A Small Secreted Virulence-Related Protein Is Essential for the Necrotrophic Interactions of Sclerotinia sclerotiorum with Its Host Plants

Abstract

Small, secreted proteins have been found to play crucial roles in interactions between biotrophic/hemi-biotrophic pathogens and plants. However, little is known about the roles of these proteins produced by broad host-range necrotrophic phytopathogens during infection. Here, we report that a cysteine-rich, small protein SsSSVP1 in the necrotrophic phytopathogen Sclerotinia sclerotiorum was experimentally confirmed to be a secreted protein, and the secretion of SsSSVP1 from hyphae was followed by internalization and cell-to-cell movement independent of a pathogen in host cells. SsSSVP1∆SP could induce significant plant cell death and targeted silencing of SsSSVP1 resulted in a significant reduction in virulence. Through yeast two-hybrid (Y2H), coimmunoprecipitation (co-IP) and bimolecular fluorescence complementation (BiFC) assays, we demonstrated that SsSSVP1∆SP interacted with QCR8, a subunit of the cytochrome b-c1 complex of mitochondrial respiratory chain in plants. Double site-directed mutagenesis of two cysteine residues (C38 and C44) in SsSSVP1∆SP had significant effects on its homo-dimer formation, SsSSVP1∆SP-QCR8 interaction and plant cell death induction, indicating that partial cysteine residues surely play crucial roles in maintaining the structure and function of SsSSVP1. Co-localization and BiFC assays showed that SsSSVP1∆SP might hijack QCR8 to cytoplasm before QCR8 targeting into mitochondria, thereby disturbing its subcellular localization in plant cells. Furthermore, virus induced gene silencing (VIGS) of QCR8 in tobacco caused plant abnormal development and cell death, indicating the cell death induced by SsSSVP1∆SP might be caused by the SsSSVP1∆SP-QCR8 interaction, which had disturbed the QCR8 subcellular localization and hence disabled its biological functions. These results suggest that SsSSVP1 is a potential effector which may manipulate plant energy metabolism to facilitate the infection of S. sclerotiorum. Our findings indicate novel roles of small secreted proteins in the interactions between host-non-specific necrotrophic fungi and plants, and highlight the significance to illuminate the pathogenic mechanisms of this type of interaction.

Fig 1. SsSSVP1 is a Sclerotinia- and Botryotinia-specific, cysteine-rich, small, secreted protein.

(A) A predicted structure diagram of SsSSVP1 which comprises 163 aa. A putative N-terminal SP (aa 1 to 17) and the position of the eight cysteine residues of SsSSVP1 are present (C³⁸, C⁴⁴, C⁵⁴, C⁷⁹, C⁸¹, C⁹², C⁹⁵ and C¹¹⁷). (B) Multiple alignments indicate the homologs of SsSSVP1 are only present in Sclerotinia- and Botryotinia in the organisms sequenced so far and the eight cysteine residues are conserved in these homologs. Red rectangle labels the sites of the eight cysteine residues in multiple alignments. Protein sequences from top to bottom are derived from B. cinerea T4, B. cinerea B05.10, B. cinerea BcDW1, Sclerotinia borealis F-4157 and S. sclerotiorum Ep-1PNA367 respectively. The protein sequences of SsSSVP1 in S. sclerotiorum Ep-1PNA367 and 1980 are the same. (C) Western blot analysis with total proteins isolated from the liquid CM culture of the wild-type strain and SsSSVP1-FLAG engineered strains. SDS-polyacrylamide gel electrophoresis shows the equal loading amount of proteins used for the west blot analysis. Horseradish peroxidase conjugated secondary antibody detected an approximate 17 kDa band in SsSSVP1-FLAG engineered strains, but not in the wild-type strain.

Fig 2. Induction of cell death and the subcellular localization of SsSSVP1^∆SP in plant cells.

(A) SsSSVP1^∆SP induces significant systemic plant cell death. A. tumefaciens containing systemic expression vector pTRV2-SsSSVP1^∆SP and pTRV1, respectively, were mixed in equal proportions and infiltrated into lower leaves of the wild-type N. benthamiana. The leaves and stems were from above the infiltrated sites. Photos were taken 15 days after A. tumefaciens infiltration. Red arrows indicate infiltration sites. (B) Laser confocal micrograph showing SsSSVP1^∆SP mainly distributed in the cytoplasm, and especially concentrated in the periphery of cytomembrane. Red particles showed chloroplast autofluorescence. Photos were taken 3 days after agroinfiltration. Maximum projections of 4 confocal images captured along the z-axis are shown.

Fig 3. Full SsSSVP1 could still induce plant cell death and it can be internalized into plant cells in the absence of a pathogen.

(A) SsSSVP1 with SP still can induce cell death in leaves. Upper leaves from above the infiltrated sites were taken photos 10 days after A. tumefaciens infiltration. (B) SsSSVP1 with SP still can induce cell death in stems. GFP alone was used as control. Photos were taken 10 days after A. tumefaciens infiltration. (C) Both SP-GFP (which was used as control) and SsSSVP1 with SP localized in ER-like structure in plant cells (details see S2 Fig), however, only full SsSSVP1 could also localize in cytoplasmic compartments in a particle like form. No particle-like form of SP-GFP was observed in cytoplasm. The SP refers in particular to the SP of SsSSVP1. Red particles show chloroplast autofluorescence. White solid arrows indicate the internalized particle-like form of SsSSVP1-GFP; White hollow arrows show the endocytic vesicle-like structure near plasma membrane. Photos were taken 3 days after agroinfiltration.

Fig 4. The nuclear targeting-based translocation assay of SsSSVP1-mCherry.

B. cinerea wild-type strain B05.10 and transformants of SP-mCherry-NLS and SsSSVP1-mCherry-NLS constructs were used to perform nuclear targeting assay, and the former two were used as controls. SsSSVP1-mCherry-NLS fluorescence occurred in the nuclei of invaded onion bulb lower epidermal cells and surrounding cells while mCherry-NLS fluorescence occurred only in the nuclei of invaded cells. No fluorescence was observed in the onion tissues infected by B05.10. The same imaging conditions were used in the three channels. Images were taken at 48 hpi using confocal laser scanning microscopy. Different layers of the intact surrounding cells were observed independently to ensure there were no hyphae in these cells. See S3 Fig for an example. The images show maximum projections of 4 confocal images captured along the z-axis. Areas within yellow dotted line indicate hyphal invaded onion epidermal cells.

Fig 5. Gene expression analysis of SsSSVP1 in the wild-type strain Ep-1PNA367 during infection.

The relative expression of SsSSVP1 is significantly up-regulated during the early stages of infection in A. thaliana (Col-0) leaves (red columns) compared to that during vegetative growth on PDA (dark columns). The expression level at 0 hpi on PDA was set as 1.0. The expression levels of β-tubulin are used to normalize the expression levels of SsSSVP1 in different samples. Three independent replicates were performed. The bars represent the mean relative expression of SsSSVP1 ± the standard deviation of the mean.

Fig 6. Phenotypes of SsSSVP1-silenced transformants of S. sclerotiorum.

(A) The colony morphology of SsSSVP1-silenced transformants. Colonies were grown on PDA for 10 days at 20°C. (B) SsSSVP1-silenced transformants showed significantly reduced virulence on detached oilseed rape (B. napus zhongyou 821) leaves. Virulence was evaluated on detached oilseed rape leaves according to the lesion diameter. Photos were taken at 48 hpi. (C) The relative expression of SsSSVP1 in silenced transformants was determined through qRT-PCR. The expression levels of β-tubulin were used to normalize the expression levels of SsSSVP1 in the different samples. The expression level in the wild-type strain was set as 1.0. (D) Comparison of the lesion diameter of silenced transformants and the wild-type strain. (E) Comparison of the growth rate of silenced transformants and the wild-type strain. In all experiments, three independent replicates were performed. The values are presented as the means±s.d. Different letters on the same graph indicate statistical significance, P = 0.05.

Fig 7. SsSSVP1^∆SP can form homo-dimer and interact with QCR8 of A. thaliana.

(A) Y2H assay showed SsSSVP1^∆SP formed a homo-dimer and interacted with QCR8. pGBKT7-53 and pGADT7-T (Clontech) were used as positive control for protein-protein interaction. “-” means there is an empty vector. The negative controls indicated SsSSVP1^∆SP and QCR8 were not self-activated. Photos were taken 2 dpi. (B) Co-IP assay for SsSSVP1^∆SP and QCR8 interaction in N. benthamiana leaves. GFP-tagged SsSSVP1^∆SP but not GFP alone interacts with 3×FLAG-tagged QCR8. IP = immunoprecipitation, IB = immunoblot. (C) BiFC confirms SsSSVP1^∆SP interacts with QCR8 in cytoplasm of plant cells. Red particles show chloroplast autofluorescence. Fluorescence was monitored 3 days after agroinfiltration using confocal laser scanning microscopy. The images show maximum projections of 4 confocal images captured along the z-axis.

Fig 8. C³⁸ and C⁴⁴ play crucial roles in the dimer formation, the SsSSVP1^∆SP-QCR8 interaction and the biological function of SsSSVP1^∆SP.

The functional analysis of the eight cysteine residues of SsSSVP1^∆SP in its dimer formation and interaction with QCR8. Y2H result showed single site-directed mutation of the eight cysteine residues had little effect on the dimer formation (A) and the interaction of SsSSVP1^∆SP with QCR8 (B). However, if C³⁸ and C⁴⁴ were simultaneously substituted with alanine, the double-point mutant SsSSVP1^{∆SP-C38A-C44A} could not form dimer (A) and interact with QCR8 (B), indicated in red rectangles. Photos were taken 2 dpi. pGBKT7-53 and pGADT7-T were used as positive controls (Clontech). (C) Single site-directed mutation of the eight cysteine residues of SsSSVP1^∆SP had little effect on induction of plant cell death but the double-point mutant SsSSVP1^{∆SP-C38A-C44A} lost the capability to induce plant cell death, indicated in black rectangle. The single- and double-point mutants of SsSSVP1^∆SP were expressed in tobacco leaves individually through A. tumefaciens-mediated plant transformation method. Photos were taken 10 days after A. tumefaciens infiltration. (D) Western blot analysis showed that horseradish peroxidase conjugated secondary antibody could detect an approximate 19 kDa band in tobacco leaf cells expressing SsSsSSVP1^∆SP-3×FLAG and SsSsSSVP1^{∆SP-C38A-C44A}-3×FLAG but not in control (tobacco leaf cells infiltrated with A. tumefaciens with empty vector). SDS-polyacrylamide gel electrophoresis shows the equal loading amount of proteins used for the west blot analysis.

Fig 9. The interaction between SsSSVP1^∆SP and QCR8 disturbs the subcellular localization of QCR8 in mitochondria.

(A) The fluorescent localization of QCR8 in mitochondria. The red particles show chloroplast autofluorescence. (B) SsSSVP1^∆SP hijacks QCR8 into cytoplasm before it targets to mitochondria. QCR8 and mitochondria marker (Mt-mCherry) co-localize in mitochondria, while QCR8 and SsSSVP1^∆SP most commonly co-localize in cytoplasm. However, the double site-directed mutant SsSSVP1^{∆SP-C38A-C44A} losing the ability to interact with QCR8 did not affect the subcellular localization of QCR8. Fluorescence was monitored 3 days after agroinfiltration using confocal laser scanning microscopy. The images show maximum projections of 4 confocal images captured along the z-axis.

Fig 10. Silencing of QCR8 leads to plant abnormal development and cell death.

(A) The relative expression levels of three QCR8 genes (QCR8-1, QCR8-2 and QCR8-3) in silenced N. benthamiana lines were determined through qRT-PCR. The expression levels of the actin gene in N. benthamiana were used to normalize the expression levels of QCR8. The QCR8 expression level in the control lines was set as 1.0. This qRT-PCR assay was performed one month after A. tumefaciens infiltration. “Up” and “M” indicated these samples were from the upper leaves and middle leaves, respectively. “CK” and “RNAi” indicated these samples were from the control lines and the silenced lines which were infiltrated with A. tumefaciens containing empty vectors and silencing vectors, respectively. (B) The phenotype of QCR8-silenced N. benthamiana lines (RNAi-QCR8) using TRV based VIGS system. The lines transformed using A. tumefaciens containing VIGS-pTRV2 empty vector were used as control. Silencing of the QCR8 in N. benthamiana plants caused growth retardation and cell death. No obvious phenotype was observed in the control. Photos were taken 10 days post A. tumefaciens infiltration.