Sclerotinia sclerotiorum SsCut1 Modulates Virulence and Cutinase Activity

The plant cuticle is one of the protective layers of the external surface of plant tissues. Plants use the cuticle layer to reduce water loss and resist pathogen infection. Fungi release cell wall-degrading enzymes to destroy the epidermis of plants to achieve the purpose of infection. Sclerotinia sclerotiorum secretes a large amount of cutinase to disrupt the cuticle layer of plants during the infection process. In order to further understand the role of cutinase in the pathogenic process of S. sclerotiorum, the S. sclerotiorum cutinsae 1 (SsCut1) gene was cloned and analyzed. The protein SsCut1 contains the conserved cutinase domain and a fungal cellulose-binding domain. RT-qPCR results showed that the expression of SsCut1 was significantly upregulated during infection. Split-Marker recombination was utilized for the deletion of the SsCut1 gene, ΔSsCut1 mutants showed reduced cutinase activity and virulence, but the deletion of the SsCut1 gene had no effect on the growth rate, colony morphology, oxalic acid production, infection cushion formation and sclerotial development. Complementation with the wild-type SsCut1 allele restored the cutinase activity and virulence to the wild-type level. Interestingly, expression of SsCut1 in plants can trigger defense responses, but it also enhanced plant susceptibility to SsCut1 gene knock-out mutants. Taken together, our finding demonstrated that the SsCut1 gene promotes the virulence of S. sclerotiorum by enhancing its cutinase activity.


Introduction
Sclerotinia sclerotiorum is a necrotrophic pathogen with a wide range of hosts, causing diseases and resulting in huge economic losses in many important crops such as oilseed rape, soybean, sunflower, and tomato [1]. Studies have shown that the pathogenicity mechanisms of S. sclerotiorum are very complicated. Many factors play important roles in the process of S. sclerotiorum infection, such as different plant cell wall-degrading enzymes and oxalic acid [2][3][4][5][6]. In addition, secreted proteins also play key roles in the pathogenic process of S. sclerotiorum, including SsCaf1, SsSSVP1, SsCP1 and SsITL [7][8][9][10]. At present, however, our understanding of the pathogenesis of S. sclerotiorum is still insufficient, and further research on its pathogenesis contributes to the development of new control strategies for Sclerotinia diseases [11][12][13].
The plant cuticle covers the leaf epidermis of higher plants and is the main barrier that isolates plants from direct contact with the atmosphere. The plant cuticle is mainly composed of cutin and wax, which is a hydrophobic property [14][15][16][17][18][19]. For different plant

Strains and Culture Conditions
The S. sclerotiorum (Lib.) de Bary wild-type strain 1980 (ATCC 18683) was obtained from diseased bean plants in Scottsbluff, NE [57]. In this study. Fungal strains were cultured on potato-dextrose agar (PDA) plates at 20 • C. The SsCut1 gene knock-out mutants and complement mutants were cultured on PDA plates amended with hygromycin B or G418 at 100 µg/mL (Sigma-Aldrich, St. Louis, MO, USA).

DNA/RNA Manipulation and RT-qPCR
Genomic DNA was isolated from mycelium by using the CTAB method [59]. Total RNA of fungi and plants were isolated with RNAiso Plus regent (Takara, Dalian, China) according to the manufacturer's protocols. To detect the expression pattern of cutinase gene in infected tissues, the hyphae of S. sclerotiorum wild-type strain were inoculated on Arabidopsis thaliana plant leaves. The hyphae were collected at 0, 1, 3, 6, 9, 12 and 24 h after inoculation. After RNA extraction, the first-strand cDNA was synthesized by the Easy Script One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China). Additionally, the gene expression was detected by qPCR analysis [60]. The β-tubulin gene in S. sclerotiorum, the UBQ5 gene in A. thaliana, and the elongation factor EF1α gene in N. benthamiana were used as internal controls, respectively. The real-time RT-qPCR analysis was repeated at least three times, with three biological replicates for each repeat.

The Yeast Secretion Trap Screen Assay
The secretion activity of SsCut1 was analyzed by using the yeast secretion trap assay [61,62]. In this study, the predicted signal peptide fragment of the SsCut1 gene was fused to the N-terminal of secretion-defective invertase gene (suc2) in the vector pSUC2 and Then transformed into the yeast competent strain YTK12. The candidate yeast transformants were cultured on YPDA medium, and selected on CMD-W (Takara, Dalian, China) medium and YPRAA (2% raffinose) medium (Macklin, Shanghai, China). TTC (2, 3, 5-Triphenyte-trazoliumchloride) was used to detect the activity of the secreted sucrase of the candidate yeast transformants. The candidate yeast transformed strain was cultured on sucrose medium, then collected by centrifugation, added with a final concentration of 0.1% TTC reagent, incubated at 35 • C for 35 min, and placed at room temperature for 5 min to observe the color change in the test tube. A positive reaction changes from colorless to dark red. The SPs of Avr1b and YTK12-pUSC2 were used as the positive and negative controls, respectively.

Gene Deletion and Complementation
The cutinase gene SsCut1 was disrupted by using the split marker system [63]. The strategy for the deletion of SsCut1 is illustrated in Supporting Information Figure S2A. To generate SsCut1-deletion mutants, the 5 and 3 flanking fragments of the ORF of SsCut1 were amplified from S. sclerotiorum wild-type strain 1980 genomic DNA, then cloned into the pMD19-T plasmid using the T/A clone ligation kit (Takara, Shiga, Japan). Two flanking sequences with a truncated hygromycin-resistant gene sequence were named 5 -HY and YG-3 , respectively. The 5 -HY and YG-3 fragments were amplified and purified. The purified fragments of 5 -HY and YG-3 were mixed in equal molarity, and used for protoplast transformation [64]. ∆SsCut1 mutants were selected on plates containing hygromycin and confirmed by PCR with genomic DNA as template. Homokaryotic ∆SsCut1 mutants were obtained by single ascospore isolation. The homokaryotic state of the ∆SsCut1 mutants was confirmed by Southern blot. The genomic DNA was extracted from mycelium and digested with SpeI. Southern blot analysis with SsCut1 as a probe was detected by using Amersham AlkPhos Direct Labelling and Detection Systems (GE Healthcare, Piscataway, NJ, USA).
For the complementation of SsCut1, the binary vector pCENTS-SsCut1 was constructed. The complementary fragment was amplified from the S. sclerotiorum genomic DNA, including the open reading frame of SsCut1 with an approximately 1200 bp native promoter region and a 200 bp terminator region. The pCENTS was cut by XhoI and SpeI to remove the EF-1α promoter and then ligated with the complementary fragment. The complementary vector was transformed into Agrobacterium strain EHA105. Agrobacterium tumefaciens-mediated transformation (ATMT) of S. sclerotiorum was performed as described [65]. Cultivate the Agrobacterium broth and the mycelium blocks of cutinase knock-out mutants in the dark condition for three days. Then, cover with the PDA medium containing hygromycin B and G418 at 100 µg/mL resistance. The DNA of candidate complementary transformants were extracted by using the CTAB method for validation.

Plant Growth and Pathogenicity Assays
The A. thaliana and N. benthamiana were grown at 20 • C in growth chambers or glasshouse under long day (16 h: 8 h, light: dark) conditions. Pathogenicity assays of S. sclerotiorum were tested with mycelia plugs (diameter = 4.0 mm) from the margins of actively growing colonies on PDA medium. Inoculated tissues were incubated at 21 • C for 48 h before image acquisition and measurement of lesion area. The treatment was replicated 10 times and the experiments were repeated three times.

Generation of Plant Expression Plasmids
In order to express SsCut1 in N. benthamiana, a plasmid with the backbone of pCNF3 and pCNG were used [9]. The recombinant plasmid pCNF3-SsCut1 and pCNG-SsCut1 were transformed into Agrobacterium strains GV3101 for expression. The protein of N. benthamiana were extracted for Western blot at 48 h post-infiltration. The leaf specimens of N. benthamiana were prepared to observe the protein subcellular localization by using the confocal microscope (Olympics).

Cutinase Activity Assay
The ELISA Reagent test kit (Sinobestbio, Shanghai, China) was used to determine the cutinase activity of fungal hypha samples. The HRP-labeled detection antibody was added to the coated microwells pre-coated with cutinase antibody, incubated and washed thoroughly. The color is developed with the substrate TMB. TMB is converted into a blue color under the catalysis of peroxidase, and converted into the final yellow under the action of acid. The intensity of the color is positively correlated with the cutinase activity in the sample. The absorbance (OD value) at a wavelength of 450 nm was measured with a microplate reader to calculate the cutinase activity in the sample.

SsCut1 Contains Cutinase Domain and Fungal Cellulose-Binding Domain
Sequence analysis showed that there are eight genes encoding cutinase proteins in the genome of S. sclerotiorum (Supplementary Figure S1). SS1G_08104 is one of the cutinaseencoding genes, which was named SsCut1 in this study. SsCut1 is 1077 bp in length and contains three exons and two introns. Its open reading frame encodes a 300 aa protein with a predicted N-terminal SP. SsCut1 possesses fourteen cysteine residues. Phylogenetic analysis showed that the homologous protein of SsCut1 was widely found in B. cinerea and other pathogenic fungi, but it did not cluster with the reported cutinase SsCut, BcCutA and BcCutB ( Figure 1A). Protein structure analysis showed that the SsCut1 had conserved GYSQG catalytic site; however, it differed from SsCut, BcCutA and BcCutB in in certain features, consistent with the results of cluster analysis. SsCut1 contains a fungal cellulosebinding (CBM1) domain which was absent in SsCut, BcCutA and BcCutB ( Figure 1B). The expression pattern of SsCut1 was detected by RT-qPCR. The results revealed that the expression of SsCut1 was relatively stable during the growth and development stage of S. sclerotiorum ( Figure 2A). However, during the infection stage, the transcript level of SsCut1 was upregulated and increased by more than 100 fold at 24 h post-inoculation (hpi) on A. thaliana ( Figure 2B), suggesting that SsCut1 may be involved in infection by S. sclerotiorum.

SsCut1 Is a Secreted Protein and Localized in the Plant Cell Wall
To investigate whether SsCut1 is normally secreted by S. sclerotiorum, the function of SsCut1 signal peptide was verified by the yeast secretion trap system [61,62]. The Avr1b was used as positive control and the YTK12 empty strain and YTK12 with pSUC2 vector were used as negative controls. The invertase mutant yeast strain YTK12 of pSUC2-Avr1b and pSUC2-SsCut1sp could regularly grow on CMD-W and YPRAA media ( Figure 3A). When treating with 2, 3, 5 triphenyltetrazolium chloride (TTC), the mutant strains of pSUC2-Avr1b and pSUC2-SsCut1sp generated the insoluble red-colored triphenylformazan, but the yeast strains YTK12 and pSUC2 remained colorless after treatment ( Figure 3B). These results indicated that the signal peptide of SsCut1 was functional in the yeast secretion trap system, and SsCut1 might be a secreted protein.
SsCut1 contains a fungal cellulose-binding domain of CBM1 superfamily at the 268-296 amino acid positions ( Figure 1B). This domain has high specific affinity to cellulosics [66]. We speculated that SsCut1 could localize in the plant cell wall. In order to clarify the localization of SsCut1 in plant cells, we performed a protein cellular localization assay in N. benthamiana. Fluorescence observation results showed that SsCut1 can be localized in the cell wall when expressed in plant cells ( Figure 3C).     β-tubulin gene was used as an internal control to normalize the data. Error bars represent the standard error (n = 3).

SsCut1 Is a Secreted Protein and Localized in the Plant Cell Wall
To investigate whether SsCut1 is normally secreted by S. sclerotiorum, the function of SsCut1 signal peptide was verified by the yeast secretion trap system [61,62]. The Avr1b was used as positive control and the YTK12 empty strain and YTK12 with pSUC2 vector were used as negative controls. The invertase mutant yeast strain YTK12 of pSUC2-Avr1b and pSUC2-SsCut1sp could regularly grow on CMD-W and YPRAA media ( Figure 3A). When treating with 2, 3, 5 triphenyltetrazolium chloride (TTC), the mutant strains of pSUC2-Avr1b and pSUC2-SsCut1sp generated the insoluble red-colored triphenylformazan, but the yeast strains YTK12 and pSUC2 remained colorless after treatment ( Figure  3B). These results indicated that the signal peptide of SsCut1 was functional in the yeast secretion trap system, and SsCut1 might be a secreted protein.
SsCut1 contains a fungal cellulose-binding domain of CBM1 superfamily at the 268-296 amino acid positions ( Figure 1B). This domain has high specific affinity to cellulosics [66]. We speculated that SsCut1 could localize in the plant cell wall. In order to clarify the localization of SsCut1 in plant cells, we performed a protein cellular localization assay in N. benthamiana. Fluorescence observation results showed that SsCut1 can be localized in the cell wall when expressed in plant cells ( Figure 3C).

SsCut1 Is Important for Virulence of S. sclerotiorum
The biological function of SsCut1 in S. sclerotiorum was investigated by generating SsCut1-deletion mutants, using a homologous recombination strategy (Supplementary Figure S2A). Homokaryotic ΔSsCut1 mutants were obtained by single ascospore isolation. Three SsCut1-deletion mutants (ΔSsCut1-1, ΔSsCut1-3 and ΔSsCut1-7) were obtained, and confirmed by PCR and Southern blot analysis. The results showed that the targeted gene was replaced by a single copy marker gene in mutants ΔSsCut1-1 and ΔSsCut1-3 (Supplementary Figure S2B,C). Two complementary transformants, ΔSsCut1-3-C8 and ΔssCut1-
Deletion of Sscut1 had no effect on the saprophytic growth and life cycle of S. sclerotiorum. ∆SsCut1 mutants showed no significant differences from the wild-type strain in colony morphology, oxalic acid production, infection cushion formation, sclerotial development and growth rate (Figure 4 and Figure S3). To determine the function of the SsCut1 in the virulence of S. sclerotiorum, the wild-type strain (1980), SsCut1 knock-out mutants, and the complementary transformants were inoculated on the leaves of oilseed rape and A. thaliana plants. Compared with the wild-type strain, both knock-out mutants caused smaller lesions on the oilseed rape and A. thaliana leaves. Complementation of SsCut1 in the knock-out strain restored virulence to the level of the wild-type strain ( Figure 5 and Figure S4). These results indicated that SsCut1 is not required for completion of life cycles, but required for full virulence of S. sclerotiorum.

of 16
Deletion of Sscut1 had no effect on the saprophytic growth and life cycle of S. sclerotiorum. ΔSsCut1 mutants showed no significant differences from the wild-type strain in colony morphology, oxalic acid production, infection cushion formation, sclerotial development and growth rate (Figures 4 and S3). To determine the function of the SsCut1 in the virulence of S. sclerotiorum, the wild-type strain (1980), SsCut1 knock-out mutants, and the complementary transformants were inoculated on the leaves of oilseed rape and A. thaliana plants. Compared with the wild-type strain, both knock-out mutants caused smaller lesions on the oilseed rape and A. thaliana leaves. Complementation of SsCut1 in the knock-out strain restored virulence to the level of the wild-type strain (Figures 5 and S4). These results indicated that SsCut1 is not required for completion of life cycles, but required for full virulence of S. sclerotiorum.

Deletion of SsCut1 Affects the Cutinase Activity of S. sclerotiorum
To investigate if SsCut1 is associated with the cutinase activity in S. sclerotiorum, wildtype strain and different mutants of SsCut1 were cultured on the PDA medium for 36 h. Then, the hyphae were collected for determination of cutinase activity. Results showed that the cutinase activity of SsCut1 knock-out transformants were significantly decreased compared with wild-type strain, while the cutinase activity of SsCut1 complementary transformants were recovered to the level of wild-type strain (Figure 6), indicating SsCut1 does encode a cutinase in S. sclerotiorum. The decrease in virulence of SsCut1 knock-out transformants should be related to the decrease in cutinase activity.

Deletion of SsCut1 Affects the Cutinase Activity of S. sclerotiorum
To investigate if SsCut1 is associated with the cutinase activity in S. sclerotiorum, wildtype strain and different mutants of SsCut1 were cultured on the PDA medium for 36 h. Then, the hyphae were collected for determination of cutinase activity. Results showed that the cutinase activity of SsCut1 knock-out transformants were significantly decreased compared with wild-type strain, while the cutinase activity of SsCut1 complementary transformants were recovered to the level of wild-type strain (Figure 6), indicating SsCut1 does encode a cutinase in S. sclerotiorum. The decrease in virulence of SsCut1 knock-out transformants should be related to the decrease in cutinase activity.

The Expression of SsCut1 in Plants Triggers Defense Responses
To evaluate whether SsCut1 participates in Sclerotinia-plant interaction, transient expression of SsCut1 was carried out in N. benthamiana. ROS burst was observed when we transiently expressed SsCut1-flag in N. benthamiana compared with that expressing mock vector ( Figure 7A). Furthermore, the expression levels of NbPR1 were also significantly upregulated when SsCut1-flag was expressed ( Figure 7B). These results suggested that the Figure 6. Cutinase activity of wild-type strain and SsCut1 transformants. All strains were cultured on PDA medium. The hyphal of wild-type strain and SsCut1 transformants was collected at 36 hpi. Cutinase activity levels were examined with the enzyme-linked immunosorbent assay (ELISA) method. Bars indicate ± SE. Statistical significance is indicated in the graph (one-way ANOVA): ** p < 0.01, *** p < 0.001.

The Expression of SsCut1 in Plants Triggers Defense Responses
To evaluate whether SsCut1 participates in Sclerotinia-plant interaction, transient expression of SsCut1 was carried out in N. benthamiana. ROS burst was observed when we transiently expressed SsCut1-flag in N. benthamiana compared with that expressing mock vector ( Figure 7A). Furthermore, the expression levels of NbPR1 were also significantly upregulated when SsCut1-flag was expressed ( Figure 7B). These results suggested that the expression of SsCut1 in N. benthamiana can trigger plant defense responses. Interestingly, the lesion size of ∆SsCut1-3 mutant on SsCut1-expressing plants was similar to that of the wild-type strain level ( Figure 7C, D), suggesting that SsCut1 expressed in plants can complement the virulence defect of the SsCut1-deletion mutant. It is another line of evidence that SsCut1 is crucial to the virulence of S. sclerotiorum and plays an important role in the Sclerotinia-plant interaction.

Discussion
Cutinase is important for the virulence of many fungal pathogens. The S. sclerotiorum genome contains eight cutinase genes [54], but only SsCut has been studied [55,56]. However, SsCut was only expressed in N. benthamiana to observe the plant defense response,

Discussion
Cutinase is important for the virulence of many fungal pathogens. The S. sclerotiorum genome contains eight cutinase genes [54], but only SsCut has been studied [55,56]. How-ever, SsCut was only expressed in N. benthamiana to observe the plant defense response, but the effect of SsCut on S. sclerotiorum phenotype has not been studied. SsCut1 encodes a cutinase protein in S. sclerotiorum and shows increased expression during the early stage of S. sclerotiorum infection (Figure 2), indicated that SsCut1 plays an important role in the early stage of infection. Therefore, in this study, we explored the effects of SsCut1 on virulence and cutinase activity of S. sclerotiorum.
Most cutinase protein sequences contain a highly conserved GYSQG motif, which is crucial for maintaining the serine activity and function of cutinase [67]. In S. sclerotiorum, all the eight cutinase genes contain the conserved GYSQG domain, but the intron and exon structures of the eight cutinase genes are different (Supplementary Figure S1). It is suggested that the cutinase gene in S. sclerotiorum genome not only keeps the core GYSQG structure, but also increases the universality and adaptability of cutinase, which is beneficial to the expansion and stability of the host range of pathogenic fungus. The amino acid sequence alignment also shows that although the protein sequences of SsCut1, SsCut, BcCutA and BcCutB are highly conserved in key domains, there are some regions with unknown functions at the C-terminal of SsCut1 (Figure 1), and it is also unknown whether these regions endow new functions to SsCut1.
Plant epidermis layers, mainly composed of cutin, wax, and cellulose, are the first barrier for fungi to infect plants [68,69]. Plant pathogenic fungi, especially necrotrophic fungi, can destroy the integrity of plant cell walls through combined actions of a large number of cell wall-degrading enzymes to facilitate invasion. During this process, cutinase plays key roles and is usually released at the early stage of infection to degrade the plant cuticle, which is the first barrier of plant [70]. Therefore, knocking out cutinase or destroying its activity often has a great influence on the virulence of plant pathogenic fungi. For example, deletion of genes encoding cutinase reduced virulence in Colletotrichum truncatum, C. gloeosporioides, Botryosphaeria dothidea and Arthrinium phaeospermum [40,45,71,72]. Using enzyme inhibitors to inhibit cutinase activity can also prevent Fusarium solani pisi from infecting plants [73]. We also found that SsCut1 knock-out transformants have no difference in growth rate, colony morphology, infection cushions formation, sclerotia formation, and oxalic acid production compared with wild-type strain (Figure 4 and Figure S3). However, the virulence and cutinase activity of the SsCut1 knock-out transformants were significantly reduced, while there was no obvious difference between complementary transformants and wild-type strain (Figures 5 and 6). Considering the high-level expression of SsCut1 in the S. sclerotiorum infection process, our results suggest that the SsCut1 has cutinase activity and is necessary for full virulence of S. sclerotiorum. There are reports that knocking out genes encoding cutinase has no obvious effect on the virulence of some plant pathogenic fungi [46][47][48][49], due to redundancy of gene functions for existence of multiple genes encoding cutinase in the genome. There are also eight cutinase genes in S. sclerotiorum [54], which is why the virulence of S. sclerotiorum only decreased after SsCut1 knock-out, but not completely lost.
During the interaction between plants and pathogenic fungi, cutinase can not only degrade the cuticle of plants, but also be recognized by plants as a PAMP (Pathogenassociated molecular patterns), thus activating the immune response [40,52,53,56,74]. For example, overexpression of the cutinase gene CUTE (cutinase-expressing) in Arabidopsis can improve the resistance of plants to B. cinerea [35]. Overexpression of a cutinase gene from B. dothidea (Bdo_10846) in N. benthamiana induced ROS burst and callose deposition, and also enhanced N. benthamiana resistance to B. cinerea [40]. The SsCut protein of S. sclerotiorum can also cause cell death in soybean and rapeseed, and induce resistance to S. sclerotiorum and B. cinerea in these plants [56]. In this study, we found that the expression of SsCut1 in N. benthamiana induced ROS burst, and upregulated expression of host pathogenesisrelated genes ( Figure 7A,B). However, when the wild-type strain 1980 was inoculated, the disease lesions on the plants expressing SsCut1 did not increase compared with the empty vector-expression plants, indicating that the resistance of these plants to S. sclerotiorum did not change significantly ( Figure 7C,D). This may be due to the strong virulence of S. sclerotiorum or the existence of multiple cutinase in the genome of S. sclerotiorum. SsCut1 and other cutinase in S. sclerotiorum may also be recognized by plants during infection, thus inducing plant immune response. For example, the expression of SsCut also can trigger plant immune response [56], so the immune response induced by SsCut1 expression in plants may not be enough to affect the virulence of S. sclerotiorum. Interestingly, the virulence of the SsCut1 knock-out mutant on SsCut1-expressing plants recovered to the level of wild-type strain 1980, indicating that SsCut1 is important to the virulence of S. sclerotiorum and plays an important role in plants during infection.

Conclusions
In summary, our study revealed that SsCut1 is important for the cutinase activity and full virulence of S. sclerotiorum, and plays important roles in triggering plant defense responses. Here, we identified a conserved cutinase gene, SsCut1 from S. sclerotiorum. SsCut1 was found to be expressed highly in the early stage of S. sclerotiorum infection. SsCut1 had secretory activity and was mainly located in the plant cell wall. Knocking out SsCut1 reduced the virulence and cutinase activity of S. sclerotiorum, but had no effect on the growth rate, colony morphology, oxalic acid production, infection cushion formation and sclerotial development. Expression of SsCut1 in plants can trigger defense responses, such as induced ROS burst and upregulate the expression of pathogenesis-related genes. However, it also enhances plant susceptibility to SsCut1 gene knock-out mutants. These findings improve our understanding of the roles of cutinase gene during the interaction of S. sclerotiorum with its host plants, and also offer clues to further reveal the mechanisms of these genes in infection.