Multiple Murine Double Minute Gene 2 (MDM2) Proteins Are Induced by Ultraviolet Light
1999; Elsevier BV; Volume: 274; Issue: 12 Linguagem: Inglês
10.1074/jbc.274.12.8161
ISSN1083-351X
AutoresLeslie Saucedo, Cena D. Myers, Mary Ellen Perry,
Tópico(s)Cancer Research and Treatments
ResumoThe mdm2 (murinedouble minute 2) oncogene encodes several proteins, the largest of which (p90) binds to and inactivates the p53 tumor suppressor protein. Multiple MDM2 proteins have been detected in tumors and in cell lines expressing high levels ofmdm2 mRNAs. Here we show that one of these proteins (p76) is expressed, along with p90, in wild-type andp53-null mouse embryo fibroblasts, indicating that it may have an important physiological role in normal cells. Expression of this protein is induced, as is that of p90, by UV light in a p53-dependent manner. The p76 protein is synthesized via translational initiation at AUG codon 50 and thus lacks the N terminus of p90 and does not bind p53. In cells, p90 and p76 can be synthesized from mdm2 mRNAs transcribed from both the P1 (constitutive) and P2 (p53-responsive) promoters. Site-directed mutagenesis reveals that these RNAs give rise to p76 via internal initiation of translation. In addition, mdm2 mRNAs lacking exon 3 give rise to p76 exclusively, and such mRNAs are induced by p53 in response to UV light. These data indicate that p76 may be an important product of the mdm2 gene and a downstream effector of p53. The mdm2 (murinedouble minute 2) oncogene encodes several proteins, the largest of which (p90) binds to and inactivates the p53 tumor suppressor protein. Multiple MDM2 proteins have been detected in tumors and in cell lines expressing high levels ofmdm2 mRNAs. Here we show that one of these proteins (p76) is expressed, along with p90, in wild-type andp53-null mouse embryo fibroblasts, indicating that it may have an important physiological role in normal cells. Expression of this protein is induced, as is that of p90, by UV light in a p53-dependent manner. The p76 protein is synthesized via translational initiation at AUG codon 50 and thus lacks the N terminus of p90 and does not bind p53. In cells, p90 and p76 can be synthesized from mdm2 mRNAs transcribed from both the P1 (constitutive) and P2 (p53-responsive) promoters. Site-directed mutagenesis reveals that these RNAs give rise to p76 via internal initiation of translation. In addition, mdm2 mRNAs lacking exon 3 give rise to p76 exclusively, and such mRNAs are induced by p53 in response to UV light. These data indicate that p76 may be an important product of the mdm2 gene and a downstream effector of p53. mouse embryo fibroblasts reverse transcription-polymerase chain reaction The mdm2 oncogene is a determinant of embryogenesis (1Montes de Oca Luna R. Wagner D.S. Lozano G. Nature. 1995; 378: 203-206Crossref PubMed Scopus (1196) Google Scholar,2Jones S.N. Roe A.E. Donehower L.A. Bradley A. Nature. 1995; 378: 206-208Crossref PubMed Scopus (1057) Google Scholar), tumorigenesis (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar, 4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar), and cell cycle progression (5Chen C.-Y. Oliner J.D. Zhan Q. Fornace A.J. Vogelstein B. Kastan M.B. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 2684-2688Crossref PubMed Scopus (290) Google Scholar, 6Dubs-Poterszman M.-C. Tocque B. Wasylyk B. Oncogene. 1995; 11: 2445-2449PubMed Google Scholar). The effects of MDM2 on these processes depend, in part, on its ability to inactivate the p53 tumor suppressor protein. For example, mice with a homozygous deletion of the mdm2 gene die during embryogenesis, and this deficiency is rescued if p53is also deleted (1Montes de Oca Luna R. Wagner D.S. Lozano G. Nature. 1995; 378: 203-206Crossref PubMed Scopus (1196) Google Scholar, 2Jones S.N. Roe A.E. Donehower L.A. Bradley A. Nature. 1995; 378: 206-208Crossref PubMed Scopus (1057) Google Scholar). In human tumors, the homologue of MDM2 is overexpressed most often in the subset lacking inactivating mutations in the p53 gene (7Leach F.S. Tokino T. Meltzer P. Burrell M. Oliner J.D. Smith S. Hill D.E. Sidransky D. Kinzler K.W. Vogelstein B. Cancer Res. 1993; 53: 2231-2234PubMed Google Scholar). Thus, high levels of MDM2 may be redundant with mutational inactivation of p53. Following exposure of cells to γ-radiation, MDM2 inactivates the G1 block mediated by p53 (5Chen C.-Y. Oliner J.D. Zhan Q. Fornace A.J. Vogelstein B. Kastan M.B. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 2684-2688Crossref PubMed Scopus (290) Google Scholar). In some cell types, MDM2 can inhibit apoptosis mediated by p53 (8Haupt Y. Barak Y. Oren M. EMBO J. 1996; 15: 1596-1606Crossref PubMed Scopus (204) Google Scholar). MDM2 binds to p53, inhibiting the transcriptional activation function of p53 (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar, 9Momand J. Zambetti G.P. Olson D.C. George D. Levine A.J. Cell. 1992; 69: 1237-1245Abstract Full Text PDF PubMed Scopus (2776) Google Scholar) and stimulating degradation of p53 (10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). Recently, MDM2 has been shown to bind to another tumor suppressor protein, p19ARF(12Zhang Y. Xiong Y. Yarborough W.G. Cell. 1998; 92: 725-734Abstract Full Text Full Text PDF PubMed Scopus (1392) Google Scholar, 13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). The ability of MDM2 to stimulate degradation of p53 is inhibited by p19ARF, indicating that the interaction of p19ARF and MDM2 may be an important determinant of cell cycle progression.In addition to interacting with p53 and p19ARF, MDM2 binds to and alters the function of other proteins that regulate cell division. For example, overexpression of MDM2 stimulates the activity of the E2F transcription factor (14Martin K. Trouche D. Hagemeier C. Sorensen T.S. La Thangue N.B. Kouzarides T. Nature. 1995; 375: 691-694Crossref PubMed Scopus (451) Google Scholar) and reverses an arrest in the cell cycle mediated by either the retinoblastoma protein (15Xiao Z.-X. Chen J. Levine A.J. Modjtahedl N. Xing J. Sellers W.R. Livingston D.M. Nature. 1995; 375: 694-698Crossref PubMed Scopus (573) Google Scholar) or a related protein, p107 (6Dubs-Poterszman M.-C. Tocque B. Wasylyk B. Oncogene. 1995; 11: 2445-2449PubMed Google Scholar). These effects of MDM2 do not depend on the presence of p53. Further evidence for p53-independent functions of MDM2 was provided by Lundgren et al. (4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar), who generated mice expressing high levels of MDM2 in the mammary gland. The mammary epithelial cells of these mice developed polyploidy, and this effect of MDM2 was also seen in mice lacking p53. Mammary tumors arose in mice overexpressing mdm2, but the dependence on p53 was not measured since mice lacking p53 die before mammary tumors would have developed. Sigalis et al. (16Sigalis I. Calvert A.H. Anderson J.J. Neal D.E. Lunec J. Nat. Med. 1996; 2: 912-917Crossref PubMed Scopus (251) Google Scholar) found that alternatively spliced mdm2 mRNAs were overexpressed in some human tumors. Introduction of these mRNAs into NIH3T3 cells resulted in morphological transformation, even though a subset of them encoded MDM2 proteins that could not bind p53. Therefore, some of the oncogenic effects of MDM2 are independent of p53.Several MDM2 proteins are detectable in cells overexpressing MDM2, including human tumor cells and transformed murine cell lines (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar, 18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). Olson et al. (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar) hypothesized that these proteins arise through alternative splicing, proteolytic processing, or post-translational modification. It has not been clear whether these proteins arise through mechanisms used in the normal regulation ofmdm2 expression. Multiple MDM2 proteins can be translated from single mRNAs in rabbit reticulocyte lysates, indicating that internal initiation may be a mechanism whereby multiple MDM2 proteins are expressed in cells (19Barak Y. Gottlieb E. Juven-Gershon T. Oren M. Genes Dev. 1994; 8: 1739-1749Crossref PubMed Scopus (285) Google Scholar). In fact, there is evidence that enhanced translation accounts for the overexpression of multiple MDM2 proteins in some human tumors (18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). In the DM3T3 cell line, in which themdm2 gene is amplified, multiple MDM2 proteins and mRNAs are expressed (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar, 20Fakharzadeh S.S. Trusko S.P. George D.L. EMBO J. 1991; 10: 1565-1569Crossref PubMed Scopus (626) Google Scholar). The most abundant protein in DM3T3 cells (p90) is the well characterized mdm2 gene product that binds to p53 and inhibits its function (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar, 9Momand J. Zambetti G.P. Olson D.C. George D. Levine A.J. Cell. 1992; 69: 1237-1245Abstract Full Text PDF PubMed Scopus (2776) Google Scholar, 10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). The second most abundant protein (p76) does not bind p53 (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar, 19Barak Y. Gottlieb E. Juven-Gershon T. Oren M. Genes Dev. 1994; 8: 1739-1749Crossref PubMed Scopus (285) Google Scholar), but does bind p19ARF (13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). The primary structure and source of p76 have not been established.Here we show that p76 is a bona fide product of the mdm2gene, as it is expressed in normal wild-type MEFs,1 but not inmdm2-null fibroblasts. Expression of p76 is induced by UV light in a p53-dependent manner. Using epitope mapping and site-directed mutagenesis, we show that p76 appears to be the product of translational initiation at codon 50 (AUG). Both internal initiation and alternate splicing can give rise to p76 in cells. The p53-responsive internal promoter of mdm2 is induced by UV light, and a fraction of the induced RNA is spliced such that it lacks exon 3 and the first two AUG codons. In this alternatively spliced mRNA, AUG codon 50 (referred to hereafter as AUG 50) is the first initiation codon. We provide evidence that the increase in this alternatively spliced mRNA accounts for the induction of p76 expression in response to UV light. Identification and characterization of p76 are important because it is likely to share some oncogenic functions with p90.DISCUSSIONWe have shown here that the p76 MDM2 protein is synthesized in normal cells. Alternatively spliced mdm2 mRNAs lacking exon 3 express p76, but not p90, and such mRNAs are induced by UV light. Synthesis of p76 is initiated at an AUG codon at position 50 of the coding sequence for p90, and p76 lacks the N-terminal domain required for interaction with p53. However, the protein retains many of the structural motifs of p90, including the nuclear localization signal, nuclear export signals, acidic domain (20Fakharzadeh S.S. Trusko S.P. George D.L. EMBO J. 1991; 10: 1565-1569Crossref PubMed Scopus (626) Google Scholar), and RING finger (38Boddy M.N. Freemont P.S. Borden K.L.B. Trends Biochem. Sci. 1994; 19: 198-199Abstract Full Text PDF PubMed Scopus (84) Google Scholar). It is likely that p76 shares some functions with p90, and its presence in normal cells indicates that it may mediate some of the physiological functions of MDM2.The ability of MDM2 to stimulate the degradation of p53 requires that MDM2 bind p53 (10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). However, the level of p76 may influence the stability of p53 because p76 could bind factors that regulate the process. For example, p76 binds the p19ARF protein, which inhibits the ability of p90 to stimulate degradation of p53 (13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). An increase in the amount of p76 relative to p90 might free p90 to be more effective at stimulating degradation of p53. Alternatively, it is possible that p76 is a dominant-negative inhibitor of the ability of p90 to stimulate degradation of p53 since p76 might titrate factors required for that function of p90. We have preliminary evidence that the second model is correct. Overexpression of p76 antagonizes the decrease in p53 levels mediated by p90. 4M. Holubar and M. E. Perry, unpublished results. Therefore, regulation of the relative amounts of p76 and p90 MDM2 may be an important mediator of p53 levels.The ratio of p90 to p76 protein is determined, at least in part, by the ratio of the four different mdm2 mRNAs described here. The most abundant mdm2 mRNA in unstressed C127 cells is expressed from the P1 promoter and contains exon 3. This mRNA gives rise to p90 predominantly. Our RT-PCR analysis indicated thatmdm2 mRNAs lacking exon 3 account for 25% of the totalmdm2 mRNA in unstressed C127 cells. Such mRNAs give rise to p76 only. The results of the transfection experiments indicate that mRNAs lacking exon 3 may be up to five times more efficient at giving rise to p76 than are normally spliced mRNAs, if equal transfection efficiencies and RNA stabilities are assumed. If this number is valid, then more than half of the p76 present in C127 cells arises from alternatively spliced mRNA. These observations indicate that the ratio of normally spliced mRNA to alternatively spliced mRNA affects the ratio of p90 to p76. Indeed, 2 h following exposure to a dose of UV light of 4 J/m2, C127 cells have a higher ratio of p90 to p76, in part because there is more normally spliced mRNA relative to alternatively spliced mRNA than in the absence of exposure. In addition, the induced transcript is more efficient than the constitutive transcript at giving rise to p90 relative to p76. Under these same conditions, p53 levels are high, and the p53 protein is active as a transcriptional activator (29Saucedo L.J. Carstens B.C. Seavey S.E. Albee L.D. Perry M.E. Cell Growth Differ. 1998; 9: 119-130PubMed Google Scholar). Later in the UV response, p53 levels return to normal. The relative ratio of p76 to p90 may change throughout the UV response and determine whether p53 is stable. Since p53 stimulates apoptosis in response to UV light (39Zeigler A. Jonason A.S. Lefell D.J. Simon J.A. Sharma H.W. Kimmelman J. Remington L. Jacks T. Brash D.E. Nature. 1994; 372: 773-776Crossref PubMed Scopus (1344) Google Scholar), its levels may be critical determinants of cell survival.Dissection of the functions of p76 may reveal p53-independent functions of MDM2. For example, p76 retains the recently identified HECT domain, which is required for the ubiquitin ligase activity of MDM2 (40Honda R. Tanaka H. Yasuda H. FEBS Lett. 1997; 420: 25-27Crossref PubMed Scopus (1587) Google Scholar). Perhaps p76 targets a protein other than p53 for degradation. Evidence suggests that some of the oncogenic effects of MDM2 are mediated by mechanisms independent of p53 (4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar, 14Martin K. Trouche D. Hagemeier C. Sorensen T.S. La Thangue N.B. Kouzarides T. Nature. 1995; 375: 691-694Crossref PubMed Scopus (451) Google Scholar, 15Xiao Z.-X. Chen J. Levine A.J. Modjtahedl N. Xing J. Sellers W.R. Livingston D.M. Nature. 1995; 375: 694-698Crossref PubMed Scopus (573) Google Scholar). In a subset of human tumors, multiple MDM2 proteins are overexpressed (7Leach F.S. Tokino T. Meltzer P. Burrell M. Oliner J.D. Smith S. Hill D.E. Sidransky D. Kinzler K.W. Vogelstein B. Cancer Res. 1993; 53: 2231-2234PubMed Google Scholar, 18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). The human gene has AUG codons at positions 1, 6, 50, 62, and 102 (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar), and a human p76 protein is expressed in normal fibroblasts.4 It may be that p76 is overexpressed in tumors and contributes to development of cancer. Identification of any oncogenic properties of p76 is important because therapeutic interventions are being designed to disrupt the interaction between p90 and p53. In tumors overexpressing p76, such therapies may not be effective. The mdm2 oncogene is a determinant of embryogenesis (1Montes de Oca Luna R. Wagner D.S. Lozano G. Nature. 1995; 378: 203-206Crossref PubMed Scopus (1196) Google Scholar,2Jones S.N. Roe A.E. Donehower L.A. Bradley A. Nature. 1995; 378: 206-208Crossref PubMed Scopus (1057) Google Scholar), tumorigenesis (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar, 4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar), and cell cycle progression (5Chen C.-Y. Oliner J.D. Zhan Q. Fornace A.J. Vogelstein B. Kastan M.B. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 2684-2688Crossref PubMed Scopus (290) Google Scholar, 6Dubs-Poterszman M.-C. Tocque B. Wasylyk B. Oncogene. 1995; 11: 2445-2449PubMed Google Scholar). The effects of MDM2 on these processes depend, in part, on its ability to inactivate the p53 tumor suppressor protein. For example, mice with a homozygous deletion of the mdm2 gene die during embryogenesis, and this deficiency is rescued if p53is also deleted (1Montes de Oca Luna R. Wagner D.S. Lozano G. Nature. 1995; 378: 203-206Crossref PubMed Scopus (1196) Google Scholar, 2Jones S.N. Roe A.E. Donehower L.A. Bradley A. Nature. 1995; 378: 206-208Crossref PubMed Scopus (1057) Google Scholar). In human tumors, the homologue of MDM2 is overexpressed most often in the subset lacking inactivating mutations in the p53 gene (7Leach F.S. Tokino T. Meltzer P. Burrell M. Oliner J.D. Smith S. Hill D.E. Sidransky D. Kinzler K.W. Vogelstein B. Cancer Res. 1993; 53: 2231-2234PubMed Google Scholar). Thus, high levels of MDM2 may be redundant with mutational inactivation of p53. Following exposure of cells to γ-radiation, MDM2 inactivates the G1 block mediated by p53 (5Chen C.-Y. Oliner J.D. Zhan Q. Fornace A.J. Vogelstein B. Kastan M.B. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 2684-2688Crossref PubMed Scopus (290) Google Scholar). In some cell types, MDM2 can inhibit apoptosis mediated by p53 (8Haupt Y. Barak Y. Oren M. EMBO J. 1996; 15: 1596-1606Crossref PubMed Scopus (204) Google Scholar). MDM2 binds to p53, inhibiting the transcriptional activation function of p53 (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar, 9Momand J. Zambetti G.P. Olson D.C. George D. Levine A.J. Cell. 1992; 69: 1237-1245Abstract Full Text PDF PubMed Scopus (2776) Google Scholar) and stimulating degradation of p53 (10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). Recently, MDM2 has been shown to bind to another tumor suppressor protein, p19ARF(12Zhang Y. Xiong Y. Yarborough W.G. Cell. 1998; 92: 725-734Abstract Full Text Full Text PDF PubMed Scopus (1392) Google Scholar, 13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). The ability of MDM2 to stimulate degradation of p53 is inhibited by p19ARF, indicating that the interaction of p19ARF and MDM2 may be an important determinant of cell cycle progression. In addition to interacting with p53 and p19ARF, MDM2 binds to and alters the function of other proteins that regulate cell division. For example, overexpression of MDM2 stimulates the activity of the E2F transcription factor (14Martin K. Trouche D. Hagemeier C. Sorensen T.S. La Thangue N.B. Kouzarides T. Nature. 1995; 375: 691-694Crossref PubMed Scopus (451) Google Scholar) and reverses an arrest in the cell cycle mediated by either the retinoblastoma protein (15Xiao Z.-X. Chen J. Levine A.J. Modjtahedl N. Xing J. Sellers W.R. Livingston D.M. Nature. 1995; 375: 694-698Crossref PubMed Scopus (573) Google Scholar) or a related protein, p107 (6Dubs-Poterszman M.-C. Tocque B. Wasylyk B. Oncogene. 1995; 11: 2445-2449PubMed Google Scholar). These effects of MDM2 do not depend on the presence of p53. Further evidence for p53-independent functions of MDM2 was provided by Lundgren et al. (4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar), who generated mice expressing high levels of MDM2 in the mammary gland. The mammary epithelial cells of these mice developed polyploidy, and this effect of MDM2 was also seen in mice lacking p53. Mammary tumors arose in mice overexpressing mdm2, but the dependence on p53 was not measured since mice lacking p53 die before mammary tumors would have developed. Sigalis et al. (16Sigalis I. Calvert A.H. Anderson J.J. Neal D.E. Lunec J. Nat. Med. 1996; 2: 912-917Crossref PubMed Scopus (251) Google Scholar) found that alternatively spliced mdm2 mRNAs were overexpressed in some human tumors. Introduction of these mRNAs into NIH3T3 cells resulted in morphological transformation, even though a subset of them encoded MDM2 proteins that could not bind p53. Therefore, some of the oncogenic effects of MDM2 are independent of p53. Several MDM2 proteins are detectable in cells overexpressing MDM2, including human tumor cells and transformed murine cell lines (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar, 18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). Olson et al. (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar) hypothesized that these proteins arise through alternative splicing, proteolytic processing, or post-translational modification. It has not been clear whether these proteins arise through mechanisms used in the normal regulation ofmdm2 expression. Multiple MDM2 proteins can be translated from single mRNAs in rabbit reticulocyte lysates, indicating that internal initiation may be a mechanism whereby multiple MDM2 proteins are expressed in cells (19Barak Y. Gottlieb E. Juven-Gershon T. Oren M. Genes Dev. 1994; 8: 1739-1749Crossref PubMed Scopus (285) Google Scholar). In fact, there is evidence that enhanced translation accounts for the overexpression of multiple MDM2 proteins in some human tumors (18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). In the DM3T3 cell line, in which themdm2 gene is amplified, multiple MDM2 proteins and mRNAs are expressed (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar, 20Fakharzadeh S.S. Trusko S.P. George D.L. EMBO J. 1991; 10: 1565-1569Crossref PubMed Scopus (626) Google Scholar). The most abundant protein in DM3T3 cells (p90) is the well characterized mdm2 gene product that binds to p53 and inhibits its function (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar, 9Momand J. Zambetti G.P. Olson D.C. George D. Levine A.J. Cell. 1992; 69: 1237-1245Abstract Full Text PDF PubMed Scopus (2776) Google Scholar, 10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). The second most abundant protein (p76) does not bind p53 (17Olson D.C. Marechal V. Momand J. Chen J. Romocki C. Levine A.J. Oncogene. 1993; 8: 2353-2360PubMed Google Scholar, 19Barak Y. Gottlieb E. Juven-Gershon T. Oren M. Genes Dev. 1994; 8: 1739-1749Crossref PubMed Scopus (285) Google Scholar), but does bind p19ARF (13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). The primary structure and source of p76 have not been established. Here we show that p76 is a bona fide product of the mdm2gene, as it is expressed in normal wild-type MEFs,1 but not inmdm2-null fibroblasts. Expression of p76 is induced by UV light in a p53-dependent manner. Using epitope mapping and site-directed mutagenesis, we show that p76 appears to be the product of translational initiation at codon 50 (AUG). Both internal initiation and alternate splicing can give rise to p76 in cells. The p53-responsive internal promoter of mdm2 is induced by UV light, and a fraction of the induced RNA is spliced such that it lacks exon 3 and the first two AUG codons. In this alternatively spliced mRNA, AUG codon 50 (referred to hereafter as AUG 50) is the first initiation codon. We provide evidence that the increase in this alternatively spliced mRNA accounts for the induction of p76 expression in response to UV light. Identification and characterization of p76 are important because it is likely to share some oncogenic functions with p90. DISCUSSIONWe have shown here that the p76 MDM2 protein is synthesized in normal cells. Alternatively spliced mdm2 mRNAs lacking exon 3 express p76, but not p90, and such mRNAs are induced by UV light. Synthesis of p76 is initiated at an AUG codon at position 50 of the coding sequence for p90, and p76 lacks the N-terminal domain required for interaction with p53. However, the protein retains many of the structural motifs of p90, including the nuclear localization signal, nuclear export signals, acidic domain (20Fakharzadeh S.S. Trusko S.P. George D.L. EMBO J. 1991; 10: 1565-1569Crossref PubMed Scopus (626) Google Scholar), and RING finger (38Boddy M.N. Freemont P.S. Borden K.L.B. Trends Biochem. Sci. 1994; 19: 198-199Abstract Full Text PDF PubMed Scopus (84) Google Scholar). It is likely that p76 shares some functions with p90, and its presence in normal cells indicates that it may mediate some of the physiological functions of MDM2.The ability of MDM2 to stimulate the degradation of p53 requires that MDM2 bind p53 (10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). However, the level of p76 may influence the stability of p53 because p76 could bind factors that regulate the process. For example, p76 binds the p19ARF protein, which inhibits the ability of p90 to stimulate degradation of p53 (13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). An increase in the amount of p76 relative to p90 might free p90 to be more effective at stimulating degradation of p53. Alternatively, it is possible that p76 is a dominant-negative inhibitor of the ability of p90 to stimulate degradation of p53 since p76 might titrate factors required for that function of p90. We have preliminary evidence that the second model is correct. Overexpression of p76 antagonizes the decrease in p53 levels mediated by p90. 4M. Holubar and M. E. Perry, unpublished results. Therefore, regulation of the relative amounts of p76 and p90 MDM2 may be an important mediator of p53 levels.The ratio of p90 to p76 protein is determined, at least in part, by the ratio of the four different mdm2 mRNAs described here. The most abundant mdm2 mRNA in unstressed C127 cells is expressed from the P1 promoter and contains exon 3. This mRNA gives rise to p90 predominantly. Our RT-PCR analysis indicated thatmdm2 mRNAs lacking exon 3 account for 25% of the totalmdm2 mRNA in unstressed C127 cells. Such mRNAs give rise to p76 only. The results of the transfection experiments indicate that mRNAs lacking exon 3 may be up to five times more efficient at giving rise to p76 than are normally spliced mRNAs, if equal transfection efficiencies and RNA stabilities are assumed. If this number is valid, then more than half of the p76 present in C127 cells arises from alternatively spliced mRNA. These observations indicate that the ratio of normally spliced mRNA to alternatively spliced mRNA affects the ratio of p90 to p76. Indeed, 2 h following exposure to a dose of UV light of 4 J/m2, C127 cells have a higher ratio of p90 to p76, in part because there is more normally spliced mRNA relative to alternatively spliced mRNA than in the absence of exposure. In addition, the induced transcript is more efficient than the constitutive transcript at giving rise to p90 relative to p76. Under these same conditions, p53 levels are high, and the p53 protein is active as a transcriptional activator (29Saucedo L.J. Carstens B.C. Seavey S.E. Albee L.D. Perry M.E. Cell Growth Differ. 1998; 9: 119-130PubMed Google Scholar). Later in the UV response, p53 levels return to normal. The relative ratio of p76 to p90 may change throughout the UV response and determine whether p53 is stable. Since p53 stimulates apoptosis in response to UV light (39Zeigler A. Jonason A.S. Lefell D.J. Simon J.A. Sharma H.W. Kimmelman J. Remington L. Jacks T. Brash D.E. Nature. 1994; 372: 773-776Crossref PubMed Scopus (1344) Google Scholar), its levels may be critical determinants of cell survival.Dissection of the functions of p76 may reveal p53-independent functions of MDM2. For example, p76 retains the recently identified HECT domain, which is required for the ubiquitin ligase activity of MDM2 (40Honda R. Tanaka H. Yasuda H. FEBS Lett. 1997; 420: 25-27Crossref PubMed Scopus (1587) Google Scholar). Perhaps p76 targets a protein other than p53 for degradation. Evidence suggests that some of the oncogenic effects of MDM2 are mediated by mechanisms independent of p53 (4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar, 14Martin K. Trouche D. Hagemeier C. Sorensen T.S. La Thangue N.B. Kouzarides T. Nature. 1995; 375: 691-694Crossref PubMed Scopus (451) Google Scholar, 15Xiao Z.-X. Chen J. Levine A.J. Modjtahedl N. Xing J. Sellers W.R. Livingston D.M. Nature. 1995; 375: 694-698Crossref PubMed Scopus (573) Google Scholar). In a subset of human tumors, multiple MDM2 proteins are overexpressed (7Leach F.S. Tokino T. Meltzer P. Burrell M. Oliner J.D. Smith S. Hill D.E. Sidransky D. Kinzler K.W. Vogelstein B. Cancer Res. 1993; 53: 2231-2234PubMed Google Scholar, 18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). The human gene has AUG codons at positions 1, 6, 50, 62, and 102 (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar), and a human p76 protein is expressed in normal fibroblasts.4 It may be that p76 is overexpressed in tumors and contributes to development of cancer. Identification of any oncogenic properties of p76 is important because therapeutic interventions are being designed to disrupt the interaction between p90 and p53. In tumors overexpressing p76, such therapies may not be effective. We have shown here that the p76 MDM2 protein is synthesized in normal cells. Alternatively spliced mdm2 mRNAs lacking exon 3 express p76, but not p90, and such mRNAs are induced by UV light. Synthesis of p76 is initiated at an AUG codon at position 50 of the coding sequence for p90, and p76 lacks the N-terminal domain required for interaction with p53. However, the protein retains many of the structural motifs of p90, including the nuclear localization signal, nuclear export signals, acidic domain (20Fakharzadeh S.S. Trusko S.P. George D.L. EMBO J. 1991; 10: 1565-1569Crossref PubMed Scopus (626) Google Scholar), and RING finger (38Boddy M.N. Freemont P.S. Borden K.L.B. Trends Biochem. Sci. 1994; 19: 198-199Abstract Full Text PDF PubMed Scopus (84) Google Scholar). It is likely that p76 shares some functions with p90, and its presence in normal cells indicates that it may mediate some of the physiological functions of MDM2. The ability of MDM2 to stimulate the degradation of p53 requires that MDM2 bind p53 (10Haupt Y. Maya R. Kazaz A. Oren M. Nature. 1997; 387: 296-299Crossref PubMed Scopus (3659) Google Scholar, 11Kubbutat M.H.G. Jones S.N. Vousden K. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2814) Google Scholar). However, the level of p76 may influence the stability of p53 because p76 could bind factors that regulate the process. For example, p76 binds the p19ARF protein, which inhibits the ability of p90 to stimulate degradation of p53 (13Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.-W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1324) Google Scholar). An increase in the amount of p76 relative to p90 might free p90 to be more effective at stimulating degradation of p53. Alternatively, it is possible that p76 is a dominant-negative inhibitor of the ability of p90 to stimulate degradation of p53 since p76 might titrate factors required for that function of p90. We have preliminary evidence that the second model is correct. Overexpression of p76 antagonizes the decrease in p53 levels mediated by p90. 4M. Holubar and M. E. Perry, unpublished results. Therefore, regulation of the relative amounts of p76 and p90 MDM2 may be an important mediator of p53 levels. The ratio of p90 to p76 protein is determined, at least in part, by the ratio of the four different mdm2 mRNAs described here. The most abundant mdm2 mRNA in unstressed C127 cells is expressed from the P1 promoter and contains exon 3. This mRNA gives rise to p90 predominantly. Our RT-PCR analysis indicated thatmdm2 mRNAs lacking exon 3 account for 25% of the totalmdm2 mRNA in unstressed C127 cells. Such mRNAs give rise to p76 only. The results of the transfection experiments indicate that mRNAs lacking exon 3 may be up to five times more efficient at giving rise to p76 than are normally spliced mRNAs, if equal transfection efficiencies and RNA stabilities are assumed. If this number is valid, then more than half of the p76 present in C127 cells arises from alternatively spliced mRNA. These observations indicate that the ratio of normally spliced mRNA to alternatively spliced mRNA affects the ratio of p90 to p76. Indeed, 2 h following exposure to a dose of UV light of 4 J/m2, C127 cells have a higher ratio of p90 to p76, in part because there is more normally spliced mRNA relative to alternatively spliced mRNA than in the absence of exposure. In addition, the induced transcript is more efficient than the constitutive transcript at giving rise to p90 relative to p76. Under these same conditions, p53 levels are high, and the p53 protein is active as a transcriptional activator (29Saucedo L.J. Carstens B.C. Seavey S.E. Albee L.D. Perry M.E. Cell Growth Differ. 1998; 9: 119-130PubMed Google Scholar). Later in the UV response, p53 levels return to normal. The relative ratio of p76 to p90 may change throughout the UV response and determine whether p53 is stable. Since p53 stimulates apoptosis in response to UV light (39Zeigler A. Jonason A.S. Lefell D.J. Simon J.A. Sharma H.W. Kimmelman J. Remington L. Jacks T. Brash D.E. Nature. 1994; 372: 773-776Crossref PubMed Scopus (1344) Google Scholar), its levels may be critical determinants of cell survival. Dissection of the functions of p76 may reveal p53-independent functions of MDM2. For example, p76 retains the recently identified HECT domain, which is required for the ubiquitin ligase activity of MDM2 (40Honda R. Tanaka H. Yasuda H. FEBS Lett. 1997; 420: 25-27Crossref PubMed Scopus (1587) Google Scholar). Perhaps p76 targets a protein other than p53 for degradation. Evidence suggests that some of the oncogenic effects of MDM2 are mediated by mechanisms independent of p53 (4Lundgren K. Montes de Oca Luna R. McNeill Y.B. Emerick E.P. Spencer B. Barfield C.R. Lozano G. Rosenberg M.P. Finlay C.A. Genes Dev. 1997; 11: 714-725Crossref PubMed Scopus (212) Google Scholar, 14Martin K. Trouche D. Hagemeier C. Sorensen T.S. La Thangue N.B. Kouzarides T. Nature. 1995; 375: 691-694Crossref PubMed Scopus (451) Google Scholar, 15Xiao Z.-X. Chen J. Levine A.J. Modjtahedl N. Xing J. Sellers W.R. Livingston D.M. Nature. 1995; 375: 694-698Crossref PubMed Scopus (573) Google Scholar). In a subset of human tumors, multiple MDM2 proteins are overexpressed (7Leach F.S. Tokino T. Meltzer P. Burrell M. Oliner J.D. Smith S. Hill D.E. Sidransky D. Kinzler K.W. Vogelstein B. Cancer Res. 1993; 53: 2231-2234PubMed Google Scholar, 18Landers J.E. Haines D.S. Strauss J.F. George D.L. Oncogene. 1994; 9: 2745-2750PubMed Google Scholar). The human gene has AUG codons at positions 1, 6, 50, 62, and 102 (3Oliner J.D. Kinzler K.W. Meltzer P.S. George D.L. Vogelstein B. Nature. 1992; 358: 80-83Crossref PubMed Scopus (1790) Google Scholar), and a human p76 protein is expressed in normal fibroblasts.4 It may be that p76 is overexpressed in tumors and contributes to development of cancer. Identification of any oncogenic properties of p76 is important because therapeutic interventions are being designed to disrupt the interaction between p90 and p53. In tumors overexpressing p76, such therapies may not be effective. We thank Moshe Oren for mdm2cDNA constructs; Arnold J. Levine for monoclonal antibodies to p53, MDM2, and T-antigen; Angie Teresky for wild-type mouse embryo fibroblasts; Guillermina Lozano for p53-null andp53/mdm2-null mouse fibroblasts; and Bill Sugden for COS-1 cells. The technical expertise of Marisa Holubar and Amy Prevost is greatly appreciated. We gratefully acknowledge Nancy Thompson and Kit Nolan for advice on typing monoclonal antibodies and John Petrini for advice on site-directed mutagenesis. We appreciate stimulating conversations about this work with Marilyn Kozak, Moshe Oren, and Jeff Ross and insightful comments on the manuscript by Chris Bradfield, Susan Mendrysa, and Jeff Ross.
Referência(s)