Host-Induced Gene Silencing of a G Protein α Subunit Gene CsGpa1 Involved in Pathogen Appressoria Formation and Virulence Improves Tobacco Resistance to Ciboria shiraiana

Hypertrophy sorosis scleroteniosis caused by Ciboria shiraiana is the most devastating disease of mulberry fruit. However, few mulberry lines show any resistance to C. shiraiana. An increasing amount of research has shown that host-induced gene silencing (HIGS) is an effective strategy for enhancing plant tolerance to pathogens by silencing genes required for their pathogenicity. In this study, two G protein α subunit genes, CsGPA1 and CsGPA2, were identified from C. shiraiana. Silencing CsGPA1 and CsGPA2 had no effect on hyphal growth but reduced the number of sclerotia and increased the single sclerotium weight. Moreover, silencing CsGpa1 resulted in increased fungal resistance to osmotic and oxidative stresses. Compared with wild-type and empty vector strains, the number of appressoria was clearly lower in CsGPA1-silenced strains. Importantly, infection assays revealed that the virulence of CsGPA1-silenced strains was significantly reduced, which was accompanied by formation of fewer appressoria and decreased expression of several cAMP/PKA- or mitogen-activated protein-kinase-related genes. Additionally, transgenic Nicotiana benthamiana expressing double-stranded RNA targeted to CsGpa1 through the HIGS method significantly improved resistance to C. shiraiana. Our results indicate that CsGpa1 is an important regulator in appressoria formation and the pathogenicity of C. shiraiana. CsGpa1 is an efficient target to improve tolerance to C. shiraiana using HIGS technology.


Introduction
Mulberry is an economically important tree with a long history in China [1]. Mulberry fruit have high nutritional value with abundant anthocyanin, flavone, resveratrol, and 1-deoxynojirimycin contents [2][3][4][5]. Mulberry leaves are the only feed for breeding silkworms, and have made great contributions to the prosperity of the Silk Road. Hypertrophy sorosis scleroteniosis caused by Ciboria shiraiana is the most devastating disease in mulberry fruit production, which causes huge economic losses in many mulberry growing areas every year, especially in southwest China [6,7]. The sclerotia of C. shiraiana fall to the ground along with the diseased mulberry fruit and remain in the soil over the winter. When the mulberry flowers bloom in spring, the sclerotia germinate and release a huge number of ascospores. The mature ascospores infect mulberry female flowers and eventually sclerotia are formed in the fruit, thereby completing its life cycle [8].

Cloning and Bioinformatics Analysis of CsGPA1 and CsGPA2 Genes
The total RNA of C. shiraiana was extracted using TRIzol reagent according to the manufacturer's procedures (Invitrogen, Carlsbad, CA, USA). A PrimeScript™ RT Reagent Kit (Takara Bio., Shiga, Japan) was used to synthesize cDNA. Base on the coding sequence obtained from the C. shiraiana gene database, primers were designed using Primer 5.0. The CsGPA1 (GenBank accession number: MZ574567) and CsGPA2 (GenBank accession number: MZ574568) genes were cloned from C. shiraiana cDNA using the primers CsGPA1-F and CsGPA1-R, and CsGPA2-F and CsGPA2-R, respectively (Table S1). The National Center for Biotechnology Information Blastp tool was used to search for homolog proteins from other fungi species (http://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on: 7 December 2021). The protein domain was predicted by SMART (http://smart.embl-heidelberg.de/, accessed on: 7 December 2021). Sequence alignment was performed using ClustalX software. The MEGA4 software using neighbor-joining method was applied to construct the phylogenetic tree.

Real-Time RT-PCR Analysis
To detect the expression of CsGPA1 and CsGPA2 in different organs, the hyphae, initial sclerotia, developing sclerotia, mature sclerotia, apothecia, and conidia of C. shiraiana were collected. Additionally, to analyze the expression of CsGPA1 and CsGPA2 genes during the infection process of N. benthamiana, fresh agar plugs of mycelia of the same size were inoculated onto tobacco leaves and the leaves were collected at time points of 0, 12,24,48,72, and 96 h post-inoculation (hpi). Total RNA was extracted using TRIzol reagent and used to synthesize cDNA with a PrimeScript™ RT Reagent Kit. Real-time quantitative PCR (qRT-PCR) was performed using a SYBR Green Reagent Kit (Takara Bio.) for 40 cycles with a final extension for 10 min. The β-tubulin gene served as the internal reference and the relative expression levels were determined using the 2 -∆∆Ct method [36].

Construction of CsGPA1 and CsGPA2 Silencing Vectors and Fungi Transformation
The pSilent-1 was digested by the SacI and XbaI and an approximately 1900 bp fragment (PtrpC-Hphr-TtrpC) was purified and then ligated into pCambia1300, which was digested by the same restriction enzymes to produce pCambia1300-PHT. Partial sense and anti-sense fragments of CsGPA1 genes were amplified from C. shiraiana cDNA using the primers SiCsGPA1-F/R and SiCsGPA1-F1/R1 with specific restriction sites. The sense fragments of CsGPA1 and CsGPA2 were cloned into psilent-1 by XhoI/HindIII, then antisense fragments were also ligated into the pSilent-1 plasmid in succession by BglII/KpnI. Subsequently, the plasmids were digested by the XbaI restriction enzyme and fragments containing the target genes were ligated with the linearized pCambia1300-PHT vector by T4 DNA Ligase (Takara Bio.) ( Figure S1a). The CsGPA1 and CsGPA2 silencing vectors were transformed into C. shiraiana via the Agrobacterium-mediated transformation method as described by Yu et al. [37]. A monoclonal colony was transferred to PDA plates with 60 µg/mL hygromycin and selected for three generations in succession. The primers for the hygromycin-resistant gene (Hyg-F and Hyg-R) were used to verify the CsGPA1and CsGPA2-silenced transformants.

Construction of HIGS Plasmids and Tobacco Transformation
Partial sense and anti-sense fragments of CsGPA1 were amplified from C. shiraiana cDNA using the primers dsCsGPA1-F/R and dsCsGPA1-F1/R1 shown in Table S1. The fragment was blasted in C. shiraiana (GenBank accession number: VNFM00000000) and N. benthamiana (https://solgenomics.net/tools/blast/, accessed on: 7 December 2021) genome and no discernible homology to off-target sequences was found. Additionally, Si-Fi software (v21) was used for off-target prediction (http://labtools.ipk-gatersleben. de, accessed on: 7 December 2021) and no off-target hits were found. The sequenced fragments were ligated with pHANNIBAL in succession by XhoI/KpnI and HindIII/BamHI. Then, the intermediate vector pHANNIBAL was digested with the restriction enzymes SacI and SpeI, and the fragments containing the target genes were purified and inserted into the destination vector pBin19 linearized by SacI and XbaI (SpeI and XbaI are a pair of isocaudarners) ( Figure S1b). The HIGS plasmid was transformed into Agrobacterium tumefaciens strain LBA4404 using a chemical method. A leaf disk co-cultivation method was used for N. benthamiana transformation [38].

Morphological Characteristics and Phenotypic Analysis of RNAi Strains
Wild-type (WT), empty vector (EV), and RNAi strains were cultured on PDA medium at 25 • C in an incubator. Hyphal growth was measured and photographed at 24 and 36 h. Sclerotia development phenotypes were photographed at 14 days post inoculation (dpi). Meanwhile, numbers and mass of sclerotia were determined. The fresh agar plugs were cultured on a hydrophobic glass slide to induce appressoria formation. The appressoria were observed using bright field microscopy (Olympus, Tokyo, Japan). For pathogenicity assays, fresh agar plugs of control and RNAi strains were inoculated onto N. benthamiana leaves. Pictures were taken at 48 hpi and the lesion size was measured using ImageJ software.

Stress Adaptation Assay
We analyzed the effects of the downregulated expression of CsGPA1 and CsGPA2 on the sensitivity of C. shiraiana to different stresses, as described by Feng et al. [39]. Fresh agar plugs of WT, EV, and RNAi strains were inoculated on CM medium supplemented with osmotic stress agents 1 M NaCl and 1 M KCl, oxidative stress agent 5 mM H2O2, and cell wall disturbing agents 0.005% sodium dodecyl sulphate (SDS) and 300 µg/mL Congo Red (CR) [39]. Pictures were taken at 48 hpi and the colony diameters were measured. All treatments were replicated three times.

Extracellular Laccase and Peroxidase Activity Assays
Extracellular laccase and peroxidase activities were determined as described by Chi et al. with slight changes [40]. The mycelia inoculated in CM liquid cultures for 3 days were removed completely by filtration and centrifugation for 10 min at 5000× g and 4 • C. The reaction mixtures (1 mL) consisted of 50 mM acetate buffer (pH 5.0) and culture filtrate mixed with the 10 mM ABTS (200 mL). The peroxidase and laccase activities were respectively determined in reaction mixtures with or without 3 mM H 2 O 2 . Then, the reaction mixtures were incubated for 5 min at 25 • C in darkness and absorbance was determined at 420 nm.

Relative Biomass and Histological Observations of Fungi in HIGS Plants
The leaves of WT and transgenic plants after inoculation with WT strains for 48 hpi were collected and then quickly homogenized into powder in liquid nitrogen. Total RNA and DNA were isolated to determine the expression of CsGPA1 and the relative biomass, respectively. The determination of relative biomass was performed according to Zhang et al. [41].
Fungal development in HIGS plants was observed as described by Redkar et al. [42]. Briefly, N. benthamiana leaves were collected at 10 hpi and then fixed with 100% ethanol and 10% KOH to remove chlorophyll. After this step, they were stained with Wheat Germ Agglutinin Alexa Fluor 488 (WGA-AF488) (Thermo Fisher Scientific, Waltham, MA, USA) as described by Redkar et al. [42]. All microscopy assays were performed using a laser scanning confocal microscope (Leica Microsystems, Wetzlar, Germany).

Statistical Analysis
All the data in this study were obtained from at least three independent repetitions. The final results are shown as means ± standard deviations (SD). SPSS Statistics 17.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. The graphs were created using GraphPad Prism 5 software (GraphPad Software Inc., La Jolla, CA, USA).

Identification and Expression Analysis of CsGPA1 and CsGPA2 Genes
A BLAST search using G protein α subunits from Magnaporthe oryzae and Saccharomyces cerevisiae as queries found two putative G protein α subunit genes in the C. shiraiana genome. The two genes were cloned from C. shiraiana cDNA and named CsGPA1 and CsGPA2. The characteristics of the predicted G protein α subunit genes are shown in Figure S2. The full-length genomic sequences of CsGPA1 and CsGPA2 were 1341 and 1365 bp, respectively. The open reading frame lengths were 1011 and 1062 bp with six and five exons, respectively ( Figure S2a). Both CsGPA1 and CsGPA2 were predicted to contain a G protein α subunit domain by SmartBLAST ( Figure S2b). Phylogenetic analysis showed that CsGPA1 and CsGPA2 had the closest relationships with S. sclerotiorum and B. cinerea, which also belong to Sclerotiniaceae ( Figure S2c).
To analyze the expression patterns of CsGPA1 and CsGPA2 at various developmental stages and plant infection process, their relative expression levels were determined using qRT-PCR. Expression levels of CsGPA1 and CsGPA2 were induced during sclerotial development stages. The expression level of CsGPA1 peaked at sclerotia 3 (mature sclerotia) and CsGPA2 was mainly expressed in hyphae and sclerotia 1 (initial sclerotia), implying that CsGPA1 and CsGPA2 may participate in sclerotia development ( Figure 1a). Both CsGPA1 and CsGPA2 showed upregulation in the late stage of plant infection (48-96 hpi). However, CsGPA2 expression was significantly decreased during the infection process compared with 0 hpi. It should be noted that CsGPA1 expression was lower in early-stage and higher in late-stage infection, indicating that it may play an important role in C. shiraiana pathogenicity ( Figure 1b).

Statistical Analysis
All the data in this study were obtained from at least three independent repetitions. The final results are shown as means ± standard deviations (SD). SPSS Statistics 17.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. The graphs were created using GraphPad Prism 5 software (GraphPad Software Inc., La Jolla, CA, USA).

Identification and Expression Analysis of CsGPA1 and CsGPA2 Genes
A BLAST search using G protein α subunits from Magnaporthe oryzae and Saccharomyces cerevisiae as queries found two putative G protein α subunit genes in the C. shiraiana genome. The two genes were cloned from C. shiraiana cDNA and named CsGPA1 and CsGPA2. The characteristics of the predicted G protein α subunit genes are shown in Fig Figure S2a). Both CsGPA1 and CsGPA2 were predicted to contain a G protein α subunit domain by SmartBLAST ( Figure S2b). Phylogenetic analysis showed that CsGPA1 and CsGPA2 had the closest relationships with S. sclerotiorum and B. cinerea, which also belong to Sclerotiniaceae ( Figure S2c).
To analyze the expression patterns of CsGPA1 and CsGPA2 at various developmental stages and plant infection process, their relative expression levels were determined using qRT-PCR. Expression levels of CsGPA1 and CsGPA2 were induced during sclerotial development stages. The expression level of CsGPA1 peaked at sclerotia 3 (mature sclerotia) and CsGPA2 was mainly expressed in hyphae and sclerotia 1 (initial sclerotia), implying that CsGPA1 and CsGPA2 may participate in sclerotia development ( Figure 1a). Both CsGPA1 and CsGPA2 showed upregulation in the late stage of plant infection (48-96 hpi). However, CsGPA2 expression was significantly decreased during the infection process compared with 0 hpi. It should be noted that CsGPA1 expression was lower in early-stage and higher in late-stage infection, indicating that it may play an important role in C. shiraiana pathogenicity (Figure 1b). Relative expression levels of CsGPA1 and CsGPA2 in different tissues (hyphae; sclerotia1, initial sclerotia; sclerotia2, developing sclerotia; sclerotia3, mature sclerotia; apothecia; and conidia) (a) and during infection (b) by qRT-PCR. Significant differences (p < 0.05) among columns were detected using a one-way ANOVA followed by Duncan's multiple range test and are labeled with different letters above the bars. Relative expression levels of CsGPA1 and CsGPA2 in different tissues (hyphae; sclerotia1, initial sclerotia; sclerotia2, developing sclerotia; sclerotia3, mature sclerotia; apothecia; and conidia) (a) and during infection (b) by qRT-PCR. Significant differences (p < 0.05) among columns were detected using a one-way ANOVA followed by Duncan's multiple range test and are labeled with different letters above the bars.

Characterization of CsGPA1-and CsGPA2-Silenced Strains
To determine the efficiency of CsGPA1 and CsGPA2 silencing, the candidate strains were screened by genomic PCR amplification of the hygromycin resistance gene ( Figure S3a,b). Additionally, the relative expression levels were determined by qRT-PCR. The expression levels of target genes were significantly lower in RNAi strains compared with WT and EV, and target gene expressions of all these mutants were reduced below 50% of WT, except for SiCsGPA2-5 ( Figure S3c,d). In CsGPA1and CsGPA2-silenced strains, three mutants with silencing efficiency of above 60% were selected for further studies.
Hyphal growth rates of CsGPA1and CsGPA2-silenced strains did not significantly differ from those of WT and EV strains ( Figure S4). However, no microsclerotia were formed in CsGPA1 and CsGPA2 RNAi strains (Figure 2a). The numbers of sclerotia were significantly lower in CsGPA1 and CsGPA2 RNAi strains, and the average weight of single sclerotia increased compared with WT and EV (Figure 2b). CsGPA1 RNAi strains had significantly higher total mass of sclerotia; however, there were only slight increases for CsGPA2 RNAi strains (Figure 2b). (a) morphologies of sclerotia; (b) number, mass, and average of single weight of sclerotia formed at 15 dpi. Three strains were selected for assay and every strain was replicated three times. Scale bars correspond to 2 mm. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. * p < 0.05, ** p < 0.01.

CsGPA1 Is Required for Compound Appressoria Formation
To further investigate the effects of CsGPA1 and CsGPA2 silencing on compound appressoria formation, hyphal morphology was observed using light microscopy at 24 hpi. The WT and EV strains formed abundant compound appressoria from vegetative hyphae (a) morphologies of sclerotia; (b) number, mass, and average of single weight of sclerotia formed at 15 dpi. Three strains were selected for assay and every strain was replicated three times. Scale bars correspond to 2 mm. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. * p < 0.05, ** p < 0.01.

CsGPA1 Is Required for Compound Appressoria Formation
To further investigate the effects of CsGPA1 and CsGPA2 silencing on compound appressoria formation, hyphal morphology was observed using light microscopy at 24 hpi. The WT and EV strains formed abundant compound appressoria from vegetative hyphae by 24 hpi. Conversely, few compound appressoria were found in CsGPA1 RNAi strains (Figure 3a). The CsGPA2-silenced strains had a similar number of compound appressoria to WT and EV (Figure 3a). The numbers of compound appressoria were significantly lower in CsGPA1-silenced strains compared with WT and EV; however, there was only a slight difference for CsGPA2-silenced strains (Figure 3b).  The numbers of compound appressoria in WT, EV, and gene-silenced strains (CsGPA1 and CsGPA2). All treatments were replicated three times. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** p < 0.001; ns, no significant difference at p < 0.05.

CsGPA1 Negatively Regulates Tolerance to Osmotic and Oxidative Stress but Is Dispensable for Extracellular Laccase and Peroxidase Activities
To investigate the roles of CsGPA1 and CsGPA2 in adaptation to various stresses, the growth rates of WT, EV, and CsGPA1 and CsGPA2 RNAi strains on CM medium supplemented with the five stress agents were compared (Figure 4a, Figure S5). The CsGPA1 RNAi strains grew faster than the WT and EV on plates supplemented with H2O2, KCl, or NaCl, but showed no differences in growth with SDS or CR (Figure 4b). Moreover, CsGPA1 silencing did not change the activities of extracellular laccase or peroxidase (Figures S6 and S7). However, the growth of CsGPA2 RNAi strains on CM supplemented with these five stress agents showed no obvious changes compared with WT and EV, although had slightly higher extracellular laccase and peroxidase activities (Figures S6 and S7). The numbers of compound appressoria in WT, EV, and gene-silenced strains (CsGPA1 and CsGPA2). All treatments were replicated three times. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** p < 0.001; ns, no significant difference at p < 0.05.

CsGPA1 Negatively Regulates Tolerance to Osmotic and Oxidative Stress but Is Dispensable for Extracellular Laccase and Peroxidase Activities
To investigate the roles of CsGPA1 and CsGPA2 in adaptation to various stresses, the growth rates of WT, EV, and CsGPA1 and CsGPA2 RNAi strains on CM medium supplemented with the five stress agents were compared (Figure 4a, Figure S5). The CsGPA1 RNAi strains grew faster than the WT and EV on plates supplemented with H 2 O 2 , KCl, or NaCl, but showed no differences in growth with SDS or CR (Figure 4b). Moreover, CsGPA1 silencing did not change the activities of extracellular laccase or peroxidase ( Figures S6 and S7). However, the growth of CsGPA2 RNAi strains on CM supplemented with these five stress agents showed no obvious changes compared with WT and EV, although had slightly higher extracellular laccase and peroxidase activities (Figures S6 and S7).

Silencing CsGPA1 Significantly Reduces C. shiraiana Virulence on Tobacco
To analyze the roles of CsGPA1 and CsGPA2 in pathogenesis, virulence was evaluated by inoculation on N. benthamiana leaves. There were significantly less necrotic lesions for CsGPA1-silenced strains than WT and EV strains (Figure 5a). The CsGPA2 RNAi strains also showed a slight non-significant reduction in virulence (Figure 5b).

Silencing CsGPA1 and CsGPA2 Downregulates Expression of cAMP and MAPK Signaling Genes
Expression levels of three MAPK signaling genes (CsSAKA, CsMPKA, and CsSMK1) and two cAMP signaling related genes (CsAC and CsPKA) in CsGPA1 RNAi strains showed significant downregulation compared with WT and EV strains ( Figure 6). Furthermore, similar decreases in expression levels of these genes were found in CsGPA2-silenced strains ( Figure S8).
In Figure 4a and Figure S5a of NaCl stress column, we mistakenly used same growth picture in WT and EV strains. So, we replaced the growth pictures of EV strain in NaCl stress in Figure  4a and Figure S5a using correct pictures, respectively.
In the original article, there was a mistake in Figure 4a and Figure S5a. In Figure 4a and Figure S5a, the EV strain growth picture in the NaCl stress is incorrect. Thus, Figure 4 and Figure S5 should be replaced with the following figure (Figure 4 and Figure S5).  The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: * p < 0.05.
for CsGPA1-silenced strains than WT and EV strains (Figure 5a). The CsGPA2 RNAi strains also showed a slight non-significant reduction in virulence (Figure 5b).
(a) (b) Figure 5. Virulence of WT, EV, and CsGPA1-and CsGPA2-silenced strains on tobacco leaves: (a) virulence and lesion areas of CsGPA1 mutants on tobacco leaves; (b) virulence and lesion areas of CsGPA2 mutants on tobacco leaves. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** significant difference at p < 0.001; ns, no significant difference at p < 0.05.

Silencing CsGPA1 and CsGPA2 Downregulates Expression of cAMP and MAPK Signaling Genes
Expression levels of three MAPK signaling genes (CsSAKA, CsMPKA, and CsSMK1) and two cAMP signaling related genes (CsAC and CsPKA) in CsGPA1 RNAi strains showed significant downregulation compared with WT and EV strains ( Figure 6). Furthermore, similar decreases in expression levels of these genes were found in CsGPA2silenced strains ( Figure S8). Figure 5. Virulence of WT, EV, and CsGPA1and CsGPA2-silenced strains on tobacco leaves: (a) virulence and lesion areas of CsGPA1 mutants on tobacco leaves; (b) virulence and lesion areas of CsGPA2 mutants on tobacco leaves. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** significant difference at p < 0.001; ns, no significant difference at p < 0.05. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: ** p < 0.01, *** p < 0.001.

CsGPA1 Silencing by Plant-Mediated RNAi Improves Plant Resistance to C. shiraiana
Silencing of CsGPA1 significantly reduced the C. shiraiana virulence, so CsGPA1 was used as the target to generate HIGS transgenic tobacco. A total of seven independent transgenic tobacco lines were obtained and three lines were used for a pathogen infection Figure 6. Relative expression levels of genes involved in cAMP/PKA and MAPK pathways in WT, EV, and CsGPA1-silenced strains. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: ** p < 0.01, *** p < 0.001.

CsGPA1 Silencing by Plant-Mediated RNAi Improves Plant Resistance to C. shiraiana
Silencing of CsGPA1 significantly reduced the C. shiraiana virulence, so CsGPA1 was used as the target to generate HIGS transgenic tobacco. A total of seven independent transgenic tobacco lines were obtained and three lines were used for a pathogen infection experiment. The transgenic tobacco lines all showed enhanced resistance to C. shiraiana to different extents (Figure 7a), mainly apparent as fewer necrotic lesions and lower relative pathogen biomass compared with WT tobacco (Figure 7b). Meanwhile, the level of CsGPA1 expression in hypha inoculated on transgenic tobacco leaves was significantly reduced (Figure 7b). The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** p < 0.001. The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** p < 0.001.
To further research the effects of HIGS tobacco on C. shiraiana development and infection, the morphology of hyphae inoculated on transgenic tobacco leaves was observed by microscope. The number of infection hyphae, indicating pathogenicity, was significantly reduced compared with WT tobacco (Figure 8a). In addition, significantly fewer compound appressoria were formed on transgenic tobacco leaves compared to WT plants, indicating that transgenic tobacco reduced the virulence of C. shiraiana (Figure 8b). The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by oneway ANOVA followed by Duncan's multiple range test. Note: *** p < 0.001.

Discussion
Heterotrimeric G-protein signaling pathways play important roles in the regulation of fungal growth, appressoria formation, and pathogenicity [13]. The G protein α subunits are the key component of trimeric G protein signals, which serve as molecular switches to regulate a series of cellular processes [18,20]. In this study, CsGPA1 and CsGPA2 were mainly expressed during sclerotia formation and late infection stages, suggesting that they were involved in the regulation of sclerotia formation and infection. Sclerotia and appressoria formation is vital for fungal spread and infection. In C. shiraiana, a maximum of 15 apothecia can germinate from a single sclerotium, and each apothecia measuring approximately 1.5 cm in diameter can release 5.6 × 10 7 to 6.3 × 10 7 ascospores [8]. In the filamentous fungus S. sclerotiorum, the Shk1-, SCD1-, and THR1-deletion mutants show significantly reduced melanin biosynthesis and sclerotia formation [43,44]. In this study, the CsGPA1-and CsGPA2-silenced mutants only formed sclerotia on the edge of the plate in lower numbers, suggesting that CsGPA1 and CsGPA2 were involved in the regulation of sclerotia formation via reduced melanin biosynthesis.
The functions of G protein α subunits in regulating vegetative growth vary in different fungal species. In M. robertsii, MrGpa1 deletion did not affect vegetative growth on PDA medium [18]. In Monascus ruber M7, the Mga2 and Mga3 knockout strains also The data represent the means ± SD of three independent replicates, and the differences between the mutants and the controls or EV were analyzed by one-way ANOVA followed by Duncan's multiple range test. Note: *** p < 0.001.

Discussion
Heterotrimeric G-protein signaling pathways play important roles in the regulation of fungal growth, appressoria formation, and pathogenicity [13]. The G protein α subunits are the key component of trimeric G protein signals, which serve as molecular switches to regulate a series of cellular processes [18,20]. In this study, CsGPA1 and CsGPA2 were mainly expressed during sclerotia formation and late infection stages, suggesting that they were involved in the regulation of sclerotia formation and infection. Sclerotia and appressoria formation is vital for fungal spread and infection. In C. shiraiana, a maximum of 15 apothecia can germinate from a single sclerotium, and each apothecia measuring approximately 1.5 cm in diameter can release 5.6 × 10 7 to 6.3 × 10 7 ascospores [8]. In the filamentous fungus S. sclerotiorum, the Shk1-, SCD1-, and THR1-deletion mutants show significantly reduced melanin biosynthesis and sclerotia formation [43,44]. In this study, the CsGPA1and CsGPA2-silenced mutants only formed sclerotia on the edge of the plate in lower numbers, suggesting that CsGPA1 and CsGPA2 were involved in the regulation of sclerotia formation via reduced melanin biosynthesis.
The functions of G protein α subunits in regulating vegetative growth vary in different fungal species. In M. robertsii, MrGpa1 deletion did not affect vegetative growth on PDA medium [18]. In Monascus ruber M7, the Mga2 and Mga3 knockout strains also showed no significant effects on vegetative growth [20]. However, in A. fumigatus, gpaB and ganA deletion resulted in decreased colony growth and increased growth of the gpaA deletion strain in minimal medium [19]. In this study, the findings suggested that CsGPA1 and CsGPA2 were not involved in vegetative growth. Therefore, the function of G protein α subunits is not completely conserved in fungal species.
In previous studies, G protein α subunits showed important roles in the regulation of sensitivity to stresses. For instance, MrGpa1 deletion strains of M. robertsii showed an increased tolerance to H 2 O 2 [18]. In Penicillium camemberti, overexpression of pga1 G42R reduced osmotic stress tolerance to 1.5 M NaCl and 1.5 M KCl [45]. In this research, similar results were obtained in CsGPA1-silenced strains, indicating that CsGPA1 negatively regulated resistance to osmotic and oxidative stresses. Laccase activity is involved in infection by some fungi [46]. Several lower-virulence mutants of M. oryzae also showed reduced laccase and peroxidase activities [47][48][49]. In our study, CsGPA2 silencing slightly improved laccase and peroxidase activities. However, CsGPA1-silenced strains had reduced virulence but normal laccase and peroxidase activity, indicating that CsGPA1-regulated pathogenicity may be independent of laccase and peroxidase in C. shiraiana.
In previous research, the G protein α subunit was shown to play an important role in appressoria formation. In M. grisea, disruption of magB significantly reduced appressoria formation, and a similar result was found in M. robertsii for MrGPA1 deletion strains [18,50]. The cAMP/PKA and MAP kinase signaling are important signal transduction cascades in regulating fungal appressoria formation and pathogenicity [51]. In M. robertsii, Ste11-, Ste7-, and Fus3-deletion strains showed significant decreases in appressoria formation and pathogenicity [52]. Similarly, the silencing of PsMPK7 from Phytophthora sojae reduced its oospore production and pathogenicity to soybean [53]. In addition to MAPK signaling, cAMP/PKA signal transduction has been well studied in the pathogenesis of several fungi, such as MaPKA1 in M. anisopliae and MAC1 from M. grisea; these genes seem to be involved in the regulation of appressoria formation and pathogenicity [54,55]. Furthermore, Gα4 regulated developmental morphogenesis in Dictyostelium by interacting with the MAPK ERK2 while the GpaB mutant had reduced pathogenicity by regulating cAMP/PKA signaling in Aspergillus flavus [56,57]. Therefore, G protein α subunits are involved in cAMP/PKA and MAP kinase signaling. In this study, silencing of CsGPA1 significantly reduced appressoria formation and virulence. Meanwhile, the expression levels of cAMP/PKA-and MAPKsignaling-related genes in CsGPA1 mutants were significantly decreased compared with WT. These results demonstrate that CsGPA1 is involved in the cAMP/PKA and MAPK signal transduction pathways in C. shiraiana, which control appressoria formation and pathogenicity.
Ciboria shiraiana is part of the Sclerotiniaceae family, and is considered to have a narrow host range due to its low amounts of secreted effector proteins [35]. Mulberry is a woody plant. It is very difficult to obtain transgenic mulberry because of its longer growth cycle and the lack of a stable transformation system. Nicotiana benthamiana is widely used as a model plant in research on biotic stress owing to its short growth period [58]. Previous studies have shown that C. shiraiana could infect N. benthamiana leaves rapidly and can serve as a model for performing infection tests [58]. Studies have also revealed that C. shiraiana can cause disease in tomato and rapeseed [35,59]. It is urgent to develop an economical, effective, and environmentally friendly method to control this disease. In some biotrophic and necrotrophic fungi, HIGS has been demonstrated as a new strategy to reveal gene function [29,32]. Cross-kingdom RNAi depends on the efficiency of fungal take-up of environmental double-stranded RNAs [60][61][62]. Both B. cinerea and S. sclerotiorum are also members of the Sclerotiniaceae, and can take-up environmental RNA with high efficiency, indicating that C. shiraiana may also be suited to HIGS [60]. Our results proved that CsGPA1 is an efficient target to increase the tolerance to C. shiraiana.

Conclusions
In conclusion, two G protein α subunit genes were isolated from C. shiraiana. Silencing CsGPA1 and CsGPA2 did not affect vegetative growth but reduced sclerotia formation. Moreover, CsGPA1-silenced strains showed reduced appressoria formation and virulence and improved tolerance to osmotic and oxidative stresses. Expressing the double-stranded RNA targeted to CsGPA1 in tobacco improved the tolerance to C. shiraiana. These data demonstrate that CsGPA1 is required for full virulence of C. shiraiana and is an efficient target to improve tolerance to C. shiraiana using HIGS technology.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/jof7121053/s1: Figure S1. Schematic diagram of RNAi (a) and HIGS (b) plasmids constructs. Figure S2: Gene features and phylogenetic analysis of CsGPA1 and CsGPA2. Figure S3: Generation of CsGPA1and CsGPA2-silenced strains. Figure S4: Hyphal growth in wild-type (WT), empty vector (EV), and gene-silenced strains (CsGPA1 and CsGPA2). Figure S5: Silencing of CsGPA2 did not affect the tolerance to osmotic, cell wall integrity, and oxidative stresses. Figure S6: Measurement of extracellular laccase activities. Figure S7: Measurement of extracellular peroxidase activity. Figure S8: Relative expression levels of genes involved in cAMP/PKA and MAPK pathways in WT, EV, and CsGPA2-silenced strains. Table S1: Primers used in this experiment.

Conflicts of Interest:
All these authors declare no competing interests.