p57KIP2 Modulates Stress-activated Signaling by Inhibiting c-Jun NH2-terminal Kinase/Stress-activated Protein Kinase*

p57KIP2, a member of the Cip/Kip family of enzymes that inhibit several cyclin-dependent kinases, plays a role in many biological events including cell proliferation, differentiation, apoptosis, tumorigenesis and developmental changes. The human p57KIP2 gene is located in chromosome 11p15.5, a region implicated in sporadic cancers and Beckwith-Wiedemann syndrome. We here report that p57KIP2 physically interacts with and inhibits c-Jun NH2-terminal kinase/stress-activated protein kinase (JNK/SAPK). The carboxyl-terminal QT domain of p57KIP2 is crucial for the inhibition of JNK/SAPK. Overexpressed p57KIP2 also suppressed UV- and MEKK1-induced apoptotic cell death. p57KIP2 expression during C2C12 myoblast differentiation resulted in repression of the JNK activity stimulated by UV light. Furthermore, UV-stimulated JNK1 activity was higher in mouse embryonic fibroblasts derived from p57–/– mice than in the cells from wild-type mice. Taken together, these findings suggest that p57KIP2 modulates stress-activated signaling by functioning as an endogenous inhibitor of JNK/SAPK.

WAF1, and p27 Kip1 , can inhibit all CDKs that regulate the G 1 /S-phase transition (2,3). p57 KIP2 shares sequence homology with p27 Kip1 and p21 CIP1/WAF1 in the NH 2 -terminal domain, which is involved in the binding to cyclin-CDK complexes (4,5). p57 KIP2 and p27 Kip1 also have unique carboxyl-terminal QT domains. The function of the QT domain is unclear. The human p57 KIP2 gene, which encodes a 316-amino acid protein, is located in chromosome 11p15.5, a region implicated in sporadic cancers and Beckwith-Wiedemann syndrome, a familial cancer syndrome (5). Interestingly, p57 KIP2 null mice show altered cell proliferation and differentiation, apoptosis, and many other phenotypes that can be observed in patients with Beckwith-Wiedemann syndrome (6,7). Thus, p57 KIP2 has been implicated in the modulation of many cellular events including cell cycle control, differentiation, apoptosis, tumorigenesis, and development. However, the mechanism by which p57 KIP2 exerts its modulatory functions is not yet fully understood.
A variety of extracellular stimuli initiate intracellular signaling through the sequential protein phosphorylations leading to the activation of mitogen-activated protein kinases (MAPKs). The mammalian MAPK family includes several subgroups such as extracellular signal-regulated kinase (ERK), p38 MAPK, and c-Jun NH 2 -terminal kinase/stress-activated protein kinase (JNK/SAPK) (8 -10). The JNK/SAPK signaling pathway is preferentially stimulated in cellular responses to a variety of stresses that include UV light, ionizing irradiation, DNA-damaging chemicals, reactive oxygen species, heat and osmotic shock, and metabolic inhibitors (9,10). The JNK/SAPK signaling cascade is composed of JNK/SAPK and its upstream kinases, MAPK kinases such as SEK1/JNKK1/MKK4 and MAPK kinase kinases such as MEKK1. Once activated, JNK/ SAPK phosphorylates its cellular substrates, which include several transcription factors such as c-Jun, ATF2, and TCF/ Elk-1, thereby enhancing their transcriptional activities (11)(12)(13). Accordingly, JNK/SAPK has been implicated in the regulation of diverse cellular activities such as cell growth, transformation, survival, and death (9,10).
JNK/SAPK activity has been shown to be modulated by other proteins through protein-protein interactions (14 -16). We previously reported that overexpressed p21 CIP1/WAF1 binds and inhibits JNK/SAPK (15). In the present study, we investigated * This work was supported by the Creative Research Initiatives Program of the Korean Ministry of Science and Technology (to E.-J. 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  whether p57 KIP2 can modulate the JNK/SAPK signaling pathway. Our findings indicate that p57 KIP2 negatively regulates the JNK/SAPK signaling cascade through direct inhibition of JNK/SAPK, independently of its well known inhibitory function on CDKs. This new function of p57 KIP2 on JNK/SAPK may be an important mechanism by which p57 KIP2 can modulate the intracellular signaling events that mediate a variety of cellular activities including cell differentiation and survival.
Cell Culture and Transfection-Human embryonic kidney HEK293, COS7, and HeLa cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum in a 5% CO 2 atmosphere at 37°C. Mouse embryonic fibroblast (MEF) cells from wild-type mice and from p57 knockout mice were used between passage three and five (17), and they were in the same passage during the experiments. Myoblast C2C12 cells were routinely grown in Dulbecco's modified Eagle's medium supplemented with 20% fetal bovine serum (growth medium). Where indicated, C2C12 cells were moved to a complete culture medium containing 2% fetal bovine serum (differentiation medium). Transfection of cultured cells was performed by the Lipo-fectAMINE (Invitrogen) or calcium phosphate method.
Protein Kinase Assay for GST-SEK1-Cultured cells were transiently transfected with an expression vector pEBG encoding GST-SEK1. Cell lysates of the transfected cells were subjected to microcentrifugation at 4°C for 10 min and solubilized with 1% Triton X-100. The soluble fraction was applied to glutathione-agarose beads, and GST-SEK1 was prepared as described previously (15). GST-SEK1 eluted from the beads was assayed for SEK1 activity using GST-SAPK␤ as substrate.
Co-immunoprecipitation of JNK and p57 KIP2 -To test the physical association of endogenous p57 KIP2 and JNK1 protein in intact cells, HeLa cells were lysed in buffer A containing 120 mM NaCl, 50 mM Tris-HCl (pH 8.0), 1 mM EDTA, 0.5% Nonidet P-40, 1 mM dithiothreitol, 10 mM NaF, 2 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1% digitonin. The lysates were then subjected to immunoprecipitation with the use of rabbit polyclonal anti-p57 KIP2 antibody or rabbit pre-immune IgG. The resulting immunopellets were subjected to SDS-PAGE on 10% polyacrylamide gel and analyzed by immunoblot with mouse monoclonal anti-JNK1 antibody. To test the interactions of ectopically expressed proteins, HEK293 cells were transfected with pcDNA3 vectors encoding HA-JNK3/SAPK␤ and p57 KIP2 . The transfected cells were lysed with buffer A, and the cell lysates were subjected to immunoprecipitation with the use of mouse monoclonal anti-HA antibody or mouse preimmune IgG as the negative control. The immunopellets were analyzed by immunoblot probed with rabbit polyclonal anti-p57 KIP2 antibody.

FIG. 1. p57 KIP2 inhibits JNK/SAPK in intact cells.
A, ectopic p57 KIP2 inhibits JNK, but not ERK or p38 MAPK, in HEK293 cells. HEK293 cells were transfected with plasmid vectors encoding HA-JNK1, HA-ERK2, or p38-FLAG along with plasmid encoding p57 KIP2 as indicated. After 48 h, the cells were exposed to either 60 J/m 2 UV light for JNK and p38 stimulation or 200 nM 12-O-tetradecanoylphorbol-13-acetate (TPA) for 30 min for ERK stimulation. When exposed to UV light, the cells were further incubated for 1 h at 37°C. After treatment, cells were lysed and cell lysates subjected to immunoprecipitation with the use of anti-HA or anti-FLAG antibody. The resulting immunoprecipitates were examined for kinase activities of JNK1, p38, and ERK2 by immune complex kinase assay. Cell lysates were also subjected to immunoblot (IB) analysis with indicated antibodies. B, p57 KIP2 does not affect SEK1 activity. HEK293 cells were transfected for 48 h with plasmids encoding GST-SEK1 and p57 KIP2 as indicated. Where indicated, the cells were exposed to 60 J/m 2 UV light and then incubated for 1 more hour. Cells were lysed and the soluble fraction mixed with glutathione-agarose beads. GST-SEK1 was eluted from the beads and assayed for SEK1 activity using GST-SAPK␤ as substrate. C, p57 KIP2 does not affect MEKK1 activity. HEK293 cells were transiently transfected for 48 h with plasmids encoding MEKK1-FLAG and p57 KIP2 as indicated. Cell lysates were subjected to immunoprecipitation with anti-FLAG antibody, and the resulting immunoprecipitates were assayed for MEKK1 activity. These data represent the results from three independent experiments.
Apoptotic Cell Death-COS7 cells were transiently transfected with pcDNA3 empty vector, pcDNA3-p57 KIP2 , pcDNA3-⌬ MEKK1, or pcDNA3-SEK(K129R) along with pEGFP. After 24 h of transfection, the cells were fixed with 70% ethanol, washed twice with phosphate-buffered saline, and stained with 4Ј,6-diamidino-2-phenylindole (DAPI). DAPI-stained nuclei were examined for apoptotic morphology with a Zeiss Axiovert fluorescence microscope. The percentage of GFP-expressing cells with apoptotic nuclei was determined from three independent dishes. TUNEL staining was performed using in situ cell death detection kit (Roche Applied Science) as described previously (14).

RESULTS
p57 KIP2 Inhibits JNK/SAPK Activity in Intact Cells-To investigate whether p57 KIP2 could modulate the JNK signaling pathway, we examined the effect of p57 KIP2 on JNK activity in intact cells. We transfected HEK293 cells with plasmids encoding p57 KIP2 and JNK1 and induced JNK stimulation by exposing transfected cells to UV light (Fig. 1). Our data indicate that ectopic p57 KIP2 inhibited the UV-stimulated activity of JNK1 (Fig. 1A) or SAPK␤/JNK3 (data not shown), whereas it did not affect either the UV-stimulated p38 MAPK activity or the phorbol ester-stimulated ERK2 activity (Fig. 1A). Because the JNK signaling system comprises JNK and its upstream kinases, which include a MAPK kinase such as SEK1/MKK4/JNKK1 and a MAPK kinase kinase such as MEKK1, we decided to examine whether p57 KIP2 would also inhibit SEK1 or MEKK1 activity. HEK293 cells were transfected with plasmid encoding GST-SEK1, and then the transfected cells were unexposed or exposed to UV light. UV irradiation enhanced the kinase activity of GST-SEK1 in the cells (Fig. 1B). Co-expressed p57 KIP2 did not affect the UV-stimulated GST-SEK1 activity. Similarly, p57 KIP2 also did not affect the UV-induced stimulation of MKK7, which is another MAPK kinase upstream of JNK (data not shown). We looked next for a possible effect of p57 KIP2 on MEKK1 activity (Fig. 1C). HEK293 cells transfected with plas- The input of 35 S-labeled proteins (33%) is also shown. B, ectopic p57 KIP2 physically associates with JNK1 in transfected HEK293 cells. HEK293 cells were transfected with plasmids expressing p57 KIP2 and HA-JNK1. Cell lysates were subjected to immunoprecipitation (IP) using mouse monoclonal anti-HA or mouse pre-immune IgG, respectively. The resulting immunoprecipitates were subjected to immunoblot (IB) analysis with rabbit polyclonal anti-p57 KIP2 antibody. Cell lysates (10% of total) were also subjected to immunoblot analysis with anti-p57 KIP2 antibody. C, physical interaction between endogenous p57 KIP2 and JNK proteins in HeLa cells. HeLa cell lysates were subjected to immunoprecipitation with rabbit pre-immune IgG or anti-p57 KIP2 antibody. Immunopellets were subjected to SDS-PAGE on a 10% polyacrylamide gel and immunoblot probed with mouse monoclonal anti-JNK1 antibody. Cell lysates (5% of total) were also subjected to immunoblot analysis with anti-JNK1 antibody.

FIG. 3. p57 KIP2 inhibits an interaction between JNK1 and c-Jun. HEK293 cells were transfected for 48 h with expression vectors
encoding c-Jun-FLAG (3 g) and HA-JNK1 (3 g) in the absence or presence of p57 KIP2 construct (1 or 2 g) as indicated. Cell lysates were subjected to immunoprecipitation (IP) with anti-HA antibody. The resulting immunoprecipitates were subjected to immunoblot (IB) analysis with anti-FLAG antibody. Amounts of p57 KIP2 , HA-JNK1, and c-Jun-FLAG present in cell lysates were also examined by immunoblotting with indicated antibodies. These data represent the results from two independent experiments. mid expressing MEKK1 showed high enzymatic activity of MEKK1 even without any further treatment, and ectopic p57 KIP2 did not alter the MEKK1 activity. A separate transfection study also indicated that p57 KIP2 did not affect the UVstimulated MEKK1 activity (data not shown). Taken together, these results suggest that p57 KIP2 suppresses the JNK/SAPK signaling pathway through acting on a site or sites downstream of MAPK kinase, probably JNK itself.
p57 KIP2 Physically Interacts with JNK-In the following experiments, we tested whether p57 KIP2 could interact with JNK/ SAPK. An in vitro binding study using in vitro translated 35 S-labeled protein kinase proteins revealed that GST-p57 KIP2 protein directly interacted with 35 S-labeled JNK1 and SAPK␤/ JNK3 but not with 35 S-labeled ERK2 or p38 ( Fig. 2A). Next we examined the physical association of p57 KIP2 and JNK in HEK293 cells after co-transfecting the cells with plasmids encoding p57 KIP2 and HA-JNK1 (Fig. 2B). The transfected HEK293 cells were subjected to immunoprecipitation with anti-HA antibody, and the resulting immunoprecipitates were subjected to immunoblot analysis with anti-p57 KIP2 antibody. The immunoblot data showed that HA-JNK1 physically associated with p57 KIP2 in the transfected cells. We also examined the physical interaction between endogenous p57 KIP2 and endogenous JNK1 in HeLa cells. HeLa cell lysates were subjected to immunoprecipitation with the use of anti-p57 KIP2 antibody. Immunoblot analysis using anti-JNK1 antibody of the p57 KIP2 immunoprecipitates revealed that endogenous JNK1 physically interacted with endogenous p57 KIP2 (Fig. 2C).
Next, we examined the effect of p57 KIP2 on the interaction between JNK1 and c-Jun, a substrate of JNK1, in intact cells. Co-immunoprecipitation data showed that the physical interaction between JNK1 and c-Jun markedly decreased in the cells expressing p57 KIP2 (Fig. 3). These results thus suggest that 57 KIP2 , by binding JNK1, inhibits an interaction between JNK1 and c-Jun.
p57 KIP2 Suppresses JNK-involved Apoptosis-The JNK/ SAPK signaling pathway has been shown to be involved in the mechanism of apoptosis (9,18,19). We therefore tested whether ectopic p57 KIP2 could suppress the JNK-involved apoptosis (Fig. 5). First, we examined the effect of p57 KIP2 on apoptosis induced by expression of ⌬MEKK1, a constitutively active mutant of MEKK1 (Fig. 5A). Apoptotic cell death was enhanced in HEK293 cells expressing ⌬MEKK1, and ⌬MEKK1-induced apoptosis was reduced when SEK1(K129R) was co-expressed. These data suggest that ⌬MEKK1 induces apoptotic cell death through the MEKK1-SEK1-JNK signaling cascade. ⌬MEKK1-induced apoptosis was suppressed in cells expressing ectopic p57 KIP2 . Because p57 KIP2 does not inhibit either MEKK1 or SEK1 activity (Fig. 1), these results suggest that JNK inhibition is the major mechanism by which p57 KIP2 suppresses JNK-mediated apoptosis. Next, we examined the effect of p57 KIP2 on UV-induced apoptosis. The DAPI staining data indicated that exposure of HEK293 cells to UV light resulted in a marked increase in apoptotic cell death (Fig. 5B). The UV-induced apoptosis was decreased in cells expressing SEK1(K129R), a dominant-negative mutant of SEK1, suggesting that the JNK signaling cascade was involved in the mechanism of UV-induced apoptosis. We found that cells transfected with plasmid encoding p57 KIP2 became more resistant to UV radiation than cells transfected with control plasmid. TUNEL staining data showed similar results (Fig. 5C).
Depletion of Endogenous p57 KIP2 Enhances Kinase Activity of Endogenous JNK in Intact Cells-Recently it has been shown that expression of p57 KIP2 is enhanced during C2C12 myoblast differentiation (20), which can be induced by withdrawing serum from the culture medium. We therefore tested whether the induction of p57 KIP2 expression by serum withdrawal would result in suppression of JNK activity in C2C12 myoblast cells. We confirmed a dramatic induction of p57 KIP2 expression in C2C12 cells after reducing the serum concentration from 20% (growth medium) to 2% (differentiation medium; data not shown). We then measured JNK1 activity in undifferentiated and differentiated C2C12 cells (Fig. 6). Our data show that the UV-stimulated JNK1 activity was inhibited in C2C12 cells in differentiation medium, compared with C2C12 cells in growth medium (Fig. 6A). The immunoblot data revealed the induction of p57 KIP2 expression in the cells in differentiation medium. To further examine whether p57 KIP2 induction was responsible for the inhibition of JNK activity during C2C12 cell differentiation, we constructed C2C12 cells that were stably transfected with p57 KIP2 antisense vector. Expression of p57 KIP2 protein was markedly reduced in the C2C12 cells harboring the p57 KIP2 antisense compared with control cells that were stably transfected with the empty vector alone (Fig. 6B). Our data thus indicate that suppression of p57 KIP2 expression prevents the inhibitory action of p57 KIP2 on UV-stimulated JNK1 activity.
To further investigate the biological relevance of the inhibitory action of p57 on JNK activity, we examined the UV-stimulated activity of endogenous JNK1 in MEF cells derived from p57 ϩ/ϩ and p57 Ϫ/Ϫ mice. The JNK1 activity at the UVstimulated state was higher in MEF p57(Ϫ/Ϫ) cells than in MEF p57(ϩ/ϩ) cells (Fig. 7). Thus, these results suggest that p57 KIP2 is an endogenous inhibitor of JNK/SAPK. DISCUSSION We have demonstrated here that p57 KIP2 can modulate stress-activated signaling through negatively regulating the JNK/SAPK pathway. p57 KIP2 appears to suppress the JNK signaling pathway by targeting JNK. p57 KIP2 did not affect other MAPK family members such as ERK and p38. p57 KIP2 physically associates with and inhibits JNK in intact cells. Thus we show in this study that p57 KIP2 functions as a natural inhibitory protein of JNK/SAPK, independently of its inhibitory action on CDKs.
p57 KIP2 contains the NH 2 -terminal CDK binding domain, the PAPA domain, and the carboxyl-terminal QT domain (4,5). p57 shares sequence homology with other members of the p21 family, p21 and p27, in the NH 2 -terminal CDK binding domain (4,5). The carboxyl-terminal QT domain of p57 KIP2 also exhibits sequence similarity to that region in p27. Although the function of the QT domain of p57 KIP2 remains unclear, this domain is thought to be involved in protein-protein interaction (5). Our present data show that a carboxyl-terminal p57 KIP2 fragment containing the QT domain is crucial for the inhibition of JNK/SAPK. Interestingly, p27, which also contains the QT domain in its carboxyl-terminal region, failed to inhibit JNK/ SAPK activity in vitro (data not shown). The difference be- A, p57 KIP2 inhibits MEKK1-dependent apoptosis. HEK293 cells were transfected for 48 h with ⌬MEKK1, p57 KIP2 , or SEK1(K129R) along with pEGFP. The transfected cells were stained with DAPI, and the DAPI-stained nuclei were examined for apoptotic morphology by fluorescence microscopy. GFP-expressing cells were scored for apoptotic nuclei with a Zeiss Axiovert microscope. B and C, p57 KIP2 inhibits UV-induced apoptosis. HEK293 cells were transfected with plasmid vectors encoding p57 KIP2 or SEK1(K129R) along with pEGFP. After 40 h, the transfected cells were exposed to 60 J/m 2 UV light and incubated further for 10 h. The cells were then analyzed for apoptotic cell death by DAPI staining (B) or TUNEL staining (C). tween the actions of p57 KIP2 and p27 on JNK/SAPK may be due to a relatively low level (about 35%) of amino acid sequence identity between these QT domains. In humans, the p57 KIP2 gene is located in chromosome 11p15.5, a region implicated in both sporadic cancers and Beckwith-Wiedemann syndrome, a familial cancer syndrome (5). Among patients with Beckwith-Wiedemann syndrome, several mutations in the p57 KIP2 gene have been detected that include deletions in the QT domain of p57 KIP2 (21,22). A deletion in the QT domain of p57 KIP2 may possibly help JNK/SAPK escape from negative regulation by p57 KIP2 . The resultant increase in JNK/SAPK activity may thus contribute to oncogenic transformation (9,23).
Previous studies using p57 KIP2 knockout mice reported an increase in apoptosis and altered differentiation during mouse development (6,7). p57 KIP2 protein also plays a role in the regulation of myoblast differentiation and apoptotic processes (20,24). It has been suggested that JNK/SAPK signaling pathway may be involved in the regulatory mechanism of apoptosis and differentiation (9,10,25,26). It is plausible, therefore, to conclude that JNK/SAPK inhibition could be an important mechanism by which p57 KIP2 exerts its functions on apoptosis and differentiation. Further studies are needed to verify this possibility. On the basis of our data, we suggest a novel function for p57 KIP2 , that it may regulate stress-activated signals by suppressing the JNK/SAPK signaling pathway.
FIG. 6. Induction of p57 KIP2 expression results in decreased JNK activity during C2C12 myoblast differentiation. A, For differentiation, C2C12 myoblast cells were moved from growth medium containing 20% fetal bovine serum to differentiation medium containing 2% fetal bovine serum where they were maintained for 2 days. Where indicated, C2C12 cells were exposed to UV light (60 J/m 2 ) and further incubated for 1 h. Cells were then harvested, lysed, and assayed for JNK1 activity by immune complex kinase assay. The cell lysates were also analyzed by immunoblot with either rabbit anti-p57 KIP2 or mouse anti-JNK1 antibody. B, C2C12 cells were stably transfected with pcDNA3 empty vector or pcDNA3 containing a full-length p57 KIP2 antisense cDNA. Stably transfected cells were selected by G418 (500 g/ml). Transfected cells were induced to differentiate by being maintaining in differentiation medium for 2 days. The cells were then exposed to UV light, as indicated, and assayed for JNK1 activity as described for panel A. These data represent the results from three independent experiments. FIG. 7. JNK/SAPK activity is higher in MEF p57؊/؊ cells than in MEF p57؉/؉ cells. MEF cells derived from p57 ϩ/ϩ and p57 Ϫ/Ϫ mice were unexposed or exposed to UV light (60 J/m 2 ) and then incubated further for 1 h. Cell lysates were subjected to immunoprecipitation with anti-JNK1 antibody. The resulting precipitates were examined for JNK1 activity by immune complex kinase assay. Cell lysates were also immunoblotted (IB) with anti-JNK1 and anti-p57 antibodies. These data represent the results from three independent experiments.