Overexpression of Arabidopsis MAP kinase kinase 7 leads to activation of plant basal and systemic acquired resistance

Authors


(fax +1 352 392 5922; e-mail zhlmou@ufl.edu).

Summary

There is a growing body of evidence indicating that mitogen-activated protein kinase (MAPK) cascades are involved in plant defense responses. Analysis of the completed Arabidopsis thaliana genome sequence has revealed the existence of 20 MAPKs, 10 MAPKKs and 60 MAPKKKs, implying a high level of complexity in MAPK signaling pathways, and making the assignment of gene functions difficult. The MAP kinase kinase 7 (MKK7) gene of Arabidopsis has previously been shown to negatively regulate polar auxin transport. Here we provide evidence that MKK7 positively regulates plant basal and systemic acquired resistance (SAR). The activation-tagged bud1 mutant, in which the expression of MKK7 is increased, accumulates elevated levels of salicylic acid (SA), exhibits constitutive pathogenesis-related (PR) gene expression, and displays enhanced resistance to both Pseudomonas syringae pv. maculicola (Psm) ES4326 and Hyaloperonospora parasitica Noco2. Both PR gene expression and disease resistance of the bud1 plants depend on SA, and partially depend on NPR1. We demonstrate that the constitutive defense response in bud1 plants is a result of the increased expression of MKK7, and requires the kinase activity of the MKK7 protein. We found that expression of the MKK7 gene in wild-type plants is induced by pathogen infection. Reducing mRNA levels of MKK7 by antisense RNA expression not only compromises basal resistance, but also blocks the induction of SAR. Intriguingly, ectopic expression of MKK7 in local tissues induces PR gene expression and resistance to Psm ES4326 in systemic tissues, indicating that activation of MKK7 is sufficient for generating the mobile signal of SAR.

Introduction

Plants, like other multicellular organisms, have innate defense mechanisms to combat microbial pathogens (Jones and Takemoto, 2004). These defense mechanisms function at different levels after the pathogen makes contact with the plant. At the infection site, defense responses are initiated by detecting two general classes of pathogen-derived molecules: pathogen-associated molecular patterns (PAMPs) and effector proteins that are delivered into the plant cell by the type-III secretion system of the pathogen (He, 1998; Nurnberger et al., 2004). Recognition of PAMPs by plant PAMP receptors activates a defense mechanism that is referred to as ‘basal’ defense (Gomez-Gomez et al., 1999; Ron and Avni, 2004). Pathogens can often overcome this initial defense mechanism by delivering effector molecules into the plant cell to interfere with normal cellular functions. However, the presence of some effectors can be detected by host resistance (R) proteins, thereby triggering a defense mechanism known as the hypersensitive response (HR) to limit pathogen growth (Dangl and Jones, 2001; Heath, 2000; Martin et al., 2003). HR at the site of infection can also activate systemic acquired resistance (SAR), which provides protection against a broad spectrum of pathogens throughout the plant (Durrant and Dong, 2004; Ryals et al., 1996).

Systemic acquired resistance requires the signal molecule salicylic acid (SA), which induces the accumulation of pathogenesis-related (PR) proteins (van Loon and van Kammen, 1970; Malamy et al., 1990; Van Loon and Van Strien, 1999). Plants expressing the SA-degrading salicylate hydroxylase (nahG) or mutants deficient in SA synthesis, such as eds5 (also known as sid1) and sid2 (also known as eds16), have impaired SAR (Gaffney et al., 1993; Nawrath and Métraux, 1999; Nawrath et al., 2002; Rogers and Ausubel, 1997; Wildermuth et al., 2001). Genetic screens in Arabidopsis have identified several SAR-related mutants (Durrant and Dong, 2004). Among them, npr1 (also known as nim1 and sai1), which exhibits enhanced susceptibility to a wide range of pathogens such as Pseudomonas syringae pv. maculicola (Psm) ES4326 and Hyaloperonospora parasitica Noco2, was found to be insensitive to SA (Cao et al., 1994; Delaney et al., 1995; Shah et al., 1997). Numerous studies have been conducted to understand the molecular function of NPR1, and to elucidate the signaling pathway downstream of SA (Dong, 2004). However, less is known about the signaling pathway upstream of SA.

There is a growing body of evidence indicating that MAPK cascades are involved in plant defense responses (Innes, 2001; Nakagami et al., 2005; Pedley and Martin, 2005; Zhang and Klessig, 2001). SA and various pathogen-derived elicitors were shown to induce the tobacco mitogen-activated protein kinases (MAPKs), SA-induced protein kinase (SIPK) and wound-induced protein kinase (WIPK) (Zhang and Klessig, 1997). Expression of a constitutively active mutant of NtMEK2, which encodes an MAPK kinase (MAPKK) upstream of SIPK and WIPK, leads to multiple defense responses, including defense gene expression and HR-like cell death (Yang et al., 2001). Both SIPK and WIPK can be activated by the Avr9/Cf-9 interaction (Romeis et al., 1999). Silencing of NPK1, which encodes an MAPKK kinase (MAPKKK), interferes with the function of the disease-resistance (R) genes N, Bs2 and Rx (Jin et al., 2002). Silencing of NTF6/NRK1 (an MAPK) or MEK1/NQK1 (an MAPKK) attenuates N-mediated resistance to tobacco mosaic virus (Liu et al., 2004). Recently, the NbMKK1–NbSIPK cascade was shown to control non-host resistance including HR cell death (Takahashi et al., 2007b).

In tomato, systemin and several oligosaccharide elicitors were shown to activate LeMPK1 and LeMPK2 (Holley et al., 2003). Silencing of genes encoding two MAPKKs (LeMKK2 and LeMKK3) and two MAPKs (LeMPK3 and one similar to Ntf6) compromises Pto-mediated resistance (Ekengren et al., 2003). Both LeMKK2 and LeMKK4 can phosphorylate LeMPK1, LeMPK2 and LeMPK3 in vitro (Pedley and Martin, 2004). Silencing of LeMAP3Kα blocks both avrPto/Pto-mediated HR and disease-associated cell death (del Pozo et al., 2004).

In Arabidopsis, a complete MAPK cascade (MEKK1, MKK4/MKK5 and MPK3/MPK6) and WRKY22/WRKY29 transcription factors were identified to function downstream of the PAMP receptor FLS2, a leucine-rich-repeat (LRR) receptor kinase (Asai et al., 2002). Several laboratories recently reported that MEKK1 is required for flg22- and/or reactive oxygen species (ROS)-induced MPK4 activation (Ichimura et al., 2006; Nakagami et al., 2006; Suarez-Rodriguez et al., 2007). MKK1 was also shown to be involved in flg22-induced activation of MPK4 (Mészáros et al., 2006). MPK4 is a negative regulator of SAR. The mpk4 mutant plants exhibit a constitutive SAR phenotype, including elevated levels of SA, constitutive expression of PR genes and increased resistance to pathogens (Petersen et al., 2000). The MPK4 protein may regulate defense responses by phosphorylation of specific WRKY transcription factors (Andreasson et al., 2005). The MKK3–MPK6 cascade was shown to play a role in jasmonate-dependent negative regulation of ATMYC2/JASMONATE-INSENSITIVE1 (Takahashi et al., 2007a). Additionally, the proteinaceous bacterial elicitor harpin can activate MPK4 and MPK6 (Desikan et al., 2001). Silencing of MPK6 by an intron-containing hairpin loop RNA (ihpRNA) compromises disease resistance (Menke et al., 2004). Phosphorylation of 1-aminocyclopropane-1-carboxylic acid synthase (ACS) by MPK6 induces ethylene biosynthesis (Liu and Zhang, 2004).

We identified a semidominant Arabidopsis activation-tagged mutant, bud1, in which the expression of the MKK7 gene is increased (t307 in Mou et al., 2002; Dai et al., 2006). Previous work has shown that the increased expression of MKK7 in bud1 or the repressed expression in MKK7 antisense transgenic plants causes deficiency or enhancement in auxin transport, indicating that MKK7 negatively regulates polar auxin transport (PAT) (Dai et al., 2006). In this study, we show that the bud1 mutant has an elevated level of SA, and exhibits constitutive PR gene expression and enhanced resistance to both Psm ES4326 and H. parasitica Noco2. Consistently, the expression of MKK7 is induced by pathogen infection in wild-type plants. Silencing of MKK7 by antisense RNA expression not only compromises basal resistance but also blocks the induction of SAR, demonstrating that MKK7 is a positive regulator required for both basal resistance and SAR. Moreover, ectopic expression of MKK7 in local tissues induces PR gene expression and resistance to the bacterial pathogen Psm E4326 in systemic tissues, indicating that MKK7 activation may be involved in generating the mobile signal for SAR.

Results

The bud1 mutant accumulates elevated levels of SA, and exhibits constitutive PR gene expression and enhanced resistance to both bacterial and oomycete pathogens

The bud1 mutant was previously generated using a sense/antisense RNA expression system (t307 in Mou et al., 2002). Increased expression of the MKK7 gene in the activation-tagged bud1 mutant causes deficiency in PAT, which in turn leads to the bushy and dwarf morphology of the bud1 mutant plants (Dai et al., 2006). The morphology of bud1 plants is reminiscent of constitutive defense response mutants such as cpr1, ssi1, and mpk4, which accumulate high levels of SA (Bowling et al., 1994; Petersen et al., 2000; Shah et al., 1999). We therefore measured the concentration of free SA in bud1 plants. As shown in Figure 1(a), bud1 plants exhibited elevated levels of free SA, indicating that BUD1/MKK7 may act upstream of SA.

Figure 1.

 Constitutive defense responses in bud1.
(a) Free salicylic acid (SA) levels in wild-type, cpr1, bud1 heterozygous and bud1 homozygous plants. Four independent leaf samples were collected from each genotype. Each leaf sample was measured once. Data represent the means of the four samples with standard deviation.
(b) Morphology of BGL2:GUS (top left) and bud1 BGL2:GUS plants (top right), and expression of the BGL2:GUS reporter gene in wild-type (bottom left) and bud1 heterozygous (bottom right) plants. Scale bars: 1 cm.
(c) RNA gel blot analysis of the expression of PR1, PR2 and PR5 in wild-type, bud1 heterozygous and bud1 homozygous plants.
(d) Growth of Pseudomonas syringae pv. maculicola (Psm) ES4326 in bud1 heterozygote, wild type and wild type treated with 0.5 mm SA. Data represent the means of eight samples with standard deviation.
(e) Growth of Hyaloperonospora parasitica Noco2 on bud1 heterozygote, wild type and wild type treated with 0.5 mm SA. H. parasitica Noco2 grew significantly less on the bud1 plants than on the wild-type plants (*P < 0.0003). Three leaf samples (each containing 25 leaves from 10 plants) were collected from six pots (8 × 8 cm) of each genotype/treatment. Each leaf sample was counted six times using a hemacytometer. An average value was obtained from the six counts for each sample, and data represent the means of the resulting three values with standard deviation.
The experiment in (a) was repeated twice with similar results, and experiments in (d) and (e) were repeated three times with similar results.

To test whether bud1 displays constitutive defense responses, we crossed a defense response reporter gene containing the BGL2 (β-1,3-glucanase 2; also known as PR2) promoter fused to the GUS coding region into the bud1 mutant background (Bowling et al., 1994). Figure 1(b) shows that the BGL2:GUS reporter gene was constitutively expressed in the bud1 mutant. The molecular marker genes of plant defense responses, PR1, PR2 and PR5, were also constitutively expressed in the bud1 mutant (Figure 1c). We then tested the growth of bacterial pathogen Psm ES4326 in bud1 and wild-type plants. As bud1 homozygous plants are significantly smaller than wild type (Dai et al., 2006), only bud1 heterozygous plants were used for the test. Figure 1(d) shows that bud1 heterozygous plants exhibited enhanced resistance to Psm ES3426. We also tested the growth of the oomycete pathogen H. parasitica Noco2 on bud1 and wild-type seedlings. As bud1 homozygous plants are sterile, a progeny population from bud1 heterozygous plants was used for the test. At the seedling stage, the size of bud1 heterozygous plants was similar to wild type, whereas bud1 homozygous seedlings were smaller than wild type (data not shown). Therefore after H. parasitica Noco2 infection, bud1 heterozygous seedlings were collected to determine the spore numbers of the pathogen. As shown in Figure 1(e), bud1 heterozygous plants also exhibited enhanced resistance to H. parasitica Noco2.

The constitutive defense responses in bud1 plants depend on SA and partially depend on NPR1

The bud1 mutant plants accumulate elevated levels of SA and exhibit constitutive defense responses. To test whether SA signaling plays a role in the constitutive defense response of bud1 plants, we crossed bud1 with the SA-deficient mutant, sid2 (eds16). The bud1sid2 double mutant plants still exhibited bud1 morphology (Figure 2a). However, the bud1-activated PR1 gene expression was completely suppressed by sid2 (Figure 2b). The sid2 mutation also suppressed bud1-conferred resistance to Psm ES4326 (Figure 2c). Therefore, bud1 activates SA-dependent defense response pathways in the mutant plants.

Figure 2.

 Epistasis analysis of bud1.
(a) Morphology of wild-type, bud1 heterozygous, bud1sid2 and bud1npr1 plants. The bud1sid2 and bud1npr1 plants are in a bud1 heterozygote background. Scale bars: 1 cm.
(b) RNA gel blot analysis of the expression of the PR1 gene in the wild-type, bud1 heterozygous, bud1npr1 and bud1sid2 plants.
(c) Growth of Pseudomonas syringae pv. maculicola (Psm) ES4326 in the bud1 heterozygous, bud1npr1 and bud1sid2 plants compared with that in wild-type, npr1 and sid2 plants. Psm ES4326 grew significantly less in the bud1 plants than in the bud1npr1 and bud1sid2 plants (*P < 0.02 and 0.001 respectively). Data represent the means of eight samples with standard deviation. The experiment was repeated three times with similar results.

NPR1 is a key component of the SA-mediated signaling pathway, and the npr1 mutation blocks SA-activated defense responses (Cao et al., 1994). To examine whether npr1 is epistatic to bud1 we generated the bud1npr1 double mutant. The double mutant retained the bud1 morphology (Figure 2a). However, the expression of PR1 and the resistance to Psm ES4326 were partially suppressed (Figure 2b,c). Thus, the bud1 mutation activates both NPR1-dependent and NPR1-independent defense responses.

The constitutive defense responses in bud1 plants is a result of increased expression of MKK7

The morphological phenotype of bud1 plants was reverted to wild type by reducing the mRNA levels of MKK7 in the bud1 plants with a 35S:MKK7 antisense transgene (Dai et al., 2006). We found that the antisense transgene also suppressed the constitutive PR1 expression in bud1 plants (Figure 3a), indicating that MKK7 overexpression is the cause for the mutant phenotype.

Figure 3.

 Overexpressing MKK7 induces defense responses.
(a) RNA gel blot analysis of the expression of the PR1 gene in wild-type, bud1 homozygous and antisense plants. The antisense plants are in a bud1 homozygote background.
(b) Quantitative PCR analysis of the expression of MKK7 and RNA gel blot analysis of the expression of PR1 in wild type, DEXin:MKK7 and pTA7001 vector transgenic plants treated either with or without dexamethasone (DEX). For MKK7, data represent the means of three samples with standard deviation.
(c) Growth of Psm ES4326 in wild-type, DEXin:MKK7,and pTA7001 vector transgenic plants treated either with or without DEX. Psm ES4326 grew significantly less in the DEXin:MKK7 plants treated with DEX than in the mock (0.1% ethanol)-treated DEXin:MKK7 plants (*P < 0.0001). Data represent the means of eight samples with standard deviation.
(d) Growth of Hyaloperonospora parasitica Noco2 on wild-type, DEXin:MKK7 and pTA7001 vector transgenic plants treated either with or without DEX. H. parasitica Noco2 grew significantly less on the DEXin:MKK7 plants treated with DEX than on the mock-treated DEXin:MKK7 plants (*P < 0.0002). Data represent the means of three average values with standard deviation as described in Figure 1(e).
Experiments in (c) and (d) were repeated three times with similar results.

We attempted to recapitulate the constitutive defense response phenotype of bud1 by overexpressing MKK7 in 35S:MKK7 transgenic plants. Unfortunately, most of the seeds produced by the 35S:MKK7 transgenic plants were not viable (Dai et al., 2006). To circumvent this problem, we generated a transgenic line expressing the MKK7 gene under the control of the dexamethasone (DEX)-inducible promoter (DEXin). DEX treatment of the DEXin:MKK7 transgenic plants not only induced the expression of MKK7, but also activated PR1 gene expression and resistance to the bacterial pathogen Psm E4326 and the oomycete pathogen H. parasitica Noco2 (Figure 3b–d). These results indicate that the constitutive defense responses in the bud1 plants are caused by increased expression of MKK7.

The kinase activity of MKK7 is required for the protein to activate defense responses

We have previously shown that the E. coli expressed recombinant myelin basic protein (MBP)-MKK7 protein has in vitro autophosphorylation activity, whereas MBP-mkk7(K74R), in which a conserved Lys residue (K) at the position 74 of the ATP binding site in the kinase domain II was replaced with an Arg residue (R), does not (Dai et al., 2006). The 35S:MKK7 transgenic plants overexpressing a wild-type MKK7 transgene mimicked the bud1 phenotype, whereas the 35S:mkk7(K74R) transgenic plants showed the same morphology as wild type, even though the mkk7(K74R) transgene was highly expressed (Dai et al., 2006). To test whether overexpression of MKK7 in the bud1 mutant increases MAPK activity in vivo, an in-gel kinase activity assay was performed using the MBP as an artificial substrate. As shown in Figure 4(a), overexpression of MKK7 in bud1 plants activated a kinase with a molecular weight (MW) of ∼45 kDa, which is different from the calculated MW of MKK7 (∼34 kDa), suggesting that a new kinase was activated in the bud1 plants. The MKK7 kinase activity in the bud1 plants was not detected in the in-gel kinase activity assay, probably because the kinase activity of wild-type MKK7 is much lower than that of the activated kinase.

Figure 4.

 The kinase activity of MKK7 is essential for activation of defense responses.
(a) Kinase activity in wild-type and bud1 plants (heterozygous) determined by an in-gel kinase activity assay using myelin basic protein (MBP) as a substrate.
(b) RNA gel blot analysis of the expression of MKK7 or mkk7(K74R) and PR1 in DEXin:MKK7 and DEXin:mkk7(K74R) transgenic plants treated either with or without dexamethasone (DEX).
(c) Growth of Pseudomonas syringae pv. maculicola (Psm) ES4326 in DEXin:MKK7 and DEXin:mkk7(K74R) transgenic plants treated either with or without DEX. Psm ES4326 grew significantly less in the DEXin:MKK7 plants treated with DEX than in the mock-treated DEXin:MKK7 plants (*P < 0.0002). Data represent the means of eight samples with standard deviation.
(d) Growth of Hyaloperonospora parasitica Noco2 on DEXin:MKK7 and DEXin:mkk7(K74R) transgenic plants treated either with or without DEX. H. parasitica Noco2 grew significantly less on the DEXin:MKK7 plants treated with DEX than on the mock-treated DEXin:MKK7 plants (*P < 0.0001). Data represent the means of three average values with standard deviation as described in Figure 1(e).
Experiment in (a) and (b) was repeated twice with similar results, and experiments in (c) and (d) were repeated three times with similar results.

To investigate whether the kinase activity of MKK7 is required for activation of defense responses, we first characterized the 35S:mkk7(K74R) transgenic plants. No PR1 gene expression or disease resistance were detected in the 35S:mkk7(K74R) transgenic plants (data not shown). To further confirm that the kinase activity of MKK7 is required for the protein to activate defense responses, we generated transgenic plants containing a DEXin:mkk7(K74R) transgene. In contrast to the DEXin:MKK7 transgenic plants, DEX treatment of the DEXin:mkk7(K74R) transgenic plants did not induce PR1 gene expression and resistance to Psm E4326 and H. parasitica Noco2, even though the mkk7(K74R) transgene was highly induced (Figure 4b–d). Note that the expression levels of the GVG gene in the DEXin:MKK7, DEXin:mkk7(K74R) and in the vector pTA7001 transgenic plants were similar (Figure S1), suggesting that the PR1 gene expression in the DEXin:MKK7 transgenic plants was not caused by expression of the GVG gene (Kang et al., 1999). These results demonstrate that the MKK7 kinase activity is essential for activation of defense responses.

The expression of the MKK7 gene is induced by pathogen infection

Although increased expression of MKK7 confers bud1 constitutive PR gene expression and enhanced disease resistance, this may not reflect the biological function of MKK7 in Arabidopsis. If MKK7 is involved in defense responses, either its expression or activity should be altered during pathogen infection. To test this hypothesis, a 1.694-kb DNA fragment of the MKK7 promoter was amplified from wild-type genomic DNA by PCR, and was fused to the GUS gene to generate MKK7:GUS transgenic plants. As shown in Figure 5(a), expression of the MKK7:GUS reporter gene was induced by infection of an avirulent pathogen P. syringae pv. tomato (Pst) DC3000/avrRpt2. This observation was confirmed by examination of the mRNA levels of MKK7 after Pst DC3000/avrRpt2 infection using both quantitative PCR and RNA gel blot analysis (Figure 5b,c). Interestingly, MKK7:GUS expression appeared to be restricted to the vascular tissues, suggesting that MKK7 may be involved in systemic signaling.

Figure 5.

 Induction of MKK7 by Pst DC3000/avrRpt2 infection.
Samples were taken 12 h after Pseudomonas syringae pv. tomato (Pst) DC3000/avrRpt2 infection.
(a) Histochemical GUS assay of the MKK7:GUS transgenic plants infected with Pst DC3000/avrRpt2, showing the expression pattern of the MKK7:GUS reporter gene. Four independent MKK7:GUS transgenic lines were analyzed and a similar induction pattern was detected in all four lines. Arrowheads indicate the inoculated leaf halves.
(b) Quantitative PCR analysis of the expression of MKK7 after Pst DC3000/avrRpt2 infection. Pst DC3000/avrRpt2 infection significantly induced MKK7 compared with the mock inoculation (10 mm MgCl2) (*P < 0.03). Data represent the means of three samples with standard deviation.
(c) RNA gel blot analysis of the expression of MKK7 after Pst DC3000/avrRpt2 infection. Note that the background expression level of MKK7 in the wild type is higher than in Figure 4(a), because of the 3-day longer exposure time.

To test whether other pathogens could induce MKK7, we infected the MKK7:GUS transgenic plants with H. parasitica Noco2 and Psm ES4326 as well as Pst DC3000/avrRpt2. As shown in Figure S2(a,b), although the avirulent pathogen Pst DC3000/avrRpt2 induced MKK7 expression 8 h after infection, both H. parasitica Noco2 and Psm ES4326 did not induce MKK7. The induction of MKK7 by Pst DC3000/avrRpt2 was further confirmed by quantitative PCR analysis of infected tissues collected at 0, 8, 16 and 24 h post-inoculation (Figure S2c). In addition, we also collected uninfected tissues (systemic tissues) of the plants inoculated with Pst DC3000/avrRpt2 at 24 and 48 h post-inoculation, and subjected the tissues to quantitative PCR analysis. Consistent with the expression pattern revealed in Figure 5(a), MKK7 was not induced in the systemic tissues (data not shown).

Silencing of MKK7 not only compromises basal resistance, but also blocks the induction of SAR after inoculation with avirulent pathogens

As increased expression of MKK7 provides bud1 with resistance to pathogens, loss-of-function mutants of MKK7 may exhibit enhanced susceptibility. We therefore obtained five dSym (SM) transposon lines (Tissier et al., 1999), in which the transposon insertions were shown in the coding region of MKK7 (an intronless gene). We confirmed the transposon insertion sites (Figure S3a and Table S1) and identified homozygous transposon insertion plants using gene-specific primers (Figure S3b). We examined the mRNA levels of MKK7 in the five SM transposon lines using both quantitative PCR and RNA gel blot analysis. The expression level of MKK7 in the five SM transposon insertion lines was similar to that of wild type (Figure S3c). To confirm this result, we performed RT-PCR using three pairs of primers covering different regions of the MKK7 cDNA (Figure S3a). RT-PCR products were detected from all the lines (Figure S3d), and the specificity of the RT-PCR reactions was confirmed by sequencing the RT-PCR products amplified with MKK7F1 and MKK7R1, a pair of primers that cover the coding region of MKK7 (data not shown). Consistent with the mRNA analysis results, no significant difference in defense responses between wild type and these transposon insertion lines was detected (data not shown).

Because we were unable to identify a knock-out mutant of MKK7, we focused on characterization of the previously generated MKK7 antisense lines that are in the wild-type background (Dai et al., 2006). As shown in Figure 6(a), although the expression of MKK9, a homolog of MKK7, was not affected in the antisense plants, the expression of MKK7 was decreased to a lower level compared with that of wild type. As the expression of MKK7 can be induced by Pst DC3000/avrRpt2, we tested the induction of MKK7 in the antisense plants. As shown in Figure S4(a), the induction of MKK7 by Pst DC3000/avrRpt2 was completely blocked in the antisense plants. Additionally, an in-gel kinase activity assay revealed that the kinase activity at ∼45 kDa, which was activated in the bud1 plants (Figure 4a), was decreased to a lower level compared with wild type (Figure S4(b)), suggesting that this kinase activity may be related to MKK7 activity. Although we cannot exclude the possibility of silencing of additional genes, these results indicated that MKK7 was silenced in the antisense plants.

Figure 6.

 Characterization of MKK7 antisense transgenic plants.
Four antisense lines were characterized with similar results, and results from a representative line were presented. Note that the antisense lines are in a wild-type background.
(a) Quantitative PCR analysis of the expression of MKK7 and MKK9 in the antisense plants. Data represent the means of three samples with standard deviation.
(b) Growth of Pseudomonas syringae pv. maculicola (Psm) ES4326 in wild-type, npr1-1 and antisense plants.
(c) Growth of Xanthomonas campestris pv. campestris (Xcc) ATCC33913 in wild-type, npr1-1 and antisense plants. Xcc ATCC33913 grew significantly more in the antisense plants than in the wild-type plants (*P < 0.0001).
(d) RNA gel blot analysis of the expression of PR1 in systemic tissues of wild-type and antisense plants inoculated with P. syringae pv. tomato (Pst) DC3000/avrRpt2 in local tissues.
(e) Growth of Psm ES4326 in systemic tissues of wild-type and antisense plants inoculated with Pst DC3000/avrRpt2 in local tissues. Psm ES4326 grew significantly more in the systemic tissues of the antisense plants than in that of the wild-type plants (*P < 0.006).
Data in (b), (c) and (e) represent the means of eight samples with standard deviation. Experiments in (b), (c), (d) and (e) were repeated three times with similar results.

To test whether silencing of MKK7 compromises basal resistance, we monitored the growth of Psm ES4326 and Xanthomonas campestris pv. campestris (Xcc) ATCC33913 in the antisense plants. Compared with wild type, the antisense plants exhibited enhanced susceptibility to both Psm ES4326 and Xcc ATCC33913 infection (Figure 6b,c), indicating that MKK7 is required for maintaining basal resistance to bacterial pathogens.

We also tested the ability of the antisense plants to develop SAR. After inoculation of lower leaves with the avirulent pathogen Pst DC3000/avrRpt2, the induction of PR1 in the systemic tissues of the antisense plants was significantly decreased (Figure 6d). Furthermore, the avirulent pathogen-induced SAR resistance to the bacterial pathogen Psm ES4326 was dramatically decreased (Figure 6e). These results demonstrate that MKK7 is essential for the establishment of SAR.

MKK7 may be involved in generating the mobile signal in SAR

Grafting experiments performed in tobacco showed that SA accumulates in systemic tissue in response to a systemic signal that is produced at the site of primary infection and is transduced systemically (Vernooij et al., 1994). The vascular expression of MKK7 in response to pathogen infection suggests that MKK7 may be involved in systemic signaling. To test this possibility, we made use of the DEXin:MKK7 transgenic plants in which the expression of MKK7 is controlled by DEX. Three lower leaves of the DEXin:MKK7 transgenic plants and the DEXin:mkk7(K74R) transgenic plants were infiltrated with DEX to induce MKK7 and mkk7(K74R) expression. As shown in Figure 7(a,c), the MKK7 and mkk7(K74R) transgenes were induced in the DEX-treated local tissues but not in systemic tissues, suggesting that DEX did not spread systemically, which is consistent with previous studies (Aoyama and Chua, 1997). Interestingly, resistance to the bacterial pathogen Psm ES4326 was induced in the systemic tissues of the DEXin:MKK7 transgenic plants, but not in those of the DEXin:mkk7(K74R) transgenic plants (Figure 7b). Additionally, PR1 gene expression was induced not only in the DEX-treated local tissues, but also in the systemic tissues of the DEXin:MKK7 transgenic plants (Figure 7c). These results showed that ectopic expression of MKK7 in local tissues is sufficient to induce SAR in systemic tissues, demonstrating a critical role for MKK7 in generating the systemic signal of SAR.

Figure 7.

 Ectopic expression of MKK7 in local tissues induces systemic acquired resistance (SAR) in systemic tissues.
(a) Quantitative PCR analysis of the expression of MKK7 or mkk7(K74R) in local tissues (top) and systemic tissues (bottom) of wild-type, DEXin:MKK7 and DEXin:MKK7(K74R) transgenic plants with or without application of dexamethasone (DEX) in local tissues. Data represent the means of three samples with standard deviation.
(b) Growth of Pseudomonas syringae pv. maculicola (Psm) ES4326 in systemic tissues of wild-type, DEXin:MKK7 and DEXin:MKK7(K74R) transgenic plants either with or without application of DEX in local tissues. Psm ES4326 grew significantly less in the systemic tissues of the DEXin:MKK7 plants treated with DEX than in that of the mock-treated DEXin:MKK7 plants (*P < 0.0003). Data represent the means of eight samples with standard deviation. The experiment was repeated three times with similar results.
(c) RNA gel blot analysis of the expression of MKK7 and PR1 in local tissues and systemic tissues of the DEXin:MKK7 transgenic plants at the indicated time points after application of DEX in local tissues. The UBQ5 gene was used as a loading control. The experiment was repeated twice with similar results.

Discussion

Plant–pathogen interaction often triggers defense responses to protect plants from further pathogen damage (Dangl and Jones, 2001; Hammond-Kosack and Jones, 1996). The signals that plants receive from pathogen infections are amplified and transduced to the nucleus to switch gene expression profiles (Nimchuk et al., 2003). The MAPK cascades have been implicated in this signal amplification and transduction process (Pedley and Martin, 2005). Characterization of the Arabidopsis mutant bud1 in this study demonstrates that the MAPK cascade, of which MKK7 is a part, plays a critical role in regulating plant basal resistance and SAR.

The activation-tagged bud1 mutant is a dwarf, exhibiting constitutive defense responses. This phenotype could be a pleiotropic effect of disturbances of normal cell function caused by increased expression of MKK7. However, four lines of evidence argue against this possibility. First, bud1 does not exhibit necrotic lesions (data not shown), suggesting that the overall cell function in bud1 is not disrupted. Second, the expression of MKK7 in wild type is induced by pathogen infection. Increased expression of MKK7 in bud1 may mimic the upregulation of MKK7 after pathogen infection. Third, the kinase activity of MKK7 is required for all the bud1 phenotypes (Dai et al., 2006; this study), indicating that an MAPK cascade(s) is activated in bud1. Finally, silencing of MKK7 by antisense not only compromises basal resistance to Psm ES4326 and Xcc ATCC33913, but also blocks induction of SAR, demonstrating that MKK7 is required for both basal resistance and SAR.

The Arabidopsis MKK7 was placed into group D of plant MAPKK, based on sequence alignment (MAPK Group, 2002). Members in group D from other plant species have been reported to play a role in plant defense responses. For example, LeMKK4, encoded by an ortholog of MKK7 in tomato, phosphorylates LeMPK1, LeMPK2 and LeMPK3 in vitro (Pedley and Martin, 2004). When overexpressed in leaves, LeMKK4 elicits cell death and activates LeMPK2 and LeMPK3. The three MAPKs, LeMPK1, LeMPK2 and LeMPK3, have been implicated in different aspects of plant defense responses (Ekengren et al., 2003; Holley et al., 2003). In tobacco, using virus-induced silencing, it has been shown that the MKK7 ortholog, NbMKK1, controls non-host resistance including HR cell death (Takahashi et al., 2007b). These results support the conclusion that MKK7 plays a function in plant defense responses.

MKK7 may affect basal resistance and SAR through SA synthesis. Consistent with this hypothesis, the defense phenotype in bud1 was completely suppressed by the SA-deficient mutation sid2. Two different mechanisms may explain how MKK7 regulates SA synthesis. One possibility is that, like EDS1 and PAD4, MKK7 may function as a component in a signal amplification loop affecting SA synthesis (Feys et al., 2001). MKK7 may also function in generating the mobile systemic signal of SAR; perception of the systemic signal leads to SA synthesis. Although these two mechanisms are not exclusive, evidence here favors the latter. First, pathogen infection of MKK7:GUS transgenic plants induces GUS gene expression in the midribs (vascular tissues) of the local tissues, but not in systemic tissues (Figure 3a). Second, ectopic expression of MKK7 in local tissues not only activates defense responses in the local tissues, but also induces SAR in systemic tissues, suggesting that activation of MKK7 in local tissues is sufficient to induce SAR. Together, these results indicate that pathogen infection activates MKK7 in the local tissues, which leads to the production of a signal that is transduced systemically to induce SAR in systemic tissues.

In Arabidopsis, there are fewer MAPKKs (10) than MAPKs (20) and MAPKKKs (60) (MAPK Group, 2002). This suggests that various signal transduction pathways may converge at the MAPKK levels in the MAPK cascades. MKK7, an MAPKK, not only functions as a negative regulator of plant PAT (Dai et al., 2006), but also functions as a positive regulator of plant basal resistance and SAR, suggesting that MKK7 may serve as a crosstalk point between auxin signaling and defense responses.

Crosstalk between auxin and plant defense responses has been known for many years. Most microbial pathogens possess the capacity to synthesize indole-3-acetic acid (IAA) (Fett et al., 1987; Wichner and Libbert, 1968). However, this capacity has been shown to be important for the pathogenicity of only a few pathogens such as P. syringae pv. savastanoi, Agrobacterium tumefaciens and Agrobacterium rhizogenes (Liu et al., 1982; Offringa et al., 1986; Surico et al., 1985). Using Xcc ATCC33913, a strain that does not synthesize IAA itself, O’Donnell et al. (2003)showed that the pathogen was able to induce the host plant to produce IAA by upregulating host genes involved in IAA synthesis. These results indicate that pathogens may perturb auxin homeostasis of the host plant to promote disease. Characterization of Arabidopsis dth9 and sgt1b/eta3 mutants suggests that auxin homeostasis is one of the components participating in the regulation of plant defense responses (Gray et al., 2003; Mayda et al., 2000). Both dth9 and sgt1b/eta3 mutants are more susceptible to pathogen invasion and insensitive to exogenous auxin application (Gray et al., 2003; Mayda et al., 2000; Tör et al., 2002). Recently, a flg22-induced microRNA (miRNA) was shown to restrict P. syringae growth by repressing auxin signaling in Arabidopsis (Navarro et al., 2006). In this study, we show that increased expression of MKK7 in the bud1 mutant not only causes deficiency in plant PAT, but also leads to constitutive defense responses, suggesting that the MKK7 MAPK cascade(s) is likely to be involved in regulating both auxin homeostasis and defense responses.

Because the kinase activity of MKK7 is required for all the bud1 phenotypes, one or more MAPK cascades may be activated in bud1. Identification of other components in the MKK7 MAPK cascade(s), and its downstream effectors, will bring us more insight into the signal amplification and transduction pathways in plant defense responses.

Experimental procedures

Plant materials and growth conditions

The wild type used was the Columbia ecotype (Col-0). The mutant alleles used were sid2-2 (eds16-1) and npr1-1. The BGL2:GUS transgenic line has been described by Bowling et al. (1994). The heterozygous bud1 plants were used to produce plants for the experiments.

The bud1 BGL2:GUS lines were generated using pollen from the BGL2:GUS transgenic plants to fertilize the heterozygous bud1 plants. The bud1 heterozygotes were identified in the F2 generation by the bud1 morphological phenotype. The BGL2:GUS homozygotes were identified in the F3 generation using a histochemical GUS assay. The bud1sid2 and bud1npr1 double mutants were generated using pollen from sid2-2 or npr1-1 plants to fertilize the heterozygous bud1 plants. The bud1 heterozygotes were identified in the F2 generation as described above. The cleaved amplified polymorphic sequence markers for sid2-2 and npr1-1 were used to confirm homozygosity at the sid2 and npr1 loci.

Arabidopsis seeds were sown on autoclaved soil (Metro-Mix 200; Grace-Sierra, Malpitas, CA, USA) and vernalized at 4°C for 3 days. Plants were germinated and grown at 22°C under a 16-h light/8-h dark regime.

Pathogen infection

Infection of plants with Psm ES4326 or H. parasitica Noco2 was performed as described previously (Clarke et al., 1998). For Psm ES4326 infection, between four and eight infected leaves were collected for each genotype, treatment or time point to determine in planta growth of the pathogen. For H. parasitica Noco2 infection, 25 leaves from 10 plants were harvested to determine the degree of infection. After vigorous vortex-mixing in 1 ml of H2O, two 10-μl aliquots from each sample were examined with a hemacytometer to determine the number of spores. Three samples for each genotype, treatment or time point were assayed to obtain a standard deviation.

For SAR induction, three lower leaves on each plant were inoculated with an avirulent bacterial pathogen Pst DC3000/avrRpt2 (OD600 = 0.02). The upper uninfected systemic leaves were collected 2 days later for PR1 gene expression analysis. After 3 days, the uninfected systemic leaves were challenge-inoculated with Psm ES4326 (OD600 = 0.001). Eight leaves were collected on day 3 to examine the growth of the pathogen.

To determine the expression pattern of MKK7, three half leaves on each MKK7:GUS plant were inoculated with Pst DC3000/avrRpt2 (OD600 = 0.02). The inoculated plants were collected for histochemical GUS assay after 24 h.

Infection of the bacterial pathogen Xcc ATCC33913 was performed following the protocol used for Psm ES4326 (Clarke et al., 1998). Briefly, the bacteria cell suspension (in 10 mm MgCl2, OD600 = 0.005) was infiltrated into leaves with a 1-ml syringe. After 4 days, eight leaves were collected to determine the in planta growth of the pathogen.

Histochemical GUS assay

Soil-grown plants (3–4-weeks old) with or without pathogen treatment were stained for GUS activity as described by Fan and Dong (2002). Briefly, plants were submerged in a solution containing 0.5 mg ml−1 5-bromo-4-chloro-3-indolyl glucuronide in 0.1 m Na2HPO4, pH 7.0, 10 mm EDTA, 0.5 mm potassium ferricyanide/ferrocyanide and 0.06% Triton X-100, and were vacuum infiltrated for 5 min. After incubation at 37°C for 16 h, the staining solution was removed and the samples were cleared of chlorophyll by sequential changes of 75% and 95% ethanol.

In-gel kinase activity assay

The in-gel kinase activity assay was performed as described by Ren et al. (2002). In brief, protein was extracted from 2- to 3-week-old plants by homogenizing in extraction buffer [50 mm Tris–HCl, pH 7.5, 5 mm EDTA, 5 mm EGTA, 10 mm Na3VO4, 10 mm NaF, 50 mm glycerophosphate, 10 mm DTT, 5% glycerol and protease inhibitors: 50 μg ml−1 L-1-tosylamido-2-phenylethylchloromethyl ketone (TPCK), 50 μg ml−1 N-alpha-p-tosyl-L-lysine chloromethyl ketone (TLCK) and 0.6 mm phenylmethanesulphonyl fluoride (PMSF)]. About 20 μg of total protein was separated by electrophoresis on 10% SDS-polyacrylamide gels embedded with 0.1 mg ml−1 myelin basic protein (MBP) in separating gel as a substrate for the kinase. After electrophoresis, the SDS was removed from the gel by washing with washing buffer (25 mm Tris-HCl, pH 7.5, 0.5 mm DTT, 0.1 mm Na3VO4, 5 mm NaF, 0.5 mg ml−1 bovine serum albumin and 0.1% Triton X-100) three times for 30 min each at room temperature (22–23°C). The proteins were then renatured in 25 mm Tris-HCl, pH 7.5, 1 mm DTT, 0.1 mm Na3VO4 and 5 mm NaF at 4°C overnight with three changes of the buffer. The gel was incubated at room temperature in 100 ml of reaction buffer (25 mm Tris-HCl, pH 7.5, 2 mm EGTA, 12 mm MgCl2, 1 mm DTT, 0.1 mm Na3VO4) for 30 min. Phosphorylation was performed for 1.5 h at room temperature in 30 ml of the same buffer with 200 nm ATP plus 50 μCi of [γ-32P]ATP (6000 Ci mmol−1). The reaction was stopped using a solution with 5% trichloroacetic acid (w/v) and 1% sodium pyrophosphate (w/v). The gel was then washed using the same solution for 6 h at room temperature with five changes of solution to remove unincorporated radioactivity. The gel was then dried with a gel dryer (Model 583; Bio-Rad, http://www.bio-rad.com) and subjected to autoradiography. Prestained size markers (Fisher Scientific, http://www.fisher.co.uk) were used to calculate the size of kinases.

SA measurement

Soil-grown plants (3–4-weeks old) were used to measure the concentration of free SA using a previously described protocol (Schmelz et al., 2003).

RNA analysis, RT-PCR and quantitative PCR

RNA extraction and RNA gel blot analysis were carried out as described by Cao et al. (1997). For reverse transcription (RT), total RNA was treated with Dnase I (Gibco, http://www.invitrogen.com) at 37°C for 30 min. After inactivation of the DNase, RT was performed using SUPERSCRIPT First-strand Synthesis System (Gibco) and 2 μg of the DNase-treated RNA in a 20-μl reaction. Aliquots of the resulting RT reaction product were used for RT-PCR and quantitative PCR. For RT-PCR, amplification of cDNA was performed with 2 μl of RT product in a 50-μl reaction. The three pairs of primers used for amplification reactions of MKK7 were MKK7F1 (5′-ATGGCTCTTGTTCGTAAACG-3′) and MKK7R1 (5′-AAGACTTTCACGGAGAAAAGG-3′), MKK7F2 (5′-GCACTTGCGCTTACAT-3′) and MKK7R2 (5′-GAAAAGGGTGACCGAGA-3′), and MKK7F3 (5′-GTAAAGAATCGAGTGAGAGG-3′) and MKK7R3 (5′-AATTGCGATTTGGGTCACCC-3′). Primers used for the GVG gene were GVGF (5′-GACAATCAAGCGGAAACCTG-3′) and GVGR (5′-TCATGCATGGAGTCCAGAAG-3′). All PCR reactions were performed under the following conditions: 94°C for 3 min, 35 cycles (94°C for 1 min, 56°C for 1 min, 72°C for 1 min), and a final extension at 72°C for 10 min.

Quantitative PCR was performed using SYBR Green protocol (Applied Biosystems, http://www.appliedbiosystems.com) with 1-μm primers and a 0.2-μl aliquot of RT product in a total of 10 μl per reaction. Reactions were run and analyzed on a Lightcycler (Roche, http://www.roche.com) according to the manufacturer’s instructions. A standard curve was made by determining the threshold cycle (Ct) values for a dilution series of the RT reaction product for each primer pair. For each reaction, the Ct was determined by setting the threshold within the logarithmic amplification phase. The relative quantity of a gene is expressed in relation to ubiquitin 5 (UBQ5) using the formula 2[Ct(UBQ5) – Ct(GENE)], where 2 represents perfect PCR efficiency. Quantitative PCR reactions were performed in triplicate to obtain a standard deviation. The primers used were MKK7F3 and MKK7R3, as described above, MKK9F (5′-AGTTTAGGAGCTTCGTTGAG-3′) and MKK9R (5′-AGTTTAGGAGCTTCGTTGAG-3′), and UBQ5F (5′-GACGCTTCATCTCGTCC-3′) and UBQ5R (5′-GTAAACGTAGGTGAGTCCA-3′).

Plasmid construction and plant transformation

To fuse the MKK7 promoter with the GUS reporter gene, a 1.694-kb DNA fragment of the MKK7 promoter was amplified from wild-type genomic DNA by PCR using primers XbaI-MKK7PF (5′-GCTCTAGAAGTGATTTGGTAGGAGCC-3′) and SmaI-MKK7PR (5′-TCCCCCGGGAGAGTGATGATGGTGATCG-3′). The PCR products were digested with XbaI and SmaI and cloned into XbaI/SmaI-digested pBI101 vector.

To generate transgenic plants expressing MKK7 under the control of DEXin, MKK7 cDNA was amplified from wild-type genomic DNA by PCR using the primers SalI-MKK7F (5′-GCGTCGACCTCTCTTCTATTTCCATGGC-3′) and SpeI-MKK7R (5′-GGACTAGTACAAGCAGTCGGATCTAAAG-3′). The PCR products were digested with SalI and SpeI, and were cloned into XhoI/SpeI-digested pTA7001 vector (Aoyama and Chua, 1997). The mkk7(K74R) mutant was generated by site-directed mutagenesis in the pTA7001-MKK7 construct using a PCR-based Quick-Change site-directed mutagenesis kit (Stratagene, http://www.stratagene.com). The presence of the expected mutation in the pTA7001-mkk7(K74R) was verified by DNA sequencing.

The T-DNA plasmids were introduced into Agrobacterium strain GV3101(pMP90) by electroporation, and were transformed into Arabidopsis plants (ecotype Columbia) using the floral-dip method.

DEX treatment

Leaves of the DEXin:MKK7 and DEXin:mkk7(K74R) transgenic plants were infiltrated with 0.01 mm DEX in 0.1% ethanol solution or 0.1% ethanol using a 1-ml syringe. After 24 h, the infiltrated leaves were either collected for PR1 gene expression analysis or inoculated with Psm ES4326 (OD600 = 0.001) for the resistance test. For H. parasitica Noco2 infection, 7-day-old DEXin:MKK7 and DEXin:mkk7(K74R) transgenic plants were sprayed with 0.01 mm DEX in 0.1% ethanol solution plus 0.01% Tween-20, or 0.1% ethanol plus 0.01% Tween-20. After 24 h the seedlings were infected with H. parasitica Noco2 as described above.

To test whether local application of DEX can induce SAR in these transgenic plants, three lower leaves on each plant were infiltrated with 0.01 mm DEX. Three days after DEX treatment, the uninfiltrated systemic leaves were challenge-inoculated with Psm ES4326 (OD600 = 0.001). Eight leaves were collected 3 days post-inoculation to examine the growth of the pathogen. For PR1 expression, after DEX treatment, both local tissues and systemic tissues were collected at different time points for RNA gel blot analysis.

Statistical methods

All statistical analyses were performed with the data analysis tools (t-test: two samples assuming unequal variances) in the Microsoft Excel program of Microsoft Office 2004 for Macintosh (Microsoft, http://www.microsoft.com).

Acknowledgements

We thank Dr Eric Schmelz (USDA, Gainesville, FL, USA) for measuring the free SA levels and Dr Jeffrey Jones (University of Florida, Gainesville, FL, USA) for providing the pathogen Xcc ATCC33913. This work was supported by a start-up fund from the University of Florida awarded to ZM. Initiation of this work was supported by an NIH fund to XD (R01-GM-069594-03). CD was supported by an Alumni Fellowship from the University of Florida.

Ancillary