Cell Death Mediated by MAPK Is Associated with Hydrogen Peroxide Production in Arabidopsis *

Rapid and localized programmed cell death, known as the hypersensitive response (HR) is frequently associated with plant disease resistance. In contrast to our knowledge about the regulation and execution of apoptosis in animal system, information about plant HR is limited. Recent studies implicated the mitogen-activated protein kinase (MAPK) cascade in regulating plant HR cell death as well as several other defense responses during incompatible interactions between plants and pathogens. Here, we report the generation of transgenic Arabidopsis plants that express the active mutants of AtMEK4 and AtMEK5, two closely related MAPK kinases under the control of a steroid-inducible promoter. Induction of the transgene expression by the application of dexamethasone, a steroid, leads to HRlike cell death, which is preceded by the activation of endogenous MAPKs and the generation of hydrogen peroxide. Both prolonged MAPK activation and reactive oxygen species generation have been implicated in the regulation of HR cell death induced by incompatible pathogens. As a result, we speculate that the prolonged activation of the MAPK pathway in cells could disrupt the redox balance, which leads to the generation of reactive oxygen species and eventually HR cell death.

The active defense mechanisms of plants against invading pathogens often include rapid cell death, known as the hypersensitive response (HR) 1 (1,2). It has long been speculated that the HR cell death and associated dehydration are directly responsible for limiting pathogen growth and development. All pathogens, including fungi, bacteria, and viruses, require a water environment to multiply and acquire nutrients from their hosts. In the case of viral pathogens, cell death/dehydration also affects the plasmodesmata, through which the viruses spread to other cells. In contrast, cell death in a compatible interaction is slow and not associated with quick dehydration. As a result, the pathogen can actually take advantage of the nutrients released during plant cell death and multiply. Recently, an Arabidopsis mutant (dnd1) was isolated that exhibits resistance to certain pathogens in the absence of HR, which suggests that disease resistance can be achieved by more than one mechanism (3). It is also possible that in some cases, a subset of defense responses may be sufficient to stop pathogen growth.
Several lines of evidence suggest that HR cell death during plant disease resistance is a form of programmed cell death because it requires active transcription and translation and is genetically defined (4 -6). Although the details for the regulation and execution of HR remain poorly understood, production of reactive oxygen species (ROS), altered ion fluxes, protein phosphorylation and dephosphorylation, and gene activation have been implicated in the process (7)(8)(9)(10). HR cell death in both suspension cells and plants induced by avirulent pathogens or elicitors can be blocked by kinase inhibitors, suggesting the involvement of protein kinase(s) (11)(12)(13). Potential targets for these kinase inhibitors could be at the recognition step because several R genes such as Pto and Xa21 and the receptor of flagellin, a non-host-specific elicitor, encode protein kinases and/or at the downstream signaling components such as calcium dependent protein kinases and MAPKs (9, 14 -18).
MAPK cascades are one of the major pathways by which extracellular stimuli are transduced into intracellular responses in all eukaryotic cells (18 -23). Increasing evidence suggests that plant MAPK cascade is one of the converging points after the perception of different pathogens and pathogen-derived elicitors (18, 24 -26). The two MAPKs implicated in the process are salicylic acid-induced protein kinase (SIPK) and wounding-induced protein kinase (WIPK) in tobacco (25)(26)(27), AtMPK6 and AtMPK3 in Arabidopsis (28,29), and salt stress-induced MAP kinase (SIMK) and stress-activated MAP kinase (SAMK) in alfalfa (30,31). Based on phylogenetic analysis, SIPK, AtMPK6, and salt SIMK are orthologs, whereas WIPK, AtMPK3, and SAMK are orthologs. Members in the WIPK subfamily are also unique in their up-regulation at the transcriptional level by stresses (15,22,32,33).
Involvement of the SIPK subfamily in the plant defense response was first demonstrated through the biochemical purification and cloning of its encoding gene (27). By using a SIPK-specific antibody, SIPK was later shown to be activated by wounding (34); osmotic and salt stresses (35); non-racespecific elicitors, including fungal cell wall elicitor (36), elicitins from Phytophthora spp. (36), and harpin from the bacterial pathogen Erwinia amylovora (37); and gene-for-gene interactions, including tobacco mosaic virus infection of N. tabacum cv. Xanthi nc tobacco plants (26) and Avr9 treatment of Cf9 transgenic tobacco (25). Although all of the above mentioned stresses activate SIPK, the kinetics of activation is different. Activation of SIPK by abiotic stresses such as wounding and high osmolarity is transient, whereas its activation by pathogens or elicitors that induce cell death is long lasting (15,26,35). It has been speculated that SIPK might function in a common signaling pathway shared by all of these stimuli and, in addition, that the duration of its activation influences the outcome (13,15,36).
More direct evidence for the role of SIPK and WIPK in HR cell death came from a conditional gain-of-function study of NtMEK2, the upstream MAPKK of SIPK and WIPK. Expression of NtMEK2 DD , a constitutively active mutant of NtMEK2, in a transient transformation analysis induces HR-like cell death in tobacco, which is preceded by the activation of endogenous SIPK and WIPK (38). Parallel studies of several other tobacco MAPKKs, which failed to activate SIPK or WIPK and induce cell death, demonstrated the specificity of the NtMEK2-SIPK/WIPK cascade. Using a similar approach, activation of SIPK alone was shown to be sufficient to induce HR-like cell death (39).
In this report, we demonstrated that AtMEK4 and AtMEK5, two closely related Arabidopsis MAPKKs, were functionally interchangeable with NtMEK2 in activating downstream MAPKs. To alleviate the defense responses induced by the Agrobacterium used in the transient transformation assay, we generated permanent transgenic Arabidopsis plants that express wild type, the active mutants, or the inactive mutants of AtMEK4, AtMEK5, or NtMEK2 under the control of a steroidinducible promoter. Activation of the endogenous Arabidopsis MAPKs by the active mutant transgenes under induced conditions leads to HR-like cell death, which is preceded by the generation of hydrogen peroxide. Both prolonged MAPK activation and ROS generation have been implicated in the regulation of HR cell death induced by incompatible pathogens. Our results presented in this report suggest that the prolonged activation of the MAPK pathway in cells could disrupt the redox balance and lead to the generation of ROS, which may eventually lead to HR cell death.

EXPERIMENTAL PROCEDURES
Plants and Growth Conditions-Arabidopsis thaliana (Columbia ecotype) plants were grown at 22°C in a growth chamber with a 12-h photoperiod at a photon flux density of 100 microeinsteins/m Ϫ2 s Ϫ1 . Unless indicated otherwise, fully expended leaves of 4-week-old plants were used for experiments. Dexamethasone (DEX; 15 M) was applied by spraying to induce the transgene expression. Samples were taken at the indicated times after the DEX treatment, quick frozen in liquid nitrogen, and stored at Ϫ80°C until use.
Preparation of Arabidopsis MAPKK Constructs-Arabidopsis At-MEK4 and AtMEK5 cDNAs were obtained by reverse transcription-PCR and cloned into pBlueScript vector (Stratagene). Primers used were 5Ј-CATATGAGACCGATTCAATCGCC-3Ј and 5Ј-ACTAGTAAAA-TTCAGAGACCCTCC-3Ј for AtMEK4 and 5Ј-CATATGAAACCGATTC-AATCTCCTTC-3Ј and 5Ј-ATCCAGCTAAAATCACTCTTAAACCA-3Ј for AtMEK5. The constitutively active and inactive mutants of the MAPKKs were generated by QuickChange site-directed mutagenesis (Stratagene) and confirmed by sequencing. The MAPKKs and their mutants with a FLAG epitope at their N termini were inserted into the XhoI/SpeI sites of the steroid-inducible pTA7002 binary vector (40). The 5Ј-untranslated region of MAPKKs was replaced with the ⍀ sequence from tobacco mosaic virus (41). These constructs were electroporated into Agrobacterium tumefaciens strain C58C1.
Agrobacterium-mediated Transformation-Agrobacterium-mediated transient transformation experiments in tobacco were performed as previously described (38). Six-week-old tobacco plants (Nicotiana tabacum cv. Xanthi nc) grown at 25°C in a growth room programmed for a 14-h light cycle were used for experiments. Permanent transgenic Arabidopsis plants were generated using the flower-dipping method (42). Transgenic lines were selected in the presence of hygromycin.
Protein Extraction and Immunoblot Analysis-Total protein was extracted from leaf tissue by grinding with small plastic pestles in extraction buffer (100 mM HEPES, pH 7.5, 5 mM EDTA, 5 mM EGTA, 10 mM Na 3 VO 4 , 10 mM NaF, 50 mM ␤-glycerophosphate, 10 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 5 g ml Ϫ1 leupeptin, 5 g ml Ϫ1 aprotinin, 5% glycerol). After centrifugation at 18,000 ϫ g for 40 min, supernatants were transferred into clean tubes, quickly frozen in liquid nitrogen, and stored at Ϫ80°C until analyses. The concentration of protein extracts was determined by using the Bio-Rad protein assay kit (Bio-Rad) with bovine serum albumin as a standard.
For immunoblot analysis, 7 g of total protein/lane were separated by electrophoresis on 10% SDS-polyacrylamide gels, and the proteins were transferred to nitrocellulose membranes (Schleicher and Schuell) by semidry electroblotting. After blocking for 1 h in Tris-buffered saline (20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 0.05% Tween 20) with 5% no-fat dried milk (Carnation) at room temperature, the membranes were incubated with anti-FLAG antibody M2 (1:10,000 dilution; Sigma). Following three washes with Tris-buffered saline buffer, the membranes were incubated with horseradish peroxidase-conjugated secondary antibody (1:10,000 dilution; PerkinElmer Life Sciences). The membranes were visualized by using an enhanced chemiluminescence kit (PerkinElmer Life Sciences), following the manufacturer's instructions.
In-gel Kinase Activity Assay-The in-gel kinase activity assay was performed as described previously (27). In brief, 7 g of protein extracts were electrophoresed on 10% SDS-polyacrylamide gels embedded with 0.1 mg ml Ϫ1 myelin basic protein in separating gel as substrate for 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 dithiothreitol, 0.1 mM Na 3 VO 4 , 5 mM NaF, 0.5 mg ml Ϫ1 bovine serum albumin, and 0.1% Triton X-100) three times for 30 min each at room temperature. The proteins were then renatured in 25 mM Tris-HCl, pH 7.5, 1 mM dithiothreitol, 0.1 mM Na 3 VO 4 , 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 MgCl 2 , 1 mM dithiothreitol, 0.1 mM Na 3 VO 4 ) for 30 min. Phosphorylation was performed for 1.5 h at room temperature in 30 ml of the same reaction buffer with 200 nM ATP plus 50 Ci of [␥-32 P]ATP (3000 Ci/mmol). The reaction was stopped by transferring the gel into solution with 5% trichloroacetic acid (w/v) and 1% sodium pyrophosphate (w/v). The unincorporated radioactivity was subsequently removed by washing the gel for 6 h at room temperature with five changes. The gel was dried on 3MM paper (Whatman) and subjected to autoradiography. Prestained size markers (Bio-Rad) were used to calculate the size of kinases.
Immunoprecipitation Kinase Assay-For immunoprecipitation kinase assays, protein extract (100 g) with or without peptide competitors (5 g/ml) was incubated with SIPK-specific antibody, Ab-p48N (2.5 g), and WIPK-specific antibody, Ab-p44N (2.5 g), that were raised against peptides corresponding to the N termini of SIPK and WIPK, respectively (26), in the immunoprecipitation buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM Na 3 VO 4 , 1 mM NaF, 10 mM ␤-glycerophosphate, 2 g/ml antipain, 2 g/ml aprotinin, 2 g/ml leupeptin, and 0.1% Tween 20) at 4°C for 2 h on a rocker. Approximately 20 l (packed volume) of protein AϪagarose was added, and the incubation was continued for another 4 h. Agarose beadϪprotein complexes were pelleted by brief centrifugation and washed three times with 1.5 ml of immunoprecipitation buffer each. Kinase activity in the complex was determined by an in-gel kinase assay as described above.
H 2 O 2 Detection by 3,3Ј-Diaminobezidine Uptake Method-H 2 O 2 was detected by an endogenous peroxidase-dependent in situ histochemical staining procedure using 3,3Ј-diaminobenzidine (DAB) (43). Leaves treated with DEX were detached and placed in a solution containing 1 mg/ml DAB (pH 5.5) for 2 h. The leaves were boiled in ethanol (96%) for 10 min and then stored in 96% ethanol. H 2 O 2 production is visualized as a reddish-brown coloration.

Arabidopsis AtMEK4 and AtMEK5
Are Functionally Interchangeable with Tobacco NtMEK2-Prolonged activation of SIPK and/or delayed activation of WIPK are associated with HR cell death elicited by pathogens or pathogen-derived elicitors (15,26). Our recent studies demonstrated that the activation of SIPK and WIPK by NtMEK2 DD , an active mutant of NtMEK2, induced HR-like cell death in a transient transformation system (38). However, the presence of agrobacterial cells in leaves by themselves induced a number of defense responses, which could complicate the interpretation of the data (44). 2 The generation of permanent transgenic plants will circumvent this shortcoming. To take advantage of the simple transformation procedure, we decided to generate transgenic Arabidopsis plants using its NtMEK2 orthologs.
A search of fully sequenced Arabidopsis genome revealed a total of 10 MAPKKs (Fig. 1A). Phylogenetic analysis using the Clustal method identified AtMEK4 and AtMEK5, two Arabidopsis MAPKKs that share the highest homology with Nt-MEK2. Their identities in amino acid sequence to NtMEK2 are 67.5 and 69.3%, respectively. The function of neither Arabidopsis MAPKK has been defined. To examine whether these two Arabidopsis MAPKKs can substitute NtMEK2 to activate downstream MAPKs, we generated active mutants of both At-MEK4 and AtMEK5 by mutating the conserved Ser/Thr in the activation loop ((S/T)XXXXX(S/T)) of plant MAPKKs to Asp (Fig. 1B) (38). Agrobacterial cells that carry AtMEK4 DD (T224D/S230D) and AtMEK5 DD (T215D/S221D) in steroid-inducible vector pTA7002 were transiently transformed into tobacco. The inactive mutants, in which the catalytic essential Lys in the kinase domain was mutated to Arg, AtMEK4 KR (K108R) and AtMEK5 KR (K99R) were used as controls. To facilitate the detection of transgene expression, a FLAG epitope tag was attached to the N termini of the transgenes.
Newly fully expanded tobacco leaves were infiltrated with agrobacterial cells that carry various constructs, and the transgene expression was induced by the application of DEX (30 M) 2 days later. As shown in Fig. 2A, leaf sections expressing the active Arabidopsis MAPKK mutants, AtMEK4 DD or AtMEK5 DD showed high level activation of two endogenous MAPKs. Their molecular masses of 48 and 44 kDa correspond to the sizes of SIPK and WIPK, respectively. SIPK-and WIPKspecific antibodies pulled down these two kinase activities as demonstrated by immunoprecipitation kinase assay (Fig. 2B). The addition of peptide competitors, p48N and p44N, to which the antibodies were raised, effectively blocked the immunoprecipitation, demonstrating the specificity of the assay. As a positive control, experiments using NtMEK2 DD , which has been shown to activate SIPK and WIPK (38), were performed in parallel. The expression of wilt-type proteins, AtMEK4 WT and AtMEK5 WT , or their inactive mutants, AtMEK4 KR and AtMEK5 KR , failed to activate any MAPK as determined by in-gel kinase activity assay; even the proteins were expressed at similar levels ( Fig. 2A). This is again in good agreement with The activation of endogenous MAPKs was determined by an in-gel kinase assay (bottom). B, immunoprecipitation (IP) kinase assays using SIPK-and WIPK-specific antibodies demonstrated that the 48-and 44-kDa kinases are encoded by SIPK and WIPK, respectively. Protein extracts (100 g) from leaf tissues transformed with NtMEK2 DD , AtMEK4 DD , or AtMEK5 DD were immunoprecipitated with Ab-p48N (2.5 g) and Ab-p44N (2.5 g) in the absence (Ϫ) or the presence (ϩ) of competitor peptides, p48N (5 g/ml) and p44N (5 g/ml). Kinase activities in the immune complex were subsequently determined by an in-gel kinase assay using myelin basic protein as a substrate. C, activation of endogenous SIPK and WIPK by the active mutants of AtMEK4 DD , AtMEK5 DD , and NtMEK2 DD leads to HR-like cell death. This photograph was taken 36 h after DEX treatment.
previous results obtained using NtMEK2 WT and NtMEK2 KR ( Fig. 2A) (38). The sizes of all proteins matched the predicted sizes based on sequences, with AtMEK5 the smallest one and AtMEK4 and NtMEK2 about the same size ( Fig. 2A). It was interesting to note that all of the active mutant proteins showed an up-shift on the SDS-PAGE, which is frequently associated with the phosphorylated forms of proteins. Within 16 h after the induction of AtMEK4 DD , AtMEK5 DD , or NtMEK2 DD expression, cell death in small areas could be seen. By 24 h, the whole area infiltrated with agrobacterial cells carrying AtMEK4 DD , AtMEK5 DD or NtMEK2 DD collapsed (Fig.  2C). No cell death was observed in leaf sections transformed with AtMEK4 WT , AtMEK4 KR , AtMEK5 WT , or AtMEK5 KR (Fig.  2C). As another control, the transformation of empty pTA7002 vector did not elicit any of the above phenotype either. These results demonstrated that Arabidopsis AtMEK4 and AtMEK5 are able to substitute tobacco NtMEK2 to activate SIPK and WIPK. This also provided additional evidence that the conserved Ser/Thr in the plant MAPKK activation motif is separated by five amino acids instead of the three amino acids found in MAPKKs from yeast and animals (19,45).

Expression of AtMEK4 DD and AtMEK5 DD in Arabidopsis Leads to a Long Lasting Activation of two MAPKs and the Death of Plants-After the relationship between Arabidopsis
AtMEK4 and AtMEK5 and tobacco NtMEK2 was established, we transformed all the constructs into Arabidopsis by flower dipping method (42). Hygromycin-resistant T1 transgenic plants were selected. A total of 78 AtMEK4 DD , 75 AtMEK5 DD , and 48 NtMEK2 DD transgenic lines were obtained. Among them, 15 AtMEK4 DD , 17 AtMEK5 DD , and 12 NtMEK2 DD lines showed transgene expression upon DEX (15 M) treatment of detached leaves. All of them showed HR-like cell death and dehydration of the tissue, which were preceded by the activation of two MAPKs (data not shown). In contrast, a similar number of wild type or inactive mutant MAPKK transgenic plants failed to activate endogenous MAPKs and show cell death phenotype; even the transgenes were expressed at similar levels as determined by anti-FLAG antibody ( Table I). The correlation between the expression of the active mutants and the HR-like cell death phenotype was also examined in the T2 generation with selected lines that contain a single copy of the transgene. Expression of transgene in randomly selected T2 plants, eight for the active mutant (DD) lines and five for the inactive mutant (KR) lines, were determined after the application of DEX. As shown in Fig. 3, all plants that carried the active mutant transgenes showed elevated levels of endogenous MAPK activities (Fig. 3, middle), which is followed by HR-like cell death (Fig. 3, bottom). In contrast expression of the inactive MAPKK mutants failed to activate endogenous MAPKs and induce cell death. In Arabidopsis, anti-FLAG an-tibody recognized a nonspecific band of about 45 kDa, which is between the sizes of FLAG-tagged AtMEK4 and AtMEK5 (Figs. 3 and 4A).
After the correlation and co-segregation between the MAPK activation by the active AtMEK4 or AtMEK5 and HR-like cell  death was established in T1 and T2 generations, detailed time course analyses were performed using T3 homozygous transgenic plants. As shown in Fig. 4A, the transgene expression was visible ϳ2 h after the application of DEX, which was accompanied by the increase of MAPK activities as revealed by an in-gel kinase activity assay. HR-like cell death appeared about 24 h after the DEX application, which always started from the tip and edge of the leaves and progressed inwards until the whole leaf died (Fig. 4B). Interestingly, the dead leaves stayed green and became brittle just as those in HR cell death elicited by incompatible pathogens. This rapid dehydration event is believed to play an important role in plant resistance, because the multiplication of pathogens requires the presence of water.

H 2 O 2 Generation Is Associated with HR-like Cell Death
Induced by AtMEK4 DD , AtMEK5 DD , or NtMEK2 DD in Arabidopsis-Reactive oxygen species have been shown to play a number of critical roles in defense responses including HR cell death during plant-pathogen interactions (8,12,46). To test whether H 2 O 2 production was involved in the cell death induced by the activation of this MAPK pathway, we used the DAB staining method to detect H 2 O 2 . As shown in Fig. 5A, the reddish brown-colored precipitants of oxidized DAB, an indication of H 2 O 2 production, were visible in leaves from AtMEK4 DD , AtMEK5 DD , or NtMEK2 DD transgenic plants treated with DEX inducer. As controls, leaves from AtMEK4 KR , AtMEK5 KR , or NtMEK2 KR transgenic plants were treated with DEX and processed side by side. No H 2 O 2 generation was observed in plants expressing the inactive MAPKKs. The DAB staining was the highest in tissues that were apparently healthy but were close to the collapsed tissues. However, after the whole leaf collapsed, it was no longer stainable with DAB. These data indicate that the DAB-detectable H 2 O 2 accumulation immediately preceded the cell death. Wild type Columbia ecotype plants inoculated with incompatible bacterial pathogen Pseudomonas syringae pv. tomato DC3000 containing the avr-Rpt2 avirulence gene also produced similar levels of H 2 O 2 (Fig.  5B). In this case, H 2 O 2 production was also detectable soon before the collapse of the tissues and was limited to the area infiltrated with pathogens. A closer examination of stained leaves under the microscope revealed that the oxidized DAB precipitants were localized inside cells in both cases, indicating the intracellular origin of H 2 O 2 (Fig. 5C). These results suggest that the DAB-detectable H 2 O 2 generation is a late event in both MAPK activation-and pathogen-induced cell death and may directly contribute to the cell death process.

DISCUSSION
Plants have evolved both preexisting and inducible defense mechanisms to protect themselves against invading pathogens. The inducible defense mechanisms of plants often include the generation of ROS, the activation of a complex array of defense genes, and rapid and localized HR cell death (8,9,47). In addition, the immunity or systemic acquired resistance to subsequent attack by a broad range of pathogens develops throughout the plant (48,49). The activation of these defense responses is initiated by the plant recognition of pathogens, which is mediated either by a gene-for-gene interaction between the products of a plant resistance (R) gene and a pathogen avirulence (Avr) gene or by the binding of a non-race-specific elicitor such as flagellin to its receptor (9, 16, 50 -52). Signals generated by such interactions are transduced into cellular responses via several interlinked pathways (8,9). In this report, we demonstrated that two Arabidopsis MAPKKs, AtMEK4 and AtMEK5, are functionally interchangeable with tobacco NtMEK2 in activating the downstream MAPKs. In transient transformation experiments performed in tobacco, the active forms of AtMEK4 and AtMEK5 activate endogenous tobacco SIPK and WIPK. These two MAP-KKs, as well as tobacco NtMEK2 also activate two endogenous MAPKs in permanently transformed Arabidopsis, which is followed by HR-like cell death of the Arabidopsis plants. Very interestingly, the appearance of cell death in AtMEK4 DD , AtMEK5 DD , or NtMEK2 DD transgenic plants under induced condition is preceded by the generation of hydrogen peroxide. Both prolonged MAPK activation and ROS generation have been implicated in the regulation of HR cell death induced by incompatible pathogens. Our results presented in this report suggest that the prolonged activation of MAPK pathway in cells cause redox imbalance and the generation of ROS, which may eventually lead to HR cell death.
The two Arabidopsis MAPKs involved are likely to be At-MPK6 and AtMPK3 based on phylogenetic analysis and their sequence homology to tobacco SIPK and WIPK (18). Direct biochemical evidence, however, is difficult to obtain because member-specific antibody is not available for Arabidopsis MAPKs. In tobacco, we consistently see the activation of SIPK and WIPK by the expression of NtMEK2 DD , AtMEK4 DD , or AtMEK5 DD (Fig. 2) (38). However, the activity of the smaller 44-kDa kinase in Arabidopsis was variable. Whether this is because of variation in its activation in vivo or artifact during sample preparation in vitro is currently unknown. Similar to SIPK, AtMPK6 has been shown to be activated by several pathogen-derived elicitors (23,53,54). In contrast, the identity of the smaller 44-kDa MAPK activated by stresses in Arabidopsis is not as clear. Experiments using epitope-tagged MAPKs in transiently transformed protoplasts demonstrated that this MAPK is likely to be encoded by AtMPK3 (23,29). However, immune complex kinase assays of protein extracts from Arabidopsis cells treated with harpin, a bacterium-derived elicitor, demonstrated that the smaller MAPK is encoded by AtMPK4 (54). This is apparently contradictory to the evidence from a genetic study, which demonstrated that AtMPK4 is a negative regulator of plant defense response (55). Loss of functional AtMPK4 in a transposon-insertional mutant leads to the constitutively activated defense responses, which predicts that the AtMPK4 activity should be down-regulated by pathogens or pathogen-derived elicitors during plant defense response (55). It is also possible that the smaller kinase band detected by the in-gel kinase assay is composed of both At-MPK3 and AtMPK4.
Reactive oxygen species play important roles in programmed cell death in both animals and plants (12,46,56). Plant cells challenged with HR-inducing pathogens display two peaks of H 2 O 2 generation (2, 10, 46). The first dramatic transient burst (phase I) of H 2 O 2 production, which occurs within minutes after the perception of pathogens, is not specific to HR, because compatible pathogens that do not induce HR also elicit this peak. Treatment of cell suspension with a number of elicitors that do not cause HR-like cell death also induces this H 2 O 2 burst. As a result, it is believed that the phase I production of H 2 O 2 , which is one of the earliest responses in plants resisting pathogen infections, is nonspecific and related to a generalized induction of defense responses that are not correlated with the induction of HR (2,46). However, there is a second increase of H 2 O 2 (phase II) that precedes the onset of HR cell death and is specific for HR-inducing pathogens.
The phase I H 2 O 2 generation is not regulated by MAPK, although a lot of circumstantial evidence points to this possibility. The activation of MAPK precedes the generation of phase I H 2 O 2 burst. In addition, the broad spectrum kinase inhibitors K-252a and staurosporine were found to block this oxidative burst (46,57), which correlates with the suppression of SIPK activation (36). 3 However, using PD98059, a specific inhibitor of MAPKKs, Tina Romeis et al. (25) demonstrated that H 2 O 2 burst does not require MAPK activation. Furthermore, our effort to detect the H 2 O 2 generation soon after the activation of SIPK and WIPK in tobacco transformed with active NtMEK2 failed to demonstrate its presence (38). As a result, we concluded that MAPK is not involved in the regulation of initial phase I H 2 O 2 burst.
In this study, we demonstrated that the phase II H 2 O 2 generation might be a result of prolonged activation of a MAPK pathway. Both the activation of MAPK and the phase II H 2 O 2 generation precede and correlate with the HR cell death, suggesting that MAPK-induced HR-like cell death might be mediated by the H 2 O 2 generation. Several possible mechanisms for the generation of H 2 O 2 during plant defense responses have been reported, including plasma membrane-located NADPHdependent oxidases, cell wall peroxidases, and apoplastic amine, diamine, and polyamine oxidase-type enzymes (46,58). Mitochondria, chloroplasts, and peroxisomes may also contribute to the generation of ROS (6, 46, 59 -61). The mechanism of H 2 O 2 generation preceding the HR-like cell death induced by the activation of MAPK is not clear. It is also possible that multiple mechanisms are involved.
How the NtMEK2-SIPK/WIPK cascade or its analogous pathways in other plant species are integrated into the HR cell death signaling pathway(s) remains unclear at this stage. In fully sequenced Arabidopsis genome, no typical homolog of the animal apoptosis regulators or executioners could be identified, suggesting that either plant programmed cell death is regulated by different mechanisms or the functional homologs in plants bear little sequence identity to their animal counterparts (62). The activation of SIPK and WIPK or their orthologs in other plant species by a number of stress stimuli is analogous to the activation of mammalian c-Jun N-terminal kinase and p38 MAPKs (18,20,21). Recently, it was shown that c-Jun N-terminal kinase is involved in regulating cytochrome c release in the mitochondrial death signaling pathway (63). It will be interesting to find out if a similar pathway is also operative in plants. We are very intrigued by the functional similarity between animal c-Jun N-terminal kinase/p38 and plant SIPK/ WIPK, because, based on molecular evolutionary analysis, all plant MAPKs are evolved from the ancient MAPK that also gave rise to mammalian extracellular signal-regulated kinase subfamily of MAPKs (64). As a result, the analogous functions 3 Y. Liu and S. Zhang, unpublished results. of plant MAPKs in regulating cell death probably evolved independently from mammalian c-Jun N-terminal kinases and p38s.