded by hand and irrigated as needed. In the pustulation growth stage of wheat, disease severity was rated according to the crown and lower stem tissues of plants on a 0-to-10 scale, where 0 = no discoloration and 10 = severe disease [24]. The control efficiency against FCR was calculated as above. In addition, grain yield was recorded at harvest time.

Transcriptome of Fusarium pseudograminearum Co-Cultured with YB-185
First, 2 mL of conidia of F. pseudograminearum from CMC broth (10 7 conidia/mL) was inoculated into 50 mL PDB and incubated at 25 • C shaking at 160 rpm. At 20 h, 5 mL of YB-185 (10 8 CFU/mL) was added and incubated at 30 • C, shaking at 160 rpm. The control was the addition of 5 mL sterile water. After 4 h and 16 h, the mycelium was filtered through two layers of sterile Miracloth (Merck Millipore, Billerica, MA, USA) and washed thoroughly with cold distilled water. Then, the samples were frozen in liquid nitrogen and stored at −80 • C. The experiment was repeated three times for each treatment. RNA extraction was done using a RNeasy Mini kit (Qiagen, Hilden, Germany), and the RNA was sent to Shanghai Meiji Biomedical Technology Company (Shanghai, China) for sequencing using an Illumina HiSeq platform.
Expression levels were calculated from fragments per kb of transcript per million reads (FPKM). Differentially expressed genes (DEGs) were identified with the DESeq2 R package (1.10.1) with |log 2 (fold change)| > 1 using a false discovery rate (FDR < 0.01) and a high statistically significant value (p < 0.05). DEGs were divided into functional categories and defined pathways. GO enrichment analysis was implemented by the GOseq R packages [27] with a corrected p-value < 0.05 as the threshold. KEGG pathway analysis was performed in the KEGG database. Global metabolic pathways were displayed using iPath 2.0 (http://pathways.embl.de, accessed on 1 April 2022).

RT-PCR of Fusarium pseudograminearum Genes in Co-Cultures with YB-185
To verify the reliability of the transcriptome data of F. pseudograminearum, 30 DEGs were selected based on low q-value, large fold difference, and annotation (Table S1). Tubulin was used as the internal reference gene (Table S2). cDNAs were synthesized from the sequencing RNAs using a PrimeScript™ RT Reagent Kit with gDNA Eraser (TaKaRa, Dalian, China), and primers were designed using Primer 5.0 (Table S2). Reverse transcription quantitative PCR (RT-qPCR) was conducted using a Step One Plus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with SYBR Green MasterMix (Applied Biosystems, Foster City, CA, USA). Relative expression was calculated using the 2 −∆∆CT method [28] in triplicate with three biological replicates.

Total Antioxidant and Glutathione-S-Transferase Activity
Fusarium pseudograminearum mycelia were prepared as per transcriptome sequencing described previously, and samples were collected at 4 h and 16 h. For antioxidant levels, a crude enzyme solution was prepared from the mycelia as per Han et al. [29]. The scavenging activity of 2,2 -Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt radical cation (ABTS·+) was determined with a Total Antioxidant Capacity (T-AOC) Assay Kit (Sangon Biotech, Shanghai, China), and activity was calculated using the equation of Han et al. [29]. Glutathione-S-transferase (GST) activity was determined from a crude enzyme solution prepared from the mycelia as per Wang et al. [30]. GST activity was determined with a Glutathione S-transferase Activity Assay Kit (Solarbio, Beijing, China).

Statistical Analysis
Data were analyzed using SPSS Statistics 26.0 (IBM, Armonk, NY, USA). After ANOVA assumptions were evaluated using the Kolmogorov-Smirnov test for normality and Levene's test for homogeneity between groups, one-way ANOVA was performed (p < 0.05).

Strain YB-185 Isolation and Identification
A total of 102 bacteria were isolated from rhizosphere soil of wheat with FCR. All of them were used for growth inhibition of F. pseudograminearum by the dual-culture method, and the results showed that 45 bacterial isolates could inhibit growth of F. pseudograminearum isolate WZ-8A on PDA. The greatest antagonistic activity was with strain YB-185 ( Figure 1A,B). Growth inhibition of F. pseudograminearum with YB-185 was 69.1%. them were used for growth inhibition of F. pseudograminearum by the dual-culture method, and the results showed that 45 bacterial isolates could inhibit growth of F. pseudograminearum isolate WZ-8A on PDA. The greatest antagonistic activity was with strain YB-185 ( Figure 1A,B). Growth inhibition of F. pseudograminearum with YB-185 was 69.1%. Colonies of YB-185 on NA were rounded, ivory white, and opaque, with irregular edges (Figure 2A). Under SEM, the cells were rod-shaped, with an average size of 3.34 μm × 0.79 μm ( Figure 2B). The cells were Gram-positive, and ellipsoidal endospores were observed in the colonies ( Figure 2C). The physiological parameters determined with the Biolog system showed that YB-185 was Bacillus velezensis with a probability of 0.936, which was consistent with the appearance of the colonies, Gram stain, and cell morphology. A phylogenetic tree based on combined 16S rRNA and gyrA sequences showed that YB-185 was most closely related to B. velezensis ( Figure 2D). Based on this and the abovementioned morphological and Biolog features, YB-185 was classified as a strain of B. velezensis. Colonies of YB-185 on NA were rounded, ivory white, and opaque, with irregular edges (Figure 2A). Under SEM, the cells were rod-shaped, with an average size of 3.34 µm × 0.79 µm ( Figure 2B). The cells were Gram-positive, and ellipsoidal endospores were observed in the colonies ( Figure 2C). The physiological parameters determined with the Biolog system showed that YB-185 was Bacillus velezensis with a probability of 0.936, which was consistent with the appearance of the colonies, Gram stain, and cell morphology. A phylogenetic tree based on combined 16S rRNA and gyrA sequences showed that YB-185 was most closely related to B. velezensis ( Figure 2D). Based on this and the above-mentioned morphological and Biolog features, YB-185 was classified as a strain of B. velezensis.
The morphology and ultrastructure of F. pseudograminearum mycelia without YB-185 filtrate showed normal growth with a smooth surface ( Figure 3A), while the mycelia grown with YB-185 filtrate was swollen and irregular ( Figure 3B). TEM showed that the F. pseudograminearum hyphae without YB-185 filtrate had a smooth surface, organized cell wall, complete plasma membrane, and uniformly distributed cytoplasm and organelles ( Figure 3C,E). However, hyphae grown with YB-185 filtrate had an irregular surface, degraded cell wall, broken plasma membrane, cytoplasm with discontinuously packed fibers and empty areas, and sparse and unevenly distributed organelles ( Figure 3D,F).
Using TUNEL staining, mycelia of F. pseudograminearum cells with YB-185 filtrate showed strong fluorescence, indicating apoptosis, mainly concentrated in the enlarged deformity at the tip of the mycelium ( Figure 1J). However, the mycelia of the control showed only slight fluorescence ( Figure 1I). In addition, mycelia treated with YB-185
The morphology and ultrastructure of F. pseudograminearum mycelia without YB-185 filtrate showed normal growth with a smooth surface ( Figure 3A), while the mycelia grown with YB-185 filtrate was swollen and irregular ( Figure 3B). TEM showed that the F. pseudograminearum hyphae without YB-185 filtrate had a smooth surface, organized cell wall, complete plasma membrane, and uniformly distributed cytoplasm and organelles ( Figure 3C,E). However, hyphae grown with YB-185 filtrate had an irregular surface, degraded cell wall, broken plasma membrane, cytoplasm with discontinuously packed fibers and empty areas, and sparse and unevenly distributed organelles ( Figure 3D,F). filtrate had cells exhibiting nuclear fragmentation based on DAP whereas the control mycelia did not show nuclear fragmentation (

YB-185 Suppression of FCR in Greenhouse
Seed treatment with 10 6 to 10 9 CFU/mL of YB-185 in the gr increasing concentration of YB-185 resulted in greater reduction in the disease index (control efficiency) ranging from 13.7 (23.9%) (66.1%) for 10 9 CFU/mL of YB-185 ( Figure 4A). Using TUNEL staining, mycelia of F. pseudograminearum cells with YB-185 filtrate showed strong fluorescence, indicating apoptosis, mainly concentrated in the enlarged deformity at the tip of the mycelium ( Figure 1J). However, the mycelia of the control showed only slight fluorescence ( Figure 1I). In addition, mycelia treated with YB-185 filtrate had cells exhibiting nuclear fragmentation based on DAPI staining ( Figure 1L), whereas the control mycelia did not show nuclear fragmentation ( Figure 1K).

YB-185 Suppression of FCR in Greenhouse
Seed treatment with 10 6 to 10 9 CFU/mL of YB-185 in the greenhouse showed increasing concentration of YB-185 resulted in greater reduction in FCR ( Figure S1), the disease index (control efficiency) ranging from 13.7 (23.9%) for 10 6 CFU/mL t (66.1%) for 10 9 CFU/mL of YB-185 ( Figure 4A).  With F. pseudograminearum alone, hyphae were visible in the root cortex, endodermis, pericycle regions, and xylem vessels, with cells in the cortex appearing highly degraded and the endodermis cells shrunken ( Figure 5A,C). However, roots infected with F. pseudograminearum and treated with 10 9 CFU/mL of YB-185 showed only a limited amount of hyphae restricted to the endodermis and pericycle, with the root tissues appearing relatively intact ( Figure 5B,D).
With F. pseudograminearum alone, hyphae were visible in the root cortex, endodermis, pericycle regions, and xylem vessels, with cells in the cortex appearing highly degraded and the endodermis cells shrunken ( Figure 5A,C). However, roots infected with F. pseudograminearum and treated with 10 9 CFU/mL of YB-185 showed only a limited amount of hyphae restricted to the endodermis and pericycle, with the root tissues appearing relatively intact ( Figure 5B,D).

YB-185 suppression of FCR in the field.
Non-treated control fields had an FCR disease index of 17.5, which was significantly higher than the 8.5 found with seed treatment of 10 9 CFU/mL YB-185 (p < 0.05) ( Figure  4B). The control efficiency of FCR by YB-185 in the field at 52.0% was not significantly different from 57.6% with seed treatment of 2 mL/kg 4.8% difenoconazole fludioxonil (p < 0.05). Grain yield with 10 9 CFU/mL YB-185 was 9040.7 kg/hm 2 , exceeding 8475.6 kg/hm 2 for the non-treated control. The yield with 10 9 CFU/mL YB-185 increased by 6.7%.

Transcriptome of F. pseudograminearum co-cultured with B. velezensis YB-185
The mycelia of F. pseudograminearum co-cultured with YB-185 in PDB began to appear swollen and irregular at 4 h and appeared cracked and melted at 16 h. However, the mycelia of the control showed normal growth with a smooth surface at both time points

YB-185 Suppression of FCR in the Field
Non-treated control fields had an FCR disease index of 17.5, which was significantly higher than the 8.5 found with seed treatment of 10 9 CFU/mL YB-185 (p < 0.05) ( Figure 4B). The control efficiency of FCR by YB-185 in the field at 52.0% was not significantly different from 57.6% with seed treatment of 2 mL/kg 4.8% difenoconazole fludioxonil (p < 0.05). Grain yield with 10 9 CFU/mL YB-185 was 9040.7 kg/hm 2 , exceeding 8475.6 kg/hm 2 for the non-treated control. The yield with 10 9 CFU/mL YB-185 increased by 6.7%.

Transcriptome of F. pseudograminearum Co-Cultured with B. velezensis YB-185
The mycelia of F. pseudograminearum co-cultured with YB-185 in PDB began to appear swollen and irregular at 4 h and appeared cracked and melted at 16 h. However, the mycelia of the control showed normal growth with a smooth surface at both time points ( Figure S2). Sequencing RNA from F. pseudograminearum alone or co-cultured with YB-185 at 4 h and 16 h in PDB resulted in a total of 88.7 Gb reads after cleaning and quality check.
The Q20 percentage of each library was 98.5% to 98.7%, and the Q30 percentage was 95.6% to 96.0% (Table S3). A mean of 86.8% clean reads was mapped to the F. pseudograminearum genome database (Table S4). The heatmap of Person's correlation coefficients showed that the values reached more than 0.989 between repeats of each treatment ( Figure S3). PCA analysis mapping also showed that the repeats of each treatment tended to aggregate together. These results showed that the replications of the transcriptomes of each treatment had a high consistency and reliability ( Figure S4). Based on the FPKM mapped read and FPKM density distributions, there were differences in the dispersion and population distribution of gene expression between the treatment and control at both 4 h and 16 h ( Figure S5). There were 5086 DEGs for F. pseudograminearum with compared to without YB-185 (|log2(fold change)| > 1, p < 0.  All the DEGs were categorized into 45 GO terms with regard to biological process, cellular component, and molecular function (Table S5). Metabolic process (GO:0008152) was the most enriched term among biological processes, followed by cellular process (GO:0009987) and single-organism process (GO:0044699) ( Figure S6). The most enriched terms among cellular components were membrane (GO:0016020), membrane part (GO:0044425), and cell (GO:0005623). The most enriched term among molecular functions were for catalytic activity (GO:0003824) and binding (GO:0005488).
The 2024 DEGs were also classified by IPATH pathway analysis. The majority of the enriched pathways were for lipid metabolism, energy metabolism, amino acid metabolism, carbohydrate metabolism, glycan biosynthesis and metabolism, metabolism of cofactors, and vitamins (Figure 7).
For fungal secondary metabolite biosynthesis, DEGs for polyketide synthetases PKS2 were down-regulated at 4 h and up-regulated at 16 h, while PKS6, PKS7, and PKS10 were up-regulated at 16 h. (Table 1). However, DEGs for PKS12 were down-regulated at both time points. Nonribosomal peptide synthetases NPS2 and NPS6 were significantly down-regulated at 16 h.

RT-PCR of Fusarium pseudograminearum Genes in Co-Cultures with YB-185
To confirm the DEG expression profiles, 36 DEGs were analyzed by qRT-PCR ( Figure S8). Overall, qRT-PCR results matched well with those from RNA-seq.

Total Antioxidant and GST Activity
Cultivation of F. pseudograminearum with B. velezensis resulted in significantly lower free radical scavenging activity as indicated by ABTS·+ clearance rates compared to the control at both 4 and 16 h ( Figure 8A). In contrast, co-cultivation resulted in significantly higher GST activities at 4 h and 16 h compared to the control ( Figure 8B).

Total Antioxidant and GST Activity
Cultivation of F. pseudograminearum with B. velezensis resulted in significantly lower free radical scavenging activity as indicated by ABTS·+ clearance rates compared to the control at both 4 and 16 h ( Figure 8A). In contrast, co-cultivation resulted in significantly higher GST activities at 4 h and 16 h compared to the control ( Figure 8B).

Discussion
Among reports of Bacillus species being used as BCAs of FCR of wheat caused by F. pseudograminearum, there have been Bacillus halotolerans [31], B. subtilis [12], and B. velezensis [32], as well as B. velezensis as a BCA of FCR of sorghum caused by F. pseudograminearum [33]. In this study, a new wheat soil bacterial strain, YB-185, was identified as B. velezensis and shown to be a promising BCA for FCR of wheat caused by F. pseudograminearum.
Seed treatment with YB-185 reduced FCR under both greenhouse and field conditions, with 10 9 CFU/mL able to reduce disease indices by 66.1% and 52.0% in the greenhouse and field, respectively. The level of control in the field was comparable to fungicide seed treatment. Using F. pseudograminearum coleoptile infection, B. halotolerans QTH8 application to wheat seedlings decreased FCR in wheat by 62.4% [31]. Using F. pseudograminearum inoculated soil, B. subtilis YB-15 wheat seed treatment reduced the FCR by 81.5% [12], priming of wheat seeds with B. velezensis UTB96 reduced FCR by 65.5% [32], and priming of sorghum plants with B. velezensis NB54 resulted in reduced disease severity by 47.6% under drought stress and 55.6% without drought stress [33]. In general,

Discussion
Among reports of Bacillus species being used as BCAs of FCR of wheat caused by F. pseudograminearum, there have been Bacillus halotolerans [31], B. subtilis [12], and B. velezensis [32], as well as B. velezensis as a BCA of FCR of sorghum caused by F. pseudograminearum [33]. In this study, a new wheat soil bacterial strain, YB-185, was identified as B. velezensis and shown to be a promising BCA for FCR of wheat caused by F. pseudograminearum.
Seed treatment with YB-185 reduced FCR under both greenhouse and field conditions, with 10 9 CFU/mL able to reduce disease indices by 66.1% and 52.0% in the greenhouse and field, respectively. The level of control in the field was comparable to fungicide seed treatment. Using F. pseudograminearum coleoptile infection, B. halotolerans QTH8 application to wheat seedlings decreased FCR in wheat by 62.4% [31]. Using F. pseudograminearum inoculated soil, B. subtilis YB-15 wheat seed treatment reduced the FCR by 81.5% [12], priming of wheat seeds with B. velezensis UTB96 reduced FCR by 65.5% [32], and priming of sorghum plants with B. velezensis NB54 resulted in reduced disease severity by 47.6% under drought stress and 55.6% without drought stress [33]. In general, B. velezensis YB-185 provided similar levels of control of FCR compared to those other Bacillus strains, but all those studies were limited to greenhouse conditions, whereas this study included a oneyear field study to verify the potential of YB-185 as a BCA under more realistic conditions. This is important as many BCAs that perform well under controlled conditions fail to perform similarly in the field [34]. The seed treatment of 10 9 CFU/mL YB-185 significantly decreased the disease index by 52.0% compared to control, and subsequently retrieved yield lost by 6.7%. To prevent the occurrence of FCR and reduce the losses of the yield to a great extent, the application methods of the biocontrol strain should be optimized in future work, and the biocontrol agent should be applied once again at the returning green stage. The use of B. velezensis YB-185 for the control of FCR under practical cultivation conditions may represent an efficient alternative to fungicides for sustainable wheat cultivation.
At least one mode of action of YB-185 against F. pseudograminearum was direct antimicrobial activity, as evidenced by growth inhibition in culture. This may be related to secreted compounds from YB-185, as culture filtrates were able to reduce spore germination and cause swollen and malformed hyphae and conidial germ tubes. The cytoplasm of fungal cells exposed to YB-185 culture filtrate contained empty and fiber-filled areas, and the organelles appeared sparse and unevenly distributed. A likely candidate for such activity would be antimicrobial peptides, which are well known in Bacillus species, especially lipopeptides, which have been involved in its biocontrol effect against many fungi [35].
Among lipopeptide studies, Liao et al. [36] reported swollen and cracked hyphae of Pyricularia oryzae exposed to fengycin from B. amyloliquefaciens BPD1, Gong et al. [37] reported that the cytoplasm of F. graminearum hyphae was disorganized and sparse after exposure to iturin A or plipastatin A from B. amyloliquefaciens S76-3, and Toral et al. [38] reported that the organelles of Botrytis cinerea hyphae were degenerated and gathered in clumps after exposure to a mixture of lipopeptides from B. methylotrophicus XT1. Thus, the effects of culture filtrate of YB-185 on F. pseudograminearum hyphae were similar to those previously reported for fungi exposed to different Bacillus lipopeptides.
In this study, the transcriptome response of F. pseudograminearum to YB-185 culture filtrate showed changes related to fungal cell wall and membrane synthesis, response to oxidative stress, cell death related to apoptosis, production of secondary metabolites, and factors potentially related to virulence to plants. Thus far, there are several transcriptome studies of plant pathogenic fungi incubated with Bacillus species or their crude culture filtrates. For example, RNA was analyzed from Colletotrichum gloeosporioides TS-09 grown on PDA with B. amyloliquefaciens SDF-005 for 9 days [39], Verticillium dahliae VdSHZ-9 grown on PDA with Bacillus N-4 for 7 days [40], and Fusarium oxysporum grown on PDA with B. subtilis HSY21 for 2 and 3 days [29]. RNA was also analyzed for Sclerotinia sclerotiorum 1980 grown on PDA containing 10% culture filtrate of B. amyloliquefaciens Bam22 for 1 day [41], and Botrytis cinerea strain B05.10 grown in PDB with B. subtilis MBI 600 culture filtrate for 0, 24, 48, and 72 h [42].
One effect of the YB-185 on F. pseudograminearum was on fungal cell wall synthesis genes. There was down-regulation of one 1,3-beta-glucan synthase and three 1,3-betaglucanosyltransferase genes at 4 and 16 h. Note that 1,3-Beta-glucan synthase is a glucosyltransferase that generates beta-glucan, a major component of fungal cell walls, and 1,3-beta-glucanosyltransferase is involved in the elongation of 1,3-beta-glucan [43]. For both S. sclerotiorum and V. dahilae with Bacillus, 1,3-beta-glucan synthase DEGs were upregulated [40,41]. It was concluded that this was an attempt to alleviate cell wall damage. In this study, there was an up-regulation of two chitin synthase genes only at 4 h, which could be an early attempt to repair cell wall injuries. Chitin synthase is a glycosyltransferase that catalyzes formation of chitin, another major component of fungal cell walls [44]. Tian et al. [40] reported that DEGs related to chitin synthase in V. dahilae were up-regulated by Bacillus. Overall, there may be more limited new cell wall synthesis of F. pseudograminearum with Bacillus stress limited to earlier in the interaction.
YB-185 also altered the expression of genes related to fungal cell membranes. There was up-regulation of two ergosterol synthase DEGs at both time points. Ergosterol is the major component of fungal cell membranes [45]. Some fungicides, such as terbinafine and naftifine, target ergosterol synthesis enzymes, resulting in fungal lysis [45]. Similarly, there was up-regulation with B. amyloliquefaciens stress for two ergosterol synthesis pathway-related genes in S. sclerotiorum [41] and C. gloeosporioides [39]. It was suggested that those changes promoted membrane fluidity, reducing the effects of antimicrobial substances. Two fatty acid synthases and two fatty acid elongase DEGs of F. pseudograminearum were down-regulated at both time points with YB-185. Fatty acids with phospholipids are major components of membranes and maintain cell membrane fluidity [46]. In contrast, two fatty acid synthesis pathway-related genes were up-regulated in S. sclerotiorum with B. amyloliquefaciens [41]. It was proposed that this was part of an attempt to reduce lipopeptide membrane damage, but this did not occur with F. pseudograminearum. Expression of a mechanosensitive ion channel membrane protein DEG was down-regulated at both time points. Mechanosensitive ion channels provide protection against hypoosmotic shock, responding to membrane tension, and reduced levels of them could result in membrane fragility and loosening [47]. However, other transcriptome studies of plant pathogenic fungi with Bacillus did not report DEGs related to mechanosensitive ion channels [29,[39][40][41][42].
Another impact of YB-185 was on the antioxidative stress response of F. pseudograminearum. Two peroxiredoxin DEGs were down-regulated at 4 h, and one SOD, three peroxiredoxin, and two thioredoxin DEGs were down-regulated at 16 h. SODs convert the damaging free radical superoxide anion into oxygen and hydrogen peroxide [48]. Peroxiredoxins are cysteine-dependent peroxidases that limit peroxide levels within cells [49], and thioredoxins reduce oxidized cysteine residues cleaving disulfide bonds, protecting proteins from oxidative aggregation and inactivation [50]. With Bacillus or culture filtrate, however, S. sclerotiorum up-regulated peroxidase and catalase DEGs [41], and V. dahilae upregulated SOD, catalase, peroxiredoxin, and thioredoxin DEGs [40]. This was proposed to be part of an ROS stress response. Evidence for an increased antioxidative stress response of F. pseudograminearum was up-regulation of one catalase, six GST, and four ABC transporter DEGs at both time points. Catalase breaks hydrogen peroxide into oxygen and water [51], GST can conjugate the reduced form of glutathione to xenobiotics for detoxification [52], and ABC transporter can transfer substrates across membranes to remove toxins from cells [53]. Similarly, C. gloeosporioides, Setosphaeria turcica, and B. cinerea DEGs for GSTs and/or ABC transporters were up-regulated with Bacillus stress [39,42,54]. However, a catalase DEG of F. oxysporum was down-regulated with Bacillus stress [29]. Thus, not all elements of fungal antioxidative stress response are up-regulated with Bacillus stress. Despite increased expression of F. pseudograminearum catalases and GSTs, the scavenging activity of ABTS·+ indicated that total antioxidant capacity was lowered, indicating YB-185-induced oxidative stress to F. pseudograminearum.
Among F. pseudograminearum DEGs related to secondary metabolites altered by YB-185 were two PKS DEGs down-regulated at 4 and/or 16 h, and four PKS DEGs up-regulated at 16 h. Fungal PKSs synthesize polyketides, a type of lipid, that have a wide range of functions [56]. By comparison to F. graminearum genome, the down-regulated PKSs of F. pseudograminearum were involved in mycelial growth and the red mycelial pigment aurofusarin [57,58], and the up-regulated PKSs were related to synthesis of the mycotoxins fusaristatin A and fusarin C, a regulator of perithecial maturation, mycelial growth, and aurofusarin [57][58][59][60][61]. This suggests that F. pseudograminearum is shifting the types of polyketides produced during the Bacillus stress response. Two NPS DEGs were downregulated at 4 and/or 16 h. NPSs produce non-ribosomal peptides that can act as fungal antibiotics, toxins, and siderophores [62]. The down-regulated NPS DEGs in this study are similar to those for synthesis of the iron-binding siderophores ferricrocin and fusarinine C/triacetylfusarinine C [63,64]. Those types of siderophores are virulence determinants of many ascomycetes, including Fusarium [65]. The down-regulation of secondary metabolites in F. pseudograminearum could limit its ability to invade wheat roots, contributing to the lower FCR severity with YB-185 treatment. However, other transcriptome studies of plant pathogenic fungi with Bacillus stress did not report DEGs related to PKS and NPS [29,[39][40][41][42].
Another impact of YB-185 culture filtrate was on DEGs for cell wall-degrading enzyme (CWDE) synthesis of F. pseudograminearum. DEGs for four cellulases, six lipases, three amylases, one endo-beta-1,4-glucanase, one endoglucanase, two endo-1,4-beta-xylanases, and two laccases were down-regulated at 16 h. Similarly, DEGs for CWDE synthesis for amylase, glucosidase, xylanase, and cellulase of F. oxysporum were significantly downregulated under the stress of B. subtilis [29]. It was proposed that this resulted in the reduction of virulence to the host plant. A reduced ability to degrade host tissues could explain the microscopy observations of F. pseudograminearum-infected root tissues remaining more intact, with hyphae being limited to the endodermis and pericycle with YB-185 treatment. However, one CWDE synthesis DEG for pectinesterase was up-regulated. Pectinesterase acts in plant cell wall modification and is a virulence factor of Botrytis cinerea [66].
Expression of two DEGs for trichothecene 3-O-acetyltransferase and two DEGs for trichothecene efflux pump were up-regulated with YB-185 culture filtrate. Trichothecenes are six-membered ring compounds with an epoxid or tricyclic ether, acting as protein synthesis inhibitors with toxicity to eukaryotes [67]. Trichothecenes are also virulence factors for spread invasion of Fusarium on spikelets [68]. However, a trichothecene 3-Oacetyltransferase can act in self-protection by converting highly toxic trichothecenes to less toxic compounds [69], and a trichothecene efflux pump can do the same by exporting trichothecene [70]. Thus, this may primarily be a self-protection mechanism in the stress response to Bacillus.
In summary, the soil bacterium, B. velezensis YB-185 was a strong inhibitor of F. pseudograminearum growth both in vitro and in vivo resulting in significant control of FCR under both controlled and field conditions. An examination of F. pseudograminearum exposed to B. velezensis or its culture filtrate showed damage to cell walls and membranes, reduced antioxidant defenses, and apoptosis. Transcriptome analysis showed that the fungus can respond and attempt to alleviate damage caused by B. velezensis. These include trying to increase cell wall and membrane synthesis, antioxidant responses, detoxification, and export of xenobiotics from its cells ( Figure 9). However, these were insufficient to prevent cell damage, and its ability to cause FCR was greatly compromised with YB-185. Supplementary Materials: The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Figure S1: Images of of wheat crown rot in the greenhouse with seed not treated (control) or treated with 10 6 to 10 9 CFU/mL of YB-185; Figure     Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jof8080763/s1, Figure S1: Images of of wheat crown rot in the greenhouse with seed not treated (control) or treated with 10 6 to 10 9 CFU/mL of YB-185; Figure Table S1: The gene information verified by RT-PCR; Table S2: The gene-specific primers used for RT-RCR; Table S3: Summary of RNA-seq data sets; Table S4: Results of clean reads mapping to