SGT1 is required in PcINF1/SRC2-1 induced pepper defense response by interacting with SRC2-1

PcINF1 was previously found to induce pepper defense response by interacting with SRC2-1, but the underlying mechanism remains uninvestigated. Herein, we describe the involvement of SGT1 in the PcINF1/SRC2-1-induced immunity. SGT1 was observed to be up-regulated by Phytophthora capsici inoculation and synergistically transient overexpression of PcINF1/SRC2-1 in pepper plants. SGT1-silencing compromised HR cell death, blocked H2O2 accumulation, and downregulated HR-associated and hormones-dependent marker genes’ expression triggered by PcINF1/SRC2-1 co-overexpression. The interaction between SRC2-1 and SGT1 was found by the yeast two hybrid system and was further confirmed by bimolecular fluorescence complementation and co-immunoprecipitation analyses. The SGT1/SRC2-1 interaction was enhanced by transient overexpression of PcINF1 and Phytophthora capsici inoculation, and SGT1-silencing attenuated PcINF1/SRC2-1 interaction. Additionally, by modulating subcellular localizations of SRC2-1, SGT1, and the interacting complex of SGT1/SRC2-1, it was revealed that exclusive nuclear targeting of the SGT1/SRC2-1 complex blocks immunity triggered by formation of SGT1/SRC2-1, and a translocation of the SGT1/SRC2-1 complex from the plasma membrane and cytoplasm to the nuclei upon the inoculation of P. capsici. Our data demonstrate that the SGT1/SRC2-1 interaction, and its nucleocytoplasmic partitioning, is involved in pepper’s immunity against P. capsici, thus providing a molecular link between Ca2+ signaling associated SRC2-1 and SGT1-mediated defense signaling.


Results
Transcriptional expression of SGT1 in response to Phytophthora capsici inoculation and PcINF1/SRC2-1 transient overexpression. Phytophthora capsici is a causal agent of phytophthora blight of pepper, one of the most important diseases in pepper production, and elicitins are conserved extracellular proteins in P. capsici and other Phytophthora and Pythium spp. triggering immunity in the host plants. Our previous study found that SRC2-1 from pepper plants perceive and interact with PcINF1 and trigger HR cell death and immunity 14 , but the underlying mechanism remains unknown. To isolate potential SRC2-1-interacting partners, a pepper cDNA library generated from P. capsici-inoculated pepper leaves was previously screened with SRC2-1 as a bait using GAL4-based yeast two-hybrid system, one positive clone acquired was identified to be SGT1. Since SGT1 have been implicated in plant immunity 40 , the expression of SGT1 was assayed by real-time PCR in P. capsici-inoculated pepper plants to test its potential role in immunity against P. capsici inoculation. The result showed that the expression of SGT1 was enhanced compared to the mock treatment (Fig. 1a). Our previous study found that PcINF1, an elicitin from P. capsici, triggered HR cell death and defense response in pepper plants by interacting with SRC2-1 14 . To test if SGT1 is involved in the immunity mediated by PcINF1/ SRC2-1, an Agrobacterium-mediated transient expression system was employed to characterize the effects of PcINF1/SRC2-1 simultaneously transient overexpression on the expression of SGT1. Agrobacterial cells containing 35S::PcINF1/35S::SRC2-1 fusion vector were infiltrated into pepper leaves, and the transcript level of SGT1 was assayed by real-time RT-PCR (Fig. 1b). The result showed that the transcript levels of PcINF1 and SRC2-1 were both significantly increased in the pepper leaves transiently expressing PcINF1 and SRC2-1 ( Supplementary  Fig. S1), additionally, SGT1 was also upregulated by the PcINF1/SRC2-1 transient overexpression, indicating that SGT1 may be involved in pepper immunity mediated by PcINF1/SRC2-1.
The Silencing of SGT1 compromises the immunity triggered by PcINF1/SRC2-1. To further confirm the potential role of SGT1 in the defense response triggered by PcINF1/SRC2-1, the silencing of SGT1 was performed by a VIGS (virus induced gene silence), which has been successfully used in Solanaceae, such as pepper, tomato, and tobacco 14,[41][42][43] . To avoid the mistargeting of any potential orthologs of SGT1, fragment of 213 bps in length corresponding to a specific region of the 3′ UTRs of SGT1 was chosen to construct VIGS vectors (The specificity of the region was confirmed by BLASTN in pepper genome databases: http://passport.pepper. snu.ac.kr/ ? t=PGENOME and http://peppersequence.genomics.cn/page/species/index.jsp). The efficiency of SGT1 silencing was examined by the real time RT-PCR and the result showed that the expression of SGT1 was significantly decreased in the TRV:SGT1 pepper plants ( Supplementary Fig. S2). In SGT1-silenced pepper plants, the HR cell death manifested with the darker trypan blue staining triggered by the transient co-overexpression of PcINF1/SRC2-1 significantly decreased compared to the control plants (Fig. 2a), and was consistent with an increased ion leakage in the PcINF1/SRC2-1 co-overexpressed and SGT1-silenced pepper plants (Fig. 2b). In addition, the H 2 O 2 accumulation triggered by the co-overexpression of PcINF1/SRC2-1 manifested by the DAB staining (Fig. 2a), which is closely related to HR cell death by previous studies 40,44 , was also found to be compromised by SGT1 silencing. Furthermore, the SA-dependent SAR82A, BPR1 and PR4, JA-dependent DEF1, HR associated marker gene HIR1 and defense-associated PO2 (triggered by the PcINF1/SRC2-1 co-overexpression) were also consistently compromised by the silencing of SGT1 (Fig. 2c). All these results suggest that SGT1 is involved in pepper immunity mediated by PcINF1/SRC2-1.
Simultaneously transient overexpression of SGT1 and SRC2-1 induced heightened cell death and defense responses in pepper leaves. Since both SRC2-1 and SGT1 are involved in the pepper defense response against P. capsici inoculation as well as the simultaneously transient overexpression of PcINF1/ SRC2-1 14 , it appears that these two proteins are functionally related. To test this possibility, the effect of the simultaneously transient co-overexpression of SRC2-1 and SGT1 on the defense response in pepper plants was assayed. Agrobacterial cells containing 35S::SGT1 and 35S::SRC2-1 were infiltrated into pepper leaves individually or simultaneously. Infiltrated pepper leaves were then harvested at either 48 HAI (for trypan blue staining, DAB staining, and conductivity measurements), or at 12, 24, and 48 HAI for total RNA isolation to detect the transcriptional expression of defense-associated marker genes. Our results showed that both SGT1 and SRC2-1 were successfully overexpressed in the pepper leaves infiltrated with agrobacteria cells carrying 35S::SGT1 and 35S::SRC2-1 ( Supplementary Fig. S3), besides, transient co-overexpression of SGT1 and SRC2-1 in pepper leaves induced significant HR-like cell death displayed by the trypan blue staining, the accumulation of H 2 O 2 , and the ion leakage measurement ( Fig. 3a/b). This was coupled with significant up-regulation of HR-marker genes HIR1, SA-dependent SAR82A, JA-dependent DEF1, and active oxygen metabolism-associated PO2. Up-regulation of these genes occurred at 12, 24, or 48 HAI with maximum expression occurring at different time points for each gene (Fig. 3c). Additionally, the heightened cell death and the accumulation of H 2 O 2 , as well as the up-regulation of HR-, SA-, and JA-associated marker genes DEF1, PO2, HIR1, and SAR82A, were triggered by the simultaneous co-overexpression of SGT1 and SRC2-1.
Simultaneous silencing of SGT1 and SRC2-1 compromises immunity to a greater extent compared to single gene silencing in pepper plants against P. capsici inoculation. To characterize the functional relationship between SGT1 and SRC2-1 by loss of function by VIGS, GV3101 cells containing the constructs TRV:SGT1 or TRV:SRC2-1 were infiltrated into 4-leaf-age pepper plants individually or simultaneously. The silencing efficiency was examined by qPCR ( Supplementary Fig. S2). The results show that SGT1 and SRC2-1 were successfully silenced in the corresponding pepper plants. The silencing of either SGT1 or SRC2-1 conferred decreased HR cell death and H 2 O 2 accumulation, manifested by weaker trypan blue and DAB staining, and also a decrement on electrolyte leakage (Fig. 4a,b). Additionally, the down-regulation of HR cell death and SA-and JA-dependent immunity-associated marker genes was observed (Fig. 4c). However, simultaneous SGT1 and SRC2-1 silencing reduced cell death indicators and the expression of immunity-associated genes (including SA-dependent SAR82A, HR-marker genes HIR1, JA-dependent DEF1, and active oxygen metabolism-associated PO2) to a much lower level compared to that resulting from the individual silencing of either gene alone or the unsilenced control plants.
Analysis of SGT1 and SRC2-1 inter-dependence in immunity induction using VIGS and transient overexpression. To investigate whether SGT1 and SRC2-1 are interdependent on each other for their roles in pepper immunity, a combination of VIGS and transient overexpression was employed to investigate the function of each protein independently. The results show that in pepper plants with SRC2-1 silenced, but with SGT1 transiently overexpressed, the cell death intensity and electrolyte leakage iron were both lower than that in control plants transiently overexpressing SGT1 (Fig. 5a,b). Additionally, the transcriptional expression of HR-associated HIR1, SA-dependent SAR82A, and JA-dependent DEF1 were significantly lower than that in control plants transiently overexpressing SGT1 (Fig. 5c). Similarly, the cell death intensity and the expression of these immunity-associated genes were also inhibited in the SGT1-silenced pepper plants transiently overexpressing SRC2-1 when compared to the control plants overexpressing SRC2-1 (Fig. 5). SGT1 interacts with SRC2-1. The above data suggests that SGT1 is involved in PcINF1/SRC2-1 mediated pepper immunity, and that SRC2-1 and SGT1 are functionally related with each other. Furthermore, it appears The transcript level of SGT1 in pepper leaves was first silenced using the VIGS technology. The GV3101 cells harboring the constructs of PcINF1/SRC2-1 or empty vector were infiltrated into the leaves of SGT1-silenced and unsilenced pepper leaves. The leaves were collected 48 h after Agro-infiltration for the electrolyte leakage measurement. (c) Relative expression levels of defense-related marker genes in pepper leaves 48 h after Agro-infiltration with GV3101 cells harboring the PcINF1/SRC2-1 fusion vector, compared to empty vector, which was set to a relative expression level of "1". CaActin and 18S ribosomal RNA were used to normalize the transcript level. (b,c) Experiments were repeated with at least four independent biological repetitions. Error bars indicate SD. Asterisks indicate a significant difference using an LSD test. *P < 0.05; **P < 0.01. that SRC2-1 and SGT1 may be functionally related in a protein-protein interacting manner. To test this possibility, the interaction between the two proteins was assayed using the yeast two-hybrid (Y2H) system. The Y2H analysis showed that SRC2-1 interacts with SGT1 (Fig. 6a).
Co-IP assays were used to investigate SGT1/SRC2-1 binding. Co-IP was performed using the same transient co-expression system for leaves of N. benthamiana. This result suggests that SRC2-1 binds to SGT1 (Fig. 6b). The interaction of PcINF1 and SRC2-1 was used as positive control. The negative control experiment, which used GFP-HA and SGT1-Flag, showed no sign of interaction (Fig. 6b).
Additionally, SGT1/SRC2-1 interaction was further confirmed by BiFC analysis. The N-and C-terminal of yellow fluorescent protein (YFP) were fused to SGT1 and SRC2-1 to generate SGT1-YFP N and SRC2-1-YFP C . The reciprocal constructs (SGT1-YFP C and SRC2-1-YFP N ) were also generated. All of the destination constructs were introduced into the Agrobacterium strain GV3101 for Agrobacterium-mediated transient overexpression in N. benthamiana leaves. 48 HAI, YFP signal in N. benthamiana leaves was detected using the leica laser confocal microscopy, with SRC2-1-YFP N /YFP C and SGT1-YFP C /YFP N being used as negative controls. Transient co-expression of SGT1-YFP C /SRC2-1-YFP N and SGT1-YFP N /SRC2-1-YFP C in N. benthamiana leaves resulted in the presence of yellow fluorescence in the membrane, cytoplasm, and nuclei, indicating that SGT1 interacts with SRC2 in the membrane, cytoplasm and the nucleus of the plant cells. The YFP signal was not detected in the negative control (Fig. 6c).
The effect of PcINF1 transient overexpression and Phytophthora capsici inoculation on SRC2-1/ SGT1 interaction. The functional relationship and the direct interaction between SRC2-1 and SGT1 suggest the involvement of a SRC2-1/SGT1 complex in the defense response triggered by PcINF1. To test its possible mode of action, the interaction between SRC2-1 and SGT1 was detected by assays of luciferase complementation imaging (FLuCI) 7 , immunoblot and fluorescent signals detection under microscope in pepper leaves inoculated with P. capsici. For the luciferase complementation imaging assay, the GV3101 cells carry the split-LUC vector (35S::SGT1-NLUC and 35S::SRC2-1-CLUC) were co-infiltrated into the pepper leaves, followed by inoculated with P. capsici. The result showed that the LUC activities was increased in response to P. capsici inoculation  Supplementary Fig. S4). Additionally, the interaction intensity was also found to be enhanced with the increase of dosage of inoculated P. capsici ( Fig. 7b and Supplementary Fig. S4A). Consistently, the interaction intensity between SRC2-1 and SGT1 was also found to be enhanced by transient overexpression of PcINF1 in pepper plants ( Fig. 7c and Supplementary Fig. S4B).

The effect of SGT1-silencing on PcINF1/SRC2-1 interaction. Since the interaction of SRC2-1 with
PcINF1 is required in the defense response triggered by PcINF1 14 . To test if PcINF1/SRC2-1 interaction is affected by SGT1, the effect of SGT1 silencing on PcINF1/SRC2-1 formation was assayed by luciferase complementation analysis and co-IP. For the luciferase complementation analysis, the GV3101 cells carrying the split-LUC vectors (35S::PcINF1-NLUC and 35S::SRC2-1-CLUC) were co-infiltrated into the leaves of SGT1-silenced pepper plants and kept in the greenhouse, 24 h after infiltration, the pepper leaves were harvested for LUC activity assay. The result showed that the interaction intensity of SGT1 and SRC2-1 was significantly decreased in the SGT1-silenced pepper leaves compared to the control plants ( Fig. 8a/b). For the immunoblot assay, GV3101 cells containing PcINF1/SRC2-1-YFP vector (BiFC vector) were infiltrated into leaves of SGT1-silenced pepper plants generated by VIGS, which were harvested at 48 HAI for protein extraction. The extracted proteins were separated by SDS-PAGE, after which the PcINF1/SRC2-1 complex was detected by immunoblotting with anti-GFP antibody (sigma). This result showed that the interaction between PcINF1 and SRC2-1 was decreased by the silencing of SGT1 (Fig. 8c).
Translocation of SGT1/SRC2-1 complex from membrane to nucleus in P. capsici inoculated N. benthamiana leaves. Agrobacterial cells containing constructs of 35S::SGT1-GFP or 35S::SRC2-1-GFP were infiltrated into N. benthamiana leaves, and the subcellular localization of SGT1-GFP and SRC2-1-GFP fused proteins in N. benthamiana leaves was detected using confocal microscopy. Consistent with the previous study 14,15 , the results show that SRC2-1 was invariably targeted to the membrane of N. benthamiana cells, whereas the fluorescent signal from SGT1-GFP in N. benthamiana leaves was found throughout whole cells, including in the ). In addition, we also detected the protein levels of SRC2-1 in the cytoplasm and the nuclei in N. benthamiana leaves with or without co-expression of SGT1, the result showed that SRC2-1 protein was detected only in the cytoplasm when it expressed alone ( Supplementary Fig. S7), but was detected in both of the cytoplasm and the nuclei when SGT1 was co-expressed ( Supplementary Fig. S7).
To test the effect of subcellular localization of the SGT1/SRC2-1 complex on pepper immunity, a set of BiFC vectors that include an NLS tag (NLS-SRC2-1-YFP C and nls-SRC2-1-YFP C ) and NES tag (NES-SGT1-YFP N and nes-SGT1-YFP N ) were constructed. BiFC analysis using different combinations of vectors was performed, and the corresponding induction of immunity associated with these combinations was also assayed. The results show that YFP signals were detected both in the nuclei and cytoplasm of cells of N. benthamiana leaf transiently overexpressing nes-SGT1-YFP N /nls-SRC2-1-YFP C . However, the YFP signals were detected in the nuclei of N. benthamiana leaf cells when NLS-SGT1-YFP N /NLS-SRC2-1-YFP C was simultaneously co-expressed, while the YFP signals were detected in the cytoplasm of N. benthamiana leaf cells when NES-SGT1-YFP N / nls-SRC2-1-YFP C was simultaneously co-expressed. The YFP signal was not detected after co-expression of NES-SGT1-YFP N /NLS-SRC2-1-YFP C (Fig. 10a). Among the vector combinations that produced YFP signals, induced immunity could only be detected (by trypan blue staining and the electrolyte leakage measurement) in leaf cells co-expressing nes-SGT1-YFP N /nls-SRC2-1-YFP C (Fig. 10a/b).

Discussion
Despite the important roles of SGT1 and SRC2-1 in plant immunity suggested by previous studies 14, 40 , a functional relationship between the two proteins and their potential roles in plant immunity have not yet been established. This study provides evidence of a novel interaction between SGT1 and SRC2-1, and a role for this interaction in pepper immunity induced by PcINF1/SRC2-1.
Using real-time PCR analysis, the response of the transcriptional expression of SGT1 to P. capsici inoculation as well as transient overexpression of PcINF1, an elicitin from P. capsici, was investigated. The qPCR result showed that the mRNA level of SGT1 can be induced in response to the P. capsici inoculation and PcINF1/SRC2-1 co-expression, implying its potential involvement in pepper immunity against P. capsici. This potentiality was further confirmed by loss-of-function assay that employed VIGS. The silencing of SGT1 significantly blocked the cell death and upregulation of immunity associated marker genes in pepper plants challenged with P. capsici or in PcINF1/SRC2-1 transiently overexpressed pepper plants, suggesting the involvement of SGT1 in pepper immunity against P. capsici and in the defense response of pepper triggered by transient overexpression of PcINF1/SRC2-1. As PcINF1 is a member of the conserved elicitin family in P. capsici, the defense response triggered by PcINF1, which is mediated by SGT1, may contribute to the pepper immunity against PTI. SGT1 was previously found to act as a positive regulator in both ETI 18,19,46,47 and PTI pathways 9 . For example, the overexpression of OsSGT1 in rice significantly increased basal resistance to the virulent bacterial blight Xanthomonas oryzae pv. oryzae PXO99 9 . In addition to its well-established positive roles, SGT1 has also been found to act as a negative regulator of immune receptor accumulation and is thought to assist in maintaining appropriate levels of immune receptor proteins in order to avoid autoimmunity 48,49 . A recent study found that SGT1, as well as PAR and HSP90, negatively regulates bacterial wilt disease caused by Ralstonia solanacearum in N. benthamiana 10 . Similarly, SGT1 was found to be induced by the potato virus X TGBp3, and enhanced virus accumulation in N. benthamiana 50 . These discrepancies might be due to genetic differences both in pathogens and host plants 51,52 . For example, Arabidopsis SGT1b can antagonize RAR1 to negatively regulate R protein accumulation before infection, and can also function in a RAR1-independent manner to regulate PCD during SGT1 is required in PcINF1/SRC2-1 induced pepper immunity by interacting to SRC2-1. SRC2-1, a C2 domain containing protein that was found to act as an interacting partner of PcINF1 in its pepper immunity induction 14 , was found to synergistically respond to P. capsici inoculation as well as to transient overexpression of PcINF1 with SRC2-1. Also in the present study, the pathogen-responsive W-box and HSE were consistently found to present in the promoters of both SGT1 and SRC2-1 (Supplementary Figs S8/S9 and supplementary Table S3).   Given that SGT1 acts as a versatile chaperone, and SRC2-1 contains a C2 domain conferring protein interaction, we speculated that SGT1 and SRC2-1 could potentially be functionally associated in a protein interaction manner. To test this speculation, BiFC experiments were performed and the result potentially demonstrated SGT1 interacts with SRC2-1 which was further confirmed through co-IP experiment, but the potential involvement of other components in the SGT1/SRC2-1 complex cannot be excluded. Additionally, the interaction between PcINF1 and SRC2-1 was found to be decreased by the silencing of SGT1; on the other hand, the transient overexpression of PcINF1 and P. capsici inoculation significantly enhanced the interaction between SGT1 and SRC2-1. All of these results indicated that SGT1 and SRC2-1 are functionally related in the response of pepper to PcINF1 transient overexpression or P. capsici inoculation.

Nucleocytoplasmic distribution of SGT1/SRC2-1 complex is required for its immunity activation.
The function of any given protein may be closely associated with its subcellular localization. Similar to the observations made in a previous study 15 , transiently overexpressed GFP-fused SRC2-1 was exclusively localized to the plasma membrane in N. benthamiana leaf epidermis cells, whereas SGT1 was localized throughout the cells including in the plasma membrane, cytoplasm, and nucleus in the present study. Similar observations of SGT1 were also previously reported in other plant species 53 . Given that no NLS is found in SGT1, its nuclear targeting is possibly due to the interaction with other nuclear-targeting proteins. Indeed, SGT1 acts as a general adaptor protein in different plant species, interacting with various classes of proteins including R proteins [Prf 54 and NLR 55 ], other chaperones [OsCYP2 56 , Hsp90 32,57 , RAR1 9 , receptor-like cytoplasmic kinase 38,40 , and type III effector AvrBsT 6 ]. Nucleocytoplasmic distributions have been recently found for TIR-NB-LRRs and CC-NB-LRRs receptors, such as N protein and Rx, which are modulated by SGT1 and are required for full disease resistance 22,23,27 . Our data showed that this nucleocytoplasmic distribution was also displayed for the SGT1/SRC2-1 interaction complex, as illustrated through its translocation from the membrane and cytoplasm into the nuclei in BiFC assays. By using the NLS and NES-tagged proteins to artificially modify the subcellular localization of SRC2-1 and SGT1, we found the exclusive nuclear or cytoplasmic targeting of SGT1/SRC2-1 complex significantly impaired pepper immunity. This result suggests that the translocation of the SGT1/SRC2-1 complex from the cytoplasm to the nucleus is required for the pepper's immunity against P. capsici, and this nucleocytoplasmic portioning is potentially mediated by other proteins, such as HSC70 and HSP90, which play roles in nucleocytoplasmic shuttling in several eukaryotic systems [58][59][60][61][62] , since SGT1 can interact with these proteins by previous studies 44,59,63,64 . Similarly, the PcINF1/SRC2-1 complex was also found to have a nucleocytoplasmic distribution by our previous study 14 . It is potential that this nucleocytoplasmic portioning may also be mediated by HSC70 or HSP90. Further experiments are needed to elucidate the potential involvement of HSC70 or HSP90 in SGT1/ SRC2-1 nucleocytoplasmic balancing.
Collectively, the data in this study suggest that SGT1 is required in PcINF1/SRC2-1 induced pepper defense response by interacting with SRC2-1, and that the nucleocytoplasmic partitioning of the resulting protein complex is required for plant immunity. Methods Plant materials and pathogen inoculation. Pepper cultivar (Capsicum annuum L. cv yanshan01) was collected from Yanshan County, Yunnan Province, China. The seeds of N. benthamiana were stored in our lab. The Phytophthora capsici strain used in this study was isolated from Phytophthora blight pepper plants in Huian County, Fujian Province, China. The pepper and N. benthamiana plants were grown as previously described 66 . For inoculation, P. capsici was cultured in V8 medium (200 g l −1 tomato juice, 3 g l −1 CaCO 3 ) overnight at 28 °C. P. capsici was collected with low speed centrifugation, suspended in sterile ddH 2 O, and then adjusted to a density of 100 zoospores ml −1 . The zoospore suspension of P. capsici was then infiltrated into the lower epidermal of the pepper leaves using a syringe without a needle. The P. capsici-infected pepper plants were then transferred into a climate chamber at 28 °C with 100 mmol photons m −2 s −1 , a relative humidity of 70%, and a 16-h light/8-h dark cycle.
Quantitative real-time PCR assay. The pepper leaves collected from treatments were frozen in liquid nitrogen immediately and stored in the − 80 °C freezer. The frozen leaf tissues were disrupted in 1.5 ml RNAse-free microcentrifuge tubes using 3 stainless steel beads and the Tissue Lyser II (Qiagen, Dusseldorf, Germany). Trizol (Invitrogen, Carlsbad, USA) was used to extract the total RNA from the disrupted tissues. Both the concentration and the quality of the total RNA were detected using the NanoDrop 2000 (ThermoScientific, Massachusetts, USA). 500 ng of RNA from each sample was used for the first cDNA strand generation using One Step PrimeScript TM cDNA Synthesis Kit (TaKaRa, Shigo, Japan), which include a genomic DNA digestion step according to the manufacturer's instructions (TaKaRa). The synthesized cDNA products were diluted ten-fold in sterilized RNAse-free ddH 2 O for further qPCR analysis. qPCR was performed as a 20-μ L reaction (10 μ L 2× SYBR Premix Ex-Taq, 0.2 μ M forward and reverse primers, and 1 μ L cDNA) in a 200-μ L PCR tube using SYBR Premix Ex Taq TM (TaKaRa) and a CFX TM Real-Time System with the following program: 95 °C, 30 s; 95 °C, 5 s; 60 °C, 34 s; 95 °C, 15 s, 40 cycles. SYBR green fluorescence was used to detect the amplification of the target genes at each cycle. The Cq (threshold cycle), defined as the real-time PCR cycle at which a statistically significant increase of reporter fluorescence was first detected, was used as a measure for the starting copy numbers of the target gene. Three replicates of each experiment were performed. The data were analyzed by the Livak method 70 and expressed as a normalized relative expression level (2 −ΔΔCT ) of the respective genes. CaActin (GQ339766) and 18S ribosomal RNA (EF564281), which have been confirmed by the published paper 66,67 , were used to normalize the transcript levels in pepper. The specificities of the qPCR primers were confirmed by the melting curve and the qPCR validation was investigated using the Bio-rad CFX manager 2.0 (Supplementary Table S1). All gene-specific primers are listed in Supplementary Table S1.

Bimolecular fluorescence complementation (BiFC) analysis. To construct vectors for BiFC anal-
ysis, the open reading frames (ORFs) of SGT1 and SRC2-1 (flanked by the attB sequence) were amplified with gene-specific primer pairs listed in Supplementary Table S2. Products were introduced into pDONR207 to yield pDONR vectors (named SGT1-207 and SRC2- , and then transferred into the satellite vectors pE3130 and pE3136 to generate SGT1-YFP N and SRC2-1-YFP C . The reciprocal constructs, SGT1-YFP C and SRC2-1-YFP N , were constructed similarly. Fragments of SGT1-YFP N and SRC2-1-YFP C or SGT1-YFP C and SRC2-1-YFP N in satellite vectors were together recombined into the destination vector PE3519, which was then transformed into the Agrobacterium strain GV3101. Agrobacterial cells harboring the BiFC constructs were Agro-infiltrated into the leaves of N. benthamiana plants. Two days after Agro-infiltration, fluorescent signals from N. benthamiana leaves were analyzed using a confocal microscope DM6000 CS (Leica, Solms, Germany). For DAPI staining, N. benthamiana leaves were immersed in DAPI solution (0.1% DAPI and 5% DMSO) and incubated at 37 °C for 1 h before observation.

Generation of Gateway-compatible split-LUC vectors and LUC activity measurement. LUC
destination vectors used in this work are purchased using the forms provided at http://www.uni-muenster.de/ Biologie.IBBP/agkudla/Plasmids.html. To construct split-LUC vectors, the attB-flanked open reading frames of SGT1 and SRC2-1 without stop codon were amplified with gene-specific primers and the PCR products were introduced into the satellite vector pDONR207. After sequencing confirmation, the SGT1 (PcINF1) and SRC2-1 entry vectors were transformed into the destination vectors (pDEST-GW NLUC and pDEST-GW CLUC) to generate 35S::SGT1-NLUC (35S::PcINF1-NLUC) and 35S::SRC2-1-CLUC, respectively. The yielded SGT1-and SRC2-LUC vectors were transformed into the Agrobacterium stain GV3101. GV3101 cells harboring the various split-LUC constructs were used for infiltration in pepper leaves. GV3101 cells carrying the indicated constructs were centrifuged and resuspended in the infiltration buffer (10 mM MgCl 2 , 10 mM MES, 150 μ M acetosyringone) at an OD595 of 0.4 for LUC activity measurement. Subsequently, the indicated split-LUC constructs and pEXP-GUS were mixed at 1:1 ratio and incubated with gentle shaking at room temperature for 4 hours before infiltration into young but fully expanded leaves of 6-7 week-old pepper plants. After infiltration, the pepper plants were maintained in the greenhouse until LUC activity measurement. Leaf pairs of each pepper plants were infiltrated with the sample in one leaf and the corresponding negative control in the other leaf. Luminescence was quantified in pepper leaves using a floated leaf disc assay 15 at appropriate times after infiltration. LUC assay solution (0.1 mM luciferin dissolved in 0.5% v/v DMSO, 10 mM MgCl 2 and 10 mM MES/KOH, pH 5.6) was prepared and infiltrated into the agro-infiltrated-pepper leaf before being cut out as leaf discs of 12 mm diameter. The leaf discs were placed in the 96-well microplate with the same LUC assay solution. The light production was measured by a microplate reader (Biotek; http://www.biotek.com). Before LUC activity measurement, the pepper plant were kept in the dark for 10 min to minimize the background fluorescence, and infiltration of LUC assay solution was performed in the low-light condition. GUS activity measurement. To normalize the LUC values for different expression levels in pepper leaves, GUS activity measurement was performed. Before Agro-infiltration, all the GV3101 cells carrying the indicated constructs were co-infiltrated with a GUS expression vector (pEXP-GUS). To quantitatively measure GUS activity, protein extraction buffer [10% glycerol, 25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 2% Triton X-100, 10 mM DTT, 1× complete protease inhibitor cocktail (Sigma-Aldrich; http://www.sigmaaldrich.com), and 2% (w/v) polyvinylpolypyrrolidone] 16 was used to extract the total protein of infiltrated pepper leave. To measure GUS activity, the rate of p-nitrophenol (γ = 415 nm) release was quantified using a microplate reader (Biotek; http://www.biotek.com).
Yeast two hybrid assay. The GAL4-based Y2H library screening system was performed according to the manufacturer's instructions (Invitrogen). The PCR-amplified, attb-containing ORFs of SGT1 and SRC2-1 were introduced into the gateway-compatible vectors pDONR201 and pDONR207, respectively, to generate entry vectors. The entry vectors were then confirmed by sequencing and transformed into the bait vector pDEST32 and the prey vector pDEST22 to generate BD-SGT1 and AD-SRC2-1, respectively. We transformed these constructs into yeast strain MaV203 with positive-control and negative-control vectors and were allowed to grow in the selection media (SD-Leu-Trp-His/+ appropriate concentrations of 3-amino-1, 2, 4-triaole).

Agrobacterium-mediated transient overexpression and subcellular localization assay.
To construct vectors for transient overexpression of SGT1 and SRC2-1, SGT1 and SRC2-1 from plasmids constructed for BiFC experiments (previously described) were transferred into gateway-compatible destination vectors pK7WG2 and pMDC45 (Invitrogen) using LR reaction to yielded pK7WG2-SGT1, pK7WG2-SRC2-1, pMDC45-SGT1, and pMDC45-SRC2-1. All of these vectors were transformed into the Agrobacterium strain GV3101, and agrobacterial cells containing the appropriate vectors were streaked on solid YEP medium (10 g Bacto peptone, 10 g yeast extract, 5 g NaCl, and 15 g agar in 1 l of water) and grown at 28 °C. Individual clones were selected two days later and further cultivated in YEP liquid medium containing appropriate antibiotics. Cells in liquid medium were collected and then suspended in MMA infiltration buffer (10 mM MgCl 2 , 10 mM MES, 200 μ m acetosyringone). The MMA-suspended cells were infiltrated into the full leaves of pepper or N. benthamiana plants using a syringe without a needle.
For subcellular localization assays, the plasmids described above were used. Green fluorescence in N. benthamiana leaves infiltrated with GV3101 cells containing pMDC45-SGT1 and pMDC45-SRC2-1 was observed using a fluorescence microscope Eclipse Ni-U (Nikon, Tokyo, Japan) at 48 HAI. For visualization of nuclei, leaf samples were stained in DAPI solution and incubated at 37 °C for 1 h before observation.
Nuclear targeting of SRC2-1 and nuclear export of SGT1 using a nuclear localization sequence (NLS) peptide and a nuclear export sequences (NES). To artificially target SRC2-1 to the nuclei, DNA for a general NLS (5′-GGCCCTAAAAAGAAGCGTAAGGTT-3′) 45 was fused to DNA corresponding to the N-terminal of the ORF of SRC2-1. A negative control sequence (nls) was produced by substituting an Asn residue for a Lys (5′-GGCCCTAAAAACAAGCGTAAGGTT-3′) 45 . To drive nuclear export of SGT1, an NES (CTTGCATTAAAGCTCGCTGGTCTTGATATT) fragment was fused to the N-terminal of the ORF of SGT1. The NLS-and NES-fused ORF of SGT1 and SRC2-1 was first cloned into the pDONR207, sequenced and transferred into the corresponding destination vectors, and eventually introduced into the agrobacterial cells for further study.
The nuclear and cytoplasmic proteins were extracted using the Eukaryotic Nuclear Protein Extraction Kit or Eukaryotic Membrane Protein Extraction Kit, respectively (Thermo scientific).

Virus-induced gene silencing (VIGS) in pepper plants.
VIGS technology was used to study the loss-function of SGT1 and SRC2-1. The VIGS assays were performed as described previously 14 . To re-describe the method briefly, specific 3′ UTR fragments of SGT1 and SRC2-1 around 200 bps in length were amplified and cloned into the vector pTRV2 using the primers listed in Supplementary Table S2. These vectors were then transformed into GV3101 cells, and agrobacterial cells harboring the TRV-derived vectors (TRV1/TRV2-SGT1, TRV1/TRV2-SRC2-1, TRV1/TRV2-SGT1/SRC2-1, and TRV1/empty) were streaked onto solid YEB medium and grown at 28 °C. Individual clones were picked two days later and cultured in liquid YEP medium containing 75 μ g ml −1 kanamycin and 75 μ g ml −1 rifampicin. GV3101 cells were collected by centrifugation and suspended in MMA infiltration buffer (10 mM MgCl 2 , 10 mM MES, 200 μ M acetosyringone, pH 5.7), and adjusted to an OD 595 of 0.6. Suspensions of GV3101 cells containing TRV1/TRV2-SGT1, TRV1/TRV2-SRC2-1, TRV1/ TRV2-SGT1/SRC2-1, or TRV1/empty were mixed at the ratio of 1:1 and incubated at room temperature in a shaker at low speed (50 rpm). After gentle shaking for 3 h, the GV3101 cells containing vectors in each combination were co-infiltrated into the cotyledons of 2-week-old pepper seedlings.
Scientific RepoRts | 6:21651 | DOI: 10.1038/srep21651 Conductivity leakage measurements. The conductivity of leaves was analyzed by measuring ion leakage following the method of Choi et al. 70 with slight modification. Six leaf discs (4 cm in diameter) were cut with a hole punch, washed with sterilized ddH 2 O for three times, and subsequently incubated in 20-ml volume of double-distilled water and maintained in a shaker with gently shaking (at 60 rpm) for 60 min at room temperature. The ion leakage of the bathing solution was measured with a conductivity meter (Mettler Toledo 326 Mettler, Zurich, Switzerland). DAB and trypan blue staining. DAB (3, 3′ -diaminobenzidine) and trypan blue staining were performed after Agro-infiltration using the method 70 to visualize H 2 O 2 accumulation and cell death. For DAB staining, the full Agro-infiltrated leaves were immersed in DAB solution (1 mg ml −1 DAB, pH 5.7), protected from light and incubated overnight (for almost 8 h). Leaves were then de-stained with 95% ethanol three times. For trypan blue staining, the full Agro-infiltrated leaves were boiled in a lactophenol-ethanol trypan blue solution (10 ml lactic acid, 10 ml glycerol, 10 g phenol, 30 ml absolute ethanol, and 10 mg of trypan blue dissolved in 10 ml of distilled water) for 5 min and allowed to rest at room temperature overnight. Leaves were de-stained with 2.5 g ml −1 chloral hydrate in distilled water.