UvHOS3‐mediated histone deacetylation is essential for virulence and negatively regulates ustilaginoidin biosynthesis in Ustilaginoidea virens

Abstract

Ustilaginoidea virens is the causal agent of rice false smut, which has recently become one of the most important rice diseases worldwide. Ustilaginoidins, a major type of mycotoxins produced in false smut balls, greatly deteriorates grain quality. Histone acetylation and deacetylation are involved in regulating secondary metabolism in fungi. However, little is yet known on the functions of histone deacetylases (HDACs) in virulence and mycotoxin biosynthesis in U. virens. Here, we characterized the functions of the HDAC UvHOS3 in U. virens. The ΔUvhos3 deletion mutant exhibited the phenotypes of retarded growth, increased mycelial branches and reduced conidiation and virulence. The ΔUvhos3 mutants were more sensitive to sorbitol, sodium dodecyl sulphate and oxidative stress/H₂O₂. ΔUvhos3 generated significantly more ustilaginoidins. RNA‐Seq and metabolomics analyses also revealed that UvHOS3 is a key negative player in regulating secondary metabolism, especially mycotoxin biosynthesis. Notably, UvHOS3 mediates deacetylation of H3 and H4 at H3K9, H3K18, H3K27 and H4K8 residues. Chromatin immunoprecipitation assays indicated that UvHOS3 regulates mycotoxin biosynthesis, particularly for ustilaginoidin and sorbicillinoid production, by modulating the acetylation level of H3K18. Collectively, this study deepens the understanding of molecular mechanisms of the HDAC UvHOS3 in regulating virulence and mycotoxin biosynthesis in phytopathogenic fungi.

UvHOS3 in Ustilaginoidea virens controls histone deacetylation at H3K9, H3K18, H3K27 and H4K8 and positively regulates fungal growth and pathogenicity, but negatively modulates mycotoxin biosynthesis.

1. INTRODUCTION

Rice false smut caused by Ustilaginoidea virens has emerged as one of the most important rice diseases and poses a significant economic impact on rice commercial production worldwide (Brooks et al., 2010; Fan et al., 2016; Tanaka et al., 2008). Ustilaginoidea virens infects and colonizes rice florets with hyphae, which are eventually transformed into false smut balls covered with powdery chlamydospores (Fan et al., 2020; Sun et al., 2020;  Yang, Zhang, Wang, et al., 2022; Zheng et al., 2022). False smut balls contain a large amount of mycotoxins that not only deteriorate grain quality but also threaten the health of both animals and people (Fu et al., 2018; Guo et al., 2012; Zhang et al., 2014). As one of the major types of mycotoxins produced by U. virens, the polyketide ustilaginoidins have at least 27 derivatives, most of which have been structurally identified (Li et al., 2019; Meng et al., 2015; Meng, Zhao, et al., 2021; Sun et al., 2017). Ustilaginoidins exhibit a variety of biological activities. For instance, ustilaginoidin D, one of the major derivatives of ustilaginoidin, has both strong antituberculosis activity and significant inhibitory activity on KB cells (Kong et al., 2013; Koyama et al., 1988; Ugaki et al., 2012). In addition, ustilaginoidin D causes liver damage and the disorder of lipid metabolism in zebrafish (Wang et al., 2021). The ustilaginoidin synthesis (ugs) gene cluster, containing at least 14 genes, has been identified to be responsible for ustilaginoidin biosynthesis and modification in U. virens. The UvPKS1‐knockout mutant produces neither ustilaginoidin derivatives nor intermediates, while the ΔugsO and ΔugsT mutants generate significantly less ustilaginoidins. UgsL functions as a laccase to dimerize monomer precursors of ustilaginoidins (Li et al., 2019; Xu et al., 2021). However, little is yet known about regulatory mechanisms of ustilaginoidin biosynthesis in U. virens.

In filamentous fungi, epigenetic modification, such as histone acetylation and deacetylation, plays an important role in the regulation of fungal growth and development, tolerance to external stresses, secondary metabolism and pathogenicity (Fan et al., 2017; Liu, Wang, et al., 2023; Macheleidt et al., 2016; Wiemann et al., 2013; Yin et al., 2019; Zhang, Guo, et al., 2022). Histone is acetylated by histone acetyltransferases (HATs), which is reversed by histone deacetylases (HDACs; Ma et al., 2013). Fungal HDACs are categorized into three major classes: class I RPD3‐type HDACs include RPD3 and HOS2; class II HDA1‐type HDACs contain HDA1 and HOS3; SIR2 and HST1‐4 belong to the class III NAD⁺‐dependent ‘sirtuin’ type HDACs (Lan et al., 2019; Pidroni et al., 2018). Both HATs and HDACs regulate gene expression and secondary metabolism through chromatin remodelling (Ma et al., 2013; Martin et al., 2021; Park & Kim, 2020). The hdaA mutants in Aspergillus nidulans produce significantly increased penicillin and carcinogenic aflatoxin precursor sterigmatocystin (Shwab et al., 2007). Similarly, the levels of phenylahistin and penicillic acid were substantially increased in Aspergillus calidoustus and Aspergillus westerdijkiae, respectively, after the treatment of an HDAC inhibitor vorinostat, illustrating the potential of HDAC modulation for improving the yield of secondary metabolites (Aldholmi et al., 2020). In contrast, inhibition of Zn²⁺‐dependent HDACs results in decreased gibberellin biosynthesis in Fusarium fujikuroi. Besides, the ffhda1 and ffhda2 mutants exhibit a significantly reduced production of gibberellins, bikaverin and fusaric acid under normally inducing conditions (Studt et al., 2013). The deletion of HosA and Hda1 causes decreased production of aflatoxin B1 (AFB1) in Aspergillus flavus (Lan et al., 2019). Consistently, the production of penicillin by the A. nidulans hosA mutants is significantly reduced and hardly detectable (Pidroni et al., 2018). Accordingly, different fungal HDACs have distinct, sometimes opposite effects on the regulation of secondary metabolism.

The class I HDAC HOSA and HOS2 are essential for vegetative growth and pathogenicity in A. flavus, Alternaria alternata and Beauveria bassiana (Cai et al., 2018; Lan et al., 2019; Ma et al., 2021). Consistently, the deletion of Rpd3 results in significantly attenuated pathogenicity in Fusarium graminearum. However, Rpd3 deletion in diverse fungi has differential effects on hyphal growth. The rpd3 mutants display severe growth defects in F. graminearum and A. alternata but show no altered growth phenotype in A. flavus (Jiang et al., 2020; Lan et al., 2019; Ma et al., 2021). The deletion of Hos3 and Hda1, encoding class II HDACs, in A. alternata has no significant effect on mycelial growth, responses to environmental stresses and virulence. The class II HDAC HdaA is not only involved in regulating mycelial growth, but also positively regulates sclerotium production in A. flavus (Lan et al., 2019; Ma et al., 2021). Therefore, the functions of certain HDAC genes are often different in vegetative growth, sexual reproduction and pathogenicity among diverse fungi.

In Saccharomyces cerevisiae, HOS3 plays an important role in life span (Pang et al., 2022). HOS3 specifically associates with nuclear pore complexes in daughter cells during mitosis and deacetylates Nup60 in G1 daughter cells, thus delaying cell cycle entry to restrain their premature commitment to a new cell division cycle (Gomar‐Alba et al., 2022; Kumar et al., 2018). Yeast HOS3 has an intrinsic HDAC activity and a distinct specificity to H2AK7, H2BK11, H3K14, H3K23, H4K5 and H4K8 in vitro. H2BK11 acetylation and de‐acetylation function as a switch from cell proliferation to cell death (Ahn et al., 2006; Carmen et al., 1999; Trojer et al., 2003). HOS3 is also involved in regulating cellular defences against oxidative, nitrosative and nutritional stresses through modulating the levels of reactive oxygen species and nitric oxide in yeast (Ahn et al., 2006; Aoyama et al., 2000; Lim et al., 2015). Collectively, HOS3 is critical for an array of biological processes in yeast. In Monascus ruber, the deletion of Mrhos3 significantly elevates the acetylation levels of H3K9, H3K18 and H4K12. MrHOS3 negatively regulates the production of the mycotoxin citrinin and pigments (Liu, Zheng, et al., 2023). However, few studies have been reported on the roles of HOS3 in phytopathogenic fungi so far.

In this study, we demonstrate that UvHOS3 is crucial for various developmental processes and virulence in U. virens. UvHOS3 negatively regulated ustilaginoidin biosynthesis in U. virens. KEGG enrichment analysis based on transcriptome data revealed that UvHOS3 regulates the expression of genes involving in vegetative growth and pathogenicity in U. virens. Combined RNA‐Seq, untargeted metabolomics and chromatin immunoprecipitation (ChIP) analyses revealed that UvHOS3 is tightly associated with the regulation of ustilaginoidin and sorbicillinoid biosynthesis. This study provides novel insights into the regulatory roles of UvHOS3 in vegetative growth, pathogenicity and mycotoxin biosynthesis in U. virens.

2. RESULTS

2.1. The HDACs inhibitor trichostatin A inhibits vegetative growth and elevates ustilaginoidin biosynthesis in U. virens

To evaluate potential roles of HDACs on mycelial growth and ustilaginoidin production in U. virens, the wild‐type strain was cultured on potato sucrose agar (PSA) plates supplemented with various concentrations of trichostatin A (TSA), a potent inhibitor of class I and II HDACs. After culturing for 7–14 days, U. virens exhibited a significantly smaller colony on TSA‐containing plates (Figure 1a, Figure S1a,b). Subsequently, high‐performance liquid chromatography (HPLC) assays revealed that the amount of ustilaginoidins generated in U. virens cultures was gradually increased when U. virens was cultured on PSA plates with elevated doses of TSA (Figure 1b and Figure S1c). Altogether, these results suggest that classical HDACs are involved in regulating mycelial growth and ustilaginoidin biosynthesis in U. virens.

FIGURE 1.

The histone deacetylase inhibitor trichostatin A (TSA) inhibits mycelial growth but promotes ustilaginoidin biosynthesis in Ustilaginoidea virens. (a) Vegetative growth of the wild‐type U. virens strain P1 after culturing on potato sucrose agar (PSA) plates supplemented with various concentrations of TSA for 7 and 14 days. (b) HPLC assays to determine the amount of ustilaginoidins generated in 28‐day‐old U. virens cultures on PSA plates with various concentrations (0, 1, 2 and 5 μM) of TSA.

2.2. Phylogenetic analysis and structural characteristics of HDAC in U. virens

Bioinformatic analysis predicted that U. virens genome encodes seven HDACs including class I members UvRPD3 and UvHOS2, class II members UvHDA1 and UvHOS3 and class III members UvSIR2, UvHST2 and UvHST4. We performed phylogenetic and domain analyses of HDAC homologues in six fungal species, including U. virens, S. cerevisiae, B. bassiana, Magnaporthe oryzae, A. alternata and A. flavus (Figure S2a). The results showed that these predicted HDACs all possess the conserved catalytic domains that are characteristic of distinct classes of HDACs (Figure S2b) and that most HDAC homologues in U. virens and B. bassiana are phylogenetically closest among the tested fungi.

Although both HDA1 and HOS3 belong to class II HDACs, HDA1 has been well studied in plant‐pathogenic fungi, whereas the function of HOS3 remains obscure, especially in regulating secondary metabolism. Genomic analysis revealed that UvHos3 in U. virens encodes a protein of 1198 amino acids. The phylogenetic tree constructed based on the conserved HDAC domains indicates that UvHOS3 is phylogenetically closest to the homologues in B. bassiana (Figure S2a,b).

2.3. UvHos3 is required for mycelial growth, conidiation and virulence in U. virens

To investigate biological functions of UvHOS3, the UvHos3 gene replacement mutants were generated in U. virens through CRISPR/Cas9‐induced homologous recombination (Liang et al., 2018) (Figure S3a). Five independent UvHos3 deletion mutants were confirmed by PCR and Southern blot analyses (Figure S3b,c). Subsequently, a plasmid‐borne UvHos3‐FLAG gene construct with its native promoter was transformed into the ΔUvhos3‐1 mutant to generate the complemented strain CΔUvhos3‐2. The expression of UvHOS3‐FLAG in CΔUvhos3‐2 was confirmed through immunoblotting (Figure S3d). The deletion mutants all showed the retarded growth phenotype compared with the wild‐type strain (Figure 2a and Figure S4a). The complemented strain CΔUvhos3‐2 restored the growth rate to the wild‐type level. Through microscopy observation and quantification of conidia, we revealed that the ΔUvhos3 mutants showed increased hyphal branches and produced significantly fewer conidia than did the wild‐type strain. The wild‐type and complemented CΔUvhos3‐2 strains showed no difference in hyphal branches and conidiogenesis (Figure 2b,d,e).

FIGURE 2.

UvHos3 is required for mycelial growth and development, conidiation and virulence. (a) Colony diameters of the wild‐type, ΔUvhos3 and complemented CΔUvhos3‐2 strains cultured on potato sucrose agar (PSA) plates for 7 days. (b) Conidiospores produced in the wild‐type (P1), ΔUvhos3‐1, ΔUvhos3‐8 and complemented CΔUvhos3‐2 strains were quantified when culturing in potato sucrose broth for 7 days. (c) The average number of false smut balls formed on rice panicles inoculated with the wild‐type, ΔUvhos3‐1, ΔUvhos3‐8 and complemented CΔUvhos3‐2 strains. A total of 30 rice panicles were inoculated and false smut balls counted for each strain. Data are shown as mean ± SD (n = 30). (d) Microscopy observation of mycelial morphology. Hyphal branches are indicated by white arrows. Scale bar: 50 μm. (e) The number of hyphal branches in the wild‐type, ΔUvhos3 and complemented CΔUvhos3‐2 strains cultured on PSA plates for 7 days. Representative data from three independent assays are shown as mean ± SD (n = 100). Different lowercase letters represent statistically significant differences according to one‐way analysis of variance followed by Duncan's multiple range test (p < 0.05).

Next, we evaluated the role of UvHos3 in U. virens virulence by injection inoculation. The ΔUvhos3 mutants produced significantly fewer false smut balls than did the wild‐type strain at about 4 weeks after panicles of the rice cultivar LYP9 were inoculated with the wild‐type, mutant and complemented strains before heading. The complemented strain restored the ability to generate false smut balls close to the wild‐type level (Figure 2c and Figure S4b). Altogether, these results indicate that UvHos3 is required for mycelial growth and development, conidiogenesis and virulence in U. virens.

2.4. UvHOS3 functions differentially in response to various external stresses and negatively modulates ustilaginoidin biosynthesis

To further investigate the role of UvHos3 on stress tolerance, the wild‐type, ΔUvhos3 and CΔUvhos3 strains were cultured on yeast extract‐tryptone (YT) medium plates supplemented with sorbitol, H₂O₂, Congo Red, TSA or sodium dodecyl sulphate (SDS) at 28°C for 14 days (Figure 3a and Figure S5a). The ΔUvhos3 mutant exhibited significantly greater inhibition rates than the wild‐type strain when these strains were cultured on the plates containing the hyperosmotic stress sorbitol, oxidative stress H₂O₂ and cytomembrane stress SDS, whereas the mutant showed a lower inhibition rate on the TSA‐containing medium plates. Interestingly, the wild‐type and mutant strains exhibited no significant difference in sensitivity to the cell wall stress Congo Red. Taken together, the results indicate that the deletion of UvHos3 in U. virens results in higher sensitivity to hyperosmotic, cytomembrane and oxidative stresses, and higher tolerance to the HDAC inhibitor TSA.

FIGURE 3.

UvHOS3 differentially regulates Ustilaginoidea virens responses to distinct external stresses and negatively modulates ustilaginoidin biosynthesis. (a) The growth inhibition rates of the wild‐type, ΔUvhos3‐1 and complemented CΔUvhos3‐2 strains in response to different stress factors. The diameters of colonies were measured and analysed after the tested U. virens strains were cultured on yeast extract‐tryptone (YT) plates and YT plates supplemented with 0.5 M sorbitol, 0.03% H₂O₂, 240 μg/mg Congo Red, 1 μM trichostatin A (TSA) or 0.03% SDS. Data are shown as mean ± SD (n = 3). Different lowercase letters indicate statistically significant differences in growth inhibition rates according to one‐way analysis of variance followed by Duncan's multiple range test (p < 0.05). (b) HPLC assays to quantify ustilaginoidins produced in the cultures of the wild‐type, mutant and complemented strains after 28 days of growth on potato sucrose agar plates.

In addition, we determined the effect of UvHos3 deletion on ustilaginoidin biosynthesis. HPLC assays demonstrated that the ΔUvhos3‐1 mutant produced significantly more ustilaginoidins than did the wild‐type strain, and the CΔUvhos3 strain restored the ability to produce ustilaginoidins when these strains were cultured on PSA plates for 28 days (Figure 3b and Figure S5b). These results indicate that UvHOS3 negatively modulates ustilaginoidin biosynthesis.

2.5. UvHos3 deletion alters metabolite profiles in U. virens

To illustrate metabolic changes caused by UvHos3 deletion, untargeted metabolomic analysis was performed to compare metabolite profiles in the wild‐type and ΔUvhos3 strains. First, data sets obtained from UHPLC–MS were subjected to Orthogonal PLS‐DA (OPLS‐DA). The samples from the wild‐type and ∆Uvhos3 strains were clearly separated in two distinct groups, indicating that the wild‐type and ∆Uvhos3 strains exhibit a significant difference in metabolite profiles (Figure 4a). The volcano map revealed 565 differentially accumulated metabolites (DAMs) including 258 downregulated metabolites and 307 upregulated metabolites in the ΔUvhos3 strain (Figure 4b). These differential metabolites contain 15 polyketide‐related metabolites, among which the majority of metabolites were elevated in the ΔUvhos3 strain (Table S2). Interestingly, multiple types of mycotoxins, including ochratoxin B, zearalenones, zeranol, mycotoxin T2, ustilaginoidin D, aflatoxin B1 and aflatoxin G2, were significantly increased in the ΔUvhos3 strain (Figure 4c). These results indicate that UvHOS3 extensively regulates the biosynthesis of secondary metabolites, particularly inhibiting mycotoxin production.

FIGURE 4.

UvHOS3 is extensively involved in the regulation of mycotoxin biosynthesis. (a) The plot generated by orthogonal PLS‐DA (OPLS‐DA) showing that the metabolomics replicates of the wild‐type and ΔUvhos3 strains were clustered in distinct groups. Each point represents a biological replicate. (b) A volcano map to exhibit differential metabolites between the wild‐type and ΔUvhos3 strains. Red and blue dots represent significantly upregulated and downregulated metabolites in the ΔUvhos3 strain, respectively. The criteria for differential metabolites are variable importance in the projection (VIP) >1.0, p < 0.05, and absolute log₂ (fold change) ≥ 1. (c) The heatmap showing differential mycotoxin metabolites ochratoxin B, zearalenones, zeranol, mycotoxin T2, ustilaginoidin D, aflatoxin B1 and aflatoxin G2 produced in the wild‐type and ΔUvhos3 strains. The relative quantitative values of differential metabolites are indicated by the colour bar.

2.6. UvHOS3 mediates histone deacetylation at H3K9, H3K18, H3K27 and H4K8 sites in U. virens

To explore whether UvHOS3 is involved in the deacetylation of histones H3 and H4, total proteins extracted from 7‐day cultures of the wild‐type, ΔUvhos3‐1 and CΔUvhos3‐2 strains were subjected to immunoblotting with site‐specific anti‐acetylated‐H3 and ‐H4 antibodies. We demonstrated that the H3K9, H3K18, H3K27 and H4K8 residues in the ΔUvhos3 mutant were hyperacetylated compared with those in the wild‐type and CΔUvhos3 strains, whereas no significant change in the acetylation level was detected at H4K5 among different strains (Figure 5a,b). Collectively, these results suggest that UvHOS3 mediates deacetylation of H3 and H4 at the H3K9, H3K18, H3K27 and H4K8 residues in U. virens.

FIGURE 5.

UvHOS3 mediates deacetylation of histones H3 and H4 at the residues H3K9, H3K18, H3K27 and H4K8 in Ustilaginoidea virens. (a) Western blot analyses to detect the acetylation levels at different lysine residues with site‐specific anti‐acetylation antibodies against H3 and H4 in the wild‐type, ΔUvhos3‐1 and complemented CΔUvhos3‐2 strains. The protein levels of β‐actin, H3, and H4 were detected via immunoblotting to indicate protein loading. (b) The relative acetylation levels of H3 and H4 at different lysine residues in the wild‐type, mutant and complemented U. virens strains. Band intensities were quantified by densitometry using ImageJ. The data from three biological replicates are shown as mean ± SD (n = 3). Different lowercase letters indicate statistically significant differences in the relative acetylation levels at different lysine residues in H3 and H4 according to one‐way analysis of variance followed by Duncan's multiple range test (p < 0.05).

2.7. UvHOS3 regulates the expression of genes related to growth and pathogenicity in U. virens

To investigate how UvHOS3 affects mycelial growth and pathogenicity in U. virens, gene expression profiles of the wild‐type and ΔUvhos3 cultures were analysed by RNA‐Seq. An average 6.77 Gb of clean reads were generated for each sample (three biological repeats for each strain) with a Q30 greater than 92%, indicating that RNA‐Seq data meet the quality requirement for subsequent analyses (Table S3).

RNA‐Seq analyses revealed 1844 differentially expressed genes (DEGs) between the wild‐type and ΔUvhos3‐mutant strains, including 887 downregulated and 957 upregulated genes (Figure S6a). KEGG pathway enrichment analyses showed that DEGs were significantly enriched in 20 different biosynthesis and metabolism pathways, including fatty acid metabolism, O‐glycan biosynthesis, pentose and glucuronate interconversions, histidine metabolism, galactose metabolism, glycerophospholipid metabolism, starch and sucrose metabolism (Figure S6b). GO enrichment analysis showed that DEGs were mostly enriched in the processes of pathogenesis, integral component of membrane, fatty acyl‐CoA synthetase activity, dolichyl‐phosphate‐mannose‐protein mannosyltransferase activity and monooxygenase activity, which are mostly involved in pathogenicity and secondary metabolism (Figure S6c).

Further analyses revealed that 27 DEGs were enriched in the term ‘growth’ in GO_level 2 and 25 out of these genes were significantly downregulated (Figure 6a and Table S4). Hyphal branching is an important indicator of the mycelial growth process of filamentous fungi (Wu et al., 2016; Yang, Zhang, He, et al., 2022). Many protein kinases and septins are involved in the regulation of lateral branch formation (Harris, 2019). Through BLAST searches and gene annotation, we revealed that 24 DEGs, the majority of which were downregulated, might be associated with hyphal branching (Table S5). Interestingly, four genes (UV8b_01520, UV8b_05473, UV8b_05580 and UV8b_04699) were simultaneously involved in hyphal growth and branching. These results reveal potential association between hyphal growth and branching. It is consistent with the finding that the deletion of UvHos3 resulted in an increased number of hyphal branches and slower growth. In addition, the DEGs involved in the process of pathogenesis (GO: 0009405) were mostly inhibited in the ΔUvhos3 mutant (Figure 6b and Table S6). For instance, the downregulated genes UV8b_01319, UV8b_02987 and UV8b_05916 related to broth growth and pathogenesis are required for fungal virulence in Ustilago maydis (Pejenaute‐Ochoa et al., 2021), which were present in the enriched KEGG pathway, mannose type O‐glycan biosynthesis (ko00515; Figure S7). These results indicate that UvHOS3 regulates the expression of genes related to growth and pathogenesis, and thereby controls mycelial growth and pathogenicity in U. virens.

FIGURE 6.

UvHOS3 is involved in the transcriptional regulation of growth‐, pathogenicity‐ and mycotoxin biosynthesis‐related genes in Ustilaginoidea virens. (a–c) The heatmaps showing expression profiles of the genes involving in growth (a), pathogenicity (b) and mycotoxin biosynthesis (c) in the wild‐type and ΔUvhos3 strains. (d) Reverse transcription‐quantitative PCR (RT‐qPCR) assay was performed to confirm elevated expression of ustilaginoidin biosynthesis‐related genes in the ΔUvhos3 mutant. The expression of ustilaginoidin biosynthesis‐related genes including UgsR1, UgsS, UgsO, UvPKS1, UgsT, UgsH, UgsJ and UgsL was detected by RT‐qPCR. The α‐tubulin gene was used as an internal reference gene. Representative data from three independent assays are shown as mean ± SD (n = 3). Different lowercase letters represent statistically significant differences in the gene expression level according to one‐way analysis of variance followed by Duncan's multiple range test (p < 0.05). (e) Chromatin immunoprecipitation‐qPCR revealed the enrichment of H3K18 acetylation on chromatin of ustilaginoidin and sorbicillinoid biosynthesis‐related genes in the ΔUvhos3 mutant. The values represent mean ± SD (n = 3). Asterisks indicate significant difference in promoter region enrichment between the wild‐type and ΔUvhos3‐mutant strains (p < 0.01, Student's t test).

2.8. Mycotoxin biosynthesis‐related genes are regulated by UvHOS3 in U. virens

The culture of U. virens generates multiple types of mycotoxins, such as ustilaginoidins and sorbicillinoids (Liu, Wang, et al., 2023; Zhang, Xu, et al., 2022). The mycotoxin biosynthesis‐related genes in DEGs are summarized in Table S7. Interestingly, the sorbicillinoid biosynthesis genes UV8b_03315 (UvSorC), UV8b_03319 (UvSorT), UV8b_03320 (UvSorR2), UV8b_03321 (UvSorA) and UV8b_03322 (UvSorB) (Lai et al., 2019; Zhang, Xu, et al., 2022), isoquinoline alkaloid biosynthesis genes UV8b_06024 and UV8b_06025 and ustilaginoidin biosynthesis‐related genes UV8b_01133 (UV_2085), UV8b_01134 (UvPKS1), UV8b_01135 (UgsZ), UV8b_01136 (UgsT), UV8b_01137 (UgsH), UV8b_01138 (UgsJ) and UV8b_01139 (UgsL) (Li et al., 2019) were significantly upregulated in the ΔUvhos3‐1 mutant (Figure 6c). Elevated expression of ustilaginoidin biosynthesis genes explains why the ΔUvhos3 mutant generates more ustilaginoidins.

Next, the expression of ustilaginoidin biosynthesis‐related genes was detected by reverse transcription‐quantitative PCR (RT‐qPCR) to validate the reliability of transcriptome data. The eight tested genes in the ugs cluster were all transcriptionally upregulated in the Uvhos3 mutant compared with the wild type, particularly the key ustilaginoidin biosynthesis gene UvPKS1 with the greatest increase (about 30‐fold; Figure 6d). The wild‐type and complemented strains exhibited a similar expression level in the majority of ugs genes (Figure 6d). The results indicate that UvHOS3 negatively regulates the expression of genes involving in mycotoxin biosynthesis, especially in ustilaginoidin biosynthesis.

To investigate whether UvHOS3 regulates mycotoxin biosynthesis through controlling the acetylation levels of histones in U. virens, we performed chromatin immunoprecipitation (ChIP) analysis using an anti‐H3K18ac antibody. The primer pairs specific for the promoters of ustilaginoidin and sorbicillinoid biosynthesis‐related genes were designed for ChIP‐qPCR analyses. The results showed that the promoters of ustilaginoidin biosynthesis‐related genes UvPKS1, UgsT, UgsH, UgsJ and UgsL, and sorbicillinoid biosynthesis‐related genes UvSorT and UvSorR2 were significantly enriched in the immunoprecipitated chromatin in the mutant strain compared to the wild‐type strain (Figure 6e). Taken together, we speculate that UvHOS3 alters the status of chromatin adjacent to the promoters of the ugs and sorbicillinoid biosynthesis gene clusters by reducing the deacetylation level of H3K18, and thereby inhibits the expression of genes involving in ustilaginoidin and sorbicillinoid biosynthesis.

2.9. WGCNA reveals the modules related to mycotoxin biosynthesis and protein acetylation

To find out the core genes in the regulatory network affected by UvHos3 deletion, we performed weighted gene co‐expression network analysis (WGCNA). A total of 17 modules with different colours were identified and summarized in Figure S8a. According to the relationship between module eigengenes (ME) and traits, the expression pattern of genes in brown module was positively correlated with traits with the highest correlation coefficient in the ΔUvhos3 mutant (Figure S8b). The genes in this module are therefore considered as the important genes that are regulated by UvHOS3.

Next, we performed correlation analysis of expression patterns between each pair of genes in the brown module to construct a weighted gene co‐expression network with an edge weight cut‐off of 0.15 (Figure 7 and Table S8). Red dots in the weighted gene co‐expression network of the brown module represent upregulated genes in the mutant, including UgsZ, UgsH, UgsJ and UgsL that are involved in the production of ustilaginoidins, and the salicylate 1‐monooxygenase gene UvSorC, MFS multidrug transporter gene UvSorT, transcription factor gene UvSorR2, the PKS genes UvSorA and UvSorB involving in sorbicillinoid biosynthesis (Lai et al., 2019; Zhang, Xu, et al., 2022). Interestingly, most of the mycotoxin biosynthesis genes formed a subset together (Figure 7). In addition, the acetyltransferase gene UV8b_03374, acyl‐CoA thioesterase2 gene UV8b_01772 and acetyl‐coenzyme A synthetase gene UV8b_08076 were upregulated in the ΔUvhos3 mutant (Figure 7), indicating that the deletion of UvHos3 also promotes the expression of some acetyltransferase genes.

FIGURE 7.

Weighted gene co‐expression network reveals some UvHOS3‐regulated mycotoxin biosynthesis genes. The weighted gene co‐expression network was constructed for the brown module. Red, blue and grey dots represent upregulated, downregulated and non‐differentially expressed genes respectively (Table S7).

To unveil the underlying regulatory mechanisms in mycotoxin metabolism, we performed correlation analyses between the DEGs in the brown module and DAMs (Figure 8). The predicted connection networks showed that nine upregulated genes encoding UV8b_01139 (laccase), UV8b_07369 (SET domain‐containing protein), UV8b_03424 (cytochrome P450), UV8b_08192 (rapamycin binding protein FKBP12), five transcription factors UV8b_00896, UV8b_01263, UV8b_03320, UV8b_06218 and UV8b_06275 and four downregulated genes encoding transcription factors UV8b_03203 (sporulation resulting in formation of a cellular spore), UV8b_04413 (carbon response regulator CreA), UV8b_04482 (nitrogen response regulator AreA) and UV8b_04588 (DNA‐binding domain with preference for A/T rich regions‐like protein) were involved in the biosynthesis of 171 DAMs including ustilaginoidin D and ochratoxin B (Figure 8, Tables S9 and S10). These results provide insights into the regulatory mechanisms underlying the biosynthesis of multiple metabolites, especially mycotoxin biosynthesis in U. virens.

FIGURE 8.

Correlation analysis between differentially expressed genes (DEGs) and differentially accumulated metabolites (DAMs). Association analysis between the DEGs in the brown module and DAMs in the ΔUvhos3 mutant (p < 0.05). Red and blue dots represent upregulated and downregulated DEGs, respectively. Grey dots indicate DAMs.

3. DISCUSSION

Ustilaginoidins, one of the major types of mycotoxins in U. virens, are toxic to humans and livestock and threaten food safety. Although HDACs have been reported to be key players in regulating gene expression and secondary metabolism in fungi, knowledge about HDAC functions in virulence and mycotoxin biosynthesis in U. virens is scarce (Liu, Wang, et al., 2023). In this study, we not only demonstrated that a class II HDAC UvHOS3 positively regulates vegetative growth and development, conidiation and pathogenicity, but also revealed that UvHOS3 inhibits biosynthesis of mycotoxins, especially ustilaginoidins and sorbicillinoids in U. virens.

As a potent inhibitor of class I and II HDACs, TSA has different effects on the production of secondary metabolites in diverse fungi. TSA induces orsellinic acid biosynthesis in A. nidulans, whereas it inhibits AFB1 accumulation in A. flavus (Pidroni et al., 2018). TSA retards the growth of B. bassiana and A. flavus (Lan et al., 2019). In this study, we revealed that TSA not only delayed the growth of U. virens but also significantly increased the production of ustilaginoidins (Figure 1a,b).

Subsequently, we investigated the functions of individual HDACs in growth, development and secondary metabolism in U. virens and revealed that deletion of UvHos3 in U. virens not only reduced mycelial growth and conidial production but also attenuated its virulence to rice (Figure 2). RNA‐Seq analysis showed that the downregulated genes associated with growth and pathogenesis were enriched in the mannose type O‐glycan biosynthesis pathway (Figure S7), including UV8b_02987, UV8b_05916 and UV8b_01319 encoding mannosyltransferase Pmt1, dolichyl‐phosphate‐mannose‐protein mannosyltransferase containing protein and dolichyl‐phosphate‐mannose‐protein mannosyltransferase 2, respectively. It has been reported that Pmt1, Pmt2 and Pmt4 are involved in fungal growth, cell wall integrity and pathogenicity in Ustilago maydis (Pejenaute‐Ochoa et al., 2021). Disruption of Pmt2 leads to defective growth, decreased conidial yield and impaired virulence in M. oryzae and Metarhizium acridum (Guo et al., 2016; Wen et al., 2021). In fungi, delayed hyphal growth can be also associated with increased hyphal branches. For example, deletions of the protein kinase gene Gil1 in A. nidulans and the septin gene AspB in F. graminearum lead to increased hyphal branching and abnormal growth (Hernández‐Rodríguez et al., 2012; Yu et al., 2017). RNA‐Seq analysis showed that the septin genes UV8b_01520, UV8b_06116 and UV8b_08253, and the majority of branching‐associated protein kinase genes were significantly downregulated in the ΔUvhos3 mutant (Table S5). These data provide an explanation for increased hyphal branches and slower growth caused by UvHos3 deletion in U. virens. The ΔRpd3 mutant in A. alternata exhibits more hyphal branches, indicating that HDACs regulate hyphal branching (Ma et al., 2021). Thus, UvHos3 deletion transcriptionally downregulates a set of genes involved in fungal branching, growth and pathogenesis, which may lead to defects in hyphal growth, conidiation and virulence. By contrast, the deletion of HosB (Hos3) in A. flavus does not affect hyphal growth, sclerotial production and aflatoxin production (Lan et al., 2019). Likewise, the Δhos3 mutant of A. alternata has no observable phenotypic changes from the wild type (Ma et al., 2021). Taken together, these findings indicate that HOS3 homologues in various fungi have different functions in regulating growth, development and virulence.

Histone acetylation often regulates secondary metabolism. Metabolomic analysis revealed that multiple types of mycotoxins, including ochratoxin B, zearalenones, zeranol, mycotoxin T2, aflatoxin B1, aflatoxin G2 (Fang et al., 2022; Janik‐Karpinska et al., 2023; Luo et al., 2022; Peng et al., 2023) and ustilaginoidin D (Wang et al., 2021) that pose a health risk to people and animals were accumulated significantly in the ΔUvhos3 mutant (Figure 4c). Interestingly, the mycotoxins ochratoxin B, zearalenone and O‐M‐sterigmatocystin are also accumulated to a higher level when the class III HDAC gene UvHst2 is deleted in U. virens (Liu, Wang, et al., 2023). The correlation analyses of the differential metabolites and DEGs in the ΔUvhos3 (Figure 8) and ΔUvhst2 mutants (Liu, Wang, et al., 2023) showed that UV8b_03320 was involved in regulating the accumulation of multiple metabolites. These results demonstrate that HDACs are important regulators of mycotoxin biosynthesis in U. virens.

In A. nidulans, H3K9 acetylation catalysed by the Saga/Ada complex is associated with the production of orsellinic acid for bacterial control (Nützmann et al., 2011). In A. flavus, the H3K14ac binding affinity and expression level of nsdD, which is important for sclerotium formation and aflatoxin production, are both reduced when the acetyltransferase gene mystB is deleted, suggesting that MystB may affect the binding of nsdD to histones by regulating the acetylation level of H3K14 (Chen et al., 2022). Interestingly, HOS3 homologues in distinct fungi have specificities to different lysine residues in histones. HOS3 in yeast has the specificity to H2AK7, H2BK11, H3K14, H3K23, H4K5 and H4K8 in vitro (Carmen et al., 1999; Gunderson et al., 2011; Trojer et al., 2003). By contrast, MrHOS3 has an obvious specificity to H3K9 and H3K18 in M. ruber (Liu, Zheng, et al., 2023). Detection of site‐specific acetylation reveals that UvHOS3 affects the acetylation levels of H3K9, H3K18, H3K27 and H4K8 in U. virens (Figure 5). Therefore, the acetylation at these sites specifically regulates expression of the genes related to growth, development, mycotoxin biosynthesis and virulence. ChIP‐qPCR assays using an anti‐H3K18 antibody showed that the promoters of the ugs and sorbicillinoid biosynthesis gene clusters were enriched in H3K18‐associated chromatin of the ΔUvhos3 mutant compared with the wild type (Figure 6e). These findings indicate that H3K18 acetylation regulated by UvHOS3 is involved in regulating ustilaginoidin and sorbicillinoid biosynthesis in U. virens.

WGCNA, which has been widely used in the analysis of disease‐related transcriptome data in humans, simplifies the interpretation of thousands of gene expression profiles to a dozen synthetic groups (or modules) and is used to identify important modules (Niemira et al., 2019; Tian et al., 2020). In the RNA‐Seq‐based weighted co‐expression gene network, we identified histone acetylation and mycotoxin biosynthesis genes in DEGs between the wild‐type and ΔUvhos3 mutant strains. One acetyltransferase gene UV8b_03374, and two acyl‐CoA related genes UV8b_01772 and UV8b_08076 were upregulated (Figure S8 and Figure 7). Acetyl‐CoA is an important substrate for HATs, and it can induce acetylation of histones at different residues, including H3K9 and H3K14 (Chen et al., 2021), substantiating that the deletion of UvHos3 alters the level of histone acetylation. In ΔUvhos3, multiple genes involved in biosynthesis of ustilaginoidins and sorbicillinoid were also upregulated (Figure 6c,d), indicating that UvHOS3 is a key negative regulator of mycotoxin biosynthesis, in particular, the production of ustilaginoidins and sorbicillinoids in U. virens (Harned & Volp, 2011; Li et al., 2019). Interestingly, correlation analysis between the DEGs in the WGCNA brown module and DAMs predicted that 13 DEGs are involved in the metabolism of 171 differentially accumulated metabolites with several types of mycotoxins such as ustilaginoidin D and ochratoxin B. In most cases, upregulated DEGs were correlated with elevated metabolites, whereas downregulated DEGs were associated with decreased metabolites (Figure 8). The analysis provides insights into the regulatory mechanisms in mycotoxin biosynthesis in filamentous fungi.

In fungi, HDACs often form as a complex to regulate gene expression. In F. graminearum, FNG3 is associated with the NUA3 HAT and FgRPD3 HDAC complexes and affects growth, conidiation, sexual reproduction and plant infection (Xu et al., 2022). SIR2/3/4 in yeast form a heterochromatin‐like structure to repress transcription. A SIR3‐HOS3 chimera can substitute for SIR2 in gene silencing and the deacetylase activity of HOS3 is required for the function of the SIR3‐HOS3 deacetylase hybrid (Chou et al., 2008). Therefore, it will be interesting to investigate whether HDACs and acetyltransferases form a complex in U. virens to synergistically regulate gene expression, which facilitates elucidating molecular mechanisms of histone acetylation in the regulation of ustilaginoidin biosynthesis.

In summary, our results demonstrated that UvHOS3 is essential for virulence, probably through positively regulating vegetative growth, conidiation and adaption to oxidative and hyperosmotic stresses, whereas the HDAC negatively modulates ustilaginoidin production in U. virens by inhibiting expression of mycotoxin biosynthesis genes. Our results deepen understanding of molecular mechanisms how the HDAC UvHOS3 and histone acetylation regulate virulence and mycotoxin biosynthesis in this important phytopathogenic fungus.

4. EXPERIMENTAL PROCEDURES

4.1. Microbial strains and culture conditions

The U. virens strain P1 was used as the wild type in this study. U. virens was cultured on PSA (decoction of 200 g potato, 20 g sucrose and 20 g agar/L) plates or in potato sucrose broth (PSB) at 28°C. For U. virens growth and sensitivity assays to TSA, the wild‐type strain was cultured on PSA plates supplemented with 1, 2, 3, 4 and 5 μM of TSA at 28°C for 7 or 14 days.

4.2. Phylogenetic analysis of HOS3 and its homologues

The amino acid sequences of HDACs in U. virens, B. bassiana, M. oryzae, Al. alternata, S. cerevisiae and A. flavus were downloaded from National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov). Phylogenetic analyses were performed using MEGA 7 based on the conserved HDAC domain sequences and maximum‐likelihood (ML) trees were constructed by the WAG+G4 model with the 1000 bootstrap replicates (Liu, Wang, et al., 2023). Protein sequences in FASTA format were used to search for conserved HDAC domains through Batch CD‐searches in NCBI website (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi).

4.3. Construction of gene‐deletion mutants and complemented strains

The UvHos3 deletion mutants were generated through homologous recombination induced by the CRISPR/Cas9 system (Li et al., 2019; Zou et al., 2023). Briefly, both flanking sequences (c. 1 kb) of the target gene were PCR‐amplified with the primers listed in Table S1. The hygromycin resistance gene was amplified from the plasmid pGKO2‐hyg. The flanking sequences of UvHos3 were then fused to 5′ and 3′ flanks of the hygromycin gene (Hyg) through fusion PCR, respectively. The CRISPR/Cas9 target site and gene‐specific sgRNA primers were designed using online tools (http://www.rgenome.net/). The synthetic sgRNA primers were annealed and inserted into the BsmBI‐digested pCAS9‐tRp (Liang et al., 2018). The resultant PCR product and pCAS9‐tRp‐gRNA construct were introduced into U. virens protoplasts by polyethylene glycol‐mediated transformation.

The gene replacement mutants were screened by PCR and were then confirmed by Southern blot analysis as described previously (Qiu et al., 2022). Briefly, genomic DNA was isolated through CTAB method. The HindIII‐digested DNA was separated on 0.7% agarose gels and blotted onto Hybond‐N⁺ membrane (GE Healthcare). The probe was amplified and labelled by DIG‐High Prime DNA Labeling and Detection Starter Kit II (Roche). The membrane was hybridized with the probe and was then visualized by NBT/BCIP chromogenic reagent.

For complementation assays, the full‐length UvHos3 gene with the 1.5‐kb native promoter was amplified by PCR and was subcloned into the vector pY2P102‐FLAG (Li et al., 2019). The construct was introduced into the ΔUvhos3‐1 mutant as described above. The complemented strains were confirmed by western blot analysis with an anti‐FLAG antibody (Sigma‐Aldrich).

4.4. Stress tolerance assays

Different U. virens strains were cultured on YT medium (1 g yeast extract, 1 g tryptone, 10 g glucose, 20 g agar and 1 L deionized water) plates supplemented with 0.5 M sorbitol, 0.03% H₂O₂, 240 μg/mL Congo Red, 1 μM TSA or 0.03% SDS at 28°C for 14 days. The colony diameter of each strain was measured for at least three plates and the growth inhibition rate was then calculated.

4.5. Conidiation assay

Five mycelial plugs (5 mm diameter) were inoculated into 100 mL of PSB and were further cultured with shaking at 150 rpm for 7 days at 28°C. The conidia produced by each strain were quantified using a haemocytometer (Paul Marienfeld GmbH & Co. KG).

4.6. Quantification of ustilaginoidins by HPLC

U. virens strains were cultured on PSA plates with cellophane at 28°C for 28 days. The mycelia (1 g) were collected and immersed in 50 mL ethyl acetate for 12 h. The extracts were condensed to dry, and were then dissolved in acetonitrile (ACN, 1 mL) followed by filtration with a 0.22‐μm filter (Lu et al., 2015). The extracts were quantified by high‐performance liquid chromatography (HPLC) on a ZORBAX Eclipse XDB‐C8 column (4.6 × 150 mm, 5 μm) with UV detector (wavelength 290 nm) using an HPLC system (LC‐20A; SHIMADZU). The extracts were eluted with ACN using the following gradient: 0–2 min, 55% ACN; 2–9 min, 55%–65% ACN; 9–10 min, 65% ACN; 10–15 min, 65%–70% ACN; 15–18 min, 70% ACN; 18–20 min, 70%–100% ACN; 20–28 min, 100% ACN; 28–30 min, 100%–70% ACN; 30–35 min, 70% ACN; 35–36 min, 70%–55% ACN; 35–36 min, 55% ACN at a flow rate of 1 mL/min. These assays were repeated at least three times.

4.7. Virulence assay

Virulence assay was performed as described previously (Li et al., 2019; Qiu et al., 2022). Briefly, the wild‐type, ΔUvhos3 and complemented CΔUvhos3 strains were cultured in PSB medium with shaking at 150 rpm at 28°C for 7 days. The culture was then smashed into a mixture of hyphae and conidiospores with a blender. Conidial suspension (1 mL, 2 × 10⁶ conidia/mL) was injected into panicles of a susceptible rice cultivar LYP9 at about 1 week before heading. At least 10 rice panicles were inoculated for each strain and were then counted for false smut balls at about 4 weeks after inoculation. These assays were independently repeated at least three times.

4.8. Metabolomics analysis

The wild‐type and ΔUvhos3‐mutant strains were cultured on PSA plates for 14 days. Accurately weighed U. virens mycelia (100 mg) were frozen in liquid nitrogen and then ground at 60 Hz for 2 min. The mixture of methanol and water (0.4 mL, 1:4 vol/vol) was added to each sample and 20 μL 2‐chloro‐L‐phenylalanine (0.3 mg/mL) dissolved in methanol was included as an internal standard. The samples were subsequently analysed by Dionex Ultimate 3000 RS UHPLC system equipped with Q‐Exactive quadrupole‐orbitrap mass spectrometer with heated electrospray ionization (ESI) source (Thermo Fisher Scientific). Quality controls were injected at regular intervals (every 10 samples) throughout the analysis to provide a set of data to assess reproducibility. Metabolomic analyses were performed with six biological repeats.

The metabolites were identified by progqenesis QI (Waters Corporation) Data Processing Software based on HMDB (http://www.hmdb.ca/), Lipidmaps (http://www.lipidmaps.org/) and self‐built databases. Principal component analysis (PCA) and (orthogonal) partial least‐squares‐discriminant analysis (O)PLS‐DA were carried out to visualize the metabolic alterations among experimental groups. The metabolites with variable importance in the projection (VIP) >1.0, p < 0.05 and absolute log₂(fold change) ≥1 were considered as differential metabolites (Li et al., 2020).

4.9. Histone acetylation assay

Total proteins were extracted from 7‐day‐old cultures with phosphate‐buffered saline containing Complete Protease Inhibitor Cocktail (Roche) and were then quantified using a BCA Protein Assay Kit (Sangon Biotech). Protein extracts (20 μg) were separated on a 10% SDS‐PAGE gel and were then electrotransferred onto PVDF membrane. The acetylation levels of H3 and H4 were detected by immunoblotting with H3K9‐, H3K18‐, H3K27‐, H4K5‐ and H4K8‐specific anti‐acetyl antibodies (Abcam). The H3, H4 and β‐actin protein levels were also probed with anti‐H3, anti‐H4 and anti‐β‐actin antibodies as protein loading controls. Band intensities were evaluated by densitometry using ImageJ (Meng, Liu, et al., 2021).

4.10. RNA isolation and RT‐qPCR

Total RNAs were isolated from 7‐day‐old U. virens cultures using an Ultrapure RNA Kit (CWBIO) and cDNAs were synthesized using a PrimeScript reverse transcription kit (TaKaRa) following the manufacturers' instructions. The transcript levels of ustilaginoidin biosynthesis‐related genes were detected by RT‐qPCR. PCR primers were designed by Primer‐BLAST (https://www.ncbi.nlm.nih.gov/tools/primer‐blast/; Table S1). RT‐qPCR was performed using Fast SYBR mixture (CWBIO, Beijing). The α‐tubulin gene was used as an internal reference. The relative gene expression levels were calculated by 2^−ΔΔCt method (Schmittgen et al., 2000).

4.11. Transcriptome analysis

For RNA‐Seq analysis, the wild‐type and ΔUvhos3‐mutant strains were cultured on PSA medium plates with cellophane at 28°C for 7 days. The mycelia were collected and frozen in liquid nitrogen. RNAs were isolated using a mirVana miRNA isolation kit (Ambion) and RNA integrity was evaluated using a 2100 Bioanalyzer (Agilent Technologies). The samples with RNA Integrity Number (RIN) ≥7 were used for construction of the libraries using a TruSeq Stranded mRNA LTSample Prep Kit (Illumina) according to the manufacturer's instructions. The libraries were then sequenced using an Illumina sequencing platform (HiSeq X Ten). RNA‐Seq analyses were performed with three biological repeats.

The gene expression levels were quantified based on normalized fragments per kilobase of transcript per million fragments mapped (FPKM). Differential expression analysis was performed using DESeq2 with the Bonferroni‐corrected p value (Q value) ≤0.05. The genes with absolute fold change values >1.5 were defined as DEGs (Cary et al., 2015). Gene ontology (GO; http://www.geneontology.org/) and Kyoto Encyclopedia of Genes and Genomics (KEGG; http://www.kegg.jp/) enrichment analyses were performed as described previously (Kanehisa et al., 2008). The GO terms and KEGG pathways enriched in DEGs were determined by Q ≤ 0.05 (Cary et al., 2015).

4.12. WGCNA and integrative analysis of metabolome and transcriptome data

The weighted gene co‐expression network was constructed using the WGCNA package in R (Langfelder & Horvath, 2008; Zhang et al., 2021). Scale‐free topology with R ² cut‐off (=0.8) was used to determine the soft‐thresholding power (&beta) of the co‐expression network. Detection of vital modules was performed with the selected power = 17 and major parameters as maxBlockSize = 5000, minModuleSize = 30, networkType = unsigned and mergeCutHeight = 0.25. Module‐trait associations were assessed based on the correlation between the module eigenvector and trait. p values were adjusted for multiple testing using the Benjamini–Hochberg (i.e., false discovery rate, FDR) approach (Jaime‐Lara et al., 2020; Zeng et al., 2021). The vital module was chosen by the correlation of module–trait value ≥|0.5| and p < 0.05. Pearson correlation coefficients (PCC) and corresponding p values (PCCP) were used to screen differential metabolites and related genes using combined metabolomic and transcriptomic analysis. The screening criteria were PCC > |0.99| and PCCP < 0.05. The network of the model with the significant correlation was visualized with the Gephi v. 0.9.7 (Amith et al., 2019).

4.13. ChIP analysis and qPCR

ChIP assay was performed with a ChIP Kit K308 (IEMED Guangzhou Biomedical Technology Co., Ltd) following the manufacturer's instructions. The proteins and DNA from 7‐day‐old cultures of the wild‐type and ΔUvhos3 strains were cross‐linked with 1% formaldehyde for 10 min followed by addition of 150 mM glycine to stop cross‐linking. The samples were centrifuged at 94 g at room temperature to remove the supernatant. The pellet was resuspended and well mixed with immunoprecipitation lysis buffer C and protease inhibitors. Genomic DNA was fragmented into 200–500 bp using a sonication machine Diagenode Bioruptor in cycles for 5 s ‘on’ and 5 s ‘off’ at 60% for 20–30 min. Subsequently, the binding enhancer buffer was added and was subject to centrifugation at 12,000 g at 4°C for 10 min. The supernatants were mixed with clean beads and were then incubated at room temperature for 20 min. After pretreating with protein A/G beads and AB binding buffer, the supernatant was incubated with anti‐acetyl‐H3K18 (Abcam; A7257) or anti‐IgG on a rotary shaker at room temperature for 1 h. The beads were then rinsed with washing buffer three times. Finally, immunoprecipitated DNA was used as template for ChIP‐qPCR.

ChIP‐qPCR was performed using Taq Pro Universal SYBR qPCR Master Mix (Vazyme) by LightCycler96 (Roche). The PCR primers specific for the promoters of ustilaginoidin and sorbicillinoid biosynthesis‐related genes were designed and listed in Table S1.

4.14. Statistical analysis

For each assessment, at least three independent repeats were performed unless noted. All values were shown as mean ± SD. All histograms were graphed with GraphPad Prism 7. Statistical significance was determined through one‐way analysis of variance followed by Duncan's multiple range test (p < 0.05; SPSS Inc.) or Student's t test.

CONFLICT OF INTEREST STATEMENT

The authors declare no potential conflict of interest.

Supporting information

Figure S1. The histone deacetylase inhibitor trichostatin inhibits mycelial growth and elevates ustilaginoidin biosynthesis in Ustilaginoidea virens. (a, b) Colony diameters of U. virens P1 after 7 days (a) and 14 days (b) of growth on potato sucrose agar plates with various concentrations of trichostatin A (TSA). (c) The relative intensities of HPLC peaks representing the major ustilaginoidin derivatives including ustilaginoidin E, ustilaginoidin D and ustilaginoidin M. Data from three technical repeats are shown as mean ± SD (n = 3) and one representative image is shown in Figure 1. Different letters (a–d) represent statistically significant differences in colony diameter according to one‐way analysis of variance followed by Duncan’s multiple range test (p < 0.05).

Figure S2. The phylogenetic tree and structural domains of putative histone deacetylases in different fungal species. (a) Phylogenetic tree of putative histone deacetylases (HDACs) in different fungal species including Ustilaginoidea virens, Beauveria bassiana, Magnaporthe oryzae, Alternaria alternata, Saccharomyces cerevisiae and Aspergillus flavus. (b) The conserved domains in the putative HDACs in the indicated fungal species. Different domains are indicated by various shades and colours.

Figure S3. Screening and confirmation of the UvHos3‐knockout mutants and complemented strains of Ustilaginoidea virens. (a) A schematic diagram to show endonuclease digestion and probe design for southern blot analysis. (b) Screening of the UvHos3‐knockout mutant candidates via PCR. Gene, Hyg, Up and Down indicate bands amplified from the target gene, Hyg gene, upstream and downstream fragments respectively. Numbers in red represent putative knockout transformants identified by PCR. WT, the wild‐type strain P1. (c) Southern blot analysis was performed to confirm the ΔUvhos3 deletion mutants. M, marker; WT, the wild‐type strain P1. Numbers in red represent the confirmed deletion mutants. (d) Western blot analysis was performed to confirm the ΔUvhos3 complemented strains. The CΔUvhos3 complemented strains were screened by immunoblotting probed with an anti‐FLAG antibody. Numbers in red represent confirmed complemented strains. The deletion mutant was used as negative control.

Figure S4. UvHos3 is required for mycelial growth and virulence. (a) The colony phenotypes of the wild‐type (P1), ΔUvhos3 mutant and complemented CΔUvhos3‐2 strains cultured on potato sucrose agar plates for 7 days. (b) Virulence assay to rice of the wild‐type, mutants and complemented strains. Disease symptoms were photographed at about 4 weeks after Ustilaginoidea virens inoculation.

Figure S5. UvHos3 differentially regulates distinct external stresses and negatively modulates ustilaginoidin biosynthesis. (a) The wild‐type (WT), mutant and complemented strains were cultured on yeast extract‐tryptone (YT) plates and YT plates supplemented with 0.5 M sorbitol, 0.03% H₂O₂, 240 μg/mg Congo Red, 1 μM trichostatin A (TSA) and 0.03% SDS. The cultures were photographed after culturing for 14 days at 28°C. (b) The relative intensities of HPLC peaks representing the major ustilaginoidin derivatives including ustilaginoidin E, ustilaginoidin D and ustilaginoidin M generated in the wild‐type, mutant and complemented strains. Data from three technical repeats are shown as mean ± SD (n = 3) and one representative image is shown in Figure 3. Different letters (a–c) represent statistically significant differences in growth inhibition rate in (a) or in ustilaginoidin amount in (b) according to one‐way analysis of variance followed by Duncan’s multiple range test (p < 0.05).

Figure S6. RNA‐Seq analysis of the Ustilaginoidea virens wild‐type and ΔUvhos3 mutant strains. (a) A volcano map to exhibit transcriptome data. Red, green and grey circles represent upregulated, downregulated and non‐differentially expressed genes (non‐DEGs) respectively (p < 0.05). (b) The top 20 enriched KEGG pathways for DEGs revealed by KEGG pathway enrichment analyses. The bubble size represents the number of genes in a specific KEGG pathway, and colour bar indicates p values. (c) The 30 enriched Gene Ontology terms for DEGs revealed by GO enrichment analyses. The horizontal axis represents different GO Terms. Green, blue and red columns represent biological processes, cellular components and molecular functions respectively. The ordinate represents −log₁₀(p value).

Figure S7. A pathway map of the mannose type O‐glycan biosynthesis. The KEGG pathway diagram shows downregulated genes. The downregulated genes UV8b_01319, UV8b_02987 and UV8b_05916 in the green box related to broth growth and pathogenesis were enriched in mannose type O‐glycan biosynthesis pathway.

Figure S8. Cluster dendrogram and module–trait relationships by weighted gene co‐expression network analysis (WGCNA). (a) Cluster dendrogram showing the module–gene relationships generated by WGCNA. Each branch represents a gene, and each colour represents a co‐expression module. (b) A heatmap illustrating module–trait relationships between module eigengenes (MEs) and the trait of wild‐type (WT) and ΔUvhos3 strains. Each module contains correlation coefficient between the gene module and the corresponding trait and p value in parentheses. The strength of the correlation is depicted by its colour.

Table S1. PCR primers used in this study.

Table S2. Differential phenylpropanoid and polyketide metabolites between the wild‐type and ΔUvhos3‐mutant strains.

Table S3. Quality of sequencing data after preprocessing.

Table S4. Genes related to growth in RNA‐Seq data.

Table S5. The genes associated with hyphal branching as revealed by RNA‐Seq.

Table S6. Genes related to pathogenesis in RNA‐Seq data.

Table S7. Genes associated with mycotoxin synthesis in RNA‐Seq data.

Table S8. The upregulated and downregulated genes are shown in Figure 7.

Table S9. The differentially accumulated metabolites associated with upregulated genes in Figure 8.

Table S10. The differentially accumulated metabolites associated with downregulated genes in Figure 8.

Wang, B. , Duan, G. , Liu, L. , Long, Z. , Bai, X. , Ou, M. et al. (2024) UvHOS3‐mediated histone deacetylation is essential for virulence and negatively regulates ustilaginoidin biosynthesis in Ustilaginoidea virens . Molecular Plant Pathology, 25, e13429. Available from: 10.1111/mpp.13429

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available within this article and its supplements. Sequences are available at NCBI GenBank (www.ncbi.nlm.nih.gov/) with accession numbers UvRpd3 (XP_042994594.1) and UvHos2 (XP_042997993.1), UvHda1 (XP_042998499.1) and UvHos3 (XP_042994466.1), UvHst4 (XP_042995395.1), UvSir2 (XP_042999091.1) and UvHst2 (XP_042995294.1).