|Home | About | Journals | Submit | Contact Us | Français|
The plant innate immune response requires a rapid, global reprogramming of cellular processes. Here we employed two complementary proteomic methods, two-dimensional differential in-gel electrophoresis (2D-DIGE) and iTRAQ, to identify differentially regulated proteins early during a defense response. Besides defense-related proteins, the constituents of the largest category of up-regulated proteins were cytoplasmic- and endoplasmic reticulum (ER)-residing molecular chaperones. Silencing of ER-resident protein disulfide isomerases, NbERp57 and NbP5, and the calreticulins, NbCRT2 and NbCRT3, lead to a partial loss of N immune receptor-mediated defense against Tobacco mosaic virus (TMV). Furthermore, NbCRT2 and NbCRT3 are required for the expression of a novel induced receptor-like kinase (IRK). IRK is a plasma membrane-localized protein required for the N-mediated hypersensitive response programmed cell death (HR-PCD) and resistance to TMV. These data support a model in which ER-resident chaperones are required for the accumulation of membrane bound or secreted proteins that are necessary for innate immunity.
Almost 20 years ago, two-dimensional gel electrophoresis (2D-E) based proteomics method was employed to identify up-regulated defense related proteins using N immune receptor-containing Nicotiana tobacum plants and Tobacco mosaic virus (TMV) as a model system (Van Loon et al., 1987). These proteins were later discovered to be highly up-regulated in a variety of plants during defense responses and termed pathogenesis-related (PR) proteins (Reviewed by van Loon et al., 2006).
Since van Loon's classical experiments, there have been significant advances in gel-based proteomics techniques. Two-dimensional differential gel electrophoresis (2D-DIGE) has been developed to use different fluorescent dyes to tag different pools of proteins (Minden, 2007). Fluorescent dyes have a high dynamic range for accurate measurement of relative protein abundance levels and allow multiple sets of proteins to be compared. Multiple sets of tagged proteins can be simultaneously separated on a single two-dimensional gel and identical proteins within the pools co-migrate. The combination of internal standards and co-migration on a single gel removes a large portion of the variation associated with traditional 2D-E method.
In addition, in-solution proteomics have been developed for large scale, quantitative proteomics. Most new technologies implement differential tags to compare protein abundance levels from different pools of proteins. Isotopes can be introduced by stable isotope labeling with amino acids in cell culture (SILAC) or by labeling protein extracts using isotope-coded affinity tags (ICAT) (Gygi et al., 1999; Ong et al., 2002). The labeled peptides from different pools of proteins will have different masses on a mass spectrometer and their relative abundance levels can be measured. A newer method labels proteins with isobaric iTRAQ reagents that have the same mass but different fragmentation patterns on a mass spectrometer (Ross et al., 2004). Differential labeling combined with chromatography techniques, such as strong cation exchange and reverse phase chromatography, provides a high throughput method for comparing relative protein abundance levels from different samples.
To investigate the multitude of biological questions in plant innate immunity, we study the N immune receptor from N. glutinosa that provides resistance to TMV (Whitham et al., 1994). The N gene is the only cloned TIR-NB-LRR resistance gene to a virus and is functional in multiple Solanaceae species, including the model plant N. benthamiana (Liu et al., 2002). N recognizes the p50 region of TMV's replicases to signal hypersensitive response programmed cell death (HR-PCD) (Erickson et al., 1999). HR-PCD is observed as lesions at the infection sites and is correlated with the restriction of TMV.
The goal of this study was to employ two complementary advanced proteomics techniques, 2D-DIGE and iTRAQ, to identify novel components that are differentially regulated directly following pathogen recognition. Indeed, we found multiple PR proteins and a wide variety of proteins required for defense signaling and reactive oxygen species (ROS) production. More interestingly, we discovered that numerous cytoplasmic and endoplasmic reticulum (ER) residing chaperones were up-regulated. The function of cytoplasmic chaperones during innate immunity has been studied in detail by multiple research groups (Shirasu, 2009). Conversely, very little is known about the function of ER chaperones during innate immunity. Therefore, we investigated the biological significance of up-regulated ER chaperones during innate immunity. We show that protein disulfide isomerases (PDIs) and calreticulins (CRTs) are required for the N immune receptor to provide complete defense against TMV. Furthermore, our data suggests that ER chaperones are up-regulated during innate immunity to aid in the accumulation of a novel induced receptor-like kinase (IRK) required for a successful innate immune response.
Differentially-regulated proteins during an innate immune response should be investigated in the context of the whole organism because immune receptors, such as N, fail to function in protoplasts and cell culture (Beachy and Murakishi, 1973; Otsuki et al., 1972). Therefore, we studied an N-mediated defense response to TMV within whole N. benthamiana plants. One difficulty of studying defense responses in whole organisms is that during a normal infection only a small number of cells are infected, which results in a mixed population of uninfected and infected cells. To overcome this, we exploited the temperature sensitive nature of N-mediated resistance (Fraser and Loughlin, 1980) to coordinate a defense response in every cell (Figure 1A). Plants were shifted to 32°C to fully inactivate N and to allow TMV to spread without inducing a defense response. Plants were then shifted to room temperature to initiate a coordinated defense response in every cell. Tissue was collected at 0, 2, 8, and 16 hours post initiation of defense (herein referred to as T=0h, T=2h, T=8h, T=16h). Plants without N did not initiate a defense response, and were used as a control for any effects caused by TMV infection or the temperature shift. We used the same tissue samples in the two complementary methods, DIGE and iTRAQ (Figure 1B).
For DIGE, soluble proteins were extracted from plants with and without the N immune receptor at T=0, T=2, T=8, and T=16h post induction of defense (Figure 1B). Wild-type plant extracts were labeled with Cy3 and N-containing plant extracts were labeled with Cy5. Extracts from N-containing plants at T=0 were labeled with Cy2 to serve as an internal control. A representative 2D-DIGE gel (T=16) was cropped to contain only spots identified by MS (Figure 2A and Table 1). Red spots indicate up-regulated proteins, green spots indicate down-regulated proteins and yellow spots indicate no change. Spot volumes and relative abundance levels of co-migrating proteins were determined using the software package Decyder MS. Only spots that showed a 1.5-fold change were considered to be differentially regulated and subjected to MS. Protein identification was conducted using SEQUEST, the Trans-Proteomic Pipeline, and our in-house Nicotiana species database (see experimental procedures and supplementary text for details). From this analysis, we identified 24 spots that were up- or down-regulated at least 1.5-fold, had a protein probability score of ≥ 0.99, and were identified by at least 2 unique peptides (Table 1). 11 spots with a 1.5-fold change in regulation did not result in a significant identification. The proteins with a change in regulation were sorted into four major categories: proteins previously implicated as defense-related, proteins required for the production or regulation of ROS and hormones, molecular chaperones and general metabolic proteins.
Since our DIGE analysis only investigated the soluble proteome during a defense response, we conducted iTRAQ analysis with the exact same tissue samples and time points to obtain global quantification data (Figure 1B). Proteins extracted from plants without N (no defense response) were labeled with iTRAQ reagent 114 and proteins extracted from plants containing N (defense response) were labeled with iTRAQ reagent 117 (see experimental procedures and supplementary text). Multidimensional Protein Identification Technology (MuDPIT) was performed on a QSTAR XL MS and identified 368 proteins at T=2h (175 single hits) and 775 proteins at T=8h (492 single hits). 114/117 ratios were calculated using the LIBRA module of the TPP software.
iTRAQ 114/117 ratios for 9124 peptides that showed >90% probability of identification confidence were graphed to choose up- and down-regulation cutoff values (Figure S1). Similar methods have been used previously for iTRAQ and ICAT experiments (Boersema et al., 2009; Pawlik et al., 2006). 91% of the peptides fell between -1.5 and 1.5 ratios of 114/117 and were designated as having no change in regulation. Approximately 7.8% and 1.3% of the peptides were up- and down-regulated >1.5-fold respectively. This corresponds to a total of 82 differentially-regulated proteins, with only 37 proteins meeting our stringent identification criteria (Table 2). More specifically, 8 proteins were up- or down-regulated in iTRAQ at both T=2 and T=8h. 11 and 18 proteins were up- or down-regulated in iTRAQ at only T=2 or T=8h respectively. An additional 45 proteins were differentially-regulated, but their identifications were supported by only one unique peptide (Table S1). 11 differentially regulated proteins identified in iTRAQ were also identified in DIGE (Figure 2B-E and Table S2). The ratios of DIGE and iTRAQ were highly similar and have a Pearson correlation of 0.89 for both time points, partially validating these two independent approaches.
Similar to the classic experiments that originally identified PR proteins (Van Loon et al., 1987), our DIGE and iTRAQ analyses also discovered pathogenesis-related proteins, PR-Q, PRp27 and PR-R. Furthermore, we observed an up-regulation of HSR203J that was previously shown to be transcriptionally up-regulated during the HR-PCD (Takahashi et al., 2004). In addition, we observed an up-regulation of alanine aminotransferase (AlaAT), which was shown to be transcriptionally up-regulated in Capsicum annuum undergoing a TMV-P0- or Xanthomonas campestris-induced HR-PCD (Kim et al., 2005). Our data suggests that the transcriptional up-regulation of HSR203J and AlaAT also results in an up-regulation of their protein abundance, and that it occurs early on during HR-PCD.
A surprising result from this study was the down-regulation of polyphenol oxidase (PPO) and chloroplastic drought-induced stress (CDSP34/fibrillin) proteins that were previously shown to be transcriptionally up-regulated during bacterial innate immune responses (Langenkamper et al., 2001; Thipyapong et al., 2004). PPOs are released from disrupted chloroplasts and oxidize phenolic compounds to quinines (Mayer, 2006). PPO overexpression enhanced defense against Pseudomonas syringae (Li and Steffens, 2002; Thipyapong et al., 2004). Surprisingly, instead of an up-regulation, we observed a down-regulation of PPO and CDSP-34/fibrillin (Table 1, ,2,2, and Table S2). These results show apparent differences in defense responses to bacteria and viruses.
One of the early hallmarks of a defense response are bursts of ROS such as hydrogen peroxide (H2O2), superoxide (O2-) and nitric oxide (NO) (Delledonne et al., 1998; Lamb and Dixon, 1997). As a result, a large percentage of the identified up-regulated proteins in our study have been implicated previously during the production or modulation of ROS (Table 1 and and22).
Interestingly, chloroplast-generated ROS are induced by a SIPK/Ntf4/WIPK MAP kinase pathway and play a crucial role during TMV-induced HR-PCD (Liu et al., 2007). The induction of H2O2 occurs in a light-dependent manner, suggesting that it is a product of the photosynthetic electron transport (PET) chain of photosystem I (PS I). The PET chain can result in the photo-reduction of O2 to O2-·, which can then be converted to H2O2 by superoxide dismutase (SOD). We observed an up-regulation of components of PS I in iTRAQ and SOD in DIGE at T=2 h, which correlates with the early burst of H2O2 during HR-PCD (Table 1 and and2).2). H2O2 also may be formed by the partial water reduction by photosystem II (PS II) (Fine and Frasch, 1992), and indeed, we found that the manganese-containing 23-kDa protein of the PS II oxygen-evolving complex was up-regulated at T=8 h in iTRAQ. Furthermore, ascorbate peroxidase, which uses ascorbate to detoxify H2O2, was down-regulated at T=8h in DIGE. All of these changes in regulation would culminate in an increase in H2O2 levels during a defense response.
Multiple heat shock proteins (HSPs), and their associated cofactors, interact with immune receptors to modulate the accumulation or folding of immune receptor complexes (Reviewed by Shirasu, 2009). Interestingly, we discovered an up-regulation of SGT1, HSP90, and HSP70. SGT1 has cochaperone features including tetratricopeptide repeats found in HSP70/HSP90 organizing proteins (Dubacq et al., 2002). HSP90 directly associates with the N immune receptor and SGT1 (Liu et al., 2004). Silencing of HSP90 or SGT1 results in a loss of N-mediated immunity to TMV (Liu et al., 2004; Peart et al., 2002). Similarly, HSP90 is required for the function of the RPM1, RPS2, and Rx immune receptors (Hubert et al., 2003; Lu et al., 2003a; Takahashi et al., 2003). HSP90 was up-regulated at T=2h after the initiation of a defense response (Table 2).
HSP70 homolog that was up-regulated in DIGE (Table 1) is similar to the Arabidopsis HSC70 homologs that associates with SGT1 and are up-regulated during a defense response (Noel et al., 2007). SGT1 is thought to function as a cochaperone of HSP90 and HSP70/Hsc70. Two peptidyl-prolyl isomerases called immunophillin rotamase FKBP (ROF1) were up-regulated in DIGE and could act as an HSP90/HSP70 organizing protein like SGT1 (Table 1). ROF1 and other immunophillins have been shown to directly interact with HSP90 in both plants and animals (Harrell et al., 2002; Owens-Grillo et al., 1996). Other FKBPs, similar to ROF1, bind to the glucocorticoid receptor (GR) and dynein to aid in the movement of GR from the cytoplasm to the nucleus (Harrell et al., 2002).
One of the most intriguing results from this study was the discovery that a large number of ER resident chaperones were up-regulated during N-mediated defense. These include protein disulfide isomerases (PDI), ERp57, P5, calreticulin 3 (CRT3), glucose regulated protein 78 (GRP78), and luminal-binding protein 5 (BiP5). Interestingly, the function of these proteins has not been examined extensively during plant innate immunity. Since the cytoplasmic chaperones HSP90, HSP70 and SGT1 have crucial functions during a defense response, we hypothesized that ER-resident chaperones, which were similarly up-regulated, might be another important class of proteins for defense. Hence, the rest of our study focused on these ER chaperones and their function during plant innate immunity.
PDIs function during the folding and the formation of disulfide bonds in the ER (Reviewed by Christis et al., 2008). We discovered that three PDI homologs were among the highest up-regulated proteins in our study (Table 1 and and2).2). Nt|TC19383 contains two PDI-b domains sandwiched by two PDI-a domains, which suggests it is a homolog to canonical PDI. The second PDI was homologous to ERp57 and was represented by the ESTs Nb|TC11094 and Nb|CK295775 that overlap by 163 nucleotides with 97% identity. Indeed, we were able to amplify the full-length sequence of NbERp57 from N. benthamiana, suggesting they are transcribed from the same gene. The third PDI, Nt|TC14539, had a domain structure similar to P5 and was identified with one unique peptide in iTRAQ at T=8 h (Table S1).
We also observed an up-regulation of N. benthamiana calreticulin 3 (NbCRT3) homolog (Table 1). Calreticulin is a lectin-like chaperone that interacts with ERp57 and mediates the association of ERp57 with its substrates (Maattanen et al., 2006). They function together during the proper folding of ER resident proteins. In addition, we discovered another chaperone, GRP78-5 that was up-regulated in iTRAQ (Table 2). A multi-chaperone complex that includes, PDI, P5, GRP78, GRP94, and ERp72 was found to function during the folding of interferon-gamma in the ER of human cell lines (Vandenbroeck et al., 2006). In this study, we discovered that 3 out of 5 homologs of these chaperones were up-regulated during a plant innate immune response.
To verify the up-regulation of some of these chaperones, we performed Western blot analyses. Analysis with CRT antibodies showed a strong up-regulation of CRT as early as T=2 (Figure 3, top panels), even though the DIGE analysis suggests that CRT3 is only significantly up-regulated at T=16. The CRT antibodies detected three bands, a nonspecific calnexin (top), NbCRT2 (middle), and NbCRT3 (bottom). Although there are three CRTs in Arabidopsis, currently only homologs to CRT2 and CRT3 are represented in the available Nicotiana sequence information. Analysis with HSP90 antibodies showed a slight up-regulation at T=2 and stronger up-regulation at T=8 (Figure 3, middle panels). These data are interesting because HSP90 transcripts were not up-regulated during a defense response in our previous study (Liu et al., 2004); hence, the increase in HSP90 must occur on the post-transcriptional level. Our analysis with Hsc70 antibodies indicates no change during a defense response (Figure 3, lower panels). This may be due to the high homology between members of the HSP70/Hsc70 super-family. Consistent with this, we observed a fairly large variation for the abundance ratios of different HSP70 homologs. Furthermore, other homologs to HSP70 showed no change in regulation, such as Nb|TC16565 (7 unique peptides; data not shown).
To determine the biological significance of the up-regulation of PDIs and CRTs, we used Tobacco rattle virus (TRV)-based virus induced gene silencing (VIGS) (Liu et al., 2002) to knock down the expression of ERp57, P5, CRT3 and CRT2. After approximately 10 days of silencing, the plants were infected with either TMV-U1 tagged with GFP (TMV-U1-GFP) or untagged TMV-U1. The loss-of-resistance phenotype was observed as spreading HR-PCD to the upper-uninoculated leaves. The partial knock down of components required for N-mediated resistance to TMV results in a loss of viral containment to the inoculated leaf, but does not fully disrupt HR-PCD. As a result, we observe HR-PCD where TMV spreads. For example, silencing the N gene (our positive control) results in spreading HR-PCD (Figure 4A, column 2). The causation of spreading HR-PCD by the movement of TMV into the upper uninoculated leaves was verified by semi-quantitative reverse transcriptase (RT) PCR (Figure 4B, column 2). We discovered that silencing NbERp57, NbP5, NbCRT3 and NbCRT2 resulted in partial movement of TMV (Figure 4 and Table S3). TMV-U1 moved in 56% of ERp57-, 71% of P5-, 80% of CRT3-, and 73% of CRT2-silenced plants using 26 – 40 biological replicates over 6 independent experimental replicates (Figures S2 and Table S3). TMV moved in 100% of N-silenced plants and 6.45% of the vector-silenced plants. Hence, the TMV movement in NbERp57, NbP5, NbCRT3 and NbCRT2 was statistically greater than our vector-silenced control, but less than N-silenced plants, suggesting the silencing partially disrupts N-mediated resistance to TMV. The partial disruption of resistance may be caused by the partial knock down of transcripts, since mRNA levels were only reduced 75.7% ± 3.1% for NbERp57, 60.4% ± 7.8% for NbP5, 78.1% ± 3.7% for NbCRT3, and 45.0% ± 5.5% for NbCRT2 (Figure S2). On the protein level, NbCRT2 abundance was reduced but still detectable in silenced plants (Figure 4C, lanes 7-9). In the NbCRT3-silenced plants, NbCRT3 protein was decreased to undetectable levels (Figure 4C lanes 4-6), which may have resulted in the more penetrant loss-of-resistance phenotype. Collectively, these results indicate that NbERp57, NbP5, NbCRT3 and NbCRT2 are required for the N immune receptor to activate a complete defense response.
We hypothesized that one possible function for the up-regulation of ER-resident chaperones was to aid in the folding or processing of membrane proteins required for plant innate immunity. A recent study suggests that ER-resident chaperones are required for the active accumulation of the receptor-like kinase, EFR, which is required for PAMP-triggered innate immunity (PTI) (Saijo et. al, and Zipfel et. al., submitted). Previously, our lab identified an induced receptor-like kinase (IRK) that was up-regulated ~5 fold within 2 hours during an N-mediated defense response in microarray studies (data not shown). We confirmed the rapid up-regulation of NbIRK during a defense response using RT-PCR (Figure 5A). Like EFR, IRK is a plasma membrane localized leucine rich repeat (LRR) receptor-like kinase (Figure 5B). IRK shows highest homology to the Arabidopsis receptor-like kinase, At4g23740, and contains 5 extracellular LRR repeats and an intracellular kinase domain (Figure S3).
To determine if IRK is required for N to function, we used VIGS to knock down IRK expression. Indeed, silencing IRK resulted in a loss-of-resistance to TMV in ~77.3% of the silenced plants in two independent experiments. Similar to above, we observed the loss-of-resistance phenotype as spreading HR-PCD to the upper leaves (Figure 5C) and TMV movement to upper leaves was confirmed by RT-PCR (Figure 5D). To further characterize the requirement of IRK, we conducted an HR-PCD assay by expressing TMV-p50 effector. In the VIGS-Vector control, HR proceeded normally as PCD leading to tissue collapse (Figure 5E, left panels). In the VIGS-IRK silenced plants, HR was attenuated to yellowing of the tissue with very little cellular collapse (Figure 5E, right panels). These results suggest that IRK is required for early signaling during the initial HR progression triggered by the N immune receptor.
Since IRK plays an important role during innate immune responses, we investigated the requirement of CRT3 and CRT2 for the accumulation of IRK. For this, we expressed IRK-Citrine in Vector-, CRT3- and CRT2-silenced plants. IRK protein was greatly reduced in CRT3-silenced plants and partially reduced in CRT2-silenced plants (Figure 5F). This suggests that both CRT3 and CRT2 have a function during the folding and accumulation of IRK. As a control, we expressed the known plasma membrane-localized protein RIN4-Citrine. The accumulation of RIN4-Citrine was not affected by silencing of CRT3 or CRT2 (Figure 5F). These results suggest that one of the reasons CRT3 and CRT2 is induced early during a defense response is to fold plasma membrane localized proteins required for innate immunity.
We implemented DIGE and iTRAQ proteomic methods to identify novel proteins that are up-regulated early during a defense response. Indeed, we identified previously discovered defense proteins that constitute the machinery and enzymes required for the rapid burst of ROS and defense hormone signals. This study also showed that cytoplasmic molecular chaperones that were previously shown to be required for plant innate immunity are up-regulated early during a defense response. More importantly, we have demonstrated that components of ER-resident multi-chaperone complexes are similarly up-regulated and are required for a successful defense response. We propose that the up-regulated ER chaperones function during the folding and subsequent accumulation of proteins required for innate immunity. To test this model, we analyzed the expression of a novel receptor-like kinase, IRK, and discovered that it requires CRTs to accumulate. Furthermore, IRK is required for the N immune receptor to fully mount HR-PCD and subsequent defense against TMV.
The lack of a full genome sequence and a complementary protein database for the model plant species N. benthamiana makes it extremely challenging to identify proteins by MS. As a result, we created a high quality protein database in this study for large-scale proteomic experiments in N. benthamiana. In our iTRAQ experiments, searching our new database resulted in an approximately 16% increase in identifications compared to searching a six-frame translated database composed of ESTs from N. benthamiana, N. tobaccum, Lycospericon esculentum, Capsicum annuum, Solanum tuberosum, and Petunia hybrida (data not shown). We also developed a relational-database system that can store, compare, and sort proteomic data. These tools will provide the framework for future proteomic experiments, such as identification of protein complex components by tandem affinity purification.
Multi-chaperone complexes have been implicated in both plant and animal defense against pathogens. In this study, we show that HSP90, SGT1 and HSP70 homologs are up-regulated early during a defense response. SGT1 directly binds to HSP90 and to HSC70, a close homolog to HSP70, possibly bridging their chaperone functions (Noel et al., 2007; Takahashi et al., 2003). SGT1 is also transcriptionally up-regulated during defense responses to Hyaloperonospora parasitica (Azevedo et al., 2006), suggesting that the up-regulation of SGT1 is not specific to N immune receptor responses. HSP90 also has been implicated during effector-triggered immunity (ETI) by multiple R immune receptors, including N (Hubert et al., 2003; Liu et al., 2004; Takahashi et al., 2003). Interestingly, HSP90 associates with N, MLAs and I-2 immune receptors (Bieri et al., 2004; de la Fuente van Bentem et al., 2005; Liu et al., 2004). Along with HSP90, SGT1 interacts with the co-chaperone RAR1 (for required for Mla12 resistance) and functions additively or independently during RPP5 immune receptor-mediated defense (Austin et al., 2002; Azevedo et al., 2002) and regulates barley MLA levels (Bieri et al., 2004).
Originally, HSP90, RAR1, and SGT1 were thought to function together during the accumulation of R immune receptors. However, it is now clear that different immune receptors have different requirements for HSP90, RAR1, and SGT1 during immune receptor accumulation and activation. HSP90, RAR1, and SGT1 are required for the accumulation of the Rx immune receptor; furthermore, the association of SGT1 and HSP90 was required for that accumulation (Azevedo et al., 2006; Boter et al., 2007). Conversely, SGT1 antagonized immune receptor accumulation by RAR1, possibly by guiding them to host cellular degradation machinery (Holt et al., 2005). This is supported by specific mutations in HSP90.2 that suppress the rar1 mutant phenotype, suggesting that in rar1 mutants HSP90 is not properly regulated and this results in a disruption of immune receptor accumulation (Hubert et al., 2009). Furthermore, SGT1 is required for N immune receptor accumulation (Mestre and Baulcombe, 2006).
The varying requirements of HSP90 and RAR1 may be caused by different pools of HSP90 having different functions. For example SGT1 may not be the only HSP90/HSP70 organizing protein involved in defense. In this study we discovered an up-regulation of two ROF1-like proteins that are known to function with HSP90, which may partially account for the varying requirements for these chaperones during innate immunity. ROF1 associates with HSP90 and dynein, which is required for the retrograde movement of the glucocorticoid receptor along microtubules from the cytoplasm to the nucleus (Galigniana et al., 2001). Furthermore, these associations and functions are conserved between plants and animals (Harrell et al., 2002). The function of the ROF1/HSP90 chaperone complex during the maturation and localization of immune receptors should be investigated further.
In this study, we show that multiple components of ER chaperone complexes are up-regulated during a defense response. PDI chaperone complexes function during the formation of disulfide bonds between cysteines, which is often the rate limiting step of protein folding (Creighton et al., 1995; Hatahet and Ruddock, 2007). More generally, they function with other ER molecular chaperones like CRTs, BiPs, GRP78, and GRP94 to form functionally competent proteins. Although the exact roles of PDIs during innate immunity were previously unknown, recent research suggests that ER chaperones may have a general role during all forms of plant innate immunity. The basal immune receptor, EFR, requires ER chaperones to accumulate during PTI (Saijo et. al. and Zipfel et al., submitted). The rapid transcriptional up-regulation of a PDI homolog during fungal-induced HR (Ray et al., 2003) and viral-induced HR (this study) suggests ER chaperones function early during effector-triggered immunity (ETI). Following PTI and ETI, plants induce systemic acquired resistance (SAR), which provides enhanced resistance to incoming pathogens. During SAR, ER chaperones, including PDIs, CRTs and GRP94, are primary target genes of NPR1 and are necessary for the secretion of PRs (Wang et al., 2005).
We discovered three distinct homologs to PDIs: archetypal PDI, ERp57, and P5. The identified plant NbERp57 is the closest homolog to mammalian ERp57 and was one of the most up-regulated proteins in our study. ERp57, and other PDIs, function with CRT in the ER as multi-chaperone molecular complexes required for the proper conformational folding and function of proteins (Reviewed by Christis et al., 2008). If a protein does not fold properly, it is removed by ER-associated degradation (ERAD) machinery (Anelli and Sitia, 2008). Hence, ERp57 and CRT may regulate the folding and maturation of immune receptors or other factors required for innate immunity. The N immune receptor does not localize or fold in the ER and disruption of ER chaperones had no effect on the accumulation of N (data not shown). However, confocal microscopy experiments examining other Arabidopsis NB-LRR immune receptors found many of them localizing to the ER (data not shown), and hence, they may require ER chaperones for maturation.
Silencing of NbCRT2 and NbCRT3 lead to a partial loss-of-resistance that was more penetrant than silencing the PDIs, NbERp57 or NbP5. However, the PDI family is quite large in N. benthamiana, and consequently, there may be more functional redundancy compared to CRTs where there are only two homologs. Alternatively, the stronger phenotype may be caused by the dual function of CRTs as molecular chaperones and as calcium buffers. The C-domain of calreticulin can store a large percentage of total ER calcium (Molinari et al., 2004), and may function during the initial calcium burst during HR-PCD. In the future, we would like to measure calcium levels in CRT-silenced plants during HR-PCD and what effect silencing may have on the rate of HR-PCD onset.
To test the possible role of ER chaperones during the folding of proteins required for innate immunity, we chose to study the function of CRTs rather than PDIs for two reasons. First, an antibody that recognizes both forms of CRT was available. This allowed us to confirm the up-regulation of CRTs and measure the level of silencing in VIGS experiments. Second, there are only two CRT homologs making it a fairly easy protein family to study. Since our data suggested that CRTs function is not required for the folding of the N immune receptor, we then took a candidate approach for substrates of CRT. IRK was our top candidate because it is transcriptionally up-regulated and is a plasma membrane protein; hence, it is processed in the ER. As expected, disrupting CRT3, and to some extent, CRT2 resulted in a decrease in the expression of IRK. This supports our model that CRTs may be up-regulated to aid in the folding and accumulation of factors required for HR-PCD and defense. Alternatively, it is possible that silencing CRTs, PDIs and IRK indirectly suppress the plant innate immune system. Irrespective of direct or indirect role, the identification of IRK as a plasma membrane localized receptor-like kinase required for viral ETI is an important discovery. Our data show that IRK is necessary for both the progression of HR-PCD and for the N to provide complete resistance to TMV. Future studies will be aimed at how a cytoplasmic and nuclear localized N immune receptor-mediated recognition of viral effector (Burch-Smith et al., 2007; Caplan et al., 2008) signals to activate or function with a plasma membrane-localized IRK.
In summary, this research has opened up two related, but distinctly different new areas of research in plant innate immunity. First, we have implicated the up-regulation and requirement of ER chaperones during a defense response to TMV. Since the up-regulation is downstream of pathogen-recognition, we hypothesize that the requirement of ER-resident chaperones may be a general, but crucial, requirement for innate immunity to a variety of pathogens. Hence, it will be exciting to see what other plant immune receptors require ER chaperones to function. Second, this study led to the identification and initial characterization of IRK, a novel RLK required for innate immunity. IRK requires ER-resident chaperones to accumulate and currently has an unknown function during a defense response. Future elucidation of IRK's biological role during plant defense will lead to new insights into how HR progresses and leads to innate immunity.
Protein was extracted from plants with or without the N gene at T=0h, T=2h, T=8h and T=16h post induction of a defense response (see supplementary experimental procedures for details). 2D-DIGE was conducted by Applied Biomics, Inc. (Hayward, CA). Wild-type plant extracts were labeled with Cy3 and NN plant extracts were labeled with Cy5 or with Cy2. N-containing plant extracts at T=0h was labeled with Cy2 as an internal control. The expression of N-containing and wild-type plants was compared at T=2h and T=8h. The samples were run on a single 2D-PAGE gel where the first dimension was performed by isoelectric focusing with pH 3-10 gel strips. The second dimension was conducted on a SDS-polyacrylamide gel. The Cy2, Cy3, and Cy5 labeled proteins were measured using excitation/emission wavelength of 488/520 nm for Cy2, 532/580 nm for Cy3, and 633/670 nm for Cy5. The relative expression levels were determined using the program Decyder MS (GE Healthcare). Identification of DIGE samples were performed by the LCQ Deca XP Plus ion trap mass spectrometer (Thermo Fisher Scientific) at the Midwest Bio Services, Inc. KS (see supplementary experimental procedures for details).
Protein was extracted with ProteoPrep (Sigma-Aldrich) chaoptropic membrane extraction buffer 3 from plants with or without the N gene at T=2h and T=8h. Protein was precipitated with 4 volumes of ice cold 100% acetone at −20°C for 1h and then washed three times with 80% acetone. 100 μg of total protein extract in digestion buffer was reduced, blocked and digested with 10 μg of trypsin overnight according to the manufacturer's protocol (Applied Biosystems). Sample from plants with and without the N gene were labeled with iTRAQ 117 and iTRAQ 114 reagents (Applied Biosystems) respectively. The samples were pooled and purified by strong cation exchange columns (Applied Biosystems), and labeled proteins were eluted with 5% ammonium hydroxide in 30% methanol. MuDPIT analysis was performed on a QSTAR XL mass spectrometer with a nano-spray II ion source (ABI, CA) at the University at Albany (see supplementary experimental procedures for details).
A high quality Nicotiana protein database (DKDB v7.1) was created and will be available upon request (see supplementary experimental procedures for details). DIGE and iTRAQ data was not pre-filtered. Proteins were identified using a pipeline consisting of SEQUEST, TPP and a custom relational database (see supplementary experimental procedures for details).
We thank members of the S.P. D-K lab for helpful comments on the manuscript and Dr. Boston and Dr. Krishna for CRT and Hsp90 antibodies. We thank Dr. Qishan Lin at the State University of NY at Albany for mass spectrometry services. Supported by NSF-DBI-0211872 and NIH-GM62625 grants to S.P. D-K.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.