Protein Kinase A Regulates Rac and Is Required for the Growth Factor-stimulated Migration of Carcinoma Cells
2001; Elsevier BV; Volume: 276; Issue: 51 Linguagem: Inglês
10.1074/jbc.m107235200
ISSN1083-351X
AutoresKathleen L. O’Connor, Arthur M. Mercurio,
Tópico(s)Coagulation, Bradykinin, Polyphosphates, and Angioedema
ResumoMembers of the Rho family of small GTPases, such as Rho and Rac, are required for actin cytoskeletal reorganization during the migration of carcinoma cells. Phosphodiesterases are necessary for this migration because they alleviate cAMP-dependent protein kinase (PKA)-mediated inhibition of RhoA (O'Connor, K. L., Shaw, L. M., and Mercurio, A. M. (1998) J. Cell Biol. 143, 1749–1760; O'Connor K. L., Nguyen, B.-K., and Mercurio, A. M. (2000),J. Cell Biol. 148, 253–258). In this study, we report that the migration of breast and squamous carcinoma cells toward either lysophosphatidic acid or epidermal growth factor involves not only phosphodiesterase activity but also cooperative signaling from PKA. Furthermore, we demonstrate that Rac1 activation in response to chemoattractant or β1 integrin clustering is regulated by PKA and that Rac1 is required for this migration. Also, we find that β1 integrin signaling stimulates the rapid and transient activation of PKA. A novel implication of these findings is that carcinoma cell migration is controlled by cAMP-dependent as well as cAMP inhibitory signaling mechanisms. Members of the Rho family of small GTPases, such as Rho and Rac, are required for actin cytoskeletal reorganization during the migration of carcinoma cells. Phosphodiesterases are necessary for this migration because they alleviate cAMP-dependent protein kinase (PKA)-mediated inhibition of RhoA (O'Connor, K. L., Shaw, L. M., and Mercurio, A. M. (1998) J. Cell Biol. 143, 1749–1760; O'Connor K. L., Nguyen, B.-K., and Mercurio, A. M. (2000),J. Cell Biol. 148, 253–258). In this study, we report that the migration of breast and squamous carcinoma cells toward either lysophosphatidic acid or epidermal growth factor involves not only phosphodiesterase activity but also cooperative signaling from PKA. Furthermore, we demonstrate that Rac1 activation in response to chemoattractant or β1 integrin clustering is regulated by PKA and that Rac1 is required for this migration. Also, we find that β1 integrin signaling stimulates the rapid and transient activation of PKA. A novel implication of these findings is that carcinoma cell migration is controlled by cAMP-dependent as well as cAMP inhibitory signaling mechanisms. phosphodiesterase cAMP-dependent protein kinase Dulbecco's modified Eagle's medium bovine serum albumin lysophosphatidic acid epidermal growth factor isobutylmethylxanthine protein kinase A inhibitor monoclonal antibody Rac binding domain of Pak1 glutathioneS-transferase major histocompatibility complex Cell migration in response to growth factors and chemoattractants is essential for embryonic development, tissue homeostasis, and the immune response (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar). It is also a major factor in the pathogenesis of many human diseases, including cancer (2Stetler-Stevenson W.G. Aznavoorian S. Liotta L.A. Annu. Rev. Cell Biol. 1993; 9: 541-573Crossref PubMed Scopus (1515) Google Scholar). This complex process involves dynamic and coordinated interactions among integrins, chemoattractant receptors, and the actin cytoskeleton, which result in actin polymerization at the leading edge and contraction of actin bundles within the cell body to promote translocation. These dynamic changes in the actin cytoskeleton are initiated by the engagement of chemoattractant and integrin receptors on the cell surface with the respective ligands. Such interactions trigger cascades of signaling events that result in the remodeling of the actin cytoskeleton and consequent directed movement (3Condeelis J. Annu. Rev. Cell Biol. 1993; 9: 411-444Crossref PubMed Scopus (396) Google Scholar). To understand cell migration at a mechanistic level, these signaling events need to be defined and linked to both cell surface receptors and actin dynamics. Members of the Rho family of GTPases, including Rho, Rac, and Cdc42 in particular, are considered to be key signaling intermediates for cell migration (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar,4Shaw L.M. Rabinovitz I. Wang H.H.-F. Toker A. Mercurio A.M. Cell. 1997; 91: 949-960Abstract Full Text Full Text PDF PubMed Scopus (542) Google Scholar, 5Keely J.P. Westwick J.K. Whitehead I.P. Der C.J. Parise L.V. Nature. 1997; 390: 632-636Crossref PubMed Scopus (649) Google Scholar, 6Sander E.E. van Delft S. ten Klooster J.P. Reid T. van der Kammen R.A. Michiels F. Collard J.G. J. Cell Biol. 1998; 143: 1385-1398Crossref PubMed Scopus (585) Google Scholar, 7Sander E.E. ten Klooster J.P. van Delft S. van der Kammen R.A. Collard J.G. J. Cell Biol. 1999; 147: 1009-1021Crossref PubMed Scopus (732) Google Scholar, 8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar). These GTPases are activated by signaling pathways initiated at the cell surface and in their activated or GTP-bound state stimulate downstream effectors that regulate actin polymerization and actin-myosin contraction (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar, 9Tapon N. Hall A. Curr. Opin. Cell Biol. 1998; 9: 86-92Crossref Scopus (690) Google Scholar). Despite considerable progress in understanding the function of the Rho GTPases, the mechanisms by which its activities are regulated by cell surface receptors and how this regulation relates to cell migration are unclear (10Symons M. Settleman J. Trends Cell Biol. 2000; 10: 415-419Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar).Recent studies have highlighted an important role for cAMP metabolism in the migration of carcinoma cells and in the regulation of RhoA function. Specifically, we found that cAMP-specific phosphodiesterases (PDEs)1 facilitate carcinoma cell migration as well as lamellae formation by lowering cAMP levels (11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar). More mechanistic studies have revealed that cAMP inhibits RhoA activity (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar, 14Laudanna C. Campbell J.M. Butcher E.C. J. Biol. Chem. 1997; 272: 24141-24144Abstract Full Text Full Text PDF PubMed Scopus (185) Google Scholar), which has been shown to be required for carcinoma migration and invasion (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 15Itoh K. Yoshioka K. Akedo H. Uehata M. Ishizake T. Narumiya S. Nat. Med. 1999; 5: 221-225Crossref PubMed Scopus (559) Google Scholar, 16Kusama T. Mukai M. Iwasaki T. Tatuta M. Matsumoto Y. Akedo H. Nakamura H. Cancer Res. 2001; 61: 4885-4891PubMed Google Scholar). In fact, based on these observations, the current study was initiated to test the hypothesis that PKA activity inhibits carcinoma cell migration. Surprisingly, we observed that chemoattractant-stimulated migration requires PKA activity as well as PDE activity. In other terms, our results infer that this migration is controlled by cAMP-dependent as well as cAMP inhibitory signaling mechanisms. We provide an explanation for this paradox by demonstrating that PKA is required for Rac activation by chemoattractants as well as β1 integrins, a function that contrasts with its inhibition of RhoA. An important implication of these findings is that localized fluctuations in the intracellular concentration of [cAMP]i may provide a spatial and temporal regulation of PKA activity that influences Rac and RhoA function.DISCUSSIONRecently, we established the importance of cAMP-dependent PDEs in the migration of carcinoma cells (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar,11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar). This PDE activity is needed to lower cAMP levels within the cell and facilitate activation of RhoA. In our current study, we show that either LPA- or EGF-stimulated migration involves not only cAMP-inhibitory signaling that requires PDE activity but also a cAMP-dependent mechanism mediated by PKA that involves Rac. Importantly, we find that the activation of Rac in carcinoma cells in response to either chemoattractant stimulation or β1integrin clustering requires PKA activity.Our results represent the first direct evidence that Rac activation can be controlled by cAMP/PKA. This finding is supported by other reports that have indirectly implicated Rac as a target of PKA. For example, PKA is necessary for cadherin-mediated cell-cell adhesion in epithelial cells (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar), a process that is dependent on Rac (25Braga V. Exp. Cell Res. 2000; 261: 83-90Crossref PubMed Scopus (106) Google Scholar). Also, Rac was shown to be an intermediate in the PKA-dependent activation of p38 mitogen-activated protein kinase (p38 MAPK) in response to thyroid-stimulating hormone (26Pomerance M. Abdullah H.-B. Kamerji S. Correze C. Blondeau J.-P. J. Biol. Chem. 2000; 275: 40539-40546Abstract Full Text Full Text PDF PubMed Scopus (83) Google Scholar). Here, we find that PKA is required for the Rac-dependent migration of carcinoma cells. In our study, we demonstrated that Rac activation in response to either β1 integrin clustering or chemoattractant stimulation is blocked by pharmacological inhibitors of PKA and not by inhibition of phosphodiesterases. In support of this link between PKA and Rac activation, we reported that the Rac-independent migration of clone A colon carcinoma cells does not require PKA (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar).An important implication of our findings is that PKA can contribute to the differential regulation of the Rac and Rho small GTPases. Recent studies (7Sander E.E. ten Klooster J.P. van Delft S. van der Kammen R.A. Collard J.G. J. Cell Biol. 1999; 147: 1009-1021Crossref PubMed Scopus (732) Google Scholar) have indicated a reciprocal relationship between Rac and Rho activation and have provided evidence that Rac can inhibit Rho activity. Our data suggest that increased PKA activity would facilitate Rac activation and impede Rho activation and vice versa, thus facilitating their reciprocal relationship. We and others have shown that cAMP/PKA-mediated signaling has an inhibitory effect on RhoA activity (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar, 14Laudanna C. Campbell J.M. Butcher E.C. J. Biol. Chem. 1997; 272: 24141-24144Abstract Full Text Full Text PDF PubMed Scopus (185) Google Scholar). The basis for this inhibition involves the direct phosphorylation of RhoA by PKA (12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar). The mechanism by which PKA stimulates Rac activation, however, is unlikely to involve its direct phosphorylation because Rac does not contain a consensus PKA phosphorylation site. PKA may regulate Rac indirectly by modifying the function of molecules that control Rac activation. For example, both Tiam-1 and Trio, which are guanine nucleotide exchange factors involved in Rac activation, have consensus PKA phosphorylation sites. Interestingly, Tiam-1 has been shown to be a target of LPA signaling (27Fleming I.N. Elliot C.M. Collard J.G. Exton J.H. J. Biol. Chem. 1997; 272: 33105-33110Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar) and, therefore, may provide a link between PKA and Rac activation.The PKA dependence of Rac activation in the cells we examined contrasts with the reported PKA-mediated inhibition of Pak1 (28Howe A.K. Juliano R.L. Nat. Cell Biol. 2000; 2: 593-600Crossref PubMed Scopus (180) Google Scholar). Pak1, which is a downstream effector of both Rac and Cdc42 (29Hall A. Science. 1998; 279: 509-514Crossref PubMed Scopus (5185) Google Scholar), has been implicated in migration (30Ren X.-D. Kiosses W.B. Schwartz M.A. EMBO J. 1999; 18: 578-585Crossref PubMed Scopus (1355) Google Scholar, 31Sells M.A. Boyd J.T. Chernoff J. J. Cell Biol. 1999; 145: 837-849Crossref PubMed Scopus (327) Google Scholar, 32Sells M.A. Pfaff A. Chernoff J. J. Cell Biol. 2000; 151: 1449-1457Crossref PubMed Scopus (134) Google Scholar). One explanation for this discrepancy is that Rac activation may require only a transient burst of PKA activity, a possibility that is supported by our data on PKA activation in response to β1 integrin clustering. The transient nature of PKA activation may permit Pak to be activated after complete Rac activation and diminution of PKA activity. Clearly, however, more studies are needed to understand the contribution of PKA to the regulation of Rac and Pak1 activation.Recent studies by our group and others have revealed that integrins are involved in the regulation of [cAMP]i and PKA activity (11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar, 24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar, 28Howe A.K. Juliano R.L. Nat. Cell Biol. 2000; 2: 593-600Crossref PubMed Scopus (180) Google Scholar, 33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar, 34Kim S. Harris M. Varner J.A. J. Biol. Chem. 2000; 275: 33920-33928Abstract Full Text Full Text PDF PubMed Scopus (188) Google Scholar). Moreover, our finding that β1 integrins can activate PKA is in agreement with studies in other systems (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar, 33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar). Most likely, β1integrin signaling stimulates adenyl cyclase activity and, consequently, a localized rise in [cAMP]i that increases PKA activity. Given the established importance of heterotrimeric (ht) G proteins in adenyl cyclase activation and cAMP signaling, integrin-mediated PKA activation could be facilitated by heterotrimeric G proteins. In fact, several recent studies have provided evidence that supports this possibility. For example, mechanical stresses applied to the cell surface stimulate cAMP signaling by modulating local release of signals generated by activated integrins in a G protein-dependent manner (33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar). Also, the αvβ3 integrin has been reported to activate ht-Gi by forming a complex with the integrin-associated protein CD47.An important conclusion from our experiments involving pharmacological inhibition of PDE and PKA is that these enzymes, whose activities are counter-opposed, function cooperatively to promote chemotactic migration. Their ability to signal cooperatively implies that these enzymes have distinct functions in migration. We suggest that they function together to create microgradients of cAMP/PKA within a cell that are important for migration. This hypothesis is supported by the finding that localized gradients of the [cAMP]i regulate growth cone movement (35Song H.-J. Ming G.-L. Poo M.-M. Nature. 1997; 388: 275-279Crossref PubMed Scopus (520) Google Scholar). The formation of cAMP/PKA gradient within migrating cells may differentially influence Rac and Rho, resulting in spatial and temporal differences in the activation of Rac and Rho. Such differences could be manifested, for example, in the Rac-mediated lamellipodial protrusion and Rho-mediated contractility necessary for migration (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar, 29Hall A. Science. 1998; 279: 509-514Crossref PubMed Scopus (5185) Google Scholar).PKA regulation of Rac activation may have another important consequence for directed migration. It has been argued that the attractant gradient across a migrating cell is insufficient to signal directionality and that the chemotactic signals at the leading edge need to be amplified to orient and polarize a cell (36Condeelis J.S. Wyckoff J.B. Bailly M. Pestell R. Lawrence D. Backer J. Segall J.E. Semin. Cancer Biol. 2001; 11: 119-128Crossref PubMed Scopus (116) Google Scholar). Our results suggest that PKA signaling may provide one mechanism for amplifying chemotactic signals. During chemotaxis, new β1 integrin contacts are formed at the leading edge of the cell in response to a chemoattractant gradient. It can be hypothesized that these new integrin contacts stimulate PKA and, as a consequence, amplify Rac activation. This scenario implies that activated Rac is localized at the leading edge of migrating cells. In fact, activated Rac has been localized at the leading edge of migrating cells using fluorescence resonance energy transfer (FRET) technology (37Kraynov V.L. Chamberlain C. Bokock G.M. Schwartz M.A. Slabaugh S. Hahn K.M. Science. 2000; 290: 333-337Crossref PubMed Scopus (560) Google Scholar). Furthermore, Rac1 has been implicated in orientingDictyostelium (38Chung C.Y. Lee S. Briscoe C. Ellsworth C. Firtel R.A. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5225-5230Crossref PubMed Scopus (113) Google Scholar) and lymphocytes (39del Pozo M.A. Vicente-Manzanares M. Tejedor R. Serrador J.M. Sanchez-Madrid F. Eur. J. Immunol. 1999; 29: 3609-3620Crossref PubMed Scopus (194) Google Scholar) during chemotaxis. In this context, it is worth noting that the activation of β1 integrins leads to the recruitment and colocalization of a PKA regulatory subunit with integrin complexes and the subsequent activation of PKA (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar). In summary, our findings highlight the importance of cAMP and PKA in the regulation of small GTPase function and migration and identify important venues for future work. Cell migration in response to growth factors and chemoattractants is essential for embryonic development, tissue homeostasis, and the immune response (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar). It is also a major factor in the pathogenesis of many human diseases, including cancer (2Stetler-Stevenson W.G. Aznavoorian S. Liotta L.A. Annu. Rev. Cell Biol. 1993; 9: 541-573Crossref PubMed Scopus (1515) Google Scholar). This complex process involves dynamic and coordinated interactions among integrins, chemoattractant receptors, and the actin cytoskeleton, which result in actin polymerization at the leading edge and contraction of actin bundles within the cell body to promote translocation. These dynamic changes in the actin cytoskeleton are initiated by the engagement of chemoattractant and integrin receptors on the cell surface with the respective ligands. Such interactions trigger cascades of signaling events that result in the remodeling of the actin cytoskeleton and consequent directed movement (3Condeelis J. Annu. Rev. Cell Biol. 1993; 9: 411-444Crossref PubMed Scopus (396) Google Scholar). To understand cell migration at a mechanistic level, these signaling events need to be defined and linked to both cell surface receptors and actin dynamics. Members of the Rho family of GTPases, including Rho, Rac, and Cdc42 in particular, are considered to be key signaling intermediates for cell migration (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar,4Shaw L.M. Rabinovitz I. Wang H.H.-F. Toker A. Mercurio A.M. Cell. 1997; 91: 949-960Abstract Full Text Full Text PDF PubMed Scopus (542) Google Scholar, 5Keely J.P. Westwick J.K. Whitehead I.P. Der C.J. Parise L.V. Nature. 1997; 390: 632-636Crossref PubMed Scopus (649) Google Scholar, 6Sander E.E. van Delft S. ten Klooster J.P. Reid T. van der Kammen R.A. Michiels F. Collard J.G. J. Cell Biol. 1998; 143: 1385-1398Crossref PubMed Scopus (585) Google Scholar, 7Sander E.E. ten Klooster J.P. van Delft S. van der Kammen R.A. Collard J.G. J. Cell Biol. 1999; 147: 1009-1021Crossref PubMed Scopus (732) Google Scholar, 8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar). These GTPases are activated by signaling pathways initiated at the cell surface and in their activated or GTP-bound state stimulate downstream effectors that regulate actin polymerization and actin-myosin contraction (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar, 9Tapon N. Hall A. Curr. Opin. Cell Biol. 1998; 9: 86-92Crossref Scopus (690) Google Scholar). Despite considerable progress in understanding the function of the Rho GTPases, the mechanisms by which its activities are regulated by cell surface receptors and how this regulation relates to cell migration are unclear (10Symons M. Settleman J. Trends Cell Biol. 2000; 10: 415-419Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar). Recent studies have highlighted an important role for cAMP metabolism in the migration of carcinoma cells and in the regulation of RhoA function. Specifically, we found that cAMP-specific phosphodiesterases (PDEs)1 facilitate carcinoma cell migration as well as lamellae formation by lowering cAMP levels (11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar). More mechanistic studies have revealed that cAMP inhibits RhoA activity (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar, 14Laudanna C. Campbell J.M. Butcher E.C. J. Biol. Chem. 1997; 272: 24141-24144Abstract Full Text Full Text PDF PubMed Scopus (185) Google Scholar), which has been shown to be required for carcinoma migration and invasion (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 15Itoh K. Yoshioka K. Akedo H. Uehata M. Ishizake T. Narumiya S. Nat. Med. 1999; 5: 221-225Crossref PubMed Scopus (559) Google Scholar, 16Kusama T. Mukai M. Iwasaki T. Tatuta M. Matsumoto Y. Akedo H. Nakamura H. Cancer Res. 2001; 61: 4885-4891PubMed Google Scholar). In fact, based on these observations, the current study was initiated to test the hypothesis that PKA activity inhibits carcinoma cell migration. Surprisingly, we observed that chemoattractant-stimulated migration requires PKA activity as well as PDE activity. In other terms, our results infer that this migration is controlled by cAMP-dependent as well as cAMP inhibitory signaling mechanisms. We provide an explanation for this paradox by demonstrating that PKA is required for Rac activation by chemoattractants as well as β1 integrins, a function that contrasts with its inhibition of RhoA. An important implication of these findings is that localized fluctuations in the intracellular concentration of [cAMP]i may provide a spatial and temporal regulation of PKA activity that influences Rac and RhoA function. DISCUSSIONRecently, we established the importance of cAMP-dependent PDEs in the migration of carcinoma cells (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar,11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar). This PDE activity is needed to lower cAMP levels within the cell and facilitate activation of RhoA. In our current study, we show that either LPA- or EGF-stimulated migration involves not only cAMP-inhibitory signaling that requires PDE activity but also a cAMP-dependent mechanism mediated by PKA that involves Rac. Importantly, we find that the activation of Rac in carcinoma cells in response to either chemoattractant stimulation or β1integrin clustering requires PKA activity.Our results represent the first direct evidence that Rac activation can be controlled by cAMP/PKA. This finding is supported by other reports that have indirectly implicated Rac as a target of PKA. For example, PKA is necessary for cadherin-mediated cell-cell adhesion in epithelial cells (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar), a process that is dependent on Rac (25Braga V. Exp. Cell Res. 2000; 261: 83-90Crossref PubMed Scopus (106) Google Scholar). Also, Rac was shown to be an intermediate in the PKA-dependent activation of p38 mitogen-activated protein kinase (p38 MAPK) in response to thyroid-stimulating hormone (26Pomerance M. Abdullah H.-B. Kamerji S. Correze C. Blondeau J.-P. J. Biol. Chem. 2000; 275: 40539-40546Abstract Full Text Full Text PDF PubMed Scopus (83) Google Scholar). Here, we find that PKA is required for the Rac-dependent migration of carcinoma cells. In our study, we demonstrated that Rac activation in response to either β1 integrin clustering or chemoattractant stimulation is blocked by pharmacological inhibitors of PKA and not by inhibition of phosphodiesterases. In support of this link between PKA and Rac activation, we reported that the Rac-independent migration of clone A colon carcinoma cells does not require PKA (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar).An important implication of our findings is that PKA can contribute to the differential regulation of the Rac and Rho small GTPases. Recent studies (7Sander E.E. ten Klooster J.P. van Delft S. van der Kammen R.A. Collard J.G. J. Cell Biol. 1999; 147: 1009-1021Crossref PubMed Scopus (732) Google Scholar) have indicated a reciprocal relationship between Rac and Rho activation and have provided evidence that Rac can inhibit Rho activity. Our data suggest that increased PKA activity would facilitate Rac activation and impede Rho activation and vice versa, thus facilitating their reciprocal relationship. We and others have shown that cAMP/PKA-mediated signaling has an inhibitory effect on RhoA activity (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar, 14Laudanna C. Campbell J.M. Butcher E.C. J. Biol. Chem. 1997; 272: 24141-24144Abstract Full Text Full Text PDF PubMed Scopus (185) Google Scholar). The basis for this inhibition involves the direct phosphorylation of RhoA by PKA (12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar). The mechanism by which PKA stimulates Rac activation, however, is unlikely to involve its direct phosphorylation because Rac does not contain a consensus PKA phosphorylation site. PKA may regulate Rac indirectly by modifying the function of molecules that control Rac activation. For example, both Tiam-1 and Trio, which are guanine nucleotide exchange factors involved in Rac activation, have consensus PKA phosphorylation sites. Interestingly, Tiam-1 has been shown to be a target of LPA signaling (27Fleming I.N. Elliot C.M. Collard J.G. Exton J.H. J. Biol. Chem. 1997; 272: 33105-33110Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar) and, therefore, may provide a link between PKA and Rac activation.The PKA dependence of Rac activation in the cells we examined contrasts with the reported PKA-mediated inhibition of Pak1 (28Howe A.K. Juliano R.L. Nat. Cell Biol. 2000; 2: 593-600Crossref PubMed Scopus (180) Google Scholar). Pak1, which is a downstream effector of both Rac and Cdc42 (29Hall A. Science. 1998; 279: 509-514Crossref PubMed Scopus (5185) Google Scholar), has been implicated in migration (30Ren X.-D. Kiosses W.B. Schwartz M.A. EMBO J. 1999; 18: 578-585Crossref PubMed Scopus (1355) Google Scholar, 31Sells M.A. Boyd J.T. Chernoff J. J. Cell Biol. 1999; 145: 837-849Crossref PubMed Scopus (327) Google Scholar, 32Sells M.A. Pfaff A. Chernoff J. J. Cell Biol. 2000; 151: 1449-1457Crossref PubMed Scopus (134) Google Scholar). One explanation for this discrepancy is that Rac activation may require only a transient burst of PKA activity, a possibility that is supported by our data on PKA activation in response to β1 integrin clustering. The transient nature of PKA activation may permit Pak to be activated after complete Rac activation and diminution of PKA activity. Clearly, however, more studies are needed to understand the contribution of PKA to the regulation of Rac and Pak1 activation.Recent studies by our group and others have revealed that integrins are involved in the regulation of [cAMP]i and PKA activity (11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar, 24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar, 28Howe A.K. Juliano R.L. Nat. Cell Biol. 2000; 2: 593-600Crossref PubMed Scopus (180) Google Scholar, 33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar, 34Kim S. Harris M. Varner J.A. J. Biol. Chem. 2000; 275: 33920-33928Abstract Full Text Full Text PDF PubMed Scopus (188) Google Scholar). Moreover, our finding that β1 integrins can activate PKA is in agreement with studies in other systems (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar, 33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar). Most likely, β1integrin signaling stimulates adenyl cyclase activity and, consequently, a localized rise in [cAMP]i that increases PKA activity. Given the established importance of heterotrimeric (ht) G proteins in adenyl cyclase activation and cAMP signaling, integrin-mediated PKA activation could be facilitated by heterotrimeric G proteins. In fact, several recent studies have provided evidence that supports this possibility. For example, mechanical stresses applied to the cell surface stimulate cAMP signaling by modulating local release of signals generated by activated integrins in a G protein-dependent manner (33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar). Also, the αvβ3 integrin has been reported to activate ht-Gi by forming a complex with the integrin-associated protein CD47.An important conclusion from our experiments involving pharmacological inhibition of PDE and PKA is that these enzymes, whose activities are counter-opposed, function cooperatively to promote chemotactic migration. Their ability to signal cooperatively implies that these enzymes have distinct functions in migration. We suggest that they function together to create microgradients of cAMP/PKA within a cell that are important for migration. This hypothesis is supported by the finding that localized gradients of the [cAMP]i regulate growth cone movement (35Song H.-J. Ming G.-L. Poo M.-M. Nature. 1997; 388: 275-279Crossref PubMed Scopus (520) Google Scholar). The formation of cAMP/PKA gradient within migrating cells may differentially influence Rac and Rho, resulting in spatial and temporal differences in the activation of Rac and Rho. Such differences could be manifested, for example, in the Rac-mediated lamellipodial protrusion and Rho-mediated contractility necessary for migration (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar, 29Hall A. Science. 1998; 279: 509-514Crossref PubMed Scopus (5185) Google Scholar).PKA regulation of Rac activation may have another important consequence for directed migration. It has been argued that the attractant gradient across a migrating cell is insufficient to signal directionality and that the chemotactic signals at the leading edge need to be amplified to orient and polarize a cell (36Condeelis J.S. Wyckoff J.B. Bailly M. Pestell R. Lawrence D. Backer J. Segall J.E. Semin. Cancer Biol. 2001; 11: 119-128Crossref PubMed Scopus (116) Google Scholar). Our results suggest that PKA signaling may provide one mechanism for amplifying chemotactic signals. During chemotaxis, new β1 integrin contacts are formed at the leading edge of the cell in response to a chemoattractant gradient. It can be hypothesized that these new integrin contacts stimulate PKA and, as a consequence, amplify Rac activation. This scenario implies that activated Rac is localized at the leading edge of migrating cells. In fact, activated Rac has been localized at the leading edge of migrating cells using fluorescence resonance energy transfer (FRET) technology (37Kraynov V.L. Chamberlain C. Bokock G.M. Schwartz M.A. Slabaugh S. Hahn K.M. Science. 2000; 290: 333-337Crossref PubMed Scopus (560) Google Scholar). Furthermore, Rac1 has been implicated in orientingDictyostelium (38Chung C.Y. Lee S. Briscoe C. Ellsworth C. Firtel R.A. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5225-5230Crossref PubMed Scopus (113) Google Scholar) and lymphocytes (39del Pozo M.A. Vicente-Manzanares M. Tejedor R. Serrador J.M. Sanchez-Madrid F. Eur. J. Immunol. 1999; 29: 3609-3620Crossref PubMed Scopus (194) Google Scholar) during chemotaxis. In this context, it is worth noting that the activation of β1 integrins leads to the recruitment and colocalization of a PKA regulatory subunit with integrin complexes and the subsequent activation of PKA (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar). In summary, our findings highlight the importance of cAMP and PKA in the regulation of small GTPase function and migration and identify important venues for future work. Recently, we established the importance of cAMP-dependent PDEs in the migration of carcinoma cells (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar,11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar). This PDE activity is needed to lower cAMP levels within the cell and facilitate activation of RhoA. In our current study, we show that either LPA- or EGF-stimulated migration involves not only cAMP-inhibitory signaling that requires PDE activity but also a cAMP-dependent mechanism mediated by PKA that involves Rac. Importantly, we find that the activation of Rac in carcinoma cells in response to either chemoattractant stimulation or β1integrin clustering requires PKA activity. Our results represent the first direct evidence that Rac activation can be controlled by cAMP/PKA. This finding is supported by other reports that have indirectly implicated Rac as a target of PKA. For example, PKA is necessary for cadherin-mediated cell-cell adhesion in epithelial cells (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar), a process that is dependent on Rac (25Braga V. Exp. Cell Res. 2000; 261: 83-90Crossref PubMed Scopus (106) Google Scholar). Also, Rac was shown to be an intermediate in the PKA-dependent activation of p38 mitogen-activated protein kinase (p38 MAPK) in response to thyroid-stimulating hormone (26Pomerance M. Abdullah H.-B. Kamerji S. Correze C. Blondeau J.-P. J. Biol. Chem. 2000; 275: 40539-40546Abstract Full Text Full Text PDF PubMed Scopus (83) Google Scholar). Here, we find that PKA is required for the Rac-dependent migration of carcinoma cells. In our study, we demonstrated that Rac activation in response to either β1 integrin clustering or chemoattractant stimulation is blocked by pharmacological inhibitors of PKA and not by inhibition of phosphodiesterases. In support of this link between PKA and Rac activation, we reported that the Rac-independent migration of clone A colon carcinoma cells does not require PKA (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar). An important implication of our findings is that PKA can contribute to the differential regulation of the Rac and Rho small GTPases. Recent studies (7Sander E.E. ten Klooster J.P. van Delft S. van der Kammen R.A. Collard J.G. J. Cell Biol. 1999; 147: 1009-1021Crossref PubMed Scopus (732) Google Scholar) have indicated a reciprocal relationship between Rac and Rho activation and have provided evidence that Rac can inhibit Rho activity. Our data suggest that increased PKA activity would facilitate Rac activation and impede Rho activation and vice versa, thus facilitating their reciprocal relationship. We and others have shown that cAMP/PKA-mediated signaling has an inhibitory effect on RhoA activity (8O'Connor K.L. Nguyen B.-K. Mercurio A.M. J. Cell Biol. 2000; 148: 253-258Crossref PubMed Scopus (180) Google Scholar, 12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar, 14Laudanna C. Campbell J.M. Butcher E.C. J. Biol. Chem. 1997; 272: 24141-24144Abstract Full Text Full Text PDF PubMed Scopus (185) Google Scholar). The basis for this inhibition involves the direct phosphorylation of RhoA by PKA (12Dong J.-M. Leung T. Manser E. Lim L. J. Biol. Chem. 1998; 273: 22554-22562Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar, 13Lang P. Gesbert F. Delespine-Carmagnat M. Stancou R. Pouchelet M. Bertoglio J. EMBO J. 1996; 15: 510-519Crossref PubMed Scopus (479) Google Scholar). The mechanism by which PKA stimulates Rac activation, however, is unlikely to involve its direct phosphorylation because Rac does not contain a consensus PKA phosphorylation site. PKA may regulate Rac indirectly by modifying the function of molecules that control Rac activation. For example, both Tiam-1 and Trio, which are guanine nucleotide exchange factors involved in Rac activation, have consensus PKA phosphorylation sites. Interestingly, Tiam-1 has been shown to be a target of LPA signaling (27Fleming I.N. Elliot C.M. Collard J.G. Exton J.H. J. Biol. Chem. 1997; 272: 33105-33110Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar) and, therefore, may provide a link between PKA and Rac activation. The PKA dependence of Rac activation in the cells we examined contrasts with the reported PKA-mediated inhibition of Pak1 (28Howe A.K. Juliano R.L. Nat. Cell Biol. 2000; 2: 593-600Crossref PubMed Scopus (180) Google Scholar). Pak1, which is a downstream effector of both Rac and Cdc42 (29Hall A. Science. 1998; 279: 509-514Crossref PubMed Scopus (5185) Google Scholar), has been implicated in migration (30Ren X.-D. Kiosses W.B. Schwartz M.A. EMBO J. 1999; 18: 578-585Crossref PubMed Scopus (1355) Google Scholar, 31Sells M.A. Boyd J.T. Chernoff J. J. Cell Biol. 1999; 145: 837-849Crossref PubMed Scopus (327) Google Scholar, 32Sells M.A. Pfaff A. Chernoff J. J. Cell Biol. 2000; 151: 1449-1457Crossref PubMed Scopus (134) Google Scholar). One explanation for this discrepancy is that Rac activation may require only a transient burst of PKA activity, a possibility that is supported by our data on PKA activation in response to β1 integrin clustering. The transient nature of PKA activation may permit Pak to be activated after complete Rac activation and diminution of PKA activity. Clearly, however, more studies are needed to understand the contribution of PKA to the regulation of Rac and Pak1 activation. Recent studies by our group and others have revealed that integrins are involved in the regulation of [cAMP]i and PKA activity (11O'Connor K.L. Shaw L.M. Mercurio A.M. J. Cell Biol. 1998; 143: 1749-1760Crossref PubMed Scopus (136) Google Scholar, 24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar, 28Howe A.K. Juliano R.L. Nat. Cell Biol. 2000; 2: 593-600Crossref PubMed Scopus (180) Google Scholar, 33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar, 34Kim S. Harris M. Varner J.A. J. Biol. Chem. 2000; 275: 33920-33928Abstract Full Text Full Text PDF PubMed Scopus (188) Google Scholar). Moreover, our finding that β1 integrins can activate PKA is in agreement with studies in other systems (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar, 33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar). Most likely, β1integrin signaling stimulates adenyl cyclase activity and, consequently, a localized rise in [cAMP]i that increases PKA activity. Given the established importance of heterotrimeric (ht) G proteins in adenyl cyclase activation and cAMP signaling, integrin-mediated PKA activation could be facilitated by heterotrimeric G proteins. In fact, several recent studies have provided evidence that supports this possibility. For example, mechanical stresses applied to the cell surface stimulate cAMP signaling by modulating local release of signals generated by activated integrins in a G protein-dependent manner (33Meyer C.J. Alenghar F.J. Rim P. Fong J.H.-J. Fabray B. Ingber D.E. Nature Cell Biol. 2000; 2: 666-668Crossref PubMed Scopus (221) Google Scholar). Also, the αvβ3 integrin has been reported to activate ht-Gi by forming a complex with the integrin-associated protein CD47. An important conclusion from our experiments involving pharmacological inhibition of PDE and PKA is that these enzymes, whose activities are counter-opposed, function cooperatively to promote chemotactic migration. Their ability to signal cooperatively implies that these enzymes have distinct functions in migration. We suggest that they function together to create microgradients of cAMP/PKA within a cell that are important for migration. This hypothesis is supported by the finding that localized gradients of the [cAMP]i regulate growth cone movement (35Song H.-J. Ming G.-L. Poo M.-M. Nature. 1997; 388: 275-279Crossref PubMed Scopus (520) Google Scholar). The formation of cAMP/PKA gradient within migrating cells may differentially influence Rac and Rho, resulting in spatial and temporal differences in the activation of Rac and Rho. Such differences could be manifested, for example, in the Rac-mediated lamellipodial protrusion and Rho-mediated contractility necessary for migration (1Horwitz A.R. Parsons J.T. Science. 2000; 286: 1102-1103Crossref Scopus (343) Google Scholar, 29Hall A. Science. 1998; 279: 509-514Crossref PubMed Scopus (5185) Google Scholar). PKA regulation of Rac activation may have another important consequence for directed migration. It has been argued that the attractant gradient across a migrating cell is insufficient to signal directionality and that the chemotactic signals at the leading edge need to be amplified to orient and polarize a cell (36Condeelis J.S. Wyckoff J.B. Bailly M. Pestell R. Lawrence D. Backer J. Segall J.E. Semin. Cancer Biol. 2001; 11: 119-128Crossref PubMed Scopus (116) Google Scholar). Our results suggest that PKA signaling may provide one mechanism for amplifying chemotactic signals. During chemotaxis, new β1 integrin contacts are formed at the leading edge of the cell in response to a chemoattractant gradient. It can be hypothesized that these new integrin contacts stimulate PKA and, as a consequence, amplify Rac activation. This scenario implies that activated Rac is localized at the leading edge of migrating cells. In fact, activated Rac has been localized at the leading edge of migrating cells using fluorescence resonance energy transfer (FRET) technology (37Kraynov V.L. Chamberlain C. Bokock G.M. Schwartz M.A. Slabaugh S. Hahn K.M. Science. 2000; 290: 333-337Crossref PubMed Scopus (560) Google Scholar). Furthermore, Rac1 has been implicated in orientingDictyostelium (38Chung C.Y. Lee S. Briscoe C. Ellsworth C. Firtel R.A. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5225-5230Crossref PubMed Scopus (113) Google Scholar) and lymphocytes (39del Pozo M.A. Vicente-Manzanares M. Tejedor R. Serrador J.M. Sanchez-Madrid F. Eur. J. Immunol. 1999; 29: 3609-3620Crossref PubMed Scopus (194) Google Scholar) during chemotaxis. In this context, it is worth noting that the activation of β1 integrins leads to the recruitment and colocalization of a PKA regulatory subunit with integrin complexes and the subsequent activation of PKA (24Whittard J.D. Akiyama S.K. J. Cell Sci. 2001; 114: 3265-3272Crossref PubMed Google Scholar). In summary, our findings highlight the importance of cAMP and PKA in the regulation of small GTPase function and migration and identify important venues for future work. We thank Steve Akiyama, Chris Carpenter, and Rick Cerione for reagents. We also thank Don Senger, Judy Glaven, Isaac Rabinovitz, Richard Bates, and Tanelli Tani for helpful discussions and Bao-Kim Nguyen for technical assistance.
Referência(s)