Portal de Periódicos da CAPES

Protein Kinase A Intersects Src Signaling in Membrane Microdomains

2003; Elsevier BV; Volume: 278; Issue: 19 Linguagem: Inglês

10.1074/jbc.m211426200

1083-351X

Hilde Abrahamsen, Torkel Vang, Kjetil Taskén,

Cell Adhesion Molecules Research

Regulation of Src kinase activity is tightly coupled to the phosphorylation status of the C-terminal regulatory tyrosine Tyr527, which, when phosphorylated by Csk, represses Src. Here, we demonstrate that activation of Csk through a prostaglandin E2-cAMP-protein kinase A (PKA) pathway inhibits Src. This inhibitory pathway is operative in detergent-resistant membrane fractions where cAMP-elevating agents activate Csk, resulting in a concomitant decrease in Src activity. The inhibitory effect on Src depends on a detergent-resistant membrane-anchored Csk and co-localization of all components of the inhibitory pathway in membrane microdomains. Furthermore, epidermal growth factor-induced activation of Src and phosphorylation of the Src substrates Cbl and focal adhesion kinase are inhibited by activation of the cAMP-PKA-Csk pathway. We propose a novel mechanism whereby G protein-coupled receptors inhibit Src signaling by activation of Csk in a cAMP-PKA-dependent manner. Regulation of Src kinase activity is tightly coupled to the phosphorylation status of the C-terminal regulatory tyrosine Tyr527, which, when phosphorylated by Csk, represses Src. Here, we demonstrate that activation of Csk through a prostaglandin E2-cAMP-protein kinase A (PKA) pathway inhibits Src. This inhibitory pathway is operative in detergent-resistant membrane fractions where cAMP-elevating agents activate Csk, resulting in a concomitant decrease in Src activity. The inhibitory effect on Src depends on a detergent-resistant membrane-anchored Csk and co-localization of all components of the inhibitory pathway in membrane microdomains. Furthermore, epidermal growth factor-induced activation of Src and phosphorylation of the Src substrates Cbl and focal adhesion kinase are inhibited by activation of the cAMP-PKA-Csk pathway. We propose a novel mechanism whereby G protein-coupled receptors inhibit Src signaling by activation of Csk in a cAMP-PKA-dependent manner. Src family kinase Src homology 2 and 3, respectively prostaglandin E2 G protein-coupled receptor adenylyl cyclase protein kinase A phosphoprotein associated with glycosphingolipid-enriched membrane domains epidermal growth factor epidermal growth factor receptor platelet-derived growth factor platelet-derived growth factor receptor detergent-resistant membrane hemagglutinin epitope human embryonic kidney 293 cells focal adhesion kinase protein-tyrosine phosphatase 4-morpholineethanesulfonic acid Src family kinases (SFKs)1 are involved in a variety of signal-transducing events leading to diverse cellular processes such as migration, adhesion, proliferation, differentiation, and survival (1Thomas S.M. Brugge J.S. Annu. Rev. Cell Dev. Biol. 1997; 13: 513-609Crossref PubMed Scopus (2166) Google Scholar). In mammals, eight different SFKs have been described (2Neet K. Hunter T. Genes Cells. 1996; 1: 147-169Crossref PubMed Scopus (125) Google Scholar). Each member has a distinct unique domain, but all SFKs have in common that they contain acylation site(s) for membrane localization and single SH3, SH2, and protein-tyrosine kinase domains (3Brown M.T. Cooper J.A. Biochim. Biophys. Acta. 1996; 1287: 121-149Crossref PubMed Scopus (1085) Google Scholar). In addition, SFKs contain two critical tyrosine residues (corresponding to Tyr527 and Tyr416 in Src) important for overall regulation of SFK activity. When phosphorylated by the ubiquitously expressed C-terminal Src kinase, Csk, the C-terminal tyrosine (Tyr527) interacts with the SH2 domain, resulting in repression of the catalytic activity (4Okada M. Nakagawa H. J. Biol. Chem. 1989; 264: 20886-20893Abstract Full Text PDF PubMed Google Scholar, 5Nada S. Okada M. MacAuley A. Cooper J.A. Nakagawa H. Nature. 1991; 351: 69-72Crossref PubMed Scopus (511) Google Scholar, 6Courtneidge S.A. EMBO J. 1985; 4: 1471-1477Crossref PubMed Scopus (214) Google Scholar, 7Matsuda M. Mayer B.J. Fukui Y. Hanafusa H. Science. 1990; 248: 1537-1539Crossref PubMed Scopus (285) Google Scholar, 8Cooper J.A. King C.S. Mol. Cell. Biol. 1986; 6: 4467-4477Crossref PubMed Scopus (159) Google Scholar, 9Cooper J.A. Gould K.L. Cartwright C.A. Hunter T. Science. 1986; 231: 1431-1434Crossref PubMed Scopus (349) Google Scholar). The other regulatory tyrosine residue (Tyr416) is located in the activation loop, and autophosphorylation of this site is required for full catalytic activity. Thus, induction of Src kinase activity involves sequential displacement of the SH2 domain from phosphorylated Tyr527 and/or dephosphorylation of Tyr527 (by PTPases such as RPTPα) (10Zheng X.M. Resnick R.J. Shalloway D. EMBO J. 2000; 19: 964-978Crossref PubMed Scopus (207) Google Scholar) and subsequent autophosphorylation of Tyr416. The importance of Csk-mediated regulation of Tyr527 is illustrated by the fact that disruption of thecsk gene is embryonically lethal, and studies of cell lines established from these animals revealed that although Src expression levels were reduced, relative Src kinase activity was increased 5–15-fold (11Imamoto A. Soriano P. Cell. 1993; 73: 1117-1124Abstract Full Text PDF PubMed Scopus (346) Google Scholar, 12Nada S. Yagi T. Takeda H. Tokunaga T. Nakagawa H. Ikawa Y. Okada M. Aizawa S. Cell. 1993; 73: 1125-1135Abstract Full Text PDF PubMed Scopus (363) Google Scholar). This was associated with very low levels of Tyr527 phosphorylation in Src, whereas the phosphorylation of Tyr416 was highly increased (11Imamoto A. Soriano P. Cell. 1993; 73: 1117-1124Abstract Full Text PDF PubMed Scopus (346) Google Scholar). Due to single or double acylations, SFKs are associated with the inner leaflet of the plasma membrane and typically partition into detergent-resistant membranes (DRMs), which is suggested as an overall term including both lipid rafts and caveolae (13Simons K. Toomre D. Nat. Rev. Mol. Cell. Biol. 2000; 1: 31-39Crossref PubMed Scopus (5162) Google Scholar). Data indicate that lipid rafts and caveolae function as platforms for the signal-transducing machinery (14Hunter T. Cell. 2000; 100: 113-127Abstract Full Text Full Text PDF PubMed Scopus (2268) Google Scholar) and serve for example to position kinases and their substrates in close proximity for signal transduction events to occur rapidly upon the appropriate signal. The functional significance of lipid rafts as signaling units has best been characterized downstream of the T cell receptor, where the integrity of lipid rafts has been shown essential for proper signaling (15Xavier R. Brennan T. Li Q. McCormack C. Seed B. Immunity. 1998; 8: 723-732Abstract Full Text Full Text PDF PubMed Scopus (837) Google Scholar). Several signal transduction pathways induce SFK activity, including immunoreceptors, cytokine receptors, G protein-coupled receptors (GPCRs), integrins, and receptor tyrosine kinases (e.g. the EGF and the PDGF receptors (EGFRs and PDGFRs, respectively)) (1Thomas S.M. Brugge J.S. Annu. Rev. Cell Dev. Biol. 1997; 13: 513-609Crossref PubMed Scopus (2166) Google Scholar). Both EGFRs and PDGFRs are activated by dimerization followed by trans-autophosphorylation, giving rise to docking sites for a variety of signaling molecules such as phosphatidylinositol 3-kinase, phospholipase Cγ, Grb2/Sos, SHC, and Cbl. Prior to ligand binding, 40–607 of the EGFRs associate with DRMs, but rapidly after ligand binding, the receptors migrate out of DRMs and into clathrin-coated pits, where receptor-mediated endocytosis will occur (16Mineo C. Gill G.N. Anderson R.G. J. Biol. Chem. 1999; 274: 30636-30643Abstract Full Text Full Text PDF PubMed Scopus (262) Google Scholar, 17Ringerike T. Blystad F.D. Levy F.O. Madshus I.H. Stang E. J. Cell Sci. 2002; 115: 1331-1340Crossref PubMed Google Scholar). Cbl plays a dual role in the sorting and degradation process of the receptors. First, Cbl binds to and ubiquitinylates the phosphorylated receptors and thus sorts them for degradation by the proteasome. Second, Src-mediated tyrosine phosphorylation of Cbl induces conformational changes in Cbl. This facilitates binding of the CIN85-endophilin complex and subsequent endocytosis and lysosomal degradation of the activated receptors (18Soubeyran P. Kowanetz K. Szymkiewicz I. Langdon W.Y. Dikic I. Nature. 2002; 416: 183-187Crossref PubMed Scopus (489) Google Scholar, 19Petrelli A. Gilestro G.F. Lanzardo S. Comoglio P.M. Migone N. Giordano S. Nature. 2002; 416: 187-190Crossref PubMed Scopus (380) Google Scholar). DRM-associated SFKs are negatively regulated through C-terminal phosphorylation catalyzed by a DRM-associated pool of Csk (20Kawabuchi M. Satomi Y. Takao T. Shimonishi Y. Nada S. Nagai K. Tarakhovsky A. Okada M. Nature. 2000; 404: 999-1003Crossref PubMed Scopus (464) Google Scholar, 21Brdicka T. Pavlistova D. Leo A. Bruyns E. Korinek V. Angelisova P. Scherer J. Shevchenko A. Hilgert I. Cerny J. Drbal K. Kuramitsu Y. Kornacker B. Horejsi V. Schraven B. J. Exp. Med. 2000; 191: 1591-1604Crossref PubMed Scopus (405) Google Scholar, 22Torgersen K.M. Vang T. Abrahamsen H. Yaqub S. Horejsi V. Schraven B. Rolstad B. Mustelin T. Tasken K. J. Biol. Chem. 2001; 276: 29313-29318Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar, 23Vang T. Torgersen K.M. Sundvold V. Saxena M. Levy F.O. Skalhegg B.S. Hansson V. Mustelin T. Tasken K. J. Exp. Med. 2001; 193: 497-507Crossref PubMed Scopus (269) Google Scholar). Csk is recruited to membrane microdomains via binding of its SH2 domain to phosphorylated Tyr317 in the phosphoprotein Cbp/PAG, which is ubiquitously expressed and partitions exclusively into the DRM fraction (20Kawabuchi M. Satomi Y. Takao T. Shimonishi Y. Nada S. Nagai K. Tarakhovsky A. Okada M. Nature. 2000; 404: 999-1003Crossref PubMed Scopus (464) Google Scholar, 21Brdicka T. Pavlistova D. Leo A. Bruyns E. Korinek V. Angelisova P. Scherer J. Shevchenko A. Hilgert I. Cerny J. Drbal K. Kuramitsu Y. Kornacker B. Horejsi V. Schraven B. J. Exp. Med. 2000; 191: 1591-1604Crossref PubMed Scopus (405) Google Scholar). A similar SH2-phosphotyrosine-mediated interaction has recently been demonstrated between Csk and Tyr14 in caveolin, which is an important constituent of caveolae (24Cao H. Courchesne W.E. Mastick C.C. J. Biol. Chem. 2002; 277: 8771-8774Abstract Full Text Full Text PDF PubMed Scopus (180) Google Scholar). In addition to spatial regulation, different mechanisms for modulation of Csk kinase activity have been described. First, Csk kinase activity is induced 2–4-fold by binding of the Csk SH2 domain to Cbp/PAG (25Takeuchi S. Takayama Y. Ogawa A. Tamura K. Okada M. J. Biol. Chem. 2000; 275: 29183-29186Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar). Second, Gऔγ can also increase Csk kinase activity 2-fold, probably via interaction with the catalytic domain (26Lowry W.E. Huang J. Ma Y.C. Ali S. Wang D. Williams D.M. Okada M. Cole P.A. Huang X.Y. Dev. Cell. 2002; 2: 733-744Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). Last, Csk kinase activity can be regulated via covalent modification as well; in T cells, we recently demonstrated a 2–4-fold induction of Csk phosphotransferase activity after protein kinase A (PKA)-mediated phosphorylation of Ser364 in Csk, with a concomitant increase in C-terminal inhibitory phosphorylation of lipid raft-associated Lck (23Vang T. Torgersen K.M. Sundvold V. Saxena M. Levy F.O. Skalhegg B.S. Hansson V. Mustelin T. Tasken K. J. Exp. Med. 2001; 193: 497-507Crossref PubMed Scopus (269) Google Scholar). PKA, Csk, and SFKs are present in all cell types. Therefore, we hypothesized that PKA-mediated Csk activation could regulate C-terminal phosphorylation and activity of other SFKs in addition to Lck. Here, we demonstrate that an adenylyl cyclase (AC)-PKA-Csk inhibitory pathway down-regulates the kinase activity of Src in both human embryonic kidney (HEK293) cells and NIH-3T3 fibroblasts, suggesting the presence of a general principle for PKA-mediated inhibition of SFKs through activation of Csk. We further suggest that the PKA-Csk-mediated Src down-regulation plays an important role in modulating signaling downstream of Src. EGF, prostaglandin E2, and methyl-औ-cyclodextrin were purchased from Sigma;n-octyl-औ-d-glucoside was from U.S. Biochemical Corp.; and H89 and forskolin were from Calbiochem. Anti-Tyr(P) monoclonal antibody (4G10; Upstate Biotechnology, Inc., Lake Placid, NY), anti-Csk, anti-LAT, anti-Grb2, anti-HA, and antibodies toward PKA subunits RIα and C were as before (22Torgersen K.M. Vang T. Abrahamsen H. Yaqub S. Horejsi V. Schraven B. Rolstad B. Mustelin T. Tasken K. J. Biol. Chem. 2001; 276: 29313-29318Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar, 23Vang T. Torgersen K.M. Sundvold V. Saxena M. Levy F.O. Skalhegg B.S. Hansson V. Mustelin T. Tasken K. J. Exp. Med. 2001; 193: 497-507Crossref PubMed Scopus (269) Google Scholar). The phosphotyrosine-specific antibodies toward Tyr416 and Tyr527 in Src were from BIOSOURCEInternational (catalog no. 44-660 and 44-662, respectively), whereas monoclonal anti-Src antibody was from Upstate Biotechnology (clone GD11). Anti-caveolin (catalog no. C37120-150/610407) was from Transduction Laboratories, and both anti-Cbl and anti-EGFR were from Santa Cruz Biotechnology, Inc. (catalog no. sc-170 and sc-120, respectively), and anti-FAK was from Upstate Biotechnology (catalog no. 05-537). PKA-substrate phosphospecific antibody (abbreviated anti-RXX(PS/PT), where pS and pT represent phosphoserine and phosphothreonine, respectively) was purchased fromCell Signaling Technology and is reactive toward phosphorylated threonine with arginine in position −3 or toward phosphoserine with arginine in position −2 or −3. The antibody that recognizes Src only when it is not phosphorylated on Tyr416(nonphospho-416) was also purchased from Cell Signaling Technology(catalog no. 2102). Anti-Cbp/PAG was kindly provided by Dr. Václav Horejsı́ (Institute of Molecular Genetics AS CR, Prague, Czech Republic). The human leukemia T cell line Jurkat TAg, a derivative of the Jurkat cell line stably transfected with the SV40 large T antigen, was kept in logarithmic growth in RPMI medium supplemented with 107 fetal calf serum and antibiotics. HEK293 (human embryonic kidney cells, purchased from ATCC (Manassas, VA), catalog no. CRL-1573) and mouse NIH-3T3 and NIH-3T3-A14 cells (mouse NIH-3T3 cells expressing insulin receptor, kindly provided by Boudewijn Burgering, University Medical Center of Utrecht, The Netherlands) were kept in logarithmic growth in minimal essential medium supplemented with glutamine, sodium pyruvate, nonessential amino acids, 107 fetal calf serum, and antibiotics (penicillin and streptomycin) and split by trypsination at less than 807 confluence. When indicated, cells were serum-starved for 12–16 h before cell stimulation. For transfections, Jurkat TAg cells (2 × 107) in 0.4 ml of Opti-MEM were mixed with 10 ॖg of each DNA construct in electroporation cuvettes with a 0.4-cm electrode gap (Bio-Rad) and subjected to an electric field of 250 V/cm with 960-microfarad capacitance. The cells were expanded in complete medium and harvested 20 h post-transfection. Transfection of HEK293, NIH-3T3, and NIH-3T3-A14 cells were performed with LipofectAMINE according to the manufacturer's instruction (1:2.5 DNA/LipofectAMINE in Opti-MEM). After incubation for 5 h at 37 °C, the transfection solution was removed, and regular medium was added. Cells were harvested by trypsination 20–24 h later. Isolation of detergent-resistant membranes was performed as described in detail elsewhere (27Zhang W. Trible R.P. Samelson L.E. Immunity. 1998; 9: 239-246Abstract Full Text Full Text PDF PubMed Scopus (752) Google Scholar). Briefly, cells were homogenized in 1 ml of an ice-cold standard lysis buffer (50 mm HEPES, pH 7.4, 100 mm NaCl, 5 mmEDTA, 0.77 Triton X-100 with 1 mm sodium orthovanadate, 1 mm phenylmethylsulfonyl fluoride, 10 mm sodium pyrophosphate, and 50 mm sodium fluoride) by 10 pestle strokes in a Dounce homogenizer, loaded at the bottom of a 40 to 57 sucrose gradient (in 25 mm MES, pH 6.5, 5 mmEDTA, 150 mm NaCl) and ultracentrifuged at 200,000 ×g for 20 h at 4 °C. Fractions (0.4 ml) were collected from the top. In the presence of ATP (1 mm) and MgCl2 (15 mm), DRM fractions were stimulated with or without forskolin (100 ॖm, 10 min if not otherwise stated) or prostaglandin E2 (100 ॖm, 2 min if not otherwise stated) at 30 °C; thereafter, either cAMP measurements or immunoprecipitations were conducted. Mixed DRM fractions subjected to immunoprecipitations were always solubilized with 50 mmn-octyl-औ-d-glucoside. Cells were stimulated with forskolin (100 ॖm, 10 min if not otherwise stated) or prostaglandin E2 (100 ॖm, 2 min if not otherwise stated) in the absence and presence of EGF (100 ng/ml) at 37 °C and then disrupted in ice-cold standard lysis buffer containingn-octyl-औ-d-glucoside (50 mm) and subjected to immunoprecipitation with the indicated antibodies. After incubation at 4 °C for a minimum of 4 h, protein A/G-Sepharose (Santa Cruz Biotechnology) was added, and the incubation continued for 1 h. Immune complexes were washed three times in lysis buffer and subjected to Western blot analysis. Src was immunoprecipitated by using monoclonal Src antibodies (clone GD11; Upstate Biotechnology) and washed three times in standard lysis buffer, and then precipitates were divided into three or four. Immunoblot analysis was either conducted on precipitates directly using phosphotyrosine-specific antibodies toward Tyr527 in Src or nonphospho-416 in Src or monoclonal Src antibodies or washed three times in a buffer containing Hepes (50 mm, pH 7.4) and MgCl2 (5 mm) and then assayed for autophosphorylation in the presence of 1 mm ATP at 30 °C for 10 min. Reactions were stopped by the addition of SDS sample buffer, and immunoblot analysis was performed by using phosphotyrosine-specific antibodies toward Tyr416 in Src. When Src phosphotransferase activity was measured toward poly(Glu,Tyr) 4:1 (Sigma) as a substrate, the protocol was as for a Csk kinase assay. Csk was immunoprecipitated with polyclonal antibodies and washed three times in standard lysis buffer. After three additional washes in Hepes buffer (50 mm, pH 7.4) with 50 MgCl2, the Csk phosphotransferase activity was measured as incorporation of [32P]phosphate into the synthetic polyamino acid poly(Glu,Tyr) 4:1 (Sigma). A standard protocol was followed (28Vang T. Tasken K. Skalhegg B.S. Hansson V. Levy F.O. Biochim. Biophys. Acta. 1998; 1384: 285-293Crossref PubMed Scopus (11) Google Scholar) with reaction volumes of 50 ॖl containing 50 mm Hepes buffer, pH 7.4, 5 mmMgCl2, 200 ॖm [γ-32P]ATP (0.15 Ci/mmol), 200 ॖg/ml poly(Glu,Tyr), and immunoprecipitated Csk. The reactions were incubated at 30 °C for 12 min. Equal amounts of immunoprecipitated Csk in each kinase reaction were verified by immunoblotting. A standard cAMP assay (cAMP kit from PerkinElmer Life Sciences; catalog no. SP004) was performed in accordance with the manufacturer's instructions. Proteins were quantified by the method of Bradford (29Bradford M.M. Anal. Biochem. 1976; 72: 248-254Crossref PubMed Scopus (216377) Google Scholar) using γ-globulin as a standard. The different Csk constructs employed have been described elsewhere (22Torgersen K.M. Vang T. Abrahamsen H. Yaqub S. Horejsi V. Schraven B. Rolstad B. Mustelin T. Tasken K. J. Biol. Chem. 2001; 276: 29313-29318Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar). The plasmids encoding wild type Src or Src-Y527F were purchased from Upstate Biotechnology (catalog nos. 21-114 and 21-115). The Src-S17A mutant was generated by site-directed mutagenesis (QuikChange; Stratagene) of Src wild type, and the resulting construct was sequenced (GATC Biotech). We have previously shown that cAMP/PKA, through phosphorylation dependent activation of Csk, specifically interfere with and inhibit Lck-mediated signaling in T cells (23Vang T. Torgersen K.M. Sundvold V. Saxena M. Levy F.O. Skalhegg B.S. Hansson V. Mustelin T. Tasken K. J. Exp. Med. 2001; 193: 497-507Crossref PubMed Scopus (269) Google Scholar). To test the hypothesis that a general pathway from GPCRs via cAMP and PKA to Csk negatively regulates different SFKs, we studied HEK293 cells. These cells are of epithelial origin (proximal tubules in kidney) and express the SFK Src. Treatment of HEK293 cells with the cAMP-elevating agent forskolin and subsequent kinase assay with immunoprecipitated Src revealed strongly reduced ability to autophosphorylate on Tyr416(densitometric scanning analysis: 4.8 ± 1.5-fold decrease, average ± S.E. n = 3), whereas there was a concomitant increase in phosphorylation of Tyr527 (1.6 ± 0.1-fold increase, average ± S.E., n = 3) (Fig. 1A). Similar results were obtained when immunoprecipitated Src not subjected to the autophosphorylation assay was analyzed for Tyr(P)416(1.5 ± 0.2-fold decrease with forskolin, average ± S.E.,n = 3) or Tyr(P)527 (1.9 ± 0.1-fold increase with forskolin, average ± S.E., n = 3) levels (data not shown). The phosphotransferase activity of Src toward the synthetic polyamino acid poly(Glu,Tyr) was also lowered upon forskolin treatment of HEK cells (Fig. 1B). This points toward a role for cAMP in down-regulation of the kinase activity of Src. Furthermore, the physiological cAMP-elevating agent prostaglandin E2 (PGE2) had an inhibitory effect on Src similar to that of forskolin (Fig. 1C). Pretreatment of cells with the PKA inhibitor H89 blocked this effect, implicating PKA in cAMP-mediated regulation of Src (Fig. 1C). Similar results were also obtained in NIH-3T3-A14 cells (fibroblasts) (Fig.1D) and in HEK293 cells treated with forskolin in the presence or absence of another PKA inhibitor, KT5720 (100 nm, 15 min, data not shown). Taken together, these results suggest that cAMP is acting through PKA via Csk to inhibit Src in these cell types. Since all SFKs have lipid modifications involved in membrane targeting, we wanted to investigate the inhibitory effect of cAMP on Src in the DRM fractions of HEK293 cells. Upon sucrose gradient fractionation of HEK293 cells, less than 0.57 of total cellular protein was found in the DRM fractions (fractions 2–5, Fig.2A) compared with the soluble fractions (fractions 9–12). Nevertheless, upon the addition of Mg-ATP to the different fractions to reconstitute AC activity, forskolin-induced cAMP production peaked in the DRM fractions (Fig.2B). Similarly, PGE2 also induced cAMP production in DRM fractions (data not shown), indicating the presence of both PGE2 receptors and AC in membrane microdomains. Furthermore, small amounts of PKA C and RIα subunits as well as Csk were found in DRM fractions of HEK293 cells, whereas substantial amounts of Src partitioned into these fractions (Fig. 2C). Since the PKA-Csk-Lck pathway in T cells has been found localized to lipid rafts, we wanted to see if a similar pattern of regulation existed for Src in DRMs of HEK293 cells. Interestingly, forskolin treatment of HEK293 DRMs reconstituted with Mg-ATP was sufficient to down-regulate the activity of Src associated with these fractions, as indicated by increased Tyr527 phosphorylation (Fig.2D), reduced Tyr416 autophosphorylation (Fig.2D), and lowered phosphotransferase activity toward poly(Glu,Tyr) (Fig. 2E). Similar results were also obtained with lipid rafts from Jurkat TAg T-cells transfected with wild type Src (Fig. 2D) and with DRMs from NIH-3T3 fibroblasts (Fig.2E). Disruption of plasma membrane organization by cholesterol depletion with methyl-औ-cyclodextrin lowered the amounts of Src, Csk, and PKA C subunit present in DRMs (Fig. 2F) but not in soluble fractions (data not shown). As a result, disruption of membrane microdomain integrity by methyl-औ-cyclodextrin completely abolished forskolin-induced Tyr527 phosphorylation of Src (Fig. 2G). Altogether, these data show that all of the components necessary for mediating the inhibitory effect of cAMP on Src are present in DRMs and need to be properly organized to be fully operative. Csk phosphorylates Tyr527 in Src, and the observation that cAMP induces phosphorylation of this site indicates a role for Csk. Therefore, we next studied the effect of increased cAMP on Csk activity in DRMs from HEK293 cells. Stimulation of Mg-ATP-reconstituted DRM fractions with forskolin or PGE2led to a 2–2.5-fold increase in Csk kinase activity (Fig.3A). This is consistent with previous findings in T cells (23Vang T. Torgersen K.M. Sundvold V. Saxena M. Levy F.O. Skalhegg B.S. Hansson V. Mustelin T. Tasken K. J. Exp. Med. 2001; 193: 497-507Crossref PubMed Scopus (269) Google Scholar). In order to identify a role for Csk in cAMP-induced Src inhibition, we overexpressed kinase-deficient Csk-SH3-SH2, an interfering mutant that can displace endogenous Csk from its anchor protein Cbp/PAG in lipid rafts (22Torgersen K.M. Vang T. Abrahamsen H. Yaqub S. Horejsi V. Schraven B. Rolstad B. Mustelin T. Tasken K. J. Biol. Chem. 2001; 276: 29313-29318Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar). When expressed in HEK293 cells, this mutant was properly recruited into DRMs (Fig.3B). Interestingly, expression of this mutant caused spontaneous hyperphosphorylation of Tyr416 in DRM-associated Src (Fig. 3C) and abolished the inhibitory effect of cAMP on Src kinase activity (Fig. 3D). In conclusion, these results show that cAMP/PKA-mediated activation of Csk down-regulates Src kinase activity in DRMs of HEK293 cells. Previous studies have indicated that Src can be phosphorylated on Ser17 by PKA, but no physiological role for this phosphorylation was demonstrated (3Brown M.T. Cooper J.A. Biochim. Biophys. Acta. 1996; 1287: 121-149Crossref PubMed Scopus (1085) Google Scholar). However, it was recently reported that direct phosphorylation of Ser17 in Src by PKA up-regulates Src kinase activity (30Schmitt J.M. Stork P.J. Mol. Cell. 2002; 9: 85-94Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). To assess direct regulation of Src versus indirect PKA-mediated regulation via Csk in DRMs, we overexpressed either wild type Src or mutant Src-S17A in HEK293 cells. The contribution of endogenous Src in these experiments was negligible due to very low expression levels compared with those of the transfected constructs (Fig.4A, upper panel). Both wild type Src and Src-S17A partitioned into DRM fractions, and compared with control transfected cells (empty vector), the levels of these proteins in DRMs exceeded endogenous Src protein levels severalfold (Fig. 4A, lower panels). Although forskolin treatment reduced Tyr416 phosphorylation of both DRM-associated wild type Src and Src-S17A, no changes in DRM association were observed (Fig.4B). Furthermore, wild type Src and Src-S17A immunoprecipitated from DRM fractions stimulated with forskolin revealed comparable reductions in Tyr416 phosphorylation, suggesting that phosphorylation of Ser17 does not appear to regulate Src activity (Fig. 4C). This is also consistent with previous findings (31Yaciuk P. Choi J.K. Shalloway D. Mol. Cell. Biol. 1989; 9: 2453-2463Crossref PubMed Scopus (19) Google Scholar). Treatment with forskolin did not change the phosphorylation status of Ser17 of immunoprecipitated wild type Src, as indicated after immunoblotting with a phosphospecific PKA-substrate antibody (Fig. 4C). However, PKA was clearly activated by forskolin in these experiments, since whole cell lysates analyzed with the phospho-specific PKA-substrate antibody showed increased phosphorylation compared with unstimulated cells (data not shown). Additionally, no significant differences were seen in phosphorylation status of Tyr416 in wild type Src compared with Src-S17A in Jurkat TAg cells stimulated with forskolin (Fig.4D). This indicates that in T cells, although normally not expressing Src, direct phosphorylation of Ser17 of Src is not likely to be involved in regulation of Src kinase activity. Constitutive phosphorylation of Ser17 in Src to ∼607 stoichiometry in vivo has been reported (3Brown M.T. Cooper J.A. Biochim. Biophys. Acta. 1996; 1287: 121-149Crossref PubMed Scopus (1085) Google Scholar). Since our results showed basal Ser17 phosphorylation that was not altered by forskolin-induced PKA activation (Fig. 4C), we speculated that PKA-mediated Ser17 phosphorylation was not physiologically relevant in this cellular setting. To fully study the impact of the PKA-mediated Ser17 phosphorylation, we incubated immunoprecipitated wild type Src or Src-S17A with increasing amounts of highly active recombinant PKA C subunit in the presence of Mg-ATP. Neither in wild type Src nor in Src-S17A did PKA change Ser17 phosphorylation, and autophosphorylation assays revealed that Tyr416 phosphorylation was unaffected (Fig.4E). PKA was clearly active, since it autophosphorylated in the same assay and was able to phosphorylate and activate recombinant Csk in a separate assay (data not shown). Thus, in the absence of the inhibitory Csk, PKA neither increased phosphorylation of Ser17 in Src nor affected the activity of Src or Src-S17A. In conclusion, DRM-associated Src is not directly phosphorylated by PKA in response to elevated cAMP in the cells examined, and the inhibitory effect of cAMP on Src in DRMs is therefore not mediated directly by PKA but rather via an indirect pathway involving Csk. Since cAMP indirectly inhibits Src kinase activity, we next wanted to assess the impact of this inhibitory pathway in a setting where Src normally would be activated. The EGFR, which partitions into DRMs (Fig. 5A) (17Ringerike T. Blystad F.D. Levy F.O. Madshus I.H. Stang E. J. Cell Sci. 2002; 115: 1331-1340Crossref PubMed Google Scholar), activates Src upon stimulation with EGF. We therefore wanted to investigate whether the AC-PKA-Csk pathway could intersect EGF-induced Src activation. Upon stimulation of HEK293 cells with EGF, Tyr416 phosphorylation of both DRM-associated Src and total cellular Src was increased. (Fig. 5, B and C, respectively). Similar results were also obtained with whole cell lysates from NIH-3T3-A14 cells (data not shown). Interestingly, EGF- induced activation of Src was significantly inhibited when cells were pretreated with forskolin (Fig. 5C). This could not be explained by lowered tyrosine phosphorylation of EGFR, since its phosphorylation status was unaffected by forskolin pretreatment in this system (Fig. 5D). However, overexpression of dominant negative Csk-SH3-SH2 fully restored EGF-induced Src activation even in the presence of forskolin (Fig. 5E). Thus, in a physiological setting cAMP can interfere with receptor tyrosine kinase-induced Src activation in a Csk-dependent manner. Since cAMP inhibits EGF-induced Src activation, we next investigated whether this could be reflected in reduced activation of proteins downstream of Src. Cbl is phosphorylated by Src upon EGFR activation (32Kassenbrock C.K. Hunter S. Garl P. Johnson G.L. Anderson S.M. J. Biol. Chem. 2002; 277: 24967-24975Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar) (our own observations in HEK293 and NIH-3T3-A14 cells; data not shown) and thereby contributes to receptor internalization (18Soubeyran P. Kowanetz K. Szymkiewicz I. Langdon W.Y. Dikic I. Nature. 2002; 416: 183-187Crossref PubMed Scopus (489) Google Scholar). Both in HEK293 cells and in A14 cells, EGF-induced tyrosine phosphorylation of Cbl was clearly inhibited when cells were pretreated with forskolin (Fig. 6, A andB). Another Src substrate, the focal adhesion kinase (FAK), was also inducibly tyrosine-phosphorylated upon EGF stimulation of NIH-3T3 transfected with wild-type Src, whereas forskolin pretreatment inhibited this phosphorylation (Fig. 6C). In contrast, the inhibitory effects of forskolin on EGF-induced FAK phosphorylation were abolished in NIH-3T3 cells transfected with Src-Y527F, which cannot be regulated by Csk (Fig. 6C), suggesting that the inhibitory effect of cAMP is dependent on Tyr527 in Src. Last, EGF-induced phosphorylation of Stat3 mediated through Src was also inhibited by cAMP (not shown). Altogether, this indicates that the cAMP/PKA pathway, through activation of Csk, intersects growth factor-induced signaling mediated by Src in DRMs. Cyclic AMP exerts inhibitory effects in many cell types and is known to interfere with mitogenic signaling from growth factors (33Burgering B.M. Pronk G.J. van Weeren P.C. Chardin P. Bos J.L. EMBO J. 1993; 12: 4211-4220Crossref PubMed Scopus (316) Google Scholar). Most of the studies published so far describing the inhibitory effect of cAMP have focused on the inhibition of growth factor-induced extracellular signal-regulated kinase signaling by cAMP, which is likely to be mediated by direct phosphorylation of Raf-1 by PKA (34Dumaz N. Light Y. Marais R. Mol. Cell. Biol. 2002; 22: 3717-3728Crossref PubMed Scopus (69) Google Scholar,35Dhillon A.S. Pollock C. Steen H. Shaw P.E. Mischak H. Kolch W. Mol. Cell. Biol. 2002; 22: 3237-3246Crossref PubMed Scopus (190) Google Scholar). Furthermore, we recently described a proximal inhibitory pathway in T cell lipid rafts whereby cAMP, through PKA type I, activates Csk by phosphorylation of Ser364 in Csk, leading to inhibition of Lck and of T cell activation through the T cell receptor complex (23Vang T. Torgersen K.M. Sundvold V. Saxena M. Levy F.O. Skalhegg B.S. Hansson V. Mustelin T. Tasken K. J. Exp. Med. 2001; 193: 497-507Crossref PubMed Scopus (269) Google Scholar). Here we demonstrate a new inhibitory pathway for regulation of growth factor-induced signaling, in which GPCRs that signal through cAMP-PKA activate Csk, which in turn directly inhibits Src activity. Several distinct mechanisms of Src activation exist. The activity of Src is tightly coupled to the phosphorylation status of the C-terminal negatively regulatory Tyr527 (11Imamoto A. Soriano P. Cell. 1993; 73: 1117-1124Abstract Full Text PDF PubMed Scopus (346) Google Scholar, 36Kmiecik T.E. Shalloway D. Cell. 1987; 49: 65-73Abstract Full Text PDF PubMed Scopus (411) Google Scholar, 37Cartwright C.A. Eckhart W. Simon S. Kaplan P.L. Cell. 1987; 49: 83-91Abstract Full Text PDF PubMed Scopus (228) Google Scholar), and a plausible model for activation of Src includes initial dephosphorylation of Tyr527. Furthermore, Src contains several other phosphorylation sites that have modulatory effects on Src kinase activity. For instance, increased Src activity at the onset of mitosis correlates with increased phosphorylation of serine (Ser72) and threonine (Thr34 and Thr46) residues in the unique domain of Src, probably catalyzed by Cdc2-cyclin complexes (38Chackalaparampil I. Shalloway D. Cell. 1988; 52: 801-810Abstract Full Text PDF PubMed Scopus (187) Google Scholar, 39Shenoy S. Choi J.K. Bagrodia S. Copeland T.D. Maller J.L. Shalloway D. Cell. 1989; 57: 763-774Abstract Full Text PDF PubMed Scopus (212) Google Scholar, 40Morgan D.O. Kaplan J.M. Bishop J.M. Varmus H.E. Cell. 1989; 57: 775-786Abstract Full Text PDF PubMed Scopus (166) Google Scholar, 41Kaech S. Covic L. Wyss A. Ballmer-Hofer K. Nature. 1991; 350: 431-433Crossref PubMed Scopus (41) Google Scholar). However, since neutralizing mutations of these residues do not abolish this activation of Src, one might speculate that these phosphorylations modulate the effects of other regulatory mechanisms acting on Src. Furthermore, Ser17phosphorylation by PKA is reported to occur in vitro(reviewed in Ref. 3Brown M.T. Cooper J.A. Biochim. Biophys. Acta. 1996; 1287: 121-149Crossref PubMed Scopus (1085) Google Scholar) and to directly activate Src in NIH-3T3 cells (30Schmitt J.M. Stork P.J. Mol. Cell. 2002; 9: 85-94Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). However, this site does not appear to have physiological impact, since (i) Ser17 appears to be constitutively phosphorylated in HEK293 cells (this report); (ii) we do not observe activation of Src in response to cAMP in HEK293 cells, not even when Csk is displaced and the inhibitory effect of PKA on Src via Csk is abolished (this report); (iii) we are not able to show PKA activation of Src in NIH-3T3 cells (this report) as reported in Ref. 30Schmitt J.M. Stork P.J. Mol. Cell. 2002; 9: 85-94Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar. In addition, the Src SH3 and SH2 domains contain other phosphorylation sites that can modulate activity, either by changing domain interactions or Src structure. Finally, phosphorylation-independent activation of Src can occur via Gαs or Gαi binding to the catalytic domain of Src, thereby inducing conformational changes that favor autophosphorylation of Tyr416 and subsequent Src activation (42Ma Y.C. Huang J. Ali S. Lowry W. Huang X.Y. Cell. 2000; 102: 635-646Abstract Full Text Full Text PDF PubMed Scopus (378) Google Scholar). Thus, several distinct mechanisms for Src regulation exist, although it remains largely unknown what is the contribution of the different regulatory mechanisms in a given physiological setting. Notwithstanding the above-discussed modulating mechanisms, C-terminal phosphorylation seems to be a crucial master switch in regulation of Src, and factors regulating Csk activity and localization would consequently be of physiological importance (43Cary L.A. Cooper J.A. Nature. 2000; 404 (947): 945Crossref PubMed Scopus (57) Google Scholar). In this study, we show that the majority of Src in HEK293 cells partitions into the DRM fraction, and given the established role of lipid rafts and caveolae in signal transduction, this probably reflects a physiologically relevant pool of Src involved in signal transduction. Thus, mechanisms regulating the activity of DRM-associated Csk and the total amount of Csk in DRMs are expected to affect Src signaling through caveolae/lipid rafts. Indeed, csk knockout cells reveal a phenotype with increased basal Tyr416 phosphorylation of Src and hardly detectable Tyr527 phosphorylation (11Imamoto A. Soriano P. Cell. 1993; 73: 1117-1124Abstract Full Text PDF PubMed Scopus (346) Google Scholar, 12Nada S. Yagi T. Takeda H. Tokunaga T. Nakagawa H. Ikawa Y. Okada M. Aizawa S. Cell. 1993; 73: 1125-1135Abstract Full Text PDF PubMed Scopus (363) Google Scholar). In accordance with this, we show that total amount of Csk associated with DRMs is important for Src regulation, since profound effects on basal Src kinase activity are observed by overexpression of dominant negative Csk, which competes endogenous Csk from its DRM association provided by Cbp/PAG. We also show that PKA-mediated phosphorylation and activation of endogenous DRM-associated Csk modulates Src kinase activity. This provides a model for a novel mechanism of cAMP-mediated inhibition of Src signaling by modulation of Csk activity. Importantly, this inhibitory pathway is able to interfere with growth factor-induced Src signaling, demonstrating its physiological relevance. Growth factor receptors such as the EGFR partition into DRMs, and our observations suggest that EGF stimulation results in activation of DRM-associated Src. Furthermore, physiological activation of Src upon EGF stimulation was clearly inhibited by pretreatment with cAMP- elevating agents. Previously, this has been explained by cAMP/PKA-mediated inhibition of EGF-induced tyrosine phosphorylation of the EGFR, resulting in reduced activation of Src (44Barbier A.J. Poppleton H.M. Yigzaw Y. Mullenix J.B. Wiepz G.J. Bertics P.J. Patel T.B. J. Biol. Chem. 1999; 274: 14067-14073Abstract Full Text Full Text PDF PubMed Scopus (47) Google Scholar). However, in the cell line studied here, no changes in tyrosine phosphorylation of the EGFR were observed, indicating that cAMP did not affect EGFR activation. In contrast, overexpression of dominant negative Csk abolished the inhibitory effect of cAMP on EGF-induced Src activation, which supports a role for Csk in this respect. Finally, the inhibitory effect of increased cAMP on EGF-induced phosphorylation of Cbl and FAK was profound. Regulation of EGF-induced phosphorylation of Stat3 was also observed (not shown). Altogether, this indicates that PKA through activation of Csk not only regulates basal activity of DRM-associated Src but potentially also intersects Src-mediated signaling in a physiological setting (model presented in Fig. 7). This also seems to be a general principle, since the same effects were seen in different cell types such as HEK293 and NIH-3T3 cells. The fact that cAMP/PKA can inhibit EGF-induced Cbl phosphorylation may have some important implications for EGFR turnover. Upon stimulation, DRM-associated EGFRs move out of DRMs and are internalized via clathrin-coated pits. Cbl directly binds to the phosphorylated EGFRs and becomes tyrosine-phosphorylated by Src (32Kassenbrock C.K. Hunter S. Garl P. Johnson G.L. Anderson S.M. J. Biol. Chem. 2002; 277: 24967-24975Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Recently, a Cbl-CIN85-endophilin complex was reported to play an important role in EGFR internalization, and tyrosine phosphorylation of Cbl is necessary for its interaction with CIN85 (18Soubeyran P. Kowanetz K. Szymkiewicz I. Langdon W.Y. Dikic I. Nature. 2002; 416: 183-187Crossref PubMed Scopus (489) Google Scholar). Therefore, it is tempting to speculate that the novel inhibitory cAMP pathway we describe in this report would affect EGFR turnover. Further studies will be necessary to address this issue. In summary, we report the presence of a novel inhibitory pathway in DRMs, whereby cAMP/PKA through activation of Csk intersects Src-mediated signaling. Since PKA, Csk, and SFKs are expressed in all cell types, this pathway may represent a general principle for regulation of SFKs. We are grateful for the technical assistance of Guri Opsahl.