A Transcriptome Prole Reveals The Regulatory Mechanism of Verticillium Dahliae Against Bacillus

Background: Verticillium dahliae, the causal agent of Verticillium wilt, is notoriously invasive in many crops and has been involved in numerous epidemics worldwide. Bacillus species, as representatives of biocontrol bacteria, produce a variety of lipopeptides (LPs), which are useful as biofungicides to many pathogenic fungi, including Verticillium dahliae. This study will explore the mechanism of resistance of V. dahliae to Bacillus and biocontrol bacteria. Results: By using in vitro confrontation bioassays, we found that under the stress induced by Bacillus, the spore vitality of V. dahliae with larger colonies was higher, and more abundant microsclerotia were formed. Then, according to the RNA-Seq analysis, the target of rapamycin (TOR) and mitophagy pathways were enriched among the signicantly upregulated 542 genes observed in two co-culture groups with different colony sizes. In addition, in the group of V. dahliae with large colonies, the pathways related to cell wall synthesis, microsclerotia formation and the clearance of reactive oxygen species were regulated, and the expression of genes was up-regulated. Conclusion: This study found that the larger colonies of V. dahliae were more resistant to the antagonistic actions of Bacillus and the likelihood of the formation of homeostasis. Therefore, the prevention of Verticillium wilt by Bacillus is more effective than the treatment of an active fungal infection. These transcriptomic insights provide direction for the use of fungicides in the prevention and treatment of diseases such as Verticillium wilt. This study showed that larger colonies of V. dahliae were less inhibited by Bacillus. When faced with biocontrol bacteria, larger colonies can regulate the TOR pathway, cell wall synthesis, microsclerotia formation, and scavenge ROS, as to preserve homeostasis in vivo and more effectively participate in niche competition. This study was based on the antagonistic effect of Bacillus on V. dahliae and revealed the self-regulatory mechanism of V. dahliae after inhibition by transcriptome analysis. This can explain why treatment using microbial agents can sometimes be ineffective and suggests that the prevention of Verticillium wilt by Bacillus is more effective than the treatment.

Bacillus species are already used in several commercial biofungicide formulations, such as Serenade® and Kodiak® [15]. They have a long history of study and use for the biological control of plant pathogens [8]. Certainly, biofungicides are not always effective and have proven to be more e cient when applied in the early stages rather than post-infection [16,17]. Peng [18] previously reported that a biofungicide produced by Bacillus is not an effective approach to control clubroot on canola. In addition, we have isolated endophytic Bacillus N-4 from Sophora alopecuroides plants. After strain screening and testing of volatile antibacterial substances, we have proved that it has a strong inhibitory effect and biocontrol ability against Verticillium dahliae.
Fungi respond to environmental changes and when faced with adverse conditions, survive via the activation of resistance pathways. Based on the yeast model, the target of rapamycin (TOR) pathway in V. dahliae has been widely studied. The TOR pathway is an evolutionarily conserved phosphoinositide-3 kinase-related kinase (PIKK)-dependent cascade that controls multiple cellular processes in response to various intracellular and extracellular signals [19,20]. Plants, animals (especially cancer cells), and fungal cells utilize the TOR pathway to thrive under stress [21][22][23]. The TOR pathway contains two distinct protein complexes: TOR Complexes 1 (TORC1) and 2 (TORC2). TORC2 can phosphorylate PKC1 and participates in the maintenance of cell wall integrity [24]. A primary determining factor of fungal stress resistance is the cell wall. The inhibition of formation of fungal cell walls has been widely studied.
The initiation of TORC1 involves many factors, such as growth factors, oxygen, and amino acid levels [25]. TORC2 primarily functions as an effector of insulin/PI3K signaling [26] and an activator associated with the ribosome and growth factors [27].
Mitochondria, as the principal source of production of cellular energy, respond to cellular oxidation levels.
They possess an antioxidant defense system that maintains the redox balance [28]. Cells utilize a selective vacuolar degradation process in mitochondria or mitophagy via autophagy-related proteins to degrade and recycle damaged and super uous mitochondria. It is also possible that mitophagy plays a role in the adaptation to environmental conditions by altering the quality and number of mitochondria [29,30].
The redox signaling pathway is often utilized to eliminate reactive oxygen species (ROS) induced by UV irradiation, high temperatures, chemical herbicides, and other environmental stressors [31,32]. The production of ROS is partially responsible for the induction of cell death by Bacillus LPs [33], but the fungi can eliminate large amounts of ROS to survive. The levels of expression of genes associated with the oxidative response and the production of superoxide ions (NADPH oxidase) were tested in response to the high concentrations of ROS induced by iturin [34]. The three main systems that play vital roles in removing ROS in cells are 1) Cu-and Zn-superoxide dismutases (SODs), catalase (CAT) or glutathione systems, 2) the peroxiredoxin system, and 3) the thioredoxin system [32].
Moreover, fungal melanin protects both the fungus from fatal radiation and the cell wall from enzymatic activity and can also bind heavy metals, thus, protecting the fungus [35][36][37]. The VMK1 gene is a mitogen-activated protein kinase gene, which affects the pathogenicity of V. dahliae by affecting the formation of microsclerotia [38,39]. Numerous fungi form structures termed sclerotia that are disseminated in infected soil and survive for long periods under stressful environmental conditions. Microsclerotia are required for virulence in V. dahliae, and strains that are defective in microsclerotia do not colonize plant roots e ciently or cause disease symptoms [40]. The fungus produces melanized microsclerotia, which assists its survival in adverse environmental conditions in both the soil and hosts. Melanin production in fungi is mediated by two main pathways: the DHN pathway and the L-DOPA pathway [41]. Both pathways are initiated from small molecules and result in melanin synthesis under the action of various enzymes.
Although some progress has been made in the biocontrol of V. dahliae, there are few studies on the stress response and self-regulatory mechanism of V. dahliae under inhibition by biocontrol bacteria. In this study, we constructed a co-culture system to control the colony size and niche occupancy of V. dahliae in the co-culture system by inoculating Bacillus at different times. The aim of this study was to explore the self-resistance mechanism of V. dahliae under Bacillus-induced stress using transcriptome analysis and to provide clues for the control of Verticillium wilt disease.

Effect of Bacillus N-4 on the V. dahliae spores
The spore number and germination rate of V. dahliae were determined from the shaking cultures of CK, Gp0, Gp1, Gp2 and Gp3. When co-cultured with N-4, the size of the V. dahliae colony increased, the total number of spores, the germination number and germination rate also visibly increased ( Fig. 1A-C).
Interestingly, although the total number of spores in the treatment group that contained N-4 was always less than that of the CK, the germination rates of Gp3 and Gp2 were higher than that of the CK (Fig. 1B, C). The spore germination rate re ects the degree of inhibition of the fungus. This indicated that larger V. dahliae colonies are less inhibited by Bacillus.

Antagonism of Bacillus N-4 against V. dahliae
In vitro confrontation bioassays were conducted for 7 days of culture in Gp0, Gp1, Gp2 and Gp3, and the corresponding strains of V. dahliae were designated Vd0, Vd1, Vd2 and Vd3, respectively. Microscopic examination showed that the hyphae in Vd0 were circular and cracked; the microsclerotia were damaged and reduced, and many cells had incomplete cell walls (Fig. 2B). The black microsclerotia could be seen in Vd3 but were barely visible in Vd0 ( Fig. 2A). The morphological characteristics of Vd3 were closer to those of the CK.

Analysis of VMK1 gene expression
Compared with the CK, the expression of VMK1 gene was downregulated in all of the co-culture groups with the exception of Vd3. When the two strains were co-cultured for 7 days, the difference of expression of VMK1 gene was the most signi cant. Therefore, we chose this stage for subsequent RNA-Seq analysis (Fig. 3).

RNA-Seq analysis of V. dahliae
Four genes related to the TOR pathway and one gene related to melanin expression were selected to verify the reliability of RNA-Seq. Comparisons of the FPKM and RT-qPCR data showed consistency and correlation between the two datasets ( Figure S1).
Nine libraries (three replicates each of the CK, Vd0, and Vd3, covering each of the key time points) were produced by RNA-Seq. Both the abundance of the genes detected and the length of transcripts were deemed su cient by assessing the number of clean reads obtained in each cDNA library after quality ltering. We obtained an average of 46,240,973 reads per sample ( Table 2). In total, 89.37% of the total reads from all nine samples were aligned to the V. dahliae genome (VdLs.17 genome) with < 2 bp mismatches. Mapping analyses revealed that 10,811 of the predicted genes were expressed. Of these, 97.78% were annotated in the database described above (Table S2).
To elucidate the dynamic changes in gene expression in the three experimental groups, the genes that were differentially expressed among the libraries (FDR < 0.05) were identi ed. Comparisons of gene expression among those groups showed that 144 (116 upregulated and 28 downregulated), 542 (158 upregulated and 384 downregulated), and 18 (15 upregulated and 3 downregulated) genes were signi cantly differentially expressed between the CK and Vd0, Vd0 and Vd3, and Vd3 and CK, respectively (FDR < 0.05) (Fig. 4A). KEGG enrichment analyses were conducted to explore the potential roles of these DEGs in more detail (Figs. 4B-F). TOR and Mitophagy pathways were signi cantly enriched (FDR < 0.05) in the 158 genes that were upregulated between Vd0 and Vd3.
The clustering a nity search technique (CAST) and screening of the DEGs were used to generate gene clusters. The expression of gene clusters was identi ed using the annotation tags in the database and included central metabolism, signaling pathway, transcription factor, biosynthesis of secondary metabolites, ROS scavenger, and cell wall ( Fig. 5A-D). It showed that in the larger-scale community of V. dahliae in Vd3, the genes involved in glycometabolism and fatty acid degradation were downregulated, while those involved in nitrogen metabolism were more active (Fig. 5A). Signaling pathways that were observed included MAPK, RGS, GPCR, and calcineurin. However, their expression was not signi cantly different among the three groups (Fig. 5B). Overall, the expression of four transcription factors in CK and Vd3 was similar (Fig. 5C). Gene expression related to secondary metabolism was downregulated in Vd0, while the expression related to ABC transporters was upregulated in Vd3 (Fig. 5D). The genes related to ROS scavenging were generally upregulated in the stress groups (Fig. 5E). The genes related to cell wall formation were upregulated in Vd3, while those that contribute to the degradation of cell walls were upregulated in Vd0 (Fig. 5F).

Analysis of the genes related to scavenging ROS
We observed that the expression of gene clusters related to scavenging ROS was generally higher in the stress group compared with that in the control (Fig. 5E). Therefore, we constructed heat maps of the key gene expression pro les that scavenge ROS (Fig. 6). It is notable that Prx1 (VDAG_06751) was upregulated in Vd3 (Fig. 6H). The stress groups showed markedly higher activity of pyridoxal reductase (VDAG_03228, VDAG_04222) and pyridoxal kinase (VDAG_04313) compared with the control (Fig. 6G). VDAG_08350 and VDAG_04529 are putative transcription factors that relate to scavenging ROS, as characterized by the EggNOG and GO databases, and their expression was upregulated in Vd3 to some extent (Fig. 6F).

Analysis of the genes related to microsclerotia formation pathways
We combined the heat maps of some key genes in the formation of microsclerotia with the two main pathways, DHN and L-DOPA, of microsclerotia formation (Fig. 7). The results showed that the VdCmr1 (VDAG_01770) gene in the DHN pathway was upregulated in Vd3, and the VDAG_03665 gene was signi cantly upregulated in Vd0 compared with that in the CK. The genes VDAG_07137 and VDAG_04798 related to tyrosinase in the L-DOPA pathway were upregulated in vd0, compared with the CK.

Analysis of the genes related to cell wall synthesis
In the phenotypic study conducted above, it was observed from microscopy that the cell wall of Vd3 microsclerotia was thicker, and the cell connection was closer than that in the CK (Fig. 2B). Several important genes related to cell wall synthesis were upregulated in Vd3 (Fig. 8F). These included VDAG_02340 (1,3-beta-glucan synthase component GLS2), VDAG_02341 (1,3-beta-glucan synthase component GLS2), and VDAG_02305 (chitin synthase regulatory factor). RhoA (VDAG_06919), PKC1 (VDAG_09909), and BCK1 (VDAG_00874) were the key genes in the CWI pathway that were also upregulated in Vd3 (Fig. 8E). We analyzed the expression of genes that are important components of the cell wall, including chitin and glucan. We found that compared with CK, Vd0 expressed more analytic enzymes, while Vd3 expressed more synthetic enzymes (Fig. 5F).

Discussion
The TOR pathway generally plays a role in the regulation of cells in severe environments. Biocontrol bacteria compete with V. dahliae for niches and nutrition and release compounds, such as LPs, that inhibit the growth of V. dahliae. This pathway should be inhibited in fungi under stress, just as observed in Vd0. However, it was enriched and upregulated in Vd3. Elements upstream of TORC activate the pathway. GATOR2 (Fig. 9B) is a central node in the activation of TORC1 activated by high levels of the amino acids leucine and arginine [20,42] with the Rag family of guanosine triphosphatases (GTPases), downstream of GATOR2, promoting the TORC1 response to amino acids [43]. Another upstream element, Rheb, acts as a signal receptor to activate TORC1 through direct interaction [44]. Alternatively, the tuberous sclerosis complex (TSC) and GATOR1 (Fig. 9D) respond to poor growth conditions and inhibit TORC1 [45,46]. Moreover, PI3K can regulate TORC2 and plays an essential role in the TOR pathway [47]. Overall, the upstream genes that activate TORC were found to be upregulated in Vd3 (Fig. 8C), particularly in the case of GATOR2, which may largely explain the upregulation of TORC. We found that Vd3 may accumulate arginine and leucine or another activator in vivo to activate the expression of GATOR2. These results demonstrate that the nutritional status of Vd3 was su cient, at least compared with that in Vd0.
The mitophagy pathway was also signi cantly enriched (FDR < 0.05) in the 158 genes that were differentially upregulated between Vd0 and Vd3. In yeast, the ATG family proteins ATG32, ATG11, and ATG8 play unique and vital roles in mitophagy [48]. In addition, the mitophagy pathway plays an integral role in the response to stress. In lamentous fungi, ATG24 is thought to be homologous to ATG32 in yeast [49]. All the ATG proteins were upregulated in Vd3 (Fig. 8G).
Vd0 and Vd3 all bind themselves to scavenging ROS. However, Vd3 was more e cient at scavenging ROS in mitochondria and thus, avoiding mitochondrial-induced apoptosis [50,51]. Mitochondria may act as the Achilles' heel of fungi [52]. The expression of Aup1, which mediates the regulation of RTG3 to participate in the retrograde signaling pathway (RTG) [53], was also prominently upregulated in Vd3. The RTG signaling-mediated cell response to mitochondrial dysfunction is involved in the enclosed induction of antioxidant defense and stress resistance [54]. Therefore, Vd3 with a larger colony can more effectively utilize the function of mitochondria and avoid damage to itself.
The peroxiredoxin isoform, Prx1, which is found in mitochondria and acts as a ROS scavenger, was signi cantly upregulated in Vd3. A recent study showed that high levels of Prx1 expression are related to niche adaptation [55] (Fig. 6H). Peroxiredoxins are key intermediates in removing ROS and quantitatively are the most abundant cellular redox defense. In yeast, there are ve isoforms encoded by independent genes, with AHP1, TSA1, and TSA2 found in the cytoplasm, Dot5 in the nucleus, and Prx1 in the mitochondria. The larger colonies (Vd3) appeared to be more effective at expressing pyridoxal kinase to produce pyridoxal phosphate, which is an active form of vitamin B6. Additionally, vitamin B6 (pyridoxine) and its derivatives are also e cient at removing ROS [56]. Lastly, we explored the response of transcription factors to the ROS (Fig. 6F). The activator protein 1 (AP-1) transcription factor governs the low level defense response to oxidative stress and is responsible for regulating the transcription of various genes involved in ROS detoxi cation [57]. VdSkn7 plays a key role in microsclerotial development and stress resistance [58].
VdCmr1 (VDAG_01770), a key transcription factor that regulates melanin synthesis associated with the DHN pathway [59] was found to be upregulated in Vd3 (Fig. 7). Other genes involved in microsclerotia formation (e.g., VdCrz1, VdMsb2, and VdMsn2) [60-62] did not show differential expression among the three stress groups (Fig. S2). In this study, the Vd3 with a larger colony adopted more effective strategies to attain better resistance. We found that maintaining the integrity of the cell wall is advantageous over melanin synthesis under the stress induced by Bacillus or LPs, since melanin is not essential for microsclerotia development. Bell had previously found genetic mutants of V. dahliae that can produce microsclerotia without pigmentation [63]. Moreover, melanin granules are deposited within the cell walls and between the cells, cross-linking with components of the cell wall, such as chitin [64,65]. In our study, although the genes directly involved in melanin production were signi cantly upregulated in Vd0, the number of microsclerotia did not increase. Observations of the microsclerotia morphology revealed that the cell walls in Vd0 were damaged (Fig. 3B). Other than in Vd0, the cell wall in Vd3 was stronger, and the expression of melanin synthesis genes was similar to that of the CK. More importantly, the key transcription factor VdCmr1 was signi cantly upregulated. This study found that in VD3, the TOR and CWI pathways were signi cantly enriched in the upregulated genes related to cell wall synthesis (Fig. 8). This indicates that the TOR pathway plays a potential role in the structural enhancement of Vd3 cell wall. The CWI pathway is the primary signaling pathway that is responsible for the regulation of cell wall stress responses and also participates in the transduction of other stimuli, such as oxidative and genotoxic stressors [66,67]. In summary, both the TOR and CWI pathways played vital roles in strengthening the cell wall of Vd3. Therefore, our study shows that under stress conditions, V. dahliae can ensure the functional integrity of its cell wall by upregulating the genes related to cell wall synthesis. The cell wall plays an important role in the production of micronuclei of V. dahliae.
This study did not seek to overturn the potential of biocontrol bacteria as a new era of biological control pathogens. Bacillus strain N-4 was selected for isolation owing to its excellent ability to inhibit V. dahliae. Extensive research has shown that Bacillus spp. or their LPs can in uence the secretion of chitinase and glucanase, damage biological membranes, suppress spore germination and mycelial growth, and ruin the formation of microsclerotia in V. dahliae [68][69][70][71]. Nevertheless, these studies were not focused on fungal colonies, but on dispersed fungal spores in culture media with added bacterial fermentation liquid or directly added biocontrol bacteria for co-culture. These approaches did not enable the formation of V. dahliae colonies characterized by the direct communication of hyphae that is indicative of colony structure.
In our study, we determined that a larger colony of V. dahliae could survive the stress of biocontrol bacteria. In fact, the aggregation of pathogenic fungi on plant is inevitable. For example, the gray mold caused by Botrytis cinerea can form infection cushions during penetration [72]. According to observations in the early stage of this experiment, the CK generally produced microsclerotia that were visible to the naked eye approximately 5 days after inoculation. Therefore, the reason for the resistance of Vd3 could be that it entered a more resistant stage earlier than Vd0. For example, it produced microsclerotia earlier. Therefore, the prevention of diseases is often more effective than treating active ones. The optimal effect can be achieved by the balanced application of biofungicides. In this process, the focus must be on the mechanism of fungal resistance. This result can provide insight into the prevention and treatment of Verticillium wilt and other diseases caused by fungi in agricultural production.

Conclusions
This study showed that larger colonies of V. dahliae were less inhibited by Bacillus. When faced with biocontrol bacteria, larger colonies can regulate the TOR pathway, cell wall synthesis, microsclerotia formation, and scavenge ROS, so as to preserve homeostasis in vivo and more effectively participate in niche competition. This study was based on the antagonistic effect of Bacillus on V. dahliae and revealed the self-regulatory mechanism of V. dahliae after inhibition by transcriptome analysis. This can explain why treatment using microbial agents can sometimes be ineffective and suggests that the prevention of Verticillium wilt by Bacillus is more effective than the treatment.

Microbial materials and culture conditions
Bacillus strain N-4 (Bacillus spp.) was isolated from Sophora alopecuroides and preserved in the Key Laboratory of Agricultural Biotechnology, Shihezi University, Xinjiang, China. The strain was cultured continuously in Luria-Bertani broth (LB) at 30°C for 2 days, and single colonies were obtained on LB agar using the spread plate method.
V. dahliae strain VdSHZ-9 (a defoliating pathotype) was isolated from cotton soil infected with Verticillium wilt in Shihezi, Xinjiang, China. The strain was cultured on potato dextrose agar (PDA) at 25°C in the dark. Strains of V. dahliae were inoculated into potato dextrose broth (PDB) and cultured at 25°C and 180 rpm in the dark for 3 days. Subsequently, the spore suspension was diluted to approximately 10 -6 , and single colonies were obtained on PDA agar plates. A single colony was inoculated into a new PDA petri dish for cultivation in the future.
Dual culture method of strains First, 8 mm diameter mycelial discs from V. dahliae isolates were placed on 9 cm PDA plates and cultured at 25℃ in the dark. On days 0, 1, 2 and 3 days of culture, Bacillus N-4 was inoculated in four diagonal directions, which were 3 cm away from the mycelial disc. According to the time of inoculation of Bacillus N-4, the co-cultured groups were designated Gp0, Gp1, Gp2 and Gp3 ( Table 1). The CK group was inoculated with a V. dahliae mycelial disc at the beginning but was not treated with Bacillus. All the groups continued their culture at 25℃ in the dark with 10 replicates per group.

Effect of Bacillus N-4 on V. dahliae spores
On days 5 to 10 of the co-culture of fungi and bacteria in each group and days 5 to 13 of the CK culture, a mycelial disc 8 mm in diameter was taken from the edge of the colony of V. dahliae and inoculated into PDB each day. A suspension of spores was obtained after 4 days of culture at 25℃ and 180 rpm ( Table  1). The spore suspension was diluted 100-fold and counted with a hemocytometer. The average value was taken as the number of spores. In addition, the spore suspension was diluted to 10 -5 , 10 -6 , and 10 -7 , respectively, and then cultured on PDA to obtain a single colony and counted. The average value was the number of germinated spores. The percentage of germination was calculated as follows: percent germination(%) = 100 × (germinated spores)/(total spores Antagonism of Bacillus N-4 against V. dahliae On day 7 of the co-culture of bacteria and fungi from the Gp0 and Gp3 groups, the front and back sides of the petri dish were photographed to record the colony morphology. To stay consistent with the culture time of V. dahliae in Gp3, the CK was photographed on day 10 ( Table 1). Hyphae and microsclerotia were collected from the V. dahliae colony, placed on a glass slide that contained sterile water, covered with a coverslip which was tapped slightly to ensure dispersion, and observed under a light microscope.

Determination of VMK1 gene expression
On days 5 to 10 of the co-culture of fungi and bacteria in each group and days 8 to 13 of the CK culture, a mycelial disc 8 mm in diameter was taken from the edge of the colony of V. dahliae and inoculated into PDB each day. A spore suspension was obtained after 7 days of culture at 25℃ and 180 rpm. This period of growth enabled the fungus to accumulate enough biomass that RNA could be extracted (Table 1). V. dahliae was then ground to a ne powder in liquid nitrogen using a mortar and pestle. The total RNA was extracted using an EZgene TM Fungal RNA Lit (Biomiga, Shanghai, China). After elution in RNase-free water, the RNA was stored at −80°C until further use.
Total RNA was used for the rst-strand cDNA synthesis using a RevertAid TM First Strand cDNA Synthesis Kit (Thermo Scientific, (EU) Lithuania), according to the manufacturer's instructions. Quantitative realtime PCR (qRT-PCR) was performed on a Light Cycler 480 system (Roche Diagnostics, Penzberg, Germany) using the Light Cycler 480 SYBR Green master mix (Roche Diagnostics). The internal reference for the qRT-PCR was the V. dahliae β-tubulin gene (VDAG_10074) The 2 -rrt method [73] was used to calculate the relative gene expression. The experiment had three operational and three biological repetitions. Primer designing tools (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) (Table S1) were used to design the primers. To determine the correlation between transcriptome sequencing and qRT-PCR, a Pearson coe cient was calculated using R software version 3.5 (BGI, Beijing, China).

RNA-Seq of V. dahliae
The library was constructed using an Illumina TruSeq TM RNA Sample Prep Kit according to the manufacturer's instructions and run on an Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA). Following sequencing by Majorbio, Inc. (Shanghai, China), dirty reads were separated from the high-quality clean reads by removing the reads that contained two or more Ns and low-quality reads with adaptors or two Ns. Clean reads from each sample were mapped to the V. dahliae VdLs.17 genome (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA28529) using TopHat [74] to predict the exon splice sites allowing < two mismatches. The DESeq Bioconductor package [75] was used to normalize the mapped reads. The sequences of genes and transcripts were compared using Diamond software [76] in three databases: NR (NCBI non-redundant protein sequences, using e-value = 1e -5 ), Swiss-Prot (using evalue = 1e -5 ), and EggNOG (evolutionary genealogy of genes: Non-supervised Orthologous Groups, using e-value =1e -5 ). The GO annotations were used in Blast2GO [77], Pfam annotations were used in Hmmer [78], and KEGG annotations were used in KOBAS2.1 [79], all with e-values equal to 1e -5 .
Analysis of differentially expressed genes FPKM (fragments per kilobase of transcript per million mapped fragments) in Cu inks [80] were used to calculate the levels of gene expression. The edgeR program [81] was used to determine the differentially expressed genes (DEGs) between the samples, using the standards of adjusted P-value ≤0.05 and fold change ≥1.5 or ≤ 0.67. Goatools [82] was used to analyze the GO enrichment of the DEGs, while KEGG pathway analysis was performed with KOBAS software [83]. Tables Table 1 The timetable for ve groups of experiments.   Day  0  1  2  3  4  5  6  7  8  9  10  11  12  13 Control > _ _ _ q q p/q/O q q q Gp3 > < q q p/q/O q q q Gp2 > < q q q q q q Gp1 > < q q q q q q Gp0 > < q q p/q/O q q q The symbol '>' represents the inoculated V. dahliae. The symbol '<' represents the inoculated Bacillus spp. The symbol '_' represents the determination of spore viability. The abbreviation q represents qRT-PCR. The abbreviation p represents the photo taken, and the O represents the RNA-Seq sampling time. qRT-PCR, quantitative real-time PCR    The expression of VMK1 gene in Verticillium dahliae in the CK, GP0, Gp1, GP2 and GP3 groups. In the CK, the X axis represents gene expression in V. dahliae cultured for 8 to 13 days and for the stress groups, the X axis represents gene expression in V. dahliae after interaction with Bacillus for 5 to 10 days.    The pro les of expression and schematic diagram related to melanin production. The heatmap data was produced by TBtools and the symbol '*' represents signi cant regulation based on the criteria of P adjusted < 0.05 and fold change ≥1.5 or≤ 0.67 between Vd0 and Vd3. (A) DHN-pathway is referred by (Howlett et al. 2012). (B) The L-DOPA pathway is based on Tyrosine Metabolism in the KEGG Pathway.