Biochemical properties of two protein kinases involved in disease resistance signaling in tomato.

In tomato plants, resistance to bacterial speck disease is mediated by a phosphorylation cascade, which is triggered by the specific recognition between the plant serine/threonine protein kinase Pto and the bacterial AvrPto protein. In the present study, we investigated in vitro biochemical properties of Pto, which appears to function as an intracellular receptor for the AvrPto signal molecule. Pto and its downstream effector Pti1, which is also a serine/threonine protein kinase, were expressed in Escherichia coli as maltose-binding protein and glutathione S-transferase fusion proteins, respectively. The two kinases each autophosphorylated at multiple sites as determined by phosphopeptide mapping. In addition, Pto and Pti1 autophosphorylation occurred via an intramolecular mechanism, as their specific activity was not affected by their molar concentration in the assay. Moreover, an active glutathione S-transferase-Pto fusion failed to phosphorylate an inactive maltose-binding protein-Pto(K69Q) fusion excluding an intermolecular mechanism of phosphorylation for Pto. Pti1 phosphorylation by Pto was also characterized and found to occur with a Km of 4.1 microM at sites similar to those autophosphorylated by Pti1. Pto and the product of the recessive allele pto phosphorylated Pti1 at similar sites, as observed by phosphopeptide mapping. This suggests that the inability of the kinase pto to confer resistance to bacterial speck disease in tomato is not caused by altered recognition specificity for Pti1 phosphorylation sites.

In tomato plants, resistance to bacterial speck disease is mediated by a phosphorylation cascade, which is triggered by the specific recognition between the plant serine/threonine protein kinase Pto and the bacterial AvrPto protein. In the present study, we investigated in vitro biochemical properties of Pto, which appears to function as an intracellular receptor for the AvrPto signal molecule. Pto and its downstream effector Pti1, which is also a serine/threonine protein kinase, were expressed in Escherichia coli as maltose-binding protein and glutathione S-transferase fusion proteins, respectively. The two kinases each autophosphorylated at multiple sites as determined by phosphopeptide mapping. In addition, Pto and Pti1 autophosphorylation occurred via an intramolecular mechanism, as their specific activity was not affected by their molar concentration in the assay. Moreover, an active glutathione S-transferase-Pto fusion failed to phosphorylate an inactive maltose-binding protein-Pto(K69Q) fusion excluding an intermolecular mechanism of phosphorylation for Pto. Pti1 phosphorylation by Pto was also characterized and found to occur with a K m of 4.1 M at sites similar to those autophosphorylated by Pti1. Pto and the product of the recessive allele pto phosphorylated Pti1 at similar sites, as observed by phosphopeptide mapping. This suggests that the inability of the kinase pto to confer resistance to bacterial speck disease in tomato is not caused by altered recognition specificity for Pti1 phosphorylation sites.
Higher plants have evolved the ability to recognize and resist invading pathogens by the activation of defense mechanisms that inhibit pathogen growth and movement in the plant (1). In many plant-pathogen interactions, the rapid activation of the defense response is mediated by a specific recognition event involving the product of an avirulence (avr) 1 gene in the pathogen and the corresponding resistance (R) gene in the plant (2).
Proteins encoded by R genes are postulated to function as receptors that bind the cognate avr gene product and activate defense mechanisms. Several R genes have been isolated to date, and structural characteristics of their encoded proteins support the proposed receptor function (3). Most R genes encode cytoplasmic or transmembrane proteins containing a region of leucine-rich repeats of variable length and content, which might be involved in protein-protein interactions (4). Some leucine-rich repeat-containing R proteins, including those encoded by the tobacco N, the flax L6, and the Arabidopsis RPP5 genes, also contain a region of homology with the cytoplasmic domains of the Drosophila Toll and the mammalian interleukin-1 receptor proteins (4).
Another class of R genes is represented by Pto, which encodes a tomato serine/threonine protein kinase and confers resistance specifically to the bacterial pathogen Pseudomonas syringae pv. tomato expressing the avirulence gene avrPto (5). Pto does not share motifs with other resistance genes, except for the kinase domain, which is also present in the product of the rice gene Xa21, conferring resistance to the bacterial pathogen Xanthomonas oryzae pv. oryzae (6). Recent studies have shown, using the yeast two-hybrid system, that a direct interaction occurs between the product of the tomato Pto resistance gene and the product of the P. syringae pv. tomato avrPto gene (7,8). Mutations of Pto or AvrPto that interfere with this interaction in yeast also abolish disease resistance in plants. In addition, for resistance to take place, Pto and AvrPto must be present simultaneously inside the plant cell, strongly suggesting that the Pto-AvrPto recognition is an intracellular event. AvrPto and other bacterial avr gene products are thought to be delivered into the plant cell by a type III secretion system encoded by the bacterial pathogen (9). In fact, a set of genes in P. syringae shows similarity to genes encoding components of the type III secretion system of mammalian pathogens, and is required both for resistance and pathogenicity (10).
The recognition event between the tomato Pto kinase and the bacterial AvrPto protein initiates a signal transduction pathway that involves downstream effectors and ultimately leads to disease resistance. The Prf gene is required for Pto-mediated resistance and encodes a protein with a leucine zipper, a nucleotide binding site, and leucine-rich repeats, common motifs in other resistance gene products (11). However, the role of Prf in the Pto pathway remains unclear. Other putative effectors were identified by their specific interaction with the Pto kinase in the yeast two-hybrid system (12,13). Among them are Pti1, a serine/threonine protein kinase that is specifically phosphorylated in vitro by Pto and is involved in the hypersensitive response (12), and Pti4, Pti5, and Pti6, putative transcription factors that are similar to the tobacco ethylene-responsive element-binding proteins (13,14).
In mammals, autophosphorylation activity plays a central role in the regulation of receptor tyrosine kinases (15). Hormones, or growth and differentiation factors, bind to receptors with tyrosine kinase activity and induce conformational alterations in the receptor extracellular domains causing oligomer-ization. Autophosphorylation by intermolecular phosphorylation in the receptor cytoplasmic domain then takes place and modulates the interaction between the activated receptor and cellular proteins (16). Induction of autophosphorylation activity by extracellular signals has also been observed in cytoplasmic protein kinases associated with receptors. For instance, autophosphorylation of the Janus kinases is activated by cytokines through cytokine receptors (17), and components of the extracellular matrix determine the integrin-mediated activation of autophosphorylation of the focal adhesion kinase FAK (18).
Because the tomato Pto serine/threonine protein kinase is thought to function as an intracellular receptor or as part of a receptor complex for the AvrPto signal molecule, we were interested in determining the role of Pto autophosphorylation activity in the elicitation of the defense response by the Pto-AvrPto interaction. As a step toward testing the possibility that AvrPto mediates Pto dimerization and activation by intermolecular autophosphorylation, we characterized the biochemical properties of Pto autophosphorylation and phosphorylation of its putative effector Pti1. Pto and Pti1 were expressed in Escherichia coli as fusion proteins and assayed for their kinase activities. Pto and Pti1 autophosphorylated at multiple sites, and their autophosphorylation proceeded via an intramolecular mechanism. We also determined the kinetics of Pti1 phosphorylation by Pto and compared the tryptic digests of Pti1 phosphorylated by Pto and by the product of the recessive pto allele.

Expression of Fusion Proteins in E. coli-For expression in bacteria,
the Pto protein kinase and its mutagenized kinase-deficient form Pto(K69Q) were fused in frame to the C terminus of the maltose-binding protein (MBP), as described previously (19). The Pto protein was also expressed as a fusion to the C terminus of glutathione S-transferase (GST). To develop the GST-Pto fusion, primers YTL8 (5Ј-ATGGGATC-CAAGTATTCTAAGGCA-3Ј) and YTL5 (5Ј-CCCTGCAGTGAAAGAAG-GATCCACAG-3Ј) were used to amplify a 1,057-base pair product encoding Pto from plasmid PTC3 (8). Primer YTL8 introduced a BamHI restriction site at the N terminus of the Pto gene, and primer YTL5 amplified a BamHI restriction site 70 base pairs after the TAA stop codon. The polymerase chain reaction product was digested by BamHI and inserted into vector pGEX-4T1 (Amersham Pharmacia Biotech) at the corresponding site. The constructs for expression of the kinase encoded by the recessive pto allele as a MBP fusion, and of the Pti1 protein kinase and its mutagenized form Pti1(K96N) as GST fusions, were prepared as described previously (12,20). MBP and GST fusions were expressed in the E. coli strain PR745 (New England Biolabs) and affinity-purified by using amylose resin (New England Biolabs) and glutathione-agarose beads (Sigma), respectively.
Kinase Activity Assays-Autophosphorylation activity of MBP-Pto and GST-Pti1 fusion proteins was assayed at different protein concentrations in 25 l of reaction buffer (50 mM Tris-HCl, pH 7.0, 1 mM dithiothreitol, 10 mM MnCl 2 , 1 mg/ml bovine serum albumin, and 20 M ATP), containing 4 Ci of [␥-32 P]ATP (6000 Ci/mmol; Amersham Pharmacia Biotech). Reactions were incubated for 15 min at room temperature and spotted on phosphocellulose P-81 filters (Whatman). The filters were extensively washed by three changes of 150 mM phosphoric acid, and by a final wash with ethanol. Phosphate incorporation was then measured using a scintillation counter (LS 6500, Beckman). The range of concentrations tested was from 0.72 to 17.9 M and from 0.06 to 8.61 M for MBP-Pto and GST-Pti1, respectively. Reactions containing the highest amount of enzyme were found to be linear over the entire reaction time (15 min).
Assays to test Pto intermolecular autophosphorylation were performed with the GST-Pto fusion protein immobilized on glutathioneagarose beads and the other components dissolved in solution. GST-Pto fusion protein (2 g) was mixed with MBP-Pto(K69Q) or GST-Pti1(K96N) fusion proteins (2 g) in 30 l of reaction buffer (described above), containing 3 Ci of [␥-32 P]ATP (6000 Ci/mmol; Amersham Pharmacia Biotech). Reactions were incubated for 10 min at room temperature and stopped by adding EDTA to a final concentration of 10 mM. Proteins were then fractionated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and stained by Coomassie Brilliant Blue R250.
The stained gel was dried and exposed to x-ray film.
To determine V max and K m values of the MBP-Pto fusion for its substrate GST-Pti1(K96N) fusion protein, 0.5 g of MBP-Pto was incubated at room temperature with different amounts of GST-Pti1(K96N) in 20 l of kinase buffer containing 2 Ci of [␥-32 P]ATP (6000 Ci/mmol; Amersham Pharmacia Biotech). The range of GST-Pti1(K96N) concentrations tested was from 0.72 to 17.9 M. Reactions were stopped after 10 min by adding EDTA to a final concentration of 10 mM. At this reaction time, phosphate incorporation was found to be linear for the highest substrate concentration used in the experiment. Proteins were then fractionated by SDS-PAGE, stained by Coomassie Brilliant Blue R250, and analyzed by Instant Imager (Packard Corp.). Values for the V max and K m parameters were estimated by nonlinear least squares fitting using SigmaPlot 4.0 for Windows software (SPSS Inc.).
Phosphopeptide Mapping-The procedure used for phosphopeptide analysis was essentially as described by Van der Geer et al. (21) with some modifications. The phosphorylated protein to be analyzed was fractionated on a SDS-polyacrylamide gel and transferred to a polyvinylidene difluoride membrane, and the corresponding piece of membrane was excised. After blocking with 100 mM acetic acid containing 0.5% polyvinyl pyrrolidone for 30 min, the membrane was extensively washed with water and twice with 50 mM ammonium bicarbonate, pH 8.2. The membrane-bound protein was then incubated in 200 l of 50 mM ammonium bicarbonate, pH 8.2, with 10 g of N-tosyl-L-phenylalanine chloromethyl ketone-treated trypsin for at least 4 h at 37°C. The sample was lyophilized to dryness and oxidized by performic acid for 60 min at room temperature. After water dilution and evaporation of performic acid in a centrifugal vacuum concentrator, the sample was resuspended in electrophoresis buffer containing n-butanol, pyridine, acetic acid, and water at a ratio of 2:1:1:36, and separated for 40 min by thin layer electrophoresis at pH 4.7 and 1.0 kV on 20 ϫ 20-cm cellulose TLC plates (EM Science). The first dimension fractionation was followed by ascending chromatography in phosphopeptide buffer containing n-butanol, pyridine, acetic acid, and water at a ratio of 15:10:3:12. After electrophoresis, the radioactive species were detected by autoradiography.
Polyacrylamide Gel Electrophoresis of Tryptic Digests-Phosphopeptides were electrophoretically separated on alkaline 40% polyacrylamide gels as described by Dadd et al. (22).

Pto and Pti1 Autophosphorylation Occur through an
Intramolecular Mechanism-The Pto and Pti1 proteins are involved in resistance to bacterial speck disease in tomato plants and have been shown previously to be active serine/threonine kinases when tested in vitro (12,19). To further characterize their kinase activity, we tested whether Pto and Pti1 autophosphorylation mechanisms are intramolecular (first order with respect to enzyme concentration) or intermolecular (second order with respect to enzyme concentration). Pto and Pti1 were expressed in bacteria as MBP (MBP-Pto) and GST (GST-Pti1) fusion proteins, respectively, and the effect of various molar concentrations on the autophosphorylation reaction was studied. As shown in Figs. 1A and 2A, the rate of autophosphorylation was linear with respect to enzyme concentration for both MBP-Pto and GST-Pti1. In addition, the phosphate incorporation per molecule was constant when MBP-Pto concentration in the reaction varied by 60-fold (Fig. 1B). Similarly, the phosphate incorporation per GST-Pti1 molecule varied by only 1.7 when the enzyme concentration in the reaction varied by 140fold (Fig. 2B). Finally, the van't Hoff plot of autophosphorylation (logarithm of phosphorylation rate versus logarithm of enzyme concentration), whose slope indicates the order of the reaction, had a slope of 1.11 Ϯ 0.038 and 0.90 Ϯ 0.010 for MBP-Pto and GST-Pti1, respectively (Figs. 1C and 2C). Taken together, these data indicate that both MBP-Pto and GST-Pti1 autophosphorylation occur predominantly via an intramolecular mechanism.
To provide further evidence for intramolecular autophosphorylation by Pto, we tested if an active GST-Pto fusion protein can phosphorylate an inactive MBP-Pto molecule in which the invariant lysine residue in kinase subdomain II was substi-tuted by a glutamine (19). As shown in Fig. 3, the GST-Pto fusion protein was able to autophosphorylate and to phosphorylate its substrate Pti1 as observed previously (12). However, it failed to phosphorylate the inactive mutant protein MBP-Pto(K69Q), strongly supporting the notion that Pto autophosphorylation occurs through an intramolecular rather than intermolecular mechanism.
Kinetic Analysis of Pti1 Phosphorylation by Pto-The Pti1 protein kinase was shown previously to be specifically phosphorylated in vitro by the Pto protein kinase (12). To study the kinetics of this reaction, the initial velocity of the phosphoryl- were resolved horizontally by thin layer electrophoresis at pH 4.7, and vertically by ascending chromatography. As shown in Fig. 5, both digestion of autophosphorylated MBP-Pto and GST-Pti1 generated one major and at least four minor phosphopeptides (Fig. 5, A and B). The ratio between the intensity of the major spot and minor spots was at least 5:1 for MBP-Pto digests, and 50:1 for GST-Pti1 digests. The variable intensity of the labeling might be a result of the degree of enzyme affinity for different sites and to the number of phosphorylation sites present in each peptide. Additional phosphopeptides were detectable after extensive exposures of the TLC plates to x-ray films and may indicate either the presence of sites in the proteins phosphorylated at a very low efficiency or, more likely, partial tryptic digestion products.
It has been shown recently that Pti1 is a substrate not only for the Pto kinase, but also for the translation product of the recessive allele pto, which is a functional protein kinase but does not confer bacterial speck resistance to tomato plants (20). To compare the Pti1 autophosphorylation sites to those phosphorylated by Pto or pto, the kinase-deficient mutant GST-Pti1(K96N) was phosphorylated by an MBP-Pto or MBP-pto fusion proteins and then digested by trypsin. Phosphopeptide maps of the Pti1 digestion products revealed the presence of one major phosphorylated peptide, similar to that observed in Pti1 autophosphorylation reactions (Fig. 5, B-D). To test whether this major spot was derived from the same peptide phosphorylated by Pti1 autophosphorylation and by Pto and recessive pto phosphorylation, tryptic digests from the three reactions were fractionated by alkaline electrophoresis. As shown in Fig. 6, the main phosphorylated peptide in all the reactions showed the same molecular weight and charge characteristics, suggesting that the same phosphorylation site(s) may be utilized by the three enzymes. Similar minor spots were observed in the phosphotryptic maps of autophosphorylated Pti1 and of Pti1 phosphorylated by Pto or recessive pto (Fig. 5,  B-D). Their intensity was variable in different experiments, and analysis of their sequence will be required to determine whether they represent phosphorylation sites, degradation products, or partial digests of the main phosphorylated peptide.

DISCUSSION
In this report, we characterized biochemical properties of two tomato serine/threonine protein kinases, Pto and Pti1, which are involved in the signaling pathway leading to resistance to the bacterial pathogen P. syringae pv. tomato expressing the avrPto gene. A specific recognition between the Pto kinase and the bacterial AvrPto protein occurs within the plant cell and triggers the activation of the pathway, defining the Pto kinase as an intracellular receptor or as part of a receptor complex (7,8). In order to investigate the molecular mechanisms taking place in this interaction and the role of autophosphorylation in Pto kinase activation, we first examined the molecular characteristics of Pto autophosphorylation in vitro. We found that Pto autophosphorylates at several sites in the protein via an intramolecular process that is not affected by protein concentration.
Induction of autophosphorylation activity by extracellular signals is well documented in mammals for receptors with tyrosine kinase activity (23). Ligand binding to the extracellular domain of receptor tyrosine kinases induces receptor dimerization. Dimerization in turn activates autophosphorylation of the receptor catalytic domain, which is mediated by an intermolecular mechanism of phosphorylation. Dimerization and intermolecular autophosphorylation of a nuclear serine/threonine kinase from Arabidopsis thaliana have been shown recently (24). However, the mechanism of activation for this protein kinase, essential for leaf and flower morphogenesis, is still unknown. In order to determine whether Pto dimerization and activation is plausible in the interaction between the signal molecule AvrPto and the Pto kinase, we tested in vitro the mechanism of Pto autophosphorylation. We found that Pto autophosphorylation occurs via an intramolecular reaction, making it unlikely that oligomerization is required for Pto activation.
Alternative modes of activation remain to be tested to elucidate the molecular mechanisms taking place during the Pto-AvrPto interaction. It is possible that AvrPto might activate Pto by causing conformational changes that possibly expose certain domains, which were not available previously, to autophosphorylation or to phosphorylation by an additional protein kinase. Such a mechanism occurs during activation of cyclindependent kinases (CDKs) by cyclins (25). The activity of CDK2, for example, which is involved in regulation of events in the eukaryotic cell cycle, is stimulated by a two-step mechanism of activation. First, the regulatory subunit cyclin A associates with CDK2, causing conformational changes in the kinase catalytic sites that make Thr-160 more accessible for phosphorylation by the CDK-activating kinase. Second, CDKactivating kinase phosphorylation of Thr-160 determines full activation of CDK2 (26).
Even in such a scenario, autophosphorylation activity of Pto still may represent a prerequisite for the interaction with AvrPto. In fact, it has been observed that forms of Pto mutated in residues essential for kinase activity do not interact with AvrPto in the two-hybrid system (7,8). The requirement of a phosphorylated residue for the interaction between a protein kinase and a regulatory subunit has been observed for the cyclic AMP-dependent protein kinase (27). Cyclic AMP-depend-ent protein kinase requires phosphorylation of a conserved threonine residue for the interaction with an associated regulatory subunit which represses its activity. Similarly, the Arabidopsis serine/threonine receptor kinase RLK5 interacts in vitro with the KAPP type 2C protein phosphatase only in its autophosphorylated form (28). In order to define the role of Pto autophosphorylation in vivo, it will be necessary to identify the residues that are phosphorylated and examine the importance of these sites on the ability of Pto to confer disease resistance. In several instances, it has been shown that sites phosphorylated in vitro by autophosphorylation or cross-phosphorylation mechanisms are also phosphorylated in vivo (29,30). Finally, examination of the quantitative and qualitative effect of AvrPto on Pto autophosphorylation may shed light on the mechanism of Pto activation.
The recognition event between Pto and AvrPto is postulated to result in activation of Pto effectors including the serine/ threonine kinase Pti1 (12) and putative transcription factors (13). Here, we have shown that Pti1, similar to Pto, autophosphorylates intramolecularly at one major and a few minor phosphorylation sites. The functional relevance of this mechanism of autophosphorylation in the Pto-signaling pathway has yet to be determined. Autophosphorylation of serine/threonine protein kinases in plants has been shown to occur intermolecularly or intramolecularly, but neither of the two mechanisms has been related to a specific form of regulation of activity (24, 31, 32).
Pti1 has been shown previously to be phosphorylated in vitro by Pto and proposed as an in vivo substrate for Pto (12). Here, we further characterized this in vitro interaction and estimated the K m of Pto for Pti1 to be 4.1 M. This K m is in the range of values observed in the interaction between kinases from different signaling pathways and their physiological or synthetic substrates. For example, the K m value of Raf-1 for MEK is 0.8 M (33), cAMP-dependent protein kinase from Dictyostelium discoideum phosphorylates a synthetic heptapeptide with a K m value of 12 M, and the ERK-2 mitogen-activated protein kinase shows a K m value of 0.12 M for Raf-1 (34).
A recessive pto allele that does not confer speck disease resistance was isolated recently and shown to encode a functional protein kinase (20). Although the pto kinase is able to use Pti1 as a substrate for phosphorylation, it shows very poor physical interaction with Pti1 and three additional Pto-interacting proteins in the yeast two-hybrid system (20). Here, we found by phosphopeptide analysis that Pti1 phosphorylation by Pto or pto and Pti autophosphorylation occur at similar sites. This observation raises questions about how the activities of different components in the speck disease resistance signaling pathway are regulated. These results also suggest that the inability of the pto kinase to mediate a resistance response to P. syringae pv. tomato expressing the avrPto gene is not related to an altered recognition specificity for Pti1 phosphorylation sites.