Comparative genomic and transcriptional analyses of the carbohydrate-active enzymes and secretomes of phytopathogenic fungi reveal their significant roles during infection and development

Our comparative genomic analysis showed that the numbers of plant cell wall (PCW)- and fungal cell wall (FCW)-degradation-associated carbohydrate-active enzymes (CAZymes) in necrotrophic and hemibiotrophic fungi are significantly larger than that in most biotrophic fungi. However, our transcriptional analyses of CAZyme-encoding genes in Melampsora larici-populina, Puccinia graminis and Sclerotinia sclerotiorum showed that many genes encoding PCW- and FCW-degradation-associated CAZymes were significantly up-regulated during the infection of both necrotrophic fungi and biotrophic fungi, indicating an existence of a universal mechanism underlying PCW degradation and FCW reorganization or modification, which are both intimately involved in necrotrophic and biotrophic fungal infection. Furthermore, our results showed that the FCW reorganization or modification was also related to the fungal development. Additionally, our transcriptional analysis of the secretome of S. sclerotiorum showed that many secreted protein-encoding genes were dramatically induced during infection. Among them, a small, cysteine-rich protein SsCVNH was experimentally confirmed to be essential for the virulence and sclerotial development, indicating that the small secreted proteins might also play crucial roles as potential effectors in host-non-specific necrotrophic fungi.


Results
Comparative genomic analysis of the CAZymes in phytopathogenic fungi. Considering the representativeness and accessibility of available genome data, nine genomes of hemibiotrophic and necrotrophic fungi (Magnaporthe oryzae, Fusarium graminearum, F. oxysporum, F. verticillioides, Phaeosphaeria nodorum, Pyrenophora tritici-repentis, Verticillium dahliae, Botrytis cinerea and S. sclerotiorum), seven genomes of biotrophic fungi (Blumeria graminis, Ustilago maydis, Cladosporium fulvum, Melampsora larici-populina, Puccinia graminis, P. triticina and P. striiformis) and two yeast genomes (Saccharomyces cerevisiae and Schizosaccharomyces pombe) were chosen for the comparative analysis of CAZymes. Our results showed that the numbers of genes encoding PCW-and FCW-degradation-associated CAZymes and their respective related CBMs in necrotrophic and hemibiotrophic fungi were both significantly larger than that in most of biotrophic fungi except for C. fulvum (Fig. 1, Table S1), indicating both PCW-and FCW-degradation-associated CAZymes may play crucial roles during the infection and (or) development of necrotrophic and hemibiotrophic fungi.
Scientific RepoRts | 5:15565 | DOi: 10.1038/srep15565 DGE of S. sclerotiorum. To explore the expression patterns of CAZyme-encoding genes during the infection and development of necrotrophic fungi, we used DGE based on deep sequencing technology to illuminate the transcriptional responses of CAZyme-encoding genes during the infection stage and different developmental stages (including vegetative growth, sclerotial development, sclerotial myceliogenic germination, sclerotial carpogenic germination and apothecium formation) of the necrotrophic fungus S. sclerotiorum, which is used as an example. Sequencing quality evaluation and sequencing saturation analyses guaranteed the data quality for further DGE analysis ( Figure S1 and S2). Differentially expressed genes were identified during the various developmental stages compared with the vegetative growth stage ( Figure S3). To validate the DGE data, twelve genes were selected randomly for quantitative reverse transcription PCR (qRT-PCR) analysis. The data were presented as fold changes in gene expression normalized to the β -tubulin gene [24][25][26] . Pearson's correlation coefficient (R value) was used to measure the consistency of the qRT-PCR and DGE data. The R values ranged from 0.88 to 1.00, indicating the qRT-PCR results were strongly correlated with the DGE data ( Figure S4). These results showed that the expression patterns of the twelve genes in the DGE and qRT-PCR data were accordant and our DGE data was reliable for further transcriptional analysis.
The expression patterns of CAZyme-encoding genes in S. sclerotiorum. Among the 501 identified CAZyme-encoding genes in S. sclerotiorum, the expression of 315 was significantly regulated during different developmental stages, compared with vegetative growth stage (Table S2a). Our result showed that the geometric mean of the expression values of the PCW-and FCW-degrading CAZyme-encoding genes and their respective related CBMs showed dramatic induction during the infection and various developmental stages of S. sclerotiorum (Fig. 2a), indicating both of the PCW-and FCW-degrading CAZymes were activated during these biological processes. As expected, gene expression cluster analysis showed that a large number of genes encoding PCW-degradation-associated CAZymes and related CBMs were significantly up-regulated during infection (Fig. 2b). For example, 73 (39.2%) out of a total of 186 genes encoding PCW-degradation-associated CAZymes and related CBMs, were significantly up-regulated during infection (Fig. 2d, Table S3). This result confirmed the important roles of plant CWDEs during infection. Additionally, a large number of genes encoding FCW-degradation-associated CAZymes and related CBMs were also dramatically induced during infection (Fig. 2c). For example, 19 (18.8%) from a total of 101 genes encoding FCW-degradation-associated CAZymes and related CBMs, were significantly up-regulated during infection (Fig. 2e, Table S3). Microscopic observation showed that the invasive hyphal tips were clearly different from the vegetative hyphal tips: the latter were thinner and their morphology was normal, whereas the former were thicker and their morphology was The numbers of PCW-(a) and FCW-(c) degradation-associated CAZymes and respective related CBMs ((b,d) respectively) were plotted for necrotrophic and hemibiotrophic fungi, biotrophic fungi and two yeasts. The numbers of PCW-and FCW-degradation-associated CAZymes and related CBMs showed significant differences among necrotrophic and hemibiotrophic fungi, biotrophic fungi and two yeasts. Differentiation was supported by the t-test at a significance level of P < 0.05. blistered, swollen and malformed (Fig. 3). This result further suggested FCW reorganization or modification was intimately involved in the infection processes of necrotrophic fungi. During the vegetative and reproductive developmental stages, there are 36 (19.4%), 37 (19.9%), 32 (17.2%) and 31 (16.7%) out of a total of 186 genes encoding PCW-degrading enzymes and related CBMs which were significantly up-regulated during sclerotial development, sclerotial myceliogenic germination, sclerotial carpogenic germination and apothecium formation, respectively, compared with the vegetative growth stage (Fig. 2d,   Figure 2. Transcriptional analysis of genes encoding CAZymes and related CBMs during different developmental stages of S. sclerotiorum. (a) The expression levels of genes encoding PCW-and FCWdegradation-associated CAZymes and respective related CBMs during different developmental stages. The expression levels were measured using the geometric mean of the TPM values of all genes in corresponding groups. To calculate the geometric mean of TPM values, all the TPM values of 0 were replaced by 0.001. (b) Gene expression cluster analysis of genes encoding PCW-degradation-associated CAZymes and related CBMs. (c) Gene expression cluster analysis of genes encoding FCW-degradation-associated CAZymes and related CBMs. The TPM values were used for the gene expression cluster analysis. Red, green and grey indicate high expression, low expression and no expression, respectively; For each heat map: top, stage tree; left, gene tree. Expression values are indicated in log2 scale. Column diagrams show the proportions of differentially expressed genes and unchanged expressed genes encoding PCW-(d) and FCW-(e) degradation-associated CAZymes and respective related CBMs, during different developmental stages, compared with the vegetative growth stage of S. sclerotiorum. The differentially expressed genes in the DGE data were identified with the threshold of |log 2 Ratio| ≥ 1 and FDR ≤ 0.001. Up, Down and Un indicate the up-regulated, down-regulated and unchanged expressed genes, respectively. Table S3), indicating the expression of these genes encoding PCW-degrading enzymes and related CBMs was induced not by specific plant factors but by the developmental regulation of S. sclerotiorum or by tough environmental factors which are unfavorable for vegetative growth (such as nutrient starvation). Notably, a large number of genes encoding FCW-degrading CAZymes and related CBMs were also significantly induced during the vegetative and reproductive developmental stages of S. sclerotiorum. For example, there are 28 (27.7%), 21 (20.8%), 18 (17.8%) and 19 (18.8%) from a total of 101 genes encoding FCW-degrading enzymes and related CBMs that were significantly up-regulated during sclerotial development, sclerotial myceliogenic germination, sclerotial carpogenic germination and apothecium formation, respectively, compared with the vegetative growth stage (Fig. 2e, Table S3). These results indicated that the FCW reorganization or rearrangement also played important roles in the vegetative and reproductive development of S. sclerotiorum, and the FCW was in a state of dynamic change during various developmental stages. Taken together, the CAZyme-encoding genes exhibited diverse expression patterns during the infection and multiple developmental stages of S. sclerotiorum, compared with the vegetative stage, indicating that the CAZymes might play versatile roles in these developmental processes.  (Table S4a), the proportion varied from 17.9% to 36.4% during different infection stages (Fig. 4e, Table S4a). There are 39 (40.2%) and 37 (38.1%) out of a total of 97 genes encoding PCW-degradation-associated CAZymes and related CBMs that were significantly induced during the infection of P. graminis urediniospores on wheat and barley, respectively, compared with the stage of urediniospores in vitro (Fig. 4g, Table S4b). Surprisingly, many genes encoding FCW-degrading CAZymes and related CBMs in M. larici-populina and P. graminis were also significantly regulated during infection. In details, there are 26 (36.1%) from a total of 72 genes encoding FCW-degradation-associated CAZymes and related CBMs which were significantly induced during at least one infection stage of M. larici-populina urediniospores on poplar leaves, compared with the stage of dried-urediniospores in vitro (Table S4a), the proportion varied from 13.9% to 23.6% during different infection stages (Fig. 4f). There are 22 (36.1%) and 22 (36.1%) from a total of 61 genes encoding FCW-degradation-associated CAZymes and related CBMs that were significantly induced during the infection of P. graminis urediniospores on wheat and barley, respectively, compared with the stage of urediniospores in vitro (Fig. 4h, Table S4b). These results indicated that the PCW degradation and FCW reorganization or modification might also be intimately involved in the infection processes of biotrophic fungi, although the numbers of genes encoding PCW and FCW-degradation associated CAZymes in biotrophic fungi were smaller than those in necrotrophic fungi.

Transcriptional analyses of CAZyme-encoding genes in
Transcriptional analysis of the secretome of S. sclerotiorum. We compared the total numbers of secreted proteins, small secreted proteins and cysteine-rich, small secreted proteins from the 18 fungi described above, respectively. The result showed P. striiformis, M. larici-populina, F. verticillioides, M. oryzae, P. graminis, F. oxysporum and P. triticina had more secreted proteins, especially small secreted proteins, compared with the other species ( Figure S5). To decipher the biological functions of the S. sclerotiorum secretome, the gene expression patterns of the secretome were clustered based on our DGE data. The result showed that many genes encoding secreted proteins were significantly regulated during the infection, sclerotial development, sclerotial myceliogenic germination, sclerotial carpogenic germination  and apothecium formation stages, compared with the vegetative growth stage ( Figure S6, Table S2b), indicating their diversified roles during these biological processes. Gene functional enrichment analysis was performed on the significantly up-regulated genes encoding secreted proteins during different developmental stages. Many Funcat functional categories and gene ontology (GO) items were enriched during various developmental stages, especially during the sclerotial development and infection stages ( Figure  S7, Table S5), indicating their important and diverse roles in these biological processes. Interestingly, compared with the vegetative growth stage, the "cell wall" associated functional modules (such as "cell wall organization", "fungal-type cell wall" and "cell wall modification") were significantly over-represented during different developmental stages ( Figure S7, Table S5), which is in accord with our hypothesis that the FCW is in a state of dynamic change during various developmental stages. Additionally, the results showed the PCW-and FCW-degradation associated functional modules (such as "extracellular polysaccharide degradation", "extracellular lignin degradation", "cellulose binding", "chitin binding", "hydrolase activity", "cellulase activity", "cutinase activity", "polygalacturonase activity", "beta-galactosidase activity" and "mannan endo-1,6-alpha-mannosidase activity") were also over-represented during different developmental stages ( Figure S7, Table S5), intimating the significant roles of CAZymes in corresponding biological processes.
SsCVNH is a predicted cysteine-rich, small secreted protein with a CVNH domain. To experimentally validate the important roles of small secreted proteins during the infection, vegetative and reproductive developmental stages of S. sclerotiorum, a predicted secreted protein-encoding gene (SS1G_02904) was selected as an example for further study. We chose this gene because of the following reasons: (i) In planta expression analysis can help to unearth candidate effector genes 18 and SS1G_02904 had the highest expression fold change during the infection in our DGE data; (ii) Many fungal effectors are cysteine-rich, small secreted proteins 17,29 , and SS1G_02904 consists of 153 aa and contains 10 cysteine residues that account for more than 6.5% of the total amino acids; (iii) Some fungal effectors show conserved protein domains, and SS1G_02904 was identified as an effector candidate that shares 35% identity with a Colletotrichum hingginsianum effector candidate 21,30 . SS1G_02904 is a predicted secreted protein with a signal peptide (1-20 aa) in its N-terminus and a CVNH domain (pfam08881, E-value = 2.65e-06) in its C-terminus (Fig. 5a); hence, it is designated as SsCVNH. The CVNH domain corresponds to a carbohydrate-binding module 31,32 . The predicted three-dimensional structure of SsCVNH is highly similar to the structure of cyanovirin-N 33 (Fig. 5b). Although the CVNH domain was found to be widely distributed in cyanobacteria, filamentous ascomycetes and the fern Ceratopteris richardii 34 , protein sequence similarity search (blastp) result showed that SsCVNH had homologs only in Sclerotinia and Botryotinia, using an E-value lower than 1e-6 as the threshold (Fig. 5c). These results indicated that the three-dimensional structure of the CVNH domain was more evolutionarily conserved than the protein sequence.
The expression patterns of SsCVNH. DGE data showed that SsCVNH was not expressed during the vegetative growth stage and was significantly induced during the stages of sclerotial development, sclerotial myceliogenic germination, sclerotial carpogenic germination, apothecium formation and infection (Fig. 6a). QRT-PCR analysis further showed that the expression of SsCVNH was significantly up-regulated during the initial stage of sclerotial development at 3 days post-inoculation (dpi) and peaked during the mature stage of sclerotial development at 7 dpi. The expression change of this gene reached more than 600-fold during sclerotial development (Fig. 6b). When pure, actively growing hyphal fragments of S. sclerotiorum without any culture medium were inoculated onto Arabidopsis thaliana (Col-0) leaves, the transcripts of SsCVNH rapidly increased by more than 15-fold during the early infection stage at 3 hours post-inoculation (hpi) and maintained at high expression levels during the later infection stages (Fig. 6c). These results were coincident with the DGE data, suggesting that SsCVNH may be involved in the infection, vegetative and reproductive development of S. sclerotiorum. SsCVNH is secreted from the hyphal tips. Bioinformatics analysis revealed that SsCVNH might be a secreted protein due to the presence of a signal peptide in its N-terminus. An immunofluorescence technique was used to confirm this hypothesis. FLAG-tagged SsCVNH strains under the control of the P EF − 1α promoter were engineered. Western blot analysis showed FLAG-tagged SsCVNH could be detected in the hyphae of these engineered strains (Fig. 7a). The mycelia of the engineered SsCVNH-FLAG strains were inoculated on the onion bulb lower epidermis prior to live cell imaging of the infected onion epidermal cells to examine the subcellular localization of SsCVNH during infection. To eliminate the possible interference of the autofluorescence of plant or fungal tissues, two different secondary antibodies conjugated with rhodamine red-X (RRX) or fluorescein isothiocyanate (FITC) were used independently to exhibit the specificity of fluorescence signal. The result showed fluorescence could be observed in the hyphal tips of the SsCVNH-FLAG engineered strains. No fluorescence was observed in the wild-type strain. Similar results were obtained with different antibodies tagged with FITC or RRX (Fig. 7b), indicating that SsCVNH was indeed a secreted protein that was secreted from the hyphal tips during the infection of S. sclerotiorum. SsCVNH-silenced transformants showed significantly reduced virulence and abnormal sclerotial development. Due to the presence of the multi-nucleated cells in S. sclerotiorum, RNAi technology was used to characterize the biological functions of SsCVNH. QRT-PCR was used to examine the transcript accumulation in SsCVNH-silenced transformants. The results showed that the sclerotial development of the transformants (SsCVNH-89, SsCVNH-84 and SsCVNH-40) with dramatically reduced SsCVNH expression was completely blocked on PDA medium at 20 °C. However, the wild-type strain and the transformant SsCVNH-46 with slightly reduced SsCVNH expression showed normal sclerotial development (Fig. 8a,d). The growth rate of the silenced transformants was also significantly reduced, although there was no obvious difference between the morphology of the hyphal tips of the silenced transformants and the wild-type strain (Fig. 8c,f). The virulence of the SsCVNH-silenced transformants was significantly reduced, and only small lesions developed on the detached leaves of Brassica napus at 2 dpi. Furthermore, the decrease in virulence was positively correlated with the silencing efficiency of different transformants (Fig. 8b,e). Similar result was observed when the SsCVNH-silenced transformants were inoculated on detached tomato leaves ( Figure S8), indicating that the virulence reduction was not host species-specific. These results indicated that SsCVNH played crucial roles in the virulence, sclerotial development and growth rate of S. sclerotiorum. In conclusion, our results indicated that small secreted proteins in typically necrotrophic fungi might function as potential effectors similar to those in hemibiotrophs and biotrophs.

Discussion
To date, whether the phytopathogenic fungi use FCW-degrading enzymes for the degradation or modification of their own cell walls, the walls of antagonistic fungi, or for plant polysaccharides during infection is unknown 35 . Our results indicated that the expression levels of many genes encoding FCW-degrading enzymes were significantly regulated during the infection of phytopathogenic fungi. Meanwhile, our results indicated different kinds of FCW-degrading enzymes participated in this biological process and their proportions varied during different infection stages, indicating the FCW reorganization or modification process is sophisticatedly and stringently regulated. The present results support the hypothesis that these FCW-degrading enzymes may play significant roles in shaping their own cell walls because the FCW reorganization or remodeling could be clearly observed during infection. The FCW reorganization or rearrangement itself may be essential for phytopathogenic fungal infection. This hypothesis may also be adapted to the development of phytopathogenic fungi because the FCW components are obviously different in various tissues during different developmental stages. Our results indicated that the FCW was in a state of dynamic change during the infection and development of phytopathogenic fungi, which is consistent with the previous studies on FCW dynamics 6,36 . Additionally, our results indicated that the PCW degradation and FCW reorganization are intimately involved in both necrotrophic and biotrophic fungal infection. But the total numbers of genes encoding PCW-and FCW-degrading CAZymes and associated CBMs in necrotrophic and hemibiotrophic fungi are significantly larger than that in most biotrophic fungi. Furthermore, the proportions of differentially expressed CAZyme-encoding genes are also obviously different during infection, indicating the capacity of PCW degradation and FCW reorganization is different in necrotrophic, hemibiotrophic and biotrophic fungi. This is consistent with the lifestyles of different types of fungi, because the nutrient acquisition of necrotrophic and hemibiotrophic fungi is based on host cell killing at the last stage of infection 37 , while biotrophic fungi absorb nutrients from living cells 38 . Our result showed a total of 235 secreted protein-encoding genes were significantly up-regulated during infection (Table S2b). These strongly induced genes are likely to represent "key offensive or repressive forces" or "effector candidates" in S. sclerotiorum that target the plant defense system during the early stages of infection. Notably, many proteinaceous fungal effectors are cysteine-rich, small secreted proteins 16,17,29 . In this study, SsCVNH was used as an example to elucidate the significant roles of these types of effector candidates during the infection, vegetative and reproductive development of S. sclerotiorum. In many reports, most effectors or small secreted proteins seemed to have little effect on pathogen development. Interestingly, this may not be the case in S. sclerotiorum because, to date, at least two small secreted proteins (SSITL and Ss-Caf1) have been reported to play important roles in the sclerotial development  22,23 . Similarly, our results demonstrated that SsCVNH also played significant roles in sclerotial development, not only in infection. Additionally, in our large-scale gene function studies of the S. sclerotiorum secretome, SsCVNH was not a unique example to show that the secreted proteins have diverse biological functions during multiple developmental stages of S. sclerotiorum (data not shown here). These results suggest a scenario in which different groups of secreted proteins synergistically function during various developmental stages.
Our result showed that SsCVNH was a secreted protein with a CVNH domain. Cyanovirin-N (CVN) has previously been identified only in the aqueous extracts from the cyanobacterium Nostoc ellipsosporum 39 . Based on our PSI-BLAST search in the NCBI non-redundant protein sequence database, CVNHs are found in a restricted range of eukaryotic organisms as diverse as the fern Ceratopteris richardii, the filamentous ascomycetes (e.g., Tuber borchii, Neurospora, Magnaporthe, Sclerotinia, Botrytis, Aspergillus, Fusarium, Colletotrichum and Trichoderma) and the basidiomycetes (e.g., Rhizoctonia solani) but are lacking in other lineages (data not shown). This patchy organism distribution suggests the acquisition of the CVNH domain by some organisms after the separation of the main evolutionary lineages and the occurrence of lateral gene transfer events. CVNH-containing organisms have strikingly different lifestyles, ranging from symbiotic, saprotrophic to pathogenic. Meanwhile, in these organisms, there are both sclerotium-producing fungi and non-sclerotium-producing fungi. These phenomena suggest that the CVNHs are not sufficient for infection and sclerotial development but are necessary in some cases. CVNH-related proteins in ferns and cyanobacteria are found from the secretome. However, owing to data limitations, previous studies reported that no secretion signals were identified in any of the fungal CVNHs in filamentous ascomycetes, and only three types of domain architectures of CVNHs have been reported 31,34 . Due to the increasing availability of genomic sequences, we find there are more than three types of domain architectures of CVNHs, including SsCVNH with a signal peptide in the N-terminus before the CVNH domain, thereby indicating that these CVNHs may also be secreted proteins in filamentous ascomycetes. Our result further confirmed this hypothesis and supplemented the knowledge concerning the domain architectures of CVNHs.

Materials and Methods
Data collection and bioinformatics analyses. The predicted proteomes of M. oryzae, F. graminearum, F. oxysporum, F. verticillioides, P. nodorum, P. tritici-repentis, V. dahliae, B. cinerea, S. sclerotiorum, U. maydis, P. graminis, P. triticina, P. striiformis, S. cerevisiae and S. pombe were downloaded from the Broad Institute (http://www. broadinstitute.org/science/projects/projects). The predicted proteomes of C. fulvum and M. larici-populina were obtained from the DOE Joint Genome Institute (JGI) site 40 and the predicted proteome of B. graminis was downloaded from BluGen (http://www. blugen.org/) 41 . The microarray data of M. larici-populina and P. graminis used in this study were downloaded from Gene Expression Omnibus (GEO) at NCBI (http://www.ncbi.nlm.nih.gov/geo/), and their accession numbers are GSE21624 and GSE25020, respectively 27,28 . Protein structure modeling was performed using the Phyre2 server 42 (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id= inde7x) and rendered using UCSF Chimera 43 . COBALT 44 (http://www.ncbi.nlm.nih.gov/tools/ cobalt/cobalt.cgi?link_loc= BlastHomeAd) was used to generate the amino acid alignment, which was viewed and edited with "Jalview" software 45 . The CAZymes were identified by the Hmmscan program in the HMMER 3.0 package 46 using the family-specific HMM profiles of CAZymes from the dbCAN database 47 . The sub-classification of CAZymes (PCW-degrading enzymes, FCW-degrading enzymes and their respective related CBMs) was performed according to Amselem et al. 35 . The SignalP 48 was used to identify secreted proteins. In this study, we defined cysteine-rich, small, secreted proteins as the secreted proteins consisting of at least 76 aa and that are less than 300 aa in length and contain at least 4% cysteine residues. Gene expression patterns were clustered using "cluster" 49 , and the hierarchical clustering algorithm was used based on the average-linkage method 50 . The dendrogram and colored image were produced by "javaTreeview" 51 . The functional enrichment analyses were performed according to FungiFun2 52 .
Bacterial and fungal strains, plant and culture conditions. The virulent S. sclerotiorum wild-type strain Ep-1PNA367 was used and routinely cultured on potato dextrose agar (PDA, Difco, Detroit, MI, USA) at a neutral pH at 20 °C. S. sclerotiorum transformants were cultured on PDA amended with 80 μ g/ml hygromycin B (Calbiochem, San Diego, CA, USA). Escherichia coli strain DH5α was used to propagate all of the plasmids, and A. tumefaciens strain EHA105 was used for the transformation of fungi. Seedlings of A. thaliana (ecotype Columbia-0) were grown in the greenhouse at 20 ± 2 °C for one month under a 12 h light/dark cycle with 70% relative humidity.
Sample preparation, RNA extraction, cDNA production and DGE analysis. For DGE analysis, the Ep-1PNA367 strain was grown or treated under different conditions, and samples were collected for RNA extraction during the following stages: (i) Vegetative stage: fresh mycelia of the Ep-1PNA367 strain were placed on a cellophane membrane overlaid onto PDA medium at 20 °C, and the mycelia were collected at 12, 24, 36, 48 and 60 h; (ii) Sclerotial formation stage: the colonies growing on the cellophane membrane overlaid onto PDA medium were further incubated under the same conditions and then the cultures were collected at 84, 96, 108, 120 and 132 h; (iii) Early stages of infection: fresh hyphal fragments of the Ep-1PNA367 strain were overlaid onto sterilized cheese cloth, which was overlaid onto the leaves of the A. thaliana ecotype Columbia-0, followed by inoculation at 20 °C with 100% relative humidity, and then, the cheese cloth with hyphae was rolled up from the leaves at 9 h and 12 h; (iv) Sclerotial myceliogenic germination stage: sclerotia were surface sterilized with sodium hypochlorite and were placed onto a PDA plate at 20 °C to induce myceliogenic germination; the sclerotia were harvested when approximately 50% had germinated; (v) Sclerotial carpogenic germination stage: sclerotia were dried at room temperature and pretreated in a freezer (4-6 °C) for up to one month; then, the sclerotia were surface sterilized and placed onto wet sterilized sands in a plate at 15 °C to induce carpogenic germination; the sclerotia were harvested when approximately 50% germinated (stipes having only emerged from the sclerotia); and (vi) Early apothecial formation stage: sclerotia were allowed to grow in the same incubator, and the stipes were cut and collected immediately before apothecium formation for RNA extraction. Finally, different time-point samples from the vegetative stage, sclerotial formation stage and infection stage were pooled together in equal quantities for RNA extraction. For qRT-PCR analysis, the samples used for validation of the DGE data were prepared independently under the same conditions described above. To evaluate the expression levels of SsCVNH in different transformants, the transformants and the wild-type strain were cultured on PDA at 20 °C for 7 dpi before total RNA extraction. This time-point was selected for qRT-PCR analysis because the expression of SsCVNH peaks at 7 dpi. For both DGE and qRT-PCR analysis, total RNA extraction was conducted according to Zhu et al. 23 . Briefly, total RNA was isolated with TriZOL reagent (Invitrogen, Carlsbad, USA) according to the manufacturer's protocols. The total RNA was incubated for 30 min at 37 °C with 10 units of DNase I (Roche Applied Science, Shanghai, China) to remove residual genomic DNA. The first cDNA strand was synthesized from 1.0 μ g of total RNA by Moloney murine leukemia virus reverse transcriptase (Promega, Madison, USA) using oligo(dT) 18 primers. For DGE analysis, tag-based transcriptome sequencing methods were used to perform the DGE (Solexa/Illumina, Shenzhen, China) analysis. The DGE raw sequences were transformed into clean tags and subsequently mapped to the S. sclerotiorum transcript database from the Broad Institute (http://www.broadinstitute.org/annotation/genome /sclerotinia_sclerotiorum/ MultiDownloads.html). All of the clean tags were mapped to the reference sequences with no more than 1 nt mismatch allowed. Clean tags mapped to reference sequences from multiple genes were filtered. The remaining clean tags were designated as unambiguous clean tags. The number of unambiguous clean tags for each gene was calculated prior to normalization to the number of transcripts per million clean tags (TPM) 53 . The complete expression datasets are available at the GEO database (NCBI) as series GSE65301.
Identification of differentially expressed genes. For DGE experiment, rigorous algorithms were developed to identify differentially expressed genes (or significantly regulated genes) between the two cDNA libraries according to Audic et al. 54 . In this study, the P-value and false discovery rate (FDR) were manipulated to determine differentially expressed genes 55 . An FDR ≤ 0.001 and an absolute value of log 2 Ratio ≥ 1 (more than 2.0 fold change) were used as thresholds to determine significant differences in gene expression. For microarray experiment, a Student t test with FDR multiple testing correction was applied to the data using the INVEX web 56 . Transcripts with an adjusted P value < 0.05 and an absolute value of log 2 Ratio ≥ 1 (more than 2.0 fold change) in transcript level were considered as significantly differentially expressed.

Quantification of gene expression by qRT-PCR.
QRT-PCR was used to validate the DGE data and evaluate the expression levels of SsCVNH in different transformants. The mRNA transcripts were measured using a SYBR Green I real-time RT-PCR assay in a CFX96 ™ real-time PCR detection systems (Applied Biosystems). The thermal cycling conditions were 95 °C for 2 min for predegeneration, 40 cycles of 95 °C for 10 s for denaturation, 58 °C for 20 s for annealing and 72 °C for 20 s for extension. All of the reactions were run in triplicate by monitoring the dissociation curve to control the dimers. The S. sclerotiorum β -tubulin gene was used as a normalizer 25 . After amplification, data acquisition and analysis were performed using the Bio-Rad CFX Manager TM Software (version 2.0). The 2 −ΔΔCT method was chosen as the calculation method 57 . The difference in the cycle threshold (CT) values of the genes and the housekeeping gene β -tubulin (called Δ C T ) was calculated as follows: Δ Δ CT = (Δ C T of genes at each time point)-(Δ C T of the initial control). See supplementary Table S6 for the primers for qRT-PCR.

Western blot analysis and the subcellular localization of SsCVNH. To generate the
SsCVNH-FLAG fusion constructs ( Figure S9a), the promoter P EF−1α was PCR amplified using the primers P EF−1α − 1 F/R and subsequently digested with Xho I and Sac I. The PCR products of SsCVNH were amplified with the primers FLAG-SsCVNH F/R and subsequently digested with Sac I and Sma I. These two fragments were sequentially ligated into the pCH vector through intermediate constructs. The primers are shown in Supplementary Table S6. The SsCVNH-FLAG construct was transformed into the wild-type strain using the Agrobacterium-mediated transformation method as previously described 58 . Briefly, the A. tumefaciens cells with the SsCVNH-FLAG constructs were diluted in minimal medium 59 amended with 50 μ g/ml kanamycin and incubated overnight at 28 °C. Then the A. tumefaciens cells were diluted in induction medium 59 and incubated for 6 h at 28 °C with gentle shaking, before they were co-cultivated with fresh S. sclerotiorum mycelial plugs on a cellophane membrane placed on co-induction medium (induction medium with agar) at 20 °C for 2 days. The cellophane membrane was then transferred to the selective medium [PDA amended with 80 μ g/ml hygromycin B and 200 μ g/ml cefotaxime Scientific RepoRts | 5:15565 | DOi: 10.1038/srep15565 sodium (DingGuo, Beijing, China)] after the mycelial plugs were removed. Colonies that were regenerated through the selective medium after incubation at 20 °C for 3-6 days were transferred to the selective medium for subcultivation. Total proteins were extracted from the transformants by cell lysis buffer for western blotting and IP (Beyotime, Wuhan, Hubei, China). A total of 5 μ l of ANTI-FLAG M2 monoclonal antibody (Sigma, St Louis, MO, USA) was added to 1 ml of the protein extracts and incubated at room temperature for 2 hours. Next, protein A + G agarose (Beyotime, Wuhan, Hubei, China) was added to the protein extracts and incubated at room temperature for 1 hour prior to collection by centrifugation. The extracts were washed five times with cell lysis buffer, and then protein loading buffer was added for the following western blot analysis. Tissues from onion epidermal cells inoculated with SsCVNH-FLAG-engineered strains were sampled at 36 hpi. The lower epidermis was peeled and cut into 1-cm pieces near the inoculation site and fixed with 4% paraformaldehyde for 10 min. Subsequently, the sample was washed 3 times with 1 × phosphate-buffered saline (PBS, pH 7.4) (5 min each wash), incubated in 0.2% Triton X-100 for 10 min and washed 3 additional times with 1 × PBS (5 min each wash). After removing the Triton X-100, the sections were blocked for 30 min in 1% (w/v) BSA and washed 3 times with 1 × PBS (5 min each wash). The sections were subsequently incubated with primary anti-FLAG-tag mouse monoclonal antibody for 2 h at 37 °C, rinsed 3 times with 1 × PBS (5 min each rinse) and incubated in secondary rhodamine red-X-conjugated (RRX) or fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG for 2 h at 37 °C. Subsequently, the onion tissues from the inoculated lower epidermis were rinsed, viewed and photographed using a confocal laser scanning microscope (OLYMPUS ® microscope FV1000). The 570 nm and 492 nm absorption laser lines with corresponding appropriate specific emission filter sets were used when imaging RRX and FITC, respectively. 60 and Yu 58 were adopted to construct the S. sclerotiorum RNAi vectors. Briefly, a 336 bp fragment from SsCVNH was amplified with the primers RNAi-SsCVNH F/R from the S. sclerotiorum cDNA library and (i) directly ligated into the pCXDPH vector digested by Xcm I (New England Biolabs, Beverly, MA, USA) to produce the pRNAi-1 vector ( Figure S9b); and (ii) digested with two pairs of suitable enzymes, followed by ligation of both of the digested fragments into the pCIT vector in the inverse orientation. Finally, the fusion fragment of PtrpC-SsCVNH-intron-SsCVNH-TtrpC was digested with Sac І and Xho І and subsequently ligated into the pCH vector 58 to produce the pRNAi-2 vector ( Figure S9c). The primers used to generate these RNAi vectors are shown in Supplementary Table S6. The Agrobacterium-mediated transformation method used to transform S. sclerotiorum is described above.

Construction of RNAi vectors and the transformation of S. sclerotiorum. Two strategies described by Nguyen
Characterization of the SsCVNH-silenced transformants. To evaluate virulence, at least nine individual, fully expanded leaves of Brassica napus and tomato were detached and inoculated with a single 0.5-cm diameter fresh mycelium-colonized PDA plug of each silenced transformant and the wild-type strain for moisture culture at 20 °C. Disease severity was measured using the average lesion diameter at 48 hpi. To assay growth rates, the virulent wild-type strain and the silenced transformants were cultivated on PDA at 20 °C for 3 days. Mycelial agar discs were collected from the active colony edge and inoculated in the center of the PDA Petri dish at 20 °C prior to examination of hyphal growth. After growth on PDA at 20 °C for 48 h, the tip hyphal morphology of the virulent wild-type strain and the silenced transformants was observed under a light microscope. Each experiment was performed at least three times.