An Ustilaginoidea virens glycoside hydrolase 42 protein is an essential virulence factor and elicits plant immunity as a PAMP

Abstract

Rice false smut, caused by the ascomycete fungus Ustilaginoidea virens, which infects rice florets before heading, severely threatens rice grain yield and quality worldwide. The U. virens genome encodes a number of glycoside hydrolase (GH) proteins. So far, the functions of these GHs in U. virens are largely unknown. In this study, we identified a GH42 protein secreted by U. virens, named UvGHF1, that exhibits β‐galactosidase activity. UvGHF1 not only functions as an essential virulence factor during U. virens infection, but also serves as a pathogen‐associated molecular pattern (PAMP) in Nicotiana benthamiana and rice. The PAMP activity of UvGHF1 is independent of its β‐galactosidase activity. Moreover, UvGHF1 triggers cell death in N. benthamiana in a BAK1‐dependent manner. Ectopic expression of UvGHF1 in rice induces pattern‐triggered immunity and enhances rice resistance to fungal and bacterial diseases. RNA‐seq analysis revealed that UvGHF1 expression in rice not only activates expression of many defence‐related genes encoding leucine‐rich repeat receptor‐like kinases and WRKY and ERF transcription factors, but also induces diterpenoid biosynthesis and phenylpropanoid biosynthesis pathways. Therefore, UvGHF1 contributes to U. virens virulence, but is also recognized by the rice surveillance system to trigger plant immunity.

UvGHF1, as a GH42‐type glycosyl hydrolase secreted by Ustilaginoidea virens, contributes to U. virens virulence, whereas it is recognized by the rice surveillance system to trigger plant immunity.

1. INTRODUCTION

Rice false smut caused by Ustilaginoidea virens (telemorph Villosiclava virens) is a prevalent and destructive disease in the majority of rice‐growing regions. U. virens primarily infects stamen filaments and hijacks host nutrients through mimicking fertilization in rice florets (Sun et al., 2020). In addition, U. virens produces a variety of toxic secondary metabolites in false smut balls, such as ustiloxin and ustilaginoidin derivatives, that impose deleterious effects on human and animal health (Li et al., 2019; Nakamura et al., 1994; Wang, Liu, et al., 2021). Few effective management methods have been developed in agricultural production to control this disease so far (Lu et al., 2018; Sun et al., 2020). Thus, understanding of the molecular mechanisms underlying U. virens virulence and pathogenicity may provide novel strategies for the prevention and control of false smut in rice.

The plant cell wall is a natural physical barrier that fulfils diverse cellular functions ranging from maintenance of structural integrity to defence against the invasion of external pathogens (Zhang, Gao, et al., 2020). More than 90% of the plant cell wall is made up of carbohydrates with cellulose, hemicelluloses, and pectic polysaccharides (Popper et al., 2010). Previous studies have shown that phytopathogenic fungi and oomycetes can secrete glycoside hydrolases (GHs) to degrade host cell walls (Wang et al., 2019). Secreted GHs in these organisms have multiple functions, with many of them acting as virulence factors (effectors) to promote infection and obtain nutrients. For example, the tomatinase FoTom1 in the GH10 family is essential for virulence of Fusarium oxysporum f. sp. lycopersici to tomato (Pareja‐Jaime et al., 2008). The tomatinase CfTom1 from Cladosporium fulvum degrades α‐tomatine into tomatidine in planta and is also required for fungal virulence (Ökmen et al., 2013). Secreted chitinase‐like GH18 proteins from Moniliophthora pernicosa and Moniliophthora roreri also play an important role in inhibiting plant immunity (Fiorin et al., 2018; Yang et al., 2021). LbGH28 with pectinase activity from Laccaria bicolor is present at hyphal tips within the Hartig net, which is related to nutrient exchange and plays an essential role in plant–fungus symbiosis (Zhang, Hua, et al., 2021).

Besides functioning as virulence factors, multiple secreted GHs of phytopathogenic fungi and oomycetes can also activate plant immunity. Knowledge on GH proteins that trigger plant immune responses is mainly obtained from the GH11 and GH12 families. For instance, VdEIX3, a secreted GH11 protein from Verticillium dahliae, triggers cell death and other immune responses (Bailey et al., 1993; Yin et al., 2021). BcXyn11A, a GH11 protein with β‐1,4‐endoxylanase activity from Botrytis cinerea, elicits defence responses (Bailey et al., 1993; Frías et al., 2019; Noda et al., 2010). Similarly, a wide range of GH12 proteins identified from plant‐associated oomycetes and fungi are capable of triggering cell death when transiently expressed in plants. XEG1, a GH12 protein with xyloglucanase and β‐glucanase activities in Phytophthora sojae, is one of the best‐studied effectors (Ma et al., 2015, 2017). XEG1 induces strong necrosis and defence responses in the model plant Nicotiana benthamiana and the host soybean (Ma et al., 2017; Xia et al., 2020). The GH12 transglycosylases VdEG1 and VdEG3, originally identified in the secretome of V. dahliae‐infected cotton leaves, trigger immune responses (Gui et al., 2017). FoEG1, a secreted GH12 protein from F. oxysporum with hydrolytic activity towards carboxymethylcellulose, functions in the plant apoplast to induce cell death (Zhang, Yan, et al., 2021). BcCrh1, a GH12 protein characterized in B. cinerea, triggers cell death in N. benthamiana and tomato (Bi et al., 2021). Interestingly, these GH proteins, in most cases, contribute to pathogen virulence.

In plants, pattern recognition receptors (PRRs) on the plasma membrane monitor conserved pathogen/microbe‐associated molecular patterns (PAMPs/MAMPs) or plant‐derived damage‐associated molecular patterns (DAMPs) to activate pattern‐triggered immunity (Ngou et al., 2021; Yuan et al., 2021). The defence‐triggered GH proteins either act as PAMPs/MAMPs or degrade the cell wall, whose hydrolysis products can be recognized as DAMPs (Bradley et al., 2022; Cook et al., 2015). The ability of CfGH17‐1/5 and MoCel12A/B to induce cell death depends upon their enzymatic activities, implying that their enzymatic hydrolysates might be perceived as DAMPs (Ökmen et al., 2019; Yang et al., 2021). By contrast, many GH proteins, such as PxEG1, VdEG1, BcXYG1, FoEG1, and VdCUT11, might function as PAMPs. The ability of these GH proteins to induce cell death depends on both of the coreceptors BAK1 and SOBIR1 but is independent of their enzymatic activity (Gui et al., 2017, 2018; Ma et al., 2017; Zhang, Yan, et al., 2021). A few GH proteins, such as VdEG3, trigger plant immunity that is only dependent on BAK1 but is independent of SOBIR1 in N. benthamiana (Gui et al., 2017). Collectively, these findings indicate that various GH proteins trigger plant immunity through different mechanisms.

Besides the GH10, GH11, GH12, GH18, and GH28 proteins, many other GH family members are secreted by fungi (Bradley et al., 2022). The GH42 enzymes are ubiquitous in microbes. At present, studies on GH42 proteins focus on the detection of their enzyme activity, but little is known on their function in plant–microbe interactions (Godoy et al., 2016). In U. virens UV‐8b, 117 secreted GH proteins, including only one GH42 protein, have been predicted to be candidate effectors (Sun et al., 2020; Zhang et al., 2014; Zhang, Zhao, et al., 2021). In this study, we identified the GH42 family protein secreted by U. virens, named UvGHF1, which can trigger plant cell death and defence responses. We further demonstrated the importance of UvGHF1 in U. virens virulence. It is a novel virulence factor that has β‐galactosidase activity. Identification and functional characterization of the secreted UvGHF1 protein is pivotal to our understanding of how UvGHF1 regulates host immunity and contributes to fungal virulence to rice.

2. RESULTS

2.1. UvGHF1 is an apoplastic effector in U. virens

UvGHF1, with an open reading frame of 1020 bp, encodes a GH42 domain‐containing protein with 339 amino acid (aa) residues. The first 16‐aa region of UvGHF1 is a putative signal peptide (SP) (Figure S1). Consequently, UvGHF1 secretion was tested through a yeast signal trap assay. The construct to express SP^UvGHF1‐SUC2 was generated by cloning the sequence encoding the SP of UvGHF1 into the vector pSUC2 and was then transformed into the yeast suc2 mutant. The SP of UvGHF1 was sufficient for the secretion of invertase in yeast because the transformants grew well on YPR agar (YPRAA) plates (Figure 1a). In addition, the activity of secreted invertase was also confirmed by the reduction of 2,3,5‐triphenyltetrazolium chloride (TTC) to insoluble red 1,3,5‐triphenylformazan (TPF) in liquid medium (Figure 1a). Thus, the yeast secretion assay showed that UvGHF1 is a secreted protein.

FIGURE 1.

UvGHF1 is a secretory protein produced by Ustilaginoidea virens. (a) Functional validation of the signal peptide (SP) of UvGHF1 by a yeast invertase secretion system. The yeast strain transformed with pSUC2‐UvGHF1 ^SP grew on YPRAA plates. When the UvGHF1^SP‐SUC2 invertase was secreted by yeast cells into liquid medium, 2,3,5‐triphenyltetrazolium chloride (TTC) was reduced into the red precipitate 1,3,5‐triphenylformazan (TPF). YTK12, untransformed yeast strain; Avr1b^SP (the SP of Phytophthora sojae effector Avr1b) and Mg87^N25 (the first 25 amino acid residues of the nonsecreted Mg87 protein in Magnaporthe oryzae) were expressed to guide SUC2 secretion as positive and negative controls, respectively. (b) UvGHF1‐mCherry and UvGHF1^ΔSP‐mCherry were localized to the cytoplasm and plasma membrane, respectively, when these proteins were transiently expressed in Nicotiana benthamiana leaves. (c) UvGHF1‐mCherry, but not UvGHF1^ΔSP‐mCherry, was localized in periplasmic spaces of N. benthamiana cells after plasmolysis. Asterisks indicate periplasmic spaces after N. benthamiana cells were plasmolysed. The leaves were detached and treated with 1 M NaCl to induce plasmolysis. Images were captured using confocal microscopy at 2 days after infiltration. Scale bars, 10 μm. (d) Green fluorescence from UvGHF1‐GFP and MoSLP1‐GFP was observed in the extra‐invasive hyphal membrane (EIHM) during M. oryzae infection. Rice sheaths were inoculated with the M. oryzae strains expressing UvGHF1‐GFP or MoSLP1‐GFP, an identified apoplastic effector. Images were taken at 36–40 h postinoculation (hpi). The arrowheads indicate invasive hyphae (IH). Scale bars, 15 μm.

To explore the subcellular localization of UvGHF1, UvGHF1‐mCherry and UvGHF1^ΔSP‐mCherry (lacking the SP) were transiently expressed under the control of the CaMV 35S promoter in N. benthamiana leaves. Red fluorescence from UvGHF1‐mCherry and UvGHF1^ΔSP‐mCherry was observed on the plasma membrane and in the cytoplasm (Figure 1b). Interestingly, when N. benthamiana cells were plasmolysed, UvGHF1‐mCherry was localized mainly in periplasmic spaces, whereas UvGHF1^ΔSP‐mCherry was retained on the plasma membrane and in the cytoplasm (Figure 1c). To further confirm the apoplastic location of UvGHF1, transiently expressed UvGHF1‐GFP in N. benthamiana was detected in the apoplastic fluid by immunoblotting, but UvGHF1^ΔSP‐GFP was not detected (Figure S2). Furthermore, the translocation system of the fungal blast pathogen Magnaporthe oryzae was exploited to investigate the secretion of UvGHF1‐GFP (Khang et al., 2010). At 36–40 h after inoculation, UvGHF1‐GFP was observed in the extra‐invasive hyphal membrane (EIHM) compartment after rice sheaths were inoculated with the transformed M. oryzae strain expressing UvGHF1‐GFP. As a positive control, the apoplastic effector MoSLP1‐GFP was also observed to be secreted through the EIHM structure (Figure 1d). Taken together, these results indicate that UvGHF1 is secreted into host apoplastic spaces during infection.

2.2. UvGHF1 triggers cell death independent of its enzymatic activity

To determine the ability of UvGHF1 to induce plant cell death, UvGHF1‐FLAG was transiently expressed in N. benthamiana leaves through Agrobacterium‐mediated gene expression. Similar to the positive control BAX, UvGHF1 induced an evident necrosis symptom, whereas green fluorescent protein (GFP) did not induce cell death in N. benthamiana (Figure 2a). Interestingly, expression of the truncated UvGHF1 without SP no longer caused cell death in N. benthamiana, suggesting that UvGHF1 might act in the apoplast to induce plant immunity.

FIGURE 2.

UvGHF1 induces cell death in Nicotiana benthamiana in a manner that is independent of its enzymatic activity. (a) Necrosis symptoms on N. benthamiana leaves caused by BAX, UvGHF1, truncated UvGHF1 lacking the signal peptide (UvGHF1^ΔSP), UvGHF1^E165A, or UvGHF1^E285A. UvGHF1 and its variant proteins were transiently expressed in N. benthamiana leaves via Agrobacterium‐mediated transient expression. BAX and green fluorescent protein (GFP) were expressed as positive and negative controls, respectively. The assay was independently repeated at least three times. The numbers below the images indicate the cell death ratio and the number of technical repeats. (b) Expression of different FLAG‐tagged proteins in the agro‐infiltrated N. benthamiana leaves was detected by immunoblotting with an anti‐FLAG antibody. Total protein loading is indicated by Ponceau S staining. (c) The severity of cell death was quantified by ion leakage assays. Conductivity was measured for five leaf discs collected from five independent plants before and after boiling. Ion leakage was calculated as the ratio of conductivity before and after boiling. Data from three independent experiments are shown as mean ± SD (n = 5). Different letters (a vs. b) indicate significant differences in ion leakage in N. benthamiana leaves (p < 0.01; Duncan's multiple range test).

BLAST searches revealed that UvGHF1 contains a typical GH42 domain and has homologues in a variety of fungi, including Beauveria bassiana, Purpureocillium lilacinum, Trichoderma longibrachiatum, Metarhizium anisopliae, and Hirsutella minnesotensis, and in the bacterial species Paludibacterium purpuratum (Figure S3a). The GH42 protein has two conserved catalytically active sites that are crucial for its hydrolase activity (Godoy et al., 2016). Multiple sequence alignment revealed that UvGHF1 contains the putative catalytic residues Glu165 and Glu285, which are highly conserved in all UvGHF1 homologues (Figure S3b). Here, we demonstrated that in vitro purified UvGHF1 has β‐galactosidase activity, whereas point mutations of the conserved Glu165 and Glu285 residues in the catalytic pocket of UvGHF1 caused the protein to lose β‐galactosidase activity (Figure S4a,b). We tested whether the GH enzymatic activity of UvGHF1 is required for its ability to induce cell death, and we found that the point‐mutated proteins UvGHF1^E165A and UvGHF1^E285A retained the ability to trigger cell death when these mutated proteins were expressed in N. benthamiana (Figure 2a,b). Ion leakage is generally coupled to cell death (Mihailova et al., 2018). We demonstrated that cell death symptoms were highly correlated with the degree of electrolyte leakage in N. benthamiana leaves (Figure 2c). These results indicate that the ability of UvGHF1 to induce cell death in N. benthamiana depends on its SP but is independent of its enzymatic activity.

2.3. UvGHF1 elevates pattern‐triggered immunity and resistance against rice diseases

To investigate whether UvGHF1 positively regulates rice immunity, the transgenic rice lines UvGHF1‐IE6 (IE6) and UvGHF1‐IE27 (IE27) with dexamethasone (DEX)‐induced expression of UvGHF1‐FLAG were generated and confirmed via immunoblotting (Figure S5). First, PAMP‐induced immune responses were detected in the wild type and IE6 and IE27 transgenic plants. The reactive oxygen species (ROS) burst in the DEX‐treated IE6 and IE27 transgenic lines was much stronger than that in the wild type and mock‐treated transgenic plants after undergoing chitin and flg22 treatments (Figure 3a–d). Likewise, expression of the defence marker genes OsPR1b and OsPR10 was significantly elevated in the DEX‐treated IE6 and IE27 transgenic seedlings compared with that in the wild type and mock‐treated transgenic seedlings in response to both chitin and flg22 treatments (Figure 3e). These results indicate that UvGHF1 enhances chitin‐ and flg22‐induced immunity in rice.

FIGURE 3.

UvGHF1 elevates PAMP‐triggered immunity in rice. (a–d) Dexamethasone (DEX)‐induced expression of UvGHF1 enhanced reactive oxygen species (ROS) burst triggered by flg22 (10 μM) and chitin (10 μg/mL) in transgenic rice plants. Oxidative burst was detected for 25 min immediately after PAMP treatments. Data from three independent assays are shown as mean ± SD from three technical repeats with nine leaf discs. (e) UvGHF1 expression elevated the transcript levels of the defence genes PR1b and PR10 triggered by flg22 (10 μM) and chitin (10 μg/mL) in transgenic rice plants. The expression levels of PR1b and PR10 were normalized to that of OsActin. Data are shown as mean ± SD (n = 3). Different letters (a vs. b) indicate significant differences in the gene expression levels (p < 0.01; Duncan's multiple range test).

Next, the transgenic lines were subjected to U. virens inoculation assays. False smut balls on the U. virens‐inoculated panicles of DEX‐treated transgenic rice lines were significantly fewer than those in the wild type and mock‐treated transgenic plants (Figure 4a,b). Furthermore, the wild type and transgenic rice lines after DEX and mock treatments were challenged with the bacterial blight pathogen Xanthomonas oryzae pv. oryzae or with M. oryzae. The disease lesions on the X. oryzae‐inoculated leaves of the DEX‐treated IE6 and IE27 transgenic lines were significantly shorter than those of the wild type and mock‐treated transgenic plants (Figure 4c,d). In addition, the DEX‐treated transgenic rice lines exhibited significantly smaller disease lesions and less fungal biomass on the M. oryzae‐inoculated leaves than did the wild type and mock‐treated transgenic plants (Figure 4e,f). Altogether, these results demonstrated that ectopic expression of UvGHF1 in rice induced plant resistance against both bacterial and fungal pathogens.

FIGURE 4.

Ectopic expression of UvGHF1 increases rice resistance against fungal and bacterial pathogens. (a, b) Disease resistance of wild‐type and transgenic rice plants to Ustilaginoidea virens. The wild type Nipponbare (Nip) and transgenic rice lines (IE6, IE27) were treated with dexamethasone (DEX, 10 μM) and mock solutions for 16 h before injection inoculation of U. virens. Disease symptoms (a) and the number of false smut balls per panicle (b) were recorded at 28 days postinoculation (dpi). Data are shown as mean ± SD (n = 10). (c, d) Disease resistance of wild‐type and transgenic rice plants to the bacterial blight pathogen Xanthomonas oryzae pv. oryzae. Typical disease symptoms were photographed (c) and at least 10 leaves were measured for disease lesion length at 14 dpi (d). Data from three independent experiments are shown as mean ± SD (n = 10). (e, f) Disease resistance of wild‐type and transgenic rice plants to M. oryzae. Disease symptoms (e) and relative fungal biomass (f) on the rice leaves after spraying inoculation of M. oryzae. Representative disease symptoms were photographed at 7 dpi. Fungal biomass was evaluated by DNA‐based quantitative PCR to detect the M. oryzae MoPot2 DNA level normalized to the rice OsUbi gene. Data from three independent experiments are shown as mean ± SD (n = 3). Asterisks indicate significant differences according to Student's t test (*p < 0.05, **p < 0.01, ***p < 0.001).

2.4. UvGHF1 regulates the expression of defence signalling genes

To understand the molecular mechanism by which UvGHF1 triggers rice immunity, transcriptome profiling of DEX‐treated wild‐type and transgenic rice plants was performed through RNA‐seq. The genes with log₂(fold change) ≥ ∣1∣ and p < 0.05 were categorized as differentially expressed genes (DEGs). A total of 4823 (2920 up‐regulated and 1903 down‐regulated) and 4957 (3116 up‐regulated and 1841 down‐regulated) DEGs were identified in transcriptomic comparisons of IE6 versus wild type and IE27 versus wild type, respectively (Figure 5a and Figure S6a). A total of 2540 up‐regulated and 1643 down‐regulated genes overlapped between the DEGs of IE6 versus wild type and IE27 versus wild type. Volcano plots visually showed the up‐regulated and down‐regulated DEGs between different genotypes (Figure 5b). Principal component analysis (PCA) showed that transcriptome data from different samples of two transgenic lines (IE6 and IE27) were grouped together, but were clearly separated from those of the wild‐type plants (Figure S6b), confirming the reproducibility of the experiment and the reliability of our transcriptome data. Next, Gene Ontology (GO) and KEGG pathway enrichment analyses were performed for the common DEGs in both transgenic lines. GO term classification showed that defence‐related GO terms such as response to fungus, terpenoid metabolic process, hydrogen peroxide metabolic process, terpenoid biosynthetic process, hormone‐mediated signalling pathway, and response to salicylic acid were among the most abundant 15 terms (Figure S7a). Moreover, KEGG pathway enrichment analysis showed that DEGs were also enriched in some defence signalling pathways, including biosynthesis of secondary metabolites, metabolic pathways, diterpenoid biosynthesis, phenylpropanoid biosynthesis, plant–pathogen interaction, MAPK signalling pathway ‐ plant, and cutin, suberine, and wax biosynthesis (Figure S7b). More specifically, the majority of DEGs related to terpenoid backbone biosynthesis and phenylalanine, tyrosine, and tryptophan biosynthesis were significantly up‐regulated (Figure 5c and Figure S8). In particular, the expression of four genes, namely OsGA2ox4 and OsCYP76M14, which are involved in diterpenoid biosynthesis, and OsPOX29 and OsPOX5.1, which are involved in phenylpropanoid biosynthesis, was most significantly enhanced in the UvGHF1‐expressing transgenic lines (Figure 5c and Figure S8). Diterpenoid phytoalexins, such as oryzalexin D, and phenylpropanoid‐derived compounds represent a wide array of secondary metabolites contributing to plant responses towards biotic and abiotic stresses (Yamane, 2013). We further found that up‐regulated genes involved in plant resistance were mainly enriched in leucine‐rich repeat (LRR) receptor‐like protein kinases, WRKY transcription factors, and ERF transcription factors. A heatmap analysis was conducted to present a general overview of the expression patterns of these three classes of up‐regulated and down‐regulated genes (Figure 5d–f). To validate the reliability of DEGs obtained from RNA‐seq analyses, the expression levels of six candidate genes were analysed using reverse transcription‐quantitative PCR (RT‐qPCR). The RNA‐seq data were consistent with the RT‐qPCR results (Figure 5g). Interestingly, we demonstrated that the six genes except OsERF102 were transcriptionally up‐regulated in rice during U. virens infection (Figure S9), substantiating that these DEGs are involved in rice immune signalling and the response to U. virens. Taken together, these analyses indicate that UvGHF1 activates the defence signalling pathways and immunity in rice.

FIGURE 5.

Comparative transcriptome analysis between wild‐type (WT) and UvGHF1‐expressing transgenic rice lines. (a) The number of differentially expressed genes (DEGs) between WT and transgenic IE6 or IE27 rice lines. (b) Volcano plots of DEGs in IE6/WT and IE27/WT comparisons. Red and green colours represent the up‐regulated and down‐regulated genes, respectively, and grey represents non‐DEGs. The genes with log₂(fold change) ≥ ∣1∣ and p < 0.05 were identified as DEGs. (c) Schematic diagram of expression profiles of genes involved in diterpenoid biosynthesis. Cubes represent the expression levels in the WT, IE6, and IE27 lines, with purple representing high fold change values and green indicating low values (Table S2). (d–f) Heatmaps showing the expression levels of LRR protein‐encoding genes (d), genes encoding WRKY transcription factors in six clades (e), and genes encoding ERF transcription factors in 10 subgroups (f). A colour bar is presented at the top right, and the colours from purple to green indicate high to low fold change values (Table S2). (g) Expression of six DEGs in the WT and transgenic rice lines was evaluated using reverse transcription‐quantitative PCR. The OsActin gene was used as an internal control. Data are shown as mean ± SD (n = 3). Asterisks (***) indicate significant differences in gene expression level between the wild type and transgenic rice lines (***p < 0.001, Student's t test).

2.5. UvGHF1 triggers BAK1‐mediated cell death in N. benthamiana

BAK1 and SOBIR1 function as coreceptors of multiple PRRs in plants (Cook et al., 2015). To investigate whether the ability of UvGHF1 to trigger plant immunity is dependent on the two coreceptors, NbBAK1‐ and NbSOBIR1‐silenced plants were obtained using tobacco rattle virus (TRV)‐mediated virus‐induced gene silencing (VIGS) (Figure S10). RT‐qPCR assays revealed that the transcript levels of NbBAK1 and NbSOBIR1 were significantly decreased by VIGS at 3 weeks after agro‐infiltration (Figure 6a,b). VIGS efficiency was also confirmed by silencing the phytoene desaturase (PDS) gene, which led to a photobleached phenotype in N. benthamiana (Figure S10). The NbBAK1‐ and NbSOBIR1‐silenced plants were then agro‐infiltrated with Agrobacterium strains carrying the constructs to express UvGHF1, BAX, or GFP. The results showed that UvGHF1 induced cell death on the leaves of NbSOBIR1‐silenced plants, but not on the leaves of NbBAK1‐silenced plants (Figure 6c). By contrast, BAX retained the ability to induce cell death in NbBAK1‐ and NbSOBIR1‐silenced N. benthamiana leaves (Figure 6c). Immunoblot analyses showed that UvGHF1 was well expressed in N. benthamiana leaves inoculated with TRV::BAK, TRV::SOBIR1, or TRV::mCherry (Figure 6d,e). Therefore, the results indicate that UvGHF1‐triggered cell death is dependent on NbBAK1 but not on NbSOBIR1 in N. benthamiana.

FIGURE 6.

UvGHF1 relies on NbBAK1 to trigger cell death in Nicotiana benthamiana. (a, b) Reverse transcription‐quantitative PCR assays to show the silencing efficiency of the immune receptor genes NbBAK1 (a) and NbSOBIR1 (b) in N. benthamiana. The Agrobacterium strain carrying TRV:mCherry was infiltrated into N. benthamiana leaves as a mock control. NbActin was used as an internal reference gene. Data from three biological replicates are shown as mean ± SD (n = 3). Asterisks (***) indicate significant differences in relative gene expression level (***p < 0.001, Student's t test). (c) The ability of UvGHF1 to trigger cell death in NbBAK1‐ or NbSOBIR1‐silenced plants. Virus‐induced gene silencing was performed by infiltration with Agrobacterium strains carrying TRV:NbBAK1 or TRV:NbSOBIR1. The TRV and TRV:mCherry constructs were used as mock controls. Transient expression of BAX and GFP in the gene‐silenced leaves was used as positive and negative controls, respectively. The necrosis phenotypes were photographed at 5 days after agro‐infiltration. The assay was independently repeated at least three times. The numbers on the images indicate the percentage of representative symptoms and the number of technical repeats. (d, e) Immunoblot analysis of UvGHF1‐FLAG transiently expressed in NbBAK1‐silenced (d) or NbSOBIR1‐silenced (e) N. benthamiana leaves. Total protein loading is indicated by Ponceau S staining.

2.6. UvGHF1 is required for virulence of U. virens to rice

Next, we investigated the expression pattern of UvGHF1 during infection. The susceptible rice cultivar LYP9 was inoculated with U. virens. RT‐qPCR showed that UvGHF1 was slightly up‐regulated at 1–3 days postinoculation (dpi) and exhibited a sharp increase at 4 dpi. The expression level peaked with an approximately 18‐fold increase at 5 dpi and was gradually reduced thereafter (Figure 7a). The up‐regulation pattern of UvGHF1 at the early stage of U. virens infection suggests its importance in U. virens infection.

FIGURE 7.

UvGHF1 is required for virulence of Ustilaginoidea virens to rice. (a) Reverse transcription‐quantitative PCR assay to show the expression pattern of UvGHF1 during U. virens infection. The expression levels of UvGHF1 were detected at the indicated time points after U. virens was inoculated into panicles of a susceptible rice variety (LYP9). The α‐tubulin gene was used as an internal reference. Data are shown as mean ± SD (n = 3). (b, c) The average number of diseased grains (b) and disease symptoms (c) on rice panicles after U. virens inoculation. False smut balls were counted for 10 panicles per strain and were photographed at 28 days postinoculation with the U. virens wild type (P1), the knockout mutant Δuvghf1‐2, and complemented strains UvGHF1‐C1 and C2. The data from three independent inoculation assays are shown as mean ± SD (n = 10). Different letters indicate significant differences in the average number of false smut balls (p < 0.05; Duncan's multiple range test).

Accordingly, gene deletion mutants and complemented strains of UvGHF1 were generated and confirmed by PCR, Southern blot, and western blot analyses (Figure S11a–d). The ∆uvghf1‐2 strain grew slightly more slowly than the wild type and complemented strains on potato sucrose agar (PSA) plates during 14 days of culturing (Figure S11d,e). Inoculation assays showed that false smut balls generated after inoculation of the gene deletion mutant Δuvghf1‐2 were significantly fewer than those produced by the wild‐type strain at 4 weeks after inoculation (Figure 7b,c). The complemented strains constructed by introducing the UvGHF1 gene into the deletion mutant largely restored U. virens virulence to rice (Figure 7b,c). These results indicate that UvGHF1 is required for virulence of U. virens to rice.

3. DISCUSSION

U. virens infects rice florets through stamen filaments and causes a unique floral disease. Given that the pathogen infects rice florets without forming observable infection structures, such as haustoria and appressoria (Tang et al., 2013), cell wall‐degrading enzymes including GH proteins secreted by U. virens might play important roles in fungal penetration (Zhang et al., 2014). Although U. virens secretes numerous GH proteins, none of them has been functionally characterized (Zhang et al., 2014). In this study, we demonstrated that a secreted protein, UvGHF1, containing a GH42 domain, functions as an essential virulence factor and contributes to U. virens pathogenicity. Transiently expressed UvGHF1 in N. benthamiana is accumulated in periplasmic spaces. Interestingly, UvGHF1, being recognized as a PAMP, induces plant immunity in nonhost N. benthamiana and host rice plants.

Similar to the GH42 protein (Godoy et al., 2016), in vitro purified UvGHF1 exhibited detectable activity on the artificial substrate p‐nitrobenzene‐β‐d‐galactopyranoside (Figure S4). This result, together with sequence alignment and point mutation analyses of UvGHF1, reveals that the GH protein is indeed a β‐galactosidase, and the conserved residues Glu165 and Glu285 are critical sites for its enzymatic activity (Figure S3). In addition, full‐length UvGHF1, but not the truncated UvGHF1 lacking the SP, induced cell death in N. benthamiana, suggesting that UvGHF1 triggers defence responses in the apoplast (Figure 2). This finding is consistent with the localization of UvGHF1 in periplasmic spaces, which was supported by different methods, including yeast secretion assays, apoplastic fluid analysis, and fluorescence observation of N. benthamiana cells after plasmolysis (Figure 1 and Figure S2). Furthermore, the translocation assay showed that ectopically expressed UvGHF1 was present in the EIHM compartment during M. oryzae infection (Figure 1d), indicating that UvGHF1 is an apoplastic effector. Notably, we showed that UvGHF1 activated plant immunity in a manner that is independent of GH activity (Figure 2). This is in line with the previous finding that VdEG1 and VdEG3 trigger cell death in a manner that is independent of their cellulolytic activity in N. benthamiana (Gui et al., 2017). Likewise, the enzymatic activity of FoEG1 is not required for cell death induction (Zhang, Yan, et al., 2021). By contrast, CfGH17‐1, an apoplastic GH17 family protein with 1,3‐β‐glucanase activity from C. fulvum, triggers cell death in a manner that is dependent on its enzymatic activity (Ökmen et al., 2019). Therefore, we speculate that these GH proteins induce cell death through different molecular mechanisms.

Some fungal effectors, particularly apoplastic effectors, are recognized by PRRs and thereby trigger plant defence responses. In most cases, the membrane‐localized immune coreceptors BAK1 and SOBIR1 are required for this recognition (van den Burgh et al., 2019). RXEG1 in N. benthamiana recognizes XEG1 from P. sojae via the LRR domain and induces PRR complex formation with the coreceptors BAK1 and SOBIR1 to transduce XEG1‐induced defence signals (Wang et al., 2018). Similar to XEG1, cell death triggered by BcXYG1 (Zhu et al., 2017), VdEG1 (Gui et al., 2017), VdCUT11 (Gui et al., 2018), and FoEG1 (Zhang, Yan, et al., 2021) is dependent on both BAK1 and SOBIR1. By contrast, VdEG3‐triggered cell death is only dependent on BAK1, but not on SOBIR1 (Gui et al., 2017). In this study, we showed that UvGHF1 no longer triggered cell death in NbBAK1‐silenced plants but continued to induce necrosis in NbSOBIR1‐silenced plants (Figure 6), suggesting that UvGHF1‐induced plant cell death is also dependent on BAK1. Therefore, we speculate that OsBAK1 forms a PRR complex with an unidentified receptor to recognize UvGHF1 to trigger rice immunity. Further investigation is required to uncover the UvGHF1 receptor.

In addition to eliciting cell death in nonhost plants, UvGHF1 induced defence responses and resistance against various pathogens in the host plant. Chitin and flg22 trigger the first layer of immune responses, including ROS burst and activation of defence marker genes (Shi et al., 2022; Zhang, Yang, et al., 2020). We revealed that the UvGHF1‐expressing transgenic plants exhibited an elevated PAMP‐triggered ROS accumulation and higher expression of OsPR1b and OsPR10a compared with wild‐type plants (Figure 3). Meanwhile, UvGHF1‐expressing transgenic plants exhibited significantly enhanced resistance to both bacterial and fungal pathogens (Figure 4), suggesting that UvGHF1 induces resistance to various pathogens. RNA‐seq data revealed that a large number of defence‐related genes were significantly up‐regulated in UvGHF1‐expressing transgenic plants (Figure 5 and Figure S9). The up‐regulated DEGs were significantly enriched in diterpenoid biosynthesis and phenylpropanoid biosynthesis, which are considered as general defence responses against a wide range of pathogens. Enhanced expression of OsGA2ox4, OsCYP76M14, OsPOX29, and OsPOX5.1 substantiates the essential role of diterpenoid and phenylpropanoid biosynthesis in plant resistance (Figure 5c and Figure S8). LRR proteins on the plant plasma membrane recognize microbe‐derived molecules to elicit pattern‐triggered immunity (Pruitt et al., 2021). A total of 19 DEGs were annotated to encode probable LRR family proteins or receptor‐like kinases. Among them, 15 DEGs were significantly up‐regulated, suggesting that these LRR proteins are involved in rice disease resistance (Figure 5d). Transcription factors are well known to regulate plant immunity against various pathogens (Hong et al., 2022; Wani et al., 2021). Here, we found that 22 WRKYs and 16 ERFs were significantly up‐regulated in UvGHF1‐expressing transgenic plants (Figure 5e,f). Among them, many WRKY factors, including OsWRKY6 and OsWRKY13, have been demonstrated to enhance resistance to rice pathogens (Im et al., 2022; Qiu et al., 2007). Therefore, we speculate that a complicated regulatory network is involved in disease resistance in UvGHF1‐expressing transgenic plants, which needs an in‐depth study.

Interestingly, deletion of UvGHF1 attenuated U. virens virulence, that is, much fewer false smut balls were generated on rice panicles, and virulence of the complemented strains was close to the wild‐type level (Figure 7b,c). Therefore, we conclude that UvGHF1 is crucial to the virulence of U. virens. It will be interesting to investigate how U. virens coordinates the seemingly opposite functions of UvGHF1 for successful infection. Here, several possibilities are proposed. First, UvGHF1 expression is precisely controlled at different U. virens infection stages. The expression of UvGHF1 is up‐regulated during the early infection stage for initial penetration (Figure 7a), but is suppressed at the late infection stage. Second, UvGHF1 might not elicit immune responses in rice floral organs to avoid inducing immunity at infection sites. Likewise, SGP1 in U. virens functions as a MAMP and activates immune responses in rice leaves, but not in floral organs (Song et al., 2021). Third, UvGHF1‐induced plant immunity might be inhibited by other U. virens effectors. U. virens secretes a large number of effectors to facilitate host colonization. Multiple effector proteins, including SCRE1, SCRE2/UV_1261, SCRE4, SCRE6, UvSec117, UvPr1a, and UvCBP1, inhibit plant immunity and enhance the pathogenesis of U. virens (Chen, Duan, et al., 2022; Chen, Li, et al., 2022; Fang et al., 2019; Li et al., 2021; Li et al., 2022; Qiu et al., 2022; Yang et al., 2022; Zhang, Yang, et al., 2020; Zheng et al., 2022). It is worthwhile to confirm these hypotheses in the future.

In addition, the ∆uvghf1 mutant showed a slightly smaller colony diameter than did the wild‐type strain (Figure S11f). Similarly, UvPr1a has been identified as a virulence effector in U. virens. The mycelial growth rate of the ∆UvPr1a mutants is slightly reduced compared with that of the wild type and complemented strains (Chen, Li, et al., 2022). Accordingly, it will be interesting to study how UvGHF1 positively affects mycelial growth and contributes to virulence in U. virens.

To the best of our knowledge, this is the first report showing that a GH42 protein from a phytopathogenic fungus induces plant immunity in host and nonhost plants. All these characteristics make UvGHF1 a good candidate immunostimulant for enhancing rice disease resistance. However, many mysteries on this intriguing class of UvGHF1‐containing GH42 proteins are still to be solved, particularly the molecular mechanisms underlying their functions.

4. EXPERIMENTAL PROCEDURES

4.1. Plant materials and microbial strains

Oryza sativa subsp. japonica ‘Nipponbare’ and transgenic rice plants were grown in the greenhouse. N. benthamiana plants were grown in a growth chamber at 25°C with a relative humidity of 60% and a 14 h light/10 h dark cycle. Escherichia coli strains DH5α and Rosetta‐gami (DE3) and A. tumefaciens GV3101 were grown in Luria Bertani (LB) medium at 37°C and 28°C, respectively. X. oryzae pv. oryzae was cultured in nutrient broth (NB) medium at 28°C. U. virens isolate P1 and M. oryzae isolate S5 were cultured on PSA and oat tomato agar plates at 28°C, respectively.

4.2. Yeast secretion assay

The yeast secretion assay was performed as described previously (Yang et al., 2022). Briefly, the sequence encoding the predicted SP of UvGHF1 was cloned into the vector pSUC2. The resultant pSUC2‐UvGHF1 ^SP plasmid was transformed into the yeast strain YTK12. The transformants screened on CMD−W medium plates (0.67% yeast N base without amino acids, 0.075% tryptophan dropout supplement, 2% sucrose, 0.1% glucose, and 2% agar) were incubated on YPRAA medium (1% yeast extract, 2% peptone, 2% raffinose, 2 mg/mL antimycin A, and 2% agar) plates with raffinose as the sole carbon source. The enzymatic activity of secreted invertase was also detected by the reduction of TTC to insoluble red TPF.

4.3. Effector translocation assay

The effector translocation assay was performed as described previously (Khang et al., 2010; Qian et al., 2022; Zhang, Yang, et al., 2020). UvGHF1 was subcloned into the pYF11‐ProRP27:GFP vector via homologous recombination in Saccharomyces cerevisiae. The plasmid was isolated from S. cerevisiae with a yeast plasmid extraction kit (Solarbo) and then transformed into M. oryzae Guy11 via polyethylene glycol (PEG)‐mediated protoplast transformation. The transformants with strong green fluorescence were selected for inoculation assays. The conidial suspension (10⁵ spores/mL) was prepared and injection‐inoculated into rice leaf sheaths. Green fluorescence was observed by confocal microscopy at 36–40 h postinoculation.

4.4. Agrobacterium‐mediated gene expression

Agrobacterium‐mediated transient expression in N. benthamiana was performed as described previously (Fang et al., 2019). UvGHF1 and UvGHF1 ^ΔSP were subcloned into pTA7001 (Qiu et al., 2022). The constructed plasmids were transformed into A. tumefaciens GV3101. Overnight cultured Agrobacterium strains were resuspended in acetosyringone solution (10 mM MES, pH 5.7, 10 mM MgCl₂, and 150 μM acetosyringone) to a cell density of OD₆₀₀ = 0.5. After incubation for 2 h, Agrobacterium cells were infiltrated into 4–5‐week‐old N. benthamiana leaves. The infiltrated leaves were sprayed with 10 μM DEX at 24 h after agro‐infiltration. Cell death symptoms were photographed at 5 days after DEX spraying. Protein expression in N. benthamiana was verified by western blotting using an anti‐FLAG antibody (Sigma‐Aldrich). Ion leakage was measured in the agro‐infiltrated leaves as described previously (Fang et al., 2019; Mittler et al., 1999).

4.5. Subcellular localization

To determine the subcellular localization of UvGHF1, the UvGHF1 coding sequence was subcloned into pGD‐mCherry (Goodin et al., 2002). The constructs to express UvGHF1‐mCherry, UvGHF1^ΔSP‐mCherry, and mCherry were transformed into A. tumefaciens GV3101 for agro‐infiltration. At 2 days after agro‐infiltration in N. benthamiana leaves, mCherry fluorescence was observed with an excitation wavelength of 514 nm under an LSM780 confocal laser scanning microscope (Zeiss). For plasmolysis, N. benthamiana leaves were detached and immersed with 1 M NaCl.

4.6. Isolation of intercellular fluids

The coding sequences of UvGHF1 and UvGHF1 ^ΔSP were amplified and subcloned into the pGD vector to express UvGHF1‐GFP and UvGHF1^ΔSP‐GFP, respectively. The constructs were transformed into A. tumefaciens GV3101 for agro‐infiltration. The transformed Agrobacterium strains were agro‐infiltrated into N. benthamiana leaves. Leaf intercellular fluids were isolated from the infiltrated N. benthamiana leaves at 48 h after agro‐infiltration according to a previously described method (He et al., 2022). UvGHF1‐GFP and UvGHF1^ΔSP‐GFP were detected by western blotting using an antibody against GFP (1:5000 dilution). The protein loading was indicated by Ponceau S staining.

4.7. In vitro protein expression and purification

In vitro protein expression and purification was performed as described previously (Godoy et al., 2016). The coding sequence of UvGHF1 ^ΔSP was amplified and subcloned into pET28a. The UvGHF1 ^E165A and UvGHF1 ^E285A constructs were generated by site‐directed mutagenesis. These constructs were transformed into E. coli BL21(DE3). Expression of recombinant proteins was induced by 0.5 mM isopropyl β‐d‐1‐thiogalactopyranoside for 20 h at 16°C. His‐tagged proteins were purified using Ni‐NTA beads (Novagen). Protein concentration was determined using a BCA protein assay kit (CWBio). The purified proteins were separated in 10% SDS‐PAGE gels and were stained with Coomassie brilliant blue.

4.8. β‐Galactosidase activity assay

In vitro β‐galactosidase activity of UvGHF1 and its variants was determined as described previously (Godoy et al., 2016).

4.9. Generation of transgenic rice lines

The pTA7001‐UvGHF1‐FLAG construct was introduced into rice cv. Nipponbare through Agrobacterium‐mediated transformation (Zhang, Yang, et al., 2020). The transgenic lines were screened on hygromycin B‐containing medium plates.

4.10. RNA sequencing

Wild‐type and transgenic rice panicles were injected with 10 μM DEX solution at 5–7 days before heading. The panicles were collected at 16 h after injection for RNA isolation. The enriched mRNA was fragmented into short fragments and then reverse transcribed into cDNA using the NEBNext Ultra Directional RNA Library Prep Kit (NEB). The constructed library was sequenced using an Illumina Novaseq 6000 platform by Gene Denovo Biotechnology Co. (Guangzhou, Guangdong). We identified the genes with log₂(fold change) ≥ ∣1∣ and p < 0.05 as significant DEGs. A Venn diagram was plotted and PCA was performed using the online OECloud tools at https://cloud.oebiotech.com. All DEGs were mapped to GO terms in the Gene Ontology database to determine the main biological functions (http://www.geneontology.org/). KEGG pathway enrichment analysis was performed to identify the metabolic pathways or signal transduction pathways significantly enriched in DEGs (Kanehisa et al., 2017).

4.11. RT‐qPCR assay

One‐week‐old rice seedlings pretreated with 10 μM DEX overnight were treated with flg22 (1 μM), chitin (8 μg/mL), or sterile double deionized water for 6 h. The seedlings were then collected for RNA isolation. Alternatively, U. virens‐inoculated or DEX‐injected rice panicles were collected at different time points. Total RNAs were isolated using TRIzol reagent (CWBio). The PrimeScript RT reagent kit with gDNA Eraser (TaKaRa) was used for reverse transcription. RT‐qPCR was performed with a LightCycler 96 System (Roche) using the FastSYBR mixture kit (CWBio). Relative gene expression was calculated using the comparative threshold cycle value (Ct) method (2^−ΔΔCt). The α‐tubulin and OsActin genes were used as internal controls for gene expression in U. virens and rice, respectively.

4.12. ROS burst assay

ROS burst assays were conducted as previously described (Shi et al., 2018; Wang, Wang, et al., 2021). Briefly, 6‐week‐old wild‐type and transgenic plants were sprayed with DEX (10 μM) or mock solution. Leaf discs were punched at 16 h after spraying and were then preincubated with sterile double deionized water overnight. ROS generation was initiated by adding 100 μM of d‐luciferin potassium salt substrate (Gold Biotechnology) with flg22 (10 μM), chitin (8 μg/mL), or sterile double deionized water. The chemiluminescence signal was immediately recorded at 1‐min intervals for a period of 25 min using a GloMax 20/20 luminometer (Promega).

4.13. Gene knockout and complementation in U. virens

CRISPR/Cas9‐based gene knockout was performed as previously described (Li et al., 2019). Briefly, sgRNA primers designed for the target gene were annealed and inserted into pmCAS9‐tRp‐gRNA. The upstream and downstream flanking sequences of UvGHF1 were amplified and fused to the hygromycin resistance gene by fusion PCR. The resultant PCR product and the plasmid construct were introduced into U. virens protoplasts through PEG‐mediated transformation. Knockout of UvGHF1 in hygromycin‐resistant transformants was screened by PCR using four sets of primers (Table S1). The UvGHF1 deletion mutants were confirmed by Southern blot analysis using the DIG‐High Prime DNA Labeling and Detection Starter Kit II (Roche) (Li et al., 2019; Qiu et al., 2022; Wang, Liu, et al., 2021). For complementation, the full‐length UvGHF1 gene and its gene variants with the native promoter were amplified and then ligated into pY2P102. The constructs were introduced into the mutant strain via PEG‐mediated transformation.

4.14. Inoculation assays

Wild‐type and transgenic plants were treated with DEX (10 μM) and mock solution for 16 h and were then evaluated for resistance to fungal and bacterial diseases. U. virens strains were injection‐inoculated into rice panicles at 5–7 days before heading (Qiu et al., 2022). False smut balls formed on rice panicles were counted at 4 weeks after inoculation. Six‐week‐old rice plants were inoculated with X. oryzae pv. oryzae PXO99A using the leaf‐clipping method and disease lesion lengths were measured at 14 dpi (Zhao et al., 2021). Rice seedlings were inoculated with M. oryzae S5 by spray inoculation (Zhang, Fang, et al., 2020). Disease symptoms on the leaves were observed and the relative fungal biomass was quantified by genomic DNA‐based RT‐qPCR at 7 dpi.

4.15. Protein extraction and immunoblotting

Total proteins were extracted using IP buffer (100 mM Tris–HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA, 5 mM EGTA, 1 mM dithiothreitol, 0.5% Triton X‐100, and 1 × protease inhibitor cocktail [Roche]). The protein extracts were separated on a 10% SDS‐PAGE gel after heating at 95°C for 10 min. The proteins were electrophoretically blotted onto nitrocellulose membranes (Millipore). After being blocked with 5% skimmed milk in TBS‐T buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, and 0.05% Tween 20) at 4°C overnight, the membranes were incubated sequentially with anti‐FLAG (1:5000 dilution) (Sigma‐Aldrich) and horseradish peroxidase‐conjugated anti‐mouse secondary antibody (1:5000 dilution) at room temperature for 1 h. After thoroughly washing with TBS‐T buffer, the immunoblots were incubated with eECL Western Substrate (CWBio) and signals were recorded using X‐ray film.

4.16. Virus‐induced gene silencing

VIGS was performed as described previously (Zhang & Thomma, 2014). Briefly, the target genes were subcloned into pTRV2. Agrobacterium strains carrying pTRV1 or pTRV2 with target genes were mixed at a 1:1 ratio and then infiltrated into N. benthamiana cotyledons of the two‐leaf stage. The efficiency of gene silencing was analysed by RT‐qPCR. The UvGHF1‐containing Agrobacterium strain was infiltrated at 16–20 days after VIGS. Cell death symptoms were monitored at 5 days after agro‐infiltration.

4.17. Statistical analysis

Significance analyses in multiple comparisons were performed with one‐way analysis of variance followed by Duncan's multiple range test using SPSS v. 23.0. Pairwise comparisons were evaluated through Student's t test.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no competing interests.

Supporting information

Figure S1 Diagram showing the structure of UvGHF1. UvGHF1, which consists of 339 amino acid residues, is predicted to contain a signal peptide and a GH42 domain. The signal peptide indicated in orange was predicted by SignalP v. 4.1. The GH42 domain is indicated in blue.

Figure S2 UvGHF1 can be detected in the apoplastic fluid collected from Nicotiana benthamiana leaves. UvGHF1‐GFP and UvGHF1^ΔSP‐GFP were transiently expressed in N. benthamiana and were determined by immunoblotting with an anti‐GFP antibody (α‐GFP). The loading control is indicated by Ponceau S staining.

Figure S3 The phylogenetic tree and multiple sequence alignment analysis of UvGHF1 homologues. (a) The phylogenetic relationships of UvGHF1 and its homologues. The phylogenetic tree of UvGHF1 and its homologues in various fungi, including Beauveria bassiana, Metarhizium anisopliae, Purpureocillium lilacinum, Trichoderma longibrachiatum, and Hirsutella minnesotensis, and in the bacterial species Paludibacterium purpuratum was constructed using the neighbour‐joining method through MEGA v. 7.0. A GH35 protein (XP014540152.1) from Metarhizium brunneum was used as an outgroup. (b) Multiple sequence alignment of UvGHF1 in Ustilaginoidea virens and its homologous proteins from the indicated species. Two highly conserved amino acid residues in the catalytic pocket are marked with red arrows.

Figure S4 The key catalytic residues are required for the β‐galactosidase activity of UvGHF1. (a) Coomassie brilliant blue (CBB) staining showing in vitro purified UvGHF1 and its point‐mutated variants. UvGHF1^E165A and UvGHF1^E285A are the mutated variants in which the two putative catalytic Glu residues were individually replaced by Ala. The purified proteins were detected and visualized by Coomassie brilliant blue staining. (b) β‐Galactosidase activity of UvGHF1 and its point‐mutated variants. Data from three independent assays are shown as mean ± SD (n = 3). Different letters (a vs. b) indicate significant differences in enzyme activity between the wild‐type and mutated proteins (p < 0.01; Duncan’s multiple range test).

Figure S5 Induced expression of UvGHF1‐FLAG in transgenic rice lines as detected by immunoblotting. The wild type (Nip) and independent transgenic lines IE6 and IE27 were treated with dexamethasone (DEX, 10 μM) and mock solutions. UvGHF1‐FLAG was detected by immunoblotting with an anti‐FLAG antibody (α‐FLAG). Total protein loading is indicated by Ponceau S staining.

Figure S6 Venn diagram and principal component analysis (PCA) confirming the reproducibility of the experiments and the reliability of transcriptome data. (a) Venn diagram showing 4183 differentially expressed genes (DEGs) overlapping between the IE6 versus wild type and IE27 versus wild type comparisons, including 2540 up‐regulated and 1643 down‐regulated genes. (b) PCA plot showing that the samples from two transgenic lines (IE6 and IE27) were grouped as a clade, but were separated from the wild‐type plant samples. Ellipses represent the 95% confidence intervals.

Figure S7 Gene Ontology (GO) and KEGG pathway enrichment analyses of differentially expressed genes (DEGs) between the wild type and transgenic rice lines. (a) GO enrichment analysis revealed the top 15 GO terms enriched in DEGs between the wild type and transgenic rice lines. The size of circles represents the number of DEGs mapped to specific GO terms. The colour bar indicates p values. (b) KEGG pathway enrichment analysis revealed the top 15 KEGG pathways enriched in DEGs between the wild type and transgenic rice lines. The size of circles represents the number of DEGs mapped to specific KEGG pathways. The colour bar indicates p values.

Figure S8 Schematic diagram of the expression profiles of genes involved in the phenylpropanoid biosynthesis pathway. Cubes represent the expression levels in the wild type, IE6, and IE27, with purple colour representing high values and green colour indicating low values based on the fold change of gene expression.

Figure S9 The transcript levels of six differentially expressed genes in rice after Ustilaginoidea virens infection. The relative expression levels of the indicated genes were determined by reverse transcription‐quantitative PCR in rice at 24 h after mock and U. virens inoculation. OsActin was used as an internal reference gene. Data are presented as mean ± SD (n = 3) (***p < 0.001, **p < 0.01; Student’s t test).

Figure S10 NbBAK1‐ and NbSOBIR1‐silenced Nicotiana benthamiana plants were generated through tobacco rattle virus‐induced gene silencing (TRV‐VIGS). The TRV:PDS construct, leading to efficient photobleaching in N. benthamiana leaves due to silencing of the phytoene desaturase (PDS) gene, was used to confirm VIGS efficiency. TRV:mCherry was used as a negative control.

Figure S11 Validation of the UvGHF1 knockout and complemented strains. (a) Screening of the UvGHF1 replacement mutants by PCR. The fragments between the hygromycin phosphotransferase (hph) gene and upstream or downstream flanking sequences of the UvGHF1 gene were amplified by PCR with designed primers (Table S1). The UvGHF1 replacement mutant candidates are indicated by the absence of the UvGHF1 gene and successful amplification of upstream and downstream sequences. (b) Schematic presentation of the strategy to confirm UvGHF1 gene replacement via Southern blot analysis. (c) The UvGHF1 knockout mutant was confirmed through Southern blot analysis. Genomic DNA isolated from the wild type P1 and different candidate gene replacement transformants was digested with SphI. After being separated on agarose gels, the digested DNA was blotted onto a nylon membrane. The digested DNA was then hybridized with a digoxigenin‐labelled UvGHF1 probe. (d) The complemented strains were confirmed by western blot analysis. Total proteins isolated from the wild type P1 and different hygromycin‐resistant strains were subjected to immunoblotting with an anti‐FLAG antibody (α‐FLAG). The confirmed complemented strains were marked in red. (e) Colony morphology of wild‐type, Δuvghf1 mutant, and complemented strains cultured on potato sucrose agar (PSA) plates at 28°C for 14 days. (f) The colony diameter of wild‐type, Δuvghf1 mutant, and complemented strains. Data are shown as mean ± SD of three biological replicates. The asterisk (*) indicates a significant difference in colony diameter between the wild type and mutant strains (*p < 0.05; Student’s t test).

Table S1 The primers used in this study.

Table S2 Differentially expressed genes as identified by RNA‐seq.

Zou, J. , Jiang, C. , Qiu, S. , Duan, G. , Wang, G. , Li, D. et al. (2023) An Ustilaginoidea virens glycoside hydrolase 42 protein is an essential virulence factor and elicits plant immunity as a PAMP . Molecular Plant Pathology, 24, 1414–1429. Available from: 10.1111/mpp.13377

DATA AVAILABILITY STATEMENT

Sequence data from this article can be found in the Rice Annotation Project (RAP) (https://rapdb.dna.affrc.go.jp/index.html). Other data that support the findings of this study are available from the corresponding author upon reasonable request.