DNA Damage-induced MDMX Degradation Is Mediated by MDM2
2003; Elsevier BV; Volume: 278; Issue: 46 Linguagem: Inglês
10.1074/jbc.m308295200
ISSN1083-351X
AutoresHidehiko Kawai, Dmitri Wiederschain, Hiroyuki Kitao, Jeremy Stuart, Kelvin K. Tsai, Zhi-Min Yuan,
Tópico(s)Ubiquitin and proteasome pathways
ResumoAlthough genetic studies have demonstrated that MDMX is essential to maintain p53 activity at low levels in non-stressed cells, it is unknown whether MDMX regulates p53 activation by DNA damage. We show here that DNA damage-induced p53 induction is associated with rapid down-regulation of the MDMX protein. Significantly, interference with MDMX down-regulation results in the suppression of p53 activation by genotoxic stress. We also demonstrate that DNA damage-induced MDMX reduction is mediated by MDM2, which targets MDMX for proteasomal degradation by a distinct mechanism that permits preferential MDMX degradation and therefore ensures optimal p53 activation. Although genetic studies have demonstrated that MDMX is essential to maintain p53 activity at low levels in non-stressed cells, it is unknown whether MDMX regulates p53 activation by DNA damage. We show here that DNA damage-induced p53 induction is associated with rapid down-regulation of the MDMX protein. Significantly, interference with MDMX down-regulation results in the suppression of p53 activation by genotoxic stress. We also demonstrate that DNA damage-induced MDMX reduction is mediated by MDM2, which targets MDMX for proteasomal degradation by a distinct mechanism that permits preferential MDMX degradation and therefore ensures optimal p53 activation. The tumor suppressor gene p53 encodes a transcription factor that controls the expression of a number of genes, the products of which mediate either cell cycle arrest or apoptosis (1Levine A.J. Cell. 1997; 88: 323-331Abstract Full Text Full Text PDF PubMed Scopus (6727) Google Scholar). Because of this growth inhibitory activity, under most physiological conditions, p53 is kept at low levels through rapid protein degradation, which ensures cell survival and proper organ development. p53 turnover is mediated primarily by MDM2, an E3 1The abbreviations used are: E3, ubiquitin-protein isopeptide ligase; MEF, mouse embryo fibroblasts; DKO, double knock-out; GFP, green fluorescent protein; EGFP, enhanced GFP; RT, reverse transcriptase; PBS, phosphate-buffered saline; NES, nuclear export sequence; NLS, nuclear localization sequence; RNAi, RNA interference; RFD, Ring finger domain; Gy, grays.1The abbreviations used are: E3, ubiquitin-protein isopeptide ligase; MEF, mouse embryo fibroblasts; DKO, double knock-out; GFP, green fluorescent protein; EGFP, enhanced GFP; RT, reverse transcriptase; PBS, phosphate-buffered saline; NES, nuclear export sequence; NLS, nuclear localization sequence; RNAi, RNA interference; RFD, Ring finger domain; Gy, grays. ubiquitin ligase, which targets p53 for ubiquitin-dependent proteasomal proteolysis, although non-proteasomal degradation may also play a role under certain conditions (2Vousden K.H. Biochim. Biophys. Acta. 2002; 1602: 47-59Crossref PubMed Scopus (302) Google Scholar). At the same time, MDM2 itself is a transcriptional target of p53, thus forming a negative feedback loop in which p53 controls the expression of its own negative regulator. The current model of p53 activation suggests that diverse stress signals converge on a single regulatory node, namely the p53-MDM2 module, and interfere with the ability of MDM2 to target p53 for degradation. Available evidence indicates that inhibition of MDM2-mediated p53 degradation can be achieved through two major mechanisms: either by preventing MDM2 from binding to p53 via DNA damage-induced phosphorylation or by functional inactivation of MDM2 through oncogene-induced expression of p14ARF (2Vousden K.H. Biochim. Biophys. Acta. 2002; 1602: 47-59Crossref PubMed Scopus (302) Google Scholar). MDMX is a close structural homologue of MDM2 (3Shvarts A. Steegenga W.T. Riteco N. van Laar T. Dekker P. Bazuine M. van Ham R.C. van der Houven van Oordt W. Hateboer G. van der Eb A.J. Jochemsen A.G. EMBO J. 1996; 15: 5349-5357Crossref PubMed Scopus (517) Google Scholar). Analogous to MDM2, it binds to p53 and inhibits its transactivation function. In contrast to MDM2, however, MDMX lacks ubiquitin E3 ligase activity in cells and is unable to target p53 for ubiquitin/proteasome-dependent proteolysis (4Stad R. Little N.A. Xirodimas D.P. Frenk R. van der Eb A.J. Lane D.P. Saville M.K. Jochemsen A.G. EMBO Rep. 2001; 2: 1029-1034Crossref PubMed Scopus (186) Google Scholar). Interestingly, MDMX ablation is associated with p53-dependent embryonic death in mice (5Parant J. Chavez-Reyes A. Little N.A. Yan W. Reinke V. Jochemsen A.G. Lozano G. Nat. Genet. 2001; 29: 92-95Crossref PubMed Scopus (412) Google Scholar, 6Finch R.A. Donoviel D.B. Potter D. Shi M. Fan A. Freed D.D. Wang C.Y. Zambrowicz B.P. Ramirez-Solis R. Sands A.T. Zhang N. Cancer Res. 2002; 62: 3221-3225PubMed Google Scholar, 7Migliorini D. Denchi E.L. Danovi D. Jochemsen A. Capillo M. Gobbi A. Helin K. Pelicci P.G. Marine J.C. Mol. Cell. Biol. 2002; 22: 5527-5538Crossref PubMed Scopus (261) Google Scholar), a phenotype similar to that of MDM2 knock-out, indicating that both MDM2 and MDMX are essential negative regulators of p53 and that neither can substitute for the loss of the other. MDM2 is an extremely unstable protein since it is capable of self-ubiquitination that leads to rapid proteasomal degradation (8Honda R. Yasuda H. Oncogene. 2000; 19: 1473-1476Crossref PubMed Scopus (313) Google Scholar). Regulation of MDM2 stability might therefore be another important mechanism by which p53 activity can be controlled. Indeed, we have recently shown that MDMX binds to and increases the stability of MDM2, thus directly contributing to the ability of MDM2 to maintain p53 at low levels (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar). We demonstrate here that MDMX contributes to the regulation of DNA damage-induced p53 activation. In addition, our study identifies MDMX as a novel substrate of MDM2. Interestingly, MDM2 utilizes distinct mechanisms to target MDMX and p53 for degradation, therefore exerting differential control over the levels of these two proteins in response to DNA damage. Plasmid Design—p53, MDM2, MDMX, and MDM2 mutants and IκBα expression plasmids have been described previously (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar, 12Kawai H. Wiederschain, D. Yuan Z.M. Mol. Cell. Biol. 2003; 23: 4939-4947Crossref PubMed Scopus (105) Google Scholar, 15Kawai H. Nie L. Yuan Z.M. Mol. Cell. Biol. 2002; 22: 6079-6088Crossref PubMed Scopus (54) Google Scholar). pBS-U6/p53 and pBS-U6/MDM2 RNAi constructs were generated essentially as described by Sui et al. (16Sui G. Soohoo C. Affar E.B. Gay F. Shi Y. Forrester W.C. Shi Y. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 5515-5520Crossref PubMed Scopus (1060) Google Scholar) using the following targeting sequences: p53 (5′-gggcttcttgcattctgggac-3′) and MDM2 (5′-gggagtgatcaaaaggacctt-3′). MDMXRNAi was described previously (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar). Cell Culture and Transfection—MCF-7 cells (American Type Culture Collection), wild type MEFs, p53 -/- MEFs, and double knock-out (DKO) MEFs (p53 -/- MDM2 -/- MEFs) were maintained in minimal essential medium (Cellgro) supplemented with 10% fetal bovine serum. HCT116 and HCT116-p53 -/- cells were maintained in McCoy's 5A medium (Cellgro). Cells were transfected by LipofectAMINE 2000 (Invitrogen) method according to the manufacturer's instructions. Preparation of Whole Cell Extracts and Western Blot Analysis—Cells were transfected in 60-mm plates with 5 μg of DNA and harvested at 24 h after transfection. Cell lysates were prepared as described previously (12Kawai H. Wiederschain, D. Yuan Z.M. Mol. Cell. Biol. 2003; 23: 4939-4947Crossref PubMed Scopus (105) Google Scholar). Protein concentrations were determined using Bio-Rad protein assay (Bio-Rad). After the addition of 5× loading buffer, the samples were incubated at 95 °C for 5 min and resolved by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes (Schleicher & Schuell) and probed with the following antibodies as indicated: anti-MDMX (6B1A), anti-FLAG (M5; Sigma), anti-p53 (Ab-6; Oncogene), anti-MDM2 (Ab-1; Oncogene), anti-MDM2 (Ab-3; Oncogene), anti-p21 (187; Santa Cruz Biotechnology), anti-actin (AC-15; Sigma), and anti-GFP (Clontech). Proteins were visualized with an enhanced chemiluminescence detection system (PerkinElmer Life Sciences). Preparation of Total RNA and Quantitative RT-PCR—Total RNA was isolated from MCF-7 cells using Trizol/chloroform extraction as directed by the manufacturer's instructions (Invitrogen). RNA was then subjected to DNase treatment (DNAfree; Ambion) and purified according to the manufacturer's instructions using the RNeasy mini kit (Qiagen). Quantitative RT-PCR was performed in two-step reactions: First, cDNA templates were created in final reaction conditions containing 2 μg of total RNA, 500 ng of oligo(dT), 100 ng of random hexamers, 0.5 mm dNTPs, 100 mm dithiothreitol, and 200 units of Superscript reverse transcriptase (Invitrogen). Quantitative PCR reactions were performed using the QuantiTect SYBR Green PCR kit (Qiagen) according to the manufacturer's instructions using specific primers (p53, 5′-aggccttggaactcaaggat-3′ and 5′-tgagtcaggcccttctgtct-3′; MDMX, 5′-agacgatatccccacactgc-3′ and 5′-tcccactcctcaaatccaag-3′; γ-actin, 5′-tccttcctgggtatggaatc-3′ and 5′-caccgtgttggcgtacag-3′). For each set of primer combinations, a standard curve was created to ensure similar efficiencies between PCR reactions. Each standard curve contained a "minus RT" reaction to ensure that there was no DNA contamination in the cDNA templates. 100 ng of cDNA template was used in the optimized PCR reactions for determining relative expression levels between MDMX and p53. Quantitative PCR reactions were carried out using the Opticon monitor thermocycler and analyzed using corresponding software (MJ Research, Inc.). Expression levels of MDMX and p53 following IR treatment were calculated relative to the γ-actin housekeeping gene. Cell Cycle Analysis—Cells were harvested by trypsinization, and after washing in phosphate-buffered saline (PBS), the cells were fixed in 70% ethanol. Following PBS washes, cells were stained with propidium iodide solution (50 μg/ml) containing 100 μg/ml RNase A at 37 °C for 1 h, and DNA content of 10,000 cells was analyzed using Beckman-Coulter ELITE flow cytometer (Kresge Center for Environmental Health Sciences, Harvard School of Public Health) and Multicycle cell cycle analysis software (Phoenix Flow Systems). Retrovirus Infection—Phoenix Ampho Φ cells were transfected with pMSCV-FLAG MDMX (10 μg), pCG-gagpol (5 μg), and pCG-VSVG (1 μg) (Dr. R. Mulligan, Harvard Medical School) by calcium-phosphate method as described previously (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar). Retroviral supernatant was harvested 48 h after transfection. HCT116 and HCT116-p53 -/- cells were infected by incubation with retroviral supernatants and polybrene (4 μg/ml) for 24 h followed by selection in puromycin (1.5 μg/ml)-containing media for 5 days. Immunoprecipitation Analysis—Immunoprecipitations were performed as described elsewhere (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar). Cell lysates were prepared in 0.5% Triton X-100 lysis buffer and incubated with anti-MDMX antibody (p55 and p56) and Protein G PLUS-Agarose (Santa Cruz Biotechnology) for 12 h. Immune complexes and whole lysates were analyzed by Western blotting. The filters were incubated with anti-MDMX (6B1A). Subcellular Distribution Assay—Cells were grown on Chamber Slides (Nunc) and transfected with the indicated FLAG-tagged vector. Cells were washed with PBS 24 h after transfection and fixed with 4% paraformaldehyde (Sigma) for 30 min at 4 °C. After washing with PBS, cells were permeabilized with ice-cold 0.2% Triton X-100 for 5 min, blocked with 0.5% bovine serum albumin for 30 min, and then incubated with the indicated antibody for 1 h. The slides were incubated with secondary antibody (Texas-Red X goat anti-mouse IgG, Molecular Probes) and 4′,6-diamidino-2-phenylindole (10 μg/ml, Sigma). Following PBS wash, the slides were mounted with Fluoromount-G (Southern Biotechnology Associates) containing 2.5 mg/ml n-propyl gallate (Sigma). Specimens were examined under the fluorescent microscope (Olympus). Contribution of MDMX to DNA Damage-induced p53 Activation—Having recently identified mutual dependence of MDM2 and MDMX in their functional inactivation of p53 in non-stressed cells (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar), we asked whether MDMX plays any role in DNA damage-induced p53 activation when the ability of MDM2 to target p53 for degradation is inhibited. Response of MCF-7 cells to IR was analyzed by Western blot using the indicated antibodies (Fig. 1A). As expected, p53 protein levels were readily induced by IR followed by the induction of p21 and MDM2 (Fig. 1A). In sharp contrast to the p53 induction, IR caused a marked decrease in endogenous MDMX levels over time. Significant and dose-dependent down-regulation of MDMX was further confirmed as a novel molecular event in response to IR in a separate experiment, as shown in Fig. 1B. To determine whether this decrease in MDMX abundance was due to either decreased protein synthesis or increased protein degradation, proteasome inhibitor MG132 was added to the culture immediately following IR treatment. Measurement of cellular levels of MDMX revealed that the addition of proteasome inhibitor almost completely blocked the IR-induced reduction in MDMX protein (Fig. 1C). Together with the failure to detect any significant effect of the dose of IR used on MDMX mRNA level (Fig. 1E), these results suggest that IR induces MDMX down-regulation via an increased proteolysis. Interestingly, in contrast to p53 accumulation, the addition of proteasome inhibitor to non-stressed cells did not result in detectable alteration in MDMX levels (Fig. 1D), suggesting that MDMX degradation is triggered by DNA damage. To substantiate this notion, we determined the half-life of MDMX in MCF-7 cells exposed to IR, as well as in mock-treated cells. We utilized cyclohexamide to inhibit de novo protein synthesis and monitored the disappearance of MDMX, MDM2, and p53 by Western blot analysis at the indicated time points (Fig. 1F). Although MDMX is a rather stable protein in non-stressed cells, the half-life of MDMX was markedly shortened in IR-treated cells, indicating that a decrease in MDMX levels is the result of IR-induced protein degradation. Notably, the shortened half-life of MDM2 and prolonged half-life of p53 tightly correlated with IR-induced MDMX degradation (Fig. 1F). Since various forms of genotoxic stress can activate p53, there are likely to be multiple modes of p53 induction. To determine whether down-regulation of MDMX is specific for IR or reflects a general response to DNA damage, we tested the response of MCF-7 cells to either UV irradiation (Fig. 1G, lanes 1-4) or adriamycin (Fig. 1G, lanes 5-8) treatment. Western analysis of the cell lysates prepared at the indicated times after treatment revealed significant reductions in MDMX levels in response to both UV and adriamycin treatment, although it was slightly delayed when compared with IR-treated cells. Once again, MDMX down-regulation was associated with p53 activation. Analogous to IR-induced MDMX degradation, the proteasome inhibitor blocked MDMX disappearance in UV- and adriamycin-treated cells (Fig. 1G, lanes 9-16). Strong correlation between p53 induction and MDMX decrease in response to IR prompted us to ask whether this MDMX down-regulation is important to p53 activation. To address this, we transiently expressed exogenous MDMX to counteract MDMX decline induced by DNA damage. Western analysis demonstrated that p53 activation in response to IR was severely compromised in MDMX-expressing cells, as evidenced by the significantly diminished induction of p53 target gene products p21 and MDM2 (Fig. 2A, lanes 4-6). To test whether this MDMX-dependent suppression of p53 activation can translate into impaired biological response, DNA damage-induced G1 cell cycle arrest was analyzed. Comparison of the S-phase population between control and IR-treated cells indicated that sustained MDMX expression resulted in impaired G1 arrest induced by IR (Fig. 2B). To further examine the MDMX-dependent p53 regulation, MCF-7 cells deficient in MDMX expression were generated by RNAi-mediated gene silencing. Consistent with our previous results (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar), inhibition of MDMX expression was associated with partial p53 activation, as demonstrated by elevated p53 and p21 levels in RNAi transfected cells (Fig. 2C, lane 4). Importantly, IR exposure resulted in greater p53 activation in MDMX-deficient cells than in control cells (Fig. 2C, lanes 2 and 3 versus lanes 5 and 6). Cell cycle analysis further demonstrated that MDMX deficiency is associated with p53 functional activation (Fig. 2D). IR-induced MDMX Down-regulation Is Mediated by MDM2—Having shown that exposure of MCF-7 cells to DNA damage agents results in the increased degradation of the MDMX protein, we went on to examine the underlying mechanism. Since our results demonstrate that MDMX is degraded through the proteasomal pathway (Fig. 1), and given the fact that MDMX binds to MDM2, which possesses an E3 ubiquitin ligase activity, we asked whether MDM2 could target MDMX for degradation. MCF-7 cells deficient in MDM2 expression were generated using RNAi. Since MDM2 expression is transcriptionally controlled by p53, efficient reduction in MDM2 levels required the use of both p53RNAi and MDM2RNAi. We first tested each RNAi for its ability to inhibit the expression of the corresponding gene and evaluated its effect on p53 activity. RNAi-mediated suppression of either MDMX (Fig. 3A, lane 2) or MDM2 (Fig. 3A, lane 4) expression resulted in p53 activation as evidenced by the induction of p21. In contrast, p53RNAi effectively reduced endogenous p53 levels, leading to down-regulation of p21 and MDM2 (Fig. 3A, lane 3). Combination of MDM2RNAi and p53RNAi inhibited the expression of both genes (Fig. 3A, lane 5). Significantly, MDMX levels remained constant following IR treatment in cells transfected with both MDM2RNAi and p53RNAi (Fig. 3B, lanes 4-6, top panel). To determine whether p53 contributes to the IR-induced MDMX reduction, we utilized p53-wild type or p53-null HCT116 cell lines retrovirally transduced with MDMX that had comparable levels of MDM2 under non-stressed conditions. Our results demonstrate that IR-induced MDMX degradation is p53-independent as both p53-wild type and p53-null HCT116 cells exhibited almost identical decrease in MDMX in response to IR (Fig. 3C, lanes 1-3 versus lanes 4-6). To further verify that MDMX levels are controlled by MDM2, we analyzed IR-induced response in p53-null HCT116 cells in which MDM2 expression was inhibited by MDM2RNAi. Western analysis indicated that deficiency in MDM2 was indeed associated with the resistance of MDMX to IR-induced degradation (Fig. 3D). Finally, MEFs isolated from wild type, p53 knock-out, or p53-MDM2 DKO mice were used to confirm the results obtained in RNAi experiments. As shown in Fig. 3E, IR exposure resulted in marked MDMX reduction in wild type and p53 -/- MEFs, but not in DKO MEFs (Fig. 3E), thus confirming that IR-induced MDMX down-regulation depends on MDM2, but not on p53. We next tested whether MDM2, which has an E3 ligase activity, could target MDMX for ubiquitination/degradation. To accomplish this, MDMX-expressing plasmid was co-transfected with either control vector or MDM2 expressing vector into DKO MEFs. Plasmids encoding p53 and IκBα were included as controls. Western analysis of cellular protein abundance at 24 h after transfection indicated that, as expected, MDM2 targeted p53, but not IκBα, for efficient degradation (Fig. 4A, lanes 1 and 2 versus lanes 5 and 6). Significantly, coexpression of MDM2 also resulted in a significant reduction in MDMX protein abundance (Fig. 4A, lanes 3 and 4). Anti-Actin and anti-GFP blots demonstrate that the observed difference in protein levels was not a result of unequal loading or transfection efficiency. Since the E3 ligase activity of MDM2 is essential to target p53 for degradation, we proceeded to test whether the Ring domain mutant of MDM2 (MDM2RM, C464A), which lacks the E3 ligase activity, is capable of down-regulating MDMX. Our data show that the E3 ligase-deficient MDM2 mutant completely lost its ability to reduce MDMX levels (Fig. 4B, lane 2 versus lane 3). To further analyze the MDM2-targeted MDMX degradation, we asked whether the steady state levels of MDMX could be affected by MDM2. To this end, the half-life of MDMX was determined in the presence of either wild type MDM2 or the Ring domain mutant of MDM2. We monitored the disappearance of MDMX as well as MDM2 by Western blot analysis in cyclohexamide-treated cells. DKO MEFs expressing the indicated vectors were analyzed at 0, 30, 60, 120, and 180 min following the addition of cyclohexamide. Substantially prolonged half-life of MDMX in the MDM2RM-expressing cells when compared with the wild type MDM2-transfected cells (Fig. 4C) supports an E3 ligase-dependent regulation of MDMX protein stability by MDM2. MDM2 is a shuttling protein that contains both the nuclear localization sequence (NLS) and the nuclear export sequence (NES). Our previous finding that MDMX is in fact a cytoplasmic protein which redistributes into the nucleus upon binding to MDM2 (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar, 10Migliorini D. Danovi D. Colombo E. Carbone R. Pelicci P.G. Marine J.C. J. Biol. Chem. 2001; 13: 13Google Scholar) led us to ask whether subcellular distribution plays a role in the MDM2-targeted MDMX degradation. MDM2 mutants deficient in either nuclear import or nuclear export were tested for their ability to degrade MDMX. Although mainly cytoplasmically localized when expressed alone, coexpression with MDM2 mutants caused MDMX to colocalize with the MDM2 protein either in the nucleus (MDM2ΔNES) or in the cytoplasm (MDM2ΔNLS) (Fig. 4D). Western analysis revealed that coexpression of either MDM2 mutant was associated with a marked decrease in cellular MDMX levels (Fig. 4E), regardless of distribution, indicating that the MDM2-mediated MDMX degradation is independent of its subcellular localization. The inverse correlation between DNA damage-induced p53 induction and MDMX down-regulation suggests the possibility that MDM2 might target MDMX for degradation by a mechanism that is distinct from p53 degradation. Since recent studies have shown that the Ring finger domain (RFD) of MDM2 is necessary but not sufficient for p53 degradation (11Fang 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 (868) Google Scholar), we asked whether it was also the case for MDMX down-regulation. Having recently demonstrated that the E3 ligase activity of MDM2 is transferable to MDMX via fusion of the MDM2RFD into MDMX backbone (12Kawai H. Wiederschain, D. Yuan Z.M. Mol. Cell. Biol. 2003; 23: 4939-4947Crossref PubMed Scopus (105) Google Scholar), we assessed the ability of MDM2RFD- or MDM2ZnRFD-bearing chimeric proteins to degrade MDMX. Consistent with our recent data (12Kawai H. Wiederschain, D. Yuan Z.M. Mol. Cell. Biol. 2003; 23: 4939-4947Crossref PubMed Scopus (105) Google Scholar), these MDMX-MDM2 chimeras were unable to target p53 for degradation as no detectable reduction in p53 levels was seen in chimera-expressing cells, whereas p53 levels were significantly diminished in the wild type MDM2-expressing cells (Fig. 5A, lane 2 versus lanes 3 and 4). Interestingly, the MDMX protein was markedly decreased by coexpression with MDM2 as well as with MDMX-MDM2 chimeras (Fig. 5A, lane 5 versus lanes 6-8). Consistent with the involvement of the zinc motif in binding to MDMX (13Sharp D.A. Kratowicz S.A. Sank M.J. George D.L. J. Biol. Chem. 1999; 274: 38189-38196Abstract Full Text Full Text PDF PubMed Scopus (237) Google Scholar), the MDMX-MDM2ZnRFD exhibited greater potency in MDMX degradation than MDMX chimera that contained only the RFD of MDM2 (Fig. 5A, lane 7 versus lane 8). In a parallel experiment, the acidic domain deletion mutant of MDM2 (MDM2Δ222-302) was tested for its ability to target MDMX for degradation. Wild type MDM2 and MDM2RM were included as controls. Measurement of cellular MDMX abundance revealed that MDM2Δ222-302 was just as efficient as wild type MDM2 in targeting MDMX for degradation, whereas the MDM2RM had no effect on MDMX levels (Fig. 5B). Similarly, ubiquitinated forms of MDMX were observed in cells expressing either wild type MDM2 or MDM2Δ222-302, but not MDM2RM (Fig. 5B, top panel). Finally, to test whether the MDM2ZnRFD is sufficient for targeting MDMX for degradation, plasmid encoding the corresponding sequence of MDM2 was coexpressed with MDMX. Our data show that this E3 ligase containing mini-protein is indeed sufficient to degrade MDMX, albeit with reduced potency (Fig. 5C, lane 2 versus lane 3), whereas not having any effect on p53 levels (Fig. 5C, lane 5 versus lane 6). Taken together, our results demonstrate that the C terminus of MDM2 is necessary and sufficient to target MDMX for ubiquitin/proteasome-dependent proteolysis. Available evidence indicates that inhibition of MDM2-mediated p53 degradation is essential for adequate p53 activation in response to various forms of genotoxic stress. Two major pathways of p53 activation have been well documented: phosphorylation-dependent disruption of p53-MDM2 complex upon DNA damage and oncogene-induced MDM2 inactivation by p14ARF. We demonstrate here that an additional layer of regulation of p53 activation in response to stress is present in the intricate interplay between MDMX and MDM2. Biochemical analysis has demonstrated that MDMX expression is necessary to render MDM2 sufficiently stable so that it can function at its full potential (9Gu J. Kawai H. Nie L. Kitao H. Wiederschain D. Jochemsen A.G. Parant J. Lozano G. Yuan Z.M. J. Biol. Chem. 2002; 277: 19251-19254Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar). In non-stressed cells, MDMX forms heterocomplex with MDM2, resulting in enhanced MDM2 protein stability and thus ensuring low p53 levels. Upon DNA damage, as our data show, MDMX is rapidly degraded via a MDM2-dependent mechanism, which leads to optimal p53 activation. Indeed, restoration of MDMX expression suppresses DNA damage-induced p53 activation, whereas down-regulation of MDMX is associated with p53 activation. It is particularly interesting that the DNA damage-induced MDMX down-regulation is mediated by MDM2 that targets MDMX for ubiquitin/proteasome-dependent proteolysis. However, MDM2 directs p53 and MDMX into the degradation pathway through distinct mechanisms. Although sufficient for MDMX degradation, the Ring domain of MDM2 is necessary but not sufficient for p53 destruction. This difference may represent a key mechanism by which DNA damage differentially modulates MDM2-mediated p53 and MDMX degradation. Our results demonstrate that in contrast to the significantly prolonged half-life of p53 in IR-treated cells, MDMX is rapidly degraded by MDM2 following stress. Therefore, MDMX down-regulation inversely correlates with p53 induction, thus ensuring rapid and maximal p53 activation. Although the importance of phosphorylation in DNA damage-induced p53 activation has been well documented, it has also been shown that a phosphorylation-resistant mutant of p53 can still be readily induced by DNA damage (14Ashcroft M. Kubbutat M.H. Vousden K.H. Mol. Cell. Biol. 1999; 19: 1751-1758Crossref PubMed Scopus (378) Google Scholar), suggesting the presence of an alternative mechanism for p53 activation. The intricate interplay between MDM2 and MDMX, as demonstrated in this study, represents a novel mode of DNA damage-induced p53 activation. We are grateful to Dr. A. Jochemsen, Leiden University, for the generous gift of anti-MDMX antibodies; Dr. Y. Shi, Harvard Medical School, for providing the pBS-U6 vector; Dr. B Vogelstein, The Johns Hopkins University, for HCT116 and HCT116-p53 -/-; Dr. Garry Nolan, Stanford University, for the gift of Phoenix Ampho Φ cells; Claire Bailey, the Center for Genomics Research at Harvard University, for training and use of the Opticon Monitor for quantitative PCR and analysis of data; Amy Imrich, Harvard School of Public Health, for help with flow cytometry analysis. The excellent technical assistance of Colleen Dionne is acknowledged.
Referência(s)