Phosphotyrosine-specific Phosphatase PTP-SL Regulates the ERK5 Signaling Pathway
2002; Elsevier BV; Volume: 277; Issue: 33 Linguagem: Inglês
10.1074/jbc.m202149200
ISSN1083-351X
AutoresMarcus Buschbeck, Jan Eickhoff, Marc N. Sommer, Axel Ullrich,
Tópico(s)Protein Kinase Regulation and GTPase Signaling
ResumoThe duration and the magnitude of mitogen-activated protein kinase (MAPK) activation specifies signal identity and thus allows the regulation of diverse cellular functions by the same kinase cascade. A tight and finely tuned regulation of MAPK activity is therefore critical for the definition of a specific cellular response. We investigated the role of tyrosine-specific phosphatases (PTPs) in the regulation of ERK5. Although unique in its structure, ERK5 is activated in analogy to other MAPKs by dual phosphorylation of threonine and tyrosine residues in its activation motif. In this study we concentrated on whether and how PTP-SL, a kinase-interacting motif-containing PTP, might be involved in the down-regulation of the ERK5 signal. We found that both proteins interact directly with each other in vitro and in intact cells, resulting in mutual modulation of their enzymatic activities. PTP-SL is a substrate of ERK5 and independent of phosphorylation binding to the kinase enhances its catalytic phosphatase activity. On the other hand, interaction with PTP-SL not only down-regulates endogenous ERK5 activity but also effectively impedes the translocation of ERK5 to the nucleus. These findings indicate a direct regulatory influence of PTP-SL on the ERK5 pathway and corresponding downstream responses of the cell. The duration and the magnitude of mitogen-activated protein kinase (MAPK) activation specifies signal identity and thus allows the regulation of diverse cellular functions by the same kinase cascade. A tight and finely tuned regulation of MAPK activity is therefore critical for the definition of a specific cellular response. We investigated the role of tyrosine-specific phosphatases (PTPs) in the regulation of ERK5. Although unique in its structure, ERK5 is activated in analogy to other MAPKs by dual phosphorylation of threonine and tyrosine residues in its activation motif. In this study we concentrated on whether and how PTP-SL, a kinase-interacting motif-containing PTP, might be involved in the down-regulation of the ERK5 signal. We found that both proteins interact directly with each other in vitro and in intact cells, resulting in mutual modulation of their enzymatic activities. PTP-SL is a substrate of ERK5 and independent of phosphorylation binding to the kinase enhances its catalytic phosphatase activity. On the other hand, interaction with PTP-SL not only down-regulates endogenous ERK5 activity but also effectively impedes the translocation of ERK5 to the nucleus. These findings indicate a direct regulatory influence of PTP-SL on the ERK5 pathway and corresponding downstream responses of the cell. mitogen-activated protein kinase amino acids epidermal growth factor extracellular signal-regulated kinase glutathione S-transferase hemagglutinin kinase-interacting motif protein-tyrosine phosphatase wild type green fluorescent protein catalytic cysteine to serine mutated phosphate-buffered saline human embryonic kidney Mitogen-activated protein kinases (MAPKs)1 are found in all eukaryotes and are expressed in virtually all mammalian cells. A broad variety of stimuli elicit MAPK activation, and MAPKs regulate a large number of distinct cellular responses (1Chang L. Karin M. Nature. 2001; 410: 37-40Crossref PubMed Scopus (4420) Google Scholar, 2Kyriakis J.M. Avruch J. Physiol. Rev. 2001; 81: 807-869Crossref PubMed Scopus (2897) Google Scholar, 3Whitmarsh A.J. Davis R.J. Nature. 2000; 403: 255-256Crossref PubMed Scopus (111) Google Scholar, 4Pearson G. Robinson F. Beers Gibson T., Xu, B.E. Karandikar M. Berman K. Cobb M.H. Endocr. Rev. 2001; 22: 153-183Crossref PubMed Scopus (3564) Google Scholar). In the very same cell, MAPKs can even be involved in the control of different functions. In these cases the duration and the magnitude of MAPK activation are critical parameters that specify signal identity (5Marshall C.J. Cell. 1995; 80: 179-185Abstract Full Text PDF PubMed Scopus (4245) Google Scholar). In rat pheochromocytoma cells, for example, epidermal growth factor (EGF) as well as nerve growth factor activate MAPKs, however, cells proliferate in response to the first stimulus and differentiate in the presence of the latter (6Traverse S. Gomez N. Paterson H. Marshall C. Cohen P. Biochem. J. 1992; 288: 351-355Crossref PubMed Scopus (807) Google Scholar). These distinct responses are due to the ability of nerve growth factor but not EGF to cause a sustained activation of MAPK. A tight and finely tuned regulation of MAPK activity is therefore critical for definition of a specific cellular response. MAPKs can be grouped into three main subfamilies: the extracellular signal-regulated kinases (ERKs), the c-Jun N-terminal kinases, and the p38 stress-activated protein kinases. ERK5, which is also termed big MAPK 1, differs considerably from the other family members in that it contains an unique loop-12 domain within the kinase region that is followed by an unusually long C-terminal tail of hitherto unknown function (7Lee J.D. Ulevitch R.J. Han J. Biochem. Biophys. Res. Commun. 1995; 213: 715-724Crossref PubMed Scopus (289) Google Scholar, 8Zhou G. Bao Z.Q. Dixon J.E. J. Biol. Chem. 1995; 270: 12665-12669Abstract Full Text Full Text PDF PubMed Scopus (540) Google Scholar). ERK5 is activated by diverse stimuli such as cellular stress and growth factors (9Abe J. Kusuhara M. Ulevitch R.J. Berk B.C. Lee J.D. J. Biol. Chem. 1996; 271: 16586-16590Abstract Full Text Full Text PDF PubMed Scopus (384) Google Scholar, 10Kato Y. Chao T.H. Hayashi M. Tapping R.I. Lee J.D. Immunol. Res. 2000; 21: 233-237Crossref PubMed Google Scholar). The MAPK kinase MEK5 has been shown to specifically phosphorylate and thereby activate ERK5 (7Lee J.D. Ulevitch R.J. Han J. Biochem. Biophys. Res. Commun. 1995; 213: 715-724Crossref PubMed Scopus (289) Google Scholar, 11Kato Y. Kravchenko V.V. Tapping R.I. Han J. Ulevitch R.J. Lee J.D. EMBO J. 1997; 16: 7054-7066Crossref PubMed Scopus (498) Google Scholar); the other components involved in the signaling cascade that ultimately leads to the activation of ERK5 are mostly unknown. Many of the ERK5 activating stimuli also affect other MAPK family members. The synergistic actions of both the ERK1/2 and the ERK5 pathways have for example been reported in the induction of cell transformation (12Pearson G. English J.M. White M.A. Cobb M.H. J. Biol. Chem. 2000; 276: 7927-7931Abstract Full Text Full Text PDF PubMed Scopus (117) Google Scholar). On the other hand, downstream effects exerted by activated ERK5 and ERK1/2 can in many cases be clearly distinguished from each other (13Watson F.L. Heerssen H.M. Bhattacharyya A. Klesse L. Lin M.Z. Segal R.A. Nat. Neurosci. 2001; 4: 981-988Crossref PubMed Scopus (384) Google Scholar, 14Cavanaugh J.E. Ham J. Hetman M. Poser S. Yan C. Xia Z. J. Neurosci. 2001; 21: 434-443Crossref PubMed Google Scholar). Kato et al. (15Kato Y. Tapping R.I. Huang S. Watson M.H. Ulevitch R.J. Lee J.D. Nature. 1998; 395: 713-716Crossref PubMed Scopus (360) Google Scholar) have for example demonstrated that ERK5 but not ERK2 is essential for proliferation and cell cycle progression in HeLa cells. More recently Karihaloo et al. (16Karihaloo A. O'Rourke D.A. Nickel C.H. Spokes K. Cantley L.G. J. Biol. Chem. 2000; 276: 9166-9173Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar) have shown that ERK5 mediates EGF-induced morphogenesis in renal epithelial cells, whereas ERK2 activity is critically involved in cell motility upon stimulation with hepatocyte growth factor. MAPKs including ERK5 are generally activated by phosphorylation of threonine and tyrosine residues in their activation motif; however, dephosphorylation of either residue is sufficient for kinase inactivation (17Canagarajah B.J. Khokhlatchev A. Cobb M.H. Goldsmith E.J. Cell. 1997; 90: 859-869Abstract Full Text Full Text PDF PubMed Scopus (626) Google Scholar). The present study concentrates on the question of whether and how PTP-SL might be involved in the down-regulation of the big MAPK 1/ERK5 signal. PTP-SL (18Hendriks W. Schepens J. Brugman C. Zeeuwen P. Wieringa B. Biochem. J. 1995; 305: 499-504Crossref PubMed Scopus (52) Google Scholar, 19Ogata M. Sawada M. Fujino Y. Hamaoka T. J. Biol. Chem. 1995; 270: 2337-2343Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, 20Shiozuka K. Watanabe Y. Ikeda T. Hashimoto S. Kawashima H. Gene (Amst.). 1995; 162: 279-284Crossref PubMed Scopus (39) Google Scholar, 21Sharma E. Lombroso P.J. J. Biol. Chem. 1995; 270: 49-53Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar), like STEP (22Lombroso P.J. Murdoch G. Lerner M. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 7242-7246Crossref PubMed Scopus (150) Google Scholar) and HePTP (23Zanke B. Squire J. Griesser H. Henry M. Suzuki H. Patterson B. Minden M. Mak T.W. Leukemia. 1994; 8: 236-244PubMed Google Scholar), belongs to the kinase interacting motif (KIM)-containing phosphatases that have previously been shown by us and others to bind, dephosphorylate, and thereby inactivate signaling by ERK1/2 (24Pulido R. Zuniga A. Ullrich A. EMBO J. 1998; 17: 7337-7350Crossref PubMed Scopus (272) Google Scholar, 25Saxena M. Williams S. Brockdorff J. Gilman J. Mustelin T. J. Biol. Chem. 1999; 274: 11693-11700Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar, 26Saxena M. Williams S. Gilman J. Mustelin T. J. Biol. Chem. 1998; 273: 15340-15344Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar). KIM-containing PTPs are characterized by a very restricted expression pattern. They are not found in the nucleus and are generally considered to play a role in the short term inactivation of MAPKs (27Saxena M. Mustelin T. Semin. Immunol. 2000; 12: 387-396Crossref PubMed Scopus (110) Google Scholar). Here we demonstrate that ERK5 and PTP-SL bind directly to each other. This interaction not only potently reduces kinase activity but also interferes with translocation of ERK5 to the nucleus. Our data show that ERK5 phosphorylates PTP-SL and that independent of phosphorylation binding the kinase stimulates phosphatase activity. Rabbit polyclonal anti-PTP-SL was obtained by immunization of rabbits with the peptide CHSMVQPEQAPKVLN coupled to keyhole limpet hemocyanin (Calbiochem). For the generation of ERK5 antibodies, rabbits were immunized with a fusion protein of GST and ERK5 aa 410–558. The anti-HA monoclonal antibody 12CA5 (Roche Molecular Biochemicals) was used for immunoprecipitation and the HA.11 (BAbCo) for Western blot analysis. Phosphorylation state-specific antibodies for ERK5 and ERK2 were from BioSource Europe and New England BioLabs, respectively. Anti-ERK2 K23 and anti-histone H1 were from Santa Cruz Biotechnology. Anti-RanGAP antibody was kindly provided by Frauke Melchior (Martinsried, Germany). Horseradish peroxidase-conjugated goat anti-rabbit antibody was from Bio-Rad, goat anti-mouse antibody was from Sigma, and donkey anti-goat antibody was from Jackson ImmunoResearch Laboratories. Chemiluminescence reagents and [γ-32P]ATP (6000 Ci/mmol) were from PerkinElmer Life Sciences. Protein A, protein G, and GSH-Sepharose beads were purchased from Amersham Biosciences. EGF was supplied by Invitrogen. The nitrocellulose membrane was from Schleicher & Schüll. All of the other reagents were obtained from Merck. ERK5 and MEK5 were amplified from human placenta cDNA by PCR using primers flanking the coding regions. For the expression in eukaryotic cells, both fragments were cloned into pcDNA3 (Invitrogen). The protein sequence of the cloned human MEK5 differs from the published sequence (8Zhou G. Bao Z.Q. Dixon J.E. J. Biol. Chem. 1995; 270: 12665-12669Abstract Full Text Full Text PDF PubMed Scopus (540) Google Scholar) by an insertion of 10 amino acids between glutamate 348 and isoleucine 349, which is also present in rat and mouse MEK5 (28Kamakura S. Moriguchi T. Nishida E. J. Biol. Chem. 1999; 274: 26563-26571Abstract Full Text Full Text PDF PubMed Scopus (458) Google Scholar). N-terminally HA-tagged ERK5 was generated by PCR. ERK5 and the truncated form ERK5kin (aa 1–409) consisting only of the kinase domain, were subcloned into pcDNA3-Fc, a modified vector containing 3′ of the multiple cloning site the coding sequence of the human Fcγ chain. The Fc-tagged proteins were purified from crude cell lysates of transfected HEK 293 cells using protein A-Sepharose. MEK5(D), the dominant active mutant of MEK5 (11Kato Y. Kravchenko V.V. Tapping R.I. Han J. Ulevitch R.J. Lee J.D. EMBO J. 1997; 16: 7054-7066Crossref PubMed Scopus (498) Google Scholar) was obtained by site-directed mutagenesis replacing serine 311 and threonine 315 by aspartate. Kinase inactive ERK5 KM was generated by exchanging lysine 83 for methionine. The expression constructs for PTP-SL WT and the catalytically inactive CS mutant in pRK5, a vector containing the cytomegalovirus early promotor, were described before (24Pulido R. Zuniga A. Ullrich A. EMBO J. 1998; 17: 7337-7350Crossref PubMed Scopus (272) Google Scholar). pcDNA3-GFP-PTP-SL WT and CS were generated by fusing cDNA coding for the GFP 5′ to the coding sequence of PTP-SL. GST-IA2-β (aa 641–1015) (29Kawasaki E. Hutton J.C. Eisenbarth G.S. Biochem. Biophys. Res. Commun. 1996; 227: 440-447Crossref PubMed Scopus (48) Google Scholar), GST-NC-PTP (aa 245–670) (GenBankTM accession number Z79693), and GST-HePTP (aa 1–339) (30Zanke B. Suzuki H. Kishihara K. Mizzen L. Minden M. Pawson A. Mak T.W. Eur. J. Immunol. 1992; 22: 235-239Crossref PubMed Scopus (85) Google Scholar) were obtained by PCR using human placenta and glioblastoma cDNA. PCR fragments were cloned into pGEX5X vectors (Amersham Biosciences) in frame with the GST gene. The GST-STEP and the GST-PTP-SL fusion proteins were described elsewhere (24Pulido R. Zuniga A. Ullrich A. EMBO J. 1998; 17: 7337-7350Crossref PubMed Scopus (272) Google Scholar). The GST-ERK5 fusion protein was constructed by subcloning a cDNA fragment encoding ERK5 aa 410–558 into the pGEX5X vector. All of the GST fusion proteins were expressed in the BL21 DE3 codon + (Stratagene) and purified with glutathione-Sepharose beads. The sequences of the primers used for the construction of all plasmids and for mutagenesis are available upon request. COS-7, HEK 293, and A431 cells were obtained from ATCC and cultivated following the supplier's instructions. PC12 cells (kindly provided by Philip Cohen) were grown in Dulbecco's modified Eagle's medium, 4500 mg/liter glucose, supplemented with 5% fetal calf serum and 10% horse serum. PC12 cells were generally grown on plastic dishes coated with collagen (Sigma). The cell culture reagents were purchased from Invitrogen. HEK 293 cells were transfected with 2 μg DNA/ml by the calcium phosphate precipitation method (31Chen C. Okayama H. Mol. Cell. Biol. 1987; 7: 2745-2752Crossref PubMed Scopus (4824) Google Scholar). For ectopic protein expression in COS-7 and PC12 cells, the cells were plated on 6-well dishes and were transfected with 1 μg of DNA/well and LipofectAMINE or LipofectAMINE Plus (Invitrogen), respectively, following the manufacturer's protocol. After 24 h, the cells were transferred to serum starvation medium and cultured for another 24 h before stimulation and lysis. Transfected PC12 cells were selected with 1 mg/ml neomycin. The cell cultures were washed with PBS and lysed with lysis buffer (50 mm HEPES, pH 7.5, 150 mm NaCl, 1 mm EDTA, 10% glycerine, and 1% Triton X-100) supplemented with phosphatase and protease inhibitors (10 mmNa4P2O7, 1 mmphenylmethylsulfonyl fluoride, 1 mm orthovanadate, 1 mm NaF, and 0.5% aprotinin). The cellular debris was removed by centrifugation. The supernatants were precleared with 20 μl of Sepharose slurry. The immunoprecipitations were carried out as described before with slight modifications (32Buschbeck M. Ghomashchi F. Gelb M.H. Watson S.P. Borsch-Haubold A.G. Biochem. J. 1999; 344: 359-366Crossref PubMed Scopus (40) Google Scholar). In brief, anti-HA, anti-PTP-SL, or anti-ERK5 antibodies were added together with 20 μl of mixed protein A- and G-Sepharose and one volume of HNTG (20 mm HEPES, pH 7.5, 150 mm NaCl, 0,1% Triton X-100, 10% glycerine, and 10 mmNa4P2O7). Fc-tagged proteins were directly precipitated with the mix of protein A- and G-Sepharose. Forin vitro binding studies, 1 μg of each Fc and GST fusion protein were incubated in 250 μl of PBS (8 mmNa2HPO4, 1.5 mmKH2PO4, 137 mm NaCl, 2.7 mm KCl, pH 7,3) for 20 min at room temperature under constant shaking. After the addition of 0.3 volumes of HNTG, the samples were precleared and finally precipitated with 20 μl of GSH-Sepharose beads. In general, the precipitation samples were incubated for 3 h on a rotation wheel at 4 °C. The precipitates were washed three times with 0.5 ml of HNTG buffer, suspended in 2× SDS sample buffer, boiled for 3 min, and subjected to gel electrophoresis. For Western blot analysis, the proteins were transferred to nitrocellulose membranes and immunoblotted. If quantification was necessary, the filters were exposed to the LAS1000 chemiluminescence camera (Fujifilm) and analyzed with the program Image Gauge 3.3 (Fujifilm). For in vitro kinase assays, the precipitates were washed twice with HNTG and twice with kinase assay buffer (20 mm HEPES, pH 7.5, 10 mmMgCl2, 1 mm dithiothreitol, and 0.5 mm orthovanadate). The samples were suspended in 30 μl of kinase assay buffer containing 50 μm ATP and 2 μCi of [γ-32P]ATP and incubated for 20 min at 30 °C under constant shaking. For the measurement of substrate phosphorylation, kinase reactions were also supplemented with 1 μg of GST fusion proteins or 10 μg of myelin basic protein. The reaction was extended to 30 min if 1 μg of purified ERK5 protein was used. The assay was stopped by the addition of 2× SDS sample buffer and boiling. The samples were resolved on SDS-PAGE and transferred to nitrocellulose. Phosphorylation was detected by phosphorus imaging using the BAS2500 Reader (Fujifilm) and quantified with Image Gauge 3.3 (Fujifilm). The amount of precipitated kinase was visualized by immunoblot analysis. For the measurement of PTP activity, 1 μg of GST-PTP-SL was diluted into 20 μl of PBS and incubated with 1 μg of different proteins for 10 min at room temperature under constant shaking. The aliquots of 10 μl were then added to 100 μl of p-nitrophenyl phosphate buffer (25 mm HEPES, pH 7.5, 1 mm dithiothreitol, 1 mm EDTA) containing 3.7 mg/ml p-nitrophenyl phosphate as unspecific PTP substrate. After 2 h at 37 °C, the absorption at 405 nm was determined. For performance of in vitro phosphatase assays (see Fig. 5, A and B), ERK5 precipitates were washed twice with HNTG and twice with PTP assay buffer (25 mm HEPES, pH 7.3, 10 mm dithiothreitol, 5 mm EDTA). The samples were resuspended in 20 μl of assay buffer containing the indicated amount of GST-PTP-SL fusion protein and were incubated for 20 min at 30 °C while shaking. Because these reaction were followed by in vitro kinase assays, they were stopped by washing with kinase assay buffer containing the phosphatase inhibitor orthovanadate. The kinase assay was carried out as described above. COS-7 cells were seeded at 2 × 104 cells/cm2 on glass cover slips. The transfections were performed as described above, and the cells were processed for immunofluorescence after 24 h of further culture. The cells were washed twice with PBS and fixed with methanol at −20 °C for 5 min, rinsed once with −20 °C cold acetone, and washed twice with PBS. All further steps were carried out at room temperature. The samples were incubated in PBG (PBS containing 0.5% bovine serum albumin and 0.045% teleostean fish gelatin) supplemented with 5% normal goat serum for 1 h, washed twice with PBG, and incubated with a 1:1000 dilution of anti-ERK5 antibody for 1 h. After three more washes with PBG, the cells were incubated with the secondary Cy3-labeled goat anti-rabbit antibody (Jackson ImmunoResearch Laboratories) in a 1:1000 dilution for 1 h. After one more PBG wash and three more PBS washes, the samples were rinsed in destilled water and mounted. DNA was stained for 10 min with 1 μg/ml bisbenzimid, which was included in the penultimate washing step. 48 h after transfection 1 × 106 COS-7 cells were trypsinated and collected by centrifugation at 500 × g. After washing twice with PBS, the cells were lysed in hypotonic lysis buffer (10 mmTris, pH 7.5, 10 mm NaCl, 3 mmMgCl2, 0,5% Nonidet P-40) on ice for 5 min. The nuclei were precipitated by centrifugation at 500 × g, washed once with hypotonic lysis buffer, and finally dissolved in Laemmli buffer. To address the relevance of PTP-SL function in the regulation of diverse MAPK cascades, we investigated the potential interaction between ERK5 and the cytosolic form of the STEP-like phosphatase. We performed co-precipitation experiments with polyclonal anti-PTP-SL antibody and lysates from transfected HEK 293 cells. Fig.1A (top panel) shows PTP-SL association with hemagglutinin-tagged ERK5 (HA-ERK5) in those cells co-overexpressing both proteins. As can be seen in the second panel of Fig. 1A, in addition to HA-ERK5 endogenous ERK1/2 was also detected in the immunoprecipitates in the presence but not the absence of overexpressed PTP-SL. It is conceivable that ERK1/2 might compete with ERK5 for PTP binding. As expected from earlier studies describing the expression of PTP-SL predominantly in cell lineages of neuroendocrine origin (19Ogata M. Sawada M. Fujino Y. Hamaoka T. J. Biol. Chem. 1995; 270: 2337-2343Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, 20Shiozuka K. Watanabe Y. Ikeda T. Hashimoto S. Kawashima H. Gene (Amst.). 1995; 162: 279-284Crossref PubMed Scopus (39) Google Scholar,33Augustine K.A. Silbiger S.M. Bucay N. Ulias L. Boynton A. Trebasky L.D. Medlock E.S. Anat. Rec. 2000; 258: 221-234Crossref PubMed Scopus (23) Google Scholar), we did not detect endogenous PTP-SL in HEK 293 cell lysates (Fig.1A, bottom panel). The interaction of ERK5 and PTP-SL was further demonstrated by performing the converse experiment as shown in Fig. 1B. When a truncated version of HA-ERK5 (HA-ERK5kin) that contained only the kinase domain but lacked the unique C-terminal tail was co-expressed with HA-tagged PTP-SL (HA-PTP-SL), we were readily able to pull down PTP-SL by precipitating ERK5kin (Fig. 1B, fourth lane). Addition of the IgG-Fc portion to the C terminus of the ERK5kin construct was necessary to visualize the bound phosphatase, which would otherwise have been masked by the heavy chain of the precipitating antibody. The question of whether the proteins interact directly was addressed by performing in vitro binding experiments with bacterially expressed PTP-SL and purified Fc-tagged ERK5 from transfected HEK 293 cells. Correct folding of the proteins was verified by determining their phosphatase or kinase activities (data not shown). Fig.2A shows that full-length ERK5 as well as the truncated ERK5kin protein directly interact with both wild type and catalytically inactive PTP-SL. Tarrega et al.(34Tarrega C. Blanco-Aparicio C. Munoz J.J. Pulido R. J. Biol. Chem. 2002; 277: 2629-2636Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar) and Tanoue et al. (35Tanoue T. Adachi M. Moriguchi T. Nishida E. Nat. Cell Biol. 2000; 2: 110-116Crossref PubMed Scopus (690) Google Scholar, 36Tanoue T. Maeda R. Adachi M. Nishida E. EMBO J. 2001; 20: 466-479Crossref PubMed Scopus (234) Google Scholar) identified docking motifs in ERK2 that mediate interaction with substrates and regulators including phosphatases. These docking motifs are also present and conserved in the ERK5 kinase domain and thus are likely to mediate the binding to PTP-SL. We then asked which domain of PTP-SL would be responsible for binding to ERK5 by incubating ERK5kin with GST fusion proteins that contained different portions of PTP-SL (Fig. 2B). As shown in Fig.2C, not only the full cytosolic form of PTP-SL but also the juxtamembrane and the phosphatase domain alone were both interacting with ERK5. Interestingly, even the PTP-SL juxtamembrane construct lacking the KIM, a motif that was shown to mediate the interaction with ERK2 (24Pulido R. Zuniga A. Ullrich A. EMBO J. 1998; 17: 7337-7350Crossref PubMed Scopus (272) Google Scholar, 25Saxena M. Williams S. Brockdorff J. Gilman J. Mustelin T. J. Biol. Chem. 1999; 274: 11693-11700Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar, 26Saxena M. Williams S. Gilman J. Mustelin T. J. Biol. Chem. 1998; 273: 15340-15344Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar), was still able to bind to ERK5 (Fig. 2D). When we further performed the binding experiment with ERK5kin and PTP-SL in the presence of crude lysates of HEK 293 and A431 cells as source for proteins that might possibly compete for PTP-binding sites, associated ERK5 decreased to a degree that correlated with the amount of endogenous ERK1/2 interacting with PTP-SL (Fig. 2E). Taken together, this set of data shows that the interaction between ERK5 and PTP-SL is direct and involves the kinase domain of ERK5 and, even though not exclusively, the KIM-containing juxtamembrane region of PTP-SL. To examine whether PTP-SL itself could serve as ERK5 substrate, we performed in vitrokinase assays using activated ERK5 that was immunoprecipitated from EGF-stimulated COS-7 cells and as substrates several GST fusion proteins containing different portions of PTP-SL. Fig.3A shows that wild type PTP-SL as well as those proteins containing the juxtamembrane domain were readily phosphorylated by ERK5. On the other hand, the phosphatase domain did not serve as a substrate. We then tested the ability of ERK5 to phosphorylate additional KIM-containing PTPs like NC-PTP, STEP (22Lombroso P.J. Murdoch G. Lerner M. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 7242-7246Crossref PubMed Scopus (150) Google Scholar), and HePTP (30Zanke B. Suzuki H. Kishihara K. Mizzen L. Minden M. Pawson A. Mak T.W. Eur. J. Immunol. 1992; 22: 235-239Crossref PubMed Scopus (85) Google Scholar). All of these phosphatases were phosphorylated; however, the cytosolic domain of the unrelated PTP IA2-β (29Kawasaki E. Hutton J.C. Eisenbarth G.S. Biochem. Biophys. Res. Commun. 1996; 227: 440-447Crossref PubMed Scopus (48) Google Scholar) was, as expected, not modified by ERK5 (Fig. 3B). We further asked whether complex formation between ERK5 and PTP-SL would alter enzymatic activity of the phosphatase. Therefore, we measured the activity of the bacterially expressed PTP-SL fusion protein upon binding to ERK5. As shown in Fig.4A, ERK5 as well as ERK5kin enhanced PTP-SL activity ∼3.5-fold, whereas GST and IgG alone had no effect. Because PTP-SL was phosphorylated, although not quantitatively, by a preparation of ERK5 protein isolated from transfected HEK 293 cells (data not shown), we tested the influence of this phosphorylation on PTP-SL activity by performing in vitro kinase reactions in the presence and absence of ATP and subsequently quantified the resulting phosphatase activity. As shown in Fig. 4B, PTP-SL activity was again strongly enhanced in the presence of either wild type ERK5, ERK5kin, or the kinase-inactive mutant ERK5 KM but was only very slightly decreased by the addition of ATP in all cases. These results indicate that binding to but not phosphorylation by ERK5 makes a major impact on PTP-SL activity. To answer the question of whether PTP-SL might regulate ERK5 activity, we performed in vitro phosphatase reactions using various amounts of GST-PTP-SL fusion proteins together with a preparation of activated ERK5 and subsequently measured kinase activity. COS-7 cells co-expressing the dominant active form of MEK5 were used as source for activated ERK5. MEK5 was shown to be the MAPK kinase specifically activating ERK5 and to possess constitutive activity if serine 311 and threonine 315 were mutated to aspartate (8Zhou G. Bao Z.Q. Dixon J.E. J. Biol. Chem. 1995; 270: 12665-12669Abstract Full Text Full Text PDF PubMed Scopus (540) Google Scholar,11Kato Y. Kravchenko V.V. Tapping R.I. Han J. Ulevitch R.J. Lee J.D. EMBO J. 1997; 16: 7054-7066Crossref PubMed Scopus (498) Google Scholar). Wild type PTP-SL almost completely abolished ERK5 kinase activity, whereas the catalytically inactive CS mutant affected neither autophosphorylation of ERK5 nor phosphorylation of myelin basic protein (Fig. 5A). As shown in Fig.5B, the degree of ERK5 inactivation correlated with the amount of PTP-SL protein utilized. We then tested whether PTP-SL would be capable of inactivating ERK5 in transfected COS-7 cells by performing immunocomplex kinase assays after stimulation of cells with hydrogen peroxide or EGF. Whereas wild type PTP-SL reduced ERK5 kinase activity to basal levels, the PTP-SL mutant that lacked phosphatase activity seemed to further enhance ERK5 autophosphorylation (Fig. 5C, upper panel). ERK5 appeared as a doublet resulting from a mobility shift of a small fraction of the enzyme that was hardly detectable in Western blot but clearly visible in the autoradiograph. This shift is probably due to phosphorylation of ERK5 because the amount of the upper band was found to correlate with the degree of kinase activation (14Cavanaugh J.E. Ham J. Hetman M. Poser S. Yan C. Xia Z. J. Neurosci. 2001; 21: 434-443Crossref PubMed Google Scholar, 37Mody N. Leitch J. Armstrong C. Dixon J. Cohen P. FEBS Lett. 2001; 502: 21-24Crossref PubMed Scopus (226) Google Scholar); when dominant active MEK5 was co-expressed this shift was virtually quantitative (Fig. 5, A and B). Because the relative activation of ERK5 after EGF and hydrogen peroxide treatment of cells was only moderate because of its high basal activity, we additionally expressed the constitutively active MEK5 construct to achieve a more pronounced ERK5 activation. Wild type PTP-SL partially counteracted the effect of MEK5 and reduced ERK5 activity to a level that may reflect the balanc
Referência(s)