Resistance to Turnip Crinkle Virus in Arabidopsis Is Regulated by Two Host Genes and Is Salicylic Acid Dependent but NPR1, Ethylene, and Jasmonate Independent

Abstract

Inoculation of turnip crinkle virus (TCV) on the resistant Arabidopsis ecotype Dijon (Di-17) results in the development of a hypersensitive response (HR) on the inoculated leaves. To assess the role of the recently cloned HRT gene in conferring resistance, we monitored both HR and resistance (lack of viral spread to systemic tissues) in the progeny of a cross between resistant Di-17 and susceptible Columbia plants. As expected, HR development segregated as a dominant trait that corresponded with the presence of HRT. However, all of the F₁ plants and three-fourths of HR⁺ F₂ plants were susceptible to the virus. These results suggest the presence of a second gene, termed RRT, that regulates resistance to TCV. The allele present in Di-17 appears to be recessive to the allele or alleles present in TCV-susceptible ecotypes. We also demonstrate that HR formation and TCV resistance are dependent on salicylic acid but not on ethylene or jasmonic acid. Furthermore, these phenomena are unaffected by mutations in NPR1. Thus, TCV resistance requires a yet undefined salicylic acid–dependent, NPR1-independent signaling pathway.

INTRODUCTION

Plants resisting pathogen attack frequently activate a variety of defense responses that are initiated by the direct or indirect interaction between the products of a plant-encoded resistance (R) gene and its corresponding pathogen-encoded avirulence (Avr) gene. A majority of such interactions also result in the induction of host cell death at the site of pathogen infection, a phenomenon known as the hypersensitive response (HR). Subsequent to the HR, the uninoculated tissues usually develop a long-lasting, enhanced resistance to further attacks by the same or even unrelated microbial pathogens; this is referred to as systemic acquired resistance (SAR) (Ryals et al., 1996; Durner et al., 1997; Dempsey et al., 1999).

Although the HR is one of the earliest manifestations of disease resistance, it remains unclear whether this phenomenon is a prerequisite for gene-for-gene-mediated disease resistance or whether it simply results from the activation of multiple defense signaling pathways. For example, the HR does not develop in potato during Rx-mediated resistance against potato virus X (Köhm et al., 1993) or in the Arabidopsis dnd1 mutant after pathogen infection (Yu et al., 1998). Similarly, in barley, the HR is dispensable in Mlg-specified resistance to the powdery mildew fungus (Koga et al., 1990; Schiffer et al., 1997). Several studies on tobacco infected with tobacco mosaic virus (TMV) have also indicated that localization of the virus is dependent on accumulation of salicylic acid (SA) and not cell death (Bi et al., 1995; Mittler et al., 1996; Ryals et al., 1996; Mur et al., 1997). In contrast, an HR is required for Mla-specified resistance to powdery mildew in barley (Koga et al., 1990; Schiffer et al., 1997). Therefore, the relationship between HR and disease resistance (cause or consequence) most probably depends on the specifics of the interaction between the resistance and avirulence gene products and the downstream components activated as a result of their interaction.

Many studies have demonstrated that SA is an important component of the signal transduction pathway leading to disease resistance (Durner et al., 1997; Dempsey et al., 1999). For example, plants unable to accumulate SA because of the expression of a bacterial nahG gene encoding salicylate hydroxylase fail to develop SAR and exhibit heightened susceptibility to pathogen infection (Gaffney et al., 1993; Delaney et al., 1994). In addition to SA, ethylene and jasmonic acid (JA) serve as important signals for the induction of various defense responses. JA and ethylene have been implicated in (1) resistance to several fungi (Knoester et al., 1998; Thomma et al., 1998; Vijayan et al., 1998), (2) induced systemic resistance triggered by Pseudomonas fluorescens (Pieterse et al., 1998; Pieterse and Van Loon, 1999), (3) activation of genes encoding thionin and defensin (Epple et al., 1995; Penninckx et al., 1996, 1998), and (4) activation of enzymes involved in phytoalexin synthesis (Boller et al., 1983; Ecker and Davis, 1987; Mauch and Staehelin, 1989). Interestingly, ethylene and JA work in conjunction with SA to signal some but not all defense responses (Dong, 1998).

Genetic analysis of the SA-regulated pathway leading to disease resistance has revealed that a key signaling component is encoded by the NPR1/NIM1 gene (Cao et al., 1994, 1997; Delaney et al., 1995; Glazebrook et al., 1996; Ryals et al., 1997; Shah et al., 1997). Arabidopsis mutants lacking a functional NPR1/NIM1 gene are unable to express the pathogenesis-related (PR) genes in response to SA or its functional analogs 2,6-dichloroisonicotinic acid and benzothiadiazole. In addition, npr1/nim1 plants show enhanced susceptibility to both bacterial and fungal pathogens (Cao et al., 1994; Delaney et al., 1995; Glazebrook et al., 1996; Shah et al., 1997). Conversely, overexpression of NPR1 confers resistance against both bacterial and fungal pathogens in a dosage-dependent manner (Cao et al., 1998). Whereas NPR1 plays a critical role in resistance to various bacterial and fungal pathogens, its involvement in viral resistance is not known.

In Arabidopsis, resistance to most viral pathogens does not involve an HR (Ishikawa et al., 1991; Leisner et al., 1993; Lee et al., 1994; Callaway et al., 1996). However, inoculation of turnip crinkle virus (TCV) (Morris and Carrington, 1988) on plants from the resistant ecotype Dijon (Di-0 or Di-17) results in both an HR and the induction of PR gene expression (Simon et al., 1992; Dempsey et al., 1993, 1997; Uknes et al., 1993). In contrast, TCV-susceptible ecotypes, including Columbia (Col-0), fail to mount an HR, exhibit delayed and weak PR gene expression, and develop systemic disease symptoms (Li and Simon, 1990; Dempsey et al., 1993). Genetic analyses revealed that HR development is conferred by a single dominant gene termed HRT (for HR to TCV) (Dempsey et al., 1997). HRT also appears to be required for resistance to TCV infection; all of the HR⁻ progeny from crosses between resistant and susceptible ecotypes developed systemic disease symptoms. However, HRT alone may not be sufficient for complete resistance because many of the HR⁺ progeny also succumbed to infection. In this study, we show that TCV resistance is influenced by a second locus, named RRT (for regulates resistance to TCV). Furthermore, we demonstrate that the HR and resistance are dependent on SA but independent of NPR1-, ethylene-, and JA-mediated defense signaling.

RESULTS

Defense Gene Expression during the Resistance Response

Several families of PR genes have been shown to be activated in Di-17 plants that resist TCV infection (Dempsey et al., 1993; Uknes et al., 1993). To determine whether other known defense genes are induced during this resistance response, we extracted total RNA from the TCV-inoculated and mock-inoculated leaves of Di-17 and Col-0 plants at 1 to 3 days postinoculation (DPI). The accumulation of gene transcripts for PR-1, phenylalanine ammonia lyase (PAL) (Wanner et al., 1995), glutathione S-transferase (GST1 and GST6) (Chen et al., 1996; Yang et al., 1999), defensin (PDF1.2 ) (Penninckx et al., 1996), thionin (THI2.1) (Epple et al., 1995), ascorbate peroxidase (APX1 and APX2) (Kubo et al., 1993; Karpinski et al., 1997), lipoxygenase (LOX2) (Bell and Mullet, 1993), NPR1 (Cao et al., 1997), 1-aminocyclopropane-1-carboxylic acid (ACC) synthase (Van der Straeten et al., 1992), and alternative oxidase (AOX) (Kumar and Soll, 1992) were then analyzed by reverse transcription–polymerase chain reaction (RT-PCR) and RNA gel blot analysis (Figure 1 and data not shown). Whereas the transcripts for ACC synthase and AOX accumulated to very low amounts (data not shown), transcripts for PR-1 and GST1 accumulated to high amounts in the TCV-inoculated leaves of Di-17 plants (Figure 1). Expression of the other genes was not induced in these leaves (Figure 1 and data not shown). In contrast, the TCV-inoculated leaves of Col-0 plants showed only basal-level expression for all of the defense genes assayed (Figure 1).

Figure 1.

PR-1, GST1, and PDF1.2 Expression and Systemic Spread of TCV in Di-17 × Col-0 F₁ and F₂ Plants.

Expression of PR-1, GST1, and PDF1.2 in the inoculated leaves of TCV (T)- and mock (M)-infected Di-17 and Col-0 plants and the F₁ and F₂ progeny obtained from a cross between them. The inoculated leaves were monitored for the presence (+) or absence (−) of the HR, and the RNA was extracted from the inoculated leaves at 4 DPI. The replication and spread of the virus were analyzed by extracting RNA from the uninoculated (TCV-U) bolt tissue at 10 DPI. Two F₁ and five F₂ plants were analyzed. The genotype of the F₂ plants at the HRT locus was determined by CAPS analysis; both homozygous and heterozygous plants were analyzed for the expression of defense genes in inoculated leaves, the presence of TCV RNA in the uninoculated bolt tissue, and the appearance (S) or absence (R) of disease symptoms. The blot (1 to 11) was sequentially probed for the indicated genes (PR-1, GST1, and PDF1.2). RNA extracted from the systemic tissues was run on a separate gel; rRNA was used as an internal control for gel loading and transfer for both blots.

Genetic Characterization of Resistance against TCV

We previously showed that the HR in TCV-infected plants is dependent on the dominant HRT gene (Dempsey et al., 1997). However, because many progeny from a Di-17 × Col-0 cross are HR⁺ but TCV susceptible, this gene may not be sufficient to confer complete resistance. To further characterize the role of HRT in TCV-infected plants, we crossed Di-17 and Col-0 plants. All of the F₁ progeny developed an HR on the TCV-inoculated leaves (Table 1) and accumulated increased amounts of PR-1 and GST1 transcripts (Figure 1). Nonetheless, these F₁ plants allowed systemic spread of the virus (Figure 1) and developed disease symptoms (crinkled leaves and drooping bolts; Table 1). One explanation for the TCV-susceptible phenotype of these F₁ plants might be that resistance requires homozygosity for the HRT allele. Alternatively, a second recessive gene, in addition to HRT, might regulate resistance.

Table 1.

Segregation of the HR and Resistance in a Cross between Di-17 and Col-0 Plants

			Phenotype of Plants
Crossa	Generation	Total No. of Plants	HR+/Sc	HR+/Rd	HR−/S	HR−/R	χ2b	P
Di-17 × Col-0	F1	6	6	0	0	0
	F2	458	280	85	93	0	0.01	>0.8

^aThe pollen-accepting plant is indicated first and the pollen donor second.

^bCalculation based on a 13:3 segregation.

^cSusceptible, disease symptoms include crinkling of leaves and drooping of the bolt.

^dResistant, no disease symptoms.

To determine whether resistance depends on HRT dosage or the presence of a second gene, we analyzed F₂ progeny from the Di-17 × Col-0 cross. Of the 458 F₂ plants, 373 were susceptible and 85 were resistant. The ratio between susceptible and resistant plants fits very closely the 13:3 segregation expected if resistance is due to at least one dose of HRT and homozygosity for a recessive gene at an unlinked locus (; 0.95 > P > 0.8), a gene we have named RRT. In contrast, if resistance were dependent on two doses of HRT, ∼33% of the HR⁺ F₂ plants (122) should have been resistant. Because the data obtained do not fit the Mendelian ratio of 2:1 (; P < 0.01), we can reject this possibility. As expected, the HR segregated as a dominant trait (Table 1; 365 HR⁺/93 HR⁻ plants; ; 0.05 > P > 0.01), although the number of HR⁺ plants (365) was somewhat higher than expected (344). This greater number of HR⁺ plants could be the result of differential transmission at the HRT locus, and indeed, genotype analysis showed that this was the case (see below).

To provide additional support for our hypothesis that both HRT and RRT regulate resistance to TCV, we monitored the HRT genotype, the ability to form an HR, and viral resistance in a population of 355 F₂ plants derived from a cross between Di-17 and Col-0 plants (Table 2). Using a cleaved amplified polymorphic sequence (CAPS) marker that can differentiate between the HRT allele of Di-17 and the hrt allele of Col-0 (Cooley et al., 2000), we identified 119 plants as HRT/HRT, 164 as HRT/hrt, and 72 as hrt/hrt. All of the HRT/− plants developed HR at 3 DPI, whereas none of the hrt/hrt plants showed any visible lesions, confirming the requirement of HRT for the HR. As was previously observed, the number of HR⁺ plants (283) was somewhat more than what was expected (267). Moreover, substantially more of these plants were homozygous for the HRT allele (119) than would be predicted (89) by a Mendelian ratio of 1:2:1 (Table 2). Thus, recovery of the dominant HRT allele (57%) in the F₂ is significantly higher than that of the recessive allele (43%) (; P < 0.01). Similar results were obtained when selfing other F₁ plants, suggesting preferential transmission of HRT by one or both gametophytes in this specific F₁ hybrid.

Table 2.

Segregation of the HRT Locus and Resistance in a Cross between Di-17 and Col-0 Plants

			Genotypeb at the HRT Locus			Resistantc Plants
Crossa	Generation	Total No. of Plants	HRT/HRT	HRT/hrt	hrt/hrt	HRT/HRT	HRT/hrt	hrt/hrt
Di-17 × Col-0	F2	355	119	164	72	42	19	0

^aThe pollen-accepting plant is indicated first and the pollen donor second.

^bThe genotype at HRT was determined by CAPS analysis.

^cResistant, no disease symptoms.

Further analysis of these F₂ plants confirmed that HRT is also required for resistance. All 72 of the hrt/hrt plants not only failed to develop an HR but also were susceptible to TCV infection. If resistance were not dependent on HRT but required only rrt, then ∼18 hrt/hrt plants should have been resistant. In addition, resistance does not correlate with the presence of two doses of the HRT allele; 42 of the 119 HRT/HRT plants and 19 of 164 HRT/hrt plants were TCV resistant (Table 2). Rather, the data are more consistent with the numbers predicted if a second locus, in addition to HRT, is required for resistance. In this scenario, 25% of the HRT/HRT plants (30) and 25% of the HRT/hrt plants (41) should be resistant. The number of HRT/HRT–resistant plants (42) was somewhat more than expected, and the number of HRT/hrt–resistant plants (19) was only half of that expected. A likely explanation for the discrepancy between the observed and expected numbers is that the increased dosage of HRT enhances the durability of resistance or reduces the extent to which environmental factors affect the expressivity of HRT. Further supporting this possibility is the observation that increased expression of HRT in transgenic Col-0 plants leads to enhanced resistance to TCV (Cooley et al., 2000).

To determine whether HRT dosage has any effect on the extent of viral spread and defense gene expression, we monitored these phenomena in a subset of the Di-17 × Col-0 F₂ progeny. All of the plants exhibiting disease symptoms were found to contain viral RNA in the uninoculated tissue. These plants accumulated similar amounts of viral RNA, regardless of whether they contained zero, one, or two doses of the HRT allele (Figure 1). In contrast, PR-1 and GST1 expression in the TCV-inoculated leaves occurred only in F₂ progeny containing at least one dose of the HRT allele and independent of their resistance phenotype. Comparable expressions of these genes were generally detected in homozygous and heterozygous plants, suggesting that these events are not sufficient to confer resistance (Figure 1).

Role of SA in HR Development and Resistance to TCV

We have previously shown that in Arabidopsis, increases in SA and its glucoside (SAG) correlate with resistance to TCV (Dempsey et al., 1997). To further explore the role of SA in TCV resistance, we first tested whether exogenously applied SA confers resistance to plants from the susceptible ecotypes Col-0 and Nössen (Nö). Two days after treatment with SA, plants were infected with TCV and then monitored for development of an HR, the appearance of disease symptoms, and the presence of TCV in uninoculated tissue. At 3 DPI, no macroscopic HR was evident on the TCV-inoculated leaves of SA-treated Col-0 plants. However, by 10 DPI, disease symptoms were evident (data not shown), and viral RNA had accumulated to very high amounts in the systemic leaves (Figure 2). No differences in the amounts of viral RNA were observed between the SA-treated and water-treated control Col-0 plants. To ensure that the SA treatment was effective in inducing defense signaling, we monitored PR-1 expression in the inoculated and systemic tissues at 10 DPI. SA treatment was observed to induce PR-1 expression in these plants (Figure 2); however, that was insufficient to confer resistance to TCV. Similar results were obtained with Nö plants (data not shown).

Figure 2.

PR-1 Expression and Systemic Spread of TCV in SA-Treated Plants.

Three-week-old Col-0 plants were treated with SA (500 μM) or water and then TCV (T)- or mock (M)-inoculated 2 days after treatment. The expression of PR-1 and the accumulation of viral genome were analyzed in both inoculated (I) and uninoculated (U) tissues at 10 DPI. The blot (1 to 9) was sequentially probed for TCV, PR-1, and rRNA (the rRNA served as an internal control for gel loading and transfer).

We next determined whether constitutively increased amounts of endogenous SA would confer TCV resistance. Several Arabidopsis mutants have been isolated that exhibit enhanced resistance to bacterial and fungal pathogens, constitutively express PR genes, and have increased SA/SAG (Durner et al., 1997; Yang et al., 1997). Of these, the cpr5 (Bowling et al., 1997), cep (Silva et al., 1999), ssi1 (Shah et al., 1999), ssi2 (J. Shah, P. Kachroo, and D.F. Klessig, unpublished results), and cpr22 (K. Yoshioka, F. Tsui, P. Kachroo, S.B. Sharma, and D.F. Klessig, unpublished results) mutants were analyzed for resistance to TCV. All of these mutants are derived from the TCV-susceptible Col-0, Nö, or Wassilewskija ecotypes. When infected, all exhibited severe disease symptoms and accumulated TCV RNA in the uninoculated tissues in amounts similar to those observed in the respective TCV-infected wild-type parents (data not shown). Thus, increased endogenous SA/SAG and the constitutive expression of PR genes are not sufficient to confer resistance to TCV.

Although the preceding studies demonstrate that increased amounts of SA do not confer resistance to susceptible plants, they do not address whether SA is involved in activating resistance in Di-17 plants. Therefore, we tested the effect of the nahG-encoded salicylate-degrading enzyme salicylate hydroxylase on HR development and TCV resistance. Reciprocal crosses were made between Di-17 and NahG transgenic (ecotype Nö) plants. Because both HRT and the nahG transgene are dominant, the phenotypes were monitored in two F₁ progeny from each cross (Table 3). As a control, HR development and resistance to TCV also were assayed in F₁ progeny from a cross between Di-17 and wild-type Nö plants. All four F₁ plants derived from the Di-17 × NahG Nö cross accumulated transcripts for the nahG transgene and failed to express the prototypical SA-induced PR-1 gene after TCV infection (Figure 3 and Table 3). These plants, like the parental NahG Nö transgenic and wild-type Nö plants, failed to develop an HR after TCV infection. In contrast, the parental Di-17 plants and three F₁ plants derived from Di-17 × Nö cross developed an HR after TCV inoculation (Figure 3 and data not shown). Based on these results, increased amounts of SA are required for the HR to TCV. All four NahG-containing F₁ plants also exhibited disease symptoms and allowed viral replication and spread from the inoculated leaves into the uninoculated tissues (Figure 3). However, because the F₁ progeny from the Di-17 × Nö cross were TCV susceptible, the Nö ecotype did not appear to contain the recessive rrt allele required for resistance. Thus, analyses of the F₂ progeny were performed so that the importance of SA in resistance could be assessed.

Table 3.

Segregation of the HR and Resistance in a Cross between Di-17 and a NahG Transgenic Plant

Crossa	Generation	Total No. of Plants	NahG+/HR+	NahG+/HR−	NahG−/HR+	NahG−/HR−
NahG Nö × Di-17	F1	2	0	2 Sb	0	0
Di-17 × NahG Nö	F1	2	0	2 S	0	0
	F2	139	0	92 S	21 Rc; 23 S	3 S

^aThe pollen-accepting plant is indicated first and the pollen donor second.

^bS, susceptible; disease symptoms include crinkling of leaves and drooping of the bolt.

^cR, resistant; no disease symptoms.

Figure 3.

PR-1 Expression and Systemic Spread of TCV in Di-17 × nahG F₁ Plants.

Two F₁ plants each from a cross of Di-17 and NahG Nö and its reciprocal cross were inoculated with TCV and monitored for the presence (+) or absence (−) of the HR. TCV (T)- or mock (M)-inoculated Di-17 parental plants were used as controls. Expression of the PR-1 gene in the inoculated leaves was analyzed at 4 DPI. The systemic spread and replication of the virus (TCV-U) were determined by analyzing RNA extracted from the uninoculated bolt tissues at 10 DPI. The blot (1 to 7) was sequentially probed for PR-1, the nahG transgene, and TCV genome. RNA extracted from the systemic tissues was run on a separate gel. rRNA was used as an internal control for gel loading and transfer for both blots.

After self-pollination of an F₁ plant, 139 F₂ progeny were analyzed for HR formation and resistance to TCV as well as the presence of the nahG transgene (Table 3). Consistent with our hypothesis that SA is required for the activation of defense responses (including the HR) and resistance to TCV, all of the NahG⁺ plants were HR⁻ and all of the HR⁺ progeny were NahG⁻ (Table 3, and Figures 4 and 5). Moreover, all of the NahG⁺ F₂ progeny were susceptible to TCV infection.

Figure 4.

Morphological Phenotypes of TCV-Infected Di-17, Nö, and F₂ Plants from a Di-17 × NahG Nö Cross.

(A) TCV-inoculated leaves at 4 DPI. Both of the F₂ plants shown contain at least one allele of HRT. The HR⁺ F₂ plant lacks the nahG transgene, whereas the HR⁻ plant contains the nahG transgene, which leads to suppression of HR.

(B) The morphological phenotypes of the corresponding TCV-inoculated plants at 14 DPI. Except for the Di-17–resistant parent (R, resistant), the rest of the plants show severe crinkling of their leaves and stunted, drooping bolts (S, susceptible).

(C) This series shows systemic cell death induced by TCV in the corresponding TCV-infected susceptible plants. Uninoculated leaves were removed from the bolt at 14 DPI and stained with trypan blue. The leaves of the plants exhibiting disease symptoms show intensely stained areas of dead cells (marked by arrowheads), which are concentrated around the veins and gradually spread to the rest of leaf. Cell death was not evident in the uninoculated leaves of resistant Di-17 infected with TCV (left).

Figure 5.

PR-1 Expression and Systemic Spread of TCV in Di-17 × nahG F₂ Plants.

F₂ plants derived from a Di-17 × NahG Nö cross and the Di-17 and Nö parents were TCV (T)- or mock (M)-inoculated and scored for the presence (+) or absence (−) of an HR at 4 DPI. RNAs extracted at 4 DPI from the inoculated leaves of three HR⁺ and four HR⁻ F₂ plants (chosen at random) and their parents were analyzed for the expression of PR-1, the nahG transgene, and the replication of TCV in the inoculated leaves (TCV-I) (1 to 12). The systemic spread and replication of TCV in the uninoculated tissue (TCV-U) were determined by analyzing RNA extracted from bolt tissues at 10 DPI. RNA extracted from the systemic tissues was run on a separate gel, and rRNA was used as an internal control for gel loading and transfer for both blots.

To ensure that the HR⁻ susceptible phenotype in at least a portion of the F₂ plants was the result of SA deficiency rather than absence of the HRT allele, the genotypes at the nahG and HRT loci were tested by PCR and CAPS analysis in 101 of the 139 F₂ plants. From this analysis, 70 NahG⁺ plants were identified; 52 of these were either homozygous or heterozygous for the HRT allele, and all were susceptible to TCV. Given that rrt segregates as an independent locus, ∼13 of these 70 plants should have been rrt/rrt and resistant to TCV if SA were not required. Of the 31 NahG⁻ plants, all contained at least one copy of the HRT allele and developed an HR after TCV infection. In addition, 16 of these 31 plants were resistant to TCV. This number of resistant plants is more than expected (eight). However, that result would be explained if the NahG transgene is loosely linked to the RRT allele in the Nö parents; selecting for NahG⁻ progeny would thereby increase the number of individuals carrying the rrt allele associated with resistance. Taken together, these results indicate that both SA and HRT are required but not sufficient for resistance to TCV.

Role of NPR1 in the HR and Resistance to TCV

NPR1 plays an essential role in the SA-mediated signal transduction pathway that leads to the activation of PR genes (Cao et al., 1994; Delaney et al., 1995; Glazebrook et al., 1996; Shah et al., 1997). Because SA appears to be required for HR development and resistance, we asked whether NPR1 plays a role in signaling these phenomena by crossing Di-17 with npr1-1 (Col-0 ecotype; Cao et al., 1994) or npr1-5 (Nö ecotype; Shah et al., 1997) plants. Both of these npr1 alleles are recessive compared with the wild-type NPR1 allele. As expected, the two F₁ plants derived from the Di-17 × npr1-1 cross exhibited an HR after TCV infection but were still susceptible to TCV (Table 4). Using CAPS analysis, we identified F₂ progeny homozygous for npr1-1 or npr1-5. These plants were subsequently monitored for their ability to develop an HR and resist TCV infection. Of 13 npr1-1/npr1-1, HRT/− segregants, all exhibited an HR and two were TCV resistant (Table 4). Likewise, of 31 npr1-5/npr1-5, HRT/− F₂ segregants, all were HR⁺ (Figure 6) and nine were TCV resistant (Table 4). Resistance in these nine plants was confirmed by RNA gel blot analysis; no TCV genomic RNA was detected in the uninoculated tissues (data not shown). Because the HR was not blocked by the presence of the npr1-1 or npr1-5 alleles and because the number of TCV-resistant progeny expected (based on digenic segregation of HRT and rrt) was unaffected in npr1 homozygous plants, NPR1 is not required for the HR or resistance to TCV, even though SA is. This surprising result argues that resistance to TCV requires a yet to be characterized SA-dependent, NPR1-independent signaling pathway.

Table 4.

Segregation of the HR and Resistance to TCV in a Cross between Di-17 and npr1-1, npr1-5, etr1-1, or coi1-1 Mutants^d

Crossa	Generation	Total No. of Plants	Genotype at NPR1, ETR1 or COI1 Locib	Phenotype of Plantsc				χ2d	Pe
				HR+/Sf	HR+/Rg	HR−/S	HR−/R
Di-17 × npr1-1 (Col-0)	F1	2	NPR1/npr1-1 (2)	2	0	0	0
	F2	121	npr1-1/npr1-1 (15)	11	2	2	0
Di-17 × npr1-5 (Nö)	F2	237	npr1-5/npr1-5 (41)	22	9	10	0	0.274	>0.8
etr1-1 (Col-0) × Di-17	F1	4	ETR1/etr1-1 (4)	4	0	0	0
	F2	102	ETR1/etr1-1 plus
			etr1-1/etr1-1 (74)	44	14	16	0	0.462	>0.8
coi1-1 (Col-0) × Di-17	F1	4	COI1/coi1-1 (4)	4	0	0	0
	F2	137	coi1-1/coi1-1 (38)	21	6	11	0	0.413	>0.8

^aThe pollen-accepting plant is indicated first and the pollen donor second.

^bThe genotype at the NPR1, COI1, and ETR1 loci were determined by CAPS analysis. The total number of plants of the given genotype that were analyzed for the HR and resistance to TCV is given in parentheses.

^cCAPS analysis indicated that all TCV-resistant plants contained at least one copy of the dominant HRT allele, whereas all HR⁻-susceptible plants were homozygous for the recessive hrt allele. Except for the Di-17 × npr1-1 cross, which showed skewed segregation of the HRT locus similar to the Di-17 × Col-0 cross (Table 2), the rest of the other crosses showed 1:2:1 Mendelian segregation of the HRT locus (data not shown).

^dCalculation based on a 9:3:4 ratio.

^eTwo degrees of freedom.

^fS, susceptible; disease symptoms include crinkling of leaves and drooping of the bolt.

^gR, resistant; no disease symptoms.

Figure 6.

The HR in TCV-Inoculated Leaves of F₂ Progeny from Various Crosses.

A comparison (at 4 DPI) of TCV-inoculated leaves from F₂ progeny derived from the crosses between Di-17 and wild-type Col-0, npr1-5, etr1-1, or coi1-1 plants. The F₂ progeny from the crosses between Di-17 and npr1-5, etr1-1, or coi1-1 were homozygous for the mutant alleles and had at least one copy of the HRT gene. The small specks seen on the inoculated leaves of the Di-17 × Col-0 hrt/hrt F₂ plant are damaged tissue or dried inoculation buffer. Those are also seen on mock-inoculated plants (data not shown; Dempsey et al., 1993).

Because activation of PR-1 expression in Arabidopsis plants that resist TCV infection correlates with increases in SA contents (Figure 1; Uknes et al., 1993; Dempsey et al., 1997), we tested whether this phenomenon depended on NPR1. RNA gel blot analysis of npr1-5/npr1-5 and npr1-1/npr1-1 F₂ plants that were homozygous or heterozygous for HRT indicated that PR-1 expression was activated in the inoculated leaves (Figure 7). Although the amount of expression in these leaves was substantial, it was not as strong as that detected in the inoculated leaves of NPR1/NPR1 F₂ plants or in the Di-17 controls. Thus, PR-1 expression appears to be regulated by both NPR1-dependent and -independent pathways in TCV-inoculated resistant plants. A similar conclusion has been drawn from studies with the npr1 mutants and bacterial pathogens (Glazebrook et al., 1996; Shah et al., 1997).

Figure 7.

PR-1 Expression in F₂ Progeny from the Crosses between Di-17 and npr1 Plants.

(A) F₂ plants from a Di-17 × npr1-5 cross.

(B) F₂ plants from a Di-17 × npr1-1 cross.

The resistant Di-17 parents were inoculated with TCV (T) or inoculation buffer (M) and monitored for the presence (+) or absence (−) of an HR at 4 DPI. The plants were subsequently scored as resistant (R) or susceptible (S) at 14 DPI, based on the absence or presence of disease symptoms, respectively. The genotype at the NPR1 locus was determined by CAPS analysis. RNA gel blot analysis was performed with RNA extracted from the inoculated leaves at 4 DPI. The blots were hybridized with probes for PR-1 and rRNA (1 to 7 in [A] and 1 to 6 in [B]), with the latter serving as an internal control for gel loading and transfer.

Role of Ethylene and JA in Signaling HR Development and Resistance to TCV

The role of ethylene and JA signaling in HR development and resistance to TCV was determined by using Arabidopsis mutants insensitive to either of these defense signaling molecules. To analyze the role of ethylene, Di-17 plants were crossed with the dominant ethylene-insensitive etr1-1 (ecotype Col-0) mutant (Chang et al., 1993). All of the F₁ progeny derived from this cross exhibited an HR but were susceptible to TCV (Table 4). Because the etr1-1 mutation is dominant, this result suggests that ethylene signaling is not required for the HR to TCV. To determine the effect of etr1-1 on resistance, 58 HRT/− F₂ progeny containing at least one dose of the dominant etr1-1 allele were identified by CAPS analysis and then infected with TCV. All of these plants exhibited an HR (Figure 6; ; ), and 14 of these were resistant to TCV (Table 4; ; 0.95 > P > 0.8). Resistance in these plants was confirmed by RNA gel blot analysis, which failed to detect TCV genomic RNA in the uninoculated tissues (Figure 8). Because the number of HR⁺ TCV-resistant plants conformed to previously observed digenic segregation ratios, despite the presence of etr1-1, we conclude that this mutation and therefore ethylene signaling had no effect on either HR formation or TCV resistance.

Figure 8.

Systemic Spread of TCV in etr1-1 × Di-17 F₂ Plants.

F₂ plants from an etr1-1 × Di-17 or a Di-17 × Col-0 cross were inoculated with TCV (T) and scored for the presence (+) or absence (−) of an HR. The genotype at the ETR1 locus was determined by CAPS analysis. The plants were scored as resistant (R) or susceptible (S) at 14 DPI, based on the absence or presence of disease symptoms, respectively. RNA was extracted from the uninoculated bolt tissues (TCV-U) at 14 DPI and analyzed for the presence of viral genome (1 to 12). This blot also was probed for rRNA as an internal control for gel loading and transfer.

The involvement of JA in the activation of defense responses against TCV was assessed by crossing the jasmonate-insensitive coi1-1 mutant (ecotype Col-0) (Xie et al., 1998) with a Di-17 plant. The F₁ progeny all developed an HR but were susceptible to TCV. To determine the effect of the recessive coi1-1 mutation on HR and resistance, 27 F₂ plants homozygous for the coi1-1 allele and containing at least one copy of the HRT allele were identified by CAPS analysis. All of these plants developed an HR (Figure 6; ; 0.8 > P > 0.5), and six were resistant to TCV (Table 4; ; 0.95 > P > 0.8). The above ratios conform to what would be expected if the coi1-1 mutation had no effect on either the HR or resistance. Hence, JA, like ethylene, is not required for development of the HR or resistance to TCV.

In the previous section, we showed that although TCV-induced PR-1 expression is SA dependent, it can be activated by way of NPR1-dependent and -independent pathways (Figure 7). Because ethylene and JA are the two other major signals for activation of defense responses after pathogen attack, we asked whether PR-1 expression after TCV inoculation was affected by the etr1-1 or coi1-1 mutations. Analysis of at least 10 HRT F₂ plants insensitive to either ethylene or JA (Figure 9) failed to detect any important difference in the amounts of TCV-induced PR-1 expression in comparison with that observed in similar F₂ progeny from a cross between Di-17 and Col-0. Thus, the pathway leading to PR-1 expression after TCV infection appears to require neither ethylene nor JA but only SA.

Figure 9.

PR-1 Expression in etr1-1 × Di-17, coi1-1 × Di-17, and Di-17 × Col-0 F₂ Plants.

F₂ plants from crosses between Di-17 and etr1-1, coi1-1, or Col-0 plants were inoculated with TCV (T) and scored for the presence (+) or absence (−) of an HR at 4 DPI. The genotype at the ETR1 or COI1 locus was determined by CAPS analysis. The plants were scored as resistant (R) or susceptible (S) at 14 DPI, based on absence or presence of disease symptoms, respectively. RNA was extracted from the inoculated leaves at 4 DPI and analyzed for PR-1 expression (1 to 13). The blot was also probed for rRNA as an internal control for gel loading and transfer.

DISCUSSION

In this article, we demonstrate that the dominant HRT allele is necessary and sufficient for development of an HR after TCV infection. HRT also is required for resistance to TCV; however, the results of our genetic analyses argue that a homozygous recessive allele of a second unlinked gene, termed rrt, is also required. HRT was recently cloned and shown to encode a classic leucine zipper/nucleotide binding site/leucine rich repeat–containing protein (Cooley et al., 2000). In contrast, the structure and function of rrt are unknown. The recessive rrt allele found in Di-17 plants could encode a protein that is unable to suppress an active resistance pathway. This situation appears to exist in wheat, where an allele of the LrT2 locus confers resistance to stem rust by encoding a nonfunctional form of protein that otherwise suppresses resistance conferred by other R genes (Dyck, 1987). Alternatively, the recessive rrt allele might encode a protein that directly or indirectly interacts with HRT or a downstream HRT-activated factor to transduce the signal leading to TCV resistance. In this scenario, the recessive nature of the rrt allele could be explained if it encodes a protein that functions as a multimer, and the allele carried by the susceptible ecotypes encodes a dominant-negative variant that destroys the activity of a multimeric complex. We currently are unable to distinguish between these possibilities.

Many studies have demonstrated that SA is a critical signal for the activation of defense responses and disease resistance (Durner et al., 1997; Dempsey et al., 1999). Indeed, although susceptibility to TCV was unaffected by treatment with exogenous SA or in mutants exhibiting high constitutive amounts of SA and PR gene expression, both HR development and TCV resistance were blocked by the presence of nahG-encoded salicylate hydroxylase, which prevents accumulation of SA (Table 3 and Figures 2 to 5). The possibility that the inability to accumulate SA reduces the size of TCV-induced lesions (resulting in micro-HR formation rather than loss of an HR) was ruled out by trypan blue staining of TCV-inoculated leaves from NahG⁺/− HRT/− F₂ progeny (data not shown). These results argue that SA alone is insufficient to confer resistance to TCV in the absence of HRT; it is required, however, for the HRT-mediated activation of an HR and resistance.

Because NPR1 encodes a critical component of the SA-mediated signal transduction pathway leading to PR gene expression and disease resistance (Cao et al., 1994; Delaney et al., 1995; Glazebrook et al., 1996; Shah et al., 1997), we tested whether mutations in this gene would affect PR expression, HR development, or disease resistance after TCV infection. Plants carrying the npr1-1 or npr1-5 mutation exhibited delayed and decreased expression of PR-1 after TCV infection. In contrast, PR-1 induction was completely inhibited in TCV-infected plants expressing the nahG transgene. Thus, PR-1 expression appears to be regulated by both an SA-dependent, NPR1-independent pathway and an SA- and NPR1-dependent pathway (Figure 10). Earlier studies using npr1 mutants and bacterial pathogens also suggested the existence of an NPR1-independent pathway leading to PR expression (Glazebrook et al., 1996; Shah et al., 1997; Reuber et al., 1998).

Figure 10.

Model for Induction of the HR and Resistance to TCV.

Inoculation of TCV on resistant Arabidopsis results in an HR at the point of inoculation and the accumulation of SA (Dempsey et al., 1997). Development of the HR is mediated by the dominant HRT gene. Resistance is regulated by HRT and a second gene, RRT. The product of the dominant RRT allele or alleles present in susceptible ecotypes may act as a suppressor (T-bars) of the HRT-mediated resistance pathway downstream of SA and HR. Alternately, the protein encoded by the recessive rrt allele in Di-17 may directly or indirectly interact with HRT or a component downstream in the HRT-mediated defense pathway to facilitate transduction of the signal leading to TCV resistance. Although HRT is required for resistance, it is unclear if the HR is necessary also (question marks). Initiation of the HR and the signaling steps leading to resistance are blocked (T-bars) in plants expressing the nahG gene encoding salicylate hydroxylase. SA-mediated signaling leading to PR-1 expression is transduced by both NPR1-dependent and NPR1-independent pathways, which are blocked by the expression of nahG. The NPR1-dependent pathway, utilized for the activation of resistance to some bacterial and fungal pathogens, is not required for the HR or resistance to TCV. Ethylene- or JA-mediated signaling that leads to PDF1.2 expression and resistance to some fungal and bacterial pathogens is not necessary for the HR or resistance to TCV because mutations resulting in insensitivity to ethylene (etr1-1) or JA (coi1-1) do not affect either of these phenotypes. Thus, the pathways required for resistance to TCV differ from those for bacterial or fungal pathogens.

In contrast, neither HR development nor resistance to TCV was altered in npr1-1 or npr1-5 mutant plants. Consistent with these results, the HR was shown to develop normally in npr1-1 mutant plants infected with an avirulent bacterial pathogen (Cao et al., 1994). However, these same npr1 mutations caused increased susceptibility to bacterial and fungal pathogens (Cao et al., 1994; Shah et al., 1997). Given that both of the npr1 mutant alleles used in this study contain a point mutation in the ankyrin repeats, perhaps they fail to disrupt a function necessary for HR formation and TCV resistance. Alternatively, these phenomena may be mediated by an NPR1-independent pathway that is distinct from the NPR1-dependent pathways leading to fungal and bacterial resistance (Figure 10).

Some precedence for an NPR1-independent viral resistance pathway comes from the studies of Carr and colleagues (Chivasa et al., 1997; Chivasa and Carr, 1998; Murphy et al., 1999), who demonstrated that in tobacco, resistance to TMV is mediated by an SA-dependent pathway that appears to require AOX. Strikingly, although resistance to TMV is blocked by salicylhydroxamic acid (SHAM), an inhibitor of AOX activity, the TMV-induced activation of PR expression is not. In addition, SHAM treatment does not affect resistance to a bacterial or a fungal pathogen (Chivasa et al., 1997). Thus, tobacco appears to contain at least two distinct SA-regulated defense pathways. One pathway presumably is NPR1 dependent and mediates PR expression and resistance to bacterial and fungal pathogens, whereas the other is probably NPR1 independent and activates resistance to a virus.

In addition to SA, the secondary signals ethylene and JA have been implicated in the induction of various defense genes and the activation of resistance to certain pathogens (Xu et al., 1994; Epple et al., 1995; Penninckx et al., 1996, 1998; Dong, 1998; Pieterse and Van Loon, 1999). Thus, their role in establishing resistance to TCV was assessed. The observation that the ethylene- or JA-inducible defense genes PDF1.2 and THI2.1 are not expressed in TCV-infected Di-17 plants raised the possibility that neither signal is involved in activating TCV resistance. This conclusion was confirmed by genetic studies with ethylene (etr1-1)- and JA (coi1-1)-insensitive mutants. Neither mutation affected the HRT-mediated development of an HR (Figure 6) or resistance to TCV (Table 4 and Figure 8). In addition, these mutations had no detectable effect on PR-1 gene induction by TCV (Figure 9). However, because SA is the primary signal for HR development and PR-1 expression, small contributions of ethylene or JA to these phenomena may have gone undetected in these studies.

In conclusion, we have identified a new gene, designated RRT, whose recessive allele works in conjunction with HRT to signal TCV resistance by way of an SA-dependent pathway that is independent of NPR1, ethylene, and JA (Figure 10). In view of these features, the TCV resistance pathway differs from the SA-dependent, NPR1-dependent and the SA-independent, ethylene- or JA-dependent pathways involved in activating resistance to bacterial and fungal pathogens. Strikingly, however, the HRT/rrt–mediated resistance pathway shares many similarities with the SA-dependent, ethylene-independent N gene–regulated pathway for TMV resistance in tobacco (Knoester et al., 1998; Murphy et al., 1999). These discoveries provide further evidence that viral resistance in plants is activated via a mechanism distinct from those used for other microbial pathogens. Future studies should help clarify the relationship between the two viral resistance pathways as well as provide new insights into how plants protect themselves from pathogens, particularly viruses.

METHODS

Conditions for Plant Growth and Viral Infections

Plants (Arabidopsis thaliana) were grown in growth chambers as described by Shah et al. (1999). Transcripts synthesized in vitro from a cloned cDNA of turnip crinkle virus (TCV) genome using T7 RNA polymerase were used for viral infections (Dempsey et al., 1993; Oh et al., 1995). For inoculations, the viral transcript was suspended at a concentration of 0.05 μg/μL in inoculation buffer, and the infection was performed as described earlier (Dempsey et al., 1993). The hypersensitive response (HR) was determined visually 3 to 5 days postinoculation (DPI). Tissues (leaves or bolt) were removed at various times for RNA extraction and analysis of defense gene activation and viral replication by RNA gel blot hybridization.

Histochemistry and Microscopy

Microscopic visualization of HR was conducted by staining TCV-inoculated leaves at 3 to 4 DPI with trypan blue. Samples were processed and analyzed as described by Bowling et al. (1997). Systemic cell death was studied by staining the uninoculated leaf at 14 DPI.

Chemical Treatment of Plants

Three-week-old plants were sprayed and subirrigated with a solution of 500 μM salicylic acid (SA). Control plants were treated with water, and at 2 DPI three leaves per plant were inoculated with TCV RNA.

RNA Extraction and Gel Analysis

Small-scale RNA extractions was performed with TRIzol reagent (Gibco BRL, Rockville, MD), according to the manufacturer's instructions. RNA gel blot analysis and synthesis of random primed probes were performed as described earlier (Shah et al., 1999). RNA gel blot hybridization was performed as described previously (Kachroo et al., 1995).

Cleaved Amplified Polymorphic Sequence Analyses for HRT, NPR1, ETR1, and COI1

The genotypes of the F₂ plants at the NPR1, ETR1, COI1, and HRT loci were determined by conducting cleaved amplified polymorphic sequence (CAPS) analysis. The polymerase chain reaction (PCR) primers used for amplifying the NPR1 locus were as described by Shah et al. (1999), and the plants containing npr1-1 and npr1-5 alleles were identified by digesting the amplified product with NlaIII or NlaIV, respectively. Because the etr1-1 (Chang et al., 1993) mutation does not lead to an alteration of any restriction site, we used the dCAPS technique (Neff et al., 1998) to generate polymorphism between the wild type and etr1-1 alleles. PCR primers were designed to amplify a 100-bp region encompassing the mutant base, where the reverse primer contained two altered bases at the 3′ end such that the presence of the mutant base in etr1-1 allele results in loss of an ApaLI restriction site. Therefore, digestion of the PCR product from plants homozygous for the wild-type allele yielded two bands at 75 and 25 bp, whereas the digested PCR product amplified from etr1-1 DNA generated only a single band at 100 bp. The PCR reaction (ETR1 Fwd, 5′-CTTTGTGAAGAAATCAGCCGTGT-3′; ETR1 Rev, 5′-CCATAAGTTAATAAGATGAGTTGGTGCA-3′) was performed at an annealing temperature of 56°C for 50 cycles. The PCR-based analysis to determine the genotype at the COI1 locus was done as described by Xie et al. (1998). The genotype at the HRT locus was determined by producing a 950-bp PCR product, followed by digestion with AluI (Cooley et al., 2000). After digestion, the PCR product of Dijon (Di-17) HRT allele generated two bands of ∼100 and 850 bp. The 850-bp fragment was further cleaved into two fragments of 300 and 550 bp when the PCR products were derived from hrt/hrt Columbia (Col-0) or Nössen (Nö) ecotypes.

Genetic Analysis

Reciprocal crosses of Di-17 plants with Nö plants homozygous for the nahG transgene were performed. The success of the crosses was determined by CAPS analysis and by analyzing expression of the nahG gene. To facilitate easy identification of F₂ plants containing the nahG transgene, two PCR primers (5′-GGCTTGCGCATCGGTATCGTCGGC-3′ and 5′-GCCATGGGCCCGATAGGCTTCTCG-3′) were designed to yield a 500-bp amplification product. PCR amplification of nahG was performed at an annealing temperature of 60°C for 35 cycles. Crosses of Di-17 with npr1-1 or npr1-5 plants were performed by using pollen from the mutant plants to pollinate the Di-17 flowers. The crosses with etr1-1 and coi1-1 plants were made by pollinating the flowers from these plants with pollen from Di-17. The F₁ plants were analyzed with NPR1-, ETR1-, or COI1- specific CAPS and with random CAPS or simple sequence length polymorphism (SSLP) markers from the genome to ensure that progeny indeed resulted from the crosses. The F₂ plants were first analyzed by gene-specific CAPS markers and subsequently were inoculated with TCV. DNA for PCR was isolated either from leaf or bolt tissue by the method of Konieczny and Ausubel (1993). CAPS or SSLP amplifications were performed as described earlier (Konieczny and Ausubel, 1993; Bell and Ecker, 1994).

NOTE ADDED IN PROOF

Interestingly, RPP8-dependent resistance to Peronospora parasitica has recently been shown to be largely independent of SA as well as NPR1 (McDowell, J.M., Dangl, J.L., and Holub, E.B. [2000]. Plant J., in press).