Portal de Periódicos da CAPES

Regulation of Plant Symbiosis Receptor Kinase through Serine and Threonine Phosphorylation

2004; Elsevier BV; Volume: 280; Issue: 10 Linguagem: Inglês

10.1074/jbc.m411665200

1083-351X

Satoko Yoshida, Martin Parniske,

Mycorrhizal Fungi and Plant Interactions

We studied the biochemical properties of a plant receptor-like kinase to gain insights into the regulatory mechanism of this largest class of plant kinases. SYMRK (symbiosis receptor kinase) is required for early signal transduction leading to plant root symbioses with nitrogen-fixing rhizobia and phosphate-acquiring arbuscular mycorrhizal fungi. Amino acid substitutions in positions critical for activity of other related kinases cause a nonsymbiotic plant phenotype, suggesting that SYMRK kinase activity is required for symbiosis. SYMRK is capable of intermolecular autophosphorylation. Nonphosphorylated SYMRK is less active than the phosphorylated version, suggesting the phosphorylation status of SYMRK determines its activity. Three Ser/Thr residues were identified as residues required for full kinase activation through targeted mutagenesis. Using quadrupole time-of-flight mass spectrometry analysis, two of these were confirmed to be phosphorylated in vitro. These crucial phosphorylation sites are conserved among various plant receptor-like kinases as well as animal Pelle/interleukin-1 receptor associated kinase. Despite the distinct domain architecture of receptor-like kinases versus Pelle/interleukin-1 receptor associated kinase, our results suggest the existence of conserved activation mechanisms. We studied the biochemical properties of a plant receptor-like kinase to gain insights into the regulatory mechanism of this largest class of plant kinases. SYMRK (symbiosis receptor kinase) is required for early signal transduction leading to plant root symbioses with nitrogen-fixing rhizobia and phosphate-acquiring arbuscular mycorrhizal fungi. Amino acid substitutions in positions critical for activity of other related kinases cause a nonsymbiotic plant phenotype, suggesting that SYMRK kinase activity is required for symbiosis. SYMRK is capable of intermolecular autophosphorylation. Nonphosphorylated SYMRK is less active than the phosphorylated version, suggesting the phosphorylation status of SYMRK determines its activity. Three Ser/Thr residues were identified as residues required for full kinase activation through targeted mutagenesis. Using quadrupole time-of-flight mass spectrometry analysis, two of these were confirmed to be phosphorylated in vitro. These crucial phosphorylation sites are conserved among various plant receptor-like kinases as well as animal Pelle/interleukin-1 receptor associated kinase. Despite the distinct domain architecture of receptor-like kinases versus Pelle/interleukin-1 receptor associated kinase, our results suggest the existence of conserved activation mechanisms. Plant receptor-like kinase (RLK) 1The abbreviations used are: RLK, receptor-like kinase; SYMRK, symbiosis receptor kinase; GST, glutathione S-transferase; CIP, calf intestinal phosphatase; Q-ToF, quadrupole time-of-flight; autorad, autoradiography; MS, mass spectrometry; IRAK, interleukin-1 receptor associated kinase. genes constitute the largest family of plant kinases, with more than 600 and 1000 genes in the Arabidopsis and the rice genome, respectively (1Shiu S.H. Karlowski W.M. Pan R. Tzeng Y.H. Mayer K.F. Li W.H. Plant Cell. 2004; 16: 1220-1234Crossref PubMed Scopus (857) Google Scholar). The hallmarks of RLKs are a signal peptide, a more or less extensive extracellular domain with a variety of sequence motifs, a transmembrane domain, and an intracellular protein kinase domain. Their overall structure suggests a role of the extracellular domain in the perception of an extracellular ligand and signal transduction through the intracellular kinase domain. This family includes genes regulating developmental processes, such as clavata 1, brassinosteroid-insensitive 1 (BRI1), and erecta, and genes involved in plant defense, such as FLS2 and Xa21 (2Shiu S.H. Bleecker A.B. Science's STKE. 2001; (http://stke.sciencemag.org/cgi/content/full/sigtrans;2001/113/RE22)PubMed Google Scholar, 3Tichtinsky G. Vanoosthuyse V. Cock J.M. Gaude T. Trends Plant Sci. 2003; 8: 231-237Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar). Several RLKs are involved in plant root symbiosis with nitrogen-fixing bacteria or phosphate-acquiring arbuscular mycorrhizal fungi (4Parniske M. Curr. Opin. Plant Biol. 2004; 7: 414-421Crossref PubMed Scopus (139) Google Scholar). The Lotus japonicus symbiosis receptor kinase (SYMRK) gene is required for a shared step between both fungal and bacterial plant symbiosis, and mutations in this gene abolish microbial interaction at an early stage (5Stracke S. Kistner C. Yoshida S. Mulder L. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Szczyglowski K. Parniske M. Nature. 2002; 417: 959-962Crossref PubMed Scopus (684) Google Scholar). Orthologues are found in Medicago truncatula (DMI2), Medicago sativa (NORK), and Pisum sativum (SYM19) (5Stracke S. Kistner C. Yoshida S. Mulder L. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Szczyglowski K. Parniske M. Nature. 2002; 417: 959-962Crossref PubMed Scopus (684) Google Scholar, 6Endre G. Kereszt A. Kevei Z. Mihacea S. Kalo P. Kiss G.B. Nature. 2002; 417: 962-966Crossref PubMed Scopus (592) Google Scholar). L. japonicus hypernodulation aberrant root formation 1 (HAR1) systemically regulates nodule number (7Nishimura R. Hayashi M. Wu G.J. Kouchi H. Imaizumi-Anraku H. Murakami Y. Kawasaki S. Akao S. Ohmori M. Nagasawa M. Harada K. Kawaguchi M. Nature. 2002; 420: 426-429Crossref PubMed Scopus (415) Google Scholar, 8Krusell L. Madsen L.H. Sato S. Aubert G. Genua A. Szczyglowski K. Duc G. Kaneko T. Tabata S. de Bruijn F. Pajuelo E. Sandal N. Stougaard J. Nature. 2002; 420: 422-426Crossref PubMed Scopus (442) Google Scholar). L. japonicus nod factor receptors 1 and 5 (NFR1 and NFR5) are supposed to act as direct receptors for nod factors (9Madsen E.B. Madsen H.L. Radutoiu S. Olbryt M. Rakwalska M. Szczyglowski K. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Nature. 2003; 425: 637-640Crossref PubMed Scopus (703) Google Scholar, 10Radutoiu S. Madsen L.H. Madsen E.B. Felle H.H. Umehara Y. Gronlund M. Sato S. Nakamura Y. Tabata S. Sandal N. Stougaard J. Nature. 2003; 425: 585-592Crossref PubMed Scopus (855) Google Scholar, 11Limpens E. Franken C. Smit P. Willemse J. Bisseling T. Geurts R. Science. 2003; 302: 630-633Crossref PubMed Scopus (600) Google Scholar). Despite the importance of RLK function in various signaling pathways involved in plant growth regulation and plant-microbe interactions, little is known about their regulatory mechanisms. Most of the characterized plant RLKs exhibit autophosphorylation activity when expressed and purified in Escherichia coli (3Tichtinsky G. Vanoosthuyse V. Cock J.M. Gaude T. Trends Plant Sci. 2003; 8: 231-237Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar). It is known that autophosphorylation of animal receptor tyrosine kinases, induced by dimerization or oligomerization upon ligand binding, leads to activation of their catalytic activity followed by signal transduction. In analogy to the receptor tyrosine kinase regulation mechanism, plant RLKs are suggested to be activated by autophosphorylation induced by dimerization or oligomerization of RLKs (2Shiu S.H. Bleecker A.B. Science's STKE. 2001; (http://stke.sciencemag.org/cgi/content/full/sigtrans;2001/113/RE22)PubMed Google Scholar, 12Li J. Curr. Opin. Plant Biol. 2003; 6: 494-499Crossref PubMed Scopus (26) Google Scholar). Arabidopsis BRI1 was found to interact with another receptor-like kinase, BRI1-associated receptor kinase 1 (BAK1). BRI1 and BAK1 transphosphorylate each other, although it is not clear whether the transphosphorylation leads to their activation (13Nam K.H. Li J. Cell. 2002; 110: 203-212Abstract Full Text Full Text PDF PubMed Scopus (885) Google Scholar, 14Li J. Wen J. Lease K.A. Doke J.T. Tax F.E. Walker J.C. Cell. 2002; 110: 213-222Abstract Full Text Full Text PDF PubMed Scopus (1055) Google Scholar). The Arabidopsis somatic embryogenesis receptor kinase autophosphorylates intermolecularly in vitro, and potential phosphorylation sites in the activation loop are important for its catalytic activity (15Shah K. Vervoort J. de Vries S.C. J. Biol. Chem. 2001; 276: 41263-41269Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar). However, the activation mechanism of plant RLKs has not yet been elucidated. In this study we used SYMRK to analyze the mechanistic properties of this important class of kinases. We show that SYMRK encodes a functional kinase and that SYMRK activity is regulated by its phosphorylation status. A combination of targeted mutagenesis and phosphopeptide analysis by mass spectrometry defined critical phosphorylation sites. Plasmid Construction—cDNA encoding the entire SYMRK intracellular domain was amplified by PCR using the forward primer TGCCGCTACAGACAA and the reverse primer ATGCATTTACTATCTCGG with additional appropriate restriction endonuclease sites. The resulting PCR products were digested with restriction endonucleases and cloned into vectors. The following combinations of restriction enzymes and vectors were used: SphI and SalI with pQE30 (Qiagen, Crawley, UK) for His-SYMRK; EcoRI and SalI with pET32a (Novagen, Nottingham, UK) for TRX-SYMRK; EcoRI and SalI with pGEX4T-1 (Amersham Biosciences) for GST-SYMRK. Point mutations were introduced by the QuikChange™ site-directed mutagenesis method from Stratagene or by a PCR-based method described previously (16Cormack B. Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Strunl K. Current Protocols in Molecular Biology. John Wiley & Sons Inc., Winston-Salem, NC1997: 8.5.7-8.5.9Google Scholar). All constructs were sequenced to exclude PCR errors and to confirm the designed mutation(s). Expression in E. coli and Protein Purification—Each construct for protein expression was transformed into E. coli strain Rosetta pLaqI (Novagen). Expression of recombinant proteins was induced by adding isopropyl 1-thio-β-d-galactopyranoside to a final concentration of 1 mm into the L broth and culturing at 22 °C for 5 h. His-tagged proteins were purified using nickel-agarose (Qiagen) according to the manufacturer's instruction. GST-tagged proteins were purified using glutathione-Sepharose 4B (Amersham Biosciences) according to the manufacturer's instructions. Purified proteins were de-salted using a PD-10 column (Amersham Biosciences), eluted with 50 mm HEPES, pH 7.4, and stored at –80 °C. In Vitro and In-gel Kinase Assay—In vitro kinase assays were carried out as follows unless otherwise specified. One μg of protein was incubated in 50 μl of kinase buffer (50 mm HEPES, pH 7.4, 10 mm MgCl2, 10 mm MnCl2, 1 mm dithiothreitol, 10 μm cold ATP, and 10 μCi [γ-32P]ATP) at 24 °C for 30 min. The reaction was stopped by adding SDS-PAGE sample buffer and boiling for 5 min. Reactions were analyzed by SDS-PAGE, and the gels were stained by Coomassie Brilliant Blue R250 or silver staining (17Shevchenko A. Wilm M. Vorm O. Mann M. Anal. Chem. 1996; 68: 850-858Crossref PubMed Scopus (7831) Google Scholar) and dried using a gel dryer. Phosphorylated proteins were analyzed by autoradiography and/or phosphorimaging (Fujifilm, Herts, UK). The sensitive in-gel kinase assay was performed as described by Romeis et al. (18Romeis T. Piedras P. Zhang S. Klessig D.F. Hirt H. Jones J.D. Plant Cell. 1999; 11: 273-287PubMed Google Scholar) using 1 mg/ml casein as substrate instead of myelin basic protein. Calf Intestinal Phosphatase (CIP) Treatment—Proteins were incubated in CIP buffer (Tris-Cl, pH 7.5, 20 mm MgCl2, 2 mm dithiothreitol) and 0.01 units/μl CIP (New England Biolabs) at 37 °C for 1 h unless otherwise indicated. Reactions were terminated by adding SDS-PAGE buffer and boiling for 5 min. Preparation of De-phosphorylated SYMRK—Purified GST-SYMRK protein was dephosphorylated as described above but using 0.1 units/μl of CIP. The CIP-treated GST-SYMRK protein was incubated with glutathione-Sepharose (Amersham Biosciences) at 4 °C for 2 h. The glutathione-Sepharose was extensively washed with phosphate-buffered saline, and protein was eluted with 1 bed volume of elution buffer (50 mm Tris, pH 8.0, 10 mm reduced glutathione) 5 times. The eluted protein was dialyzed against 50 mm HEPES, pH 7.4, and stored at –80 °C. Trypsin Digest of Recombinant Protein—Purified GST-SYMRK was used for trypsin digests with or without preincubation in a buffer allowing autophosphorylation (50 mm HEPES, pH 7.4, 10 mm MgCl2, 1 mm dithiothreitol, and 2 mm ATP) for 1 h at 37 °C.A1-μl aliquot of a 30 mm solution of iodoacetamide (Sigma-Aldrich) in 50 mm Tris, pH 8.0, was added to 5 μl of each protein sample. After 15 min, 5 μl of a solution containing 50 ng of porcine trypsin (Promega, Southampton, UK) in 50 mm Tris, pH 8.0, was added and incubated for 16 h at 37 °C. Finally, each sample was acidified using formic acid. Quadrupole Time-of-flight (Q-ToF) Mass Spectrometry and Data Base Searching—Peptides generated from tryptic digestion were loaded at a high flow rate onto a reverse-phase trapping column (0.3 mm inner diameter × 1 mm containing 5-μm C18 100-Å PepMap packing, LC Packings, Amsterdam, The Netherlands) and eluted through a reverse-phase capillary column (75-μm inner diameter × 150-mm column containing Symmetry C18 300-Å packing, Waters Ltd., Elstree, UK) directly into the nanoelectrospray ion source of a Q-ToF2 (Micromass UK Ltd., Manchester, UK). Fragment ion spectra generated from the liquid chromatography-MS/MS were analyzed using the MASCOT search tool (Matrix Science Ltd., London, UK) against the weekly updated SPTrEMBL data base using appropriate parameters. The criteria for protein identification were based on the manufacturer's definitions (Matrix Science Ltd.). Candidate peptides with probability-based Mowse scores exceeding the threshold (p < 0.05), and thus indicating a significant or extensive homology, were referred to as "hits." Protein scores were derived from peptide ion scores as a non-probabilistic basis for ranking proteins. The SYMRK Intracellular Domain Encodes a Functional Kinase—The SYMRK intracellular domain contains a conserved Ser/Thr protein kinase motif, but its kinase activity has not been shown previously. For this, the N-terminal His6-tagged SYMRK intracellular domain (His-SYMRK) or vector alone was expressed in E. coli and purified with nickel-agarose. The purified proteins were incubated with kinase buffer containing [γ-32P]ATP and separated by SDS-PAGE followed by Coomassie staining. Phosphorylated proteins were visualized by autoradiography or phosphorimaging. We detected a radioactive band in the His-SYMRK sample but not in the vector control, indicating that the SYMRK intracellular domain has autophosphorylation activity in vitro (Fig. 1A). To characterize the enzymatic properties of the SYMRK kinase, the autophosphorylation activity of His-SYMRK was analyzed in the presence of either 10 mm Mg2+, Mn2+, Mg2+ and Mn2+, or Ca2+. A similar extent of autophosphorylation activity was detected in the presence of Mg2+, Mn2+, or both cations, but no phosphorylation activity was observed when no metal or only Ca2+ was added, indicating His-SYMRK requires Mg2+ or Mn2+ but not Ca2+ for activity (Fig. 1B). Additional Ca2+ did not affect the Mg2+-dependent kinase activity (data not shown), indicating that Ca2+ does neither inhibit nor promote activity. The specificity of SYMRK was tested using known kinase substrates. His-SYMRK was incubated in kinase buffer with either 2 mg of casein, histone IIA, or myelin basic protein. Autoradiography detected bands corresponding to these substrates, but casein and myelin basic protein gave stronger signals than histone (Fig. 1C). Therefore, casein was used as an artificial substrate in subsequent experiments. Correlation between Kinase Activity and Biological Function of SYMRK—The domain structure of SYMRK supports the hypothesis that the extracellular domain perceives an unknown signal and the intracellular domain transduces the perception event to downstream targets via phosphorylation. A critical role for the kinase activity is suggested by several missense and nonsense point mutation alleles in the SYMRK kinase domain from L. japonicus and P. sativum (Refs. 5Stracke S. Kistner C. Yoshida S. Mulder L. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Szczyglowski K. Parniske M. Nature. 2002; 417: 959-962Crossref PubMed Scopus (684) Google Scholar and 19Perry J.A. Wang T. Welham T.J. Gardner S. Pike J.M. Yoshida S. Parniske M. Plant Physiol. 2003; 131: 866-871Crossref PubMed Scopus (266) Google Scholar and Table I), which result in plants that cannot establish root nodule symbiosis. Mutations are overrepresented in conserved kinase residues known to be critical for the catalytic activity. L. japonicus symrk-6 (G603R) and P. sativum sym19 P4 (G604E) mutations replace the second G of the GXGXXGXV ATP binding motif of domain I. symrk-9 (G604R) is in the same region, although it does not affect a conserved amino acid. The symrk-10 (D738N) and sym19 P55 (D739N) mutations destroy the conserved DFG motif of domain VII at the beginning of the activation loop, which is crucial for kinase activity (20Huse M. Kuriyan J. Cell. 2002; 109: 275-282Abstract Full Text Full Text PDF PubMed Scopus (1370) Google Scholar). Another mutation, symrk-8 (G759R), is also located in the activation loop. To determine the effects of these mutations on protein structure, a model for the SYMRK kinase domain was generated, and the mutation sites were mapped on the model. As depicted in Fig. 2, most of the mutations generating non-symbiotic SYMRK alleles are located close to the activation loop.Table ISYMRK mutant alleles from L. japonicus and P. sativum with mutations in the kinase domainAlleleLine designationAmino acid positionAmino acid changeDomainReferenceL. japonicussymrk-6EMS34603Gly to ArgI19Perry J.A. Wang T. Welham T.J. Gardner S. Pike J.M. Yoshida S. Parniske M. Plant Physiol. 2003; 131: 866-871Crossref PubMed Scopus (266) Google Scholarsymrk-9SL605-2,3604Gly to ArgI19Perry J.A. Wang T. Welham T.J. Gardner S. Pike J.M. Yoshida S. Parniske M. Plant Physiol. 2003; 131: 866-871Crossref PubMed Scopus (266) Google Scholarsymrk-10SL1951-2,3,4,5,6,7738Asp to AsnVII19Perry J.A. Wang T. Welham T.J. Gardner S. Pike J.M. Yoshida S. Parniske M. Plant Physiol. 2003; 131: 866-871Crossref PubMed Scopus (266) Google Scholarsymrk-8SL160-2,3,4,5,6759Gly to ArgVII19Perry J.A. Wang T. Welham T.J. Gardner S. Pike J.M. Yoshida S. Parniske M. Plant Physiol. 2003; 131: 866-871Crossref PubMed Scopus (266) Google Scholarsymrk-11SL3472-2793Gly to AspIX19Perry J.A. Wang T. Welham T.J. Gardner S. Pike J.M. Yoshida S. Parniske M. Plant Physiol. 2003; 131: 866-871Crossref PubMed Scopus (266) Google Scholarsymrk-7EMS61806Trp to stopX5Stracke S. Kistner C. Yoshida S. Mulder L. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Szczyglowski K. Parniske M. Nature. 2002; 417: 959-962Crossref PubMed Scopus (684) Google ScholarP. sativumsym19P4604Gly to GlnI5Stracke S. Kistner C. Yoshida S. Mulder L. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Szczyglowski K. Parniske M. Nature. 2002; 417: 959-962Crossref PubMed Scopus (684) Google Scholarsym19P55739Asp to AsnVII5Stracke S. Kistner C. Yoshida S. Mulder L. Sato S. Kaneko T. Tabata S. Sandal N. Stougaard J. Szczyglowski K. Parniske M. Nature. 2002; 417: 959-962Crossref PubMed Scopus (684) Google Scholar Open table in a new tab From the information above it can be expected that most of these symrk mutations result in either loss or impairment of kinase activity. To confirm this, we introduced a subset of the mutations corresponding to plant mutant alleles into the E. coli-expressed SYMRK intracellular domain. His-W806*, His-G793D, and TRX-D738N contain the same mutation as the plant alleles, symrk-7, symrk-11, and symrk-10, respectively (Fig. 3A). These proteins were purified and subjected to an autophosphorylation assay. Neither of the mutant proteins showed autophosphorylation activity (Fig. 3B; for TRX-D738N data are not shown). These results indicate a correlation between plant symbiotic phenotype and kinase activity, suggesting that the kinase activity is important for SYMRK function in symbiosis. Intermolecular Phosphorylation of SYMRK—To investigate whether the SYMRK autophosphorylation proceeds via an intermolecular or an intramolecular mechanism, kinase-deficient SYMRK mutant proteins were used as a potential substrate of SYMRK. The wild-type SYMRK intracellular domain was cloned in the vector pET32a (TRX-SYMRK), which adds a 27-kDa thioredoxin tag, to generate a protein with a distinct molecular weight from the His-tagged mutant proteins. The TRX-SYMRK protein was purified from E. coli and incubated in kinase buffer with or without the kinase-dead mutants His-W806* or His-G793D. Autoradiography analysis showed TRX-SYMRK has autophosphorylation activity. When His-W806* and His-G793D were incubated with TRX-SYMRK kinase, phosphorylated bands were detected with sizes corresponding to these kinase-deficient proteins (Fig. 3C). These results show that the SYMRK intracellular domain catalyzes intermolecular phosphorylation. The Phosphorylation Status of SYMRK Affects Kinase Activity—When E. coli-expressed His-SYMRK was analyzed by SDS-PAGE, we observed multiple bands of apparent molecular mass between 57 and 50 kDa, although the calculated molecular mass of the tagged protein is 44 kDa. Because these multiple bands were recognized by an anti-His antibody (data not shown), we investigated whether these low mobility bands represented phosphorylated species of His-SYMRK. After incubation with CIP, only one band with a similar size to the smallest of untreated His-SYMRK was observed (Fig. 4A). This result indicates that His-SYMRK protein is phosphorylated in E. coli. Because kinase-deficient mutant protein shows a single band that corresponds to the size of the CIP-treated sample (data not shown), His-SYMRK is likely to be autophosphorylated and not phosphorylated by E. coli-derived kinases. The differential phosphorylation status may affect the catalytic activity of SYMRK. To analyze the kinase activity of the different phosphorylated forms of SYMRK intracellular domain, we performed an in-gel kinase assay of His-SYMRK preincubated with or without CIP. Interestingly, we detected a radioactive band in the lane of non-CIP-treated His-SYMRK, but this band was barely visible in the lane containing CIP-treated protein. This result suggests that non-phosphorylated SYMRK has a much lower kinase activity than phosphorylated SYMRK. A time course experiment confirmed that the in-gel kinase activity decreased with progression of SYMRK dephosphorylation (Fig. 4B). Because the in-gel kinase assay includes procedures of protein denaturation and renaturation, the reduced activity might be due to an inability of the non-phosphorylated SYMRK to refold properly. To investigate this possibility, we purified the non-phosphorylated SYMRK intracellular domain under a native condition. GST-tagged SYMRK intracellular domain (GST-SYMRK) purified from E. coli was dephosphorylated by CIP for 1 h at 37 °C and repurified with glutathione-Sepharose. The repurified de-phosphorylated SYMRK was subjected to an in vitro kinase assay. GSTSYMRK showed a strong autophosphorylation band within 5 min and reached a maximum level after 30–60 min of incubation, whereas de-phosphorylated SYMRK showed a weak signal after 10 min of incubation that gradually increased during longer incubation periods (Fig. 5A). To test if this difference in the kinase activity is due to a contamination with remaining CIP activity, GST-SYMRK and de-phosphorylated SYMRK were mixed at 1:1 ratio and subjected to an in vitro kinase assay for 10 min. GST-SYMRK activity was not inhibited by the de-phosphorylated SYMRK, suggesting that there was no remaining phosphatase activity in the de-phosphorylated SYMRK sample (Fig. 5B). Taken together, these results suggest that phosphorylation is necessary for full kinase activity of SYMRK. Site-directed Mutagenesis of Predicted Phosphorylation Sites— To identify Ser/Thr residues decisive for the regulation of kinase activity, we performed site-directed mutagenesis. In the SYMRK intracellular domain 25 potential phosphorylation sites were predicted by the NetPhos2.0 program (www.cbs.dtu.dk/services/NetPhos). Among these we selected residues conserved between legume and non-legume SYMRK homologues as mutagenesis targets (data not shown). To narrow down the number of target residues, we especially selected the Ser/Thr residues placed at the following three regions; the activation segment, the juxtamembrane region preceding the conserved kinase domain, and the C-terminal region located after the conserved kinase domain. Previous studies suggested that phosphorylation of the activation segment is required for full kinase activity in many kinases, and some kinases contain regulatory phosphorylation sites in the juxtamembrane and the C-terminal region (20Huse M. Kuriyan J. Cell. 2002; 109: 275-282Abstract Full Text Full Text PDF PubMed Scopus (1370) Google Scholar, 23Adams J.A. Biochemistry. 2003; 42: 601-607Crossref PubMed Scopus (186) Google Scholar). We selected two Ser (Ser-751 and Ser-754) and 1 Thr (Thr-760) residues from the activation segment, one Ser (Ser-580) and one Thr (Thr-593) from the juxtamembrane region, and two Ser (Ser-870 and Ser-888) from the C-terminal region. These Ser/Thr residues were substituted by Ala in the GST-SYMRK construct, and the mutated proteins were analyzed by in vitro and in-gel kinase assay (Fig. 6, A and B). Mutations T760A in the activation segment and T593A in the juxtamembrane region reduced kinase activity significantly, although both proteins still showed basal in vitro kinase activity. The activity of these mutant proteins could not be detected by in-gel kinase assays. The S754A mutation slightly reduced auto phosphorylation and substrate phosphorylation activity in vitro, whereas no activity was detected by in-gel kinase assay. The mutations at S570A, S870A, and S888A did not affect either in vitro or in-gel kinase activity. In brief, we found three Ser/Thr residues that affect autophosphorylation and transphosphorylation activity of SYMRK. Determination of Phosphorylation Sites of SYMRK Expressed in E. coli—Although crucial Ser/Thr residues for regulation of SYMRK kinase activity were identified, it was unclear whether these sites were indeed phosphorylated. To answer this question, we determined the presence of phospho-amino acids by Q-ToF/MS in E. coli-expressed GST-SYMRK proteins with or without preincubation in an autophosphorylation buffer. Although several tryptic SYMRK peptides were not detected by Q-ToF/MS, the peptides that included each of Ser-754, Thr-760, and Thr-593 were detected. MS-Spectra identified phosphorylated Ser-754 in the GST-SYMRK sample without preincubation (Fig. 7A) and Thr-760 from the preincubated sample (Fig. 7B). Phosphorylation of Thr-593 could not be detected in either sample. Because no phosphorylated peptides were detected on GST-K622E, a kinase-dead mutant, the observed phosphorylation of GST-SYMRK is likely due to autophosphorylation and not mediated by an E. coli-derived kinase. SYMRK is a member of the large family of leucine-rich repeat receptor-like kinases in plants. In this study we show that the SYMRK intracellular domain encodes a functional kinase that is regulated by phosphorylation. Several L. japonicus and P. sativum mutants defective in symbiosis were shown to carry mutations in the conserved kinase motifs of SYMRK, which suggest that the kinase activity of SYMRK is required for its symbiotic function. Although kinase domain sequences of various plant receptor kinases are highly conserved, phosphorylation properties are variable. The disease-resistance gene Xa21 from Rice encodes a leucine-rich repeat-type receptor kinase, and the Xa21 kinase domain autophosphorylates in an intramolecular manner (24Liu G.Z. Pi L.Y. Walker J.C. Ronald P.C. Song W.Y. J. Biol. Chem. 2002; 277: 20264-20269Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). In contrast, Arabidopsis somatic embryogenesis receptor kinase is reported to autophosphorylate intermolecularly (15Shah K. Vervoort J. de Vries S.C. J. Biol. Chem. 2001; 276: 41263-41269Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar). We found that SYMRK autophosphorylates intermolecularly as the SYMRK kinase domain phosphorylates kinase-dead mutant versions, although we cannot exclude the possibility SYMRK also autophosphorylates intramolecularly. In animal systems, intermolecular phosphorylation is particularly important for activation of some receptor-tyrosine kinases that recognize growth factors. Upon ligand binding, dimerization or oligomerization of receptor-tyrosine kinases triggers intermolecular phosphorylation of regulatory residues that are essential for activation of kinases (25Schlessinger J. Cell. 2000; 103: 211-225Abstract Full Text Full Text PDF PubMed Scopus (3557) Google Scholar). A similar mechanism is also reported for the Drosophila kinase Pelle, a toll receptor-associated kinase. The Pelle kinase is activated by concentration-dependent autophosphorylation, and this activation is supposed to be triggered by dimerization or oligomerization of Toll receptor complexes (26Shen B. Manley J. Development. 2002; 129: 1925-1933Crossref PubMed Google Scholar). All plant RLKs belong to a monophyletic gene family, which includes Drosophila Pelle (2Shiu S.H. Bleecker A.B. Science's STKE. 2001; (http://stke.sciencemag.org/cgi/content/full/sigtrans;2001/113/RE22)PubMed Google Scholar). Based on analogy to animal kinases, phosphorylation-dependent activation of plant RLKs was repeatedly proposed (2Shiu S.H. Bleecker A.B. Science's STKE. 2001; (http://stke.sciencemag.org/cgi/content/full/sigtrans;2001/113/RE22)PubMed Google Scholar, 12Li J. Curr. Opin. Plant Biol. 2003; 6: 494-499Crossref PubMed Scopus (26) Google Scholar). However, this has not yet been shown. Here we demonstrate that SYMRK kinase activity is regulated by its phosphorylation status and probably via autophosphorylation. Results from both in-gel and in vitro kinase assays suggest that SYMRK needs to be phosphorylated to gain full kinase activity. To identify the phosphorylation sites that affect the SYMRK activation, we mutagenized seven computer-predicted phosphorylation sites that are conserved between SYMRK orthologues from different plant species. We found three crucial residues, Thr-593, Ser-754, and Thr-760, that are required for full kinase activity. Residues Ser-754 and Thr-760 are located within the activation segment. The T760A mutation reduced kinase activity significantly but not completely. The S754A mutant protein showed only slightly less autophosphorylation and substrate phosphorylation activity than wild-type GST-SYMRK protein, and the kinase activity was not detected by the in-gel kinase assay. In the Arabidopsis somatic embryogenesis receptor kinase protein the Thr residues Thr-462 and Thr-468 located within the activation loop are important for its catalytic activity (15Shah K. Vervoort J. de Vries S.C. J. Biol. Chem. 2001; 276: 41263-41269Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar) and correspond to Ser-754 and Thr-760 in SYMRK, respectively. Although the effects of mutations in these residues were generally more pronounced in somatic embryogenesis receptor kinase than in SYMRK, the trend was the same, with Thr-468 having a stronger defect than Thr-462. Residue Thr-593 is located just before the ATP binding site in the conserved kinase domain of SYMRK. A tomato disease-resistance gene, Pto, encodes a kinase. The sequence of Pto is closely related to the kinase domain of RLKs. The autophosphorylation sites in Pto were extensively analyzed, and Thr-38 was found to be required for kinase activation and signal transduction (27Sessa G. D'Ascenzo M. Martin G.B. EMBO J. 2000; 19: 2257-2269Crossref PubMed Scopus (87) Google Scholar). Alignment of the SYMRK intracellular domain and Pto revealed that Thr-38 in Pto is equivalent to SYMRK Thr-593 (Fig. 8). Mutation of this residue leads to a dramatic reduction in kinase activity, suggesting that this critical residue has a conserved role in SYMRK and Pto. Recently, the phosphorylation-dependent activation mechanism of IRAK-1, the mammalian orthologue of Pelle, was reported (28Kollewe C. Mackensen A.C. Neumann D. Knop J. Cao P. Li S. Wesche H. Martin M.U. J. Biol. Chem. 2004; 279: 5227-5236Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar). IRAK-1 requires phosphorylation of two different sites, Thr-209 and Thr-387, to be activated. Kollewe et al. (28Kollewe C. Mackensen A.C. Neumann D. Knop J. Cao P. Li S. Wesche H. Martin M.U. J. Biol. Chem. 2004; 279: 5227-5236Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar) suggested that phosphorylation of Thr-209 "kicks off" the activation of the kinase and subsequent Thr-387 phosphorylation precedes hyperphosphorylation of the ProST region followed by dissociation of IRAK-1 from the interleukin-1 receptor complex. The amino acid alignment indicates Thr-209 and Thr-387 in IRAK-1 correspond to Thr-593 and Thr-760 of SYMRK, respectively (Fig. 8). The combined results of the present study with that of Kollewe et al. (28Kollewe C. Mackensen A.C. Neumann D. Knop J. Cao P. Li S. Wesche H. Martin M.U. J. Biol. Chem. 2004; 279: 5227-5236Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar) suggest similarities in the regulatory mechanisms of IRAK-1 and SYMRK, although SYMRK has additional transmembrane and extracellular domains that IRAK-1 lacks. Moreover, with the exception of Ser-754 in Pelle, the three crucial phosphorylation sites identified in this study are largely conserved among various plant receptor-like kinases and Pto as well as Pelle and IRAK-1, suggesting the existence of an evolutionary conserved activation mechanisms. By using Q-ToF/MS analysis of trypsin-digested SYMRK, we could prove that some of these crucial Ser and Thr are actual phosphorylation sites at least in E. coli-expressed protein. We observed peaks corresponding to phosphorylated Ser-754 and T760 residues in the sample with and without pre-autophosphorylation treatment, respectively. These results suggest that the phosphorylation of Ser-754 and Thr-760 may occur under different conditions. We could not detect phospho-Thr-593 by Q-ToF. This could either indicate that Thr-593 has a role in kinase activation independent of phosphorylation or that the phosphorylation of Thr-593 is transient and, hence, difficult to detect. Alternatively, we could not detect it for technical reasons. Recently, a phosphoproteome analysis of plasma membrane proteins from Arabidopsis cell cultures discovered in vivo phosphorylation sites of RLKs (29Nühse T.S. Stensballe A. Jensen O.N. Peck S.C. Plant Cell. 2004; 16: 2394-2405Crossref PubMed Scopus (411) Google Scholar). These data of Nühse et al. (29Nühse T.S. Stensballe A. Jensen O.N. Peck S.C. Plant Cell. 2004; 16: 2394-2405Crossref PubMed Scopus (411) Google Scholar) are consistent with our in vitro results; Ser and Thr residues corresponding to SYMRK Thr-593, Ser-754, and Thr-760 were found to be phosphorylated in several Arabidopsis RLKs in vivo. Taken together, the combined results highlight the relevance of these sites not only for the regulation of SYMRK but also of other members of the RLK gene family. We are grateful to Dr. Andrew Bottrill and Dr. Mike Naldrett of the joint IFR-JIC Proteomics Facility for the Q-ToF/MS analysis of SYMRK. We thank Dr. Marcus C. Durrant for SYMRK kinase structural modeling and Dr. Judith Müller for carefully reading the manuscript. Research at The Sainsbury Laboratory is supported by The Gatsby Charitable Foundation.