Portal de Periódicos da CAPES

Anaplastic Lymphoma Kinase Is Activated Through the Pleiotrophin/Receptor Protein-tyrosine Phosphatase β/ζ Signaling Pathway

2007; Elsevier BV; Volume: 282; Issue: 39 Linguagem: Inglês

10.1074/jbc.m704505200

1083-351X

Pablo Pérez‐Piñera, Wei Zhang, Yunchao Chang, José A. Vega, Thomas F. Deuel,

Wnt/β-catenin signaling in development and cancer

Anaplastic lymphoma kinase (ALK) is a receptor tyrosine kinase (RTK) first discovered as the constitutively active nucleophosmin-ALK oncoprotein in anaplastic large cell lymphomas (ALCL). Full-length ALK has a critical role in normal development and differentiation. Activated full-length ALK also is found in different malignant cancers. Nevertheless, the ligand to activate ALK remained unknown until recently, when ALK was proposed to be the physiological receptor of the cytokine pleiotrophin (PTN, Ptn). However, earlier studies had demonstrated that receptor protein tyrosine phosphatase (RPTP) β/ζ is a physiological PTN receptor. We now demonstrate that phosphorylation of ALK in PTN-stimulated cells is mediated through the PTN/RPTPβ/ζ signaling pathway. ALK is phosphorylated independently of a direct interaction of PTN with ALK. The data thus support a unique model of ALK activation. In cells not stimulated by PTN, RPTPβ/ζ dephosphorylates ALK at the site(s) in ALK that is undergoing autophosphorylation through autoactivation. In contrast, when RPTPβ/ζ is inactivated in PTN-stimulated cells, the sites that are autophosphorylated in ALK no longer can be dephosphorylated by RPTPβ/ζ; thus, autoactivation and tyrosine phosphorylation of ALK rapidly increase. The data indicate that the PTN/RPTPβ/ζ signaling pathway is a critical regulator of the steady state levels of tyrosine phosphorylation and activation of ALK; the data support the conclusion that ALK phosphorylation and activation in PTN-stimulated cells are increased through a unique "alternative mechanism of RTK activation." Anaplastic lymphoma kinase (ALK) is a receptor tyrosine kinase (RTK) first discovered as the constitutively active nucleophosmin-ALK oncoprotein in anaplastic large cell lymphomas (ALCL). Full-length ALK has a critical role in normal development and differentiation. Activated full-length ALK also is found in different malignant cancers. Nevertheless, the ligand to activate ALK remained unknown until recently, when ALK was proposed to be the physiological receptor of the cytokine pleiotrophin (PTN, Ptn). However, earlier studies had demonstrated that receptor protein tyrosine phosphatase (RPTP) β/ζ is a physiological PTN receptor. We now demonstrate that phosphorylation of ALK in PTN-stimulated cells is mediated through the PTN/RPTPβ/ζ signaling pathway. ALK is phosphorylated independently of a direct interaction of PTN with ALK. The data thus support a unique model of ALK activation. In cells not stimulated by PTN, RPTPβ/ζ dephosphorylates ALK at the site(s) in ALK that is undergoing autophosphorylation through autoactivation. In contrast, when RPTPβ/ζ is inactivated in PTN-stimulated cells, the sites that are autophosphorylated in ALK no longer can be dephosphorylated by RPTPβ/ζ; thus, autoactivation and tyrosine phosphorylation of ALK rapidly increase. The data indicate that the PTN/RPTPβ/ζ signaling pathway is a critical regulator of the steady state levels of tyrosine phosphorylation and activation of ALK; the data support the conclusion that ALK phosphorylation and activation in PTN-stimulated cells are increased through a unique "alternative mechanism of RTK activation." Anaplastic lymphoma kinase (ALK) 5The abbreviations used are: ALKanaplastic lymphoma kinaseRTKreceptor tyrosine kinasePTNpleiotrophinEGFepidermal growth factorEGFREGF receptorGSTglutathione S-transferaseshRNAshort hairpin RNA. is a receptor-type transmembrane tyrosine kinase (RTK) of the insulin receptor superfamily (1Morris S.W. Naeve C. Mathew P. James P.L. Kirstein M.N. Cui X. Witte D.P. Oncogene. 1997; 14: 2175-2188Crossref PubMed Scopus (422) Google Scholar). ALK was first identified as the oncogenic chimeric nucleophosmin-ALK fusion protein that results from the (2;5)(p23;q35) chromosomal translocation in anaplastic large cell lymphomas (2Le Beau M.M. Bitter M.A. Larson R.A. Doane L.A. Ellis E.D. Franklin W.A. Rubin C.M. Kadin M.E. Vardiman J.W. Leukemia. 1989; 3: 866-870PubMed Google Scholar, 3Mason D.Y. Bastard C. Rimokh R. Dastugue N. Huret J.L. Kristoffersson U. Magaud J.P. Nezelof C. Tilly H. Vannier J.P. Hemet J. Warnke R. Br J Haematol. 1990; 74: 161-168Crossref PubMed Scopus (285) Google Scholar, 4Morris S.W. Kirstein M.N. Valentine M.B. Dittmer K.G. Shapiro D.N. Saltman D.L. Look A.T. Science. 1994; 263: 1281-1284Crossref PubMed Scopus (1980) Google Scholar); nucleophosmin-ALK contains the N-terminal domain of nucleophosmin fused with the C-terminal cytoplasmic (catalytic) domain of ALK and is constitutively active. Wild type ALK is required for normal embryonic development (5Iwahara T. Fujimoto J. Wen D. Cupples R. Bucay N. Arakawa T. Mori S. Ratzkin B. Yamamoto T. Oncogene. 1997; 14: 439-449Crossref PubMed Scopus (564) Google Scholar) and in the determination of cell survival and cell fate (6Mourali J. Benard A. Lourenco F.C. Monnet C. Greenland C. Moog-Lutz C. Racaud-Sultan C. Gonzalez-Dunia D. Vigny M. Mehlen P. Delsol G. Allouche M. Mol. Cell. Biol. 2006; 26: 6209-6222Crossref PubMed Scopus (86) Google Scholar); however, it is also involved in the pathogenesis of different cancers. Nevertheless, the mechanisms through which ALK is activated in these different contexts are not clear. anaplastic lymphoma kinase receptor tyrosine kinase pleiotrophin epidermal growth factor EGF receptor glutathione S-transferase short hairpin RNA. ALK is known to be activated through autoactivation in vitro, as is the case with other RTKs (7Schlessinger J. Cell. 2000; 103: 211-225Abstract Full Text Full Text PDF PubMed Scopus (3557) Google Scholar, 8Schlessinger J. Harvey Lect. 1993; 89: 105-123PubMed Google Scholar). Autoactivation is the consequence of ligand-enforced dimerization of RTKs that results from induced conformational changes in the active site and leads to autophosphorylation, which further enhances catalytic activity and introduces sites for different adapter proteins to engage and activate downstream signaling pathways. However, the ligand to enforce ALK dimerization and autoactivation of ALK was unknown; thus, ALK has been regarded as an orphan receptor. Recently, ALK was proposed to be the physiological receptor of pleiotrophin (PTN, Ptn) (9Bowden E.T. Stoica G.E. Wellstein A. J. Biol. Chem. 2002; 277: 35862-35868Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar, 10Powers C. Aigner A. Stoica G.E. McDonnell K. Wellstein A. J. Biol. Chem. 2002; 277: 14153-14158Abstract Full Text Full Text PDF PubMed Scopus (147) Google Scholar, 11Stoica G.E. Kuo A. Aigner A. Sunitha I. Souttou B. Malerczyk C. Caughey D.J. Wen D. Karavanov A. Riegel A.T. Wellstein A. J. Biol. Chem. 2001; 276: 16772-16779Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar). Pleiotrophin is a 136-amino acid cytokine that has diverse functions in both normal cell growth and differentiation and in growth of different malignant cells (12Li Y.S. Milner P.G. Chauhan A.K. Watson M.A. Hoffman R.M. Kodner C.M. Milbrandt J. Deuel T.F. Science. 1990; 250: 1690-1694Crossref PubMed Scopus (453) Google Scholar, 13Milner P.G. Li Y.S. Hoffman R.M. Kodner C.M. Siegel N.R. Deuel T.F. Biochem. Biophys. Res. Commun. 1989; 165: 1096-1103Crossref PubMed Scopus (201) Google Scholar). The proposal that ALK is the receptor for PTN was surprising, since earlier studies had demonstrated that the receptor protein tyrosine phosphatase (RPTP) β/ζ is a functional receptor of PTN (14Meng K. Rodriguez-Pena A. Dimitrov T. Chen W. Yamin M. Noda M. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 2603-2608Crossref PubMed Scopus (371) Google Scholar). In those studies, PTN was shown to signal through enforced dimerization of RPTPβ/ζ, which, in turn, results in a loss of the RPTPβ/ζ catalytic tyrosine phosphatase activity. The PTN-induced inactivation of RPTPβ/ζ in turn leads to increased tyrosine phosphorylation of each of the substrates of RPTPβ/ζ (14Meng K. Rodriguez-Pena A. Dimitrov T. Chen W. Yamin M. Noda M. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 2603-2608Crossref PubMed Scopus (371) Google Scholar, 15Kawachi H. Fujikawa A. Maeda N. Noda M. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 6593-6598Crossref PubMed Scopus (110) Google Scholar, 16Pariser H. Ezquerra L. Herradon G. Perez-Pinera P. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 332: 664-669Crossref PubMed Scopus (89) Google Scholar, 17Pariser H. Herradon G. Ezquerra L. Perez-Pinera P. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 12407-12412Crossref PubMed Scopus (56) Google Scholar, 18Pariser H. Perez-Pinera P. Ezquerra L. Herradon G. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 335: 232-239Crossref PubMed Scopus (66) Google Scholar, 19Tamura H. Fukada M. Fujikawa A. Noda M. Neurosci. Lett. 2006; 399: 33-38Crossref PubMed Scopus (77) Google Scholar), since the sites that normally are dephosphorylated by RPTPβ/ζ continue to be phosphorylated by constitutive tyrosine kinases when RPTPβ/ζ is inactivated in PTN-stimulated cells. The proposal that ALK is the physiological receptor for PTN also was surprising, since, in Drosophila, Jelly Belly (Jeb) was identified as a ligand of ALK (20Englund C. Loren C.E. Grabbe C. Varshney G.K. Deleuil F. Hallberg B. Palmer R.H. Nature. 2003; 425: 512-516Crossref PubMed Scopus (131) Google Scholar, 21Lee H.H. Norris A. Weiss J.B. Frasch M. Nature. 2003; 425: 507-512Crossref PubMed Scopus (144) Google Scholar). Jeb has no apparent sequence or structural relationship with PTN, and the human genome data bases do not contain a human homolog of Jeb. Furthermore, neither Miple1, the Drosophila homolog of Ptn, nor Miple2, the Drosophila homolog of midkine (Mk), which also signals through RPTPβ/ζ (22Maeda N. Ichihara-Tanaka K. Kimura T. Kadomatsu K. Muramatsu T. Noda M. J. Biol. Chem. 1999; 274: 12474-12479Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar), were found to be spatially localized with DAlk or temporally co-expressed with DAlk during Drosophila development. Furthermore, in yet another study, Moog-Lutz et al. (23Moog-Lutz C. Degoutin J. Gouzi J.Y. Frobert Y. Brunet-de Carvalho N. Bureau J. Creminon C. Vigny M. J. Biol. Chem. 2005; 280: 26039-26048Abstract Full Text Full Text PDF PubMed Scopus (128) Google Scholar) developed two anti-ALK monoclonal antibodies, each of which effectively activated ALK. However, under the same conditions, PTN failed to activate ALK; thus, in this study also, the data did not support the possibility that ALK is a physiological receptor of PTN. The conflicting views of ALK as the physiological receptor of PTN have now led us to test the possibility that ALK is activated by PTN through an alternative mechanism. In the present studies, the data demonstrate that PTN stimulates tyrosine phosphorylation of ALK through the PTN/RPTPβ/ζ signaling pathway. It stimulates phosphorylation of ALK at the same site that is autophosphorylated when ALK is autoactivated in vitro; this site is also recognized by RPTPβ/ζ and dephosphorylated by RPTPβ/ζ. The data suggest that the PTN/RPTPβ/ζ signaling pathway functions to activate ALK through an "alternative mechanism of RTK activation" that is independent of a direct interaction of PTN with RPTPβ/ζ. Cell Lines—HeLa, U87MG, U373, and MCF-7 cells were obtained from the American Tissue Collection Center (ATCC) and grown in Dulbecco's modified Eagle's medium or minimal essential medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37 °C in a 5% CO2 atmosphere. HeLa cells express RPTPβ/ζ and low levels of ALK, U87MG cells express RPTPβ/ζ and ALK, U373 cells express RPTPβ/ζ but not ALK, and MCF-7 cells express neither ALK nor RPTPβ/ζ. Reagents—Rabbit anti-ALK antibodies (catalog number 51-3900) were obtained from Zymed Laboratories Inc. (currently Invitrogen), mouse anti-phosphotyrosine antibodies (catalog number 05-321) were obtained from Upstate (Charlottesville, VA), mouse anti-RPTPβ/ζ antibodies (catalog number R20720) were obtained from BD Biosciences (San Diego, CA), PTN was obtained from R&D Systems (Minneapolis, MN), and horseradish peroxidase-conjugated anti-mouse and anti-rabbit antibodies were obtained from San Cruz Biotechnology, Inc. (Santa Cruz, CA). Plasmids, Transfections—The epidermal growth factor receptor (EGFR)/RPTPβ/ζ chimeric protein was constructed by fusing the sequences to encode the extracellular and transmembrane domains of EGFR (amino acids 1-671) with the cytoplasmic domain of RPTPβ/ζ (amino acids 1662-2315), which then were inserted into the vector pcDNA3.1. Human full-length ALK cDNA (GenBank™ accession number NM_004304), human full-length RPTPβ/ζ cDNA (Gen-Bank™ accession number NM_002851), and a truncated ALK cDNA to encode the eight membrane-proximal extracellular amino acids and the intact transmembrane and intracellular domains of ALK (amino acids 1027-1620) with the IgGκ signal peptide sequences were inserted in the pcDNA3.1 vector. Transfections were performed using the Fugene 6 transfection reagent (Roche Applied Science) following the manufacturer's instructions. The vectors containing shRNA targeting RPTPβ/ζ and a control empty vector were obtained from Open Biosystems (Huntsville, AL) (catalog numbers RHS1764-9687266 (shRNA RPTPβ/ζ-1) and RHS1764-9218363 (shRNA RPTPβ/ζ-2)) and delivered into cells using Arrest-In transfection reagent following the manufacturer's recommendations. Ninety-six hours after transfection, the levels of expression of RPTPβ/ζ were analyzed using Western blots and real time reverse transcription-PCR as described before (24Perez-Pinera P. Alcantara S. Dimitrov T. Vega J.A. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 17795-17800Crossref PubMed Scopus (50) Google Scholar). The vector pC4-Fv1E to enforce dimerization of RPTPβ/ζ through AP20187 was a kind gift from ARIAD Pharmaceuticals, Inc. (Cambridge, MA). The cDNA sequence encoding the intracellular domain of RPTPβ/ζ was inserted in frame with the Fv fragment of FKBP12. RPTPβ/ζ D1, D1 (C1932S), D1 (D1900A) GST "Capture"—Proteins recognized by the active site (D1) domain of RPTPβ/ζ (residues 1663-2034) in lysates of HeLa cells transfected with full-length ALK were captured by GST-RPTPβ/ζ D1, GST-RPTPβ/ζ inactivated D1 (C1932S), and GST-RPTPβ/ζ "substrate trap" D1 (D1900A) as described previously (18Pariser H. Perez-Pinera P. Ezquerra L. Herradon G. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 335: 232-239Crossref PubMed Scopus (66) Google Scholar) and analyzed in Western blots probed with anti-ALK antibodies and anti-GST antibodies. Dephosphorylation by RPTPβ/ζ of ALK Phosphorylated in Tyrosine—MCF-7-EGFR/RPTPβ/ζ cells, which express very limited levels of EGFR into which EGFR/RPTPβ/ζ was introduced were stimulated with 150 ng/ml EGF (R&D Systems) for 1 min or with sodium pervanadate (20 μg/ml) for 20 min. Cell lysates were prepared and immunoprecipitated with anti-ALK antibodies, and the immunoprecipitates were incubated with GST-RPTPβ/ζ D1 or GST-RPTPβ/ζ (C1932S) and analyzed by scanning densitometry of Western blots probed with antiphosphotyrosine antibodies as described before (18Pariser H. Perez-Pinera P. Ezquerra L. Herradon G. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 335: 232-239Crossref PubMed Scopus (66) Google Scholar). In Vitro Kinase Assays—ALK (catalog number 14-555; Millipore, Billerica, MA) was incubated with 200 μm ATP in a buffer containing 25 mm Tris-HCl (pH 7.5), 5 mm β-glycerophosphate, 2 mm dithiothreitol, 0.1 mm Na3VO4, and 10 mm MgCl2 for 20 min at 37 °C. The reaction was stopped with boiling loading buffer, and the samples were analyzed in Western blots probed with anti-phosphotyrosine antibodies. As negative control, samples without ATP were analyzed. Pleiotrophin Requires RPTPβ/ζ to Stimulate Tyrosine Phosphorylation of ALK—As cited above, pleiotrophin signals through enforced dimerization and inactivation of the cytoplasmic domain of RPTPβ/ζ (14Meng K. Rodriguez-Pena A. Dimitrov T. Chen W. Yamin M. Noda M. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 2603-2608Crossref PubMed Scopus (371) Google Scholar, 25Fukada M. Fujikawa A. Chow J.P. Ikematsu S. Sakuma S. Noda M. FEBS Lett. 2006; 580: 4051-4056Crossref PubMed Scopus (77) Google Scholar), leading to increased tyrosine phosphorylation of each of its substrates. To first determine if ALK is phosphorylated in PTN-stimulated cells through the PTN/RPTPβ/ζ signaling pathway, MCF-7 cells that do not express ALK or RPTPβ/ζ were transfected with cDNAs to encode full-length ALK and RPTPβ/ζ. The cells were stimulated with PTN for 5, 30, and 60 min. Lysates from stimulated and nonstimulated (control) cells were then prepared and immunoprecipitated with anti-ALK antibodies. The levels of tyrosine phosphorylation of ALK in the immunoprecipitates were measured by scanning densitometry of Western blots probed with anti-phosphotyrosine antibodies. It was found that base-line levels of tyrosine phosphorylation of ALK in non-stimulated cells were low, as previously shown (26Lamant L. Pulford K. Bischof D. Morris S.W. Mason D.Y. Delsol G. Mariame B. Am. J. Pathol. 2000; 156: 1711-1721Abstract Full Text Full Text PDF PubMed Scopus (176) Google Scholar, 27Dirks W.G. Fahnrich S. Lis Y. Becker E. MacLeod R.A. Drexler H.G. Int. J. Cancer. 2002; 100: 49-56Crossref PubMed Scopus (108) Google Scholar, 28Motegi A. Fujimoto J. Kotani M. Sakuraba H. Yamamoto T. J. Cell Sci. 2004; 117: 3319-3329Crossref PubMed Scopus (128) Google Scholar). Tyrosine phosphorylation of ALK was minimally increased at 5 min and increased ∼2.5-fold at 30 min and ∼5.2-fold at 60 min after stimulation with PTN. MCF-7 cells were then transfected with ALK alone but not with RPTPβ/ζ. The cells were then stimulated with PTN and analyzed as above; it was found that PTN failed to increase tyrosine phosphorylation of ALK in MCF-7 cells that do not express RPTPβ/ζ (Fig. 1, compare A and B). The data demonstrate that PTN effectively stimulates tyrosine phosphorylation of ALK in MCF-7 cells but only when MCF-7 cells express RPTPβ/ζ; it is concluded that RPTPβ/ζ is required for PTN to stimulate tyrosine phosphorylation of ALK in PTN-stimulated cells. The data also support the important conclusion that the level of catalytic activity of RPTPβ/ζ regulates the steady state levels of tyrosine phosphorylation of ALK when RPTPβ/ζ is not inactivated in PTN-stimulated cells. MCF-7 cells that express both ALK and RPTPβ/ζ were then stimulated with the general inhibitor of tyrosine phosphatases sodium pervanadate for 20 min. Stimulation of the MCF-7 cells expressing both ALK and RPTPβ/ζ with sodium pervanadate increased levels of tyrosine phosphorylation of ALK by ∼7-fold (Fig. 1A). MCF-7 cells that do not express RPTPβ/ζ were then stimulated with sodium pervanadate; incubation of these cells with sodium pervanadate also sharply increased tyrosine phosphorylation of ALK (Fig. 1, compare A and B). These data thus support the conclusion that tyrosine phosphatases other than RPTPβ/ζ also regulate the steady state levels of tyrosine phosphorylation of ALK when their catalytic activities are not inhibited by sodium pervanadate, presumably at tyrosine phosphorylation sites that are different from the site(s) regulated through the PTN/RPTPβ/ζ signaling pathway. shRNAs to "Knock Down" RPTPβ/ζ Prevent PTN-stimulated Tyrosine Phosphorylation of ALK—The requirement of RPTPβ/ζ for PTN to stimulate tyrosine phosphorylation of ALK in PTN-stimulated cells was investigated further. Two shRNAs to knock down RPTPβ/ζ were expressed in U87MG cells. U87MG cells express both endogenous RPTPβ/ζ and endogenous ALK (29Lu K.V. Jong K.A. Kim G.Y. Singh J. Dia E.Q. Yoshimoto K. Wang M.Y. Cloughesy T.F. Nelson S.F. Mischel P.S. J. Biol. Chem. 2005; 280: 26953-26964Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). Both the two shRNAs were first shown to effectively knock down expression of RPTPβ/ζ; it was found in these control experiments that the levels of RPTPβ/ζ mRNA were reduced ∼66% in cells expressing shRNA RPTPβ/ζ-1 and ∼73% in cells expressing shRNA RPTPβ/ζ-2, as measured by quantitative reverse transcription-PCR. Pleiotrophin failed to increase tyrosine phosphorylation of ALK in the U87MG cells that expressed either of the two shRNAs to nearly the levels of increase in cells that expressed the control shRNA. However, PTN sharply increased tyrosine phosphorylation of ALK in U87MG cells into which a "scrambled" shRNA (control) was expressed (Fig. 1C). These data thus demonstrate that knockdown of RPTPβ/ζ prevents PTN-stimulated tyrosine phosphorylation of ALK in PTN-stimulated U87MG cells; the data indicate that endogenous RPTPβ/ζ is required for PTN to stimulate tyrosine phosphorylation of ALK in U87MG cells that express both endogenous ALK and RPTPβ/ζ. Epidermal Growth Factor (EGF) Stimulates Increased Tyrosine Phosphorylation of ALK in MCF-7 Cells through the EGFR/RPTPβ/ζ Chimeric Receptor—To further pursue the requirement of enforced dimerization of the cytoplasmic domain of RPTPβ/ζ for PTN to stimulate tyrosine phosphorylation of ALK, a chimeric EGFR/RPTPβ/ζ receptor was stably expressed in MCF-7 cells (MCF-7-EGFR/RPTPβ/ζ cells). ALK was introduced into MCF-7-EGFR/RPTPβ/ζ cells, and, 48 h later, the cells were stimulated with EGF. The levels of tyrosine phosphorylation of ALK in EGF-stimulated cells were measured as described above. EGF was found to stimulate tyrosine phosphorylation of ALK ∼2.2-fold after MCF-7-EGFR/RPTPβ/ζ cells were stimulated with EGF for 1 and 5 min (Fig. 2A). The data thus provide additional support for the possibility that it is enforced dimerization of the intracellular domain of RPTPβ/ζ that has been induced by EGF through the EGFR extracellular domain that stimulates tyrosine phosphorylation of ALK. As anticipated, an ∼4-fold increase in levels of tyrosine phosphorylation of ALK was found when MCF-7-EGFR/RPTPβ/ζ cells were stimulated with sodium pervanadate (Fig. 2A), again supporting the possibility that multiple tyrosine phosphatases regulate levels of tyrosine phosphorylation of ALK, presumably at sites that are different from the site(s) regulated through the PTN/RPTPβ/ζ signaling pathway. To exclude the possibility that EGF stimulates tyrosine phosphorylation of ALK through the endogenous EGFR, MCF-7 cells that do not express EGFR/RPTPβ/ζ were transfected with ALK, stimulated with EGF, and analyzed as above. EGF did not stimulate tyrosine phosphorylation of ALK (Fig. 2A). Chemically Enforced Dimerization of RPTPβ/ζ Stimulates Tyrosine Phosphorylation of ALK—To directly demonstrate that enforced dimerization of RPTPβ/ζ alone is sufficient to stimulate tyrosine phosphorylation of ALK, MCF-7 cells were transfected with a vector pC4-Fv1E that contains the cDNA sequence to encode the Fv domain of FKBP12 in frame with the intracellular domain of RPTPβ/ζ (see "Experimental Procedures"). The FV domain of FKBP12 is needed to enforce dimerization of RPTPβ/ζ that is induced chemically by AP20187 (25Fukada M. Fujikawa A. Chow J.P. Ikematsu S. Sakuma S. Noda M. FEBS Lett. 2006; 580: 4051-4056Crossref PubMed Scopus (77) Google Scholar). The MCF-7 cells were then transfected with a cDNA to encode full-length ALK, and, 48 h later, the cells were stimulated with AP20187 (2 μm) for 30 and 60 min. The levels of tyrosine phosphorylation of ALK were then analyzed in Western blots of cell lysates probed with anti-phosphotyrosine antibodies. Enforced dimerization induced by AP20187 was found to stimulate tyrosine phosphorylation of ALK ∼12- and ∼25-fold when analyzed 30 and 60 min after the cells were stimulated with AP20187 (Fig. 2B). In the control cells that expressed ALK but contained an "empty" pC4-Fv1E vector, stimulation of the cells with AP20187 failed to increase tyrosine phosphorylation of ALK. These data thus directly demonstrate that enforced dimerization of RPTPβ/ζ alone in MCF-7 cells that express pC4-Fv1E in frame with the intracellular domain of RPTPβ/ζ and ALK is sufficient to initiate tyrosine phosphorylation of ALK. The data further support the possibility that the activity levels of RPTPβ/ζ in cells in which RPTPβ/ζ has not been inactivated by PTN regulate the steady state levels of tyrosine phosphorylation of ALK. The very high level of increase of tyrosine phosphorylation of ALK in cells stimulated with AP20187 suggests that AP20187 is more effective in inducing dimerization of RPTPβ/ζ expressed with the Fv domain of FKBP12 than is found in PTN-stimulated cells. It was found that the levels of tyrosine phosphorylation of ALK in nonstimulated cells that express pC4-Fv1E alone were very high. As will be demonstrated subsequently, ALK is a substrate of RPTPβ/ζ. Thus, it is predicted that the "empty" vector that lacks RPTPβ/ζ cannot effectively reduce tyrosine phosphorylation of ALK through the activity of RPTPβ/ζ.Itis noticed also that the levels of tyrosine phosphorylation of ALK in nonstimulated cells that express pC4-Fv1E-RPTPβ/ζ are very low, suggesting the possibility that RPTPβ/ζ with the Fv domain of FKBP12 may function as a more effective tyrosine phosphatase than RPTPβ/ζ "tethered" through its transmembrane and extracellular domains. ALK That Lacks an Extracellular Domain (Truncated ALK) Is Phosphorylated in PTN-stimulated Cells; PTN Requires RPTPβ/ζ but Not the Extracellular Domain of ALK to Stimulate Tyrosine Phosphorylation of ALK—To exclude the possibility that PTN interacts directly with ALK alone or with ALK and RPTPβ/ζ to stimulate tyrosine phosphorylation of ALK, MCF-7 cells were co-transfected with plasmids to encode RPTPβ/ζ and "truncated ALK" that lacks the extracellular domain; truncated ALK contains only the first 8 amino acids of the extracellular domain but retains the intact transmembrane and intracellular domains (amino acids 1027-1620). Thus, in cells expressing truncated ALK but not endogenous ALK, PTN cannot directly interact with ALK to stimulate tyrosine phosphorylation of ALK. MCF-7 cells transfected together with RPTPβ/ζ and truncated ALK were therefore stimulated with PTN, and the levels of tyrosine phosphorylation of ALK were measured as above. PTN was shown to stimulate tyrosine phosphorylation of truncated ALK ∼5-, ∼13-, and ∼26-fold after stimulation for 5, 30, and 60 min (Fig. 3A). The data thus demonstrate directly that PTN stimulates an increase in tyrosine phosphorylation of ALK independently of an interaction of PTN with the extracellular domain of ALK. The data furthermore demonstrate that PTN requires RPTPβ/ζ for PTN to stimulate tyrosine phosphorylation of ALK without an extracellular domain. As anticipated, PTN failed to stimulate tyrosine phosphorylation of the truncated ALK in MCF-7 cells that expressed truncated ALK but not RPTPβ/ζ. It was found that the base-line levels of tyrosine phosphorylation of ALK in nonstimulated cells were low. Again, as shown subsequently, ALK is a substrate of RPTPβ/ζ; the low base-line level of ALK phosphorylation suggests the possibility that the extracellular domain of ALK may restrict the availability of ALK as a substrate of RPTPβ/ζ in some manner, which is not the case in truncated ALK that lacks the extracellular domain. The remarkably high levels of tyrosine phosphorylation of truncated ALK in PTN-stimulated MCF-7 cells transfected with both RPTPβ/ζ and truncated ALK suggest the possibility that the levels of ALK phosphorylation are negatively regulated by the extracellular domain of ALK; ALK phosphorylation may be facilitated when ALK dimerization is not encumbered by the extracellular domain of ALK. As also anticipated, sodium pervanadate stimulates increased tyrosine phosphorylation of truncated ALK to levels in excess of tyrosine phosphorylation stimulated by PTN (Fig. 3B). ALK Co-immunoprecipitates with EGFR/RPTPβ/ζ—To address the mechanism through which enforced inactivation of RPTPβ/ζ increases tyrosine phosphorylation of ALK in PTN-stimulated cells, the possibility suggested above that ALK is a substrate of RPTPβ/ζ was pursued. To first determine if RPTPβ/ζ and ALK associate together, lysates of MCF-7-EGFR/RPTPβ/ζ cells that transiently express ALK were immunoprecipitated with anti-ALK antibodies. The immunoprecipitates were then analyzed in Western blots probed with anti-RPTPβ/ζ antibodies as above. It was found that EGFR/RPTPβ/ζ and ALK co-immunoprecipitate together (Fig. 4A). However, an ∼60% loss of the EGFR/RPTPβ/ζ that co-immunoprecipitated with ALK from lysates of unstimulated cells was found in lysates from MCF-7-EGFR/RPTPβ/ζ cells that were stimulated with EGF for 1 min. An ∼75% loss of association was found at 2 min in lysates of stimulated cells. Thus, inactivation of RPTPβ/ζ by enforced dimerization induced by EGF in MCF-7-EGFR/RPTPβ/ζ cells sharply reduced the levels of association of ALK and EGFR/RPTPβ/ζ. The data suggest the possibility that ALK associates with RPTPβ/ζ through an interaction of ALK with the active site of RPTPβ/ζ in cells that have not been stimulated with EGF. When the active site of RPTPβ/ζ is no longer accessible to ALK because of enforced dimerization, access of ALK to the active site of RPTPβ/ζ is denied. The data thus are consistent with the possibility that ALK is a substrate of RPTPβ/ζ. To further pursue the question of whether ALK associates with RPTPβ/ζ in vivo, U373 cells that express high levels of endogenous RPTPβ/ζ (14Meng K. Rodriguez-Pena A. Dimitrov T. Chen W. Yamin M. Noda M. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 2603-2608Crossref PubMed Scopus (371) Google Scholar) were then transfected with a cDNA to encode ALK and stained with fluorescein isothiocyanate-tagged anti-ALK antibodies (green), Texas Red-tagged anti-RPTPβ/ζ antibodies (red), and 4′,6-diamidino-2-phenylindole (blue). Using confocal microscopy, ALK and RPTPβ/ζ were seen to co-localize (Fig. 4B), thus supporting the likelihood that ALK and RPTPβ/ζ associate together in vivo. ALK Is a Substrate of RPTPβ/ζ—To more directly test whether ALK is a substrate of RPTPβ/ζ, HeLa cells, which express RPTPβ/ζ, were transfected with a cDNA to encode ALK and stimulated with PTN. Lysates were then prepared from the PTN-stimulated cells and incubated with the GST-D1 domain of RPTPβ/ζ or with the GST-tagged RPTPβ/ζ D1 domain mutants, GST-RPTPβ/ζ D1 (C1932S), or GST-RPTPβ/ζ D1 (D1900A). Proteins that associated with GST-RPTPβ/ζ D1, GST-RPTPβ/ζ D1 (C1932S), and GST-RPTPβ/ζ D1 (D1900A) were then captured with glutathione-Sepharose, eluted, and analyzed in Western blots probed with anti-ALK antibodies (18Pariser H. Perez-Pinera P. Ezquerra L. Herradon G. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 335: 232-239Crossref PubMed Scopus (66) Google Scholar). The D1 domain contains the active tyrosine phosphatase site of RPTPβ/ζ and itself is an active tyrosine phosphatase. RPTPβ/ζ D1 (C1932S) is an inactivated active site D1 domain with the catalytic cysteine mutated to serine. RPTPβ/ζ D1 (D1900A) is the "substrate trap" D1 mutant; the substrate trap mutant "captures" substrates of RPTPβ/ζ with very high affinity and specificity (30Dewang P.M. Hsu N.M. Peng S.Z. Li W.R. Curr. Med. Chem. 2005; 12: 1-22Crossref PubMed Scopus (47) Google Scholar). It was found that the GST-tagged RPTPβ/ζ D1 and each of the GST-tagged RPTPβ/ζ D1 domain mutants captured ALK (Fig. 5A); the capture of ALK by the D1 (D1900A) substrate trap mutant establishes with near certainty that ALK is a substrate of RPTPβ/ζ. To demonstrate directly that ALK is dephosphorylated by RPTPβ/ζ and thus a substrate of RPTPβ/ζ, ALK was then immunoprecipitated from lysates of EGF-stimulated MCF-7-EGFR/RPTPβ/ζ cells. The immunoprecipitates were resuspended and incubated with the active site D1 domain of RPTPβ/ζ or with the inactive RPTPβ/ζ D1 (C1932S) mutant. After 2 h at 37 °C, the proteins were then analyzed in Western blots probed with anti-phosphotyrosine antibodies and scanning densitometry. It was found that levels of tyrosine phosphorylation of ALK were reduced by ∼43% in lysates derived from EGF-stimulated MCF-7-EGFR/RPTPβ/ζ cells incubated with the active GST-RPTPβ/ζ D1 domain (Fig. 5B, second lane) when compared with levels of tyrosine phosphorylation of ALK that had been incubated with the inactive GST-RPTPβ/ζ D1 (C1932S) mutant (Fig. 5B, first lane). The data thus strongly support the possibility that RPTPβ/ζ dephosphorylates ALK at the site in ALK phosphorylated when RPTPβ/ζ was inactivated in EGF-stimulated MCF-7-EGFR/RPTPβ/ζ cells. Importantly also, and as anticipated, tyrosine phosphorylation of ALK in sodium pervanadate-stimulated cells was reduced by only ∼16% by incubation with GST-RPTPβ/ζ D1 (Fig. 5B, fourth lane) compared with incubation with the inactive RPTPβ/ζ mutant (Fig. 5B, third lane), consistent with data cited above indicating that multiple tyrosine phosphatases dephosphorylate ALK in sites that are not recognized by RPTPβ/ζ. ALK Is Autoactivated and Autophosphorylated in Tyrosine in Vitro—ALK autophosphorylated and autoactivated is dephosphorylated by RPTPβ/ζ. To pursue the mechanism through which ALK is phosphorylated in PTN-stimulated cells, ALK was incubated in an in vitro kinase assay. The levels of tyrosine phosphorylation of ALK were measured in Western blots probed with anti-phosphotyrosine antibodies (Fig. 6A); tyrosine phosphorylation of ALK was readily apparent 1 min after initiation of the in vitro kinase reaction and increased in levels for up to 20 min. ALK was then immunoprecipitated with anti-ALK antibodies, and the immunoprecipitates were incubated with either GST-RPTPβ/ζ D1 or with GST-RPTPβ/ζ D1 (C1932S) and analyzed in Western blots probed with anti-phosphotyrosine antibodies as described above (Fig. 6B). It was found that ALK autophosphorylated in tyrosine in vitro was dephosphorylated by GST-RPTPβ/ζ D1 and thus likely to be the same site in ALK phosphorylated in tyrosine in PTN-stimulated cells that also is dephosphorylated by RPTPβ/ζ. The data thus support the important conclusion that ALK is phosphorylated at the same site in PTN-stimulated cells that is phosphorylated when ALK is autoactivated in vitro. In both cases, this site(s) is a substrate of RPTPβ/ζ. RTKs are activated through engagement of ligands (7Schlessinger J. Cell. 2000; 103: 211-225Abstract Full Text Full Text PDF PubMed Scopus (3557) Google Scholar, 8Schlessinger J. Harvey Lect. 1993; 89: 105-123PubMed Google Scholar) that enforce dimerization and lead to autoactivation and autophosphorylation of the cognate receptors. In the present study, it is demonstrated that PTN stimulates tyrosine phosphorylation and activation of ALK through a mechanism that is independent of ligand-enforced dimerization of ALK; it is suggested that ALK is activated through a unique alternative mechanism of RTK activation. The data indicate that a mechanism of activation of ALK is autoactivation and autophosphorylation of ALK that continues to advance when the autophosphorylation sites in ALK no longer are dephosphorylated by RPTPβ/ζ that has been inactivated in PTN-stimulated cells; the steady state activity of ALK is regulated through the levels of activity of RPTPβ/ζ that dephosphorylates the critical tyrosine phosphorylation site(s) phosphorylated during ALK activation. Activation thus is not the result of a direct interaction of PTN with ALK. The model does not preclude the possibility that unidentified tyrosine kinases also may phosphorylate the site in ALK that is autophosphorylated through autoactivation of ALK. The model thus underscores that the steady-state levels of tyrosine phosphorylation of ALK are regulated through the activity of the PTN/RPTPβ/ζ signaling pathway. The "alternative mechanism of RTK activation" of ALK raises the important possibility that through the PTN/RPTPβ/ζ signaling pathway, PTN may regulate multiple RTKs and thus effectively regulate different important cellular functions; it potentially also effectively expands the range of different systems regulated through this pathway. In studies of Mourali et al. (6Mourali J. Benard A. Lourenco F.C. Monnet C. Greenland C. Moog-Lutz C. Racaud-Sultan C. Gonzalez-Dunia D. Vigny M. Mehlen P. Delsol G. Allouche M. Mol. Cell. Biol. 2006; 26: 6209-6222Crossref PubMed Scopus (86) Google Scholar), ALK was shown to have proapoptotic activity that is relieved when engaged by agonists that activate its catalytic activity. Our experiments thus implicate PTN as a potential regulator of the low levels of tyrosine phosphorylation of ALK that are critical to cell survival. Since many transformed cells are known to constitutively express Ptn, PTN may be antiapoptotic through maintaining levels of activation of ALK sufficient to support survival of malignant cells. This study identifies ALK as a new and important kinase downstream of the PTN/RPTPβ/ζ signaling pathway whose levels of tyrosine phosphorylation are regulated by PTN. Additional targets of the PTN/RPTPβ/ζ signaling pathway thus far identified include β-catenin (14Meng K. Rodriguez-Pena A. Dimitrov T. Chen W. Yamin M. Noda M. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 2603-2608Crossref PubMed Scopus (371) Google Scholar), GIT1/Cat-1 (15Kawachi H. Fujikawa A. Maeda N. Noda M. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 6593-6598Crossref PubMed Scopus (110) Google Scholar), Fyn (16Pariser H. Ezquerra L. Herradon G. Perez-Pinera P. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 332: 664-669Crossref PubMed Scopus (89) Google Scholar), β-adducin (18Pariser H. Perez-Pinera P. Ezquerra L. Herradon G. Deuel T.F. Biochem. Biophys. Res. Commun. 2005; 335: 232-239Crossref PubMed Scopus (66) Google Scholar), and P190RhoGAP (19Tamura H. Fukada M. Fujikawa A. Noda M. Neurosci. Lett. 2006; 399: 33-38Crossref PubMed Scopus (77) Google Scholar). In each case, PTN stimulates a sharp increase in the levels of tyrosine phosphorylation of the substrates of RPTPβ/ζ in PTN-stimulated cells. Furthermore, in each case studied (Fyn, β-catenin, β-adducin), RPTPβ/ζ dephosphorylates each of the substrates of RPTPβ/ζ whose levels of tyrosine phosphorylation have been increased in PTN-stimulated cells. Through this pathway, PTN thus coordinately regulates steady state levels of tyrosine phosphorylation of critical proteins important in diverse systems, such as cytoskeletal stability and function and intracellular and transmembrane tyrosine kinase activities. It is suggested that this pathway profoundly influences cellular functions at different levels and coordinately influences different cellular systems to alter the cellular phenotype. The view is supported in studies that demonstrate that PTN stimulates an epithelial to mesenchymal transition, in which loss of cell-cell adhesion, profound cytoskeleton alternations, and a morphological transition to a more motile and invasive phenotype are induced through the PTN/RPTPβ/ζ signaling pathway (24Perez-Pinera P. Alcantara S. Dimitrov T. Vega J.A. Deuel T.F. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 17795-17800Crossref PubMed Scopus (50) Google Scholar). The epithelial to mesenchymal transition is a critical event in early development and a critical component of progression of malignant cells to a more aggressive phenotype (31Thiery J.P. Nat. Rev. Cancer. 2002; 2: 442-454Crossref PubMed Scopus (5531) Google Scholar, 32Thiery J.P. Curr. Opin. Cell Biol. 2003; 15: 740-746Crossref PubMed Scopus (1451) Google Scholar). Lu et al. (29Lu K.V. Jong K.A. Kim G.Y. Singh J. Dia E.Q. Yoshimoto K. Wang M.Y. Cloughesy T.F. Nelson S.F. Mischel P.S. J. Biol. Chem. 2005; 280: 26953-26964Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar) isolated two forms of PTN (one an 18-kDa form (PTN18) and the second a 15-kDa form (PTN15)) that result from posttranslational modifications. The studies that we report use the 136-amino acid 18-kDa PTN initially reported (12Li Y.S. Milner P.G. Chauhan A.K. Watson M.A. Hoffman R.M. Kodner C.M. Milbrandt J. Deuel T.F. Science. 1990; 250: 1690-1694Crossref PubMed Scopus (453) Google Scholar, 13Milner P.G. Li Y.S. Hoffman R.M. Kodner C.M. Siegel N.R. Deuel T.F. Biochem. Biophys. Res. Commun. 1989; 165: 1096-1103Crossref PubMed Scopus (201) Google Scholar). Pleiotrophin 15 (residues 1-124) lacks the 12 C-terminal amino acids (KKEGKKQEKMLD) of PTN18, and, thus, it lacks the 5 lysine residues that establish the strong net positive charge of the C-terminal domain of PTN. In their studies, immobilized PTN18 was shown to stimulate migration of glioblastoma cells in an RPTPβ/ζ-dependent manner, whereas immobilized PTN15 promoted proliferation of glioblastoma cells in an ALK-dependent fashion. The data of Lu et al. (29Lu K.V. Jong K.A. Kim G.Y. Singh J. Dia E.Q. Yoshimoto K. Wang M.Y. Cloughesy T.F. Nelson S.F. Mischel P.S. J. Biol. Chem. 2005; 280: 26953-26964Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar) raise the possibility that PTN18 and PTN15 signal through different receptors. However, the PTN15-stimulated proliferation response that was reported to occur through ALK was determined by demonstrating phosphorylation of Akt; thus, it is possible that ALK is activated either through an interaction directly with PTN15 or, alternatively, through a different signaling pathway; thus, these data do not distinguish the mechanism of activation of ALK and are not inconsistent with our data. However, the studies of Lu et al. (29Lu K.V. Jong K.A. Kim G.Y. Singh J. Dia E.Q. Yoshimoto K. Wang M.Y. Cloughesy T.F. Nelson S.F. Mischel P.S. J. Biol. Chem. 2005; 280: 26953-26964Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar) are consistent with previous studies in this laboratory in which the N- and C-terminal domains of PTN were independently expressed with the endogenous PTN signal peptide; in those studies, the different truncated PTNs were found to stimulate transformation and angiogenesis, respectively (33Zhang N. Zhong R. Perez-Pinera P. Herradon G. Ezquerra L. Wang Z.Y. Deuel T.F. Biochem. Biophys. Res. Commun. 2006; 343: 653-658Crossref PubMed Scopus (32) Google Scholar). These data thus clearly demonstrate that ALK is activated through an alternative mechanism of RTK activation. To the best of our knowledge, this study is the first to demonstrate that a cytokine-dependent inactivation of a transmembrane tyrosine phosphatase is the mechanism to activate a transmembrane receptor tyrosine kinase. The vector pC4-Fv1E was generous gift from ARIAD Pharmaceuticals, Inc.