Overproduction of MDM2 in vivo disrupts S phase independent of E2F1.

Expression of a beta-lactoglobulin (BLG)/mdm2 transgene (BLGmdm2) in the epithelial cells of the mouse mammary gland causes an uncoupling of S phase from M phase, resulting in polyploidy and tumor formation. The cell cycle defects are independent of interactions with p53. Because MDM2 also binds and activates the S phase-specific transcription factor E2F1, we hypothesized that increased E2F1 activity causes the development of the BLGmdm2 phenotype. We, therefore, generated BLGmdm2 mice that were null for E2F1. We observed no notable differences in histology or cyclin gene expression between BLGmdm2 and BLGmdm2/E2F1-/- mice, indicating that endogenous E2F1 activity was not required for the BLGmdm2 phenotype. Because, depending on the experimental system, either loss of E2F1 function or overexpression of E2F1 results in transformation, we also tested whether overexpression of E2F1 augmented the severity of the BLGmdm2 phenotype by generating mice that were bitransgenic for BLGmdm2 and BLGE2F1. We observed a unique mixture of the two single transgenic phenotypes histologically and found no significant changes in cyclin levels, indicating that overexpression of E2F1 had no effect on the BLGmdm2 transgenic phenotype. Thus, increased expression or absence of E2F1 does not affect the ability of MDM2 to disrupt the cell cycle.

Targeted expression of MDM2 uncouples S phase from mitosis and inhibits mammary gland development independent of p53.

MDM2 is a cellular protein that binds to and inactivates the p53 tumor suppressor protein. Although mdm2 has been shown to function as an oncogene in vitro, all studies to date have assessed MDM2 activities in the presence of p53, implicating p53 inactivation in MDM2-directed transformation. To determine the role of MDM2 in the cell cycle and in tumorigenesis and whether or not this role is dependent on p53, an MDM2 minigene was expressed during gestation and lactation in the mammary gland of both wild-type p53 (p53+/+) and p53 knockout (p53-/-) mice using the bovine beta-lactoglobulin promoter. In six different transgenic mouse lines, deregulated expression of MDM2 inhibited normal development and morphogenesis of the mammary gland, and caused cellular hypertrophy and nuclear abnormalities. These abnormalities included both multinucleated cells and enlarged cells with giant nuclei. Although there were fewer epithelial cells present in the transgenic mammary gland, no apoptosis was observed. Instead, BrdU incorporation and PCNA staining showed that 12%-27% of the transgenic mammary epithelial cells were in S phase at a time when normal cells were terminally differentiated. Analysis of DNA content showed that 30%-45% of the cells were polyploid, with DNA contents up to 16N, indicating that overexpression of MDM2 caused mammary epithelial cells to undergo multiple rounds of S phase without cell division. This phenotype was similar in the p53+/+ and p53-/- background, demonstrating a role for MDM2 in the regulation of DNA synthesis that is independent of the ability of MDM2 to inhibit p53 activity. Additionally, multiple lines of BLGMDM2 transgenic mice developed mammary tumors, confirming that overproduction of MDM2 contributes to tumorigenesis in epithelial cells in vivo.

The organization and expression of the mdm2 gene.

The mdm2 gene encodes a zinc finger protein that negatively regulates p53 function by binding and masking the p53 transcriptional activation domain. Two different promoters control expression of mdm2, one of which is also transactivated by p53. We cloned and characterized the mdm2 gene from a murine 129 library. It contained at least 12 exons and spanned approximately 25 kb of DNA. Sequencing of the mdm2 gene revealed three nucleotide differences that resulted in amino acid substitutions in the previously published mdm2 sequence. Sequencing of normal BalbC/J DNA and the original cosmid clone isolated from the 3T3DM cell line revealed that they are identical, suggesting that the published sequence is in error at these three positions. In addition, we analyzed the expression pattern of mdm2 and found ubiquitous low-level expression throughout embryo development and in adult tissues. Analysis of mRNA from numerous tissues for several mdm2 spliced variants that had been identified in the transformed 3T3DM cell line revealed that these variants could not be detected in the developing embryo or in adult tissues.

A functionally inactive p53 Li-Fraumeni syndrome mutant.

Germline mutations in the tumor-suppressor p53 have been recently identified in Li-Fraumeni syndrome patients. We analysed the function of one of these mutations, an arg-to-trp substitution at amino acid 245 in the murine p53 gene. This p53LFS mutant could not, unlike wild-type p53, suppress foci formation of rat embryo-fibroblasts. Like other p53 mutants it cooperated with activated ras to transform rat embryo fibroblasts. Overexpression of p53LFS thus resulted in a phenotype similar to other mutant p53s. The p53LFS protein was also transcriptionally inactive in contrast to previous studies using a p53LFS/GAL4 fusion protein. To better understand the functional domain disrupted in p53LFS, we developed a dimerization assay and showed that p53LFS still dimerized. In addition, p53LFS retained its ability to bind SV40 large T antigen and not hsc70, both characteristics of wild-type p53. Using immunofluorescence, we localized p53LFS to the nucleus. From these results we conclude that p53LFS represents an unusual p53 mutant in that it retains many characteristics of wild-type p53, however activities critical for growth suppression are lost.

The mdm-2 oncogene can overcome wild-type p53 suppression of transformed cell growth.

Expression of a p53-associated protein, Mdm-2 (murine double minute-2), can inhibit p53-mediated transactivation. In this study, overexpression of the Mdm-2 protein was found to result in the immortalization of primary rat embryo fibroblasts (REFs) and, in conjunction with an activated ras gene, in the transformation of REFs. The effect of wild-type p53 on the transforming properties of mdm-2 was determined by transfecting REFs with ras, mdm-2, and normal p53 genes. Transfection with ras plus mdm-2 plus wild-type p53 resulted in a 50% reduction in the number of transformed foci (relative to the level for ras plus mdm-2); however, more than half (9 of 17) of the cell lines derived from these foci expressed low levels of a murine p53 protein with the characteristics of a wild-type p53. These results are in contrast to previous studies which demonstrated that even minimal levels of wild-type p53 are not tolerated in cells transformed by ras plus myc, E1A, or mutant p53. The mdm-2 oncogene can overcome the previously demonstrated growth-suppressive properties of p53.

p53 loss of function: implications for the processes of immortalization and tumorigenesis.

The complex process of cell immortalization and transformation is likely to involve the inactivation of growth regulatory genes. Mutations (deletions, missense mutations) in the p53 gene are the most frequently observed genetic alteration in human tumors, making p53 a candidate for a cellular protein involved in the control of cell growth. Two recent studies have examined the role of p53 in immortalization and tumorigenesis. In the first study, p53 expression was examined in both mortal and immortal chick embryo fibroblasts. All mortal clones expressed p53 but the loss of wild-type p53 expression was observed in every immortal cell line examined. In the second study, a line of mice carrying two null p53 alleles has been created and characterized. Although these mice develop normally, they show a predisposition to develop a variety of neoplasms at an early age (< 6 months). Although it is unclear whether p53 regulates the same, different, or overlapping pathways in the two experimental systems, these data demonstrate that p53 function is critical for the maintenance of normal growth control and support the current classification of p53 as a growth suppressive or tumor suppressor gene.

The p53 tumour suppressor gene.

The cell cycle is composed of a series of steps which can be negatively or positively regulated by various factors. Chief among the negative regulators is the p53 protein. Alteration or inactivation of p53 by mutation, or by its interactions with oncogene products of DNA tumour viruses, can lead to cancer. These mutations seem to be the most common genetic change in human cancers.

Mutant p53 DNA clones from human colon carcinomas cooperate with ras in transforming primary rat cells: a comparison of the "hot spot" mutant phenotypes.

The majority of the p53 genes derived from human colorectal carcinomas contain point mutations. A significant number of these mutations occur in or around amino acids 143, 175, 273, or 281. Experiments presented here demonstrate for the first time that p53 DNA clones containing any one of these mutations cooperate with the activated ras oncogene to transform primary rat embryo cells in culture. These transformed cells produce elevated levels of the human p53 protein, which has extended half-lives (1.5-7 h), as compared to the wild-type human p53 protein (20-30 min). The p53 mutant with an alteration at residue 175 (p53-175H) binds tightly to the cellular heat shock protein, hsc70. In contrast, the p53 mutants possessing mutations at either residue 273 or 281 (p53-273H/281G) do not bind detectably to this heat shock protein and generally are less efficient at forming transformed foci in culture. The transformed cell lines are tumorigenic in nude mice. Thus, two classes of p53 mutant proteins can be distinguished: p53-175H, which cooperates with ras efficiently and binds to hsc70, and p53-273H/281G, which has a reduced efficiency of transformed foci formation and does not bind hsc70. This demonstrates that complex formation between mutant p53 and hsc70 is not required for p53-mediated transformation, but rather it facilitates this function, perhaps by ensuring sequestration of the endogenous wild-type p53 protein. The positive effect on cell proliferation by these mutant p53 proteins is consistent with a role for activated p53 mutants in the genesis of colorectal carcinomas.

The p53 proto-oncogene can act as a suppressor of transformation.

DNA clones of the wild-type p53 proto-oncogene inhibit the ability of E1A plus ras or mutant p53 plus ras-activated oncogenes to transform primary rat embryo fibroblasts. The rare clones of transformed foci that result from E1A plus ras plus wild-type p53 triple transfections all contain the p53 DNA in their genome, but the great majority fail to express the p53 protein. The three cell lines derived from such foci that express p53 all produce mutant p53 proteins with properties similar or identical to transformation-activated p53 proteins. The p53 mutants selected in this fashion (transformation in vitro) resemble the p53 mutants selected in tumors (in vivo). These results suggest that the p53 proto-oncogene can act negatively to block transformation.

Activating mutations for transformation by p53 produce a gene product that forms an hsc70-p53 complex with an altered half-life.

The 11-4 p53 cDNA clone failed to transform primary rat fibroblasts when cotransfected with the ras oncogene. Two linker insertion mutations at amino acid 158 or 215 (of 390 amino acids) activated this p53 cDNA for transformation with ras. These mutant cDNAs produced a p53 protein that lacked an epitope, recognized by monoclonal antibody PAb246 (localized at amino acids 88 to 110 in the protein) and preferentially bound to a heat shock protein, hsc70. In rat cells transformed by a genomic p53 clone plus ras, two populations of p53 proteins were detected, PAb246+ and PAb246-, which did or did not bind to this monoclonal antibody, respectively. The PAb246- p53 preferentially associated with hsc70, and this protein had a half-life 4- to 20-fold longer than free p53 (PAb246+). These data suggest a possible functional role for hsc70 in the transformation process. cDNAs for p53 derived from methylcholanthrene-transformed cells transform rat cells in cooperation with the ras oncogene and produce a protein that bound with the heat shock proteins. Recombinant clones produced between a Meth A cDNA and 11-4 were tested for the ability to transform rat cells. A single amino acid substitution at residue 132 was sufficient to activate the 11-4 p53 cDNA for transformation. These studies have identified a region between amino acids 132 and 215 in the p53 protein which, when mutated, can activate the p53 cDNA. These results also call into question what the correct p53 wild-type sequence is and whether a wild-type p53 gene can transform cells in culture.

Immunological evidence for the association of p53 with a heat shock protein, hsc70, in p53-plus-ras-transformed cell lines.

A rabbit antiserum was prepared against the C-terminal peptide of 21 amino acids from the human heat shock protein hsp70. These antibodies were shown to be specific for this highly inducible heat shock protein (72 kilodaltons [kDa] in rat cells), and for a moderately inducible, constitutively expressed heat shock protein, hsc70 (74 kDa). In six independently derived rat cell lines transformed by a murine cDNA-genomic hybrid clone of p53 plus an activated Ha-ras gene, elevated levels of p53 were detected by immunoprecipitation by using murine-specific anti-p53 monoclonal antibodies. In all cases, the hsc70, but not the hsp70, protein was coimmunoprecipitated with the murine p53 protein. Similarly, antiserum to heat shock protein coimmunoprecipitated p53. Western blot (immunoblot) analysis demonstrated that the hsc70 and p53 proteins did not share detectable antigenic epitopes. The results provide clear immunological evidence for the specific association of a single heat shock protein, hsc70, with p53 in p53-plus-ras-transformed cell lines. A p53 cDNA clone, p11-4, failed to produce clonable cell lines from foci of primary rat cells transfected with p11-4 plus Ha-ras. A mutant p53 cDNA clone derived from p11-4, SVKH215, yielded a 2- to 35-fold increase in the number of foci produced after transfection of rat cells with SVKH215 plus Ha-ras. When cloned, 87.5% of these foci produced transformed cell lines. SVKH215 encodes a mutant p53 protein that binds preferentially to the heat shock proteins of 70 kDa compared with binding by the parental p11-4 p53 gene product. These data suggest that the p53-hsc70 protein complex could have functional significance in these transformed cells.

Autocrine stimulation of WI38 cell proliferation in the presence of glucocorticoids. Characteristics of the stimulatory factor(s) involved in this response.

Chronic exposure to hydrocortisone (HC) or dexamethasone (DEX) results in a 20-40% extension in the proliferative lifespan of WI38 cells. Within a single growth cycle, the addition of HC or DEX at seeding results in saturation densities 20-40% higher than in control cultures. We have recently reported that, within a single growth cycle, the proliferative response of WI38 cells to glucocorticoids is mediated by a stimulatory factor(s) present in medium conditioned by cells in the presence of the hormone. We report here that chronic exposure to medium conditioned in the presence of HC for the first 24 h after seeding (24-h HC-conditioned medium (24-h HC-CM)) results in a 25% extension in the proliferative lifespan of these cultures. The generation of the stimulatory factor(s) present in glucocorticoid-conditioned medium is apparently dependent upon undefined cellular alterations which result from the subcultivation-procedure; confluent or low-density quiescent cultures did not generate media stimulatory to cell growth in the presence of glucocorticoids. This response was not trypsin-dependent, since cultures subcultivated in the absence of proteolytic treatment generated media equally stimulatory to cell growth. A further characterization of this glucocorticoid-induced activity revealed the stimulatory factor(s) was of low MW (dialyzable and recoverable in the less than 10,000 MW fraction following ultrafiltration), heat-stable (95 degrees C), and resistant to treatment with trypsin, chymotrypsin, or protease (S. griseus).

Enhanced proliferation of WI38 cells in the presence of glucocorticoid-conditioned medium.

The addition of hydrocortisone (HC) or dexamethasone (DEX) to WI38 cells at subcultivation is known to result in increased saturation densities (20-40%). We report that maximal increases in saturation density are, however, only observed if HC is added to the culture shortly after subcultivation. We have found that the proliferative response of WI38 cells to glucocorticoids is mediated by a secondary stimulatory factor(s) present in medium-conditioned by cells in the presence of the hormone. Control cultures refed with medium conditioned in the presence of HC for the first 24 h after seeding (24 h HC-CM) achieve saturation densities 20-40% higher than cultures refed with either medium conditioned in the absence of the hormone (24 h CM) or 24 h CM supplemented with fresh HC. Furthermore, WI38 cultures remain responsive to the stimulatory activity present in 24 h HC-CM throughout the growth cycle. The stimulatory effects of 24 h HC-CM are enhanced by repeatedly refeeding cultures; WI38 cells refed every 2 days with 24 h HC-CM demonstrate an extended period of logarithmic growth and achieve densities 2-3 times higher than controls. A preliminary characterization of this activity shows it to be of low molecular weight (MW) (dialyzable using 12 000 MW cut-off tubing) and heat-stable (75 degrees C).