MEK-ERK-mediated Phosphorylation of Mdm2 at Ser-166 in Hepatocytes

Mdm2 inactivates the tumor suppressor p53 and Akt has been shown to be a major activator of Mdm2 in many cell types. We have investigated the regulation of Mdm2 in hepatocytes. We found that growth factor-induced Ser-166 phosphorylation of Mdm2 was inhibited by the MEK inhibitors U0126 and PD98059 in HepG2 cells and in a rat liver cell line, TRL 1215. Also, bile acids and oxidative stress induced phosphorylation of Mdm2 at Ser-166 by an apparently MEK-ERK-dependent mechanism. In contrast, Ser-166 phosphorylation of Mdm2 in lung cells was mediated by Akt. Further studies revealed that phosphatidylinositol 3-kinase inhibitors LY294002 and wortmannin induced phosphorylated ERK Tyr-204 and pMdm2 Ser-166 phosphorylations in hepatocytes in culture and in rat hepatocytes in vivo. In HepG2 cells, this effect was inhibited by U0126 and PD98059. LY294002 also reduced the level of pRaf Ser-259. Furthermore, we have shown that myr-Akt-induced overexpression of pAkt suppressed the levels of pMdm2 Ser-166 in hepatocytes. These data indicate a reversed relationship between Akt and Mdm2 in hepatocytes and suggest that Akt is a negative regulator of Raf-MEK-ERK-Mdm2 in this cell type. Ser-166 phosphorylation of Mdm2 has been shown to increase its ubiquitin ligase activity and increase p53 degradation, and our data indicated an attenuated p53 response to DNA damage in hepatocytes exhibiting high levels of pMdm2 Ser-166. Taken together, our data indicate that Mdm2 phosphorylation is regulated via MEK-ERK in hepatocytes. This Mdm2 signaling might be important for the regeneration of hepatocytes after centrilobular cell death. Mdm2 inactivates the tumor suppressor p53 and Akt has been shown to be a major activator of Mdm2 in many cell types. We have investigated the regulation of Mdm2 in hepatocytes. We found that growth factor-induced Ser-166 phosphorylation of Mdm2 was inhibited by the MEK inhibitors U0126 and PD98059 in HepG2 cells and in a rat liver cell line, TRL 1215. Also, bile acids and oxidative stress induced phosphorylation of Mdm2 at Ser-166 by an apparently MEK-ERK-dependent mechanism. In contrast, Ser-166 phosphorylation of Mdm2 in lung cells was mediated by Akt. Further studies revealed that phosphatidylinositol 3-kinase inhibitors LY294002 and wortmannin induced phosphorylated ERK Tyr-204 and pMdm2 Ser-166 phosphorylations in hepatocytes in culture and in rat hepatocytes in vivo. In HepG2 cells, this effect was inhibited by U0126 and PD98059. LY294002 also reduced the level of pRaf Ser-259. Furthermore, we have shown that myr-Akt-induced overexpression of pAkt suppressed the levels of pMdm2 Ser-166 in hepatocytes. These data indicate a reversed relationship between Akt and Mdm2 in hepatocytes and suggest that Akt is a negative regulator of Raf-MEK-ERK-Mdm2 in this cell type. Ser-166 phosphorylation of Mdm2 has been shown to increase its ubiquitin ligase activity and increase p53 degradation, and our data indicated an attenuated p53 response to DNA damage in hepatocytes exhibiting high levels of pMdm2 Ser-166. Taken together, our data indicate that Mdm2 phosphorylation is regulated via MEK-ERK in hepatocytes. This Mdm2 signaling might be important for the regeneration of hepatocytes after centrilobular cell death. p53 suppresses tumor development by activating growth arrest, apoptosis, or DNA repair. In unstressed cells, levels of p53 are maintained low by its antagonist murine double minute 2 (Mdm2). 2The abbreviations used are: Mdm2, murine double minute 2; E3, ubiquitin-protein isopeptide ligase; MAPKAP, mitogen-activated protein kinase-activated protein; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; ERK, extracellular signal-regulated kinase; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; TGFα, transforming growth factor α; myr-Akt, myristoylated-Akt; DEN, diethylnitrosamine; PI3K, phosphatidylinositol 3-kinase. 2The abbreviations used are: Mdm2, murine double minute 2; E3, ubiquitin-protein isopeptide ligase; MAPKAP, mitogen-activated protein kinase-activated protein; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; ERK, extracellular signal-regulated kinase; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; TGFα, transforming growth factor α; myr-Akt, myristoylated-Akt; DEN, diethylnitrosamine; PI3K, phosphatidylinositol 3-kinase. Mdm2 is an E3 ubiquitin ligase and a major regulator of p53, targeting p53 for proteasomal degradation (1Moll U.M. Petrenko O. Mol. Cancer Res. 2003; 1: 1001-1008PubMed Google Scholar, 2Fang S. Jensen J.P. Ludwig R.L. Vousden K.H. Weissman A.M. J. Biol. Chem. 2000; 275: 8945-8951Abstract Full Text Full Text PDF PubMed Scopus (864) Google Scholar, 3Vousden K.H. Lu X. Nat. Rev. Cancer. 2002; 2: 594-604Crossref PubMed Scopus (2703) Google Scholar). Genotoxicity and other stressful stimuli can induce phosphorylations of both Mdm2 and p53 and thus inhibit the ability of Mdm2 to degrade p53. When accumulated, p53 transactivates target genes, including Mdm2, and the proteins interact in an autoregulatory feedback loop (4Levine A.J. Cell. 1997; 88: 323-331Abstract Full Text Full Text PDF PubMed Scopus (6714) Google Scholar, 5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar, 6Momand J. Zambetti G.P. Olson D.C. George D. Levine A.J. Cell. 1992; 69: 1237-1245Abstract Full Text PDF PubMed Scopus (2780) Google Scholar, 7Iwakuma T. Lozano G. Mol. Cancer Res. 2003; 1: 993-1000PubMed Google Scholar, 8Wahl G.M. Cell Death Differ. 2006; 13: 973-983Crossref PubMed Scopus (27) Google Scholar). Mdm2 phosphorylations at Ser-395 and Tyr-394 have been shown to decrease the capacity of Mdm2 to target p53 for degradation. In contrast, phosphorylations of Mdm2 at Ser-166 and Ser-186 activate Mdm2. They increase the E3 ligase activity and increase the degradation of p53 (9Gottlieb T.M. Leal J.F. Seger R. Taya Y. Oren M. Oncogene. 2002; 21: 1299-1303Crossref PubMed Scopus (379) Google Scholar, 10Mayo L.D. Donner D.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603Crossref PubMed Scopus (946) Google Scholar, 11Zhou B.P. Liao Y. Xia W. Zou Y. Spohn B. Hung M. Nat. Cell Biol. 2001; 3: 973-982Crossref PubMed Scopus (783) Google Scholar, 12Ogawara Y. Kishishita S. Obata T. Isazawa Y. Suzuki T. Tanaka K. Masuyama N. Gotoh Y. J. Biol. Chem. 2002; 277: 21843-21850Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar). These phosphorylations have also been associated with an increased nuclear localization of Mdm2 (10Mayo L.D. Donner D.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603Crossref PubMed Scopus (946) Google Scholar). A major regulator of Mdm2, protein kinase B/Akt, phosphorylates Mdm2 at Ser-166 and -186 (10Mayo L.D. Donner D.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603Crossref PubMed Scopus (946) Google Scholar). Akt is an important anti-apoptotic signaling molecule, and the phosphorylation of Mdm2 may serve to protect cells from p53-induced apoptosis (12Ogawara Y. Kishishita S. Obata T. Isazawa Y. Suzuki T. Tanaka K. Masuyama N. Gotoh Y. J. Biol. Chem. 2002; 277: 21843-21850Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar). Also, other kinases have been identified to induce Ser-166 phosphorylation of Mdm2 and increase its ability to target p53 for degradation. Thus, Mdm2 has been shown to be phosphorylated by MAPKAP kinase and dampen the duration of p53 response induced by UV light (13Weber H.O. Ludwig R.L. Morrison D. Kotlyarov A. Gaestel M. Vousden K.H. Oncogene. 2005; 24: 1965-1972Crossref PubMed Scopus (50) Google Scholar). MEK-ERK signaling has also been shown to control levels of Mdm2 by regulating the mRNA export to the cytoplasm in some cell types (14Phelps M. Phillips A. Darley M. Blaydes J.P. J. Biol. Chem. 2005; 280: 16651-16658Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar, 15Phillips A. Jones C.J. Blaydes J.P. FEBS Lett. 2006; 580: 300-304Crossref PubMed Scopus (10) Google Scholar). Recently, we reported that the dioxin model compound 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (16Pääjärvi G. Viluksela M. Pohjanvirta R. Stenius U. Högberg J. Carcinogenesis. 2005; 26: 201-208Crossref PubMed Scopus (59) Google Scholar) and cholesterol-lowering drugs, statins, induced Ser-166 phosphorylation of Mdm2 and an attenuated p53 response in hepatocytes (17Pääjärvi G. Roudier E. Crisby M. Högberg J. Stenius U. FASEB J. 2005; 19: 476-478Crossref PubMed Scopus (46) Google Scholar). Furthermore, we provided evidence for an involvement of an Akt-independent but mammalian target of rapamycin-ERK-dependent signaling in the statin-induced Ser-166 phosphorylation (17Pääjärvi G. Roudier E. Crisby M. Högberg J. Stenius U. FASEB J. 2005; 19: 476-478Crossref PubMed Scopus (46) Google Scholar). Our findings raised questions about the liver-specific regulation of Mdm2 and p53. Considering the role of the liver in xenobiotic metabolism and the DNA reactivity of many intermediary metabolites formed in the liver (5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar), it may be speculated that there are safeguarding mechanisms that protects hepatocytes from excess apoptosis. From this perspective, we have characterized the Mdm2 regulation further. Our data indicate that growth factors and toxic stress induce Ser-166 phosphorylation of Mdm2 via MEK-ERK signaling in hepatocytes. Furthermore, an up-regulation of Akt down-regulated Mdm2 in hepatocytes. Cell Lines and Reagents—Human hepatocellular carcinoma cells, HepG2 cells, and human lung carcinoma cells, A549 cells, were purchased from the American Type Culture Collection. TRL 1215 cells were provided as a generous gift from Dr. Michael P. Waalkes from the National Cancer Institute. This non-tumorigenic cell line was originally derived from the livers of 10-day-old Fischer F344 rats (18Idoine J.B. Elliott J.M. Wilson M.J. Weisburger E.K. In Vitro. 1976; 12: 541-553Crossref PubMed Scopus (56) Google Scholar). HepG2 cells were grown on collagen-coated dishes in minimal essential medium with Earle's salts and l-glutamine. Minimal essential medium was also supplemented with sodium pyruvate (1 mm), non-essential amino acids, penicillin/streptomycin, and 10% inactivated fetal bovine serum. The HepG2 cells were serum-starved (0.5% serum) 48 h before exposure. A549 cells were grown in Dulbecco's modified Eagle's medium with glucose (4500 mg/liter) and l-glutamine. Dulbecco's modified Eagle's medium was also supplemented with sodium pyruvate (1 mm), penicillin/streptomycin, and 10% inactivated fetal bovine serum. The A549 cells were serum-starved (0.1% serum) 24 h before exposure. TRL 1215 cells were grown in Williams' medium E with Glutamax. The medium was also supplemented with penicillin/streptomycin and 10% inactivated fetal bovine serum. The TRL 1215 cells were serum-starved (no serum) 24 h before exposure. LY294002, wortmannin, rapamycin, sodium deoxycholate, chenodeoxycholic acid, anisomycin, 5-fluorouracil, benzo-(a) pyrene, and etoposide were purchased from Sigma-Aldrich and dissolved in Me2SO. U0126 and PD98059 were purchased from Cell Signaling Technology (Beverly, MA) and also dissolved in Me2SO. The Me2SO concentration on the dishes was <0.2%. Transforming growth factor α (TGFα) was from Nordic Biosite (Täby, Sweden), insulin and H2O2 from Merck (Darmstadt, Germany), and leptomycin B from Calbiochem (Darmstadt, Germany). Small Interference RNA Transfection—HepG2 cells were transfected the day after plating by using SignalSilence Akt small interference RNA (Cell Signaling Technology) and TransIT-TKO transfection reagent (Mirus Bio Corporation, Madison, WI). The cells were transfected for 48 h according to the manufacturer's protocol. The medium was changed to low serum medium (containing 0.5% serum) 24 h before exposure. Transfection of HepG2 Cells with Myristoylated-Akt (myr-Akt)—For transfection, LNCX plasmid encoding myr-Akt, which is a constitutive active form of Akt, was used. Plasmid encoding myr-Akt was a generous gift from Raphael A. Nemenoff from the University of Colorado Health Science Center (Denver, CO). Myr-Akt plasmid was amplified and purified by using the Qiagen Maxi kit (Qiagen, West Sussex, UK). Twenty-four hours after plating, HepG2 cells were transfected for 48 h with Lipofectin (Invitrogen) using 1 or 2 μg of DNA of the myr-Akt plasmid or an empty vector, according to the manufacturer's protocol. Western Blot Analysis—The cells were washed with phosphate-buffered saline and lysed in IPB7 (1 mm phenylmethylsulfonyl fluoride, 1 mg/ml leupeptin, 1 mg/ml pepstatin, 1 mm NaF, 1 mm NaVO3, 0.1 mg/ml trypsin inhibitor, and 1 mg/ml aprotinin). Rats were injected intraperitoneally with diethylnitrosamine (DEN), 1.32 mmol/kg body weight (Sigma-Aldrich), 4 or 24 h before sacrifice. The livers were homogenized and subfractionated. The protein was quantificated by using Coomassie Plus, the better Bradford assay kit (Pierce). The samples were subjected to SDS-PAGE. The separated proteins were transferred to a polyvinylidene difluoride membrane (Bio-Rad), and the protein bands were subsequently probed with antibodies. Primary antibodies used in the analyses were phospho-Mdm2 (Ser-166), Mdm2, phospho-p53 (Ser-46), phospho-Akt (Ser-473), phospho-Raf (Ser-259) (Cell Signaling Technology), Mdm2 (4B11), Mdm2 (4B2) (gifts from Dr. A. Levine), pAkt1/2/3 (Ser-473)-R catalog number sc 7985-R, Akt1 (B-1) sc-5298, p-ERK (E-4) sc-7383, ERK1 (K-23) sc-94, actin (C-11)-R sc1615, Cdk2 (M2) sc-163, p21 (F-5) sc-6246 (Santa Cruz Biotechnology, Santa Cruz, CA), and p53 CM-1 (Novocastra, Newcastle, UK). Secondary antibodies used in the analyses were goat anti-rabbit IgG-horseradish peroxidase sc-2004, goat anti-mouse IgG-horseradish peroxidase sc-2005 (Santa Cruz Biotechnology), and anti-rabbit IgG horseradish peroxidase P0217 (DAKO, Glostrup, Denmark). The results were visualized by the ECL detection kit (Amersham Biosciences AB, Uppsala, Sweden). Cdk2, used as a loading control, did not vary between different treatments and correlated with actin, total ERK, and total Mdm2. The results were analyzed with NIH Image version 1.62 software and ImageJ version 1.34s software. Statistical Analysis—Statistical analysis was performed using Student's t test. The statistical analysis was based on at least three different experiments, and the results were considered to be statistically significant when p < 0.05. Immunohistochemical Staining—Female Sprague-Dawley rats were injected intraperitoneally with LY294002 (5 mg/kg body weight) or wortmannin (15 μg/kg body weight) 90 min before sacrifice or with DEN (0.99 mmol/kg body weight) 3 or 48 h before sacrifice. Livers were fixed and slices were stained as described previously (5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar). In brief, to achieve a rapid fixation, the livers were perfused in 3.7% buffered formaldehyde for 1.5 h and subsequently fixed for 24 h. The sections were incubated overnight with primary antibodies, phospho-Mdm2 (Ser-166) (Cell Signaling Technology), p-ERK (E-4) catalog number sc-7383 (Santa Cruz Biotechnology), and p53 (Ab-3) (Calbiochem). Primary antibodies were visualized using the EnVision+™ peroxidase kit (DAKO). No staining was detected when the primary antibodies were omitted. All animals received humane care, and the experimental protocol was approved by the Swedish Board for Laboratory Animals and was in accordance with National Institutes of Health guidelines. Akt has been shown to activate Mdm2 in many cell types by inducing phosphorylation at Ser-166 and -186 (9Gottlieb T.M. Leal J.F. Seger R. Taya Y. Oren M. Oncogene. 2002; 21: 1299-1303Crossref PubMed Scopus (379) Google Scholar, 10Mayo L.D. Donner D.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603Crossref PubMed Scopus (946) Google Scholar, 11Zhou B.P. Liao Y. Xia W. Zou Y. Spohn B. Hung M. Nat. Cell Biol. 2001; 3: 973-982Crossref PubMed Scopus (783) Google Scholar, 12Ogawara Y. Kishishita S. Obata T. Isazawa Y. Suzuki T. Tanaka K. Masuyama N. Gotoh Y. J. Biol. Chem. 2002; 277: 21843-21850Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar). In the present study, we have investigated the effect of Akt activation on Mdm2 phosphorylation in HepG2 cells. We found that insulin phosphorylated Akt at Ser-473 (pAkt Ser-473) and Mdm2 at Ser-166 (pMdm2 Ser-166) in serum-starved HepG2 cells (Fig. 1A). The phosphatidylinositol 3-kinase (PI3K) inhibitor LY294002 reduced the insulin-induced pAkt Ser-473 without affecting pMdm2 Ser-166 (Fig. 1A). The same result was obtained by using another PI3K inhibitor, wortmannin (not shown), or by silencing Akt with small interference RNA (Fig. 1B). These results suggested that the phosphorylation of Mdm2 at Ser-166 was not mediated by PI3K/Akt in HepG2 cells. We tested the involvement of MEK by using the MEK inhibitors PD98059 and U0126. Data presented in Fig. 1A show that PD98059 and U0126 inhibited the insulin-induced pMdm2 Ser-166. As expected, MEK inhibitors effectively reduced levels of phosphorylated ERK (pERK) (Fig. 1A). Fig. 1C shows a densitometric analysis of pMdm2 Ser-166 data from three different experiments. Although the effect of insulin on pERK Tyr-204 was weak in the experiment shown in Fig. 1A, the densitometric analysis of results from three experiments shows that pERK Tyr-204 levels increased compared with untreated controls (123 ± 15%, ratio pERK Tyr-204:Cdk2 after 1 h with insulin). Total Mdm2, Akt, or ERK were not affected (Fig. 1A). Next, we tested the effect of TGFα. We found that TGFα induced the same response as insulin in HepG2 cells. Thus, PI3K inhibitors LY294002, wortmannin, and a mammalian target of rapamycin inhibitor, rapamycin, did not inhibit the TGFα-induced pMdm2 Ser-166, although the MEK inhibitors U0126 and PD98059 did so (Fig. 1D). The phosphorylation of ERK, the downstream target of MEK, was completely inhibited by these inhibitors (Fig. 1D). Together, these findings suggest that growth factor-induced phosphorylation of Mdm2 at Ser-166 in HepG2 cells was mediated by the MEK-ERK pathway. We also analyzed levels of total Mdm2, as it has been indicated that MEK-ERK can control nuclear export of Mdm2 mRNA, at least in some cell types (14Phelps M. Phillips A. Darley M. Blaydes J.P. J. Biol. Chem. 2005; 280: 16651-16658Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar, 15Phillips A. Jones C.J. Blaydes J.P. FEBS Lett. 2006; 580: 300-304Crossref PubMed Scopus (10) Google Scholar). However, Mdm2 levels were not altered by the MEK inhibitors (Fig. 1A) under our experimental conditions. The non-small cell lung cancer cell line A549 was used as a control. In this cell line, we found that insulin induced pAkt Ser-473 and pMdm2 Ser-166 and that the PI3K inhibitor LY294002 inhibited these effects (Fig. 1E). In this case, inhibition of MEK by PD98059 or U0126 did not affect insulin-induced Ser-166 phosphorylation. These results thus deviated from those obtained with HepG2 cells and indicated an involvement of Akt in pMdm2 Ser-166 formation in lung cells. These data are in line with previous publications (9Gottlieb T.M. Leal J.F. Seger R. Taya Y. Oren M. Oncogene. 2002; 21: 1299-1303Crossref PubMed Scopus (379) Google Scholar, 10Mayo L.D. Donner D.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603Crossref PubMed Scopus (946) Google Scholar, 11Zhou B.P. Liao Y. Xia W. Zou Y. Spohn B. Hung M. Nat. Cell Biol. 2001; 3: 973-982Crossref PubMed Scopus (783) Google Scholar, 19Feng J. Tamaskovic R. Yang Z. Brazil D.P. Merlo A. Hess D. Hemmings B.A. J. Biol. Chem. 2004; 279: 35510-35517Abstract Full Text Full Text PDF PubMed Scopus (219) Google Scholar) showing an Akt-mediated phosphorylation of Mdm2 at Ser-166 in many cell types. As indicated in Fig. 1, increases in pMdm2 Ser-166 levels were paralleled by increases in pERK levels. To further investigate the role of ERK in the phosphorylation of Mdm2, known ERK-activating stimuli were tested. Rosseland et al. (20Rosseland C.M. Wierød L. Oksvold M.P. Werner H. Østvold A.C. Thoresen G.H. Paulsen R.E. Huitfeldt H.S. Skarpen E. Hepatology. 2005; 42: 200-207Crossref PubMed Scopus (58) Google Scholar) have recently shown that oxidative stress phosphorylates ERK in hepatocytes. To investigate whether oxidative stress also phosphorylates Mdm2, we treated HepG2 cells with H2O2. It was found that H2O2 induced pERK Tyr-204, pAkt Ser-473, as well as pMdm2 Ser-166 in a concentration-dependent manner in HepG2 cells (Fig. 2A). Elevated levels of pERK and pMdm2 were registered for at least 2 h (Fig. 2B); whereas the Akt phosphorylation was more transient (data not shown). Fig. 2C shows the densitometric analysis of H2O2-induced pMdm2 and pERK in four experiments. Importantly, pretreatment with the MEK inhibitor PD98059 abolished the H2O2-induced phosphorylation of Mdm2 at Ser-166 (Fig. 2D). The MEK inhibitor U0126 also inhibited the H2O2-induced pMdm2 Ser-166 (data not shown). This indicates that oxidative stress also induced pMdm2 via MEK-ERK in HepG2 cells. In A549 cells, H2O2 induced pERK and pAkt but not pMdm2 (Fig. 2E). Bile acids have been shown to activate ERK (21Dent P. Fang Y. Gupta S. Studer E. Mitchell C. Spiegel S. Hylemon P.B. Hepatology. 2005; 42: 1291-1299Crossref PubMed Scopus (95) Google Scholar). As shown in Fig. 3A, both deoxycholic acid and chenodeoxycholic acid induced pERK levels, which correlated with increased pMdm2 levels. Also, this effect was inhibited by U0126, whereas levels of total ERK and Mdm2 were unaffected (Fig. 3B).FIGURE 3Bile acids and anisomycin induce phosphorylation of Mdm2 at Ser-166 and phosphorylation of ERK at Tyr-204 in HepG2 cells. HepG2 cells were incubated with deoxycholic acid (200 μm) or with chenodeoxycholic acid (200 μm) for 20 min (A). In B, HepG2 cells were preincubated with PD98059 (20 μm) or U0126 (10 μm) for 1 h and 20 min and thereafter incubated with deoxycholic acid (200 μm) for 20 min. C, HepG2 cells were incubated with anisomycin (25 ng/ml) for the times indicated. The samples were analyzed by Western blotting using antibodies against pMdm2 Ser-166, Mdm2, pERK Tyr-204, ERK1, and pAkt Ser-473. Cdk2 was used as a loading control.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Anisomycin has been shown to induce phosphorylation of Mdm2 at Ser-166 in a MAPKAP-dependent manner (13Weber H.O. Ludwig R.L. Morrison D. Kotlyarov A. Gaestel M. Vousden K.H. Oncogene. 2005; 24: 1965-1972Crossref PubMed Scopus (50) Google Scholar, 22Hazzalin C.A. Cano E. Cuenda A. Barratt M.J. Cohen P. Mahadevan L.C. Curr. Biol. 1996; 6: 1028-1031Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar). We found that anisomycin induced pMdm2 and pERK (but not pAkt) in HepG2 cells (Fig. 3C). In A549 cells, anisomycin induced pAkt and pMdm2 but not pERK (data not shown). PI3K phosphorylates Akt and its inhibitors decreased pAkt levels in growth factor-stimulated cells without inhibiting pMdm2 Ser-166 (Fig. 1, A and D). Next, we investigated whether constitutive levels of pMdm2 Ser-166 were affected by PI3K inhibitors. To our surprise, we found that LY294002 and wortmannin increased pMdm2 levels (Fig. 4, A and E). The increase in pMdm2 levels in HepG2 cells was preceded by decreased levels of pAkt and increased levels of pERK, whereas total ERK, Mdm2, and Akt were unaffected (Fig. 4A). Densitometric analysis of results from three experiments shows that pMdm2 Ser-166 levels were significantly increased after 15 min of exposure (Fig. 4B). The peak pERK level was seen within 2 h (data not shown). The LY294002-induced pMdm2 Ser-166 was inhibited by the MEK inhibitor PD98059 (Fig. 4C). Raf can regulate ERK activity (23Rommel C. Clarke B.A. Zimmermann S. Nuñez L. Rossman R. Reid K. Moelling K. Yancopoulos G.D. Glass D.J. Science. 1999; 286: 1738-1741Crossref PubMed Scopus (661) Google Scholar, 24Zimmermann S. Moelling K. Science. 1999; 286: 1741-1744Crossref PubMed Scopus (905) Google Scholar, 25Reusch H.P. Zimmermann S. Schaefer M. Paul M. Moelling K. J. Biol. Chem. 2001; 276: 33630-33637Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar, 26Myhre O. Sterri S.H. Bogen I.L. Fonnum F. Toxicol. Sci. 2004; 80: 296-303Crossref PubMed Scopus (20) Google Scholar), and we found that LY294002 decreased the inactivating Ser-259 phosphorylation of Raf (Fig. 4D). These data are in line with a previously discussed cross-talk between Akt and Raf-ERK (23Rommel C. Clarke B.A. Zimmermann S. Nuñez L. Rossman R. Reid K. Moelling K. Yancopoulos G.D. Glass D.J. Science. 1999; 286: 1738-1741Crossref PubMed Scopus (661) Google Scholar, 24Zimmermann S. Moelling K. Science. 1999; 286: 1741-1744Crossref PubMed Scopus (905) Google Scholar, 25Reusch H.P. Zimmermann S. Schaefer M. Paul M. Moelling K. J. Biol. Chem. 2001; 276: 33630-33637Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar, 26Myhre O. Sterri S.H. Bogen I.L. Fonnum F. Toxicol. Sci. 2004; 80: 296-303Crossref PubMed Scopus (20) Google Scholar). The data presented here extend previously published data and suggest a cross-talk regulating Mdm2 activation in hepatocytes. In A549 cells, the levels of pMdm2 were unaffected or decreased by LY294002 and wortmannin, whereas pERK levels increased by LY294002 (Fig. 4F). Insulin was used as a positive control (Fig. 4, E and F). To study these effects in another liver cell line, TRL 1215 cells were employed. As shown in Fig. 5A, LY294002 increased pMdm2 and pERK levels in these cells as well. This effect was inhibited by the MEK inhibitor PD98059 (Fig. 5B). Also, growth factor-induced pMdm2 Ser-166 was inhibited by MEK inhibitors in TRL 1215 cells (data not shown). Next, the effect of Akt overexpression was studied. In non-starved myr-Akt-overexpressing cells, the levels of pMdm2 Ser-166 and pERK Tyr-204 were suppressed (Fig. 6). Interestingly, a dose-effect relationship was indicated, so that increasing doses of myr-Akt gave decreasing levels of pMdm2.FIGURE 5PI3K inhibitor LY294002 induces MEK-mediated phosphorylation of Mdm2 at Ser-166 in TRL 1215 liver cell line. A, TRL 1215 cells were incubated with LY294002 (25 μm) for the times indicated. B, TRL 1215 cells were preincubated with PD98059 (20μm) for 30 min and thereafter incubated with LY294002 (25 μm) for 90 min. The samples were analyzed by Western blotting using antibodies against pMdm2 Ser-166, Mdm2, pERK Tyr-204, ERK1, pAkt Ser-473, and Akt. Cdk2 was used as a loading control.View Large Image Figure ViewerDownload Hi-res image Download (PPT)FIGURE 6Myr-Akt transfection in HepG2 cells increases Akt expression and the phosphorylation of Akt at Ser-473 but decreases the phosphorylation of Mdm2 at Ser-166. Non-starved HepG2 cells were transfected for 48 h with control plasmid or with plasmid for myr-Akt (1 or 2 μg). The samples were analyzed by Western blotting using antibodies against pMdm2 Ser-166, Mdm2, Akt, pAkt Ser-473, pERK Tyr-204, and ERK1. Cdk2 was used as a loading control.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The effect of PI3K inhibitors on phosphorylated Mdm2 and ERK was studied in vivo in rat liver. In control livers, no or very weak nuclear staining for pMdm2 Ser-166 or pERK Tyr-204 was detected (Fig. 7, A and D). Ninety minutes after an intraperitoneal injection of LY294002 (5 mg/kg) or wortmannin (15 μg/kg), a nuclear staining for pERK in hepatocytes was markedly increased (Fig. 7, E and F). There was also an increased staining in other cell types than just hepatocytes. No clear zonal distribution was observed. Both treatments also induced nuclear staining for pMdm2 Ser-166, affecting a major part of the lobule including centrilobular and midzonal areas (Fig. 7, B and C). Of interest is that this increase only affected hepatocytes. For example, non-hepatocytes stained for pERK Tyr-204 were not stained for pMdm2 Ser-166 (Fig. 7, B, C, E and F). These results corroborate the data obtained with cell lines and indicate that PI3K inhibitors induced phosphorylation of Mdm2 at Ser-166 in hepatocytes in vivo. Other cell types in the liver were not affected. We have previously described the p53/Mdm2 autoregulatory loop and the induction of apoptosis in rat liver in response to DNA-damaging agents (5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar). We subsequently tested the possible involvement of pMdm2 Ser-166 in this response. As shown in Fig. 8, treatment with DEN induced nuclear staining for pMdm2 Ser-166 in hepatocytes in the midzonal area already within 3 h (Fig. 8A). The zonal distribution of pMdm2 Ser-166-positive hepatocytes seen at 3 h was the same as the previously described zonal distribution of cytoplasmic and nuclear staining for Mdm2 seen in the control liver (5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar). At 24 h, the pMdm2 Ser-166 staining correlated with the staining seen for p53; that is, the pMdm2 Ser-166-positive hepatocytes roughly overlapped the p53-positive cells (data not shown). However, 48 h after DEN administration, almost all p53-positive hepatocytes had disappeared (Fig. 8C), whereas several centrilobular hepatocytes were positive for pMdm2 Ser-166 (Fig. 8B). Seventy-two hours after DEN treatment, no clear nuclear staining for p53 was detected, but still single pMdm2-positive cells were seen close to the central vein (not shown). Liver samples were also analyzed employing Western blotting. A clear pMdm2 Ser-166 response was seen 4 and 24 h after DEN treatment (Fig. 8D). This confirms the histological study. Mdm2 phosphorylation at Ser-166 has been shown to enhance p53 degradation (11Zhou B.P. Liao Y. Xia W. Zou Y. Spohn B. Hung M. Nat. Cell Biol. 2001; 3: 973-982Crossref PubMed Scopus (783) Google Scholar, 12Ogawara Y. Kishishita S. Obata T. Isazawa Y. Suzuki T. Tanaka K. Masuyama N. Gotoh Y. J. Biol. Chem. 2002; 277: 21843-21850Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar). Therefore, we studied the effects of LY294002-induced pMdm2 on the p53 response to DNA damage. We also analyzed the downstream protein p21. Serum-starved HepG2 cells were incubated for 21 h with agents selected to stabilize p53 and p21 through variable mechanisms. We found that pretreatment with LY294002 attenuated the p53 response induced by all three genotoxic compounds 5-fluorouracil, benzo(a)pyrene, and etoposide (Fig. 9A). This was associated with attenuated p21 response for 5-fluorouracil and benzo(a)pyrene at that time point (Fig. 9A). Similar results were obtained with leptomycin B, which is a nuclear export inhibitor and a potent inducer of p53 accumulation without causing DNA damage. (Fig. 9A) (27Kudo N. Wolff B. Sekimoto T. Schreiner E.P. Yoneda Y. Yanagida M. Horinouchi S. Yoshida M. Exp. Cell Res. 1998; 242: 540-547Crossref PubMed Scopus (706) Google Scholar). Benzo(a)pyrene also induced phosphorylation of p53 at Ser-46, which has been implicated in apoptosis responses (28Mayo L.D. Seo Y.R. Jackson M.W. Smith M.L. Rivera Guzman J. Korgaonkar C.K. Donner D.B. J. Biol. Chem. 2005; 280: 25953-25959Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). Pretreatment with LY294002 strongly inhibited this phosphorylation of p53 (Fig. 9A). As indicated in Fig. 4, A–C, the pMdm2 Ser-166 response was rapid. In Fig. 9B, it is shown that 2 h of incubation with LY294002 (in combination with leptomycin B) induced Mdm2 phosphorylation. We also confirmed the more short term effect of LY294002 on Ser-166 phosphorylation in etoposide-exposed HepG2 cells. As expected, LY294002 induced pMdm2 within 3 h, and this effect was inhibited by MEK inhibitors (Fig. 9C). In A549 cells, LY294002 decreased basal levels of pMdm2 (Fig. 9D). A recent paper shows that LY294002 prevents p53 induction by DNA-damaging agents (cisplatin, camptothecin, and 5-fluorouracil) in mouse fibroblasts and some human cell types. Cells were cultured under rich serum conditions, which stimulated PI3K (29Bar J. Lukaschuk N. Zalcenstein A. Wilder S. Seger R. Oren M. Cell Death Differ. 2005; 12: 1578-1587Crossref PubMed Scopus (35) Google Scholar). This apparently PI3K-dependent effect was reported to be independent of ERK, Akt, and mammalian target of rapamycin, and neither Mdm2 phosphorylations nor hepatocytes were studied. We conclude that the effects studied here in hepatocytes differ from those studied by Bar et al. (29Bar J. Lukaschuk N. Zalcenstein A. Wilder S. Seger R. Oren M. Cell Death Differ. 2005; 12: 1578-1587Crossref PubMed Scopus (35) Google Scholar). Thus, we used starved cells, and we found that MEK inhibitors blocked a pERK phosphorylation and inhibited an Mdm2 phosphorylation in hepatocytes but not in lung cells. In this study, we have shown that bile acids, H2O2, genotoxic compounds such as the liver carcinogen DEN, and PI3K inhibitors all induce Ser-166 phosphorylation of Mdm2 in hepatocytes. We also have provided evidence that this phosphorylation is mediated by MEK-ERK. This effect was registered in HepG2 cells and in TRL 1215 cells. The effect was also seen in rat hepatocytes but not in non-hepatocytes in vivo. By using lung cells under similar culturing conditions, we have confirmed previous studies showing that Ser-166 phosphorylation of Mdm2 is mediated by Akt in many cell types. In hepatocytes, the Mdm2 phosphorylation attenuated the p53 response, and our data suggest that MEK-ERK-Mdm2 can limit p53 downstream effects. Bile acids and oxidative stress have been shown to induce ERK activation in hepatocytes (20Rosseland C.M. Wierød L. Oksvold M.P. Werner H. Østvold A.C. Thoresen G.H. Paulsen R.E. Huitfeldt H.S. Skarpen E. Hepatology. 2005; 42: 200-207Crossref PubMed Scopus (58) Google Scholar, 21Dent P. Fang Y. Gupta S. Studer E. Mitchell C. Spiegel S. Hylemon P.B. Hepatology. 2005; 42: 1291-1299Crossref PubMed Scopus (95) Google Scholar). In the present study, we have provided evidence that toxic stress induced by deoxycholic acid, chenodeoxycholic acid, and oxidative stress accumulated pMdm2 Ser-166 via MEK-ERK signaling. Thus, the specific MEK1 and MEK2 inhibitors, U0126 and PD98059, inhibited the effects in HepG2 cells. In contrast, in A549, non-small cell lung cancer cell H2O2 induced ERK phosphorylation but did not affect pMdm2 Ser-166 levels. Furthermore, in this cell type, Mdm2 phosphorylated on Ser-166 accumulated in response to insulin in an Akt-dependent manner. Surprisingly, we also found that the PI3K inhibitors induced Ser-166 phosphorylation of Mdm2 in HepG2 cells via MEK-ERK. Thus, LY294002-induced pMdm2 Ser-166 was completely inhibited by the MEK inhibitor PD98059. Similar effects were also seen in TRL 1215 cells. These data clearly document an increased Mdm2 phosphorylation in hepatocytes, even though Akt was down-regulated. Furthermore, transfection of cells with constitutively active myr-Akt decreased Mdm2 phosphorylation. We conclude that, in many cell types, pAkt has been shown to phosphorylate and activate Mdm2 (9Gottlieb T.M. Leal J.F. Seger R. Taya Y. Oren M. Oncogene. 2002; 21: 1299-1303Crossref PubMed Scopus (379) Google Scholar, 10Mayo L.D. Donner D.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603Crossref PubMed Scopus (946) Google Scholar, 11Zhou B.P. Liao Y. Xia W. Zou Y. Spohn B. Hung M. Nat. Cell Biol. 2001; 3: 973-982Crossref PubMed Scopus (783) Google Scholar, 12Ogawara Y. Kishishita S. Obata T. Isazawa Y. Suzuki T. Tanaka K. Masuyama N. Gotoh Y. J. Biol. Chem. 2002; 277: 21843-21850Abstract Full Text Full Text PDF PubMed Scopus (472) Google Scholar), whereas data presented here indicate that Akt rather inactivates Mdm2 in hepatocytes. It is perhaps important to add that we used cell lines with wild-type p53 function in this study; it is not known how non-functional p53 influences this regulation. The activation of the MEK-ERK pathway by PI3K inhibitors can be explained by a cross-talk between Akt and Raf. Thus, high levels of pAkt can inhibit Raf-1-MEK-ERK signaling by inactivating Raf-1 (30Baccarini M. FEBS Lett. 2005; 579: 3271-3277Crossref PubMed Scopus (107) Google Scholar), and our data show that LY294002 might activate Raf. This cross-talk has been documented in vitro in several cell types (23Rommel C. Clarke B.A. Zimmermann S. Nuñez L. Rossman R. Reid K. Moelling K. Yancopoulos G.D. Glass D.J. Science. 1999; 286: 1738-1741Crossref PubMed Scopus (661) Google Scholar, 24Zimmermann S. Moelling K. Science. 1999; 286: 1741-1744Crossref PubMed Scopus (905) Google Scholar, 25Reusch H.P. Zimmermann S. Schaefer M. Paul M. Moelling K. J. Biol. Chem. 2001; 276: 33630-33637Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar, 26Myhre O. Sterri S.H. Bogen I.L. Fonnum F. Toxicol. Sci. 2004; 80: 296-303Crossref PubMed Scopus (20) Google Scholar) including the Hep3B cell line (31Wang Z. Wang M. Carr B.I. J. Cell. Physiol. 2005; 203: 510-519Crossref PubMed Scopus (10) Google Scholar). It has also been shown that an inhibition of Akt by LY294002 up-regulates B-Raf in a human kidney epithelial cell line (32Guan K.L. Figueroa C. Brtva T.R. Zhu T. Taylor J. Barber T.D. Vojtek A.B. J. Biol. Chem. 2000; 275: 27354-27359Abstract Full Text Full Text PDF PubMed Google Scholar). Interestingly, our data show that both LY294002 and wortmannin induced increased levels of pERK in rat hepatocytes and in non-parenchymal liver cells in situ, indicating an in vivo relevance of our findings. These results and the dose-effect relationship obtained with transfected myr-Akt suggest that Akt maintains a negative control and modulates Raf-MEK-ERK in quiescent rat liver cells in situ. Many cell types have been employed in studies showing Mdm2 activation by Akt, and hepatocytes seem to be the only cell type, so far, in which ERK-dependent Mdm2 activation is documented. Furthermore, we have shown that pMdm2 Ser-166 accumulated in hepatocytes but not in lung cells or in non-parenchymal liver cells. Both toxicological stress and growth signals were mediated to Mdm2 via MEK-ERK (Fig. 10). In a previous publication (17Pääjärvi G. Roudier E. Crisby M. Högberg J. Stenius U. FASEB J. 2005; 19: 476-478Crossref PubMed Scopus (46) Google Scholar), we showed that a statin-induced pMdm2 Ser-166 accumulation was prevented by MEK-inhibitors as well as by rapamycin, suggesting an involvement of MEK and mammalian target of rapamycin in this response in hepatocytes. It is unknown why pAkt in hepatocytes, even in high levels, was unable to phosphorylate Mdm2 or why pERK was unable to induce pMdm2 Ser-166 in lung cells or non-parenchymal liver cells. It is well established that Mdm2 ablation is embryonically lethal and that the apoptotic phenotype is rescued by p53 knock-out (33Montes de Oca Luna R. Wagner D.S. Lozano G. Nature. 1995; 378: 203-206Crossref PubMed Scopus (1197) Google Scholar). Also Raf-1 knock-out is lethal, and embryonic liver cell apoptosis seems to be a primary event in this early death, at least in some genetic backgrounds (34Galabova-Kovacs G. Matzen D. Piazzolla D. Meissl K. Plyushch T. Chen A.P. Silva A. Baccarini M. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 1325-1330Crossref PubMed Scopus (80) Google Scholar). Further studies have indicated that Raf-1 can prevent apoptosis by at least three different mechanisms (35O'Neill E.E. Matallanas D. Kolch W. Cancer Res. 2005; 65: 5485-5487Crossref PubMed Scopus (46) Google Scholar, 20Rosseland C.M. Wierød L. Oksvold M.P. Werner H. Østvold A.C. Thoresen G.H. Paulsen R.E. Huitfeldt H.S. Skarpen E. Hepatology. 2005; 42: 200-207Crossref PubMed Scopus (58) Google Scholar). However, none of these mechanisms can explain why liver cells are selectively targeted by the Raf-1 knock-out. We provide evidence for a MEK-ERK-Mdm2-dependent mechanism in hepatocytes that may prevent apoptosis. Thus, a down-regulation of Mdm2 may explain why hepatocytes were targeted in Raf-1 knock-out embryos. We have shown here that PI3K-inhibitors induced pMdm2 Ser-166. They also attenuated p53 accumulation following treatment with leptomycin B or DNA-damaging agents in HepG2 cells. This suggests a role for pMdm2 Ser-166 in attenuating the p53 response, although we cannot exclude a direct effect on p53. DEN metabolism is associated with both oxidative stress and DNA adduct formation, and hepatocytes are targeted by DEN toxicity as a consequence of a high expression of xenobiotic metabolizing enzymes (5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar). We have found that pMdm2 Ser-166 was induced in rat liver by DEN treatment, and our results indicate a role for pMdm2 Ser-166 in the p53/Mdm2 response to DNA damage in hepatocytes. We have reported previously on a curtailed p53 response in midzonal hepatocytes upon DEN treatment (5Finnberg N. Silins I. Stenius U. Högberg J. Carcinogenesis. 2004; 25: 113-122Crossref PubMed Scopus (25) Google Scholar). Ser-166 phosphorylation activates Mdm2 (36Brazil D.P. Park J. Hemmings B.A. Cell. 2002; 111: 293-303Abstract Full Text Full Text PDF PubMed Scopus (487) Google Scholar), and further studies may reveal whether the increased levels of pMdm2 Ser-166, seen here in midzonal hepatocytes at 3 and 4 h (earlier than any p53 accumulation can be detected) is related to this curtailed p53 response. Our data thus suggest a role for ERK-Mdm2-p53 signaling in midzonal hepatocytes stressed by xenobiotic metabolism and DNA damages. These cells can replicate in response to centrilobular cell death, and it can be speculated that an ERK-Mdm2 signaling is needed for coordinating this replication under genotoxic stress. Another reasoning suggests that an ERK-Mdm2 signaling is important for hepatocytes to endure chronic stress; it has been reported that chronic ethanol intake is associated with a down-regulated Akt signaling (37He L. Simmen F.A. Mehendale H.M. Ronis M.J. Badger T.M. J. Biol. Chem. 2006; 281: 11126-11134Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar). Our data raise the question whether this chronic down-regulation of pAkt levels may serve to up-regulate Raf-MEK-ERK signaling. In summary, we suggest that hepatocytes respond to low pAkt levels with an up-regulation of Mdm2 phosphorylation at Ser-166 via the MEK-ERK pathway. This ERK-mediated effect on Mdm2 may attenuate the duration and intensity of p53 responses in liver.