Glycogen Synthase Kinase 3 Is a Natural Activator of Mitogen ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 278, No. 16, Issue of April 18, pp. 13995–14001, 2003 Printed in U.S.A.

Glycogen Synthase Kinase 3␤ Is a Natural Activator of Mitogen-activated Protein Kinase/Extracellular Signal-regulated Kinase Kinase Kinase 1 (MEKK1)* Received for publication, January 9, 2003, and in revised form, February 10, 2003 Published, JBC Papers in Press, February 12, 2003, DOI 10.1074/jbc.M300253200

Jin Woo Kim‡, Ji Eun Lee‡, Myung Jin Kim‡, Eun-Gyung Cho‡§, Ssang-Goo Cho‡¶, and Eui-Ju Choi‡储 From the ‡National Creative Research Initiative Center for Cell Death, Graduate School of Life Science and Biotechnology, Korea University, Seoul 136-701, Korea and ¶Department of Animal Science, Konkuk University, Seoul 143-702, Korea

Glycogen synthase kinase 3␤ (GSK3␤) is implicated in many biological events, including embryonic development, cell differentiation, apoptosis, and insulin response. GSK3␤ has now been shown to induce activation of the mitogen-activated protein kinase kinase kinase MEKK1 and thereby to promote signaling by the stressactivated protein kinase pathway. GSK3 ␤ -binding protein blocked the activation of MEKK1 by GSK3␤ in human embryonic kidney 293 (HEK293) cells. Furthermore, co-immunoprecipitation analysis revealed a physical association between endogenous GSK3 ␤ and MEKK1 in HEK293 cells. Overexpression of axin1, a GSK3␤-regulated scaffolding protein, did not affect the physical interaction between GSK3␤ and MEKK1 in transfected HEK293 cells. Exposure of cells to insulin inhibited the activation of MEKK1 by GSK3␤, and this inhibitory effect of insulin was abolished by the phosphatidylinositol 3-kinase inhibitor wortmannin. Furthermore, MEKK1 activity under either basal or UV- or tumor necrosis factor ␣-stimulated conditions was reduced in embryonic fibroblasts derived from GSK3␤ knockout mice compared with that in such cells from wild-type mice. Ectopic expression of GSK3␤ increased both basal and tumor necrosis factor ␣-stimulated activities of MEKK1 in GSK3␤ⴚ/ⴚ cells. Together, these observations suggest that GSK3␤ functions as a natural activator of MEKK1. Glycogen synthase kinase 3␤ (GSK3␤)1 is a serine/threonine kinase that is thought to contribute to a variety of biological events such as embryonic development, metabolism, tumorigenesis, and cell death (1– 4). GSK3␤ is a constitutively active * This work was supported by the Creative Research Initiatives Program of the Korean Ministry of Science and Technology and in part by a Korea University Grant (to E.-J. C.), and in part by the faculty research fund of Konkuk University (to S.-G. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § Supported by a postdoctoral fellowship from the BK21 program of the Korean Ministry of Education. 储 To whom correspondence should be addressed. Tel.: 82-2-3290-3446; Fax: 82-2-3290-4741; E-mail: ejchoi@korea.ac.kr. 1 The abbreviations used are: GSK3␤, glycogen synthase kinase 3␤; GBP, GSK3␤-binding protein; PI3K, phosphatidylinositol 3-kinase; MAPK, mitogen-activated protein kinase; JNK, c-Jun N-terminal kinase; SAPK, stress-activated protein kinase; ASK1, apoptosis signalregulating kinase 1; HA, hemagglutinin; MEF, mouse embryonic fibroblast; GFP, green fluorescence protein; TNF-␣, tumor necrosis factor ␣; HEK293, human embryonic kidney 293; GST, glutathione Stransferase. This paper is available on line at http://www.jbc.org

kinase and regulates many intracellular signaling pathways by phosphorylating substrates such as ␤-catenin. The phosphorylation of ␤-catenin by GSK3␤ is facilitated by the scaffold protein axin and is inhibited either by GSK3␤-binding protein (GBP), also known as Frat (Frequently rearranged in advanced T-cell lymphomas), or by Dishevelled (5–7). The GSK3␤-catalyzed phosphorylation of ␤-catenin results in its ubiquitinmediated proteolysis (8, 9). The function of GSK3␤ is also regulated through phosphorylation by other protein kinases including Akt (10 –13). Akt is a serine/threonine kinase that is activated by phosphatidylinositol 3-kinase (PI3K) signaling and phosphorylates GSK3␤ on Ser9, thereby inactivating it (10, 14, 15). Mitogen-activated protein kinase (MAPK) signaling cascades are evolutionarily conserved from yeast to humans and are involved in a variety of cellular functions, including cell growth, differentiation, and cell death (16 –19). Each MAPK signaling pathway includes three components, a MAPK, a MAPK kinase, and a MAPK kinase kinase. An activated MAPK kinase kinase phosphorylates serine or threonine residues of a MAPK kinase, which in turn phosphorylates and thereby activates a MAPK (16 –18). The mammalian family of MAPKs includes at least three subgroups, extracellular signal-regulated kinases, c-Jun NH2-terminal kinases (JNKs), also known as stress-activated protein kinases (SAPKs), and p38 MAPK (17, 18). The extracellular signal-regulated kinase pathway is activated by various mitogens (19), whereas the JNK/SAPK and p38 pathways are activated by proinflammatory cytokines such as tumor necrosis factor-␣ (TNF-␣) as well as cellular stresses (16 –18). The JNK/SAPK pathway, which is associated with stress-induced cellular events including cell death (20 – 27), is composed of JNK/SAPK, a MAPK kinase such as MKK4 (also known as SEK1 or JNKK1) and MKK7, and a MAPK kinase kinase such as MEKK1 (16 –18). To provide further insight into the regulatory role of GSK3␤ in intracellular signaling cascades, we have now investigated the effect of this enzyme on the JNK/SAPK pathway. We show that GSK3␤ physically associates with and activates MEKK1, thereby stimulating the JNK/SAPK pathway. The PI3K-Akt signaling inhibited the GSK3␤-induced activation of MEKK1. Furthermore, the activity of endogenous MEKK1 was reduced in mouse embryonic fibroblasts (MEFs) derived from GSK3␤ knockout (GSK3␤⫺/⫺) mice compared with that in MEFs from GSK3␤⫹/⫹ mice. These findings suggest that GSK3␤ functions as an endogenous activator of MEKK1. EXPERIMENTAL PROCEDURES

Antibodies—Mouse monoclonal antibodies to GSK3␤ and to axin as well as rabbit polyclonal antibodies against MEKK1 and apoptosis

13995

13996

Activation of MEKK1 by GSK3b

FIG. 1. GSK3␤ induces the stimulation of MEKK1, SEK1, and JNK. A, HEK293 cells were transfected for 48 h with expression vectors encoding HA-SAPK␤ (left panel) or HA-extracellular signal-regulated kinase (HA-ERK2) (right panel) either alone or together with vectors for FLAG-GSK3␤ or FLAG-GSK3␤(K85A) as indicated. The cells were then either unexposed or exposed to UV light (80 J/m2) (left panel) or 200 nM phorbol 12-myristate 13-acetate (right panel). After 30 min, cell lysates were subjected to immunoprecipitation with mouse monoclonal anti-HA antibody. The resulting immunoprecipitates were examined for SAPK␤ activity (left panel) or extracellular signal-regulated kinase 2 activity (right panel) by immunocomplex kinase assays with GST-c-Jun-(1–79) or myelin basic protein (MBP), respectively, as substrates. The fold increase in activity relative to that of control cells is indicated. Cell lysates were also subjected to immunoblot analysis (IB) with antibodies to HA or to FLAG. Graphs in the bottom show the means ⫾ average deviation of three independent experiments. B and C, HEK293 cells were transfected for 48 h with a vector encoding GST-SEK1 (B) or HA-MEKK1 (C) either alone or together with a vector for GSK3␤-FLAG or GSK3␤(K85A)-FLAG, as indicated. The cells were then untreated or treated with UV light (80 J/m2) and further incubated for 30 min. In B, GST-SEK1 expressed in the transfected cells was isolated from cell lysates with the use of glutathione-agarose beads and then assayed for kinase activity with GSTSAPK␤(K55R) as substrate. In C, cell lysates were subjected to immunoprecipitation with antibody to HA, and the resulting precipitates were assayed for MEKK1 activity with GST-SEK1(K129R) as substrate. Graphs in the bottom show the means ⫾ average deviation for SEK1 and MEKK1 activity, respectively, of three independent experiments. D, HEK293 cells were transfected for 48 h with an expression vector encoding GST-JNK1 either alone or together with the indicated combinations of vectors for GSK3␤-FLAG, HA-MEKK1, HA-MEKK1(K1253M), HA-ASK1, or HA-ASK1(K709R). The cells were then unexposed or exposed to UV light (80 J/m2) and incubated for an additional 30 min. GST-JNK1 was precipitated from cell lysates with glutathione-agarose beads and assayed for kinase activity with GST-c-Jun-(1–79) as substrate.


13997

FIG. 2. Effects of GBP and LiCl on the JNK/SAPK signaling pathway. A, HEK293 cells were transfected for 48 h with plasmid encoding HA-MEKK1 either alone or together with plasmids for GSK3␤-FLAG and GBP as indicated. Cell lysates were then subjected to immunoprecipitation with anti-HA antibody, and the resulting precipitates were assayed for MEKK1 activity with GST-SEK1(K129R) as substrate. Cells were also subjected to immunoblot analysis with antibodies to HA or to FLAG. B, HEK293 cells were treated for 30 min with LiCl at the indicated concentrations. Cell lysates were then subjected to immunoprecipitation with antibodies to JNK1 or to MEKK1. The resulting immunoprecipitates were assayed for JNK1 or MEKK1 activity with GST-c-Jun-(1–79) or GST-SEK1(K129R), respectively, as substrates. IB, immunoblot. signal-regulating kinase 1 (ASK1) were purchased from Santa Cruz Biotechnology. Mouse monoclonal antibody to JNK1 was from BD Pharmingen. Mouse monoclonal antibodies to the hemagglutinin epitope (HA) and to the FLAG epitope were obtained from Roche Molecular Biochemicals and Stratagene, respectively. Plasmid Constructs—MEKK1 constructs used were prepared as described previously (28). GSK3␤, GSK3␤(K85A), and GSK3␤(S9A) cDNAs were isolated by the polymerase chain reaction using primers (5⬘-GCGGAATTCATGTCAGGGCGGCCCAGA-3⬘ and 5⬘-CGCCTCGAGGGTGGAGTTGGAAGCTGATGC-3⬘) and cloned into pCMV5-FLAG (Eastman Kodak Co.). cDNAs for GSK3␤ deletion mutants (NT, CEN, ⌬N, and ⌬C) were made by the polymerase chain reaction and cloned into the EcoRI/XhoI sites of pcDNA3-HA (Invitrogen). Cell Culture and Transfection—Human embryonic kidney 293 (HEK293) and rat neuroblastoma B103 cells were maintained under a humidified atmosphere of 5% CO2 at 37 °C in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal bovine serum and 100 units/ml penicillin/streptomycin (Invitrogen). Cells were transiently transfected with the use of GenePORTER II (Gene Therapy Systems), by the calcium phosphate method, or by electroporation. Apoptotic Cell Death—B103 cells were transfected for 48 h with pEGFP (Clontech) and the appropriate vector constructs, fixed with 4% paraformaldehyde, and then stained with 4⬘,6-diamidino-2-phenylindole (10 ␮g/ml) for 10 min. The 4⬘,6-diamidino-2-phenylindole-stained nuclei of cells expressing green fluorescence protein (GFP) were examined for apoptotic morphology with a Zeiss Axiovert fluorescence microscope. The percentage of apoptotic cell death was determined as the number of GFP-expressing cells with apoptotic nuclei divided by the total number of GFP-expressing cells. More than 200 cells were counted in each experiment. Co-immunoprecipitation—Cells were lysed in buffer A consisting of 20 mM Tris-HCl (pH 7.4), 150 mM sodium chloride, 1% Triton X-100, 1% deoxycholate, 12 mM ␤-glycerophosphate, 10 mM sodium fluoride, 5 mM EGTA, and 1 mM phenylmethylsulfonyl fluoride. Cell lysates were subjected to centrifugation at 12,000 ⫻ g for 15 min at 4 °C, and the resulting supernatants were subjected to immunoprecipitation with the appropriate antibodies. The resulting precipitates were washed four times with buffer A and then examined by SDS-PAGE and immunoblot analysis. Immunocomplex Kinase Assays—Cells were lysed in buffer A, and the cell lysates were subjected to immunoprecipitation as described above. The resulting immunoprecipitates were assayed for enzymatic activities of the indicated protein kinases as described previously (29, 30). Phosphorylated proteins were separated by SDS-PAGE, and the extent of phosphorylation was quantified with a Fuji BAS2500 imager. Bacterially expressed glutathione S-transferase (GST) fusion proteins of c-Jun-(1–79), SAPK␤(K55R), SEK1(K129R), and ␤-catenin were used as substrates for JNK/SAPK, SEK1, MEKK1, and GSK3␤, respectively. In Vitro Binding Analysis—GSK3␤ and its variants were translated in vitro in the presence of [35S]methionine with the use of a rabbit reticulocyte lysate system (Promega). The resulting 35S-labeled GSK3␤ proteins were then incubated for 3 h at 4 °C with bacterially expressed GST fusion proteins of MEKK1 variants in a solution containing 50 mM Tris, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1 mM dithiothreitol, 0.1% Nonidet P-40, and 5 mg/ml bovine serum albumin. The GST fusion

FIG. 3. Axin1⌬-(217–353) does not inhibit GSK3␤-induced activation of MEKK1. HEK293 cells were transfected for 48 h with an expression vector encoding HA-MEKK1 and the indicated combinations of vectors for GSK3␤-FLAG, axin1, and axin1⌬-(217–353). Cell lysates were then subjected to immunoprecipitation with anti-HA antibody, and the resulting immunoprecipitates were assayed for MEKK1 activity with GST-SEK1(K129R) as substrate. Cell lysates were also subjected to immunoblot (IB) analysis with anti-HA or anti-FLAG antibody to show expression of HA-MEKK1, HA-axin1, HA-axin1⌬-(217–353), or GSK3␤-FLAG. proteins were recovered with the use of glutathione-agarose beads. The beads were then washed three times with washing buffer (50 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 0.1% Tween 20). The associated 35S-labeled proteins were eluted from the beads and analyzed by SDS-PAGE and autoradiography. RESULTS

GSK3␤ Induces the Activation of MEKK1—To investigate the possible effect of GSK3␤ on the SAPK/JNK signaling pathway, we first transfected HEK293 cells with plasmid encoding HA-tagged SAPK␤ either alone or together with a vector for FLAG epitope-tagged GSK3␤ or GSK3␤(K85A). Exposure of the transfected cells to UV resulted in activation of the recombinant SAPK␤ (Fig. 1A). This effect of UV irradiation was enhanced by ectopic expression of GSK3␤ but not by that of the mutant GSK3␤(K85A), which is catalytically inactive (31). Rather, GSK3␤(K85A) inhibited the UV-induced activation of SAPK␤. Whereas overexpressed GSK3␤ also increased the basal activity of SAPK␤, it did not enhance either the basal activity or the phorbol 12-myristate 13acetate-stimulated activity of extracellular signal-regulated kinase 2 (Fig. 1A). We next examined whether GSK3␤ affects the activities of SEK1 and MEKK1, which function as MAPK kinase and MAPK kinase kinase of JNK/SAPK, respectively. Overexpressed GSK3␤ increased in both the basal and UV-stimulated

13998


FIG. 4. GSK3␤ physically associates with MEKK1 in intact cells. A, interaction of ectopic GSK3␤ and MEKK1 in transfected HEK293 cells. Cells were transfected for 48 h with an expression vector encoding GSK3␤-FLAG either alone or together with vectors for HA-MEKK1 or HA-axin1, as indicated. Cell lysates were then subjected to immunoprecipitation (IP) with anti-FLAG antibody, and the resulting precipitates were examined by immunoblot analysis with anti-HA antibody. Cell lysates were also subjected to immunoblot analysis with anti-HA or anti-FLAG antibody. B, physical interaction between endogenous GSK3␤ and MEKK1 in HEK293 cells. Cells were unexposed or exposed to UV (80 J/m2) and then incubated for 15 min before immunoprecipitation with either mouse preimmune IgG or anti-GSK3␤ antibody. The resulting immunoprecipitates were subjected to immunoblot analysis with anti-MEKK1 or anti-ASK1 antibody. IgGH, the heavy chain of immunoglobulin G.

FIG. 5. Direct interaction between GSK3␤ and MEKK1 in vitro. A, 35S-labeled GSK3␤ was produced by in vitro translation and incubated for 3 h at 4 °C with GST alone or the indicated GST-fused deletion mutants of MEKK1. The GST fusion proteins were recovered with the use of glutathione-agarose beads, and the bead-bound proteins were eluted and analyzed by SDS-PAGE and autoradiography. B, in vitro translated 35 S-labeled GSK3␤ variants were incubated for 3 h at 4 °C with GST or GST-MEKK1-N1, after which the 35S-labeled GSK3␤ variants associated with GST-MEKK1-N1 were detected as in A. C, phosphorylation of MEKK1 by GSK3␤ in vitro. HEK293 cells were transfected for 48 h with plasmids encoding GSK3␤-FLAG, HA-MEKK1, or HA-MEKK1(K1253M). Cell lysates were then subjected to immunoprecipitation with anti-FLAG or anti-HA antibody. In vitro phosphorylation assays were performed for 30 min at 30 °C with the indicated combinations of HA-MEKK1, HA-MEKK1(K1253M), and GSK3␤-FLAG immunoprecipitates and 10 mM LiCl in the presence of [␥-32P]ATP (100 ␮Ci/ml). Phosphorylated proteins were analyzed by SDS-PAGE and autoradiography.

activities of SEK1 and MEKK1 (Fig. 1, B and C). In contrast, GSK3␤ did not increase the activity of ASK1 (data not shown), which also functions as a MAPK kinase kinase of JNK/SAPK (32). Furthermore, the GSK3␤-induced activation of JNK1/SAPK␥ was inhibited by expression of MEKK1(K1253M) (Fig. 1D), a catalytically inactive mutant of MEKK1, but not by ASK1(K709R), a kinase-inactive mutant of ASK1. Collectively, these results suggest that GSK3␤ induces the activation of MEKK1, thereby triggering JNK/ SAPK activation. We then examined the possible effect of GBP, an intracellu-

lar inhibitor of GSK3␤ (7), on the activation of MEKK1 by GSK3␤. Overexpressed GBP inhibited the GSK3␤-induced increase in MEKK1 activity (Fig. 2A). Furthermore, exposure of HEK293 cells to LiCl, which functions as an inhibitor of GSK3␤ (33, 34), resulted in a concentration-dependent reduction in the activities of endogenous JNK1 and MEKK1 (Fig. 2B). GSK3␤ Induces MEKK1 Activation in an Axin-independent Manner—Axin, a GSK3␤-regulated scaffolding protein (4), has been shown to activate MEKK1 through a direct protein-protein interaction (35, 36). Furthermore, axin interacts with both GSK3␤ and MEKK1, implying that it might function as a


FIG. 6. PI3K and Akt inhibit GSK3␤-induced activation of MEKK1. A, HEK293 cells were transfected for 48 h with a vector encoding HA-MEKK1 and the indicated combinations of vectors for GSK3␤-FLAG, GSK3␤(S9A)-FLAG, Akt-CA, and Akt(K129R). Cell lysates were then subjected to immunoprecipitation with anti-HA antibody, and the resulting immunoprecipitates were assayed for MEKK1 activity with GST-SEK1(K129R) as substrate. Cell lysates were also subjected to immunoblot (IB) analysis with antibodies to HA and to FLAG. B, HEK293 cells were transfected for 48 h with a vector encoding HA-MEKK1 either alone or together with a vector for GSK3␤-FLAG. The cells were incubated in the absence or presence of 200 nM wortmannin for 30 min and then without or with 1 ␮M insulin for 30 min. Cell lysates were then assayed for MEKK1 activity as in A.

FIG. 7. Role of MEKK1 activation in GSK3␤-induced apoptosis in neuroblastoma B103 cells. Cells were transfected for 24 h with pEGFP and the indicated combinations of expression vectors encoding GSK3␤, GSK3␤(S9A), and MEKK1(K1253M). The cells were then untreated or treated with 1 ␮M insulin for 16 h, fixed, and stained with 4⬘,6-diamidino-2-phenylindole. GFP-expressing cells were examined for the apoptotic nuclei by fluorescence microscopy. The percentage of apoptotic cells was determined, and data are the means ⫾ S.E. of values from the representative from two independent experiments.

molecular linker for MEKK1 and GSK3␤ (36). We therefore examined whether axin mediates the GSK3␤-induced activation of MEKK1. HEK293 cells were cotransfected with an HAMEKK1 vector and various combinations of GSK3␤-FLAG, axin1, and axin1⌬-(217–353) constructs, and then cell lysates were examined for MEKK1 activity. Axin1⌬-(217–353), which lacks the MEKK1-interacting domain, exerts a dominant-negative effect on axin-induced MEKK1 activation (36). Overexpression of either GSK3␤ or axin1 induced MEKK1 activation in HEK293 cells, and these effects of GSK3␤ and axin1 were additive (Fig. 3). Whereas axin1⌬-(217–353) inhibited the activation of MEKK1 by axin1, it did not affect that induced by GSK3␤. These data suggest that GSK3␤ induces MEKK1 activation in an axin-independent manner.

13999

FIG. 8. MEKK1 activity is higher in MEF cells from GSK3␤(ⴙ/ⴙ) mice than in MEF cells from GSK3␤-null mice. A, MEF cells from either GSK3␤⫹/⫹ or GSK3␤⫺/⫺ mice were left untreated, exposed to UV (80 J/m2), and then incubated for 30 min or treated with TNF-␣ (20 ng/ml) for 15 min. Cell lysates were then subjected to immunoprecipitation with anti-MEKK1 or anti- GSK3␤ antibody. The resulting immunoprecipitates were examined for MEKK1 or GSK3␤ activity by immunocomplex kinase assay. B, MEFGSK3␤(⫺/⫺) cells were transfected for 48 h with a vector encoding HA-MEKK1 either alone or together with a vector for GSK3␤-FLAG or GSK3␤(K85A)-FLAG. The cells were then incubated for 15 min in the absence or presence of TNF-␣ (20 ng/ml). Cell lysates were subjected to immunoprecipitation with anti-HA antibody, and the resulting immunoprecipitates were assayed for MEKK1 activity as in A. IB, immunoblot.

GSK3␤ Physically Associates with MEKK1—We next investigated whether GSK3␤ interacts with MEKK1 in intact cells. HEK293 cells were transfected with a GSK3␤-FLAG vector either alone or together with vectors for HA-MEKK1 or HAaxin1. Co-immunoprecipitation analysis revealed the presence of HA-MEKK1 in GSK3␤-FLAG immunoprecipitates (Fig. 4A). HA-axin1 also associated with GSK3␤-FLAG, but it did not affect the extent of the physical interaction between GSK3␤FLAG and HA-MEKK1. Similar analysis of non-transfected cells with antibodies to GSK3␤ and to MEKK1 also revealed an interaction of the two endogenous proteins (Fig. 4B). This interaction was not markedly affected by exposure of the cells to UV radiation. In comparison, endogenous GSK3␤ did not interact with endogenous ASK1. Next, we examined whether GSK3␤ interacts directly with MEKK1 in vitro. In vitro translated 35S-labeled GSK3␤ was incubated with GST-MEKK1 fusion proteins, and an interaction between the two proteins was examined by GST pull-down analysis. 35S-Labeled GSK3␤ bound to MEKK1-N1 (amino acids 1– 401 of MEKK1) and MEKK1-N2 (amino acids 402– 821) but not to MEKK1-⌬NC (amino acids 822–1172) or ⌬MEKK1 (amino acids 1173–1493) (Fig. 5A). The ⌬MEKK1 mutant is a constitutively active form of MEKK1. In a separate in vitro binding analysis with GST-MEKK1-N1 and 35S-labeled GSK3␤ variants, MEKK1-N1 bound to full-length GSK3␤, GSK3␤-NT (amino acids 1–150), and GSK3␤-⌬C (amino acids 1–300), but not to GSK3␤-CEN (amino acids 151–300) or GSK3␤-⌬N (amino acids 151– 420) (Fig. 5B). These results thus suggest that the NH2-terminal region that contains amino acids 1–150 of GSK3␤ is required for binding to MEKK1. Neither GSTMEKK1-⌬NC nor GST-⌬MEKK1 interacted with the various GSK3␤ deletion mutants (data not shown). GSK3␤ phosphorylated both MEKK1 and the kinase-inactive mutant MEKK1(K1253M) in vitro, and these phosphorylation reac-

14000


tions were inhibited by LiCl (Fig. 5C). LiCl also inhibited the autophosphorylation of GSK3b. Akt Inhibits GSK3␤-induced MEKK1 Activation—Akt phosphorylates GSK3␤ on Ser9 and thereby inactivates this enzyme (4, 10, 15). We therefore examined whether Akt negatively regulates the activation of MEKK1 by GSK3␤. HEK293 cells were transfected with an HA-MEKK1 vector either alone or together with various combinations of vectors for GSK3␤FLAG, GSK3␤(S9A)-FLAG, Akt-CA, and Akt(K179A). Akt-CA is a myristoylated form of Akt that is constitutively active (37), and Akt(K179A) is a kinase-inactive mutant of Akt. Ectopic GSK3␤ induced MEKK1 activation in the transfected cells, and this activation was inhibited by Akt-CA but not by Akt(K179A) (Fig. 6A). In contrast, Akt-CA failed to inhibit MEKK1 activation induced by GSK3␤(S9A), which is resistant to the Aktmediated phosphorylation as a result of the substitution of Ser9 with Ala. Insulin initiates the activation of PI3K that in turn induces Akt activation (38 – 40). We thus investigated whether insulin is also able to block the GSK3␤-induced activation of MEKK1. The activation of MEKK1 by GSK3␤ in transfected HEK293 cells was inhibited by an exposure of the cells to insulin (Fig. 6B). Furthermore, the inhibitory effect of insulin was abolished by the PI3K inhibitor wortmannin. These results suggest that the insulin-activated PI3K negatively regulates the activation of MEKK1 by GSK3␤. The Kinase-inactive Mutant MEKK1(K1253M) Blocks GSK3␤-induced Apoptosis—GSK3␤ has been shown to induce cell death in several studies (31, 41). Persistent activation of MEKK1 also induces apoptosis (42, 43). We therefore investigated the possibility that MEKK1 activation mediates apoptosis induced by GSK3␤. Neuroblastoma B103 cells were transiently transfected with vectors encoding GSK3␤ or GSK3␤(S9A) in the absence or presence of a vector for the dominant-negative mutant MEKK1(K1253M), after which the cells were examined for apoptosis (Fig. 7). Overexpressed GSK3␤ or GSK3␤(S9A) increased apoptosis in the transfected cells, and co-expression of MEKK1(K1253M) inhibited this effect. These results thus suggest that activation of MEKK1 contributes to the induction of apoptosis by GSK3␤. Insulin blocked the induction of apoptosis by GSK3␤ but not that by the Akt-resistant mutant GSK3␤(S9A). Reduced MEKK1 Activity in MEFs Derived from GSK3␤ Null Mice—To confirm the biological relevance of GSK3␤-induced MEKK1 activation, we examined the basal, UV-stimulated, and TNF␣-stimulated activities of MEKK1 in MEFs derived from GSK3␤⫹/⫹ and GSK3␤⫺/⫺ mice (Fig. 8A). The MEKK1 activities under all three conditions were higher in MEFGSK3␤(⫹/⫹) than MEFGSK3␤(⫺/⫺). Furthermore, ectopic expression of GSK3␤ increased both the basal and TNF␣stimulated activities of MEKK1 in MEFGSK3␤(⫺/⫺) (Fig. 8B). In comparison, expression of the kinase-inactive mutant GSK3␤(K85A) did not affect the basal or TNF␣-stimulated activities of MEKK1 in MEFGSK3␤(⫺/⫺). These data thus suggest that GSK3␤ functions as a natural activator of MEKK1. DISCUSSION

We have shown that GSK3␤ physically interacts with and activates MEKK1, thereby triggering the JNK/SAPK signaling pathway. Insulin-activated PI3K and Akt inhibited the GSK3␤-induced activation of MEKK1. Furthermore, MEKK1 activity was reduced in MEFs from GSK3␤⫺/⫺ mice compared with that in MEFs from wild-type animals, and forced expression of GSK3␤ increased MEKK1 activity in the cells from GSK3␤⫺/⫺ mice. Our data thus indicate that GSK3␤ is an activator of MEKK1. Axin, a GSK3␤-interacting protein, was also recently shown

to interact directly with and to activate MEKK1 (36). It implies that axin might mediate the activation of MEKK1 by GSK3␤. Our results, however, indicate that GSK3␤ physically associates with and activates MEKK1 in a manner independent of axin. MEKK1 is activated as a result of phosphorylation by upstream kinases or through binding to oligomeric forms of upstream regulators such as TNF receptor associated factor 2 (44, 45). GSK3␤ might thus induce the activation of MEKK1 by phosphorylation given that MEKK1 contains several potential sites for GSK3␤-catalyzed phosphorylation. Indeed, the kinaseinactive mutant GSK3␤(K85A) failed to activate MEKK1. Moreover, GSK3␤ phosphorylated MEKK1 in vitro. Further studies are needed to identify a phosphorylation site(s) in MEKK1 targeted by GSK3␤. We cannot also exclude the possibility that the physical interaction between GSK3␤ and MEKK1 induces the activation of MEKK1 through a mechanism independent of protein phosphorylation. In this study we demonstrate that insulin inhibited the activation of MEKK1 by GSK3␤ and that this effect of insulin was abolished by the PI3K inhibitor wortmannin. Furthermore, Akt inhibited GSK3␤-induced MEKK1 activation. These observations thus suggest that PI3K and Akt modulate the JNK/SAPK signaling pathway through inhibition of GSK3␤-induced MEKK1 activation. Many lines of evidence have previously demonstrated PI3K and Akt as negative regulators of the JNK/ SAPK cascade (46 –50). Akt phosphorylates and inhibits both ASK1 (47) and SEK1 (50), both of which actions result in inhibition of JNK/SAPK. Thus, PI3K and Akt appear to tightly regulate the JNK/SAPK signaling pathway by means of multiple mechanisms. GSK3␤ is implicated in many biological events, including signaling activated by Wnt and glycogen metabolism (4). GSK3␤-deficient mice, however, do not exhibit any marked alterations in the Wnt-induced embryonic development or glycogen metabolism (51). Interestingly, the activation of the transcription factor NF-␬B by TNF-␣ is impaired in MEFs from GSK3␤-deficient mice (51). We now show that MEKK1 activity under either basal or UV- or TNF-␣-stimulated condition is reduced in MEFs from GSK3␤-deficient mice compared with that in MEFs from wild-type mice, consistent with the notion that GSK3␤ potentiates the function of MEKK1. Thus, our results demonstrate that GSK3␤ is an endogenous activator of MEKK1. This function of GSK3␤ may be an important aspect of the mechanism by which this enzyme modulates intracellular signaling, including that triggered by cellular stress. Acknowledgments—We thank Dr. J. W. Woodgett for providing wildtype and GSK3␤⫺/⫺ MEFs and SAPK␤ cDNA. We also thank Dr. R. J. Davis, Dr. G. Johnson, Dr. H. Ichijo, Dr. L. I. Zon, Dr. J. Chung, Dr. S. C. Lin, and Dr. D. Kimelman for providing JNK1, MEKK1, ASK1, SEK1, Akt, axin1, and GBP cDNA vector constructs and Dr. W. A. Toscano, Jr. for critical reading of the manuscript. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Bienz, M. (1999) Curr. Opin. Genet. Dev. 9, 595– 603 Ferkey, D. M., and Kimelman, D. (2000) Dev. Biol. 225, 471– 479 Kim, L., and Kimmel, A. R. (2000) Curr. Opin. Genet. Dev. 10, 508 –514 Woodgett, J. R. (2001) Science’s STKE http://www.stke.org/cgi/content/full/ OC_sigtrans;2001/100/re12 Fagotto, F., Jho, E., Zeng, L., Kurth, T., Joos, T., Kaufmann, C., and Costantini, F. (1999) J. Cell Biol. 145, 741–756 Li, L., Yuan, H., Weaver, C. D., Mao, J., Farr, G. H., III, Sussman, D. J., Jonkers, J., Kimelman, D., and Wu, D. (1999) EMBO J. 18, 4233– 4240 Yost, C., Farr, G. H., III, Pierce, S. B., Ferkey, D. M., Chen, M. M., and Kimelman, D. (1998) Cell 93, 1031–1041 Hart, M. J., de los Santos, R., Albert, I. N., Rubinfeld, B., and Polakis, P. (1998) Curr. Biol. 8, 573–581 Rubinfeld, B., Albert, I., Porfiri, E., Fiol, C., Munemitsu, S., and Polakis, P. (1996) Science 272, 1023–1026 Cross, D. A., Alessi, D. R., Cohen, P., Andjelkovich, M., and Hemmings, B. A. (1995) Nature 378, 785–789 Fang, X., Yu, S. X., Lu, Y., Bast, Jr., R. C., Woodgett, J. R., and Mills, G. B. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 11960 –11965 Hartigan, J. A., Xiong, W. C., and Johnson, G. V. (2001) Biochem. Biophys. Res.

Activation of MEKK1 by GSK3b Commun. 284, 485– 489 13. Oriente, F., Formisano, P., Miele, C., Fiory, F., Maitan, M. A., Vigliotta, G., Trencia, A., Santopietro, S., Caruso, M., Condorelli, G., Van Obberghen, E., and Beguinot, F. (2001) J. Biol. Chem. 276, 37109 –37119 14. Hemmings, B. A. (1997) Science 275, 628 – 630 15. Shaw, M., Cohen, P., and Alessi, D. R. (1997) FEBS Lett. 416, 307–311 16. Davis, R. J. (2000) Cell 103, 239 –252 17. Ip, Y. T., and Davis, R. J. (1998) Curr. Opin. Cell Biol. 10, 205–219 18. Minden, A., and Karin, M. (1997) Biochim. Biophys. Acta 1333, 85–104 19. Schaeffer, H. J., and Weber, M. J. (1999) Mol. Cell. Biol. 19, 2435–2444 20. Chen, Y. R., Wang, X., Templeton, D., Davis, R. J., and Tan, T. H. (1996) J. Biol. Chem. 271, 31929 –31936 21. Dong, C., Yang, D. D., Wysk, M., Whitmarsh, A. J., Davis, R. J., and Flavell, R. A. (1998) Science 282, 2092–2095 22. Kuan, C. Y., Yang, D. D., Samanta Roy, D. R., Davis, R. J., Rakic, P., and Flavell, R. A. (1999) Neuron 22, 667– 676 23. Seimiya, H., Mashima, T., Toho, M., and Tsuruo, T. (1997) J. Biol. Chem. 272, 4631– 4636 24. Tournier, C., Hess, P., Yang, D. D., Xu, J., Turner, T. K., Nimnual, A., Bar-Sagi, D., Jones, S. N., Flavell, R. A., and Davis, R. J. (2000) Science 288, 870 – 874 25. Verheij, M., Bose, R., Lin, X. H., Yao, B., Jarvis, W. D., Grant, S., Birrer, M. J., Szabo, E., Zon, L. I., Kyriakis, J. M., Haimovitz-Friedman, A., Fuks, Z., and Kolesnick, R. N. (1996) Nature 380, 75–79 26. Xia, Z., Dickens, M., Raingeaud, J., Davis, R. J., and Greenberg, M. E. (1995) Science 270, 1326 –1331 27. Zanke, B. W., Boudreau, K., Rubie, E., Winnett, E., Tibbles, L. A., Zon, L., Kyriakis, J., Liu, F. F., and Woodgett, J. R. (1996) Curr. Biol. 6, 606 – 613 28. Kim, J. W., Joe, C. O., and Choi, E.-J. (2001) J. Biol. Chem. 276, 27064 –27070 29. Kim, J. W., Chang, T. S., Lee, J. E., Huh, S. H., Yeon, S. W., Yang, W. S., Joe, C. O., Mook-Jung, I., Tanzi, R. E., Kim, T. W., and Choi, E.-J. (2001) J. Cell Biol. 153, 457– 463 30. Park, H. S., Lee, J. S., Huh, S. H., Seo, J. S., and Choi, E.-J. (2001) EMBO J. 20, 446 – 456 31. Pap, M., and Cooper, G. M. (1998) J. Biol. Chem. 273, 19929 –19932 32. Ichijo, H., Nishida, E., Irie, K., ten Dijke, P., Saitoh, M., Moriguchi, T., Takagi,

14001

M., Matsumoto, K., Miyazono, K., and Gotoh, Y. (1997) Science 275, 90 –94 33. Klein, P. S., and Melton, D. A. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 8455– 8459 34. Stambolic, V., Ruel, L., and Woodgett, J. R. (1996) Curr. Biol. 6, 1664 –1668 35. Zhang, Y., Neo, S. Y., Han, J., and Lin, S. C. (2000) J. Biol. Chem. 275, 25008 –25014 36. Zhang, Y., Neo, S. Y., Wang, X., Han, J., and Lin, S. C. (1999) J. Biol. Chem. 274, 35247–35254 37. Kohn, A. D., Summers, S. A., Birnbaum, M. J., and Roth, R. A. (1996) J. Biol. Chem. 271, 31372–31378 38. Alessi, D. R., and Cohen, P. (1998) Genet. Dev. 8, 55– 62 39. Burgering, B. M., and Coffer, P. J. (1995) Nature 376, 599 – 602 40. Downward, J. (1998) Curr. Opin. Cell Biol. 10, 262–267 41. Hetman, M., Cavanaugh, J. E., Kimelman, D., and Xia, Z. (2000) J. Neurosci. 20, 2567–2574 42. Cardone, M. H., Salvesen, G. S., Widmann, C., Johnson, G., and Frisch, S. M. (1997) Cell 90, 315–323 43. Johnson, N. L., Gardner, A. M., Diener, K. M., Lange-Carter, C. A., Gleavy, J., Jarpe, M B., Minden, A., Karin, M., Zon, L. I., and Johnson, G. L. (1996) J. Biol. Chem. 271, 3229 –3237 44. Baud, V., Liu, Z. G., Bennett, B., Suzuki, N., Xia, Y., and Karin, M. (1999) Genes Dev. 13, 1297–1308 45. Schlesinger, T. K., Fanger, G. R., Yujiri, T., and Johnson, G. L. (1998) Front. Biosci. 3, 1181–1186 46. Cerezo, A., Martinez-A, C., Lanzarot, D., Fischer, S., Franke, T. F., and Rebollo, A. (1998) Mol. Biol. Cell 9, 3107–3118 47. Kim, A. H., Khursigara, G., Sun, X., Franke, T. F., and Chao, M. V. (2001) Mol. Cell. Biol. 21, 893–901 48. Levresse, V., Butterfield, L., Zentrich, E., and Heasley, L. E. (2000) J. Neurosci. Res. 62, 799 – 808 49. Okubo, Y., Blakesley, V. A., Stannard, B., Gutkind, S., and Le Roith, D. (1998) J. Biol. Chem. 273, 25961–25966 50. Park, H. S., Kim, M. S., Huh, S. H., Park, J., Chung, J., Kang, S. S., and Choi, E.-J. (2002) J. Biol. Chem. 277, 2573–2578 51. Hoeflich, K. P., Luo, J., Rubie, E. A., Tsao, M. S., Jin, O., and Woodgett, J. R. (2000) Nature 406, 86 –90