- © 1999 American Society of Plant Physiologists
Abstract
The Arabidopsis NPR1 gene was previously shown to be required for the salicylic acid (SA)– and benzothiadiazole (BTH)-induced expression of pathogenesis-related (PR) genes and systemic acquired resistance. The dominant ssi1 (for suppressor of SA insensitivity) mutation characterized in this study defines a new component of the SA signal transduction pathway that bypasses the requirement of NPR1 for expression of the PR genes and disease resistance. The ssi1 mutation caused PR (PR-1, BGL2 [PR-2], and PR-5) genes to be constitutively expressed and restored resistance to an avirulent strain of Pseudomonas syringae pv tomato in npr1-5 (previously called sai1) mutant plants. In addition, ssi1 plants were small, spontaneously developed hypersensitive response–like lesions, accumulated elevated levels of SA, and constitutively expressed the antimicrobial defensin gene PDF1.2. The phenotypes of the ssi1 mutant are SA dependent. When SA accumulation was prevented in ssi1 npr1-5 plants by expressing the SA-degrading salicylate hydroxylase (nahG) gene, all of the phenotypes associated with the ssi1 mutation were suppressed. However, lesion formation and expression of the PR genes were restored in these plants by the application of BTH. Interestingly, expression of PDF1.2, which previously has been shown to be SA independent but jasmonic acid and ethylene dependent, was also suppressed in ssi1 npr1-5 plants by the nahG gene. Furthermore, exogenous application of BTH restored PDF1.2 expression in these plants. Our results suggest that SSI1 may function as a switch modulating cross-talk between the SA- and jasmonic acid/ethylene–mediated defense signal transduction pathways.
INTRODUCTION
In plants, the outcome of an interaction with a pathogen is governed by multiple factors, including the genotypes of the plant and pathogen as well as the complex exchange of signals between the host and the intruder (Yang et al., 1997). Over years of coevolution with pathogens, plants have evolved complex mechanisms to defend themselves against disease. Whereas some defense responses are constitutive, others are induced upon pathogen attack. One such induced response is systemic acquired resistance (SAR), which is triggered by prior exposure to pathogens that cause cell death (Ross 1961; Kuc, 1982; Ryals et al., 1996). SAR is long lasting and provides the plant with protection against a broad spectrum of pathogens (Dempsey and Klessig, 1995; Hunt and Ryals, 1996). A more rapid defense response that precedes SAR is the hypersensitive response (HR), which occurs at sites of pathogen entry and is characterized by programmed host cell death and restriction of pathogen growth and spread (Matthews, 1991; Hammond-Kosack and Jones, 1996). Increased expression of a subset of the pathogenesis-related (PR) genes, many of which encode proteins possessing antimicrobial activities, is tightly correlated with the development of the HR and SAR. Hence, the induction of these PR genes serves as a good molecular marker for a resistance response (Klessig and Malamy, 1994; Hunt and Ryals, 1996).
Considerable effort has been directed toward identifying signaling molecules responsible for activating the HR and SAR. Salicylic acid (SA) has emerged as a key signaling component in the manifestation of these phenomena (Hammond-Kosack and Jones, 1996; Ryals et al., 1996; Durner et al., 1997). In both tobacco and Arabidopsis, exogenous SA induces the expression of PR (PR-1, PR-2, and PR-5) genes (Antoniw and White, 1980; Ward et al., 1991; Uknes et al., 1992) and resistance (White, 1979; Uknes et al., 1993). Several studies have shown a good correlation between increases in the endogenous levels of SA and its conjugates in infected plants and both the expression of PR genes and the development of disease resistance (Malamy et al., 1990; Métraux et al., 1990; Uknes et al., 1993; Summermatter et al., 1995; Dempsey et al., 1997). Furthermore, when SA accumulation is prevented in tobacco and Arabidopsis plants because of the constitutive expression of an nahG transgene, which encodes the SA-degrading enzyme salicylate hydroxylase, PR gene expression and resistance to several pathogens is compromised (Gaffney et al., 1993; Delaney et al., 1994). Likewise, preventing SA synthesis by specifically inhibiting the activity of phenylalanine ammonia–lyase (PAL), the first enzyme in the SA biosynthetic pathway, makes otherwise resistant Arabidopsis plants susceptible to Peronospora parasitica (Mauch-Mani and Slusarenko, 1996). The ability of two synthetic functional analogs of SA, 2,6-dichloroisonicotinic acid (INA) and benzothiadiazole (BTH) (Vernooij et al., 1995; Görlach et al., 1996; Du and Klessig, 1997; Wendehenne et al., 1998), to restore resistance in SA-depleted NahG tobacco (Friedrich et al., 1996) and Arabidopsis (Lawton et al., 1996) plants further confirms the importance of SA in disease resistance.
To better understand the mechanisms of disease resistance and SAR, several Arabidopsis mutants with altered resistance responses have been identified. They can be broadly classified into two groups. The first group contains mutants that exhibit constitutive SAR, such as acd2 (for accelerated cell death; Greenberg et al., 1994), lsd (for lesion-simulating disease; Dietrich et al., 1994; Weymann et al., 1995), cpr (for constitutive expresser of PR genes; Bowling et al., 1994, 1997; Clarke et al., 1998), and cep (for constitutive expression of PR genes; Klessig et al., 1996). These mutants constitutively accumulate high levels of SA and express the PR genes. They also show enhanced resistance to pathogens. The second group comprises the eds (for enhanced disease susceptibility; Glazebrook et al., 1996; Parker et al., 1996; Rogers and Ausubel, 1997), ndr1 (for non-race-specific disease resistance; Century et al., 1995), and the allelic npr1 (for nonexpresser of PR genes; Cao et al., 1994; Glazebrook et al., 1996), nim1 (for noninducible immunity; Delaney et al., 1995), and sai1 (for salicylic acid–insensitive; Shah et al., 1997) mutants, all of which exhibit compromised resistance to pathogens. The enhanced disease susceptibility phenotype of the allelic npr1, nim1, and sai1 (renamed npr1-5) mutants is due to their inability to respond to SA. These mutants are incapable of expressing the PR (PR-1, BGL2 [PR-2], and PR-5) genes or developing SAR in response to SA, INA, and BTH.
The recessive nature of most of the NPR1 mutant alleles strongly suggests that NPR1 is a positive regulator of the SA signal transduction pathway. The NPR1 gene was recently cloned (Cao et al., 1997; Ryals et al., 1997) and shown to encode a protein containing ankyrin repeat motifs. These ankyrin repeats appear to be important for NPR1 function because they contain missense mutations in three of the npr1 mutants (npr1-1, nim1-2, and npr1-5; Cao et al., 1997; Ryals et al., 1997; H. Cao, J. Shah, D.F. Klessig, and X. Dong, unpublished results). In animals, ankyrin repeats have been implicated in mediating protein–protein interactions, such as those between 53BP2 and the tumor suppressor p53 (Gorina and Pavletich, 1996) and between IκBα and NF-κB (Krappmann et al., 1996). Interestingly, it has been suggested that NPR1 is the plant homolog of IκBα (Ryals et al., 1997), which inhibits the activity of mammalian immune and inflammatory responses by binding to NF-κB. By analogy, NPR1 may regulate the SA-mediated activation of plant defense responses by interacting with another protein(s).
Expression of the PR genes can be induced in response to bacterial and fungal pathogens independently of NPR1. In npr1-2, npr1-3 (Glazebrook et al., 1996), and npr1-5 (Shah et al., 1997) plants, expression of the PR-1, BGL2, and PR-5 genes was induced upon infection with a bacterial pathogen. Likewise, infection of nim1-1 mutant plants with P. parasitica also induced expression of the PR-1 gene (Delaney et al., 1995). Although the induction of PR-1 expression in these mutants was delayed and never reached the maximal levels seen in pathogen-infected wild-type plants (Glazebrook et al., 1996; Shah et al., 1997), the BGL2 and PR-5 genes were induced with kinetics and magnitude similar to those observed in pathogen-infected wild-type plants (Glazebrook et al., 1996).
Even though SA is required for resistance to various pathogens in several plant species, some defense responses appear to be activated by an SA-independent pathway(s). For example, systemic resistance to Fusarium oxysporum as well as Pseudomonas syringae pv tomato can be induced in Arabidopsis, in the absence of SA accumulation and SA-mediated PR gene expression, by initially inoculating the roots with P. fluorescens (Pieterse et al., 1996). Cf-2 and Cf-9 gene–mediated resistance to Cladosporium fulvum species in tomato also seem to be SA independent because fungal growth is restricted equally well in wild-type and nahG transgenic plants (Hammond-Kosack and Jones, 1996). Likewise, the systemic induction in Arabidopsis of cysteine-rich antimicrobial peptides called defensins after Alternaria brassicicola infection appears to be mediated by an SA-independent pathway (Penninckx et al., 1996). Neither the presence of the nahG transgene nor mutations in NPR1 adversely affected the A. brassicicola–mediated induction of the defensin PDF1.2 gene. Systemic induction of PDF1.2 has further been shown to be dependent on jasmonic acid (JA) and ethylene signaling.
Both JA and ethylene have been implicated as important signals during plant defense responses (Yang et al., 1997). Several lines of evidence suggest that there may be cross-talk between the JA, ethylene, and SA signaling pathways. Ethylene has been shown to potentiate the SA-mediated induction of PR-1 in Arabidopsis (Lawton et al., 1995). Likewise, simultaneous application of methyl jasmonate and SA superinduces the expression of the SA-inducible PR-1 gene in tobacco (Xu et al., 1994). The functional SA analog INA has also been shown to elevate JA levels and stimulate the expression of JA-responsive genes in rice (Schweizer et al., 1997) and the thionin gene in barley (Wasternack et al., 1994). Very recently, NPR1 has been shown to be required for the activation of ethylene- and JA-mediated systemic resistance induced by P. fluorescens (Pieterse et al., 1998). Antagonistic effects between these signaling pathways have also been reported. For example, JA biosynthesis and signaling are inhibited by SA and its derivative acetyl SA (aspirin) in tomato (Peña-Cortés et al., 1993; Doares et al., 1995), and ethylene biosynthesis is inhibited by SA in apple (Pennazio et al., 1985; Leslie and Romani, 1988).
To identify other components of the SA signal transduction pathway, we set up a mutant screen for Arabidopsis to identify genetic suppressors of the npr1-5 mutation. The npr1-5–conferred phenotype is due to a C-to-T transition mutation that causes a proline-to-serine change at amino acid 342 in one of the ankyrin repeats of NPR1 (H. Cao, J. Shah, D.F. Klessig, and X. Dong, unpublished results). Here, we report the identification and characterization of the dominant suppressor of SA insensitivity (ssi1) mutation, which confers constitutive expression of the PR genes and restores disease resistance in plants homozygous for the npr1-5 mutant allele. Furthermore, ssi1 plants spontaneously develop HR-like lesions and constitutively express the defensin PDF1.2 gene. All of these ssi1-conferred phenotypes, including the expression of PDF1.2, are dependent on the ability of the mutant to accumulate high levels of SA. BTH application restores all of the ssi1-conferred phenotypes in SA-depleted ssi1 npr1-5 nahG plants, including the expression of the PDF1.2 gene, suggesting that the SSI1 protein may be involved in one of the key steps regulating signaling through the SA-dependent and the JA/ethylene–dependent defense pathways.
RESULTS
The ssi1 Mutation Causes PR Gene Expression and Spontaneous Development of HR-like Lesions in npr1-5 Plants
To isolate suppressors of the SA-insensitive npr1-5 mutation, seeds from npr1-5 plants were mutagenized with ethyl methanesulfonate (EMS), as previously described (Shah et al., 1997). Three- to 4-week-old M2 progeny of these EMS-mutagenized M1 seeds were screened by RNA gel blot analysis for mutants that constitutively accumulated elevated levels of the PR-1 gene transcript. Seven ssi mutants of npr1-5 were identified among the 2400 M2 plants screened. The ssi1 mutant was further characterized. As shown in Figure 1A, unlike the wild-type (SSI1 NPR1) and the parental npr1-5 plants (SSI1 npr1-5), the ssi1 npr1-5 double mutants constitutively accumulated elevated levels of the PR-1, BGL2, and PR-5 transcripts. In addition, the ssi1 npr1-5 double mutants were smaller than was the parental npr1-5 plant (SSI1 npr1-5; Figure 2A), and they developed macroscopic lesions on their leaves (Figure 2B). Trypan blue staining showed a heavy concentration of intensely stained dead cells in these necrotic areas (Figure 2C). Similar patterns and intensity of staining were not observed in wild-type or npr1-5 leaves (data not shown). In addition to cell death and PR gene expression, the accumulation of autofluorescent material at lesion sites is associated with the HR. UV microscopy showed that the necrotic areas on the leaves of ssi1 npr1-5 plants were associated with enhanced levels of autofluorescence (Figure 2D). This result indicates that the spontaneous lesions that develop on the ssi1 mutant plants are HR-like.
Genetic Analysis of ssi1
M3 progeny from the ssi1 npr1-5 double mutant segregated in a ratio of three plants constitutively expressing PR-1 (PR+) to one plant lacking constitutive PR-1 expression (PR−), suggesting that the ssi1 mutant allele is dominant over the wild-type allele. To confirm the dominant nature of the ssi1 mutant allele, we backcrossed the ssi1 npr1-5 double mutant to the parental npr1-5 mutant (wild type for SSI1), which is in the ecotype Nössen background. The resulting F1 and F2 progeny were then monitored for constitutive PR-1 gene expression. All of the F1 plants constitutively expressed the PR-1 gene, and this ssi1-conferred phenotype segregated in a 3 PR+:1 PR− (160 PR+ plants to 61 PR− plants; χ2 = 0.79; 0.5 > P > 0.3) Mendelian ratio in the F2 progeny. This indicates that ssi1 is a dominant mutation at a single genetic locus. The spontaneous lesion formation and reduced size phenotypes cosegregated with constitutive PR-1 expression, suggesting that they are due to a dominant mutation in either the SSI1 gene or a gene(s) tightly linked to the SSI1 locus. Approximately one-third of the plants exhibiting constitutive PR-1 expression and lesions were very small in size (ssi1 npr1-5; Figure 2A) compared with the wild-type SSI1 plants. The other two-thirds of the constitutive PR-1–expressing, lesion-bearing plants were intermediate in size (ssi1[het] npr1-5), suggesting that the reduced size phenotype of the ssi1 mutant is dependent on the dose of the mutant ssi1 allele. Analyses of F3 progeny confirmed that the very small F2 plants and the intermediate-sized F2 plants were homozygous and heterozygous for the ssi1 mutant allele, respectively.
A second site mutation within the npr1-5 allele could potentially suppress the npr1-5–conferred phenotype. If so, this intragenic suppressor mutation should cosegregate with the npr1-5 allele. To determine whether the ssi1 mutation is an intragenic suppressor of npr1-5, the ssi1 npr1-5 double mutant was crossed to wild-type (SSI1 NPR1) plants of ecotype Nössen, and F2 progeny were analyzed for constitutive PR-1 expression. This ssi1-conferred phenotype segregated in a 3 PR+:1 PR− ratio (33 PR+ plants to 11 PR− plants). Spontaneous lesion formation cosegregated with the PR+ phenotype in these plants. Using codominant cleaved amplified polymorphic sequence (CAPS) analysis, we determined the genotype of these 33 phenotypically ssi1 plants at the NPR1 locus. The mutant npr1-5 allele can be detected by the absence of an NlaIV restriction site that is present in the wild-type NPR1 allele. One-quarter of the F2 plants with the ssi1 phenotype (eight of 33) were homozygous for the NPR1 wild-type allele. Analysis of PR-1 expression in the F3 progeny of these ssi1 NPR1 plants (Figure 1B) confirmed that the ssi1-conferred phenotype is unlinked to the npr1-5 allele and therefore is not an intragenic suppressor of npr1-5. Furthermore, the ssi1-conferred phenotype is not dependent on the npr1-5 mutation.
PR and PDF1.2 Expression in ssi1.
(A) Expression of PR-1, BGL2, PR-5, and PDF1.2 genes in water-treated or SA-treated wild-type (SSI1 NPR1) and npr1-5 (SSI1 npr1-5) plants and in untreated heterozygous (ssi1[het] npr1-5) and homozygous (ssi1 npr1-5) ssi1 plants. The heterozygous as well as homozygous ssi1 plants analyzed were all homozygous for the npr1-5 mutant allele. RNA was extracted from leaves of untreated ssi1(het) npr1-5 and ssi1 npr1-5 plants and, as controls, from water (W)- or SA-treated (SA) SSI1 NPR1 and SSI1 npr1-5 plants 24 hr after treatment.
(B) A comparison of the expression of PR-1 and PDF1.2 genes in ssi1 npr1-5 double mutants and ssi1 NPR1 plants homozygous (ssi1 npr1-5 and ssi1 NPR1) or heterozygous (ssi1[het] npr1-5 and ssi1[het] NPR1) for the ssi1 mutant allele.
(C) PR-1 and PDF1.2 expression in npr1-5 and nim1-3 mutants homozygous for the wild-type SSI1 (SSI1 npr1-5 and SSI1 nim1-3) or the ssi1 mutant allele (ssi1 npr1-5 and ssi1 nim1-3). Two plants of each genotype were investigated.
The blots were sequentially probed for the indicated genes and rRNA as an internal control for gel loading and transfer. Plants were grown in soil and were 3 weeks old when sampled.
To determine whether the ssi1-conferred phenotype requires NPR1, we analyzed the ssi1-conferred phenotype in the nim1-3 (allelic with npr1) mutant background. The nim1-3 allele contains a single base pair deletion causing a frame-shift at amino acid 172, thus encoding a truncated protein lacking the C-terminal two-thirds of NPR1 (Ryals et al., 1997). nim1-3 plants are insensitive to SA and are defective in activating SAR. The ssi1 npr1-5 double mutant was crossed with a SSI1 nim1-3 plant. F2 plants homozygous for the nim1-3 allele were identified using CAPS analysis, and the expression of the PR-1 gene was analyzed in these plants. Figure 1C shows the expression of PR-1 in two of these F2 segregants. Three-fourths of the plants homozygous for the nim1-3 allele constitutively expressed the PR-1 gene and spontaneously developed lesions, thus strongly arguing that the ssi1 mutant phenotypes do not require NPR1.
The ssi1 Mutant Constitutively Accumulates High Levels of SA and SAG
Several studies have demonstrated the presence of an SA-dependent potentiation and feedback amplification loop in the expression of defense genes and the development of HR-like lesions (Weymann et al., 1995; Fauth et al., 1996; Mur et al., 1996; Shirasu et al., 1997; Thulke and Conrath, 1998). Therefore, we analyzed the endogenous levels of SA and its glucoside (SAG) in npr1-5 plants that were either homozygous or heterozygous for the ssi1 mutant allele. As shown in Figure 3, SA (7.3 ± 0.4 μg per gram fresh weight of tissue) and SAG (80.0 ± 4.2 μg per gram fresh weight of tissue) levels in plants homozygous for the ssi1 mutant allele (ssi1 npr1-5) were ~20- and 200-fold higher, respectively, than in the parental npr1-5 plants (SSI1 npr1-5). In comparison, plants heterozygous for the ssi1 mutant allele (ssi1[het] npr1-5) accumulated slightly less SA (6.3 ± 1.3 μg per gram fresh weight of tissue) and approximately twofold less SAG (34.4 ± 4.9 μg per gram fresh weight of tissue) than did the homozygous ssi1 plants.
Morphological Phenotypes of ssi1.
(A) A comparison of the dominant small-size phenotype of ssi1 npr1-5 double mutants heterozygous (ssi1[het] npr1-5) or homozygous (ssi1 npr1-5) for the ssi1 mutant allele with an npr1-5 mutant (SSI1 npr1-5) plant.
(B) A comparison of leaves from SSI1 npr1-5 double mutants, heterozygous (ssi1[het] npr1-5) or homozygous (ssi1 npr1-5) for the ssi1 mutant allele, showing chlorosis and spontaneous lesions with a leaf from an npr1-5 mutant (SSI1 npr1-5) plant.
(C) Microscopy of a trypan blue–stained leaf containing lesions from an ssi1 npr1-5 plant showing an intensely stained area of dead cells.
(D) UV microscopy of a leaf from an ssi1 npr1-5 plant showing increased autofluorescence, above background, at the site of a lesion.
All plants were grown in soil and photographed when 3 weeks old.
Comparison of SA and SAG Levels in the Wild Type (SSI1 NPR1), the npr1-5 Mutant (SSI1 npr1-5), and ssi1 npr1-5 Double Mutants Heterozygous (ssi1[het] npr1-5) or Homozygous (ssi1 npr1-5) for the ssi1 Mutant Allele.
Leaves from 3-week-old soil-grown plants were harvested, extracted, and analyzed by HPLC, as described by Bowling et al. (1994). The SA and SAG values ±sd, presented as micrograms of SA per gram fresh weight (FW) of tissue, are averages of three to five sets of samples per line.
Expression of an nahG Transgene Suppresses Constitutive PR Gene Expression and Spontaneous Lesion Formation in ssi1 Plants
To determine whether high endogenous levels of SA and SAG are required for the mutant phenotypes exhibited by the ssi1 plants, ssi1 npr1-5 double mutants were crossed with NahG (ecotype Nössen) plants, which are unable to accumulate elevated levels of SA. The resulting F1 plants (ssi1 × nahG) did not constitutively express the PR-1, BGL2, or PR-5 genes (Figure 4). Furthermore, they were morphologically similar to wild-type plants and did not develop lesions or possess the reduced-size phenotype associated with the dominant ssi1 mutant allele. These results were further confirmed in the F2 progeny (data not shown). Thus, the elevated levels of endogenous SA detected in the ssi1 mutant appear to be required for all of the ssi1-conferred phenotypes.
BTH Restores PR Gene Expression and Spontaneous Lesion Formation in SA-Depleted ssi1 npr1-5 nahG Plants
BTH, a functional analog of SA (Görlach et al., 1996; Lawton et al., 1996; Du and Klessig, 1997; Wendehenne et al., 1998), induces PR gene expression and disease resistance in both wild-type and nahG-expressing transgenic Arabidopsis plants (Lawton et al., 1996). However, BTH requires a functional SA signaling pathway because it is unable to induce PR expression in npr1-5 plants (Shah et al., 1997). Because the ssi1 mutation restores SA-mediated expression of the PR genes in npr1-5 plants, we tested whether it could also restore BTH-induced PR gene expression in npr1-5 plants. ssi1 npr1-5 nahG plants (homozygous for the npr1-5 allele) were used for this experiment because they fail to accumulate elevated levels of SA and do not possess any of the phenotypes associated with the ssi1 mutation. BTH treatment was found to restore lesion formation in the existing and newly emerging leaves of these plants (Figures 5A and 5B). It also induced to high levels the expression of PR-1 (Figure 6), BGL2, and PR-5 (data not shown). In comparison, BTH was unable to induce high levels of PR gene expression or lesion formation in npr1-5 nahG plants homozygous for the wild-type SSI1 allele (SSI1 npr1-5 nahG). Trypan blue staining confirmed the absence of lesions in these plants after BTH treatment (Figure 5B). These results argue that the mere accumulation of high levels of SA may not necessarily cause the ssi1-conferred phenotypes. Rather, the SSI1 gene functions as a component of the SA signal transduction pathway, and the ssi1 mutation bypasses the requirement of NPR1 function for expression of the PR genes.
ssi1 Restores Resistance to an Avirulent Bacterial Pathogen in npr1-5 Plants
Because the ssi1 mutation restores SA- and BTH-inducible PR gene expression in npr1-5 plants, we tested whether it would also restore disease resistance. Wild-type Nössen plants contain the resistance gene RPS2 (Bent et al., 1994) and are resistant to P. s. tomato DC3000 carrying the avrRpt2 avirulence gene. In contrast, npr1-5 plants show enhanced susceptibility to this pathogen (Shah et al., 1997). Because plants homozygous for the ssi1 mutant allele are very small and difficult to infiltrate with P. s. tomato, we chose to infect plants heterozygous for the ssi1 mutant allele. As shown in Figure 7A, the growth of P. s. tomato was 14-fold lower in ssi1(het) npr1-5 double mutants than in SSI1 npr1-5 plants at 3 days postinfiltration (DPI). The presence of the wild-type NPR1 gene led to an additional two-fold decrease in bacterial growth in ssi1(het) NPR1 plants as compared with ssi1(het) npr1-5 double mutant plants.
Expression of PR Genes in ssi1 × nahG F1 Plants.
RNA was extracted from SSI1 × nahG and ssi1 × nahG F1 plants and a plant heterozygous for the ssi1 mutant allele (ssi1[het]). These F1 plants contain a wild-type NPR1 allele derived from the wild-type or NahG parents. As a control, RNA was also extracted from an untreated wild-type plant (SSI1) and a wild-type plant 24 hr after SA (500 μM) treatment (SSI1 + SA). All plants were grown in soil and sampled when 3 weeks old. The blot was sequentially probed for the Arabidopsis PR-1, BGL2, and PR-5 gene transcripts and the nahG transgene transcript (NahG); rRNA was used as an internal control for gel loading and transfer.
The accumulation of PR-1 gene transcript was also monitored in these plants after P. s. tomato infection (Figure 7B). Unlike the wild-type (SSI1 NPR1) and the SSI1 npr1-5 plants, the PR-1 gene was constitutively expressed in the uninfected ssi1(het) npr1-5 and ssi1(het) NPR1 plants. Furthermore, as in the wild-type plants, PR-1 expression increased after P. s. tomato infection and reached maximal levels by 1 DPI in the ssi1(het) npr1-5 and ssi1(het) NPR1 plants. In comparison, as previously demonstrated in SSI1 npr1-5 plants (Shah et al., 1997), PR-1 expression was delayed and never attained the maximal levels seen in the P. s. tomato–infected wild-type or the ssi1 plants.
Constitutive Defensin Gene Expression in the ssi1 Mutant Is SA Dependent
Expression of the PDF1.2 gene, which encodes defensin, an antifungal peptide, has previously been shown to be independent of both SA and NPR1 (Penninckx et al., 1996). This gene is constitutively expressed in an SA- and NPR1-independent manner in the Arabidopsis cpr5 and cpr6-1 mutants that, like ssi1, constitutively express the PR genes and exhibit SAR (Bowling et al., 1997; Clarke et al., 1998). Therefore, we examined the accumulation of PDF1.2 transcripts in ssi1 plants. As shown in Figures 1A and 1B, plants homozygous for both the ssi1 and npr1-5 alleles (ssi1 npr1-5) constitutively accumulated elevated levels of PDF1.2 transcripts. Similarly, plants homozygous for the ssi1 and nim1-3 alleles also constitutively expressed the PDF1.2 gene (Figure 1C). Interestingly, the expression of PDF1.2, unlike that of the PR genes, was observed to be higher in plants heterozygous for the ssi1 allele (ssi1[het] npr1-5) compared with those homozygous for the ssi1 allele (ssi1 npr1-5; Figure 1A). Constitutive PDF1.2 expression was also repeatedly observed to be higher in ssi1 plants homozygous for the npr1-5 mutant allele compared with ssi1 plants homozygous for the wild-type NPR1 allele (Figures 1B and 7B). Unlike expression of the PR-1 gene, PDF1.2 expression was not induced by P. s. tomato infection in either the wild-type or the ssi1 plants (Figure 7B). Instead, steady state levels of the PDF1.2 transcript in the ssi1 mutants transiently decreased 1 and 2 DPI, returning to the basal levels seen in uninfected plants by 3 DPI. In contrast, in two of four experiments, infection with P. s. tomato was found to induce PDF1.2 expression in the SSI1 npr1-5 mutant.
To determine whether the constitutive expression of PDF1.2 observed in ssi1 plants was associated with elevated SA levels, we analyzed its expression in ssi1 npr1-5 nahG plants. Expression of the nahG transgene was observed to suppress PDF1.2 expression. BTH treatment of these nahG-expressing ssi1 npr1-5 mutants, however, restored PDF1.2 expression (Figure 6). In contrast, BTH treatment did not induce PDF1.2 expression in nahG plants homozygous for the wild-type SSI1 and mutant npr1-5 alleles (SSI1 npr1-5 nahG).
Comparison of Lesion Formation in Untreated and BTH-Treated ssi1 npr1-5 nahG and SSI1 npr1-5 nahG Plants.
(A) Leaves from an untreated and BTH-treated ssi1 npr1-5 nahG plant showing absence of macroscopic lesions in untreated plants and restoration of lesions when BTH was applied.
(B) Trypan blue–stained, BTH-treated leaves of an ssi1 npr1-5 nahG plant showing intensely stained areas of dead cells. As a control, leaves from a BTH-treated SSI1 npr1-5 nahG plant are shown; they do not have areas of intensely stained dead cells.
Three-week-old plants were treated with 100 μM BTH, and leaf tissue from untreated and BTH-treated plants was analyzed 6 days later.
ssi1 Maps to Chromosome 4 and Defines a Novel Gene
To determine the map position of the SSI1 locus, we crossed ssi1 plants (ecotype Nössen) with wild-type plants from ecotype Columbia. As expected, the F2 progeny segre-gated in a 3 PR+:1 PR− ratio (88 PR+ plants to 28 PR− plants) when scored for constitutive PR-1 expression. The genotype at the SSI1 locus was determined for these 116 F2 plants by monitoring constitutive PR-1 expression in the F3 families. CAPS (Konieczny and Ausubel, 1993) and simple sequence length polymorphism (SSLP; Bell and Ecker, 1994) marker analyses were subsequently performed on 24 PR+ (homozygous ssi1) F2 plants. Based on these preliminary analyses, ssi1 was mapped to chromosome 4 in the 28-centimorgan (cM) interval between the SSLP marker nga8 and the CAPS marker SC5.
Because the ssi1 mutant allele is dominant and homozygous ssi1 plants are small and poor seed producers, the map position of the SSI1 locus was further determined using 182 phenotypically wild-type (PR−) F2 plants. Based on this analysis, the SSI1 gene was mapped within a 16.1-cM interval, 9.8 cM from the SSLP marker AthDET1 and 6.3 cM from the CAPS marker SC5. The CPR1, LSD1, ACD2, and CEP1 genes, which have recessive mutations and confer constitutive expression of PR genes, also map to chromosome 4 (Bowling et al., 1994; Dietrich et al., 1994; Greenberg et al., 1994; Klessig et al., 1996) but distal to ssi1. CPR1 is the closest to SSI1; however, the cpr1 mutation is recessive, and unlike the dominant ssi1 mutation, cpr1 plants do not constitutively express the PDF1.2 gene or spontaneously develop lesions. From these results, we conclude that cpr1 and ssi1 are most likely mutations in two distinct genes.
DISCUSSION
To identify components of the SA-mediated defense signaling pathway, we isolated and characterized a suppressor of the npr1-5 mutant designated ssi1. The dominant ssi1 mutation suppresses all of the known phenotypes of the npr1-5 mutant, including lack of SA-mediated expression of the PR-1, BGL2, and PR-5 genes and reduced disease resistance. In addition, ssi1 npr1-5 plants constitutively express these PR genes and the JA- and ethylene-responsive PDF1.2 gene, spontaneously develop HR-like lesions, are smaller in size than either the wild-type or parental npr1-5 plants, and accumulate elevated levels of SA and SAG.
The ssi1-conferred phenotypes appear to be dependent on high levels of SA because they are suppressed in SA-depleted ssi1 npr1-5 plants expressing the nahG gene. Moreover, the ability of BTH to restore the phenotype conferred by ssi1 in these plants suggests that SSI1 is a component of the SA-mediated signaling pathway leading to defense responses. Like the dominant ssi1 mutant, the recessive cpr5 (Bowling et al., 1997) and the dominant cpr6-1 (Clarke et al., 1998) mutants, which map to chromosomes 5 and 1, respectively, also constitutively express the PR and PDF1.2 genes, accumulate elevated levels of SA, show enhanced resistance to bacterial pathogens, and are small in size. In addition, like ssi1, cpr5 plants spontaneously develop HR-like lesions. The constitutive expression of PR genes in ssi1, cpr5, and cpr6-1 is dependent on SA. However, whereas PR gene expression in cpr5 requires NPR1 function, it is independent of NPR1 in the ssi1 and cpr6-1 mutants. Moreover, in contrast to the ability of ssi1 to restore resistance against bacterial pathogens in npr1-5 plants, resistance to a bacterial pathogen in cpr5 and cpr6-1 was dependent on NPR1 function. Finally, unlike ssi1, constitutive PDF1.2 expression in cpr5 and cpr6-1 is independent of SA.
The NPR1 protein has previously been shown to be required for SA-mediated expression of PR genes and the establishment of SAR in Arabidopsis (Cao et al., 1994; Delaney et al., 1995; Glazebrook et al., 1996; Shah et al., 1997). These studies have demonstrated that NPR1 functions downstream of SA. Because ssi1 plants accumulate elevated levels of SA and require SA for constitutive PR expression, it is possible that SSI1 functions upstream of NPR1 in the SA-signaling pathway. This scenario would require npr1-5 to be a leaky mutation. However, the ssi1 mutation can confer constitutive PR expression in the nim1-3 (allelic with npr1) mutant background (Figure 1C), which potentially expresses a protein lacking the C-terminal two-thirds of NPR1 (Ryals et al., 1997). This confirms that the ssi1-conferred phenotype does not depend on NPR1 function; hence, SSI1 does not function upstream of NPR1.
If SSI1 is not upstream of NPR1, where does it function in the defense signaling pathway? NPR1 contains ankyrin repeats, which are involved in protein–protein interactions in animals. These repeats appear to be important for NPR1 function because three mutant alleles of the NPR1 gene contain missense mutations in the ankyrin repeat region (Cao et al., 1997; Ryals et al., 1997; H. Cao, J. Shah, D.F. Klessig, and X. Dong, unpublished results). Thus, it is possible that SSI1 physically interacts with NPR1 to propagate the SA signal. The mutation in npr1-5 could adversely affect this interaction, thereby preventing the SA signal from being transmitted further. In the ssi1 mutant, a compensatory mutation in the SSI1 protein could allow partial interaction between the mutant SSI1 and NPR1 proteins. This would then restore the SA signaling pathway leading to expression of the PR genes and disease resistance. However, the inability of the nim1-3 mutant to suppress the ssi1-mediated constitutive PR expression phenotype strongly argues against SSI1 functioning as an NPR1-interacting protein.
A more plausible hypothesis is that SSI1 might either directly or indirectly influence the SA signaling pathway downstream of NPR1 (Figure 8A). SSI1 could function either downstream of NPR1 or produce a signal that, in conjunction with an NPR1-derived signal, activates PR gene expression and disease resistance. A mutation in the SSI1 protein might allow the requirement for the NPR1-derived signal to be bypassed while still requiring SA. SA either would be required to activate the SSI1 protein or may be required in any one of the downstream steps leading to expression of the PR genes and disease resistance. The recent observation that PR gene expression is not constitutively activated in Arabidopsis overexpressing NPR1 (Cao et al., 1998) can also be explained by this hypothesis. Although these plants overexpress NPR1, they lack the SSI1-activated signal, which is also required for PR gene induction. Alternatively, SSI1 might function completely independently of NPR1 and might be a component of an SA-dependent but NPR1-independent pathway (Figure 8B) involved in the expression of PR genes and resistance. Previous studies of various npr1 mutants have suggested the existence of such an NPR1-independent resistance pathway (Glazebrook et al., 1996; Shah et al., 1997). Expression of PR genes was activated in npr1 plants upon pathogen infection. Because SA by itself cannot activate PR expression in SSI1 npr1 plants, this scenario would require SA plus another pathogen-induced signal for the activation of SSI1-dependent PR gene expression. A mutation in the SSI1 protein could bypass the requirement for this unknown pathogen-induced signal, although SA would still be required.
Expression of PR-1 and PDF1.2 Genes in BTH-Treated ssi1 npr1-5 nahG Plants.
RNA was extracted 6 days after BTH treatment (+; 100 μM) from leaves of 3-week-old ssi1 npr1-5 nahG and SSI1 npr1-5 nahG plants. As controls, RNA was also extracted from untreated (−) ssi1 npr1-5 nahG, SSI1 npr1-5 nahG, and ssi1 npr1-5 plants. The blot was sequentially probed for the Arabidopsis PR-1 and PDF1.2 gene transcripts, the nahG transgene transcript (NahG), and rRNA as an internal control for gel loading and transfer.
Growth of P. s. tomato DC3000 Containing the avrRpt2 Avirulence Gene and PR-1 and PDF1.2 Expression in Wild-Type (SSI1 NPR1), npr1-5 Mutant (SSI1 npr1-5), ssi1 Mutant (ssi1[het] NPR1), and ssi1 npr1-5 Double Mutant (ssi1[het] npr1-5) Plants.
(A) P. s. tomato DC3000 containing the avrRpt2 avirulence gene (OD600 nm = 0.001 in 10 mM MgCl2) was infiltrated into the abaxial surface of leaves with a syringe. Four leaf discs were harvested 3 DPI from the P. s. tomato–infected leaves and ground in 10 mM MgCl2, and bacterial numbers were titered. The bacterial numbers ±sd, presented as colony-forming units (cfu) per leaf disc (0.2 cm2), are averages of three samples. The ssi1 NPR1 (ssi1[het] NPR1) and ssi1 npr1-5 double mutant (ssi1[het] npr1-5) plants were heterozygous for the ssi1 mutant allele.
(B) Expression of the PR-1 and PDF1.2 genes in P. s. tomato–infected wild-type (SSI1 NPR1), npr1-5 mutant (SSI1 npr1-5), ssi1 mutant (ssi1[het] NPR1), and ssi1 npr1-5 double mutant (ssi1[het] npr1-5) plants. P. s. tomato–infected leaf samples were taken at the indicated times, and RNA was extracted. The blot was sequentially probed for Arabidopsis PR-1 and PDF1.2, and rRNA was used as an internal control for gel loading and transfer.
Irrespective of where SSI1 acts in the SA signal transduction pathway, based on the dominance of the ssi1 mutant allele over the wild-type SSI1 allele, the ssi1-conferred phenotype could be due either to a gain-of-function mutation in a positive regulator of the SA signal transduction pathway or, alternatively, to a dominant loss-of-function mutation in a negative regulator. The available data do not allow us to classify SSI1 as an activator or a repressor of the SA signaling pathway.
Even though SSI1 does not function upstream of NPR1, the ssi1 mutant accumulates elevated levels of SA. In addition, SA is required to activate the ssi1-conferred phenotype in the SA-insensitive npr1-5 background. This implies that SSI1-activated components of the defense pathway might subsequently regulate SA accumulation through a feedback loop. A feedback or SA-dependent potentiation loop has previously been identified in the pathways leading to several defense responses including cell death and the expression of PAL (Weymann et al., 1995; Fauth et al., 1996; Mur et al., 1996; Shirasu et al., 1997; Thulke and Conrath, 1998). PAL activity is required for SA biosynthesis in Arabidopsis (Mauch-Mani and Slusarenko, 1996), and induction of PAL expression by pathogen infection or elicitor treatment is potentiated by SA (Shirasu et al., 1997; Thulke and Conrath, 1998). In the ssi1 mutant, the pathogen-inducible PAL1 gene is constitutively expressed at elevated levels (J. Shah and D.F. Klessig, unpublished data). An SSI1-activated defense component(s) in the ssi1 plant could activate expression of the PAL1 gene, which in turn could lead to the increased production of SA. SA in turn would then activate SSI1-dependent signaling leading to expression of PR genes and disease resistance.
ssi1 plants spontaneously develop HR-like lesions. Lesion formation in ssi1 is dependent on the accumulation of elevated levels of SA. However, cell death in ssi1 plants is not a direct result of toxicity due to the enormously high levels of SA that accumulate in the mutant but rather is a result of the activation of an SSI1-dependent cell death pathway. This is evident from the fact that BTH application at concentrations that are known not to induce cell death in wild-type SSI1 plants activates cell death in SA-depleted ssi1 npr1-5 nahG plants but not in SSI1 npr1-5 nahG plants (Figures 5A and 5B).
HR-like cell death in several cases has been shown to be mediated through reactive oxygen species (ROS), and SA has been shown to potentiate pathogen-induced as well as ROS-induced cell death (Levine et al., 1994; Dangl et al., 1996). In the lsd1 mutant, cell death was shown to be dependent on the accumulation of elevated levels of superoxide, and SA or its functional analog INA could induce cell death in lsd1 under nonpermissive conditions (Dangl et al., 1996; Jabs et al., 1996). Whether activation of cell death in ssi1 is a result of the production of ROS needs to be investigated. Although PR and PDF1.2 gene expression and cell death in the ssi1 mutant are tightly correlated, a cause and effect relationship between cell death and PR and PDF1.2 expression is at present unclear. However, in several other cases, PR gene expression, SA accumulation, and resistance can occur independently of cell death. The Arabidopsis cpr1 (Bowling et al., 1994), cpr6-1 (Clarke et al., 1998), and dnd1 (Yu et al., 1998) mutants constitutively accumulate elevated levels of SA and express PR genes without any evident spontaneous cell death. cpr6-1 plants, like the ssi1 mutant plants, also constitutively express the PDF1.2 gene. Furthermore, the dnd1 mutant also demonstrates gene-for-gene disease resistance in the absence of HR-associated cell death (Yu et al., 1998).
A. brassicicola–induced expression of the PDF1.2 gene in Arabidopsis has previously been shown to occur independent of both SA and NPR1 (Penninckx et al., 1996). However, based on our analyses of ssi1 npr1-5 nahG plants, SA is required for constitutive PDF1.2 expression in ssi1 mutant plants. Possibly, SA is required in these plants to activate the mutant SSI1 protein or a downstream component and thus PDF1.2 expression. Supporting this hypothesis, BTH treatment restores high levels of PDF1.2 expression in ssi1 npr1-5 nahG plants. Interestingly, our studies of the ssi1 mutant suggest that NPR1 negatively regulates PDF1.2 expression. PDF1.2 expression was repeatedly observed to be higher in ssi1 npr1-5 plants compared with ssi1 NPR1 plants (Figures 1B and 7B). Similarly, Bowling et al. (1997) have observed elevated steady state levels of PDF1.2 transcript in the npr1-1 mutant grown on agar plates compared with wild-type plants.
Possible Relationships among SA and the NPR1, SSI1, PR, and PDF1.2 Genes.
(A) SSI1 or an SSI1-generated signal enters the SA signaling pathway downstream of NPR1. The mutant SSI1 protein bypasses the requirement of the NPR1-generated signal for activation of the signaling pathway.
(B) SSI1 activates expression of the PR genes and resistance via an SA-dependent but NPR1-independent pathway. A second pathogen-activated signal is required for activation of SSI1, in addition to SA, to account for the inability of exogenous SA to induce expression of PR genes in npr1 mutant plants. The mutant SSI1 protein bypasses the requirement for this pathogen-derived signal, although it still requires SA.
In both (A) and (B), SSI1 is shown to regulate the accumulation of SA via an autoregulatory loop (dashed arrows). SSI1-mediated cell death and SA accumulation have been shown to be independent of each other in light of the accumulating genetic evidence that cell death may not be required for SA accumulation. Cell death is also shown to be independent of the NPR1 gene because the HR develops normally in the npr1 mutant when infected with avirulent bacterial pathogens (Cao et al., 1994; J. Shah and D.F. Klessig, unpublished results). Pathogen-induced expression of PDF1.2 is dependent on JA and ethylene signaling. At the present time, it is not known whether SSI1-induced expression of PDF1.2 in the ssi1 mutant bypasses the requirement of JA and ethylene. However, inhibition of JA and ethylene biosynthesis by SA could partly explain the inability of exogenous SA to induce PDF1.2 expression in wild-type plants.
If SSI1 is involved in the expression of both the PR genes and PDF1.2, why does the application of SA or BTH fail to induce PDF1.2 expression in plants homozygous for the wild-type SSI1 gene? One possible explanation is that a second signal, in addition to SSI1, is required for PDF1.2 expression in SA/BTH–treated wild-type plants. In ssi1 plants, the mutant SSI1 protein might be able to bypass the need for this second signal. An alternative, although not mutually exclusive, explanation is that SA, which is required for PR expression, inhibits the synthesis of JA (Peña-Cortés et al., 1993) and ethylene (Pennazio et al., 1985; Leslie and Romani, 1988). Both JA and ethylene, along with their corresponding signal transduction pathways, are known to be required for pathogen-induced activation of the PDF1.2 gene. Supporting this possibility, defensin accumulation after A. brassicicola infection was shown to be higher in nahG transgenic Arabidopsis plants than in nontransgenic plants (Penninckx et al., 1996). Similarly, PDF1.2 expression was repeatedly observed to be higher in heterozygous ssi1 plants (Figure 1), which accumulate twofold lower levels of total SA, than in the homozygous ssi1 plants (Figure 3). The ssi1 mutant might constitutively accumulate high levels of JA and/or ethylene as well as SA. Elevated levels of JA and ethylene would lead to activation of PDF1.2 expression in ssi1 plants. In support of this possibility, the JA-responsive thionin (THI2-1), lipoxygenase (LOX2), and vegetative storage protein (VSP) genes and the ethylene-responsive basic PR-3 gene are constitutively expressed at elevated levels in the ssi1 mutant (J. Shah and D.F. Klessig, unpublished results). Interestingly, the acd2 mutant of Arabidopsis, which like ssi1 plants spontaneously develops lesions, accumulates high levels of SA, and constitutively expresses both PR and PDF1.2 genes (Greenberg et al., 1994; Penninckx et al., 1996), also accumulates ninefold higher levels of JA than do wild-type plants (Penninckx et al., 1996).
An interesting outcome of our study with the ssi1 mutant is the finding that the SA-responsive and the JA- and ethylene-responsive defense pathways do not appear to function completely independently of each other. Rather, each might regulate the temporal expression and/or amplitude of the other pathway. The SSI1 gene might encode an important switch that, depending on the input signal(s), may differentially regulate these pathways. Several lines of evidence support the existence of cross-talk between the SA-mediated pathway and the JA- and ethylene-dependent, wounding-responsive pathway(s). For example, whereas many wounding responses are mediated by JA and ethylene, overexpression of the rice Ras-like G protein gene rgp1 renders SA accumulation and PR gene expression wounding responsive in tobacco (Sano et al., 1994). Similarly, overexpression of the tobacco WIPK (for wounding-induced protein kinase) mitogen-activated protein (MAP) kinase gene leads to elevated levels of SA and PR gene expression upon wounding (Seo et al., 1995). In addition, the tobacco MAP kinase SIPK (for SA-induced protein kinase) is rapidly and transiently activated by wounding as well as SA (Zhang and Klessig, 1998), further suggesting the existence of shared components between these different defense pathways. Finally, NPR1 has recently been shown to be required for the JA- and ethylene-mediated activation of systemic resistance induced by P. fluorescens (Pieterse et al., 1998). Thus, cloning the SSI1 gene and identifying suppressors of the ssi1 mutation should help elucidate not only the signaling components associated with these different defense pathways but also the mechanisms through which they interact.
METHODS
Growth Conditions for Plants and Bacteria
Plants (Arabidopsis thaliana) were grown in soil at 22°C in growth chambers programmed for a 16-hr-light (8000 to 10,000 lux) and 8-hr-dark cycle, unless otherwise stated. Pseudomonas syringae pv tomato DC3000 carrying a plasmid-borne avirulence avrRpt2 gene (Bent et al., 1994) was propagated at 30°C on King's B medium (King et al., 1954) containing rifampicin (100 μg/mL) and kanamycin (25 μg/mL).
Bacterial Infection of Plants
Infection of plants with P. s. tomato DC3000 carrying a plasmid-borne avrRpt2 gene (Bent et al., 1994) was performed as described previously (Shah et al., 1997). Four leaves per plant were infiltrated with a suspension (OD600 nm of 0.001) in 10 mM MgCl2. Twelve leaf discs, 0.5 cm in diameter (0.20 cm2), were harvested at the indicated times and processed for bacterial counts and RNA extraction, as described previously (Shah et al., 1997). The average bacterial count in the leaves immediately after infection was 2000 colony-forming units per leaf disc.
Chemical Treatment of Plants
Three-week-old plants were sprayed and subirrigated with a solution of salicylic acid (SA; 500 μM) or benzothiadiazole (BTH; 100 μM active ingredient) in water, as previously described (Shah et al., 1997). Wherever possible, another set of control plants was similarly treated with water. Leaves were harvested at the indicated times after treatment and quick frozen in liquid nitrogen. Leaf samples were stored at −80°C. For analysis of individual plants, two fully expanded leaves were harvested before any chemical treatment. This sample served as the untreated control.
RNA Extraction and Gel and Dot Blot Analyses
Large-scale preparation of RNA from Arabidopsis was conducted according to Das et al. (1990). Small-scale extraction of RNA from one or two leaves was performed in the TRIzol reagent (Gibco BRL, Gaithersburg, MD), following the manufacturer's instructions. RNA gel blot analysis and synthesis of random primed probes for PR-1, BGL2, PR-5, and rDNA were done as described previously (Shah et al., 1997). Probes specific for the nahG and PDF1.2 gene transcripts were synthesized by random primed 32P-labeling of gel-purified DNA fragments containing the nahG insert and a 400-bp polymerase chain reaction (PCR) fragment of PDF1.2 (Penninckx et al., 1996). RNA gel blot hybridization and dot blot analysis were performed according to Sambrook et al. (1989).
Histochemistry and Microscopy
Leaf samples for trypan blue staining and epifluorescence microscopy were obtained from 3-week-old plants. Samples were processed and analyzed as described by Bowling et al. (1997).
SA and SA Glucoside Estimations
SA and SA glucoside (SAG) were extracted and estimated from 0.25 to 0.5 g of fresh weight leaf tissue, as described by Bowling et al. (1994).
Mutagenesis and Selection of ssi Mutants
Five thousand seeds from plants homozygous for the npr1-5 mutant allele (ecotype Nössen) were mutagenized with 0.3% ethyl methanesulfonate (EMS; Sigma), as previously described (Shah et al., 1997). M2 seeds were harvested as pools; each pool contained M2 seeds derived from ~10 EMS-mutagenized M1 seeds. Approximately 80 M2 seeds from each pool were germinated in soil in four 144-cm2 pots (20 seeds per pot). Leaf samples from the 20 plants in each pot were pooled, and RNA was extracted using the TRIzol reagent. Five micrograms of total RNA was analyzed on a gel blot for expression of the PR-1 gene. Pools with high constitutive levels of PR-1 transcript were identified, and RNA was extracted from each individual plant in these pools. Five micrograms of total RNA from these pools was analyzed for PR-1 expression by RNA gel blot analysis. Individual plants constitutively expressing the PR-1 gene were thus identified and allowed to set seed. The mutant phenotype was confirmed in the M3 generation.
Cleaved Amplified Polymorphic Sequence Analysis for npr1-5
The EMS-induced npr1-5 mutation caused a C-to-T transition in the NPR1 gene (H. Cao, J. Shah, D.F. Klessig, and X. Dong, unpublished results), resulting in the substitution of serine for proline at amino acid 342 in the mutant NPR1 protein. This single base pair mutation also abolished an NlaIV restriction site present in the wild-type NPR1 gene. PCR primers were designed to amplify a 691-bp region covering amino acids 200 to 430 (5′-GAGGACACATTGGTTATACTC-3′; 5′-CAAGATCGAGCAGCGTCATCTTC-3′). Restriction analysis of the PCR-amplified products with NlaIV generated two fragments of 266 and 425 bp for the npr1-5 allele and three fragments of 182, 243, and 266 bp for the wild type and nim1-3 allele. PCR amplifications were performed as described by Konieczny and Ausubel (1993).
Genetic Analysis
Backcrosses were performed by pollinating flowers of the npr1-5 parental line (SSI1 npr1-5) with pollen from an ssi1 npr1-5 double mutant. For all other genetic analyses, progeny from a backcrossed line homozygous for the ssi1 and npr1-5 mutant alleles was used. To generate ssi1 plants homozygous for the NPR1 wild-type allele, pollen from an ssi1 npr1-5 double mutant was used to pollinate flowers from Arabidopsis ecotype Nössen line 1/8 E/5 (Shah et al., 1997), that is, wild type at both the SSI1 and NPR1 loci. Likewise, to generate ssi1 plants homozygous for the nim1-3 mutant allele, pollen from an ssi1 npr1-5 double mutant was used to pollinate flowers from an SSI1 nim1-3 plant (ecotype Wassilewskija). Success of the cross was confirmed by cleaved amplified polymorphic sequence (CAPS) analysis on F1 plants for heterozygosity at the NPR1 locus. Segregation of the ssi1 mutant allele was monitored in the F2 progeny by RNA gel blot or dot blot analysis for constitutive PR-1 gene expression. CAPS analysis was performed on DNA from these phenotypically ssi1 plants to identify plants homozygous for the wild-type NPR1 or the nim1-3 mutant allele. For mapping analysis, pollen from an ssi1 npr1-5 double mutant (ecotype Nössen) was used to pollinate flowers from a wild-type plant of ecotype Columbia. F2 progeny plants from the above cross were monitored for spontaneous lesion and constitutive PR-1 expression phenotype by dot blot analysis. DNA for PCR was isolated from leaf tissue by the method of Konieczny and Ausubel (1993) and used for CAPS or simple sequence length polymorphism (SSLP) marker analysis, as described previously (Konieczny and Ausubel, 1993; Bell and Ecker, 1994).
Crosses with Arabidopsis Plants Expressing the nahG Gene
Transgenic NahG plants of ecotype Nössen were generated by using the Agrobacterium tumefaciens–mediated whole plant infiltration protocol (Bechtold et al., 1997). After three generations of selfing, a transgenic line having the T-DNA integrated at a single locus was identified and used for all experiments. Pollen from this transgenic NahG plant was used to pollinate flowers from an ssi1 npr1-5 plant. Success of the cross was confirmed by analyzing expression of the nahG gene in the F1 plants.
ACKNOWLEDGMENTS
We thank Eric Ward and John Ryals for providing the nahG clone and Barbara Kunkel for providing P. s. tomato DC3000. We are grateful to Frank Tsui for conducting HPLC analyses for SA and SAG quantitation. We gratefully acknowledge D'Maris Dempsey for critical reading of the manuscript. This work was made possible by grants from the National Science Foundation (Nos. MCB9319371 and MCB-9723952) to D.F.K.
Footnotes
-
↵1 Current address: Division of Biology, Kansas State University, 303 Ackert Hall, Manhattan, KS 66506.
- Received October 9, 1998.
- Accepted November 30, 1998.
- Published February 1, 1999.