The cellular oncogene p53 can be activated by mutagenesis

P53 is a cellular phosphoprotein of short half-life (t1/2) which is present at elevated levels in cells transformed by various stimuli including viruses, chemicals and radiation. p53 forms specific stable complexes with simian virus 40 (SV40) large-T antigen and an adenovirus E1b protein of relative molecular mass (Mr) 57,000. A number of reports have associated p53 with cell proliferation, and p53 complementary DNA expression constructs immortalize primary cells in vitro and render them sensitive to transformation by an activated ras oncogene. We have examined the biological properties of a set of p53 expression constructs, and report here that cellular immortalization by a wild-type p53 cDNA gene is conditional upon the promoter/enhancer construction used, but that p53 can extend cellular lifespan by a second distinct mechanism involving rearrangements of the coding sequence which give rise to stable protein products. Cells immortalized by one of these mutants are refractory to subsequent transformation by a ras oncogene, indicating that cellular immortalization and ras cooperation are separate activities.     

Cellular immortalization by a cDNA clone encoding the transformation-associated phosphoprotein p53

Malignant transformation of primary cells requires at least two distinct and characteristic alterations in cellular behaviour. The first, cellular immortality, can be induced by chemical carcinogens or by cloned oncogenes such as polyoma large T (ref. 4), adenovirus early region 1A (E1A) or the oncogene from avian (MC29) myelocytomatosis virus, v-myc. Cells whose in vitro life-span has been extended by these procedures can be fully transformed by transfection with oncogenes belonging to a different complementation group, including genes of the ras family, adenovirus E1b and polyoma virus middle T (refs 4, 5). The unstable cellular phosphoprotein p53 is frequently present at elevated levels in transformed cells and is stabilized by the formation of complexes with simian virus 40 (SV40) large T or adenovirus E1b 57K protein. Although several reports have associated p53 with cell proliferation, its role remains obscure. We have cloned complementary DNA sequences encoding murine p53 and report here that transfection of p53 expression constructs into cells of finite lifespan in vitro results in cellular immortality and susceptibility to transformation by a ras oncogene.     

Specific interaction between the p53 cellular tumour antigen and major heat shock proteins

The protein p53 is capable of participating in neoplastic transformation and can form specific complexes with the large-T antigen of simian virus 40 (SV40). This interaction probably results in the stabilization of p53 (refs 7,8) and may contribute to SV40-mediated transformation. Several non-SV40-transformed cells also exhibit a stabilized p53 which is present in elevated levels. Recently, this stabilization was shown to coincide with the ability to precipitate a polypeptide (p68) of relative molecular mass (Mr) 68,000-70,000 by anti-p53 monoclonal antibodies. We now report that this co-precipitation indeed represents a specific complex between the two proteins; the complex sediments on a sucrose gradient as a relatively broad peak of 10-14S and can be dissociated in vitro. Furthermore, p68 is the HSP70 heat shock protein cognate, found in elevated levels in a p53-overproducing cell line. On heat-shock treatment of such overproducers, p53 also forms a complex with the related highly inducible HSP68.     

p53 and DNA polymerase alpha compete for binding to SV40 T antigen

The large T antigen (T) of simian virus 40 is a multifunctional protein required for both viral DNA replication and cellular transformation. T antigen forms specific protein complexes with the host protein p53 in both virus-infected and transformed cells. p53 has recently been shown to be an oncogene, but its normal function is not clear. We previously established a radioimmunoassay to study the newly described complex between T antigen and DNA polymerase alpha, and have noted a similarity between the antigenic changes induced in T by the binding of both p53 and polymerase. We now extend this analysis to a larger collection of anti-T antibodies and formally establish that p53 and DNA polymerase alpha can compete for binding to the SV40 T antigen. At a critical concentration of the three components it is possible to detect a trimeric complex of T, p53 and DNA polymerase alpha. Our observations have important implications for the control by these nuclear oncogenes of viral and cellular DNA synthesis and viral host range in both normal and transformed cells. We present a model for the action of p53 in growth control.     

An embryo protein induced by SV40 virus transformation of mouse cells

A specific protein of molecular weight (MW) approximately 55,000 (55K) was found recently by immunoprecipitation in all SV40 virus-transformed mammalian cells, in addition to the SV40 large T antigen (appoximately 94K) and small antigen (approximately 17K), which are the only proteins coded by the 'early half' of the SV40 genome. The 55K protein is encoded by cellular DNA; its peptide pattern is different from that of the SV40 antigens and it is species specific in mouse, rat, hamster, monkey and human SV40-transformed (or infected) cells. A 55K protein with a similar peptide pattern was found in mouse embryonal carcinoma cells not exposed to SV40. Similar proteins were reported in mouse sarcomas and leukaemias induced by a great variety of aetiological agents and also in a spontaneously transformed mouse fibroblast cell line, and it has been suggested that the protein may be a general correlated of cellular tumorigenicity. We now report that the approximately 55K protein is present in primary cell cultures from 12-14 day old mouse embryos, but not in 16-day old mouse embryos. The embryo protein has a peptide pattern virtually indistinguishable from that of the SV40-induced protein. We also show by comparing closely related cell families that spontaneously transformed highly tumorigenic mouse cells do not possess the 55K protein.     

Cooperation between gene encoding p53 tumour antigen and ras in cellular transformation

The protein p53 is highly expressed in a large variety of transformed cell types originating from diverse species. These include cells transformed by Simian virus 40 (SV40), adenovirus and Abelson virus, as well as a variety of chemically transformed cells. Substantial amounts of p53 are also present in certain non-transformed cells, for example, some embryonic tissues. The protein may be localized in different cellular compartments in normal and transformed cells. The strong correlation between tumorigenicity and high levels of p53 suggests an important role of p53 in tumorigenesis. We report here experiments in which we have co-transfected the murine cellular gene encoding for p53 with a ras gene into primary rat embryo fibroblasts. Our results indicate that the p53-encoding gene can play a causal role in the conversion of normal fibroblasts into tumorigenic cells.     

Virus-mediated killing of cells that lack p53 activity

A major goal of molecular oncology is to identify means to kill cells lacking p53 function. Most current cancer therapy is based on damaging cellular DNA by irradiation or chemicals. Recent reports support the notion that, in the event of DNA damage, the p53 tumour-suppressor protein is able to prevent cell death by sustaining an arrest of the cell cycle at the G2 phase. We report here that adeno-associated virus (AAV) selectively induces apoptosis in cells that lack active p53. Cells with intact p53 activity are not killed but undergo arrest in the G2 phase of the cell cycle. This arrest is characterized by an increase in p53 activity and p21 levels and by the targeted destruction of CDC25C. Neither cell killing nor arrest depends upon AAV-encoded proteins. Rather, AAV DNA, which is single-stranded with hairpin structures at both ends, elicits in cells a DNA damage response that, in the absence of active p53, leads to cell death. AAV inhibits tumour growth in mice. Thus viruses can be used to deliver DNA of unusual structure into cells to trigger a DNA damage response without damaging cellular DNA and to selectively eliminate those cells lacking p53 activity.     

DNA-dependent protein kinase is not required for the p53-dependent response to DNA damage.

Damage to DNA in the cell activates the tumour-suppressor protein p53, and failure of this activation leads to genetic instability and a predisposition to cancer. It is therefore crucial to understand the signal transduction mechanisms that connect DNA damage with p53 activation. The enzyme known as DNA-dependent protein kinase (DNA-PK) has been proposed to be an essential activator of p53, but the evidence for its involvement in this pathway is controversial. We now show that the p53 response is fully functional in primary mouse embryonic fibroblasts lacking DNA-PK: irradiation-induced DNA damage in these defective fibroblasts induces a normal response of p53 accumulation, phosphorylation of a p53 serine residue at position 15, nuclear localization and binding to DNA of p53. The upregulation of p53-target genes and cell-cycle arrest also occur normally. The DNA-PK-deficient cell line SCGR11 contains a homozygous mutation in the DNA-binding domain of p53, which may explain the defective response by p53 reported in this line. Our results indicate that DNA-PK activity is not required for cells to mount a p53-dependent response to DNA damage.     

Initiation of a G2/M checkpoint after ultraviolet radiation requires p38 kinase

Response to genotoxic stress can be considered as a multistage process involving initiation of cell-cycle arrest and maintenance of arrest during DNA repair. Although maintenance of G2/M checkpoints is known to involve Chk1, Chk2/Rad53 and upstream components, the mechanisms involved in its initiation are less well defined. Here we report that p38 kinase has a critical role in the initiation of a G2 delay after ultraviolet radiation. Inhibition of p38 blocks the rapid initiation of this checkpoint in both human and murine cells after ultraviolet radiation. In vitro, p38 binds and phosphorylates Cdc25B at serines 309 and 361, and Cdc25C at serine 216; phosphorylation of these residues is required for binding to 14-3-3 proteins. In vivo, inhibition of p38 prevents both phosphorylation of Cdc25B at serine 309 and 14-3-3 binding after ultraviolet radiation, and mutation of this site is sufficient to inhibit the checkpoint initiation. In contrast, in vivo Cdc25C binding to 14-3-3 is not affected by p38 inhibition after ultraviolet radiation. We propose that regulation of Cdc25B phosphorylation by p38 is a critical event for initiating the G2/M checkpoint after ultraviolet radiation.     

ATP-dependent assembly of double hexamers of SV40 T antigen at the viral origin of DNA replication

Simian virus 40 (SV40) replicates in nuclei of human and monkey cells. One viral protein, large tumour (T) antigen, is required for the initiation of DNA replication. The development of in vitro replication systems which retain this property has facilitated the identification of the cellular components required for replication. T antigen recognizes the pentanucleotide 5'-GAGGC-3' which is present in four copies within the 64 base-pairs (bp) of the core origin. In the presence of ATP it binds with increased affinity forming a distinctive, bilobed structure visible in electron micrographs. As a helicase, it unwinds SV40 DNA bidirectionally from the origin. We report here that in vitro and in the presence of ATP, T antigen assembles a double hexamer, centred on the core origin and extending beyond it by 12 bp in each direction. The assembly of this dodecamer initiates an untwisting of the duplex by 2-3 turns. In the absence of ATP, a tetrameric structure is the largest found at the core origin. In the absence of DNA, but in the presence of ATP or its non-hydrolysable analogues, T antigen assembles into hexamers. This suggests that ATP effects an allosteric change in the monomer. The change alters protein-protein interactions and allows the assembly of a double hexamer, which initiates replication at the core origin.     

Functional identity of proliferating cell nuclear antigen and a DNA polymerase-delta auxiliary protein

The mechanism of replication of the simian virus 40 (SV40) genome closely resembles that of cellular chromosomes, thereby providing an excellent model system for examining the enzymatic requirements for DNA replication. Only one viral gene product, the large tumour antigen (large-T antigen), is required for viral replication, so the majority of replication enzymes must be cellular. Indeed, a number of enzymatic activities associated with replication and the S phase of the cell cycle are induced upon SV40 infection. Cell-free extracts derived from human cells, when supplemented with immunopurified SV40 large-T antigen support efficient replication of plasmids that contain the SV40 origin of DNA replication. Using this system, a cellular protein of relative molecular mass 36,000 (Mr = 36K) that is required for the elongation stage of SV40 DNA replication in vitro has been purified and identified as a known cell-cycle regulated protein, alternatively called the proliferating cell nuclear antigen (PCNA) or cyclin. It was noticed that, in its physical characteristics, PCNA closely resembles a protein that regulates the activity of calf thymus DNA polymerase-delta. Here we show that PCNA and the polymerase-delta auxiliary protein have similar electrophoretic behaviour and are both recognized by anti-PCNA human autoantibodies. More importantly, both proteins are functionally equivalent; they stimulate SV40 DNA replication in vitro and increase the processivity of calf thymus DNA polymerase-delta. These results implicate a novel animal cell DNA polymerase, DNA polymerase-delta, in the elongation stage of replicative DNA synthesis in vitro.     

Degradation of the cyclin-dependent-kinase inhibitor p27Kip1 is instigated by Jab1

The proliferation of mammalian cells is under strict control, and the cyclin-dependent-kinase inhibitory protein p27Kip1 is an essential participant in this regulation both in vitro and in vivo. Although mutations in p27Kip1 are rarely found in human tumours, reduced expression of the protein correlates well with poor survival among patients with breast or colorectal carcinomas, suggesting that disruption of the p27Kip1 regulatory mechanisms contributes to neoplasia. The abundance of p27Kip1 in the cell is determined either at or after translation, for example as a result of phosphorylation by cyclinE/Cdk2 complexes, degradation by the ubiquitin/proteasome pathway, sequestration by unknown Myc-inducible proteins, binding to cyclinD/Cdk4 complexes, or inactivation by the viral E1A oncoprotein. We have found that a mouse 38K protein (p38) encoded by the Jab1 gene interacts specifically with p27Kip1 and show here that overexpression of p38 in mammalian cells causes the translocation of p27Kip1 from the nucleus to the cytoplasm, decreasing the amount of p27Kip1 in the cell by accelerating its degradation. Ectopic expression of p38 in mouse fibroblasts partially overcomes p27Kip1-mediated arrest in the G1 phase of the cell cycle and markedly reduces their dependence on serum. Our findings indicate that p38 functions as a negative regulator of p27Kip1 by promoting its degradation.     

SV40 enhancer and large-T antigen are instrumental in development of choroid plexus tumours in transgenic mice

We have shown recently that choroid plexus tumours frequently develop in transgenic mice which have developed from fertilized eggs injected with DNA molecules containing both simian virus 40 (SV40) early-region genes and metallothionein (MT) fusion genes, and several lines of mice have now been established in which all of the offspring that inherit the foreign DNA succumb to these tumours at 3-5 months of age (ref. 1 and our unpublished data). Several other tissues, notably thymus and kidney, occasionally also show pathological changes. SV40 large-T antigen protein and messenger RNA are always present in affected tissues at much greater concentrations than in unaffected tissues, suggesting that SV40 early-region genes are preferentially activated in choroid plexus, thymus and kidney and that this activation frequently leads to tumorigenesis in the choroid plexus. To determine which regions of the original constructs are important for this tumorigenesis, we have now tested several derivatives and report here that the large-T antigen is sufficient, that the MT fusion gene is dispensable and that the SV40 enhancer (72-base-pair repeat region) has an important role in directing tumours to the choroid plexus. Deletion of the SV40 enhancer region alone commonly leads to peripheral neuropathy, as well as liver and pancreatic tumours, which are the subject of the accompanying paper. Evidence is presented that these pathologies may result from an enhancing effect of the MT sequences on large-T antigen genes, made possible by removal of the otherwise dominant SV40 enhancer.