Report

Induction of Protein Secretory Pathway Is Required for Systemic Acquired Resistance

See allHide authors and affiliations

Science  13 May 2005:
Vol. 308, Issue 5724, pp. 1036-1040
DOI: 10.1126/science.1108791

Abstract

In plants, systemic acquired resistance (SAR) is established as a result of NPR1-regulated expression of pathogenesis-related (PR) genes. Using gene expression profiling in Arabidopsis, we found that in addition to controlling the expression of PR genes, NPR1 also directly controls the expression of the protein secretory pathway genes. Up-regulation of these genes is essential for SAR, because mutations in some of them diminished the secretion of PR proteins (for example, PR1), resulting in reduced resistance. We provide evidence that NPR1 coordinately regulates these secretion-related genes through a previously undescribed cis-element. Activation of this cis-element is controlled by a transcription factor that is translocated into the nucleus upon SAR induction.

SAR is a plant immune response that is induced after a local infection and confers resistance throughout the plant to a broad spectrum of pathogens (1). Induction of SAR requires accumulation of the endogenous signaling molecule salicylic acid (SA), which activates gene expression mediated by the master regulator protein NPR1 [Nonexpressor of pathogenesis-related (PR) genes 1, also known as NIM1] (2). Exogenous application of SA triggers the translocation of NPR1 into the nucleus. Once in the nucleus, NPR1 interacts with TGA transcription factors to mediate gene expression. Treatment of plants with SA alters the expression of a large array of genes (3-5). Among these, only some are regulated by NPR1 and therefore specific to SAR. The most-studied NPR1 targets are PR genes, which encode small, secreted or vacuole-targeted proteins with antimicrobial activities (6).

To identify additional NPR1 target genes, we used the 35S::NPR1-GR transgenic line generated in the npr1-3 mutant (7). In this transgenic plant, nuclear translocation of NPR1-GR (GR, glucocorticord receptor) requires not only SA but also dexamethasone (Dex). Treating 35S::NPR1-GR plants first with SA and then with Dex specifically activates NPR1 target genes. With the use of a known NPR1 primary target, PR1, experimental conditions were optimized with the inclusion of the translation inhibitor cycloheximide (Chx) to achieve maximal induction of NPR1 target genes in the absence of de novo protein synthesis (fig. S1).

Using Affymetrix GeneChips (8200 genes), we identified putative NPR1 primary target genes by comparing transcriptional profiles of npr1 and npr1/35S::NPR1-GR that were both treated with SA and then Chx/Dex (fig. S1). Duplicate experiments were performed independently and the data were analyzed with both MAS5.0 and dChip programs. Genes that showed a consistent difference in their pattern of expression (induction or repression) and low P-values (<0.05) in both replicates were considered for further analysis (table S1). Many of the induced genes can be classified into groups according to their known or deduced functions. One group (Table 1) contains genes known to be involved in defense, including several PR genes. Another group encodes members of the protein secretory pathway (9 of the 49 genes with >2-fold induction or 18 of the 120 genes with >1.6-fold induction), most of which are endoplasmic reticulum (ER)–localized proteins (8, 9). These secretion-related genes include those encoding the Sec61 translocon complex, which provides a channel for proteins to cross the ER membrane, and a signal recognition particle (SRP) receptor, which directs proteins with a signal peptide to the translocon complex. NPR1 also regulates many genes encoding ER-resident chaperones, such as BiP2 and glucose regulated protein 94 (GRP94), as well as co-chaperones including defender against apoptotic death 1 (DAD1) (10), calnexins (CNXs), calreticulins (CRTs), and protein disulfide isomerases (PDIs). These proteins function in the cotranslational folding and modification (e.g., disulfide bond formation and glycosylation) of nascent polypeptides destined for the apoplast or various organelles. Other genes in this group encode a Golgi-associated membrane trafficking protein; a clathrin, which is involved in packaging secretory proteins into small vesicles; and a vacuolar sorting receptor.

Table 1.

A partial list of primary target genes of NPR1 identified in the microarray experiments. Data from two independent biological replicates are presented. Fold changes (F. C.) and P-values were calculated with the MAS 5.0 package, and similar results were obtained with dChip (21). Both groups of genes are shown to be significantly induced by a mixed-model ANOVA test with P = 7.6 10-14 for the defense genes and P = 0.003 for the secretory pathway genes.

Gene description Locus Replicate 1 Replicate 2
F. C. P-value F. C. P-value
Defense
PR-1 At2g14610 42.2 1.0 × 10-6 256.0 0
PR-5 At1g75040 14.9 0 2.6 0
Endochitinase At2g43570 2.1 1.8 × 10-4 2.5 0
Putative disease resistance protein At4g12010 6.1 5.2 × 10-3 2.6 2.4 × 10-4
Disease resistance protein RPP8 At5g43470 2.5 9.2 × 10-5 2.0 0
Beta-1,3-glucanase At4g34480 2.0 9.9 × 10-5 2.8 1.0 × 10-6
Chitinase At2g43570 2.1 1.8 × 10-4 2.5 0
Peroxidase At5g64120 2.1 4.7 × 10-5 2.6 0
Endoxyloglucan transferase At5g57550 3.5 1.1 × 10-5 29.9 0
Protein folding and secretion
Signal recognition particle receptor At2g45770 4.9 3.1 × 10-5 6.1 0
Sec61α subunit At2g34250 2.1 5.4 × 10-5 1.5 0
Sec61β subunit At2g45070 1.4 4.0 × 10-2 1.3 1.0 × 10-2
BiP2 At5g42020 3.2 7.0 × 10-6 2.5 0
GRP94 At4g24190 7.5 0 2.3 0
DAD1 At1g32210 3.0 1.0 × 10-6 2.0 0
Protein disulfide-isomerase (PDI) At2g47470 3.2 3.1 × 10-5 2.0 1.0 × 10-6
Protein disulfide-isomerase (PDI) At3g54960 1.6 8.0 × 10-4 3.0 1.0 × 10-6
Calreticulin 3 At1g08450 4.0 0 2.3 0
Calreticulin At1g09210 4.3 1.0 × 10-6 3.1 0
Calnexin 1 At5g61790 6.1 4.3 × 10-5 2.6 0
Calnexin 2 At5g07340 1.4 5.0 × 10-6 2.1 1.2 × 10-5
Ribophorin I At2g01720 3.7 8.0 × 10-6 1.5 1.0 × 10-6
Tetratricoredoxin At4g22670 3.2 0 1.9 0
Cyclophilin At2g47320 4.0 2.0 × 10-6 1.4 7.9 × 10-5
Clathrin-coat assembly protein At1g10730 3.7 5.8 × 10-2 2.0 4.7 × 10-5
Transmembrane trafficking protein At1g14010 4.3 2.2 × 10-4 2.0 1.0 × 10-6
Vacuolar sorting receptor At1g30900 7.0 1.0 × 10-5 3.5 0

Taken as a group, the secretion-related genes showed statistically significant up-regulation in both experiments using a mixed-model analysis of variance (ANOVA) (P = 0.003) (11). Furthermore, the induction of many of these genes by SA via the endogenous NPR1 was confirmed by RNA blot analysis, real-time reverse transcription–polymerase chain reaction (RT-PCR) (fig. S2), and with data from public microarray databases, such as the Stanford Microarray Database (12).

Although several previous studies have noted the induction of a few individual secretion-related genes by plant defense elicitors, the importance of this induction has only been speculated on (13-17) and the regulatory mechanism is not known. During SAR, there is a massive buildup of PR proteins in vacuoles and the apoplast. The basal activity of the protein secretory pathway may not be sufficient to accommodate the marked increase in PR protein synthesis. Therefore, we hypothesized that a coordinated up-regulation in the protein secretory machinery is required to ensure proper folding, modification, and transport of PR proteins. Consistent with this hypothesis, the ER-resident gene, BiP, was shown to be induced before the accumulation of PR1 (fig. S3) (16).

To provide genetic evidence that the up-regulation of the secretion-related genes is essential for SAR, we identified knockout mutants in five secretion-related genes from the Salk Institute transferred DNA (T-DNA) insertion collection (18). Mutants of a calnexin (Salk_044381; At5g07340) and a PDI (Salk_046705; At2g47470) gene showed no significant change in induced resistance. Because the Arabidopsis genome contains six calnexin and the related calreticulin genes, and more than 20 PDI genes, the lack of a phenotype in these mutants is likely due to functional redundancy. On the other hand, T-DNA insertions in BiP2 (Salk_047956), DAD1 (Salk_046070), and SEC61α (Salk_034604) all compromised the plant's ability to efficiently secrete PR1 after treating with benzothiadiazole S-methylester (BTH, an SA analog) (Fig. 1A). The reduction in PR protein secretion directly correlates with impaired resistance against the bacterial pathogen Pseudomonas syringae pv. maculicola ES4326 (Fig. 1B). In the sec61α bip2 and dad1 bip2 double mutants, less PR1 was secreted and more pathogen growth was detected compared with the single mutants. Thus, an intact and responsive protein secretory pathway is required for SAR.

Fig. 1.

Effects of mutations in secretion-related genes on PR1 protein secretion and resistance against Pseudomonas syringae pv. maculicola ES4326 (Psm ES4326). (A) Immunoblot of PR1 protein in the secretion-related gene mutants. Intercellular wash fluid (IWF) was collected from equal amounts of plant tissue treated with BTH or untreated. As a control, total protein was extracted. Secreted (E) and total (T) PR1 was examined using separate immunoblots with an antibody raised against this protein. An antibody against α-tubulin was then used to probe the total protein blot to confirm equal loading of the samples. The same antibody was also used to probe the IWF blot to demonstrate that no intracellular protein (α-tubulin) leaked out of the cell during IWF preparation (21). This experiment was repeated three times with similar results. (B) Growth of Psm ES4326 in the secretion-related gene mutants after BTH induction. Plants were induced with 60 μM BTH 24 hours before infiltration with Psm ES4326 (optical density at 600 nm = 0.001). Uninduced WT plants were infiltrated at the same time. Bacterial growth was monitored 2 days after infection (27). Error bars represent 95% confidence limits of log-transformed data from eight independent samples. cfu, colony-forming units.

The importance of this coordinated induction of PR and secretion-related genes by NPR1 was further demonstrated by additional characterization of the bip2 mutant. In Arabidopsis, there are three BiP genes (19, 20). Knocking out the NPR1-regulated BiP2 gene resulted in reduced accumulation of total BiP protein after BTH induction (to ∼50% of the wild-type level) (Fig. 2A). As a result, the bip2 mutant is not only impaired in BTH-induced resistance (Fig. 1B), but also is hypersensitive to treatment of BTH or 2,6-dichoroisonicotinic acid (INA, another SA analog). Application of these chemicals at high concentrations to the bip2 mutant resulted in a rapid tissue collapse not seen in wild-type (WT) or untreated mutant (Fig. 2B) (21). We believe that in the bip2 plants, the increased PR protein synthesis is not accompanied by a sufficient increase in BiP protein, causing intracellular accumulation of unfolded proteins, leading to an acute unfolded protein response (UPR) in the form of cell death. In mammalian cells, free BiP binds the UPR sensor Ire1 (a kinase and endonuclease on the ER membrane) and prevents it from dimerizing and activating the UPR. When unfolded proteins accumulate, BiP dissociates from Ire1 and binds the unfolded proteins, thus freeing Ire1 to activate UPR (22). Indeed, in the bip2 mutant, several UPR marker genes, which are also NPR1-responsive, are hyperactivated after BTH treatment (Fig. 2C). The bip2 mutant plants are also more sensitive to inducers of UPR, such as tunicamycin, which causes misfolding of proteins by inhibiting glycosylation (23). Whereas WT plants recovered from tunicamycin treatment, bip2 plants failed to do so (Fig. 2B).

Fig. 2.

UPR in the bip2 mutant after BTH, INA, and tunicamycin treatment. (A) Total BiP protein levels in WT and the bip2 knockout mutant after BTH treatment. BiP protein was detected with a polyclonal antibody (anti-BiP, Santa Cruz Biotechnology) and α-tubulin (TUB) was probed as a loading control. (B) Upper panel: Leaf collapse, marked by arrows, observed overnight after INA treatment. Lower panel: Three-week-old seedlings treated with tunicamycin (0.3 μg ml-1) for 5 days during germination. (C) Induction of UPR marker genes in bip2 and bip2 npr1 (b2n1). Real-time RT-PCRs were performed to examine the relative mRNA levels (Rel. mRNA level) of GRP94, CNX (At5g61790), and PDI (At1g21750) normalized to that of ubiquitin. Error bars represent standard deviations from three PCR results. (D) Rescue of the BTH-induced leaf collapse phenotype in bip2 by npr1.

To determine whether the UPR observed in bip2 was caused by insufficient processing of PR proteins, we introduced the bip2 mutation into the npr1 background, in which BTH-induced PR gene expression is blocked. Both BTH-induced UPR marker gene expression and cell death were diminished in the bip2 npr1 double mutant (Fig. 2, C and D). These genetic data clearly illustrate the detrimental effects that can occur when SAR is induced in plants without a sufficient increase in protein folding and secretion capability.

Because the entire protein secretory pathway is coordinately up-regulated by NPR1 during SAR, the regulatory mechanism may involve common elements. To search for such elements, we focused on the 13 ER-resident genes listed in Table 1 and analyzed their promoter regions (1 kb upstream of the start codon) using the MEME program (24). A consensus sequence, designated TL1 (CTGAAGAAGAA), was overrepresented in the promoter regions of all 13 NPR1-responsive ER-resident genes surveyed (Fig. 3A) (P = 0.02), but was absent from related genes not up-regulated by NPR1, such as the other DAD (At2g35520) and Sec61α (At1g78720).

Fig. 3.

Characterization of the TL1 element. (A) Discovery of a conserved sequence (TL1) in the promoters of all 13 NPR1-regulated ER-resident protein genes (1 kb upstream of the translation start codon) using MEME (24). Blue and gray letters represent highly and moderately conserved nucleotides identified by the program, respectively. Variations from the consensus sequence are marked in red. (B) EMSA using whole-cell protein extracts and 32P-labeled oligonucleotides TL1, m1, m2, and TLm. The arrow indicates the specific DNA-protein band, the single asterisk marks a nonspecific band, and double asterisks indicate the free probes. FP, free probe without protein extract. (C) EMSA of TL1 using whole-cell extracts prepared from WT and npr1-1 with and without SA induction. (D) EMSA of TL1 using nuclear extracts prepared from WT and npr1-1 with and without SA treatment. (E) Mutant analysis of the BiP2 promoter. The three putative TL1 elements found in the BiP2 promoter (Sites 1, 2, and 3) are indicated by small rectangles and their orientations shown by black arrows. The coding sequence is on the right (green arrow). WT and mutant (m1a, m1b, m2a, m2b, m3a, m3b, and m2a3a; the altered nucleotides are underlined) BiP2 promoters were cloned upstream of the GUS coding sequence and transformed into plants. Eight independent T2 lines were pooled for each construct, and the inducibility of the promoter by INA was measured by quantitative GUS assay. Fold changes were determined using GUS activity ratios between induced and uninduced samples. Error bars represent standard deviations of three measurements.

In an electrophoretic mobility shift assay (EMSA), the TL1 element was shown to have a specific protein-binding activity, which was completely abolished with changes in the core sequence (m1, m2, and TLm in Fig. 3B). When whole-cell extracts were used, this binding activity was not affected by SA treatment in either WT or npr1 mutants (Fig. 3C; only npr1-1 is shown). When nuclear extracts were used instead, the specific binding was enhanced in SA-treated samples, suggesting that the TL1-binding protein translocates into the nucleus upon SAR induction (Fig. 3D). Moreover, this translocation was facilitated by NPR1, as indicated by a less profound enhancement of TL1 binding after SA treatment in npr1. Because the induced binding was not completely abolished in npr1, we do not exclude the possibility that NPR1 also controls the activation of the TL1-binding protein. All of these data are consistent with the facts that the induction of secretion-related genes does not require de novo protein synthesis and that NPR1 is also translocated to the nucleus after SAR induction.

To demonstrate the biological activity of TL1, we generated reporter constructs in which the coding sequence of β-glucuronidase (GUS) is driven by either the WT BiP2 promoter or mutant constructs with changes in each of the promoter's three TL1 elements (Fig. 3E; Sites 1, 2, and 3). These constructs were transformed into WT plants, and the effect of each mutation on promoter activity was analyzed through a GUS assay. As expected, the reporter driven by the WT BiP2 promoter showed a 4.8-fold increase in expression after treatment with INA. When transformed into npr1, the WT BiP2 promoter::GUS reporter showed no induction by INA, consistent with the result from the RNA blot analysis (fig. S2) (21). Whereas mutations in Site 1 had little effect on the inducibility of the promoter, mutations in Sites 2 and 3 significantly reduced the induction of the GUS gene. Because Sites 2 and 3 are adjacent (30 base pairs apart) (Fig. 3E), we examined whether they can function cooperatively to confer full induction of BiP2. Indeed, when both sites were mutated, the reporter showed no induction after INA treatment (Fig. 3E, m2a3a).

Taken together, our microarray analysis, EMSA, and promoter mutagenesis data suggest that TL1 is indeed a cis-element involved in SA induction of secretion-related genes via NPR1. The transcription factor that controls TL1 is unlikely a TGA factor because TL1 is distinct from the TGA-binding as-1 element. Furthermore, in a tga2 tga5 tga6 triple mutant, the induction of PR genes is diminished (25), whereas secretion-related genes are still induced (fig. S4). Therefore, we believe that NPR1 regulates secretion-related and PR genes through different transcription factors and cis-elements.

Our finding sheds new light on the induction mechanism of SAR by demonstrating that NPR1 not only directly induces the PR genes but also prepares the cell for secretion of the PR proteins by first making more secretory machinery components. A similar phenomenon is also observed in mammals in which the secretory machinery in B cells is up-regulated before the B cells start secreting antibodies (26). Further study may clarify whether this commonality reflects any conserved regulatory mechanisms.

Supporting Online Material

www.sciencemag.org/cgi/content/full/308/5724/1036/DC1

Materials and Methods

Figs. S1 to S4

Table S1

References and Notes

References and Notes

View Abstract

Navigate This Article