Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Regulation of Casein Kinase Iϵ Activity by Wnt Signaling

2004; Elsevier BV; Volume: 279; Issue: 13 Linguagem: Inglês

10.1074/jbc.m304682200

ISSN

1083-351X

Autores

Wojciech Swiatek, I-Chun Tsai, Laura Klimowski, Andrea Pepler, Janet E. Barnette, H. Joseph Yost, David M. Virshup,

Tópico(s)

Digestive system and related health

Resumo

The Wnt/β-catenin signaling pathway is important in both development and cancer. Casein kinase Iϵ (CKIϵ) is a positive regulator of the canonical Wnt pathway. CKIϵ itself can be regulated in vitro by inhibitory autophosphorylation, and recent data suggest that in vivo kinase activity can be regulated by extracellular stimuli. We show here that the phosphorylation state and kinase activity of CKIϵ are directly regulated by Wnt signaling. Coexpression of XWnt-8 or addition of soluble Wnt-3a ligand led to a significant and rapid increase in the activity of endogenous CKIϵ. The increase in CKIϵ activity is the result of decreased inhibitory autophosphorylation because it is abolished by preincubation of immunoprecipitated kinase with ATP. Furthermore, mutation of CKIϵ inhibitory autophosphorylation sites creates a kinase termed CKIϵ(MM2) that is significantly more active than CKIϵ and is not activated further upon Wnt stimulation. Autoinhibition of CKIϵ is biologically relevant because CKIϵ(MM2) is more effective than CKIϵ at activating transcription from a Lef1-dependent promoter. Finally, CKIϵ(MM2) expression in Xenopus embryos induces both axis duplication and additional developmental abnormalities. The data suggest that Wnt signaling activates CKIϵ by causing transient dephosphorylation of critical inhibitory sites present in the carboxyl-terminal domain of the kinase. Activation of the Wnt pathway may therefore stimulate a cellular phosphatase to dephosphorylate and activate CKIϵ The Wnt/β-catenin signaling pathway is important in both development and cancer. Casein kinase Iϵ (CKIϵ) is a positive regulator of the canonical Wnt pathway. CKIϵ itself can be regulated in vitro by inhibitory autophosphorylation, and recent data suggest that in vivo kinase activity can be regulated by extracellular stimuli. We show here that the phosphorylation state and kinase activity of CKIϵ are directly regulated by Wnt signaling. Coexpression of XWnt-8 or addition of soluble Wnt-3a ligand led to a significant and rapid increase in the activity of endogenous CKIϵ. The increase in CKIϵ activity is the result of decreased inhibitory autophosphorylation because it is abolished by preincubation of immunoprecipitated kinase with ATP. Furthermore, mutation of CKIϵ inhibitory autophosphorylation sites creates a kinase termed CKIϵ(MM2) that is significantly more active than CKIϵ and is not activated further upon Wnt stimulation. Autoinhibition of CKIϵ is biologically relevant because CKIϵ(MM2) is more effective than CKIϵ at activating transcription from a Lef1-dependent promoter. Finally, CKIϵ(MM2) expression in Xenopus embryos induces both axis duplication and additional developmental abnormalities. The data suggest that Wnt signaling activates CKIϵ by causing transient dephosphorylation of critical inhibitory sites present in the carboxyl-terminal domain of the kinase. Activation of the Wnt pathway may therefore stimulate a cellular phosphatase to dephosphorylate and activate CKIϵ The Wnt signaling pathway is critical for many aspects of development and proliferation including dorsal axis formation, tissue pattering, and establishment of cell polarity (1Polakis P. Genes Dev. 2000; 14: 1837-1851Crossref PubMed Google Scholar, 2Moon R.T. Bowerman B. Boutros M. Perrimon N. Science. 2002; 296: 1644-1646Crossref PubMed Scopus (887) Google Scholar). The canonical Wnt signaling pathway regulates growth and development in part by controlling of the formation of a heterodimer containing β-catenin and a member of the Lef/Tcf family of transcriptional regulators. Mutations in the pathway cause inappropriate up-regulation of β-catenin signaling and underlie numerous human cancers (1Polakis P. Genes Dev. 2000; 14: 1837-1851Crossref PubMed Google Scholar, 3Morin P.J. Sparks A.B. Korinek V. Barker N. Clevers H. Vogelstein B. Kinzler K.W. Science. 1997; 275: 1787-1790Crossref PubMed Scopus (3517) Google Scholar, 4Rubinfeld B. Robbins P. El-Gamil M. Albert I. Porfiri E. Polakis P. Science. 1997; 275: 1790-1792Crossref PubMed Scopus (1137) Google Scholar, 5Groden J. Thliveris A. Samowitz W. Carlson M. Gelbert L. Albertsen H. Joslyn G. Stevens J. Spirio L. Robertson M. White R. Leppert M. Cell. 1991; 66: 589-600Abstract Full Text PDF PubMed Scopus (2412) Google Scholar). This signaling pathway has therefore been the focus of intense study. Three isoforms of casein kinase I (CKI) 1The abbreviations used are: CKI, casein kinase I; APC, adenomatous polyposis coli; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; Dvl, dishevelled; GSK, glycogen synthase kinase; HA, hemagglutinin; HEK, human embryonic kidney; mAb, monoclonal antibody; MBP, maltose-binding protein; PP, protein phosphatase; WT, wild type. have been found to be essential regulators of β-catenin abundance and Wnt signaling (6Polakis P. Curr. Biol. 2002; 12: R499-R501Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar). Casein kinase I α (CSNK1A, CKIα) is a constitutively active negative regulator of the pathway, interacting with axin and directly phosphorylating β-catenin on serine 45, priming it for regulated phosphorylation by GSK3 (7Amit S. Hatzubai A. Birman Y. Andersen J.S. Ben-Shushan E. Mann M. Ben-Neriah Y. Alkalay I. Genes Dev. 2002; 16: 1066-1076Crossref PubMed Scopus (595) Google Scholar, 8Liu C. Li Y. Semenov M. Han C. Baeg G.H. Tan Y. Zhang Z. Lin X. He X. Cell. 2002; 108: 837-847Abstract Full Text Full Text PDF PubMed Scopus (1687) Google Scholar, 9Yanagawa S. Matsuda Y. Lee J.S. Matsubayashi H. Sese S. Kadowaki T. Ishimoto A. EMBO J. 2002; 21: 1733-1742Crossref PubMed Scopus (158) Google Scholar). Knockdown of CKIα by RNA interference blocks phosphorylation of Ser-45 and hence stabilizes β-catenin. CKIϵ (CSNK1E) and the closely related CKIδ (CSNK1D) have been identified as positive regulators of the canonical Wnt-β-catenin pathway. Overexpression of CKIϵ in Xenopus embryos promotes ectopic dorsal axis formation and expression of Wnt-responsive genes, whereas CKI inhibitors and knockdown of CKIϵ by RNA interference blunt the accumulation of β-catenin in response to Wnt signaling (10Peters J.M. McKay R.M. McKay J.P. Graff J.M. Nature. 1999; 401: 345-350Crossref PubMed Scopus (385) Google Scholar, 11Sakanaka C. Leong P. Xu L. Harrison S.D. Williams L.T. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 12548-12552Crossref PubMed Scopus (189) Google Scholar, 12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar, 13Hino S.-i. Michiue T. Asashima M. Kikuchi A. J. Biol. Chem. 2003; 278: 14066-14073Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar). Diverse substrates for CKIϵ in the pathway have been identified, including dishevelled (Dvl), axin, the adenomatous polyposis coli protein APC, and Tcf-3. CKIϵ has been shown to enhance recruitment of GBP/Frat-1 to Dvl, to dissociate the β-catenin destruction complex, and to enhance the interaction between β-catenin and Tcf-3 (12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar, 14Lee E. Salic A. Kirschner M.W. J. Cell Biol. 2001; 154: 983-993Crossref PubMed Scopus (128) Google Scholar, 15Zhang Y. Qiu W.J. Chan S.C. Han J. He X. Lin S.C. J. Biol. Chem. 2002; 277: 17706-17712Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar, 16Kishida M. Hino Si S. Michiue T. Yamamoto H. Kishida S. Fukui A. Asashima M. Kikuchi A. J. Biol. Chem. 2001; 276: 33147-33155Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 17McKay R.M. Peters J.M. Graff J.M. Dev. Biol. 2001; 235: 378-387Crossref PubMed Scopus (39) Google Scholar, 18McKay R.M. Peters J.M. Graff J.M. Dev. Biol. 2001; 235: 388-396Crossref PubMed Scopus (98) Google Scholar, 19Rubinfeld B. Tice D.A. Polakis P. J. Biol. Chem. 2001; 276: 39037-39045Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar). CKIϵ may participate in additional developmental pathways including regulation of the Dvl-JNK signaling pathway and the proteolysis of Cubitus interruptus protein in the Hedgehog signaling pathway (15Zhang Y. Qiu W.J. Chan S.C. Han J. He X. Lin S.C. J. Biol. Chem. 2002; 277: 17706-17712Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar, 20Schwarz-Romond T. Asbrand C. Bakkers J. Kuhl M. Schaeffer H.J. Huelsken J. Behrens J. Hammerschmidt M. Birchmeier W. Genes Dev. 2002; 16: 2073-2084Crossref PubMed Scopus (162) Google Scholar, 21Price M.A. Kalderon D. Cell. 2002; 108: 823-835Abstract Full Text Full Text PDF PubMed Scopus (326) Google Scholar). Understandably, mutations in the Drosophila CKIϵ homolog double-time/discs overgrown therefore result in diverse developmental (as well as circadian rhythm) abnormalities (15Zhang Y. Qiu W.J. Chan S.C. Han J. He X. Lin S.C. J. Biol. Chem. 2002; 277: 17706-17712Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar, 17McKay R.M. Peters J.M. Graff J.M. Dev. Biol. 2001; 235: 378-387Crossref PubMed Scopus (39) Google Scholar, 21Price M.A. Kalderon D. Cell. 2002; 108: 823-835Abstract Full Text Full Text PDF PubMed Scopus (326) Google Scholar, 22Zhang Y. Qiu W.J. Liu D.X. Neo S.Y. He X. Lin S.C. J. Biol. Chem. 2001; 276: 32152-32159Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 23Zilian O. Frei E. Burke R. Brentrup D. Gutjahr T. Bryant P.J. Noll M. Development. 1999; 126: 5409-5420PubMed Google Scholar, 24Vielhaber E.L. Virshup D.M. IUBMB Life. 2001; 51: 273-278Crossref Scopus (66) Google Scholar). Notably, whereas CKIϵ interacts with Dvl in the absence of Wnt signaling, a Dvl electrophoretic mobility shift apparently resulting from direct CKIϵ phosphorylation only occurs during Wnt signaling (10Peters J.M. McKay R.M. McKay J.P. Graff J.M. Nature. 1999; 401: 345-350Crossref PubMed Scopus (385) Google Scholar, 12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar, 18McKay R.M. Peters J.M. Graff J.M. Dev. Biol. 2001; 235: 388-396Crossref PubMed Scopus (98) Google Scholar). The positive role of CKIϵ and the closely related CKIδ suggests that their kinase activity might in turn be regulated by Wnt signaling. CKIϵ and CKIδ differ from the 35-kDa CKIα isoform in that they are negatively regulated by an additional 13-kDa carboxyl-terminal autoregulatory domain. Intramolecular autophosphorylation of this domain strongly inhibits CKIϵ and CKIδ kinase activity. Kinase autoinhibition can be relieved in vitro by dephosphorylation of the autoinhibitory domain by diverse serine/threonine phosphatases (25Cegielska A. Gietzen K.F. Rivers A. Virshup D.M. J. Biol. Chem. 1998; 273: 1357-1364Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar, 26Rivers A. Gietzen K.F. Vielhaber E. Virshup D.M. J. Biol. Chem. 1998; 273: 15980-15984Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar, 27Graves P.R. Roach P.J. J. Biol. Chem. 1995; 270: 21689-21694Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). Recently, Liu et al. (28Liu F. Virshup D.M. Nairn A.C. Greengard P. J. Biol. Chem. 2002; 277: 45393-45399Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar) found that CKIϵ in neostriatal neurons could be activated by stimulation of metabotropic glutamate receptors. Receptor activation led to a rise in intracellular calcium, resulting in activation of the calcium-regulated phosphatase calcineurin, with subsequent dephosphorylation and activation of CKIϵ. The observation that CKIϵ activity could be regulated both in vitro and in vivo led us to test whether Wnt signaling controls CKIϵ activity as well. Our studies demonstrate that Wnt signaling directly regulates CKIϵ by inducing dephosphorylation of its autoregulatory domain. Wnt ligand rapidly activated both endogenous and overexpressed CKIϵ, an effect that was mimicked by phosphatase treatment of immunoprecipitated CKIϵ and reversed by preincubation of immunoprecipitated CKIϵ with ATP. Furthermore, Wnt signaling did not further activate a CKIϵ mutant (termed CKIϵ(MM2)) (29Gietzen K.F. Virshup D.M. J. Biol. Chem. 1999; 274: 32063-32070Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar) with serine and threonine to alanine point mutations at the inhibitory autophosphorylation sites. Consistent with the hypothesis that dephosphorylation of CKIϵ is a required step in the Wnt signaling pathway, the CKIϵ(MM2) mutant was more active in both transactivation of a Lef-1 reporter and in inducing axis duplication and additional developmental abnormalities in Xenopus embryogenesis. Interestingly, multiple distinct phosphatases appear to affect the activation of CKIϵ. Together, these data suggest that CKIϵ activation is an important consequence of canonical Wnt signaling. The data further imply that an unidentified protein phosphatase upstream of CKIϵ is activated by Wnt signaling to dephosphorylate and activate CKIϵ. Antibodies, Plasmids, Reagents—The expression plasmids pCS2-Myc-CKIϵ and pCS2-Myc-CKIϵ(MM2) with six amino-terminal Myc epitope tags were prepared by cloning into pCS2-MT as described previously (12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar, 29Gietzen K.F. Virshup D.M. J. Biol. Chem. 1999; 274: 32063-32070Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar). Expression plasmids pCS2+XWnt-8, TOPflash, pEV3S-Lef-1, pCMV-β-Gal, and pCS-PP2Cα were kindly provided by Randy Moon, Hans Clevers, Marian Waterman, Andrew Thorburn, and Daniel Sussman, respectively. The expression construct for MBP-tagged axin was a gift from Adrian Salic and has been described previously (30Salic A. Lee E. Mayer L. Kirschner M.W. Mol. Cell. 2000; 5: 523-532Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar). Wnt-3a-conditioned medium was produced as described previously (12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar, 31Shibamoto S. Higano K. Takada R. Ito F. Takeichi M. Takada S. Genes Cells. 1998; 3: 659-670Crossref PubMed Scopus (231) Google Scholar). Anti-HA (12CA5) was obtained from Roche Applied Science. The anti-CKIϵ affinity-purified rabbit polyclonal antibody UT31 has been described previously (32Fish K. Cegielska A. Getman M. Landes G. Virshup D.M. J. Biol. Chem. 1995; 270: 14875-14883Abstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar). Monoclonal anti-CKIϵ antibody was from Transduction Laboratories. Protease inhibitor mixture tablets were obtained from Roche Applied Science. Lambda protein phosphatase was obtained from Upstate Biotechnology. MBP-axin fusion protein was expressed in Escherichia coli and purified using amylose resin (New England Biolabs) according to the published protocol with minor modification (30Salic A. Lee E. Mayer L. Kirschner M.W. Mol. Cell. 2000; 5: 523-532Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar). Cell Culture, Maintenance, and Transient Transfection—Human embryonic kidney (HEK) 293 cells and mouse L cells were grown in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum and 10% calf serum, respectively, and antibiotics (100 IU/ml penicillin and 100 μg/ml streptomycin). Cultures of cells were maintained in a humidified incubator at 37 °C and 5% CO2. For transient transfections, cells were seeded in 6-well dishes treated with poly-l-lysine (Sigma). Cells were transiently transfected when they reached 60-70% confluence using LipofectAMINE and PLUS reagent (Invitrogen) according to the manufacturer's instructions. For immunoprecipitation experiments 0.5 μg of the CKIϵ expression plasmids was transfected. The total amount of transfected DNA was adjusted to 1 μg with empty vector where necessary. Cells were lysed at 24 h post-transfection as described below. Cell Lysis and Immunoprecipitation—Prior to lysis, plated cells were washed once with phosphate-buffered saline. 200 μl of lysis buffer (50 mm Hepes, pH 8.0, 1% Nonidet P-40, 0.1% SDS, 150 mm NaCl, 1 mm EDTA, 1 mm dithiothreitol) supplemented with 1× Complete protease inhibitor mixture (Roche Applied Science) and phosphatase inhibitors (1 mm sodium fluoride, 1 mm β-glycerophosphate, 1 mm sodium orthovanadate) was then added to cells. The cells were then mechanically sheared, and the lysates were centrifuged at 16,000 × g for 10 min at 4 °C. Myc epitope-tagged proteins, CKIϵ(WT) and CKIϵ(MM2), were quantitatively immunoprecipitated from the lysates (0.5-1 mg of protein in total) with 4 μg of anti-Myc antibody (9E10). Rabbit anti-CKIϵ antibody UT31 was used where indicated (25Cegielska A. Gietzen K.F. Rivers A. Virshup D.M. J. Biol. Chem. 1998; 273: 1357-1364Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). The protein lysates and antibodies were incubated in lysis buffer (0.5 ml final volume) for 1 h on ice to allow formation of antibody-antigen complexes. Immune complexes were then mixed with 25 μl of protein A-agarose beads (Invitrogen) for 1 h at 4 °C with gentle agitation. The isolated beads were then collected by brief centrifugation at 185 × g in a tabletop centrifuge and subsequently washed once each with lysis buffer, and lysis buffer lacking both SDS and dithiothreitol, and then twice in protein storage buffer/kinase buffer (10% sucrose, 50 mm Hepes, pH 7.2, 15 mm NaCl, 1 mm dithiothreitol, 0.5 mm EDTA). CKI activity was determined as described previously except that MBP-axin was used as the substrate (28Liu F. Virshup D.M. Nairn A.C. Greengard P. J. Biol. Chem. 2002; 277: 45393-45399Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). Briefly, kinase reactions containing immunoprecipitates and 18 pmol of MBP-axin were incubated for 15 min at 37 °C in 40 μl of kinase reaction mixture (50 mm Tris-HCl, pH 7.5, 10 mm MgCl2, 1 mm dithiothreitol, and 250 μm ATP) containing 5 μCi of [γ-32P]ATP. The reactions were terminated by adding 25 μl of hot 5× SDS-PAGE loading buffer. The boiled samples were separated by 8% SDS-PAGE, then transferred to nitrocellulose membrane (Amersham Biosciences), visualized by autoradiography, and subsequently subjected to immunodetection of CKI. Quantification of 32P-MBP-axin bands was performed using a Storm 860 gel imaging system (Molecular Dynamics). -Fold activation of the kinase was normalized to the CKIϵ expression levels estimated by Western blotting analysis. The background kinase activity present in the absence of specific antibody was subtracted. Data are expressed as -fold kinase activation compared with that in vector-transfected cells. In the indicated experiments, immunocomplexes of CKIϵ from HEK 293 cells were incubated either with 500 μm ATP in kinase buffer or 200 units of lambda phosphatase in 50 mm Tris-HCl, 0.1 mm Na2EDTA, 5 mm dithiothreitol, 2 mm MnCl2. Control reactions without lambda protein phosphatase or additional ATP were also performed. Dephosphorylation/phosphorylation reactions were incubated at 37 °C for 20 min. To stop the reactions, beads with CKIϵ were washed three times with protein storage buffer/kinase buffer. Kinase activity was measured as described above. Two-dimensional Gel Electrophoresis—HEK 293 cells were transfected as described above with expression plasmids for 4HA-CKIϵ without or with XWnt-8. Cells were lysed in 200 μl of hypotonic buffer (40 mm Tris, 4 μm okadaic acid), homogenized, and centrifuged at 14,000 rpm at 4 °C for 10 min. Hypotonic lysates (25 μg) were diluted into IEF buffer (8 m urea, 4% CHAPS, 40 mm Tris, 0.2% Bio-Lyte 3/10 ampholyte (Bio-Rad) and used in active rehydration (12 h) of 7-cm ReadyStrip IPG Strips, pH 3-10. Samples were focused using a Bio-Rad Protean IEF Cell under the following conditions at 20 °C: rapid 250 V, 1 h; rapid 500 V, 1 h; slow 8,000 V, 1 h; rapid 20,000 V, 1 hr. Second dimension SDS-PAGE was performed in 4-20% gradient gels and subjected to immunoblot analysis with anti-HA antibodies. Lef1-luciferase Reporter Gene Assay—200 ng of pCS2-CKIϵ(WT), pCS2-CKIϵ(MM2), or pCS2 empty vector was transfected into HEK 293 cells in 35-mm dishes along with 500 ng of TOPflash, 100 ng of pEV3S-Lef-1, and 25 ng of pCMV-β-Gal. At 46 h post-transfection, cells were harvested with lysis buffer (Dual-light kit, Tropix), and luciferase and β-galactosidase activity was measured with the Dual-Light kit and Microtiter Plate Luminometer (Dynex Technologies) according to the manufacturer's instructions. To standardize for transfection efficiency, the luciferase activities of all transfected cells were divided by the β-galactosidase activities. The -fold increase of Lef-1-dependent activity was defined by comparing the normalized level observed from cells transfected with pCS2. Data are presented as the mean ± S.E. from two separate experiments. Xenopus Injections and Analysis of Phenotypes—Sense mRNA of CKIϵ(WT) and CKIϵ(MM2) was prepared with the mMessage mMachine kit (Ambion) using linearized plasmid as templates. RNA was purified with RNAeasy (Qiagen). 0.5-1 ng of RNA was microinjected into the ventral side of four-cell stage embryos as described previously (12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar, 33Li X. Yost H.J. Virshup D.M. Seeling J.M. EMBO J. 2001; 20: 4122-4131Crossref PubMed Scopus (131) Google Scholar). The embryos were scored 1 and 3 days after injection. Previous studies indicated that CKIϵ kinase activity could be regulated both in vitro and in vivo by intramolecular autophosphorylation of its carboxyl-terminal inhibitory domain (25Cegielska A. Gietzen K.F. Rivers A. Virshup D.M. J. Biol. Chem. 1998; 273: 1357-1364Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar, 26Rivers A. Gietzen K.F. Vielhaber E. Virshup D.M. J. Biol. Chem. 1998; 273: 15980-15984Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar, 28Liu F. Virshup D.M. Nairn A.C. Greengard P. J. Biol. Chem. 2002; 277: 45393-45399Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). Given the role of CKIϵ as a positive regulator of the Wnt pathway, we tested the hypothesis that Wnt signaling functions in part by activation of CKIϵ. HEK 293 cells, which retain many aspects of the Wnt signal transduction cascade, were transfected with a plasmid expressing XWnt-8. Endogenous CKIϵ was immunoprecipitated from lysates 24 h later, and kinase activity was assessed in the immunoprecipitate. As Fig. 1A demonstrates, XWnt-8 expression led to almost a 6-fold increase in kinase activity compared with vector-only transfected cells. Activation of CKIϵ was independent of the substrate used because it was seen with casein and recombinant axin and Dvl-1 (data not shown). Notably, CKIϵ activity immunoprecipitated from cells not expressing XWnt-8 was only 10% above the background activity obtained in the absence of immunoprecipitated kinase, suggesting that basal CKIϵ activity is low. Furthermore, XWnt-8 expression had no effect on the amount of CKIϵ immunoprecipitated, demonstrating that the activation of CKIϵ was due solely to an increase in the specific activity of the kinase. To analyze further the influence of Wnt signaling on CKI activity, we assessed the response of CKIϵ to the treatment of the cells with Wnt-3a-conditioned medium. Wnt-3a-conditioned medium induces the accumulation of β-catenin in mouse L cells (Fig. 1C) as described previously (31Shibamoto S. Higano K. Takada R. Ito F. Takeichi M. Takada S. Genes Cells. 1998; 3: 659-670Crossref PubMed Scopus (231) Google Scholar). Addition of Wnt-3a-conditioned medium to HEK 293 cells for 30 min prior to lysis increased CKIϵ activity severalfold compared with cells treated with control medium (Fig. 1B). Hence, both transfected and exogenous Wnt ligands are capable of activation of CKIϵ kinase. CKIϵ can be strongly inhibited by intramolecular autophosphorylation. Wnt signaling may activate CKIϵ by stimulating intracellular phosphatases to remove inhibitory phosphoryl groups. This model predicts that inactive CKIϵ immunoprecipitated from resting cells should be activated by phosphatase treatment and more importantly, that activated CKIϵ immunoprecipitated from Wnt-treated cells should be subsequently inactivated by reautophosphorylation. Finally, Wnt-regulated dephosphorylation of CKIϵ should be reflected in an appropriate shift in the isoelectric point of CKIϵ. To test these predictions, endogenous CKIϵ was immunoprecipitated from HEK 293 cells treated with either control medium or Wnt-3a-conditioned medium. Prior to the in vitro kinase assay, the immunoprecipitated CKIϵ immune complexes were preincubated in buffer alone or in buffer containing added ATP or protein phosphatase λ for 20 min. As Fig. 2A confirms, Wnt-3a activated CKIϵ (lanes 1 and 2). Furthermore, dephosphorylation of immunoprecipitated kinase led to its activation (lanes 5 and 6) to a level similar to that seen with Wnt ligand, consistent with previous reports (25Cegielska A. Gietzen K.F. Rivers A. Virshup D.M. J. Biol. Chem. 1998; 273: 1357-1364Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar, 28Liu F. Virshup D.M. Nairn A.C. Greengard P. J. Biol. Chem. 2002; 277: 45393-45399Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). Most importantly, the activation of CKI by Wnt signaling was fully reversed when the immunoprecipitated activated kinase was allowed to reautophosphorylate during a preincubation with ATP (lanes 3 and 4). These data strongly suggest that Wnt signaling activates CKIϵ by stimulating dephosphorylation of its inhibitory domain. Isoelectric focusing was used to test directly whether the Wnt-induced increase in CKIϵ activity is accompanied by changes in the phosphorylation state of CKIϵ (Fig. 2B). A shift toward the basic end of the gel should be observed after removal of negatively charged phosphate groups. Wild type HA epitope-tagged CKIϵ migrates with an isoelectric point of ∼6, with a few less abundant species consistent with the existence of multiple phosphorylation states. Importantly, coexpression of XWnt-8 led to a marked shift of the CKIϵ to a pI of ∼7, consistent with the loss of one or more phosphoryl groups. Notably, most of the 4HA-CKIϵ was altered by the coexpression of XWnt-8, suggesting that the majority of the expressed kinase was dephosphorylated after Wnt signaling. If Wnt signaling activates CKIϵ by stimulating site-specific dephosphorylation, then a mutant form of CKIϵ, CKIϵ(MM2), which lacks inhibitory autophosphorylation sites, ought to be more active than CKIϵ in resting cells and not be further stimulated by Wnt signaling. To test this, HEK 293 cells were transfected with plasmids expressing Myc epitope-tagged CKIϵ or CKIϵ(MM2), and XWnt-8 or empty vector. CKI activity was measured after immunoprecipitation of the kinase with 9E10 monoclonal antibody specific for the Myc epitope. As Fig. 3A shows, overexpressed CKIϵ is activated by XWnt-8 expression (Fig. 3A, lane 1 versus lane 2), indicating that CKIϵ is a specific target of the Wnt pathway. CKIϵ(MM2) has significantly higher basal activity than wild type CKIϵ and is not activated further by coexpression of XWnt-8. Similar results were obtained when cells expressing Myc-CKIϵ or Myc-CKIϵ(MM2) were treated with Wnt-3a supernatant (Fig. 3B). To investigate the kinetics of CKIϵ activation, HEK 293 cells were treated with Wnt-3a-conditioned medium or control medium for different periods of time prior to lysis and immunoprecipitation-kinase assay. As shown in Fig. 4A, Wnt ligand activates endogenous CKIϵ after as little as 15 min with no further activation after 30 min. This observation suggests that CKIϵ activation is a direct downstream consequence of Wnt signaling and is not caused by long term activation or changes in downstream gene expression. To assess whether the activity of CKIϵ correlates with the presence of Wnt ligand, we also determined the rate of kinase inactivation following washout of Wnt-3a ligand. HEK 293 cells expressing CKIϵ(WT) and CKIϵ(MM2) were treated with Wnt-3a-conditioned medium for 30 min. The conditioned medium was removed and replaced either with Dulbecco's modified Eagle's medium, or as a control, the conditioned medium. Kinase activity was then assessed at the indicated times. As demonstrated in Fig. 4B, washout of Wnt ligand was followed by a steady decline in CKIϵ activity, reaching base line after 90 min. CKIϵ(MM2) activity remained high throughout the entire washout period, consistent with the hypothesis that CKIϵ is inactivated by autophosphorylation as Wnt signaling decreases. The data suggest that Wnt signaling activates CKIϵ by transient dephosphorylation of carboxyl-terminal inhibitory sites on the kinase. CKIϵ(MM2), the CKIϵ mutant lacking inhibitory autophosphorylation sites, has a marked increase in kinase activity in the absence of Wnt ligand. If activation of CKIϵ is an important step in Wnt signaling, then we speculated that CKIϵ(MM2) might be more active than wild type CKIϵ in signaling to downstream molecules in the pathway. Activation of the Wnt signaling pathway leads to transcription from Lef-1/Tcf-responsive promoters. Therefore, to test whether loss of autophosphorylation of CKIϵ increase its ability to transduce Wnt signaling, we compared the ability of wild type and CKIϵ(MM2) expression to activate a Lef-1 reporter plasmid. As Fig. 5A shows, expression of CKIϵ and CKIϵ(MM2) caused a 3- and 6-fold increase in the Lef-1 reporter activity, respectively, indicating that CKIϵ(MM2) is more active than wild type kinase. Immunoblots showed the that wild type and mutant kinase were expressed to the same level (data not shown). This result suggests that activation of CKIϵ is an important step in the Wnt pathway. A second readily assayable consequence of Wnt signaling is induction of secondary dorsal axis during Xenopus embryogenesis. CKIϵ expression in ventral cells causes axis duplication and has been epistatically positioned to be downstream of Dvl and upstream of axin and PP2A (10Peters J.M. McKay R.M. McKay J.P. Graff J.M. Nature. 1999; 401: 345-350Crossref PubMed Scopus (385) Google Scholar, 11Sakanaka C. Leong P. Xu L. Harrison S.D. Williams L.T. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 12548-12552Crossref PubMed Scopus (189) Google Scholar, 12Gao Z.-H. Seeling J.M. Hill V. Yochum A. Virshup D.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1182-1187Crossref PubMed Scopus (192) Google Scholar). We hypothesized that the loss of autoinhibition of CKIϵ would lead to constitutive rather than time-limited activation of Wnt and potentially additional CKIϵ-regulated signaling pathways. To test this, we microinjected the RNA encoding wild type or CKIϵ(MM2) into ventral cells of four-cell stage Xenopus embryos. As expected, ventral injection of wild type CKIϵ RNA induced the formation of ectopic dorsal axes (Fig. 5C). However, as Fig. 5D shows, microinjection of CKIϵ(MM2) RNA causes a distinct phenotype. The embryos microinjected with CKIϵ(MM2) RNA formed an ectopic axis, but the secondary axis was not fully developed. In addition, these embryos commonly had a shorter and bent tail and more prominent anterior structures. This distinct phenotype was only rarely seen at the highest dose of wild type CKIϵ expression. For scoring purposes, this phenotype was termed "extreme" (Fig. 5, B-D). Importantly, immunoblot of Xenopus embryos showed equal expression of CKIϵ and CKIϵ(MM2) (data not shown), demonstrating that the differences in development were caused by changes in kinase activity rather than kinase abundance. Thus, expression of CKIϵ(MM2) leads to increased transcription from a Wnt-responsive promoter and a distinctive Wnt-related phenotype in Xenopus embryogenesis, consistent with the hypothesis that regulation of CKIϵ activity is a critical element of normal embryogenesis. The dephosphorylation-mediated activation of CKIϵ by Wnt ligand implies that an intracellular phosphatase is upstream of CKIϵ. In vitro, PP1, PP2A, and PP2B (calcineurin) are all capable of dephosphorylating and activating CKIϵ (25Cegielska A. Gietzen K.F. Rivers A. Virshup D.M. J. Biol. Chem. 1998; 273: 1357-1364Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). To test which of these phosphatases might mediate Wnt-induced activation of CKIϵ in vivo, HEK 293 cells transiently expressing Myc-CKIϵ or CKIϵ(MM2) were treated with Wnt-3a-conditioned medium followed by further addition for 30 min of the indicated phosphatase inhibitor prior to lysis and immunoprecipitation-kinase assay. Calyculin A inhibits both PP1 and PP2A, whereas cyclosporine A inhibits calcineurin (PP2B). Interestingly, addition of either inhibitor blocked Wnt reversed the activation of CKIϵ (Fig. 6). The effect of the inhibitors was specific because they did not inhibit CKIϵ(MM2). Overexpression of an unrelated serine/threonine phosphatase, PP2C, also stimulated CKIϵ activity. These data do not allow identification of the specific CKIϵ phosphatase but further support the model that Wnt signaling regulates CKIϵ activity via an intracellular protein serine/threonine phosphatase. CKIϵ and the closely related CKIδ are positive regulators of the Wnt signaling pathway, functioning downstream of Wnt and upstream of the APC·axin·GSK3·PP2A complex. The data presented here show that Wnt signaling results in both dephosphorylation of CKIϵ and a significant increase in CKIϵ kinase activity. CKIϵ activation is detectable within 15 min after the application of Wnt ligand, indicating that activation is a direct effect rather than the result of downstream changes in gene expression. CKIϵ activation is caused by dephosphorylation of its inhibitory carboxyl-terminal domain because 1) the isoelectric point of the kinase shifted toward the basic region after Wnt treatment; 2) Wnt-induced activation was reversed by reautophosphorylation and mimicked by in vitro dephosphorylation of CKIϵ; 3) a mutant CKIϵ that lacks inhibitory autophosphorylation sites was active and not further activated by Wnt, and 4) inhibition of cellular serine/threonine phosphatases blocked the activation of CKIϵ. Finally, activation of CKIϵ is biologically relevant because constitutively active CKIϵ was a more active regulator of transcription from a Wnt-responsive promoter and led to marked phenotypic differences in Xenopus development compared with wild type CKIϵ. Thus, although the CKI family has generally been considered to contain a group of constitutively active kinases, it appears that a subset of the CKI family, those with autoinhibitory domains, may be inactive in vivo until specifically activated. The data suggest that regulated activation of CKIϵ may be one mechanism to terminate a Wnt signal after ligand dissociates from its receptors. CKIϵ, unlike CKIα, appears to be inactive in vivo until specifically activated by dephosphorylation of its autoinhibitory domain. Following removal of Wnt ligand, kinase once again autophosphorylates and autoinhibits, and hence its activity returns to base line. The relatively slow decrease in kinase activity could be the result of slow dissociation of Wnt ligand from the cell surface or prolonged activation of the phosphatase that activates CKI. Signal-regulated time-limited activation of CKIϵ may be a general mechanism that regulates phosphorylation of additional substrates such as p53, the circadian regulators Per1 and Per2, and centrosome proteins (34Knippschild U. Milne D.M. Campbell L.E. DeMaggio A.J. Christenson E. Hoekstra M.F. Meek D.W. Oncogene. 1997; 15: 1727-1736Crossref PubMed Scopus (143) Google Scholar, 35Sillibourne J.E. Milne D.M. Takahashi M. Ono Y. Meek D.W. J. Mol. Biol. 2002; 322: 785-797Crossref PubMed Scopus (91) Google Scholar, 36Vielhaber E. Eide E. Rivers A. Gao Z.-H. Virshup D.M. Mol. Cell. Biol. 2000; 20: 4888-4899Crossref PubMed Scopus (246) Google Scholar). The ability of CKI isoforms to interact specifically with a diverse array of scaffold-like proteins such as axin, Dvl, Per2, and AKAP450 may provide a mechanism to restrict the activated kinase to phosphorylate only a subset of substrates, i.e. those also interacting with the scaffold protein (37Pawson T. Scott J.D. Science. 1997; 278: 2075-2080Crossref PubMed Scopus (1904) Google Scholar). A similar function for axin-bound GSK3 in the Wnt pathway is well described (38Ding V.W. Chen R.H. McCormick F. J. Biol. Chem. 2000; 275: 32475-32481Abstract Full Text Full Text PDF PubMed Scopus (378) Google Scholar). For Wnt signaling to activate CKIϵ, it must somehow regulate a cellular serine/threonine phosphatase upstream of CKIϵ. In vitro and in vivo, diverse protein phosphatases have been shown to regulate CKIϵ autophosphorylation. For example, we have shown previously that CKIϵ can be activated in vitro by incubation with PP1, PP2A, and calcineurin, whereas in vivo, the glutaminergic pathway appears to turn on CKIϵ by activating calcineurin (25Cegielska A. Gietzen K.F. Rivers A. Virshup D.M. J. Biol. Chem. 1998; 273: 1357-1364Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar, 28Liu F. Virshup D.M. Nairn A.C. Greengard P. J. Biol. Chem. 2002; 277: 45393-45399Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). There is no evidence that calcineurin activation is involved in the canonical Wnt pathway, although Wnt-regulated calcium signaling has been described with Wnt-5 and other noncanonical Wnts (39Kuhl M. Sheldahl L.C. Park M. Miller J.R. Moon R.T. Trends Genet. 2000; 16: 279-283Abstract Full Text Full Text PDF PubMed Scopus (748) Google Scholar). However, a number of other protein phosphatases have been shown to influence the Wnt-β-catenin pathway. PP2A has been shown to play both a positive and a negative role (reminiscent of the opposing roles of CKI isoforms), and PP2C interacts with Dvl and plays a positive role (33Li X. Yost H.J. Virshup D.M. Seeling J.M. EMBO J. 2001; 20: 4122-4131Crossref PubMed Scopus (131) Google Scholar, 40Seeling J.M. Miller J.R. Gil R. Moon R.T. White R. Virshup D.M. Science. 1999; 283: 2089-2091Crossref PubMed Scopus (367) Google Scholar, 41Ikeda S. Kishida M. Matsuura Y. Usui H. Kikuchi A. Oncogene. 2000; 19: 537-545Crossref PubMed Scopus (165) Google Scholar, 42Ratcliffe M.J. Itoh K. Sokol S.Y. J. Biol. Chem. 2000; 275: 35680-35683Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar, 43Strovel E.T. Wu D. Sussman D.J. J. Biol. Chem. 2000; 275: 2399-2403Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar). Inhibition of CKIϵ activation by both calyculin A and cyclosporine suggests that intracellular CKIϵ may be regulated by diverse phosphatases in different pathways. When any one of the phosphatases is inhibited by exogenous compounds, the balance of autophosphorylation and dephosphorylation may be perturbed sufficiently to prevent kinase dephosphorylation and activation by physiologic stimuli. The extreme phenotype of the Xenopus embryos expressing constitutively active CKIϵ (CKIϵ(MM2)) raises the question of whether activated CKIϵ functions solely in the canonical Wnt pathway to stabilize β-catenin or whether sustained activation also causes changes in a number of other pathways. Forced expression of wild type CKIϵ may not have extreme consequences because the signals that activate it are lacking or tightly regulated. However, overexpressed constitutively active CKIϵ has the ability to alter additional diverse pathways outside of the Wnt pathway in the absence of external stimuli. Finally, the results from these studies indicate that CKIϵ is an additional example of a group of protein kinases including Cdks, GSK3, Src family members, and Raf-1, which are inhibited by phosphorylation and require dephosphorylation for signaling to occur. However, CKIϵ is inhibited by intramolecular autophosphorylation, whereas the other kinases require upstream kinases to regulate their activity. Taken together, the data support the conclusion that CKIϵ is a Wnt-regulated kinase that phosphorylates multiple substrates, including Dvl, APC, and Tcf-3, thereby stimulating β-catenin signaling by multiple mechanisms including, but not limited to, inhibition of GSK3 activity, dissociation of the β-catenin destruction complex, and stabilization of the β-catenin/Tcf-3 transcription factor. We thank members of the Virshup and Yost laboratories, Andrew Thorburn, Rich Dorsky, Don Ayer, and David Jones for productive discussions; Jeff Brown, Hans Clevers, Marian Waterman, and Andrew Thorburn for plasmids; and S. Takada for Wnt-3a-expressing cells.

Referência(s)