Disease resistance (R) proteins, as central regulators of plant immunity, are tightly regulated for effective defense responses and to prevent constitutive defense activation under non-pathogenic conditions. Here we report the identification of an F-box protein CPR1/CPR30 as a negative regulator of an R protein SNC1 likely through SCF (Skp1-cullin-F-box) mediated protein degradation. The cpr1-2 (cpr30-1) loss-of-function mutant has constitutive defense responses, and it interacts synergistically with a gain-of function mutant snc1-1 and a bon1-1 mutant where SNC1 is upregulated. The loss of SNC1 function suppresses the mutant phenotypes of cpr1-2 and cpr1-2 bon1-1, while overexpression of CPR1 rescues mutant phenotypes of both bon1-1 and snc1-1. Furthermore, the amount of SNC1 protein is upregulated in the cpr1-2 mutant and down-regulated when CPR1 is overexpressed. The regulation of SNC1 by CPR1 is dependent on the 26S proteosome as a protease inhibitor MG132 stabilizes SNC1 and reverses the effect of CPR1 on SNC1. Interestingly, CPR1 is induced after infection of both virulent and avirulent pathogens similarly to the other negative defense regulator BON1. Thus, this study reveals a new mechanism in R protein regulation likely through protein degradation and suggests negative regulation as a critical component in fine control of plant immunity.
Disease resistance proteins are central regulators of plant immunity. In addition to employing pattern recognition receptors (PRR) to detect common features of pathogens and induce basal defense, plants utilize disease resistance (R) proteins to perceive directly or indirectly race-specific effectors from pathogens (Nurnberger et al., 2004; Chisholm et al., 2006; Jones and Dangl, 2006). This specific recognition leads to rapid and efficient defense responses including a form of programmed cell death named hypersensitive response to restrict the growth of pathogens (Hammond-Kosack and Jones, 1996). R gene activation also induces systemic acquired resistance mediated by salicylic acid (SA) in distal regions to prepare plants for later pathogen attacks (Durrant and Dong, 2004).
Most of the R proteins in Arabidopsis are nucleotide binding-leucine rich repeat (NB-LRR) proteins with structural similarity to animal immune receptor NOD like receptor (NLR) proteins (Ausubel, 2005). R genes are tightly regulated because overexpression or activation of R proteins lead to compromised plant growth and/or cell death (Shirano et al., 2002; Stokes et al., 2002; Zhang et al., 2003). A number of plant host genes have been identified as negative regulators of defense responses or cell death (Lorrain et al., 2003; Moeder and Yoshioka, 2008) and some of these genes might impact defense responses through R genes. The loss-of-function mutation of the evolutionarily conserved copine gene BON1 (BONZAI1) in Arabidopsis has an autoimmune phenotype resulting from activation of the NB-LRR R gene SNC1 (Suppressor of npr1-1 constitutive 1) (Zhang et al., 2003; Yang and Hua, 2004). No direct protein-protein interaction between BON1 and SNC1 proteins has been observed, and SNC1 has a higher transcript level in bon1-1 than in the wild-type Col-0 (referred as Col). SA plays an important role in this upregulation through a feedback regulation (Yang and Hua, 2004), and BON1 can weakly repress SNC1 transcription in the absence of a feedback regulation (Li et al., 2007). Thus BON1 likely represses the expression of SNC1 transcript and the derepression of SNC1 in bon1-1 is amplified by a feedback regulation mediated by SA. SNC1 is also negatively regulated by MKP1 (MAPK phosphatase 1) and SRFR1 (SUPPRESSOR OF rps4-RLD1) (Bartels et al., 2009; Kim et al., 2010; Li et al., 2010a). MKP1 regulates the activity of MPK6, a component in the MAPK kinase cascade that is involved in pathogen-associated molecular pattern (PAMP) signaling (Pitzschke et al., 2009). SRFR1 encodes a conserved tetratricopeptide repeat protein with similarity to transcriptional repressors (Kwon et al., 2009). Both the SNC1 transcript and the SNC1 protein accumulate to a higher level in the srfr1 mutant than in the wild type (Kim et al., 2010). The interaction between SRFR1 and SGT1b suggests that it may directly regulate the accumulation of SNC1 protein (Li et al., 2010a).
A missense mutation snc1-1 (named snc1 in Zhang et al., 2003) induces constitutive defense responses that are EDS1 (ENHANCED DISEASE SUSCEPTIBILITY1) and PAD4 (PHYTOALEXIN DEFICIENT4) dependent. Defense responses triggered by the active form of snc1-1 has been used to dissect R-mediated disease resistance through modifier of snc1 (mos) mutants that suppresses the snc1-1 phenotype. A large number of MOS genes have been isolated, and their molecular identities have implicated the involvement of transcriptional and post-transcriptional regulation, protein stability, and nucleus-cytosol shuttling control in plant immunity (Monaghan et al., 2010).
Key regulators of developmental or environmental responses are often subject to regulation at the level of protein stability. This regulation ensures that they can be rapidly turned on and off depending on whether or not their functions are needed. One of the prevalent ways to turn proteins off is through the ubiquitin mediated 26S proteosome system (Hershko and Ciechanover, 1998). Ubiquitin is activated by a covalent linkage to a ubiquitin-activating enzyme E1, transferred to a ubiquitin-conjugation enzyme E2, and added to the substrate with an isopeptide bond by a ubiquitin–protein ligase E3. Polyubiquitinated proteins are subsequently recognized and degraded by the 26S proteosome. The E3 ligases conferring specificity of target selection belong to four subfamilies with the SCF as the most abundant E3 ligase subfamily. SCF contains four subunits and the F-box subunit directly interacts with the substrate. There are more than 700 F-box proteins in Arabidopsis, indicating that a large number of proteins and processes are regulated by this system (Bachmair et al., 2001; Lechner et al., 2006).
Like other developmental and environmental response processes, plant immunity is subject to regulation by the ubiquitin system (Zeng et al., 2006). A few genes in the ubiquitin system are implicated in plant pathogen interaction. Among these, SON1, SPL11, and CPR30 are target recognition subunits of the E3 ligase proteins, and they appear to be negative regulators of defense responses (Kim and Delaney, 2002; Zeng et al., 2004; Gou et al., 2009). CPR30 encodes a functional F-box protein which interacts with multiple ASK proteins. The loss-of-function cpr30 mutant showed a dwarf phenotype and enhanced resistance to both virulent and avirulent bacterial pathogens. These phenotypes are dependent on EDS1, PAD4, and NDR1 (NONRACE-SPECIFIC DISEASE RESISTANCE1). CPR30 may directly target a positive regulator of plant defense for degradation through the 26S proteasome (Gou et al., 2009). The cpr30 mutant is recently found to be allelic to cpr1 as this manuscript is being revised (Cheng et al., 2011), and therefore we will refer CPR30 as CPR1 and the cpr30-1 mutant (Gou et al., 2009) as cpr1-2. R proteins have been suggested to be regulated at the level of protein stability. A regulator of SCF complex, SGT1b, has been proposed to be required for R function and it appears to have dual functions in R gene mediated resistance (Holt et al., 2005). In contrast, it antagonizes RAR1 to negatively regulate the accumulation of R proteins, and on the other hand, it acts as co-chaperons with HSP90 and RAR1 to help R proteins to fold, assemble, and mature (Holt et al., 2005; Kadota et al., 2010). The requirement of SGT1b differs for different R proteins, which suggested a difference in relative effects of the two functions of SGT1b. SGT1b does not appear to be required for SNC1 function (Goritschnig et al., 2007), but the SNC1 protein accumulates to a high level in the sgt1b null mutant (Li et al., 2010b), suggesting that SGT1b is involved in the degradation of SNC1.
Here we described the regulation of R genes by the F-box protein CPR1/CPR30. CPR1 is found to negatively regulate SNC1, likely at the level of SNC1 protein accumulation. This finding suggests that R proteins are under the control of protein degradation to ensure fine control of plant immunity.
The cpr1-2 mutant phenotype is enhanced by bon1-1 and suppressed by snc1-11
The CPR1 gene encodes an F-box protein, and the loss of CPR1 function leads to constitutive defense responses and dwarfism dependent on EDS1, PAD4, NDR1, and SA (Gou et al., 2009). The cpr1-2 mutant is similar to the bon1-1 mutant in its temperature-sensitive growth and defense phenotypes (Figure 1a,b). The resemblance of the cpr1-2 phenotype to that of bon1-1 promoted us to analyze the genetic interactions between these mutants. The double mutant of cpr1-2 bon1-1 had an extremely dwarf phenotype at 22°C and eventually died at two-leaf stage (Figure 1a), indicating a synergistic interaction between cpr1-2 and bon1-1. Most strikingly, the lethality of the cpr1-2 bon1-1 double mutant was rescued at 28°C where it had an almost wild-type looking growth phenotype (Figure 1a). As R genes have been shown to be temperature sensitive (Zhu et al., 2010), this rescue suggests that R genes may mediate the phenotype of cpr1-2 bon1-1.
The R gene SNC1 is upregulated in the bon1-1 mutant (Yang and Hua, 2004), and the fact that the lethality of cpr1-2 bon1-1 can be suppressed by a high growth temperature suggests that SNC1 might be the R gene activated in cpr1-2 as well. We therefore introduced into cpr1-2 a loss-of-function SNC1 allele snc1-11 (Yang and Hua, 2004). In contrast to the cpr1-2 mutant, the cpr1-2 snc1-11 had a wild-type looking phenotype at 22°C (Figure 1c). The transcripts of both PR1 and PR2 were highly induced in cpr1-2, but they were greatly reduced in the cpr1-2 snc1-11 double mutant (Figure 1d). In addition, the snc1-11 mutation largely rescued the cpr1-2 bon1-1 phenotype (Figure 1c), indicating the R gene SNC1 is the common target of BON1 and CPR1.
The cpr1-2 mutation has a synergistic interaction with the snc1-1 mutation
We subsequently analyzed the interaction between cpr1-2 and the constitutive active snc1-1 mutant (Zhang et al., 2003). Both single mutants were recessive in growth phenotype, but the double heterozygous F1 plants showed a dwarf phenotype similar to that of the single homozygous mutants (Figure 1e). In 200 individual plants genotyped in the F2 population, nine were cpr1-2 snc1-1/+ and 5 were cpr1-2/+ snc1-1 and all plants with these two genotypes were severely dwarf and could not set seeds at 22°C (Figure 1e). The double homozygous plants were not found from this F2 population, most likely due to the fact that CPR1 and SNC1 are physically linked, at a distance of 2000 kb and 7 cM (calculated from this F2 segregation). As neither cpr1-2 snc1-1/+ nor cpr1-2/+ snc1-1 could set seeds, we were unable to determine if the double homozygous mutants would have an even more severe phenotype. In any case, the dominant interaction between cpr1-2 and snc1-1 suggests a very close interaction between CPR1 and SNC1.
Requirements of the MOS genes for the mutant phenotypes in cpr1-2
To further examine the genetic relationship between CPR1 and SNC1, we crossed cpr1-2 to a few mos mutants that suppressed the snc1-1 phenotype. MOS7 encodes a nucleoporin Nup88 homologue residing at the nuclear envelope (Cheng et al., 2009). Similar to snc1-1, cpr1-2 could be partially but significantly rescued by a mutation in the MOS7 gene (Figure 2a). However, cpr1-2 was not affected by a mutation in MOS3 that encodes another nucleoporin (Zhang and Li, 2005; Figure 2b). MOS5 encodes UBA1, one of the two ubiquitin-activating enzyme (E1) in Arabidopsis (Goritschnig et al., 2007). In contrast to the partial suppression of snc1-1 by mos5-1, the mos5-1 mutation even slightly enhanced the cpr1-2 phenotype (Figure 2c). Neither did the mutation in the MOS2 gene that likely encodes an RNA binding protein suppressed the cpr1-2 phenotype (Zhang et al., 2005; Figure 2d). Taken together, the genetic requirements of MOS genes for cpr1-2 and snc1-1 are mostly different, although the SNC1 activity is upregulated in both mutants.
Overexpression of CPR1 suppresses the phenotypes of bon1-1 and snc1-1
We hypothesize that if SNC1 is negatively regulated by CPR1, overexpression of CPR1 might suppress bon1 or snc1-1 phenotype where SNC1 activity is upregulated. A CPR1 overexpression construct (p35S::CPR1, named as p35S::CPR30 in Gou et al., 2009) was transformed into cpr1-2, bon1-6 and snc1-1 plants. The bon1-6 allele was used instead of bon1-1 because the CPR1 construct would confer Basta resistance that already exists in the bon1-1 mutant. bon1-6 was obtained from a screen for insensitive to temperature (int) mutants in the snc1-1 background (Zhu et al., 2010). This allele is a loss-of-function allele as it contains a mutation at the splicing site that leads to protein truncation. More than 10 transgenic lines were each generated in cpr1-2, bon1-6, and snc1-1. RNA blot analysis revealed a much higher level of the CPR1 transcripts in transgenic plants (Figure 3a). Eleven out of 14 transgenic lines in cpr1-2 showed wild-type phenotype indicating an intact function of the transgene (Figure 3b). Fifteen out of 19 and 12 out of 14 transgenic plants in bon1-6 and snc1-1 respectively also showed rescued phenotypes to different degrees (Figure 3b). These data further support that SNC1 is a target gene of CPR1.
The SNC1 protein amount is inversely correlated with the CPR1 amount
Because the CPR1 gene encodes an F-box protein which is a subunit of the SCF E3 ligase involved in ubiquitin mediated protein degradation, we asked if SNC1 could be a direct target of CPR1 for protein degradation. Because the endogenous SNC1 protein could not be detected with antibodies raised against SNC1, we resorted to the pSNC1::SNC1:GFP transgenic plants (Li et al., 2007; Zhu et al., 2010) for protein detection. The pSNC1::SNC1:GFP plants show dwarfism because of higher expression of the transgene than the endogenous SNC1 gene (Li et al., 2007). Transgenic plants of pSNC1::SNC1:GFP in cpr1-2 were identified from the F2 progenies of a cross between cpr1-2 and a pSNC1::SNC1:GFP line, and they were severely dwarfed (Figure 4a). The SNC1:GFP protein detected by western blot accumulated to a much higher level in cpr1-2 than in the wild type (Figure 4b), suggesting that the SNC1 protein level is affected by the cpr1-2 mutation.
To further determine the regulation of the SNC1 protein by CPR1, we overexpressed CPR1 in pSNC1::SNC1:GFP transgenic plants. Fifteen out of the 18 T1 plants showed rescued phenotypes (Figure 4c), similarly to the suppression of snc1-1 by CPR1 overexpression. A decrease of SNC1:GFP protein level was observed in these lines compared to those without CPR1 overexpression (Figure 4d), further supporting a regulation of the SNC1 protein level by CPR1.
Although the SNC1 transcript was not significantly upregulated in cpr1-2 in plate growing seedlings (Gou et al., 2009), we found that its RNA transcript was expressed at a higher level in cpr1-2 than in the wild type in seedlings grown in soil (Figure 4e). This upregulation was reduced by eds1-1 and pad4-1 mutations (Figure 4e), suggesting a feedback regulation on the SNC1 transcript similar to that in the bon1-1 mutant (Yang and Hua, 2004). Therefore, the primary regulatory effect of CPR1 on SNC1 might be at the transcript or the protein level.
Expression of CPR1 in Nicotiana benthamiana inhibits cell death triggered by SNC1
To bypass the feedback regulation on SNC1 transcript, we utilized transient expression in Nicotiana benthamiana to examine the regulation of SNC1 by CPR1. Under optimized conditions, expression of SNC1:GFP and SNC1-1:GFP by the strong 35S promoter could trigger cell death characterized by collapsed tissues distinct from wounding caused cell damage (Figure 5a). We took advantage of this system to analyze the regulation of SNC1 at the protein level because the 35S promoter is unlikely subject to feedback control. N. benthamiana plants were infiltrated with p35S::SNC1:GFP or p35S::SNC1-1:GFP together with a GFP fusion of a control E3 ligase gene (At5g45100) on one half of the leaf and with p35S::CPR1 on the other half of the same leaf. While SNC1:GFP or SNC1-1:GFP each triggered strong cell death when co-expressed with the control gene, they triggered significantly less cell death when co-expressed with CPR1 (Figure 5b). Similar results were obtained when another F-box protein COI1 or the empty vector carrying the GFP alone were used as control (Figure S1).
The extent of cell death inhibition is dependent on the expression level of CPR1. A varying amount of Agrobacteria containing p35S::CPR1 were co-infiltrated with p35S::SNC1:GFP or p35S::SNC1-1:GFP in one half of the leaf, and the same varying amount of the control construct was infiltrated in the other half of the leaf. Cell death was quantified by ratio of weight of CPR1 infiltrated half leaf versus the control infiltrated half leaf. A higher amount of Agrobacteria containing CPR1 leads to a higher weight ratio indicative of a stronger suppression of cell death, and this was found for cell death triggered by either SNC1WT or SNC1-1 (Figure 5c and Figure S2a).
We found that the amount of the SNC1:GFP and SNC1-1:GFP proteins was reduced when it was co-expressed with CPR1 compared with the control in N. benthamiana (Figure 5d). Furthermore, the reduction of the SNC1:GFP amount is largely correlated with the amount of CPR1 co-infiltrated. As the infiltration amount of Agrobacteria strains containing CPR1 was progressively increased, the SNC1:GFP protein level was progressively reduced (Figure S2b).
MG132 treatment stabilizes SNC1 protein in N. benthamiana
Because CPR30 is an F-box protein that is a component of the SCF E3 complex, it is possible that CPR1 directly targets SNC1 for degradation through 26S proteosome. To test this hypothesis, we sprayed the leaves of N. benthamiana with the proteosome inhibitor MG132 after Agroinfiltration with SNC1:GFP. A higher accumulation of both SNC1:GFP and SNC1-1:GFP proteins was found in the MG132 treated samples (Figure 6), indicating that the SNC1 protein is more stable when the proteosome activity is inhibited. Furthermore, the reduction of both SNC1:GFP and SNC1-1:GFP accumulation induced by CPR1 overexpression was suppressed by MG132 (Figure 6), suggesting that the inhibition of SNC1 accumulation by CPR1 is dependent on the 26S proteosome.
CPR1 potentially regulates activities of other R genes
We further examined in Arabidopsis if the loss of CPR1 function could alter resistance mediated by other R genes. To bypass the secondary effect caused by higher SNC1 activity, we used both cpr1-2 and cpr1-2 snc1-11 in this analysis. Consistent with previous findings (Gou et al., 2009), the cpr1-2 mutant exhibited an enhanced resistance to virulent pathogen Pseudomonas syringae pv tomato (Pst) DC3000 (Figure 7a). While snc1-11 was as susceptible to Pst DC3000 as the wild-type Col, cpr1-2 snc1-11 had a stronger resistance than the wild-type Col or snc1-11 (Figure 7a), suggesting that CPR1 regulates additional defense genes other than SNC1. Both cpr1-2 and cpr1-2 snc1-11 were significantly more resistant to Pst DC3000 AvrRpt2 compared with the wild-type Col and snc1-11 (Figure 7b), suggesting that CPR1 regulates RPS2 mediated resistance as well.
CPR1 is induced by virulent and avirulent pathogens
CPR1 is ubiquitiously expressed throughout the plants from analysis of the pCPR1::GUS transgenic plants (Gou et al., 2009). But very little information of its regulation by various stimuli is available from public database because this gene is not on the commonly used microarrays. As CPR1 is a negative regulator of an R protein, we examined the expression of CPR1 during the course of pathogen invasion. After inoculation of Arabidopsis plants with bacterial pathogen strains Pst DC3000 and Pst DC3000 AvrRpt2, the transcripts of CPR1 were significantly elevated at 24 or 36 hours post inoculation (hpi) as revealed by the RNA blot analysis (Figure 7c). This induction was also observed in transgenic lines with the pCPR1::GUS reporter gene. These lines were infiltrated with Pst DC3000 and Pst DC3000 AvrRpt2 respectively, and a stronger staining was observed close to the infiltration sites at 24 hpi for both bacterial strains compared with the buffer control (Figure 7d). Therefore, CPR1 is itself regulated after pathogen invasion perhaps to fine tune the progression of defense responses.
R gene activation is an essential step in plant immunity, however, activation in the absence of pathogens or over-activation in response to pathogen invasion is costly or detrimental to plant fitness. Here we present evidence indicating that an R protein is regulated at the level of protein accumulation likely through the 26S proteosome. The F-box protein CPR1/CPR30 negatively regulates the activity and accumulation of the R proteins SNC1. Negative regulation of R genes could serve two purposes: one is to keep the basal level of R proteins down when there is no pathogen invasion and the other is to keep the immune response under check so that it does not over react. Protein degradation has emerged as a universal mechanism in controlling the amount of key regulators to keep the regulatory systems under fine control. CPR1 could be such a regulator of R gene activities at the level of protein accumulation. CPR1 was previously shown to be ubiquitously expressed in different organs at different growth stages (Gou et al., 2009), and it is moderately induced when immune responses are fully triggered (Figure 7). A similar upregulation by pathogen invasion has been observed for BON1 and BAP1, two other negative regulators of SNC1 (Jambunathan and McNellis, 2003; Yang et al., 2006). Synergistic interaction between bon1-1 and cpr1-2 suggests that BON1 and CPR1 might regulate SNC1 activity through different mechanism, one at the RNA transcript level and the other at the level of protein accumulation. Therefore, negative regulation of R gene activity might be an essential component during defense responses.
The simplest model for CPR1 function is that the CPR1 protein directly targets the R protein SNC1 for poly-ubiquitination and subsequent protein degradation. CPR1 and SNC1 are both localized to the cytosol and nucleus (Cheng et al., 2009; Gou et al., 2009; Zhu et al., 2010), so they should have chance to interact with each other directly or indirectly under certain physiological conditions. So far we have not detected a direct interaction between these two proteins by co-immunoprecipitation or ubiquitination of SNC1:GFP in the transient expression system of N. benthamiana. Neither did we detect an interaction between CPR1 and SNC1 in the yeast two-hybrid system. Both the full-length and the F-box deleted CPR1 were tested against either the full-length or the TIR, NB, and LRR domains of SNC1 (data not shown). These negative results might be due to low expression of the proteins in the systems we were analyzing. We could not exclude the possibilities that the interaction between SNC1 and CPR1 requires a trigger or a small molecule, similarly to the requirement of JA-Ile in the interaction between the F-box protein COI1 and its substrate JAZ1 (Thines et al., 2007) or that the direct target of CPR1 is a shared signaling component of R proteins. The recent independent study shows that SNC1 is a direct target of CPR1 (Cheng et al., 2011), thus making the later hypotheses unlikely.
The snc1-1 mutation confers a higher activity to the SNC1 protein. However, the SNC1-1 mutant protein is subject to a regulation by CPR1 similarly to the wild-type SNC1, indicating that the gain-of-function activity of SNC1-1 is not through blocking CPR1 regulation. Defense responses induced in cpr1-2 and snc1-1 appear to have both similar and different genetic requirements. Both depend on the function of EDS1, PAD4, and MOS7, but MOS3, MOS5 and MOS2 are only required for the snc1-1 phenotype but not the cpr1-2 phenotype. Further investigating the difference between the two mutants might reveal the mechanism of snc1-1 activation and CPR1 function.
CPR1 might regulate R genes other than SNC1 as both the cpr1-2 and cpr1-2 snc1-11 mutants exhibited an enhanced resistance to the virulent Pst DC3000 and the avirulent Pst DC3000 AvrRpt2 (Figure 7). It is possible that CPR1 has multiple R protein targets, among which SNC1 is the major one. Alternatively, SNC1 might have a larger effect on plant growth and development than other R proteins.
SNC1 is also under a negative regulation of SRFR1 at least in part at the protein level (Kim et al., 2010; Li et al., 2010a). SRFR1 is found to interact directly with SGT1 proteins (Li et al., 2010a), and CPR1 interacts with multiple SKP1 (ASK) proteins in the yeast two-hybrid system (Gou et al., 2009). Since SGT1b can directly interact with SKP1 proteins in yeast, barley and N. benthamiana (Kitagawa et al., 1999; Azevedo et al., 2002; Liu et al., 2002), it is possible that SRFR1 and SGT1b work together with the SCF complex including CPR1 to regulate SNC1. Future studies should further reveal the mechanism and regulation of R protein accumulation in plant immunity.
Plant growth conditions
Arabidopsis plants were grown under 24 h light or 12 h light (for pathogen test) per day at 22 or 28°C with light intensity at 100 μmol m−2 sec−1 and relative humidity at 50–70%. Seeds were planted either on 0.5× MS (Murashige and Skoog, Sigma, http://www.sigmaaldrich.com/) medium containing 0.8% agar and 2% sucrose or directly in soil (Metro-Mix 200; Sungo). Nicotiana benthamiana plants were grown in the green house at 24°C for 3–4 weeks before use.
Double mutant constructions
The cpr1-2 mutant was crossed to bon1-1, snc1-11, snc1-1, mos7-1, mos2-2, mos3-1 and mos5-1 to make the double mutants, and cpr1-2 snc1-11 was crossed to bon1-1 snc1-11 to make the triple mutant.
The full-length genomic fragments of At5g45100 and COI1 were amplified from the genomic DNA and cloned in the Gateway entry vector pCR®8 TOPO® TA vector (Invitrogen, http://www.invitrogen.com/ ). The At5g45100 gene was then cloned into a modified pHPT vector (Tzfira et al., 2005) to generate p35S::At5g45100:GFP. The COI1 gene was cloned to the PMDC32 binary vector (Curtis and Grossniklaus, 2003) to generate the construct of p35S::COI1.
Agroinfiltration in N. benthamiana and MG132 treatment
One-month-old N. benthamiana plants were used for infiltration. All of the constructs were transformed into Agrobacterium tumefaciens strain GV3101. Agrobacteria strains were grown overnight in liquid media and bacteria were collected and resuspended in induction medium (10 mm MES, pH5.7, 10 mm MgCl2, 200 mm acetosyringone). The final concentration of Agrobacteria strains of SNC1 and SNC1-1 was 0.2 at OD600, and that for the CPR1 or the control was 0.2, 0.4, 0.6, and 0.8 at OD600. Bacteria were infiltrated into N. benthamiana leaves using 1ml needle-less syringe. Leaves were cleared in 75% ethanol with shaking for several days. For MG132 treatment, N. benthamiana leaves were sprayed with MG132 at a final concentration of 20 μm [2 mm stock in dimethyl sulfoxide (DMSO)] or 1% DMSO only, and leaf tissues were collected 2 h after spraying.
For protein extraction, leaf tissues were collected and ground in extraction buffer (50 mm Tris–HCl pH 7.5, 2 mm EDTA, 150 mm NaCl, 10% glycerol, 5 mm DTT, 1 mm phenylmethanesulfonylfluoride (PMSF), 1 μg ml−1 leupeptin, 1 μg ml−1 pepstatin, 1 μg ml−1 aprotinin) at 1.5 ml for 1 g of tissue. The homogenate was spun at 2000 g for 2 min, and the supernatant was boiled in 1× sodium dodecyl sulfate (SDS) protein loading buffer for 10 min. The total protein of 25 μl was loaded for each lane. Western blot was performed using a monoclonal anti-GFP antibody (Convance, http://www.covance.com/).
Plant transformation and transgenic plants selection
The p35S::CPR30 construct previously described (Gou et al., 2009) was transformed into cpr1-2, bon1-6, snc1-1 and pSNC1::SNC1WT:GFP plants using the floral dipping method (Clough and Bent, 1998). Transgenic T1 seeds were planted in soil and seedlings were sprayed with Basta for selection.
Bacterial growth assay
Pseudomonas syringae pv. tomato (Pst) strains DC3000 and DC3000 AvrRpt2 were grown on plates with 100 μg ml−1 kanamycin and 25 μg ml−1 rifamycin for 1 day. Two-week-old plants grown under constant light were infected by dipping with Pst DC3000 and DC3000 AvrRpt2 respectively at a concentration of 106 colony forming units (cfu) ml−1. Pathogen growth was assayed as previously described (Yang et al., 2006).
RNA blot analysis
Total RNAs were extracted from 3-week-old seedlings using Trizol reagent (Invitrogen) as instructed. 20 μg of RNA was resolved in a 1.2% agarose gel containing 1.8% formaldehyde. RNA blots were hybridized with gene specific, 32P labeled, single strand DNA probes. For the CPR1, the full-length sequences of its coding region were used as a probe. Probes for SNC1, PR1, and PR2 genes were as previously described (Gou et al., 2009).
GUS activity analysis
Strains of Pst DC3000 and DC3000 AvrRpt2 were grown as described above. Bacteria were collected, resuspended and diluted in 10 mm of MgCl2 to a final concentration of 0.05 at OD600. Leaves of 3-week-old plants were infiltrated with bacteria or 10 mm of MgCl2. Leaves were collected 24 h after infiltration and GUS activity analysis was performed as previously described (Gou et al., 2009).
We thank Yongqing Li for the SNC1 yeast two-hybrid constructs, Xin Li for the mos mutants, and Brian Staskawicz for bacterial strains. We thank Xin Li and Donglei Yang for discussions. This research was supported by a grant from the National Science Foundation to J.H. and Department of Plant Biology at Cornell University.