Genetic Deletion of Glycogen Synthase Kinase-3β Abrogates Activation of IκBα Kinase, JNK, Akt, and p44/p42 MAPK but Potentiates Apoptosis Induced by Tumor Necrosis Factor

Glycogen synthase kinase (GSK)-3β is a constitutively active, proline-directed serine/threonine kinase that controls growth modulation and tumorigenesis through multiple intracellular signaling pathways. How GSK-3β regulates signaling pathways induced by cytokines such as tumor necrosis factor (TNF) is poorly understood. In this study, we used fibroblasts derived from GSK-3β gene-deleted mice to understand the role of this kinase in TNF signaling. TNF induced NF-κB activation as measured by DNA binding in wild-type mouse embryonic fibroblasts, but deletion of GSK-3β abolished this activation. This inhibition was due to suppression of IκBα kinase activation and IκBα phosphorylation, ubiquitination, and degradation. TNF-induced NF-κB reporter gene transcription was also suppressed in GSK-3β gene-deleted cells. NF-κB activation induced by lipopolysaccharide, interleukin-1β, or cigarette smoke condensate was completely suppressed in GSK-3β–/– cells. Deletion of GSK-3β also abolished TNF-induced c-Jun N-terminal kinase and p44/p42 mitogen-activated kinase activation. Most surprisingly, TNF-induced Akt activation also required the presence of GSK-3β. TNF induced expression of the NF-κB-regulated gene products cyclin D1, COX-2, MMP-9, survivin, IAP 1, IAP 2, Bcl-xL, Bfl-1/A1, TRAF1, and FLIP in wild-type mouse embryonic fibroblasts but not in GSK-3β–/– cells, and this correlated with potentiation of TNF-induced apoptosis as indicated by cell viability, annexin V staining, and caspase activation. Overall, our results indicate that GSK-3β plays a critical role in TNF signaling and in the signaling of other inflammatory stimuli and that its suppression can be exploited as a potential target to inhibit angiogenesis, proliferation, and survival of tumor cells.

Unlike most kinases, GSK-3␤ is active in resting cells. Akt is activated in response to various mitogens or growth factors and then phosphorylates GSK-3␤ at serine residue 9, inactivating the kinase (21,(25)(26)(27). Some reports (28 -31) indicate that overexpression of GSK-3␤ leads to apoptosis. Apoptosis induced by human immunodeficiency virus-tat, platelet activating factor, or staurosporin has been shown to be mediated through activation of GSK-3␤ (17,32,33). Most interestingly, however, disruption of the murine GSK-3␤ gene results in embryonic lethality caused by severe liver degeneration during development (18); moreover, cells with the GSK-3␤ gene deleted are more sensitive to apoptosis. These results indicate that GSK-3␤ has a role in cell survival.
Cell Lines-The mouse embryonic fibroblast (MEF) derived from GSK-3␤ Ϫ/Ϫ C57Bl/6J mice and its wild type were kindly provided by Dr. James R. Woodgett (Ontario Cancer Institute, Toronto, Canada). These cells have been well characterized and described (18). The MEF derived from p65 Ϫ/Ϫ C57Bl/6J mice and its wild type were kindly provided by Dr. David Baltimore (California Institute of Technology, Pasadena, CA). These cells have been well characterized and described (34). Cells were cultured in DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin.
Electrophoretic Mobility Shift Assays (EMSA)-To determine NF-B activation, we performed EMSA as described (35). Briefly, nuclear extracts prepared from TNF-treated cells (1 ϫ 10 6 /ml) were incubated with 32 P-end-labeled 45-mer double-stranded NF-B oligonucleotide (15 g of protein with 16 fmol of DNA) from the human immunodeficiency virus-long terminal repeat, 5Ј-TTGTTACAAGGGACTTTCCGCTGGG-GACTTTCCAGGGAGGCGTGG-3Ј (boldface indicates NF-B-binding sites), for 30 min at 37°C, and the DNA-protein complex formed was separated from free oligonucleotide on 6.6% native polyacrylamide gels. A double-stranded mutated oligonucleotide, 5Ј-TTGTTACAACT-CACTTTCCGCTGCTCACTTTCCAGGGAGGCGTGG-3Ј, was used to examine the specificity of binding of NF-B to the DNA. The specificity of binding was also examined by competition with the unlabeled oligonucleotide. For supershift assays, nuclear extracts prepared from TNFtreated cells were incubated with antibodies against either p50 or p65 of NF-B for 30 min at 37°C before the complex was analyzed by EMSA. Antibodies against cyclin D1 and preimmune serum were included as negative controls. The dried gels were visualized and radioactive bands quantitated by a PhosphorImager (Amersham Biosciences) using ImageQuant software.
Western Blot Analysis-To determine the levels of protein expression in the cytoplasm or nucleus, we prepared extracts of each from TNFtreated cells (36) and fractionated them by SDS-PAGE. After electrophoresis, the proteins were electrotransferred to a nitrocellulose membrane, blotted with each antibody, and detected by ECL reagents (Amersham Biosciences). The bands obtained were quantitated by using an NIH image (National Institutes of Health, Bethesda).
IKK Assay-The IKK assay was performed by a method described previously (37). Briefly, IKK complexes from whole-cell extracts were precipitated with antibody against IKK-␣ and then were treated with protein A/G-Sepharose beads (Pierce). After a 2-h incubation, the beads were washed with lysis buffer and then assayed in a kinase assay mixture containing 50 mM HEPES, pH 7.4, 20 mM MgCl 2 , 2 mM dithi-othreitol, 20 Ci of [␥-32 P]ATP, 10 M unlabeled ATP, and 2 g of substrate GST-IB␣ (amino acids . After incubation at 30°C for 30 min, the reaction was terminated by boiling with SDS sample buffer for 5 min. Finally, the protein was resolved by 10% SDS-PAGE; the gel was dried, and the radioactive bands were visualized by PhosphorImager.
To determine the total amounts of IKK-␣ and IKK-␤ in each sample, 50 g of the whole-cell protein was resolved by 7.5% SDS-PAGE, electrotransferred to a nitrocellulose membrane, and blotted with either anti-IKK-␣ or anti-IKK-␤ antibody.
NF-B-dependent Reporter Gene Expression Assay-To examine TNF-induced reporter gene expression, cells (3 ϫ 10 5 ) were plated in 6-well plates and transiently transfected by the calcium phosphate method with pNF-B-secretory alkaline phosphatase (SEAP, 0.5 g) and the control plasmid pCMVFLAG1 DNA (2 g). After 24 h, cells were washed, exposed to 1 nM TNF for 24 h, harvested from the cell culture medium, and then analyzed for SEAP activity according to the protocol essentially as described by the manufacturer (Clontech) using a 96-well fluorescence plate reader (Fluoroscan II; Labsystems, Chicago) with excitation set at 360 nm and emission at 460 nm.
Real Time (Quantitative) PCR-Real time PCR assays were performed essentially as described previously (38). The primers used in the reaction had a concentration of 100 nM. A threshold cycle (C t ) value was obtained from each amplification curve using the software provided by the manufacturer. A C t value was calculated by subtracting the C t value for the 2% input sample from the C t value for the immunoprecipitated sample, i.e. ⌬C t ϭ C t(input) Ϫ C t(IP) . The percentage of the total input amount for the immunoprecipitated sample was then calculated by raising 2 to the ⌬C t power, i.e. the total percentage of immunoprecipitated sample ϭ 2 ⌬Ct ϫ 2 as described previously. The following primers, corresponding to the human MMP-9 or COX-2 promoter sequence, were used for real time PCR: forward (5-TGTCCCTTTACTGCCCTGA-3), reverse (5-ACTCCAGGCTCTGTCCTCCTCTT-3) for MMP-9 promoter and forward (5-AAAGACATCTGGCGGAAACCT-3), reverse (5-AG-GAAGCTGCCCCAATTTG-3) for Cox-2. These primers correspond to sequences Ϫ657/Ϫ638 and Ϫ484/Ϫ504 of the MMP-9 promoter or Ϫ434/ Ϫ414 and Ϫ319/Ϫ337 of the COX-2 promoter, respectively. For conventional PCR, the identical primer pair was used with the following PCR cycle parameters: denaturation at 95°C for 15 s, annealing and extension at 60°C for 60 s, with a total of 40 cycles. To verify the amplicon size, PCR products were checked in a 10% polyacrylamide gel.
Cell Transfection-To introduce IB␣-DN or p65 plasmids, 3 ϫ 10 5 cells were seeded, and 2 g of plasmid was diluted in 200 ml of DMEM (without serum and antibiotics) and then mixed with 200 ml of DMEM containing 4 ml of LipofectAMINE 2000. This mixture was added to the cells and incubated for 12 h. After incubation, cells were washed and maintained for an additional 48 h before using for experiments.
Immunocytochemistry for NF-B p65 Localization-The effect of TNF on the nuclear translocation of p65 was examined by an immunocytochemical method as described (39). Briefly, treated cells were plated on a poly-L-lysine-coated glass slide by centrifugation using a cytospin 4 (Thermoshendon, Pittsburgh, PA), air-dried, and fixed with 4% paraformaldehyde following permeabilization with 0.2% Triton-X-100. After being washed in phosphate-buffered saline, slides were blocked with 5% normal goat serum for 1 h and then incubated with rabbit polyclonal anti-human p65 antibody at a 1:100 dilution. After overnight incubation at 4°C, the slides were washed, incubated with goat antirabbit IgG-Alexa 594 at a 1:100 dilution for 1 h, and counter-stained for nuclei with Hoechst 33342 (50 ng/ml) for 5 min. Stained slides were mounted with mounting medium purchased from Sigma and analyzed under a fluorescence microscope (Labophot-2, Nikon, Tokyo, Japan). Pictures were captured by using a Photometrics Coolsnap CF color camera (Nikon, Lewisville, TX) and MetaMorph version 4.6.5 software (Universal Imaging, Downingtown, PA).
Cytotoxicity Assay-The cytotoxic effects of TNF were determined by the MTT uptake method as described (40). Briefly, 2 ϫ 10 4 cells were seeded in triplicate in 96-well plates, pretreated with 1 g/ml cycloheximide, and then treated with various concentrations of TNF for 24 h at 37°C. Thereafter, MTT solution was added to each well. After a 2-h incubation at 37°C, extraction buffer (20% SDS, 50% dimethylformamide) was added, and the cells were incubated overnight at 37°C, and the absorbance was measured at 570 nm by using a 96-well multiscanner (Dynex Technologies, MRX Revelation, Chantilly, VA).
Annexin V Assay-An early indicator of apoptosis is the rapid translocation and accumulation of the membrane phospholipid phosphatidylserine from the cytoplasmic interface to the extracellular surface. This loss of membrane asymmetry can be detected by utilizing the binding properties of annexin V. To identify the apoptosis, we employed annexin V antibody, which conjugated with FITC fluorescence dye. Briefly, 1 ϫ 10 5 cells were pretreated with 1 g/ml cycloheximide, treated with 1 nM TNF for 12 h at 37°C, and subjected to annexin V staining. Cells were washed in phosphate-buffered saline, resuspended in 100 l of binding buffer containing FITC-conjugated anti-annexin V antibody, and then analyzed by flow cytometry (FACSCalibur, BD Biosciences).
TUNEL Assay-We also assayed cytotoxicity by the terminal deoxynucleotidyltransferase-mediated deoxyuridine triphosphate nick end-labeling (TUNEL) method, which examines DNA strand breaks during apoptosis, using the Roche Applied Science in situ cell death detection reagent. Briefly, 1 ϫ 10 5 cells were pretreated with 1 g/ml cycloheximide and then treated with 1 nM TNF for 16 h at 37°C. Thereafter, the cells were plated on a poly-L-lysine-coated glass slide by centrifugation using a cytospin 4, air-dried, fixed with 4% paraformaldehyde, and permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate. After being washed, cells were incubated with reaction mixture for 60 min at 37°C. Stained cells were mounted with mounting medium purchased from Sigma and analyzed under a fluorescence microscope (Labophot-2).
Live and Dead Assay-To measure the apoptosis induced by TNF, we also used the Live and Dead assay (Molecular Probes, Eugene, OR), which determines intracellular esterase activity and plasma membrane integrity. This assay employs calcein, a polyanionic dye, which is retained within live cells and produces a green fluorescence. It also employs the ethidium bromide homodimer dye (red fluorescence), which can enter cells through damaged membranes and bind to nucleic acids but is excluded from live cells by the intact plasma membrane. Briefly, 1 ϫ 10 5 cells were pretreated with 1 g/ml cycloheximide and then treated with 1 nM TNF for 16 h at 37°C. Cells were stained with the Live and Dead® reagent (5 M ethidium bromide homodimer, 5 M calcein-AM) and then incubated at 37°C for 30 min. Cells were analyzed under a fluorescence microscope (Labophot-2).

RESULTS
The goal of this study was to investigate the effects of GSK-3␤ on TNF signaling. This proinflammatory cytokine has been shown to activate NF-B through IKK and to activate p44/p42 MAPK and apoptosis. To investigate the role of GSK-3␤, we used wild-type murine fibroblasts (MEF) and GSK-3␤ gene-deleted fibroblasts (MEF/GSK-3␤ Ϫ/Ϫ ). We first examined the expression of GSK-3␤ protein in both MEF and MEF/GSK-3␤ Ϫ/Ϫ by Western blot analysis. GSK-3␤ was expressed in wild-type MEF but not in GSK-3␤ Ϫ/Ϫ MEF (Fig. 1A). Furthermore, as revealed by Western blot analysis, both TNF receptor 1 and TNF receptor 2 were expressed equally in both cell lines (Fig. 1B).
GSK-3␤ Is Required for TNF-dependent NF-B Activation-We first examined the effect of GSK-3␤ deletion on TNFinduced NF-B activation. Cells were stimulated with TNF, prepared the nuclear extracts, and analyzed them for NF-B activation by EMSA. As shown in Fig. 1C, TNF induced NF-B activation in a time-dependent manner in wile-type fibroblasts but abrogated its activation in GSK-3␤ Ϫ/Ϫ fibroblasts.
Because activation of NF-B by TNF is more robust at higher concentrations (41), we determined the effect of GSK-3␤ deletion on NF-B activation induced by higher concentrations of TNF (Fig. 1D). At a concentration of 10 nM TNF activated NF-B activity stronglyin in wild-type cells but not in GSK-3␤ Ϫ/Ϫ cells. These results show that GSK-3␤ is required for TNF-induced NF-B activation.
Lithium Chloride Suppresses TNF-induced NF-B Activation-Lithium ions have been shown to inhibit GSK-3␣ and -3␤ both in vitro and in intact cells (42). To investigate the effect of lithium chloride (LiCl) on TNF-induced NF-B activation, we pretreated cells with LiCl and then stimulated them with TNF, and we analyzed the nuclear extract by EMSA. KCl and NaCl were used as control. As shown in Fig. 2A, LiCl suppressed TNF-induced NF-B activation, but KCl and NaCl did not. These results suggest that the activity of GSK-3␤ is required for suppression of NF-B.

GSK-3␤ Is Required for NF-B Activation Induced by LPS, IL-1␤, or Cigarette Smoke Condensate-NF-B
is activated by a wide variety of carcinogens and inflammatory stimuli other than TNF, through a mechanism that may differ from that of TNF (43)(44)(45)(46). Therefore, we investigated the role of GSK-3␤ in NF-B activation induced by LPS, IL-1␤, or CSC. As shown in Fig. 2B, all these agents activated NF-B, and GSK-3␤ deletion suppressed this activation in every case. These results suggest that GSK-3␤ must act at a step in the NF-B activation pathway that is common to all these agents.
Because various combinations of Rel/NF-B protein can constitute an active NF-B heterodimer that binds to a specific sequence in the DNA (47), we next showed that the retarded band visualized by EMSA in TNF-treated cells was indeed NF-B. We incubated nuclear extracts from TNF-stimulated cells with antibodies to either the p50 (NF-B1) or the p65 (RelA) subunit of NF-B. Both shifted the band to a higher molecular mass (Fig. 2C), suggesting that the TNF-activated complex consisted of p50 and p65 subunits. Neither preimmune serum nor irrelevant antibody (anti-cyclin D1) had any effect. Excess unlabeled NF-B (100-fold) caused complete disappearance of the band, but the mutant oligonucleotide of NF-B did not affect NF-B binding activity.
GSK-3␤ Is Required for TNF-dependent IB␣ Degradation-Translocation of NF-B to the nucleus is preceded by proteolytic degradation of IB␣ (47). To determine whether inhibition of TNF-induced NF-B activation in GSK-3␤ Ϫ/Ϫ cells was due to inhibition of IB␣ degradation, we exposed cells to TNF for the indicated intervals and assayed degradation of IB␣ by Western blot analysis. TNF induced IB␣ degradation, which precedes NF-B translocation in control cells as early as 15 min in wild-type fibroblasts. In GSK-3␤ Ϫ/Ϫ fibroblasts, however, TNF had no effect on IB␣ degradation (Fig. 3A, upper panel). Thus GSK-3␤ is required for degradation of IB␣.
GSK-3␤ Is Required for TNF-dependent IB␣ Phosphorylation-The proteolytic degradation of IB␣ is known to require phosphorylation at serine residues 32 and 36 (47). To determine whether GSK-3␤ deletion affects TNF-induced IB␣ phosphorylation, we assayed the TNF-induced phosphorylated form of IB␣ by Western blot analysis, using antibody that recognizes the serine-phosphorylated form of IB␣. TNF induced IB␣ phosphorylation in wild-type fibroblasts, but in GSK-3␤ Ϫ/Ϫ fibroblasts, the IB␣ phosphorylation induced by TNF was almost completely suppressed (Fig. 3A, middle  panel).
GSK-3␤ Deletion Inhibits TNF-induced IKK Activation-Because IKK is required for TNF-induced phosphorylation of IB␣ (47), we next determined the effect of GSK-3␤ deletion on TNF-induced IKK activation. As shown in Fig. 3B, GSK-3␤ deletion completely suppressed TNF-induced activation of IKK without any effect on the expression of IKK-␣ or IKK-␤. These results suggest that GSK-3␤ is required for TNF-induced IKK activation.
GSK-3␤ Is Required for TNF-induced Phosphorylation and Nuclear Translocation of p65-Degradation of IB␣ leads to nuclear translocation of the p65 subunit of NF-B. Therefore, we also analyzed the effect of the GSK-3␤ deletion on TNFinduced nuclear translocation of p65 by Western blot analysis. As shown in Fig. 3C, TNF induced nuclear translocation of p65 in a time-dependent manner; as early as 5 min after TNF stimulation, nuclear p65 was noted in wild-type MEF cells. In GSK-3␤ Ϫ/Ϫ fibroblasts, TNF failed to induce nuclear translocation of p65.
TNF induced the phosphorylation of p65, which is required for its transcriptional activity (48). Therefore, we also analyzed the effect of GSK-3␤ deletion on TNF-induced phosphorylation of p65 by Western blot analysis. As shown in Fig. 3C, TNF induced phosphorylation of p65 in a time-dependent manner; as early as 5 min after TNF-stimulation, p65 was phosphorylated in wild-type fibroblasts. In GSK-3␤ Ϫ/Ϫ fibroblasts, TNF failed to induce phosphorylation of p65.
To confirm further the effect of GSK-3␤ deletion on the suppression of nuclear translocation of p65, we performed an immunocytochemical assay. The results showed that p65 is localized in the cytoplasm, that TNF induced nuclear translocation of p65 in wild-type fibroblasts, and that TNF failed to induce p65 translocation to the nucleus in GSK-3␤ Ϫ/Ϫ fibroblasts (Fig. 3D). Whole-cell extracts were prepared, resolved by SDS-PAGE, and electrotransferred to a nitrocellulose membrane, and we then performed Western blot analysis using anti-GSK-3␤ antibody. B, effect of GSK-3␤ deletion on expression levels of TNF receptors. Wholecell extracts were prepared, resolved by SDS-PAGE, and electrotransferred to a nitrocellulose membrane, and we then performed Western blot analysis using anti-TNFR1 and anti-TNFR2 antibodies. C, time-dependent effects of NF-B activation on GSK-3␤ Ϫ/Ϫ fibroblasts by TNF. One million cells were treated with 0.1 nM TNF for the indicated times; nuclear extracts were prepared, and NF-B activation was analyzed by EMSA. D, dose-dependent effect of NF-B activation in GSK-3␤ Ϫ/Ϫ fibroblasts by TNF. One million cells were treated with the indicated concentrations of TNF for 30 min; Nuclear extracts were prepared, and NF-B activity was analyzed by EMSA.

GSK-3␤ Is Required for TNF-induced NF-B-dependent Re-
porter Gene Expression-DNA binding does not always correlate with NF-B-dependent gene transcription (49 -51). To determine the role of GSK-3␤ in TNF-induced NF-B-dependent reporter gene expression, we transiently transfected cells with the NF-B-regulated SEAP reporter construct and then stimulated the cells with different concentrations of TNF. NF-Bregulated reporter gene expression was activated by TNF in a dose-dependent manner in wild-type MEF, but minimal activation was seen in GSK-3␤-deleted cells (Fig. 4). These results suggest that GSK-3␤ is required not only for activation of IKK, nuclear translocation of p65, or p65 binding to the DNA but also for NF-B-regulated reporter gene expression.
GSK-3␤ Is Required for TNF-induced Activation of JNK-We investigated the effect of GSK-3␤ on other signals transduced by TNF. Activation of JNK is one of the earliest events induced by TNF (52). To explore the specific role of GSK-3␤ in TNF-induced JNK activation, we treated cells with TNF for the indicated intervals, prepared whole-cell extracts, and analyzed them for JNK activity by immune complex kinase assay. TNF induced time-dependent activation of JNK in wildtype fibroblasts but not in GSK-3␤ Ϫ/Ϫ fibroblasts (Fig. 5A). These results suggest that GSK-3␤ is required for TNF-induced JNK activation.  2. A, effects of LiCl on TNF-induced NF-B activation. One million cells of wild-type MEF were pretreated with 30 mM KCl, NaCl, or LiCl for 4 h and treated with 0.1 nM TNF for 30 min. Nuclear extracts were prepared, and NF-B activity was analyzed by EMSA. B, effects of various activators on NF-B activation in GSK-3␤ Ϫ/Ϫ fibroblasts. One million cells were pretreated with 0.1 nM TNF, 100 ng/ml IL-1␤, 10 g/ml LPS for 30 min, or 1 g/ml CSC for 1 h; nuclear extracts were prepared, and NF-B activity was analyzed by EMSA. C, composition of NF-B induced by TNF. Nuclear extracts from untreated or TNF-treated MEF cells were incubated with different antibodies, preimmune sera (PIS), or unlabeled NF-B oligoprobe or mutant oligoprobe, and then NF-B activity was analyzed by EMSA.

GSK-3␤ Required for TNF Signaling
GSK-3␤ Is Required for TNF-induced Activation of p44/p42 MAPK-TNF has been shown to activate p44/p42 MAPK through the Ras/Raf/MAPK kinase cascade, (52). To explore the specific role of GSK-3␤ in TNF-induced p44/p42 MAPK activation, we treated the fibroblasts with TNF for the indicated intervals, prepared whole-cell extracts, resolved them by SDS- FIG. 3. A, effects of GSK-3␤ deletion on TNF-induced phosphorylation and degradation of IB␣. One million cells were treated with 0.1 nM TNF for the indicated times. Cytoplasmic extract was prepared, resolved by SDS-PAGE, and electrotransferred onto a nitrocellulose membrane, and then Western blot analysis using anti-IB␣ and anti-phospho-specific IB␣ antibodies was performed. B, effects of GSK-3␤ deletion on TNFinduced activation of IKK. One million cells were pretreated with 50 g/ml proteosome inhibitor ALLN for 1 h and then stimulated with 1 nM TNF for the indicated times. Whole-cell extracts were incubated with anti-IKK-␣ antibody for 2 h, immunoprecipitated with protein A/G-Sepharose beads, and then analyzed by immunocomplex kinase assay. To examine the level of expression of IKK proteins, the same whole-cell extracts were resolved by SDS-PAGE and performed Western blot analysis using anti-IKK-␣ and anti-IKK-␤ antibodies. C, effects of GSK-3␤ deletion on TNF-induced nuclear translocation of the p65 subunit of NF-B. One million cells were treated with 0.1 nM TNF for the indicated times; nuclear extract was prepared, resolved by SDS-PAGE, and electrotransferred to a nitrocellulose membrane, and then Western blot analysis using anti-p65 and anti-phospho-specific p65 antibodies was performed. D, immunocytochemical analysis of p65 localization. Cells were treated with 1 nM TNF for 20 min and then cells were subjected to immunocytochemical analysis as described under "Materials and Methods." PAGE, and performed Western blot analysis by using phosphospecific anti-p44/p42 MAPK antibody. Time-dependent phosphorylation of p44/p42 MAPK occurred in wild-type cells, but p44/p42 MAPK activation was abolished in GSK-3␤ Ϫ/Ϫ fibroblasts (Fig. 5B). These results suggest that GSK-3␤ is required for TNF-induced p44/p42 MAPK activation also.  5. A, effects of GSK-3␤ deletion on TNF-induced activation of JNK. One million cells were treated with 1 nM TNF for the indicated times; whole-cell extracts were prepared, incubated with anti-JNK1 antibody for 2 h, and then immunoprecipitated with protein A/G-Sepharose beads. The beads were washed and subjected to kinase assay as described under "Materials and Methods." The same protein extracts were resolved by SDS-PAGE, electrotransferred to a nitrocellulose membrane, and then Western blot analysis using anti-JNK1 antibody was performed. B, effects of GSK-3␤ deletion on TNF-induced activation of p44/p42 MAPK. One million cells were treated with 1 nM TNF for the indicated times; whole-cell extracts were prepared, resolved by SDS-PAGE, and electrotransferred to a nitrocellulose membrane, and then Western blot analysis using phospho-specific anti-p44/p42 MAPK antibody was performed. The same membrane was reblotted with anti-p42 MAPK antibody. C, effects of GSK-3␤ deletion on TNF-induced activation of Akt. One million cells were treated with 1 nM TNF for the indicated times; whole-cell extracts were prepared, resolved by SDS-PAGE, and electrotransferred to a nitrocellulose membrane, and then Western blot analysis using phospho-specific anti-Akt antibody was performed. The same membrane was reblotted with anti-Akt antibody.

GSK-3␤ Is Required for TNF-induced Activation of Akt-
Activation of Akt is also the earliest event induced by TNF (53). Activated Akt has been shown to inactivate GSK-3␤ through phosphorylation (54 -56). To explore the specific role of GSK-3␤ in TNF-induced Akt activation, we treated cells with TNF for the indicated intervals, prepared whole-cell extracts, resolved them by SDS-PAGE, and performed Western blot analysis by using phospho-specific anti-Akt antibody. Time-dependent activation of Akt occurred in wild-type fibroblasts, but not in GSK-3␤ Ϫ/Ϫ fibroblasts (Fig. 5C). Thus although Akt can inactivate GSK-3␤, GSK-3␤ is needed for TNF-induced Akt activation.

GSK-3␤ Is Required for Expression of TNF-induced NF-Bdependent Antiapoptotic Proteins-Because
GSK-3␤ and NF-B Are Required for TNF-induced COX-2 Promoter Activity-It has been shown that NF-B activation is required for COX-2 expression. Whether the COX-2 promoter is regulated by GSK-3␤ through modulation of NF-B was examined. As shown in Fig. 7A, TNF activated COX-2 promoter activity in wild-type MEF but not in GSK-3␤-deleted cells, thus suggesting that GSK-3␤ is needed for COX-2 expression. Whether COX-2 promoter is regulated by NF-B was examined by using p65 NF-B-deleted cells. As shown in Fig. 7A, TNF activated COX-2 promoter activity in wild-type MEF but not in p65-deleted cells, thus suggesting that NF-B is needed for COX-2 expression.
IB␣-DN Transfection Down-regulates and p65 NF-B Upregulates Expression of COX-2 and Cyclin D1-Whether suppression of NF-B by IB␣-DN mimics the loss of GSK-3␤ in NF-B-regulated gene expression was also examined. To determine this, wild-type MEF cells were transfected with IB␣-DN plasmid and then treated with 1 nM TNF and analyzed COX-2 and cyclin D1 protein expressions (Fig. 7B). The exogenously transfected IB␣-DN down-regulated TNF-induced expression of COX-2 and cyclin D1 in wild-type MEF cells. TNF-induced COX-2 expression was also abolished in MEF/p65 Ϫ/Ϫ cells.
Whether overexpression of p65 reverses the loss of GSK-3␤ was also examined. To determine this, MEF/GSK-3␤ Ϫ/Ϫ cells were transfected with p65 plasmid, and then after 48 h cells were treated with 1 nM TNF and analyzed expression of COX-2 FIG. 6. A, effects of GSK-3␤ deletion on expression of NF-B-regulated proteins. Cells (5 ϫ 10 5 cells/well) were treated with 1 nM TNF for the indicated times. Whole-cell extracts were prepared and analyzed by Western blot analysis using antibodies against COX-2, MMP-9, cyclin D1, and ␤-actin. B, effects of GSK-3␤ deletion on expression of NF-B-regulated antiapoptotic proteins. Cells (5 ϫ 10 5 cells/well) were incubated with 1 nM TNF for the indicated times. Whole-cell extracts were prepared and analyzed by Western blot analysis using antibodies against survivin, IAP 1, IAP 2, XIAP, Bcl-2, Bfl-1/A1, TRAF1, FLIP, and ␤-actin. and cyclin D1 proteins (Fig. 7C). Ectopic expression of p65 reversed the TNF-induced COX-2 and cyclin D1 expression in GSK-3␤-deleted cells. Similarly, COX-2 and cyclin D1 expres-sions were induced by TNF in MEF/p65 ϩ/ϩ cells but not in MEF/p65 Ϫ/Ϫ cells. Taken together, these results provide the evidence that GSK-3␤ mediates its effect on TNF-induced gene Cells (wild-type and GSK-3␤-deleted) were treated with 1 nM TNF for the indicated times; the proteins were cross-linked with DNA by using formaldehyde and subjected to ChIP assay by using an anti-p65 antibody with the indicated primers. Reaction products were resolved by electrophoresis (A1 and A2) or quantified by real time PCR (B1 and B2). After analysis, a melting curve was performed to confirm the amplification of a single product (C1 and C2). IP, immunoprecipitation; Ab, antibody.

GSK-3␤ Required for TNF Signaling 39550 expression through activation of NF-B.
GSK-3␤-mediated NF-B Activation Regulates the Expression of MMP-9 and COX-2 in Vivo-TNF has been shown to induce MMP-9 and COX-2 (39,57), both of which have NF-Bbinding sites in their promoters (58 -60). Whether the lack of TNF-induced gene expression in GSK-3␤-deleted cells is due to suppression of NF-B activation in vivo, we performed chromatin immunoprecipitation (ChIP) assay. NF-B-binding sites in the promoters of MMP-9 and COX-2 were examined. Cells were treated with 1 nM TNF, cross-linked in situ DNA-protein complexes, and isolated chromatin and sheared. Subsequently the chromatin was immunoprecipitated with the anti-p65 antibody, purified the DNA, and subjected to real time PCR for quantification using MMP-9 or COX-2 promoter-specific primers. The predicted size of the DNA was found in the wild-type MEF cells (Fig. 8, A1 and A2). By using quantitative PCR, we found that the TNF induced NF-B binding to MMP-9 and COX-2 promoters (Fig. 8, B1 and B2) isolated from wild-type MEF but not from GSK-3␤-deleted cells. These results suggest that NF-B bound in vivo to the regulatory portion of the MMP-9 and COX-2 promoter in wild-type MEF. A single amplification product was evident in the melting curve following the real time PCR, again confirming specificity of the assay for the MMP-9 and COX-2 promoters (Fig. 8, C1 and C2). Overall, these results also confirm the requirement of GSK-3␤ for the activation of NF-B leading to expression of COX-2 and MMP-9.
Deletion of GSK-3␤ Potentiates TNF-induced Apoptosis-Activation of NF-B has been shown to inhibit TNF-induced apoptosis (69 -73). Our results suggest that deletion of GSK-3␤ might enhance apoptosis induced by TNF through suppression of NF-B-regulated antiapoptotic gene products. Whether sup- FIG. 9. Effect of GSK-3␤-deletion on TNF-induced cytotoxicity. A, twenty thousand cells were seeded in triplicate in 96-well plates. Cells were pretreated with 1 g/ml cycloheximide (CHX) for 1 h and then incubated with the indicated concentrations of TNF for 24 h. Thereafter, cell viability was analyzed by the MTT method as described under "Materials and Methods." B, cells (1 ϫ 10 5 cells/well) were pretreated with 1 g/ml cycloheximide for 1 h and then incubated with 1 nM TNF for 16 h. Cells were stained with Live and Dead assay reagent for 30 min and then analyzed under a fluorescence microscope as described under "Materials and Methods." C, cells (1 ϫ 10 5 cells/well) were pretreated with 1 g/ml cycloheximide for 1 h, incubated with 1 nM TNF for 12 h, and then subjected to annexin V staining. Cells were washed, incubated with FITC-conjugated anti-annexin V antibody, and then analyzed by flow cytometry. D, one million cells were pretreated with 1 g/ml cycloheximide for 1 h and then incubated with 1 nM TNF for 16 h. Whole-cell extracts were prepared, resolved by SDS-PAGE, and then Western blot analysis using anti-PARP and ␤-actin antibodies was performed. E, ectopic expression of IB␣-DN enhances TNF-induced cytotoxicity. Cells were transfected with mock or IB␣-DN plasmid for 48 h and then treated with 1 nM TNF. After 24 h, cell viability was examined by the MTT method. F, ectopic expression of p65 suppresses TNF-induced cytotoxicity. Cells were transfected with mock or p65 plasmid for 48 h and then treated with 1 nM TNF. After 24 h, cell viability was examined by the MTT method.
pression of NF-B by GSK-3␤ deletion affects TNF-induced apoptosis was therefore investigated. MTT assay showed that TNF was cytotoxic to cells and that GSK-3␤ deletion enhanced that cytotoxicity (Fig. 9A). A Live and Dead assay indicated that GSK-3␤ deletion up-regulated TNF-induced apoptosis from 32.4 to 86.9% (Fig. 9B). Whether enhanced cytotoxicity was due to apoptosis was further investigated. Annexin V staining results showed that TNF-induced apoptosis was enhanced from 23.5% in the wild-type MEF to 87.2% in GSK-3␤ Ϫ/Ϫ MEF (Fig. 9C). The PARP-cleavage assay showed that GSK-3␤ deletion potentiated TNF-induced caspase activity (Fig. 9D). All these assay results together suggest that GSK-3␤ deletion also potentiates TNF-induced apoptosis.
Ectopic Expression of IB␣-DN Enhances and of p65 NF-B Blocks TNF-induced Cytotoxicity-Whether enhancement of TNF-induced apoptosis in GSK-3␤-deficient cells is due to lack of activation of NF-B was further investigated. To determine this, wild-type MEFs were transfected with DN-IB␣ plasmid and then examined for TNF-induced cytotoxicity. Results in Fig. 9E show that suppression of NF-B by DN-IB␣ enhances TNF-induced cytotoxicity. Similarly, transfection of GSK-3␤deficient cells with the p65-expressing plasmid inhibited TNFinduced cytotoxicity (Fig. 9F). These results clearly demonstrate the antiapoptotic role of NF-B activated through GSK-3␤. DISCUSSION The goal of the study presented here was to investigate the role of GSK-3␤ in TNF signaling. Our findings show that TNFinduced NF-B activation, IKK activation, IB␣ phosphorylation, IB␣ ubiquitination, IB␣ degradation, and NF-B reporter gene transcription were all suppressed in GSK-3␤ genedeleted fibroblasts. NF-B activation induced by LPS, IL-1␤, or CSC was abrogated in GSK-3␤ Ϫ/Ϫ fibroblasts. Deletion of GSK-3␤ also abolished TNF-induced JNK, p44/p42 MAPK, and Akt activation. TNF induced the expression of NF-B-regulated gene products cyclin D1, COX-2, MMP-9, survivin, IAP 1, IAP 2, Bcl-x L , Bfl-1/A1, TRAF1, and FLIP and were all down- GSK-3␤ Required for TNF Signaling modulated in the GSK-3␤ Ϫ/Ϫ fibroblasts, and this correlated with the potentiation of TNF-induced apoptosis.
Our results clearly show that deletion of the GSK-3␤ gene abolishes TNF-induced NF-B activation. These results are in agreement with a previous report (18). Our results, however, appear to differ from those of Hoeflich et al. (18) in the mechanism of how GSK-3␤ suppresses TNF-induced NF-B activation. They found that deletion of GSK-3␤ had no effect on TNF-induced IB␣ degradation or on the nuclear translocation of p65, but they placed the defect downstream of IB␣ phosphorylation and the nuclear translocation of NF-B. Hoeflich et al. (18) also showed that IB␣ resynthesis, which is under the control of NF-B, is unchanged by deletion of GSK-3␤, and suggested that partial inhibition of NF-B DNA binding is not sufficient for down-regulation of the IB␣ gene. In contrast, we found that GSK-3␤ is needed for activation of IKK, IB␣ phosphorylation, IB␣ degradation, and NF-B-dependent gene expression. Similar to our results, the recent findings of Sanchez et al. (31) showed that TNF-induced NF-B-dependent reporter activity, NF-B-dependent COX-2 expression, and DNA binding of NF-B are completely blocked in the GSK-3␤ mutant S9A in primary astrocytes. Also similar to our results were the findings of Sanchez et al. (31) showing that TNF-induced IKK activation and IB␣ degradation are suppressed by the GSK-3␤ mutant S9A. Moreover, Sanchez et al. (31) showed that the N-terminal region of NEMO (also called IKK-␥), a subunit of IKK, physically interacts with GSK-3␤ S9A and competes for binding to IKK-␣ and IKK-␤.
We found that deletion of GSK-3␤ abolishes the TNF-induced phosphorylation of p65. These results are in agreement with a report by Schwabe et al. (74) that recombinant GSK-3␤ phosphorylates p65 between residues 354 and 551.
We found that the TNF-induced JNK and p44/p42 MAPK activations are abolished on deletion of GSK-3␤. It was reported recently (23) that GSK-3␤ physically associates with and activates MEKK1, thereby stimulating the JNK pathway. Because MEKK1 is required for activation of JNK (75-77), p44/p42 MAPK (78 -80), and NF-B (81,82), it is possible that abrogation of TNF-induced activation of IKK, JNK, and p44/ p42 MAPK was due to the lack of activation of MEKK1 by GSK-3␤. Indeed, Kim et al. (23) found that MEKK1 activation was lower in GSK-3␤ Ϫ/Ϫ MEF than in wild-type MEF. Unlike our results, however, those of Sanchez et al. (31) showed that TNF-induced p44/p42 MAPK activity was unaffected in GSK-3␤ mutant S9A primary astrocytes. Sanchez et al. (31) used an adenoviral construct to transfect the cells with GSK-3␤, and they showed that adenovirus alone activated p44/p42 MAPK. This may explain why they failed to document the suppression of p44/p42 MAPK by GSK-3␤ mutant S9A.
Our results indicate that deletion of GSK-3␤ abolishes TNFinduced Akt activation. To our knowledge, there is no prior report about the effect of GSK-3␤ on the activation of Akt. Akt is, however, known to inactivate GSK-3␤. This suggests that there exists a positive feedback loop between GSK-3␤ and Akt.
The down-modulation of cyclin D1, survivin, IAP 1, IAP 2, Bcl-x L , Bfl-1/A1, TRAF1, and cFLIP, all antiapoptotic gene products, suggests that deletion of GSK-3␤ should enhance apoptosis. Indeed, we found that TNF-induced apoptosis was significantly enhanced by GSK-3␤ deletion. These results are in agreement with a recent report (85) that showed that suppression of GSK-3␤ sensitizes prostate cancer cells to TRAIL. Our results are also in agreement with those of Sanchez et al. (31) showing that GSK-3␤-mediated apoptosis in primary astrocytes requires inhibition of NF-B signaling. Inhibition of GSK-3␤ inhibited Wnt-1-dependent NF-B activation leading to inhibition of growth of PC-12 cells (86). All of these results indicate that GSK-3␤ has a prosurvival role. Our results demonstrate that this role is mediated through the expression of cyclin D1, survivin, IAP 1, IAP 2, Bcl-x L , Bfl-1/A1, TRAF1, and cFLIP. Overall, our results demonstrate that GSK-3␤ plays a critical role not only in TNF-induced NF-B activation but also in TNF-induced activation of IKK, JNK, p44/p42 MAPK, Akt, and apoptosis (Fig. 10). Because of the critical role of TNF and NF-B in cancer, inflammation, diabetes, and neurodegenerative diseases, inhibitors of GSK-3␤ have therapeutic potential in these diseases (87).