WRKY70 modulates the selection of signaling pathways in plant defense


  • Jing Li,

    1. Viikki Biocenter, Division of Genetics, Department of Biological and Environmental Sciences, Faculty of Biosciences, University of Helsinki, PO Box 56, FI-00014, Helsinki, Finland
    Search for more papers by this author
  • Günter Brader,

    1. Viikki Biocenter, Division of Genetics, Department of Biological and Environmental Sciences, Faculty of Biosciences, University of Helsinki, PO Box 56, FI-00014, Helsinki, Finland
    Search for more papers by this author
  • Tarja Kariola,

    1. Viikki Biocenter, Division of Genetics, Department of Biological and Environmental Sciences, Faculty of Biosciences, University of Helsinki, PO Box 56, FI-00014, Helsinki, Finland
    Search for more papers by this author
  • E. Tapio Palva

    Corresponding author
      *(fax +358 9 191 59076; e-mail tapio.palva@helsinki.fi).
    Search for more papers by this author

*(fax +358 9 191 59076; e-mail tapio.palva@helsinki.fi).


Cross-talk between signal transduction pathways is a central feature of the tightly regulated plant defense signaling network. The potential synergism or antagonism between defense pathways is determined by recognition of the type of pathogen or pathogen-derived elicitor. Our studies have identified WRKY70 as a node of convergence for integrating salicylic acid (SA)- and jasmonic acid (JA)-mediated signaling events during plant response to bacterial pathogens. Here, we challenged transgenic plants altered in WRKY70 expression as well as WRKY70 knockout mutants of Arabidopsis with the fungal pathogens Alternaria brassicicola and Erysiphe cichoracearum to elucidate the role of WRKY70 in modulating the balance between distinct defense responses. Gain or loss of WRKY70 function causes opposite effects on JA-mediated resistance to A. brassicicola and the SA-mediated resistance to E. cichoracearum. While the up-regulation of WRKY70 caused enhanced resistance to E. cichoracearum, it compromised plant resistance to A. brassicicola. Conversely, down-regulation or insertional inactivation of WRKY70 impaired plant resistance to E. cichoracearum. Over-expression of WRKY70 resulted in the suppression of several JA responses including expression of a subset of JA- and A. brassicicola-responsive genes. We show that this WRKY70-controlled suppression of JA-signaling is partly executed by NPR1. The results indicate that WRKY70 has a pivotal role in determining the balance between SA-dependent and JA-dependent defense pathways.


In order to survive, plants must rapidly sense and adapt to continual changes in their environment. One of the major environmental challenges is exposure to potential pathogens. This has led to the evolution of complex strategies to allow rapid modulation of cellular functions and mounting of an active defense response. Induced defenses play a major role in plant disease resistance and are regulated by a network of interconnecting signal transduction pathways, with ethylene (ET), jasmonic acid (JA) and salicylic acid (SA) as crucial signaling molecules. Accumulation of these plant hormones in response to pathogen recognition will trigger specific downstream signaling cascades and induce distinct sets of defense-related genes and production of defense proteins as well as secondary metabolites with antimicrobial activity leading to enhanced resistance (Brader et al., 2001; Glazebrook, 2001; Pieterse and Van Loon, 1999; Reymond and Farmer, 1998; Thomma et al., 2001).

Several of these induced defenses have been characterized in detail (Glazebrook, 2001; Kunkel and Brooks, 2002). Perhaps the best studied of these responses is the systemic acquired resistance (SAR) triggered by necrotizing pathogens and often associated with the hypersensitive response (HR) in incompatible plant–pathogen interactions (Dangl et al., 1996; Durrant and Dong, 2004; Greenberg, 1997; Nimchuk et al., 2003). One of the physiological hallmarks of SAR is an increase in endogenous SA, which is required to mediate the response. Transgenic plants or mutants deficient in SA accumulation are compromised in SAR (Glazebrook, 2001). For example, plants carrying the NahG transgene encoding the SA-degrading enzyme salicylate hydroxylase exhibit enhanced susceptibility to viral, fungal and bacterial pathogens (Delaney et al., 1994). Similarly, mutants in SA biosynthesis such as in SID2 encoding an isochorismate synthase (Wildermuth et al., 2001) cause increased susceptibility to a number of pathogens including the fungal biotroph Erysiphe cichoracearum (Dewdney et al., 2000).

The role of the ankyrin-repeat protein NPR1 (NIM1) as a central positive regulator of SAR transducing the SA signal to activate downstream pathogenesis-related (PR) gene expression has been well demonstrated (Dong, 2004; Pieterse and Van Loon, 2004). Mutations in NPR1 lead to loss of PR gene expression and compromised disease resistance (Cao et al., 1994, 1997; Chern et al., 2001; Delaney et al., 1995; Glazebrook et al., 1996; Ryals et al., 1997). During SAR, nuclear translocation of NPR1 is a prerequisite for the activation of PR-1 expression (Kinkema et al., 2000). Recently, Mou et al. (2003) demonstrated that cytosolic oligomers of NPR1 undergo monomerization, which is required for nuclear translocation in response to SA/INA (2,6 dichloroisonicotinic acid) or pathogen attack. Yeast two-hybrid studies have indicated that NPR1-dependent activation of PR genes requires interaction with the TGA subclass of bZIP transcription factors (Després et al., 2000). Such interaction has been directly visualized in protoplasts and also verified in planta (Fan and Dong, 2002).

Jasmonic acid (JA), an oxylipin hormone derived from linolenic acid, regulates not only diverse developmental processes but also many defense responses, which contribute to resistance against insect herbivores and distinct pathogens (Turner et al., 2002). Tissue damage caused by wounding or attack by necrotrophic pathogens will induce JA accumulation. JA, often together with ET, is a key mediator of SA-independent defenses and triggers a subset of defense-related genes such as b-CHI and PDF1.2 (Kunkel and Brooks, 2002; Penninckx et al., 1998; Thomma et al., 1998). Blocking the response to JA generally renders plants more susceptible to a variety of pathogens such as the bacterial pathogen Erwinia carotovora (Norman-Setterblad et al., 2000) and the fungal pathogens A. brassicicola, Botrytis cinerea (Thomma et al., 1998) and Pythium sp. (Staswick et al., 1998; Vijayan et al., 1998). The JA-insensitive coi1-1 mutant displays a pleiotrophic phenotype and is blocked in diverse JA-dependent responses such as expression of a specific set of genes, JA-mediated resistance to necrotrophic pathogens, the inhibition of root growth, and production of stress indicators such as anthocyanins (Feys et al., 1994; Penninckx et al., 1998; van Wees et al., 2003). Recent evidence suggests that COI1 is a component of the ubiquitin-ligase complex SCFCOI1 that regulates JA-responsive gene expression by recruiting repressors of these genes for ubiquitin-mediated destruction (Turner et al., 2002; Xu et al., 2002).

Global gene expression profiling suggests that induced disease resistance largely results from activation of partly overlapping sets of defense responses upon recognition of a particular type of pathogen (Schenk et al., 2000, 2003). Interestingly, growing evidence indicates that plants can selectively activate or repress signal transduction pathways controlling defenses against particular intruders (Kunkel and Brooks, 2002). Cross-talk between SA- and JA-dependent signaling pathways is thought to be involved in fine-tuning plant defenses, eventually leading to the activation of an optimal system of defense responses (Felton and Korth, 2000; Pieterse et al., 2001). The activation of SAR has been shown to suppress JA signaling, therefore favoring SA-dependent resistance to microbial biotrophs over JA-dependent resistance against insect herbivores and microbial necrotrophs (Felton and Korth, 2000; Pieterse et al., 2001). The mutual antagonism between SA- and JA-dependent defenses (Gupta et al., 2000; Kunkel and Brooks, 2002) appears to involve MPK4 and NPR1 (Petersen et al., 2000; Spoel et al., 2003; reviewed by Pieterse and Van Loon, 2004). In contrast to the requirement for nuclear localization of NPR1 during SAR, cytosolic NPR1 protein is implicated in SA-dependent inhibition of JA-mediated defenses (Spoel et al., 2003). We have recently shown that the transcription factor WRKY70 activates expression of SAR-related genes but suppresses expression of JA-responsive genes (Li et al., 2004). Expression of WRKY70 itself is up-regulated by SA but down-regulated by JA, suggesting that WRKY70 acts as a node of convergence for integrating SA- and JA-signaling events during plant defense. Modulation of WRKY70 expression altered Arabidopsis resistance to bacterial pathogens and indicated that WRKY70 could control the balance between plant defense pathways (Li et al., 2004).

To explore the role of WRKY70 in modulating the balance between plant defense pathways, we employed the two fungal pathogens Alternaria brassicicola and E. cichoracearum, which have distinct pathogenicity strategies and are subject to different plant resistance responses. Resistance to the necrotroph A. brassicicola requires JA-mediated signal transduction, and mutants with defects in JA signaling, e.g. coi1-1, are considerably more susceptible to this fungal pathogen (Thomma et al., 1998). In contrast, resistance to the fungal biotroph E. cichoracearum is triggered by SA-dependent pathways and involves activation of the SAR response (Dewdney et al., 2000). Here, we show that gain or loss of WRKY70 function cause opposite effects on Arabidopsis resistance to A. brassicicola or E. cichoracearum. Our results demonstrate that insertional inactivation of the WRKY70 gene leads to constitutive up-regulation of some JA-responsive genes such as AtCLH1 and enhanced MeJA- or pathogen-induced induction of others. In contrast, WRKY70 over-expression appears partly to mimic the coi1-1 mutant phenotype by blocking JA-dependent developmental and defense responses. We further show that WRKY70 acts as a repressor for a distinct subset of JA-responsive genes. This repression seems to be partly mediated by NPR1-dependent mechanisms. These data indicate that WRKY70 can modulate the balance between SA-and JA-dependent defenses.


Characterization of T-DNA insertion mutants of WRKY70

Our previous studies using transgenic Arabidopsis where WRKY70 was constitutively over-expressed or antisense-silenced (Li et al., 2004) suggested a function for WRKY70 in integrating SA and JA signaling in plant defense. To further elucidate the role of WRKY70 in defense signaling, we characterized two T-DNA lines (accession numbers SALK_025198 and GABI_324D11), which were predicted to contain a T-DNA insertion in the WRKY70 locus. The SALK line and GABI line were designated as wrky70-1 and wrky70-2, respectively. Both T-DNA insertion lines harbored a single T-DNA located either in the first intron (wrky70-1) or the second exon (wrky70-2) of the WRKY70 gene (Figure 1a). The homozygous progeny were selected by PCR using a WRKY70-specific primer pair. PCR fragments with the expected size were obtained from the DNA templates of wild-type and the genomic DNA mixture of wild-type and homozygous mutant plants, but not from that of homozygous mutant progeny because of the T-DNA insertion (Figure 1b). PCR with a primer pair from the WRKY70 gene and T-DNA left border was used to confirm the existence of the insert (data not shown). Homozygous progeny were used in subsequent experiments. Both the basal and SA-induced expression of WRKY70 was completely blocked in the insertion mutants (Figure 1c), indicating that each line might be a null allele of WRKY70. Unlike WRKY70 antisense plants, which were slightly larger than wild-type (Li et al., 2004), wrky70-1 and wrky70-2 mutants did not exhibit similar morphological differences from the wild-type under normal growth chamber conditions. One possible explanation for the observed phenotypic differences between the antisense lines and WRKY70 knockout mutants is that antisense silencing of WRKY70 might in addition cause suppression of an unknown gene affecting plant size.

Figure 1.

Molecular identification of two WRKY70 T-DNA insertion lines.
(a) Schematic picture of the T-DNA insertions in the WRKY70 locus. Exons (white boxes) and introns (gray boxes) are indicated. The numbers 1, 404/652 and 1328 represent the translation start site, the position of T-DNA insertion confirmed by sequencing PCR fragments with the primer pairs from the WRKY70 gene and T-DNA left border, and the stop codon of WRKY70, respectively.
(b) Homozygous lines were selected by PCR. Equal amounts of DNA templates from wild-type plants (WT) and the T-DNA tagged lines (wrky70-1 or wrky70-2) were mixed to confirm the appearance of the insert with the WRKY70-specific primer pairs. PCR fragments of expected size were amplified from DNA templates of WT and the DNA mixture of WT and mutants, but not from mutants only.
(c) The basal expression and SA induction of WRKY70 in homozygous T-DNA tagged plants. Each RNA sample was isolated from a leaf pool of six plants treated with SA. As a control for equal loading, rRNA was visualized after staining with ethidium bromide.

Insertional inactivation and over-expression of WRKY70 results in opposite responsiveness to MeJA

JA signaling is involved in plant responses to many biotic and abiotic stresses as well as in various developmental processes (Feys et al., 1994). To elucidate whether JA-mediated responses were affected by WRKY70, we firstly characterized the role of WRKY70 in root growth and accumulation of anthocyanins. As shown in Figure 2(a), root growth of vector control seedlings germinated on MS plates was inhibited by increasing concentrations of MeJA. The root lengths of control seedlings under 25, 50 or 250 μm MeJA treatments were significantly shorter (P < 0.05 by t-test in each case) compared to untreated seedlings. This observation is consistent with a previous report (Feys et al., 1994). This inhibition is blocked in the JA-insensitive mutant coi1-16 (Figure 2a). Interestingly, root growth of the over-expressing line S55 was not affected by MeJA at concentrations up to 500 μm (P < 0.05 by t-test in each case). This relief of root growth inhibition in the WRKY70-over-expressing line S55 (Figure 2a) suggested reduced sensitivity to JA, and that WRKY70 over-expression partially represses JA-mediated responses. WRKY70 also appeared to control JA-induced accumulation of anthocyanins. WRKY70 over-expression inhibited JA-induced accumulation of anthocyanins (Figure 2b,c), while insertional inactivation of WRKY70 had the opposite effect (Figure 2d,e). The anthocyanin content in seedlings germinated on MS plates containing 20 μm MeJA was significantly reduced in several over-expression lines when compared to control plants (Figure 2c), suggesting that WRKY70 over-expressors are less sensitive to JA and are hence compromised in JA-controlled responses. In contrast, enhanced JA-induced accumulation of anthocyanins was evident in both wrky70 mutants, suggesting that wrky70-1 and wrky70-2 mutants exhibit enhanced sensitivity to MeJA compared to wild-type plants (Figure 2d,e). Similar levels of accumulation of anthocyanins were detected in these two independent knockout lines (Figure 2e), strongly indicating that the anthocyanin phenotype results from the mutation in WRKY70. Taken together, these data from both WRKY70-over-expressing plants and wrky70 mutants suggest that WRKY70 is a negative regulator of the JA-mediated responses in Arabidopsis.

Figure 2.

Inhibitory effect of MeJA on root growth and accumulation of anthocyanins in WRKY70-over-expressing seedlings and wrky70 knockout plants.
(a) Fifty seeds of the WRKY70over-expression line S55 (upper panel), the vector control pCP60 (middle panel) and JA-insensitive mutant coi1-16 (bottom panel) were germinated on MS plates containing different concentrations of MeJA as indicated. Representative seedlings were photographed after a 4-day germination period. The experiments were independently performed three times with similar results.
(b) WRKY70-over-expressing lines S27, S48 and S55 (bottom panel) as well as pCP60 (upper panel) were germinated on MS plates containing 2.5 μm MeJA. Representative seedlings were photographed after a 4-day germination period.
(c) Comparison of accumulation of anthocyanins in seedling leaves from different WRKY70-over-expressing lines. Seeds were germinated on MS plates containing 20 μm MeJA, and 7-day-old seedlings were divided into three samples. Each sample contained seedling leaves from 50 1-week-old plants, and anthocyanins were individually extracted. The amount of anthocyanins in each sample was normalized as a percentage of the vector control. The data represent the average of three replicates ±SD.
(d) Seeds of wild-type (WT, upper panel), wrky70-1 (middle panel) and wrky70-2 (bottom panel) were germinated on MS plates containing 1 μm MeJA. Three-day-old seedlings were photographed.
(e) Comparison of anthocyanin accumulation in wrky70 mutant seedlings. For each sample, 100 seedlings were collected after a 3-day germination period.

Enhanced disease susceptibility to bacterial pathogens in the wrky70-1 mutant

As both wrky70-1 and wrky70-2 mutants appeared phenotypically similar, wrky70-1 was employed to further survey the consequences of loss-of-function phenotypes in subsequent experiments. The decrease in resistance to the bacterial necrotroph Ecc SCC1 observed earlier in WRKY70 antisense plants (Li et al., 2004) was also evident in the wrky70-1 mutant (Table 1). Similar to the previous observation using antisense suppression of WRKY70 (Li et al., 2004), the loss of WRKY70 function in wrky70-1 did not lead to a substantially altered resistance to Pst DC3000 (Table 1).

Table 1.  The effect of WRKY70 on plant resistance to Ecc (Erwinia carotovora subsp. carotovora) SCC1 and Pst (Pseudomona syringae pv tomato) DC3000
  1. aResistance of WRKY70 T-DNA insertional mutants to the virulent bacterial pathogen Ecc SCC1. Four-week-old soil-grown plants were surface-inoculated locally with 10 μl drops of SCC1 suspension (106 cells ml−1 in 0.9% NaCl). The survival of the plants after 1 week of infection was examined. Each experiment consisted of 28 individual plants from each line. The data represent the mean values from three independent experiments ±SD.

  2. bGrowth of Pst DC3000 in planta. Each fully expanded leaf from pCP60 and wrky70-1 was infiltrated with 10 μl suspension of Pst DC3000 (106 cells ml−1 in 10 mm MgCl2), and samples were taken 2 days after infection. Data represent the means of three pools derived from 10 individual plants ±SD. The experiments were repeated twice with similar results. CFU, colony-forming units.

Survival (%)a21.4 ± 3.60
CFU per leafb(9.89 ± 1.59) × 107(1.52 ± 0.40) × 108

Insertional inactivation or over-expression of WRKY70 cause opposite resistance phenotypes to Erysiphe cichoracearum

Our previous work and the data presented above indicate that WRKY70 controls plant resistance to bacterial pathogens and suggest that this could be accomplished by altering the balance between SA- and JA-dependent defenses (Li et al., 2004). To elucidate the generality of this control and whether it also applied to other types of pathogens, we employed the fungal biotroph E. cichoracearum. Arabidopsis thaliana ecotype Columbia (Col-0) is susceptible to this fungus, developing macroscopic disease symptoms of powdery mildew (Adam and Somerville, 1996). The role of SA-dependent defenses in limiting growth of this pathogen has been well documented (Dewdney et al., 2000; Xiao et al., 2001). Both WRKY70-over-expressing plants and the wrky70-1 mutant were used to assess the contribution of WRKY70 to Arabidopsis resistance to the powdery mildew E. cichoracearum strain UCSC1. Eight days after inoculation with asexual spores of E. cichoracearum, vector control plants displayed typical disease symptoms with abundant conidiophores on mature leaves combined with some chlorosis, but no necrosis (Figure 3a). These symptoms were drastically enhanced in the wrky70-1 insertion mutant, which displayed a clearly increased susceptibility to E. cichoracearum. Approximately 30% of leaves became chlorotic, and about 20% of theses developed extensive necrosis and tissue collapse. About 10% of leaves were completely consumed by the spreading necrosis (Figure 3a). In contrast, WRKY70-over-expressing plants were highly resistant to E. cichoracearum UCSC1 (Figure 3a). The differences in disease development between WRKY70 over-expressors, vector control and wrky70-1 plants were scored 10 days after inoculation (Figure 3b) according to the standard protocol (Weigel and Glazebrook, 2002). Taken together, these results show that WRKY70 is required for Arabidopsis resistance to E. cichoracearum and are consistent with the requirement for SA-dependent defenses in this resistance.

Figure 3.

WRKY70 expression is required for resistance of Arabidopsis to the obligate biotroph E. cichoracearum.
(a) Disease symptoms in control (pCP60, bottom panel), WRKY70 over-expressors (S55, middle panel) and knockout plants (wrky70-1, upper panel). Plants were sprayed with an asexual spore suspension of E. cichoracearum UCSC1 (5 × 105 spores per ml H2O), moved to transparent containers and kept at 100% relative humidity. Representative leaves were photographed 8 days after inoculation.
(b) Disease rating was scored on 72 plants of each genotype 10 days after inoculation according to a standard protocol (Weigel and Glazebrook, 2002), where the range from 0 to 5 is an indicator of conidiophore presence (0, no conidiophores; 5, more than 20 conidiophores on infected leaves). The experiments were independently repeated twice with similar results. dpi, days post-inoculation.

WRKY70 promotes susceptibility to Alternaria brassicicola

In contrast to E. cichoracearum, JA but not SA signaling is essential for Arabidopsis resistance to the necrotrophic fungal pathogen A. brassicicola. The JA-insensitive coi1-1 mutant is susceptible to this necrotrophic fungus, whereas plants impaired in accumulation of SA or SA-mediated signal transduction retain strong resistance (Thomma et al., 1998). Therefore, elucidating the potential contribution of WRKY70 to A. brassicicola resistance or susceptibility could define whether WRKY70 is involved in the mutual antagonism between SA and JA signaling. With this aim, we first determined the susceptibility of WRKY70-over-expressing plants to A. brassicicola. When challenged with this fungus, 4-week-old control plants were essentially resistant and showed only non-spreading local necrotic lesions at the inoculation sites. In contrast, spreading necrosis and more severe disease symptoms were observed in plants (line S55) over-expressing WRKY70 as shown in Figure 4(a). The average diameter of lesions caused by A. brassicicola was determined 6 days after inoculation and found to be at least fivefold larger than in control plants. To verify that the more severe disease symptoms were the direct consequence of increased fungal colonization, we estimated the difference of in fungal biomass between WRKY70-over-expressing and control plants using real-time quantitative PCR with gene-specific primers for A. brassicicola. As shown in Figure 4(b), there was a more than 100-fold difference in fungal biomass between the over-expressing line S55 and the control plants. Similar results were also obtained in two additional over-expression lines tested (S27 and S48, data not shown). The compromised resistance to A. brassicicola in WRKY70-over-expressing lines suggests that WRKY70 might suppress the JA-induced defenses that are essential for plant resistance to A. brassicicola.

Figure 4.

WRKY70 enhances susceptibility to A. brassicicola.
(a) Four-week-old Arabidopsis plants were inoculated with 5 μl spore suspension of A. brassicicola (5 × 105 spores per ml H2O). The images show lesion formation 6 days after inoculation. The smaller image on the right side is a 25× amplification of the framed area.
(b) Quantification of disease symptoms by determining the relative amount of fungal DNA. Forty-five inoculated leaves were pooled for collecting total DNA. Values represent the means ± SE of six measurements for each sample.
(c) wrky70 mutants show increased resistance to A. brassicicola. Two leaves of 4-week-old axenic Arabidopsis plants were inoculated with A. brassicicola (5 × 105 spores per ml H2O) 24 h after pre-treatment with H2O or SA (final concentration in MS medium: 450 μm). Seven days post-inoculation, fungal growth was measured by quantitative real-time PCR. Inoculated leaves of six plants were pooled for collecting total DNA. Values represent the means ± SE of six measurements for each sample.

To confirm the role of WRKY70 in modulating JA-dependent defense pathways, we assessed whether compromising JA-dependent defense by exogenous application of SA (Kariola et al., 2005) was altered in wrky70 mutants. To this end, axenically grown wrky70-1 and wild-type plants were treated with SA 1 day before inoculation with this pathogen. As shown in Figure 4(c), there was a substantial increase in the susceptibility in the vector control following SA treatment. This increase was partly blocked by the wrky70-1 mutation, with the mutants being more resistant to A. brassicicola than the vector control plants. These results indicate that WRKY70 is required for SA-controlled suppression of JA-mediated defense against A. brassicicola, and suggest that WRKY70 can promote plant susceptibility to this fungus.

WRKY70 represses biosynthesis of the phytoalexin camalexin and indol-3-ylmethyl glucosinolate (IGS)

Arabidopsis resistance to A. brassicicola requires camalexin biosynthesis (Thomma et al., 1999; van Wees et al., 2003). To examine whether the compromised A. brassicicola resistance of WRKY70 over-expressors could be due to impaired production of camalexin, we compared the camalexin contents of over-expressors 48 h after A. brassicicola inoculation with those of control plants. Intriguingly, the camalexin levels in WRKY70 over-expressors were higher than those in control plants (Figure 5a). This apparent contradiction could be due to the increased camalexin accumulation as a consequence of the more severe colonization by A. brassicicola, as seen in the JA-insensitive mutant coi1-1 (Thomma et al., 1999). To resolve this dilemma, we assessed the effect of WRKY70 on silver nitrate-induced camalexin accumulation. Significantly, this chemical induction of camalexin biosynthesis was suppressed in WRKY70-over-expressing lines, and the accumulation of camalexin in over-expressors displayed a clear reduction of 66% after 24 h (Figure 5b).

Figure 5.

Production of camalexin and indol-3-ylmethyl glucosinolate following pathogen infection or chemical treatments.
(a) Camalexin accumulation 48 h after A. brassicicola inoculation.
(b) Camalexin accumulation 24 h after silver nitrate treatment.
(c) Induction of indol-3-ylmethyl glucosinolate 48 h after spraying with 100 μm MeJA.
In each case the contents of the defense metabolites were determined within six pooled plants. The data shown are the means ± SE of six measurements. All experiments are repeated at least twice with similar results. FW, fresh weight of tissue.

Glucosinolates are implicated in plant defense (Wittstock and Halkier, 2002), and accumulating evidence suggests that these secondary metabolites have a defensive role in the resistance of Arabidopsis to A. brassicicola (Schenk et al., 2003). Earlier reports demonstrated a JA-dependent induction of biosynthesis of IGS in Arabidopsis (Brader et al., 2001; Mikkelsen et al., 2003). To examine the effect of WRKY70 on regulation of IGS, we assayed IGS contents in the WRKY70-over-expressing line S55 and control plants following MeJA treatment. A substantial reduction in MeJA-induced accumulation of IGS was detected in the WRKY70-over-expressing line (Figure 5c), indicating that WRKY70 negatively regulates JA-induced production of IGS.

WRKY70 suppresses induction of JA-responsive genes

To explore the molecular mechanism of the altered JA sensitivity of WRKY70 mutants and over-expressors, we first characterized the effect of WRKY70 on the selected JA-responsive markers PDF1.2 and AtCLH1 (AtCOR1). As shown in Figure 6, PDF1.2 expression is observed in control plants induced by exogenous MeJA, elicitors derived from two E. carotovora strains (SCC1 and SCC3193) and A. brassicicola infection. Consistent with the previous data (Li et al., 2004), MeJA-induced expression of PDF1.2 was drastically reduced in WRKY70 over-expressors (Figure 6a) while it was enhanced in the wrky70-1 mutant (Figure 6c). Over-expression of WRKY70 essentially abolished both the elicitor-induced and A. brassicicola-induced expression of PDF1.2 (Figures 6a,b and 7a). In contrast, the accumulation of PDF1.2 transcripts after both MeJA treatment and A. brassicicola infection was clearly increased in the wrky70-1 mutant (Figures 6c and 7b). These results indicate that WRKY70 is a repressor of PDF1.2.

Figure 6.

WRKY70-mediated suppression of JA-responsive genes.
(a) Northern blot showing MeJA or CF induction of PDF1.2 in the vector control pCP60 and the WRKY70 over-expression line S55. CFSCC1 and CF3193 culture filtrates from the two E. carotovora subsp. carotovora strains SCC1 and SCC3193 and MeJA (50 μm) were applied as four 5 μl drops on leaves.
(b) A. brassicicola induction of PDF1.2 in WRKY70 over-expression line S55 (see Figure 4 for fungal inoculation).
(c) PDF1.2 expression in the wrky70-1 mutant.
(d) SA-mediated suppression of the basal expression of AtCLH1 is reduced in wrky70-1 mutants.
All samples represent 20 μg of total RNA from six individual plants harvested at each designated time point. Ethidium bromide staining of rRNAs is shown to demonstrate equal loading. hpt, hours post-inoculation. *, 1 h exposure time. #, 4 h exposure time.

Figure 7.

MeJA- and pathogen-induced expression of JA-responsive markers and JA production following pathogen infection in different genotypes.
(a) Induction of PDF1.2 and AtCHL1 in different genotypes after spraying with 50 μm MeJA or A. brassicicola inoculation (see Figure 4 for fungal inoculation). Inoculated leaves from six individual plants were harvested and pooled at indicated time points. Total RNA was probed with gene specific probes. All images for the same gene in different genotypes derive from a single blot. Ethidium bromide staining of rRNAs indicates the loading.
(b) Northern blots of 20 μg total RNA showing accumulation of AtCHL1 in the wild-type and in the wrky70-1 mutant. To ensure equal loading, the RNA gel was stained with ethidium bromide.
(c) Time course of endogenous JA content after inoculation with A. brassicicola. JA levels in leaves collected 0 h and 1, 2 and 3 days after inoculation were determined by GC-MS. Each value represents the average of six replicates ±SE. The experiments were repeated once with similar results. FW, fresh weight of tissue.

Similarly, both the basal and induced expression of AtCLH1 was altered by insertional inactivation or over-expression of WRKY70 (Figures 6d and 7a,b). The basal expression of AtCLH1 was clearly elevated in the wrky70-1 mutant (Figures 6d and 7b), in agreement with the previous results using antisense-silenced plants (Li et al., 2004), suggesting that WRKY70 indeed down-regulates AtCLH1. Moreover, the SA-induced suppression of AtCLH1 appears to be mediated by WRKY70 and is relieved in the wrky70-1 mutant (Figure 6d). In agreement with this notion, a reduction in AtCLH1 expression was observed in WRKY70 over-expressors following MeJA application or A. brassicicola infection (Figure 7a).

To identify putative downstream factors involved in WRKY70 suppression of PDF1.2, we used RNA gel blot analysis to characterize the MeJA- or A. brassicicola-induced expression of PDF1.2 in NahG and npr1-1 plants over-expressing the WRKY70 transgene. As expected, the pathogen-induced expression of PDF1.2 is enhanced in NahG and npr1-1 plants when compared to vector control lines (Figure 7a). Interestingly, the robust suppression of MeJA- or pathogen-induced expression of PDF1.2 in WRKY70 over-expressors (S55) is partially relieved in both S55/NahG and S55/npr1-1 plants (Figure 7a). Suppression of MeJA- or pathogen-induced expression of AtCLH1 is also evident in WRKY70 over-expressors, albeit not as strongly as with PDF1.2 (Figure 7a). This suppression was also partially or even completely relieved by introduction of NahG or npr1-1 into the S55 background (Figure 7a).

To explore whether the relief from WRKY70 suppression could have been caused by altered JA levels, we determined the JA content in all genotypes following A. brassicicola infection (Figure 7c). The JA levels were responsive to pathogen infection in all genotypes, but particularly in NahG where the pathogen-stimulated JA production was more pronounced compared to control plants during a 3-day period (Figure 7c). In npr1-1-containing lines, the JA levels were even slightly lower than in controls and could not explain the increased expression of PDF1.2 or AtCHL1 in these backgrounds (Figure 7c). Furthermore, A. brassicicola infection did not substantially alter the production of endogenous SA in the tested lines (data not shown). Taken together, these results suggest that over-expression of WRKY70 suppresses basal and induced expression of JA-responsive genes, and this suppression is at least partly mediated by NPR1, but does not appear to involve major changes in pathogen-induced JA synthesis.

A subset of MeJA- and pathogen-inducible genes is controlled by WRKY70

The suppression of JA-dependent disease resistance and expression of two JA-induced marker genes by WRKY70 prompted us to explore the effect of WRKY70 on other JA-controlled genes. We used a cDNA macroarray containing 379 Arabidopsis expressed sequence tags (ESTs) with a bias toward putative defense-related and regulatory genes. We detected a total of 23 genes that showed consistent MeJA induction in independent experiments (Tables 2 and 3). These genes could be divided into two categories based on their sensitivity to WRKY70-mediated suppression. The first group includes nine genes, for which MeJA induction was neither affected in the WRKY70 over-expression background nor in the wrky70-1 mutant (Tables 2 and 3), suggesting that this subset of genes is not controlled by WRKY70. Interestingly, this group contains the gene AOS encoding allene oxide synthase that catalyzes the entrance reaction in JA biosynthesis, as well as the ASA1 gene encoding the alpha subunit of anthranilate synthase involved in biosynthesis of tryptophan (Trp) and Trp-derived metabolites. The second group consisted of 12 (Table 2) or 13 (Table 3) genes. MeJA induction of these genes was substantially reduced or blocked in WRKY70 over-expressors, whereas both background and MeJA-induced expression of this subset of genes (except for At1g56250) either remained unchanged or was enhanced in the wrky70-1 mutant background compared to wild-type plants (Table 2). These results confirm that WRKY70 is a negative regulator of a subset of JA-responsive genes. This WRKY70-mediated suppression was relieved by introduction of the npr1-1 mutation except in At4g21580 (Table 3). A similar relief of WRKY70 suppression was also observed for many but not all genes in S55/NahG plants (Table 3). These data suggest the existence of two types of JA-responsive genes: genes that are WRKY70-suppressed and genes that are insensitive to this suppression. Moreover, WRKY70-mediated suppression of most of the JA-responsive genes appears to require functional NPR1.

Table 2.  Effect of WRKY70 on MeJA- or A. brassicicola-induced gene expression
Locus identifierA. brassicicola MeJAGene function
0 h24 h48 h8 h0 h24 h48 h8 h0 h24 h48 h8 h
  1. Leaves of six plants of wild-type, WRKY70 over-expressors (line S55) and wrky70-1 mutants were collected 0, 24 and 48 h post-inoculation with A. brassicicola or 8 h after MeJA treatment. Expression of each gene is indicated as fold induction compared to wild-type (0 h) set to 1.0.

Group I (MeJA induction not affected by WRKY70)
 At5g210901.010.07.827.63.610.28.933. repeat protein like
 At5g426501. oxide synthase AOS
 At2g371301. ATP2a
 At1g735001. MAP kinase
 At5g057301. synthase alpha ss ASA1
 At4g218501. domain-containing protein
 At3g109201. dismutase MSD1
 At1g113101. protein Mlo2
 At3g059701. long-chain-fatty-acid-CoA ligase
Group II (MeJA induction suppressed by WRKY70)
 At2g044001.015.39.716. phosphate synthase
 At3g164201. myrosinase-binding protein
 At5g096601. malate dehydrogenase
 At3g546401. synthase alpha ss TSA1
 At1g241001. S-GT
 At5g247801. storage protein VSP1
 At4g020801. protein Sar1 homolog
 At1g562501. interacting partner 3- related protein
 At5g139301. synthase
 At4g392601. RNA-binding protein
 At1g050101. oxidase EAT1
 At1g732601. trypsin inhibitor
Group III (not induced by MeJA)
 At4g341501. domain-containing protein
 At2g411001. protein TCH3
 At3g443201. 3
 At1g315801. to pathogen-inducible protein cxc750 precursor
 At2g324401. acid hydroxylase
 At4g031901. protein, F-box protein with LRR
 At2g021201. proteinase inhibitor gamma-thionin family
 At2g238101. senescence-associated protein 5
 At2g029901. RNS1
Table 3.  Effect of WRKY70 on MeJA-induced gene expression
Locus identifierMeJA/controlGene function
  1. Leaves from six plants of the indicated genotypes were taken 8 h after spraying with MeJA. Expression data is given as fold induction compared to mock-treated (control) plants.

Group I (MeJA-inducible, not suppressed by WRKY70)
 At5g210909. repeat protein like
 At5g426508. oxide synthase AOS
 At2g371306. ATP2a
 At1g735003. MAP kinase
 At5g057303. synthase alpha ss ASA1
 At4g218502. domain-containing protein
 At3g109202. dismutase MSD1
 At1g113102. protein Mlo2
 At3g059702. long-chain-fatty-acid–CoA ligase
Group II (MeJA-inducible, WRKY70-suppressed)
 At2g04400111.11.558.339.827.413.4Indole-3-glycerol phosphate synthase
 At3g1642066.60.61.825.83.529.9Putative myrosinase binding protein
 At5g0966062. malate dehydrogenase
 At3g5464038.40.612.018.712.718.0Tryptophan synthase alpha ss TSA1
 At1g241008.11.23.511.25.32.3UDP-glucosyltransferase S-GT
 At4g020805. binding protein Sar1 homolog
 At1g562504. interacting partner 3-related protein
 At5g139304.10.82.810.52.78.6Chalcone synthase
 At4g392604. RNA binding protein
 At4g215804. dehydrogenase
 At1g050103. oxidase (=EAT1)
 At1g732603. trypsin/proteinase inhibitor
 At1g659302. Isocitrate dehydrogenase (NADP+)

To identify potential candidate genes controlling resistance to A. brassicicola that were suppressed by WRKY70 over-expression, we characterized the expression profiles following this fungal infection. Of the 379 genes tested, 30 showed over twofold induction by A. brassicicola in wild-type, WRKY70 over-expressors or wrky70-1 mutant backgrounds (Table 2). Most of the MeJA-induced genes identified (groups I and II in Tables 2 and 3) were also activated by this fungal pathogen in all three backgrounds. The MeJA-inducible genes that were not suppressed by WRKY70 were similarly induced by Alternaria in control and WRKY70-over-expressing plants and wrky70-1 mutants (group I in Table 3), confirming that this subset of JA-responsive genes is not controlled by WRKY70. In contrast, many but not all of the MeJA-inducible genes that were suppressed by WRKY70 also showed suppression of Alternaria induction in the WRKY70-over-expressing plants but not in wrky70-1 mutants. This suggests that WRKY70-mediated suppression of these genes could contribute to the compromised resistance of WRKY70 over-expressors to A. brassicicola. In addition, A. brassicicola induced expression of additional genes that were not responsive to MeJA (group III in Table 2). None of these were suppressed by WRKY70 nor affected by the wrky70-1 mutation.

NPR1 is required for the enhanced susceptibility of WRKY70 over-expressors to A. brassicicola

WRKY70-mediated suppression of JA-induced defense genes was partly relieved in the npr1-1 background, prompting us to characterize the possible role of NPR1 in WRKY70-determined susceptibility to A. brassicicola. To elucidate the NPR1 dependence of this enhanced susceptibility in WRKY70 over-expressors, we assessed A. brassicicola resistance of npr1-1 plants carrying the WRKY70 transgene. The enhanced susceptibility to A. brassicicola due to WRKY70 over-expression was essentially lost in these lines (Figure 4a,b). Of all tested leaves from the S55/npr1-1 genotype, around 85% produced restricted necrosis as observed in control or npr1-1 plants, whereas only roughly 15% exhibited slightly spreading but limited lesions (Figure 4a). This is in clear contrast to the spreading necrotic lesions observed in WRKY70 over-expressor lines. The results were confirmed by measuring the total fungal biomass in all infected leaves, which revealed only slightly increased colonization in S55/npr1-1 plants, similar to control and npr1-1 plants and in contrast to the S55 plants (Figure 4b). Taken together, these results indicate that WRKY70-mediated suppression of A. brassicicola resistance is mainly through an NPR1-dependent mechanism.


WRKY70, a unique member of the WRKY transcription factor family, has been identified as an important regulatory component in plant defense (Li et al., 2004). Our previous data show that WRKY70 is an activator of SA-dependent defense genes and appears to repress JA-regulated genes. These data suggest that WRKY70 might regulate the balance between JA- and SA-mediated defense and fine-tune the plant resistance response to effectively combat specific pathogens. In this paper, we employed two fungal pathogens A. brassicicola and E. cichoracearum with different pathogenesis strategies to further elucidate the function of WRKY70 in integrating signals from both JA- and SA-mediated defense pathways.

Evidence reported here shows that gain or loss of WRKY70 function by over-expression or insertional inactivation causes opposite effects on Arabidopsis resistance to the obligate biotroph E. cichoracearum and the necrotroph A. brassicicola. WRKY70 over-expressors are significantly less sensitive to the biotrophic pathogen, whereas an insertional mutation in WRKY70 locus results in enhanced susceptibility (Figure 3). We have previously demonstrated that WRKY70 over-expression causes enhanced resistance to the bacterial necrotroph E. c. carotovora and the hemi-biotroph Pseudomonas syringae pv tomato DC3000 (Li et al., 2004). Resistance to both of these virulent pathogens can be triggered by SA-mediated mechanisms (Kariola et al., 2003, 2005; Nawrath and Métraux, 1999; Palva et al., 1994). Similar to these bacterial pathogens, A. thaliana eco-type Col-0 is susceptible to E. cichoracearum (Adam and Somerville, 1996). The involvement of SA signaling in response to E. cichoracearum infection has been well documented. For instance, deficiency in SA accumulation results in increased susceptibility to E. cichoracearum (Dewdney et al., 2000). In contrast, transferring the R-gene PRW8 from the resistant ecotype Ms-0 confers gene-for-gene resistance through SA-dependent mechanisms in Col-0 wild type (Xiao et al., 2001). Our data showing that WRKY70 promotes resistance to E. cichoracearum are consistent with and support the previous notion that WRKY70 is a key positive regulator of the SA-dependent plant disease resistance response.

In contrast to resistance to E. cichoracearum, SA signaling is not required for Arabidopsis resistance to the necrotroph A. brassicicola; neither SA-deprived transgenic plants (NahG) nor SA-insensitive mutants (npr1) exhibit enhanced susceptibility (Thomma et al., 1998), and application of SA even increases susceptibility to A. brassicicola (Kariola et al., 2005). In contrast, JA signaling is essential for Arabidopsis resistance to this fungal necrotroph (Thomma et al., 1998). Our data demonstrate that this resistance is severely compromised by over-expression of WRKY70 (Figure 4a,b) but enhanced in the wrky70 mutant background (Figure 4c), indicating that WRKY70 has the capability to suppress JA-mediated defenses. Furthermore, the SA-compromised and JA-dependent resistance to A. brassicicola is less affected in wrky70 mutants (Figure 4c). Suppression of JA signaling by WRKY70 was not limited to defense pathways but was also evident in other JA-controlled processes such as JA-induced inhibition of root growth and accumulation of anthocyanins. For example, accumulation of anthocyanins was markedly reduced in WRKY70 over-expressors and significantly increased in wrky70 mutants (Figure 2). Our results show that WRKY70 down-regulates both JA- and pathogen (Alternaria)-induced genes such as the JA-responsive molecular markers PDF1.2 and AtCLH1 (Figures 6a–c and 7a,b).

The WRKY70-mediated suppression of JA-dependent defense pathways was further confirmed by macroarray analysis of a number of defense-related genes. Induction of the majority of the JA- and/or Alternaria-responsive genes on these arrays was blocked in WRKY70 over-expressors but showed enhanced responsiveness in wrky70 mutants (Tables 2 and 3). Several of the WRKY70-suppressed genes have potential roles in plant resistance to A. brassicicola. For example, in agreement with the reduced induction of two particular genes (At2g04400 and At3g54640) involved in Trp biosynthesis, we observed that WRKY70 suppressed induction of the corresponding Trp-derived defense compounds camalexin and IGS in WRKY70 over-expressors (Figure 5b,c) (Brader et al., 2001; Glawischnig et al., 2004). Our results argue that the transcription factor WRKY70 represses JA-dependent defenses and acts as a negative regulator to JA-dependent disease resistance. We propose that down-regulation of a specific set of A. brassicicola-inducible, JA-responsive genes by WRKY70 is responsible for the enhanced susceptibility to this pathogen in WRKY70 over-expressors.

How is the suppression by WRKY70 of JA-dependent defenses executed? Our previous data (Li et al., 2004) and results presented here indicate that WRKY70-mediated suppression of the JA-dependent defense pathway is not executed through inhibiting JA biosynthesis. Indeed, both MeJA and pathogen induction of AOS, encoding allene oxide synthase, which catalyzes the conversion of 13-hydroperoxylinolenic acid to 12,13-epoxy-octadecatrienoic acid in the octadecanoid pathway, is not affected by WRKY70 over-expression or mutation (Tables 2 and 3). The same expression pattern of this gene in wild-type plants, over-expressors and wrky70-1 mutants was further confirmed by Northern blotting (data not shown). Moreover, JA levels are also induced in WRKY70 over-expressors after A. brassicicola infection (Figure 7c). Consequently, our results argue that WRKY70 mainly affects JA signaling. This suppression of JA signaling is not due to constitutively elevated SA levels, as WRKY70 over-expressor plants have similar levels of free SA as vector control and wild-type plants and even a reduced content of conjugated SA (Li et al., 2004). However, our results demonstrate that SA is required for WRKY70-mediated suppression of MeJA-responsive genes, as evidenced by relief of this suppression in the NahG background (Figure 7a and Table 3).

Our results show that the compromised A. brassicicola resistance of WRKY70 over-expressors is largely abolished by introduction of the npr1-1 mutation (Figure 4), suggesting a potential involvement of the functional NPR1 protein in WRKY70-mediated suppression of JA-dependent defenses. In support of this hypothesis, our data show that JA induction of a subset of WRKY70-repressed genes is either fully or partially relieved in the S55/npr1-1 background (Table 3 and Figure 7a). This would suggest that WRKY70-mediated suppression is probably either completely or partially controlled by NPR1. At least seven genes (group II in Table 3) exhibited similar or even increased JA induction in S55/npr1-1 plants in comparison to npr1-1 mutants, indicating that WRKY70-mediated suppression of this type of genes might be fully NPR1-dependent.

How is WRKY70 suppression of JA-responsive genes through NPR1 executed? While nuclear localization of NPR1 is important for SA-dependent PR gene induction, a cytosolic function of NPR1 has been implicated in mediating suppression of JA-dependent genes such as AtVSP and PDF1.2 (Spoel et al., 2003). Dong (2004) has recently proposed two possible modes of action for the NPR1-modulated suppression, either through blocking nuclear translocation of a positive regulator of JA signaling or through activating a negative regulator of JA signaling. Possibly, WRKY70 could trigger these negative regulators of JA signaling, subsequently activated by cytosolic NPR1. Alternatively, WRKY70 might induce the expression of factors that block positive regulators of JA signaling in concert with cytosolic NPR1.

In conclusion, the differential modulation of resistance by WRKY70 to distinct fungal pathogens provides strong evidence that WRKY70 controls the balance between JA- and SA-dependent defense pathways. Distinct defense responses can be optimized in the plant through the correct adjustment of expression of WRKY70 depending on invading pathogens.

Experimental procedures

Biological material and infections with fungi

WRKY70 transgenic lines and vector controls were grown as described previously (Li et al., 2004). Confirmation of the localization of the T-DNA insertion of the SALK line (accession number SALK_025198) and the selection of homozygous lines was performed with the gene-specific primers 5′-TAAGATACCACTCACCAAAAACTTCCTCAA-3′ and 5′-CTCATGGTCTTAGTCCTAATGTAGTGGT-3′, as well as an LB primer from pROK2 (5′-AGTTGCAGCAAGCGGTCCACGC-3′). Alternaria brassicicola, strain CBSnr 567.77, was obtained from the Centraalbureau voor Schimmelcultures, Utrecht, The Netherlands. The fungus was cultivated on potato carrot extract agar (PCA) plates. After a growth period of at least 10 days, spores were washed from the plates with water, and the concentration was adjusted to 5 × 105 ml−1 after counting the spores microscopically. To monitor the susceptibility to this fungal pathogen, the 5th to 9th true leaves of 4-week-old soil-grown plants were infected with 5 μl spore suspension after making a small wound with a pipette tip. Inoculated plants were incubated at 23°C at 100% relative humidity in trays covered with a transparent polystyrene lid for the time periods indicated. To analyze gene expression, JA content, camalexin levels and the content of indole-3-ylmethyl glucosinolate, four 5 μl drops of the spore suspension were placed on fully expanded leaves. Strain UCSC1 of Erysiphe cichoracearum was obtained from Hans Thordal-Christensen (Risø National Laboratory, Roskilde, Denmark). Asexual spores of E. cichoracearum from infected squash leaves were suspended in water, adjusted to 5 × 105 spores per ml, and 2-week-old plants were infected by spraying, moved to transparent containers and kept at 100% relative humidity.

Bacterial inoculation and treatments with MeJA and Erwinia elicitors

The Erwinia carotovora subsp. carotovora strain SCC1 was cultured overnight at 28°C in LB medium. To evaluate bacterial virulence in planta, plants were inoculated as described previously (Li et al., 2004). Preparation of Erwinia elicitors was performed as described by Vidal et al. (1998). Four 5 μl drops of Erwinia elicitors were placed on fully expanded leaves (four leaves per plant). As indicated in the figures, MeJA treatments were performed with 50 μm MeJA (Sigma, Cat 392707) in 0.1% v/v ethanol in the same manner as CF treatments (Li et al., 2004) or plants were sprayed with 100 μm MeJA. All MeJA-treated plants were immediately covered with a transparent lid.

RNA gel blot analysis

Total RNA extraction and RNA gel blot analysis were performed as previously described (Li et al., 2004). The gene-specific probes for AtCHL1, PDF1.2 and WRKY70 were synthesized as previously described (Li et al., 2004).

Quantitative PCR assay for the determination of fungal biomass

Fungal DNA levels were determined on the seventh day of infection relative to the Arabidopsis DNA levels by quantitative PCR using primers for the genomic Alternaria sp. 5.8S ribosomal RNA region (GenBank accession number U05198; 5′-CGGATCTCTTGGTTCTGGCA and 5′-AATGACGCTCGAACAGGCAT) and primers for Arabidopsis genomic Actin2 (5′-CTCCCGCTATGTATGTCGCC and 5′-CGGTTGTACGACCACTGGC). DNA extraction was performed as described by Tierens et al. (2001). Real-time quantitative PCR was performed on an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA, USA), with serial dilutions of DNA of in vitro-grown A. brassicicola and axenic-grown Arabidopsis plants as standards. Each reaction (25 μl) contained 12.5 μl SYBR Green PCR Master mix (Applied Biosystems, Warrington, UK), 0.75 μl of both primers (10 μm) and 100 ng DNA template. Forty cycles of amplification (15 sec at 95°C, 60 sec at 60°C) after an initial 10 min at 95°C were carried out in 96-well optical reaction plates (Applied Biosystems). The relative amount of fungal DNA in relation to plant DNA in a given sample corresponds to 2−ΔΔT with ΔT as the difference in cycle numbers required to reach a given fluorescence threshold level between A. brassicicola and Arabidopsis-specific amplification reactions. −ΔΔT is set to 0 in the average of the reference samples. The combined error ranges from ΔΔT+s to ΔΔT-s where s is the standard error of the Δt values of the reactions of independent samples.

Camalexin and indole gluocosinolate content, and quantification of JA and SA

Camalexin content (Hammerschmidt et al., 1993) and indole gluocosinolate content (Brader et al., 2001) were determined as described previously. JA and SA were extracted and quantified with (±)-9,10-dihydroJA and 13C1-SA as internal standards on a GC-MS (Trace DSQ, Thermo Finnigan, Waltham, MA, USA) using the protocol described by Baldwin et al. (1997).

Extraction and measurement of anthocyanins

One-week-old seedlings were grown in MS plates containing 20 μm MeJA under a 12 h photoperiod and a light intensity of 40 μmol m−2 sec−1. Samples were extracted in a buffer containing isopropanol:HCl:H2O (18:1:81 v/v). Anthocyanins were measured as described previously (Rabino and Mancinelli, 1986). The mean of three parallel samples and their standard deviations were calculated.

Macroarray hybridization

The expression of 379 selected genes was studied with macroarrays using cDNA clones from the Arabidopsis Biological Resource Center (Columbus, Ohio, USA) the Kazusa DNA Research Institute (Japan), and a collection of cDNA clones obtained by subtractive hybridization (Brader et al., 2001). As positive or negative controls, five Caenorhabditis elegans clones (obtained from the National Institute of Genetics, Mishima, Japan) and a Bluescript SK + vector (Stratagene, La Jolla, CA, USA) containing the 35S promoter and nos terminator were used. Inserts were amplified by PCR, purified and sequenced with M13 forward and reverse universal primers. The products (100 ng) were spotted in quadruplets onto positively charged nylon membranes (Genetix, New Milton, Hampshire, UK) with a QPix robot (Genetix), and the DNA was cross-linked to the membranes under UV light. Poly(A)+ RNA was isolated from 30–50 μg total RNA using oligo(dT) magnetic beads (Dynal A.S., Oslo, Sweden) according to the manufacturer's instructions. 32P-labeled probes were prepared from mRNA using a modified protocol of Fedorova et al. (2002). The reaction mixture containing 500 ng mRNA, 1–100 pg of three C. elegans transcripts produced with T3 RNA polymerase (Promega, Madison, WI, USA), 1 μl dNTP mix (2.5 mm dATP, 2.5 mm dGTP, 2.5 mm dTTP, 25 μm dCTP) and 2 μl oligo(dT)16 (50 μm) was incubated at 65°C for 5 min and chilled on ice. After adding 2 μl DTT (100 mm), 20 U RNAse inhibitor (Roche Diagnostics, Basel, Switzerland), 4 μl 5× first-strand buffer, 2.5 μl [α-32P]dCTP (10 mCi ml−1), water (to 19 μl) and 1 μl (200 U) SuperScript II reverse transcriptase (Invitrogen Life Technologies, Carlsbad, CA, USA), the mixture was incubated for 75 min at 42°C, and incubation continued for 30 min after adding 1 μl dCTP (5 mm). The probes were purified on ProbeQuant G-50 Micro Columns (Amersham Biosciences, Piscataway, NJ, USA) and denatured at 95°C (2 min). After wetting the membranes in 2 × SSC and at least 4 h pre-hybridization, the membranes were hybridized at 65°C overnight in a sodium phosphate buffer (Church and Gilbert, 1984) containing 0.5 m sodium phosphate pH 7.2, 7% w/v SDS, 1% w/v BSA, 1 mm EDTA, sheared herring sperm DNA (10 μg ml−1), and the labeled probes at a concentration of 106 cpm ml−1. The membranes were washed at 65°C once in 40 mm sodium phosphate pH 7.2, 5% w/v SDS, 1 mm EDTA, and twice in 40 mm sodium phosphate pH 7.2, 1% w/v SDS, 1 mm EDTA. The macroarray was quantified by using an Imaging plate Bas-MP 2040S (Fuji Photo Film, Tokyo, Japan), a phosphor imager (Bas-1500; Fuji Photo Film) and an imaging analysis program. Gene Spring software (Silicon Genetics, Redwood City, CA, USA) was used for normalization and further analyses. Expression was normalized by calculating the 50% percentile of positive control gene expression.


We wish to thank Yuji Kohara, Center for Genetic Resource Information, National Institute of Genetics, Japan, for the C. elegans EST clones, the Kazusa DNA Research Institute and the Arabidopsis Biological Resource Center for Arabidopsis EST clones, the European Arabidopsis Stock Centre for seeds of the SALK T-DNA insertion line, the Max Planck Institute for Plant Breeding Research for seeds of the GABI-Kat T-DNA insertion line, and Hans Thordal-Christensen for the UCSC1 strain of Erysiphe cichoracearum. This study was supported by the Academy of Finland (projects 79776, 202886; Finnish Centre of Excellence Programme 2000–2005), the Helsinki Graduate School in Biotechnology and Molecular Biology, and Biocentrum Helsinki.