INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plants mount a range of ameliorative responses during biotic and abiotic stresses that are dependent on signal perception and activation of both overlapping and distinct signaling pathways. In response to potentially pathogenic microorganisms, several early defense reactions are initiated in plant cells, including rapid changes in ion fluxes, generation of reactive oxygen species (ROS), and reversible protein phosphorylation (Yang et al., 1997). In turn, these signaling pathways culminate in the expression of defense-related genes, formation of phytoalexins, and localized host cell death, constituting the hypersensitive response (HR; Heath, 1999).

Reversible protein phosphorylation is a key process regulating many aspects of cellular function in eukaryotes, including elicitor-induced defense responses. For example, several pharmacological studies have shown that both protein kinase and phosphatase activities are involved in regulation of the oxidative burst, in which hydrogen peroxide (H₂O₂) is generated (Levine et al., 1994; Chandra and Low, 1995; Kauss and Jeblick, 1995; Desikan et al., 1996).

Considerable interest is currently focussed on mitogen-activated protein kinase (MAPK) signaling in plants. MAPKs are Ser/Thr kinases with both cytoplasmic and nuclear substrates, and are themselves activated via dual phosphorylation on Thr and Tyr residues by an MAPK kinase (MAPKK or MEK). In turn, this kinase is activated by an MAPKK kinase (MAPKKK). A large number of MAPK signaling components have now been cloned from plants, with considerable evidence accumulating for their roles in mediating adaptive responses to environmental stresses such as drought, cold, and osmotic stress (Hirt, 1997; Mizoguchi et al., 1997; Hirt and Asard, 2000), as well as pathogen challenge (Somssich, 1997). There are now several examples in the literature where MAPKs have been implicated during defense responses in a wide variety of plant-elicitor interactions (Adam et al., 1997; Ligterink et al., 1997; Lebrun-Garcia et al., 1998; Zhang and Klessig, 1998; Desikan et al., 1999a; Romeis et al., 1999; Suzuki et al., 1999; Yang et al., 2001).

Harpins are heat-stable Gly-rich proteins that are encoded by hrp genes present in several phytopathogenic bacteria (He, 1996). Harpin is one of the first bacterial elicitors characterized as inducing HR in several non-host species (Lindgren, 1997). In addition to eliciting several active defense responses in plants leading to the HR, which include rapid ion fluxes, membrane depolarization, and generation of ROS (Baker et al., 1993; He et al., 1994), harpin has also been shown to contribute to disease resistance in plants by reducing bacterial growth (Dong et al., 1999). In previous work, we have shown that harpin from Pseudomonas syringae pv syringae induces a number of defense responses in Arabidopsis cell suspension cultures, including generation of H₂O₂ (Desikan et al., 1996), activation of defense gene expression, and programmed cell death (PCD; Desikan et al., 1998). We have demonstrated recently that harpin induces the activation of two MAPK-like enzymes in Arabidopsis cells (Desikan et al., 1999a), whereas exogenous H₂O₂ activates a single MAPK-like enzyme (Desikan et al., 1999b). Several MAPK homologs have been identified in Arabidopsis (Mizoguchi et al., 1997), but as yet there is only limited information available on the role of specific MAPKs in defense responses (Nuhse et al., 2000; Yang et al., 2001).

In this study, we identify the two MAPK-like enzymes activated by harpin as AtMPK4 and AtMPK6. Harpin-induced activation of AtMPK4 and AtMPK6 is independent of the presence of H₂O₂, although H₂O₂ activates AtMPK6 but not AtMPK4. We show that harpin and H₂O₂ also induce a similar activation profile of AtMPK4 and AtMPK6 in Arabidopsis leaves. Treatment with the MAPKK inhibitor PD98059 reduces the harpin-induced activation of AtMPK4 in suspension cultures, but has no effect on the activation of AtMPK6. Together, these data suggest that harpin activates several signaling pathways, one leading to the oxidative burst and others leading to the activation of AtMPK4 or AtMPK6. Neither harpin nor H₂O₂ altered the expression of the genes encoding AtMPK4 and AtMPK6, nor did they have any effect on the expression of genes encoding AtMEK1, ATMEKK1, or ATMKK2, likely upstream components in a functional cascade activating AtMPK4 (Ichimura et al., 1998; Mizoguchi et al., 1998).

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Harpin and H₂O₂ Activate Myelin Basic Protein (MBP) Kinases in Arabidopsis Leaves

In previous work, we have shown that harpin and H₂O₂ activate MAPK-like enzymes in Arabidopsis cell suspension cultures (Desikan et al., 1999a, 1999b). To determine if similar responses would be reproduced in leaves, harpin (5 µg mL¹) or H₂O₂ (20 mM) was vacuum infiltrated into leaves for various times. Subsequent in-gel kinase assays of extracts from these leaves demonstrated that harpin induced the activation of two MBP kinases of 43 and 47 kD within 15 min, and that after 30 min the activation of these kinases diminished (Fig. 1A). Exogenous H₂O₂ also induced the activation of an MBP kinase at about 47 kD after 15 min (Fig. 1B). Mock infiltration of leaves with water did not induce the activation of any MBP kinase (Con, Fig. 1, A and B). The activation kinetics seen with leaves were similar to those of suspension cultures (Desikan et al., 1999a, 1999b).

View larger version (36K): [in this window] [in a new window]   Figure 1.   Harpin- and H2O2-induced activation of MBP kinases in Arabidopsis leaves. A, Protein extracts from control- (Con) or harpin- (hrp, 5 µg mL1) treated leaves for various times (indicated in minutes) were subjected to in-gel protein kinase assay using MBP as substrate. The molecular masses of the 43- and 47-kD kinases are indicated. B, Protein extracts from control- (Con) or H2O2- (20 mM) treated leaves for various times (in minutes) were subjected to in-gel protein kinase assay using MBP as substrate. The molecular mass of the 47-kD protein is indicated.

AtMPK4 and AtMPK6 Proteins Are Present in Arabidopsis Cell Cultures

AtMPK4 and AtMPK6 proteins have been shown to be present in Arabidopsis leaves (Ichimura et al., 2000). To determine whether these MAPKs are similarly present in Arabidopsis suspension cultures, immunoblot analysis was performed on protein extracts from control-, harpin-, or H₂O₂-treated cells using antibodies specifically raised against the C and N terminus of AtMPK4 and AtMPK6, respectively (Ichimura et al., 2000). Figure 2A shows that the anti-AtMPK4 antibody reacted strongly with a protein of molecular mass of about 43 kD in cell extracts, and also, but to a lesser extent, with a larger protein. In leaf extracts, the anti-AtMPK4 antibody reacts with AtMPK4 at an apparent molecular mass of 43 kD (Ichimura et al., 2000); some cross-reactivity with a higher molecular mass non-MAPK protein was also apparent, as observed here. The anti-AtMPK6 antibody recognized a single protein of molecular mass of about 47 kD (Fig. 2B), as reported for Arabidopsis leaves (Ichimura et al., 2000). The estimated molecular mass of the proteins detected by anti-AtMPK4 and -AtMPK6 antibodies is dependent on the migration behavior of the M_r markers used during SDS-PAGE. In this study, Bio-Rad (Hertfordshire, UK) markers were used, whereas NEB (Hertfordshire, UK) markers were used in a previous report from our laboratory (Desikan et al., 1999a, 1999b). We reported previously that harpin activated two MAPK-like enzymes of molecular mass of about 39 kD and 44 kD (Desikan et al., 1999a) and H₂O₂ activated a single MAPK-like enzyme of about 44 kD (Desikan et al., 1999b) in Arabidopsis cells. However, in this report, the apparent molecular masses of AtMPK4 and AtMPK6 in suspension cultures were calculated as 43 and 47 kD, respectively, based on Bio-Rad markers, which is in agreement with other work (Ichimura et al., 2000). This apparent discrepancy in the sizes of MAPKs has been reported by other workers (Romeis et al., 1999; Nuhse et al., 2000). Figure 2 also shows that both harpin (1 µg mL¹) or H₂O₂ (20 mM) had little effect on the relative abundance of AtMPK4 and AtMPK6 proteins in Arabidopsis cell extracts over the time course of this experiment.

View larger version (43K): [in this window] [in a new window]   Figure 2.   Immunodetection of AtMPK4 and AtMPK6 in Arabidopsis cells. A, Protein extracts from control- (Con), harpin- (Hrp, 1 µg mL1) or H2O2- (20 mM) treated cells (for 15 min) were fractionated by SDS-PAGE, and western blotting performed using anti-AtMPK4 antibody. B, Western blotting was performed on extracts as above using anti-AtMPK6 antibody. The molecular masses of the 43- and 47-kD proteins are indicated.

Harpin Induces the Activation of AtMPK4 and AtMPK6

To determine if the harpin or H₂O₂-activated MAPK-like enzymes in suspension cultures (Desikan et al., 1999a, 1999b) and leaves (Fig. 1) are AtMPK4 and AtMPK6, immunoprecipitation was performed on harpin- and H₂O₂-treated extracts using anti-AtMPK4 and -AtMPK6 antibodies. The precipitated proteins were fractionated on MBP-embedded gels and in-gel kinase assays performed. As can be seen in Figure 3, the anti-AtMPK4 antibody immunoprecipitated only a 43-kD kinase. Furthermore, this kinase was only seen in extracts from harpin-treated cells; little or no immunoprecipitable kinase activity was seen in extracts from control- or H₂O₂-treated cells. Using anti-AtMPK6 antibody to determine if harpin or H₂O₂ activates AtMPK6, similar experiments revealed that immunoprecipitable kinase activity was seen in extracts from both harpin- and H₂O₂-treated cells; however, only the 47-kD kinase was precipitated (Fig. 3, AtMPK6).

View larger version (41K): [in this window] [in a new window]   Figure 3.   Activation of AtMPK4 and AtMPK6 in Arabidopsis suspension cultures. Protein extracts from control- (Con), harpin- (Hrp, 2 µg mL1 for 15 min), or H2O2- (20 mM for 15 min) treated cells were immunoprecipitated with anti-AtMPK4 antibody (AtMPK4, using 500 µg protein extract) or anti-AtMPK6 antibody (AtMPK6, using 100 µg protein extract) and in-gel kinase assay performed. Extracts that were not immunoprecipitated (IP, using 40 µg protein extract) were also subjected to in-gel kinase assay. The molecular masses of the 43- and 47-kD MAPKs are indicated.

The activity of kinases was also determined in extracts from control and treated cells that were not immunoprecipitated. These experiments revealed that a 43-kD kinase activated only after harpin treatment (Fig. 3; IP lanes) aligned with that precipitated by anti-AtMPK4 antibody. A 47-kD kinase that did possess MBP kinase activity and was clearly activated by harpin and H₂O₂ as described earlier (Desikan et al., 1999a, 1999b), was also seen (Fig. 3; IP lanes), although it was not precipitated by anti-AtMPK4 antibody. However, the 47-kD kinase was immunoprecipitated by anti-AtMPK6 antibody, although the 43-kD kinase was not (Fig. 3).

The identity of the MAPKs activated by harpin and H₂O₂ in Arabidopsis leaves was also determined using immunoprecipitation and in-gel assays. Extracts from harpin-treated leaves immunoprecipitated with anti-AtMPK6 and -AtMPK4 antibodies possessed kinase activities at 47 and 43 kD, respectively (Fig. 4, lanes 2 and 5). H₂O₂-treated leaves showed AtMPK6-immunoprecipitable kinase activity, but only very weak AtMPK4-immunoprecipitable kinase activity (Fig. 4, lanes 3 and 6), compared with extracts from harpin-treated leaves (Fig. 4, lane 5).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Activation of AtMPK4 and AtMPK6 in Arabidopsis leaves. Protein extracts from control- (lanes 1 and 4), harpin- (lanes 2 and 5; 5 µg mL1, 15 min), or H2O2- (lanes 3 and 6; 20 mM, 30 min) treated leaves were subjected to in-gel kinase assay after immunoprecipitation with anti-AtMPK6 (lanes 1-3) or anti-AtMPK4 (lanes 4-6) antibodies. The molecular masses of the 43- and 47-kD MAPKs are indicated.

Harpin-Induced Activation of AtMPK4 and AtMPK6 Occurs Independent of the Oxidative Burst via Different Pathways

Previous work in our laboratory has shown that harpin induces the generation of ROS such as H₂O₂ (Desikan et al., 1996), and that both harpin and H₂O₂ induce differential defense responses in Arabidopsis suspension cultures (Desikan et al., 1998). To investigate whether the effects of harpin on the activation of AtMPK4 and AtMPK6 were dependent on H₂O₂ generation, cells were challenged with harpin in the presence of catalase, which scavenges H₂O₂ and ameliorates its effects (Desikan et al., 1998), and in-gel kinase assays subsequently performed. Catalase pretreatment did not diminish harpin-induced activation of the 43- and 47-kD kinases (Fig. 5A). However, catalase pretreatment did cause a slight increase in kinase activitythis was a nonspecific effect of catalase (Fig. 5A) because even boiled catalase caused this effect (data not shown). To determine whether harpin-induced AtMPK4 and AtMPK6 activation occurred independently of the oxidative burst, cells were pretreated with DPI, an inhibitor of ROS generation via NADPH oxidase (Desikan et al., 1996). DPI pretreatment did not inhibit activation of the two kinases by harpin (Fig. 5B), demonstrating that such activation is independent of the oxidative burst.

View larger version (34K): [in this window] [in a new window]   Figure 5.   Harpin-induced activation of AtMPK4 and AtMPK6 occurs independently of the oxidative burst. A, Protein extracts from control- (Con), catalase- (Cat, 0.5 mg mL1), harpin- (Hrp, 2 µg mL1, 15 min) or harpin plus catalase- (Hrp + cat, 0.5 mg mL1) treated cells were subjected to in-gel kinase assay. The molecular masses of the 43- and 47-kD MAPKs are indicated. B, Protein extracts from control- (Con), harpin- (Hrp, 2 µg mL1), or harpin plus diphenylene iodonium- (DPI; Hrp + DPI, 10 µM) treated cells were subjected to in-gel kinase assay. The molecular masses of the 43- and 47-kD MAPKs are indicated.

MAPKs are typically activated via phosphorylation by an upstream MAPK kinase (MEK). To address the possibility of such a requirement for activation of AtMPK4 and AtMPK6, we adopted a pharmacological approach. In a previous report, we described the inhibitory effects of an inhibitor of MAPKK activation, PD98059, on harpin-induced activation of the 43-kD kinase, with little or no effects on the activation of the 47-kD kinase (Desikan et al., 1999a). To determine the effects of PD98059 on the activation of AtMPK4 and AtMPK6, extracts from cells treated with harpin in the absence or presence of PD98059 were immunoprecipitated with anti-AtMPK4 and -AtMPK6 antibodies and analyzed by in-gel kinase assay (Fig. 6, A and B). PD98059 treatment substantially reduced the activation of AtMPK4 by harpin. However, using the same extracts in immunoprecipitation experiments with anti-AtMPK6 antibody, it was observed that harpin-induced activation of AtMPK6 was not inhibited by PD98059, in accordance with our previous report (Fig. 6B; Desikan et al., 1999a).

View larger version (23K): [in this window] [in a new window]   Figure 6.   Harpin-induced AtMPK4 but not AtMPK6 activation is inhibited by PD98059, an MAPKK inhibitor. A, Cells were either treated with dimethyl sulfoxide (DMSO; Con), harpin + DMSO (Hrp, 2 µg mL1, 15 min), or harpin in the presence of PD98059 (PD; 1 × 104 M) and protein extracts subjected to in-gel kinase assay after immunoprecipitation with anti-AtMPK4 antibody. B, Cells were either treated with DMSO (Con), harpin + DMSO (Hrp, 2 µg mL1, 15 min), or harpin in the presence of PD98059 (PD, 1 × 104 M) and protein extracts subjected to in-gel kinase assay after immunoprecipitation with anti-AtMPK6 antibody.

Effect of Harpin and H₂O₂ on AtMPK4, AtMPK6, AtMEK1, AtMKK2, and AtMEKK1 mRNA

Various abiotic stresses have been shown to induce the expression of genes encoding MAPKs, MAPKKs, and MAPKKKs in Arabidopsis (Mizoguchi et al., 1996; Morris et al., 1997; Ichimura et al., 1998). Furthermore, yeast two-hybrid analysis has demonstrated that AtMPK4, AtMEK1/AtMKK2, and AtMEKK1 represent a potential cognate MAPK-MAPKK-MAPKKK cascade in Arabidopsis (Ichimura et al., 1998; Mizoguchi et al., 1998). To determine if harpin and H₂O₂ had any effect on the transcription of genes encoding AtMPK4 and AtMPK6, northern analysis was performed using RNA from cells treated for 2 h with harpin or H₂O₂ (Fig. 7). As a positive control, blots were hybridized with a PAL1 cloneboth harpin and H₂O₂ induced an increase in PAL1 mRNA, as described previously (Desikan et al., 1998). It is clear that neither harpin nor H₂O₂, at the concentrations used, had any effect on the expression of AtMPK4 or AtMPK6 mRNA (Fig. 7). The effects of harpin and H₂O₂ on the transcription of AtMEKK1, AtMEK1, and AtMKK2 were also determined. Neither treatment appeared to have any effect on the expression of these genes (Fig. 7).

View larger version (34K): [in this window] [in a new window]   Figure 7.   Effects of H2O2 and harpin on the expression of AtMEKK1, AtMEK1, AtMKK2, AtMPK4, and AtMPK6 mRNA. Arabidopsis cells were treated with either H2O2 (10 mM) or harpin (Hrp, 1 µg mL1) for 2 h and RNA isolated from the harvested cells subjected to northern analysis using 32P-labeled AtMEKK1, AtMEK1, AtMKK2, AtMPK4, and AtMPK6 cDNAs as probes. As a control, the blot was stripped and probed with a PAL1 genomic clone (Desikan et al., 1998).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Suspension cultures of Arabidopsis are a good model system with which to elucidate signaling processes required for defensive responses induced by pathogens or elicitor challenge. In recent work, we have shown that Arabidopsis cells challenged with phytopathogenic bacteria (Clarke et al., 2000), or the bacterial elicitor harpin, initiate a series of responses that includes the generation of H₂O₂ and PCD (Desikan et al., 1996, 1998). PCD is a characteristic of the HR, a complex suite of cellular responses that requires the interplay of numerous signal transduction pathways and culminates in host cell death and retardation of pathogen growth (Heath, 1999).

Harpin is a nonspecific bacterial elicitor, one of the few bacterial elicitors characterized to date. It induces a number of HR-related events, such as changes in ion fluxes, reversible protein phosphorylation, and the oxidative burst (Baker et al., 1993; He et al., 1994; Desikan et al., 1996). Harpin has been shown to activate an MAPK-like enzyme in tobacco (Nicotiana tobacum L. var. Samsun NN) leaves (Adam et al., 1997), and two MAPK-like enzymes in Arabidopsis cell suspension cultures (Desikan et al., 1999a). Molecular characterization of MAPKs activated by various stresses is clearly imperative, so that the biochemical and biological roles of such enzymes can be subsequently elucidated. The activation profile of AtMPK4 (Ichimura et al., 2000) suggested that AtMPK4 may be one of the two MAPKs activated by harpin in Arabidopsis suspension cultures. The immunological data presented here demonstrate that harpin does activate AtMPK4. The demonstrations elsewhere that AtMPK6 is activated by H₂O₂ (Kovtun et al., 2000), or bacterial flagellins (Nuhse et al., 2000), in Arabidopsis protoplasts and cells, respectively, suggested that the other MAPK (at 47 kD) activated by harpin and H₂O₂ (Desikan et al., 1999a, 1999b) is probably AtMPK6. In the present study, we confirm that H₂O₂ does activate AtMPK6 in suspension cultures, and show that AtMPK6 can also be activated independently by harpin. We also demonstrate that harpin and H₂O₂ activate AtMPK4/6 and AtMPK6, respectively, in leaves in a similar manner to that seen with suspension cultures. This finding lends experimental support to the concept of suspension cultures as model systems with which to elucidate some of the signaling mechanisms observed in planta.

It is clear that activation of AtMPK4 and AtMPK6 by harpin can occur independently of the oxidative burst, as both catalase and DPI, which remove H₂O₂ and inhibit ROS generation, respectively (Desikan et al., 1996), did not inhibit harpin-induced AtMPK4 and AtMPK6 activation. This is similar to the activation of MAPKs in other systems that have been reported to be upstream or independent of the oxidative burst (Ligterink et al., 1997; Romeis et al., 1999; Yang et al., 2001). Thus, harpin must activate several signaling pathways, one leading to the activation of the oxidative burst, and others leading to activation of AtMPK4 or AtMPK6. The effects of the MEK inhibitor PD98059 suggest that the activation of AtMPK4 and AtMPK6 by harpin is also differentially regulated. PD98059 was originally reported to be an inhibitor of MAPKKs in mammalian systems (Cohen, 1997), and has been shown recently to inhibit the activation of several plant MAPK systems (Desikan et al., 1999a; Romeis et al., 1999; Burnett et al., 2000; Samuel et al., 2000). Here, the activation of AtMPK4 is inhibited, as previously suggested (Desikan et al., 1999a). This implies that the MAPKK responsible for activating AtMPK4 (potentially AtMEK1, see below) is sensitive to PD98059, whereas the activating enzyme for AtMPK6, probably ANP1 (Kovtun et al., 2000), is not. It is, perhaps, also significant that the inhibition of AtMPK4 activation by PD98059 correlates with inhibition by PD98059 of harpin-induced PCD (Desikan et al., 1999a), suggesting a role for AtMPK4 in the signaling pathway leading to cell death during harpin-induced HR in Arabidopsis. Recent data indicate that AtMPK4 negatively regulates pathogen-induced systemic acquired resistance in Arabidopsis (Petersen et al., 2000); it is likely that there is considerable cross talk involving MAPK signaling during HR and systemic acquired resistance that still requires clarification.

Phylogenetic analysis of plant MAPKs indicates several distinct groups (Zhang and Klessig, 1997; Jonak et al., 1999), with kinases within the same group having potentially similar functions. AtMPK4 falls within a third subgroup of a major group of MAPKs, and is related to MMK2 of alfalfa, an enzyme that can complement a mutant yeast MAPK required for high temperature tolerance (Jonak et al., 1995). Two other subgroups of this major group of MAPKs include the tobacco wound-induced protein kinase (WIPK) and salicylic acid-induced protein kinase (SIPK), as well as the Arabidopsis AtMPK3 and AtMPK6. WIPK and SIPK were originally shown to be wound- (Seo et al., 1995) and salicylic acid-induced (Zhang and Klessig, 1997), respectively. SIPK is also activated by fungal elicitors (Zhang et al., 1998) and abiotic stress (Samuel et al., 2000), and both SIPK and WIPK have been shown to be activated in a fungal race-specific interaction in tobacco (Romeis et al., 1999). Importantly, a functional role for both SIPK and WIPK in HR has recently been demonstrated in tobacco, where defense responses such as cell death and defense gene activation have been shown to be directly regulated by these MAPKs (Yang et al., 2001).

It is now apparent that components of MAPK cascades in plants can be regulated not only at the posttranslational level, but also at the level of transcription (Hirt, 1999). For example, members of the WIPK subgroup of MAPKs are induced at the mRNA level by wounding, elicitor challenge, drought, and fungal pathogens (Seo et al., 1995; Mizoguchi et al., 1996; Ligterink et al., 1997; Romeis et al., 1999). However, those of the SIPK subgroup are not induced at the mRNA level by elicitors, salicylic acid, or fungal pathogens (Zhang et al., 1998; Romeis et al., 1999). In this study, neither AtMPK4 nor AtMPK6 mRNA were induced by either harpin or H₂O₂. Other work has demonstrated that AtMPK4 and AtMPK6 transcripts are not induced by low temperature, low humidity, touch, or wounding (Ichimura et al., 2000).

It is also possible that upstream components of the MAPK signaling cascade are transcriptionally regulated. AtMEKK1 has been shown to interact physically with AtMEK1/AtMKK2 and AtMPK4, suggesting that this collection of enzymes may represent a functional complex (Mizoguchi et al., 1998). AtMEK1 phosphorylates AtMPK4 in vivo only on Thr residues (Huang et al., 2000), with auto-phosphorylation being an important regulatory event in activation of AtMPK4. AtMEKK1 and AtMEK1 transcripts accumulated following various stresses or wounding (Mizoguchi et al., 1996; Morris et al., 1997), but in the present study, neither harpin nor H₂O₂ appeared to have any effect on the levels of the transcripts encoding these proteins. These data suggest that different stimuli have different effects on MAPK cascades, at both the enzyme activation and transcription levels.

A novel dual-specificity protein phosphatase has recently been identified in Arabidopsis and shown to dephosphorylate and inactivate AtMPK4 (Gupta et al., 1998). Whether this enzyme is constitutively active or inducible during biotic interactions remains to be demonstrated. Scaffold proteins that mediate physical associations between MAPK components have been identified in yeast and mammals. Such interactions represent obvious regulatory junctures in MAPK signaling (Whitmarsh and Davis, 1998). It is probable that such a level of regulation also exists in Arabidopsis, as has been suggested for AtMEKK1 (Ichimura et al., 1998).

In conclusion, we have shown that two of the MAPKs activated by the bacterial elicitor harpin in Arabidopsis cells and leaves are AtMPK4 and AtMPK6, and that H₂O₂ does not have any effect on the activation of AtMPK4. Furthermore, harpin-induced AtMPK6 activation occurs independently of the oxidative burst, implying divergent signaling pathways. AtMPK4 and its potential upstream interacting partners are not regulated at the level of transcription by either harpin or H₂O₂ treatment. The interactions of AtMPK4 and AtMPK6 with other Arabidopsis MAPK signaling pathways and their roles in specific plant-pathogen interactions await elucidation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Treatment of Arabidopsis Suspension Cultures

Cell suspension cultures of Arabidopsis var. Landsberg erecta were maintained as described in Desikan et al. (1996). Harpin from Pseusomonas syringae pv syringae, isolated as described previously (Desikan et al., 1998), and H₂O₂ were added to the cells at the indicated concentrations and times. Controls were represented by mock treatment of cells with appropriate volumes of sterile distilled water. For inhibitor experiments, cells were treated 20 min prior to the addition of harpin with the MAPKK inhibitor PD98059 (Calbiochem, Nottingham, UK) at the indicated concentration. Controls for these experiments involved treating the cells with equivalent amounts of dimethylsulfoxide. Following treatments, Arabidopsis cells were harvested by vacuum filtration and frozen in liquid nitrogen.

Growth and Treatment of Arabidopsis Plants

Arabidopsis var. Landsberg erecta plants were grown in Levington's F₂ compost in a controlled plant growth cabinet (60% [v/v] humidity, 8-h light period, 600 µE m² s¹ at 16°C, and 16-h dark period at 12°C). The seedlings were grown for 5 weeks, after which the shoots were harvested. For treatments, leaves were chopped and incubated in water overnight with gentle shaking, to allow completion of any wound-induced responses. Following this, leaves were vacuum infiltrated for 30 s with either harpin or H₂O₂ at the indicated concentrations, and left incubating in the appropriate solution with gentle shaking. Leaves were then frozen in liquid nitrogen at the appropriate times after infiltration.

Extraction and Immunoblotting of Proteins

Frozen cells or plant leaves (about 0.5 g) were ground with a mortar and pestle using liquid nitrogen, followed by homogenisation at 4°C with 2 volumes of protein extraction buffer {100 mM HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], pH 7.5, 5 mM EDTA, 5 mM EGTA, 10 mM dithiothreitol [DTT], 10 mM Na₃VO₄, 10 mM NaF, 50 mM -glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, 5 µg mL¹ aprotinin, and 5 µg mL¹ leupeptin}. The ground slurry was centrifuged at 12,000g for 20 min at 4°C in a microcentrifuge. Supernatants were aliquotted into clean tubes, snap frozen in liquid nitrogen, and stored at 80°C for later use. Protein concentrations were estimated using the method described by Bradford (1976).

Immunoblotting and detection were performed as described in Desikan et al. (1996) using an enhanced chemiluminescence western blotting detection kit (Amersham, Little Chalfont, UK) with a 1:5,000 or 1:1,000 dilution of the primary antibodies (anti-AtMPK4 and anti-AtMPK6, respectively) and a 1:3,000 dilution of the secondary antibody (peroxidase-conjugated anti-rabbit IgG; Amersham). Generation and characterization of the anti-AtMPK4 (AtMPK4CT, raised against the C-terminal 16 amino acids of AtMPK4) and anti-AtMPK6 (Ab6NT1, raised against the N terminus of AtMPK6) antibodies are described elsewhere (Ichimura et al., 2000)

In-Gel Protein Kinase Assays

Forty micrograms of Arabidopsis protein from cell or leaf extracts were electrophoresed on 10% (w/v) SDS-polyacrylamide gels embedded with 0.5 mg mL¹ MBP from bovine brain (Sigma, Poole, UK) in the resolving gel as substrate for the kinase. Prestained M_r markers (Bio-Rad) were used as standards. After electrophoresing at 100 V for 2 h, SDS was removed from the gel by washing the gel with 100 mL of washing buffer (25 mM Tris-HCl, pH 7.5, 0.5 mM DTT, 0.1 mM Na₃VO₄, 5 mM NaF, 0.5 mg mL¹ bovine serum albumin, and 0.1% [v/v] Triton X-100) three times for 30 min each at room temperature with gentle shaking. The proteins were then denatured by incubating the gel in 100 mL of denaturation buffer (6 M guanidine-HCl, 50 mM Tris-HCl, pH 8, and 5 mM 2-mercaptoethanol) for 1 h at room temperature. The proteins were subsequently renatured overnight at 4°C in 200 mL of renaturation buffer (25 mM Tris-HCl, pH 8, 1 mM DTT, 0.1 mM Na₃VO₄, and 5 mM NaF) with at least three changes of the buffer. The gel was then incubated at room temperature in 30 mL of reaction buffer (25 mM Tris-HCl, pH 8, 2 mM EGTA, 12 mM MgCl₂, 1 mM DTT, and 0.1 mM Na₃VO₄) for 30 min. Phosphorylation was performed for 1 h at room temperature in 15 mL of the same reaction buffer supplemented with 50 µM ATP (Sigma) and 50 µCi [-³²P] ATP (specific activity 3,000 Ci mmol¹; Amersham). Unincorporated radioactivity subsequently was removed by washing the gel for 5 to 6 h at room temperature with several changes of 5% (w/v) trichloroacetic acid and 1% (w/v) sodium pyrophosphate. The gel was dried onto paper (3 MM, Whatman, BDH-Merck, Poole, UK) and subjected to autoradiography.

Immunoprecipitation and In-Gel Kinase Assay

Five hundred (for AtMPK4) or 100 (for AtMPK6) µg of protein from harpin- or H₂O₂-treated cells or 100 µg protein from leaf extracts were incubated by shaking for 2 h at 4°C with 8 µg of anti-AtMPK4 or -AtMPK6 antibody in immunoprecipitation buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM Na₃VO₄, 1 mM NaF, 10 mM -glycerophosphate, 5 µg mL¹ aprotinin, 5 µg mL¹ leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 0.5% [v/v] Triton X-100). Approximately 30 µL packed volume of protein G sepharose (Sigma) was added and incubated for a further 2 h. The sepharose bead-protein complexes were pelleted by gentle centrifugation (1,000g) and subsequently washed twice in wash buffer (20 mM Tris, pH 7.5, 5 mM EDTA, 100 mM NaCl, and 1% [v/v] Triton X-100) and once in kinase assay buffer (25 mM Tris, pH 7.5, 5 mM MgCl₂, 1 mM EGTA, 1 mM DTT, and 0.1 mM Na₃VO₄). Following washing, the immunoprecipitated proteins were released from the immunocomplex by boiling the samples with SDS sample loading buffer. In-gel kinase assay was then performed on the samples using MBP as the substrate, as described above.

RNA Isolation and Northern Analysis

For RNA isolation, Arabidopsis cells were treated with harpin (1 µg mL¹) or H₂O₂ (10 mM) for 2 h, harvested, and frozen in liquid nitrogen. RNA isolation and northern analysis were performed as described by Desikan et al. (1998) with the following modifications. RNA blotted onto nylon membrane was prehybridized and hybridized at 42°C overnight in a formamide buffer containing 5× SSPE (20× SSPE solution contains 3.6 M NaCl, 0.2 M NaH₂PO₄, and 0.02 M Na₂EDTA, pH 7.4), 5× Denhardt's solution (50× Denhardt's solution contains 1% [w/v] polyvinylpyrrolidone, 1% [w/v] bovine serum albumin Fraction V, and 1% [w/v] Ficoll 400), 1% (w/v) SDS, 50 mM NaH₂PO₄, pH 6.8, 10% (w/v) dextran sulfate, 100 µg mL¹ denatured salmon sperm DNA, and 50% (v/v) formamide. cDNA products were obtained as described below. AtMEK1 cDNA (Morris et al., 1997) or a PAL1 genomic clone (Desikan et al., 1998) were used as hybridization probes. Post-hybridization washes were carried out as described by Desikan et al. (1998). The blot was stripped after each hybridization using boiling 0.1% (w/v) SDS, and subsequently used for the next round of hybridization. Equivalent RNA loadings were confirmed by ethidium bromide staining of the gel.

Reverse Transcription-PCR

PCR primers were designed against known DNA sequences of AtMEKK1 (accession no. D50468), AtMKK2 (accession no. AB015313), AtMPK4 (accession no. D21840), and AtMPK6 (accession no.D21842). Primer sequences were: AtMEKK1 (forward), 5' TCGCTCTTTGGAGTTTCCGG-3'; AtMEKK1 (reverse), 5' ATCTGCAAGTTTGACGGCGC- 3'; AtMKK2 (forward), 5' TGATCAGCTGAGCTTGTCGG 3'; AtMKK2 (reverse), 5' ATGGTGATATTATGTCTCCC 3'; AtMPK4 (forward), 5' GCTACAAACTCAGAGACTGG 3'; AtMPK4 (reverse), 5' TTTCACGGTATATAAGCTCC 3'; AtMPK6 (forward), 5' AAACATCTTCGAGGTCACCG 3'; and AtMPK6 (reverse), 5' AAGCTCTGGTGCACGGTACC 3'.

mRNA was isolated from total RNA using Dynabeads Oligo(dT)₂₅ (Dynal, Wirral, UK), as described by the manufacturers. mRNA was reverse transcribed to single-stranded cDNA using SuperscriptII RNase H reverse transcriptase (Gibco-BRL, Paisley, UK) and random hexanucleotides (Pharmacia, Little Chalfont, UK) using the following conditions: 94°C, 5 min; 42°C, 40 min; and 94°C, 5 min. The cDNA thus synthesized was used in a PCR reaction using the designed primers under the following conditions: denaturation at 94°C for 5 min, followed by 35 cycles of 94°C (denaturation) for 45 s; 56°C (annealing), 1 min; 72°C, (extension) 1 min, and an additional extension step of 72°C for 5 min. PCR products of the expected sizes were obtained (922, 543, and 933 bp and 1 kb for AtMPK4, AtMPK6, AtMKK2, and AtMEKK1, respectively), and subsequently were cloned and sequenced to confirm identities. PCR products were gel purified using a Qiaex kit (Qiagen, Crawley, UK) to be used as probes for northern analysis.