FUNCTIONAL ACTIVATION OF THE p53 TUMOR SUPPRESSOR IN NON- TUMORIGENIC VARIANTS OF THE HELA CERVICAL CARCINOMA CELL LINE

Two non-transformed revertants of the HeLa (ATCC CCL2) cell line have been isolated using a selection procedure based on prolonged retention of the fluorescent dye, rhodamine 123, within the mitochondria of carcinoma cells versus normal epithelial cells. Unlike the parental HeLa, the revertant cells exhibited a flat non-refractile morphology, failed to grow in suspension culture, had more than 100-fold lower cloning efficiencies in semi-solid medium, and failed to induce subcutaneous tumors when injected into nude mice. Both clones have retained a revertant phenotype after more than three years of continuous culture. Somatic cell fusion experiments suggested that revertant cells had sustained a mutation or mutations that resulted in the activation of a dominant suppressor of transformation. Molecular characterization of these revertants demonstrated that the HPV-18 sequences present in the parental HeLa cells are retained in the revertants. Expression of the HPV-18 E6 and E7 at both the mRNA and protein levels were comparable to those seen in the parental HeLa line, suggesting that loss of the viral oncogenes was not responsible for loss of the tumorigenic phenotype in either clone. Among other possibilities, it was reasoned that a change in a regulator of the HPV oncoproteins or of the Rb and p53 genes could have interrupted the oncogene-suppressor association and have thus led to an activation of either Rb or p53. Using immunoaffinity purification and immunoblotting experiments no difference in Rb steady state levels were observed but increased p53 protein levels were found in both revertants as compared to Hela.. It is shown that functional activation of the p53 pathway is an obligatory step in the reversion process. The HeLa cells express wild-type p53 whose protein levels are maintained too low for its growth supressory functions to be manifested, due to the presence and function of the E6 protein. Increased steady state p53 levels in the revertants despite the presence of the E6 protein correlate precisely with the reversion of the transformed phenotype. Ectopic expression of the wild-type p53 gene by retroviral integration was shown to decrease the clonigenicity if the Hela cell in soft agar by a hundred fold. The p53 gene in HeLa ATCC CCL2 has been previously characterized as wild-type. Mutation in the p53 cDNA's from HeLa and revertants has not been detected by RNAase protection assay or direct sequencing. Increased protein levels of p53 protein in the revertants as compared to HeLa were thus anticipated and found to result in a correspondingly differential manifestation of wild-type p53 function between the three cell lines. First, in the revertants as in non-transformed cells carrying the wild-type p53 gene, p53 protein is inducible upon treatment with Actinomycin D and yirradiation with kinetics of induction and induced p53 protein levels both being at least ten-fold higher than in Hela. Secondly, p53-specific transcriptional activity is detectable in the revertants. Specifically, transcription of the chloramphenicol acetylatransferase (CAT) gene driven from a promoter with multiple p53 consensus binding sites and a TATA box is detectable by CAT enzyme assay only in the revertants and not in HeLa. Treatment with Actinomycin D and yirradiation inreases the transactivating activity of endogenous p53 in the revertants thirty-fold. Moreover, downstream effectors of the transcriptionally active p53, namely the waf-l and gadd45 genes, are expressed in the revertants but not in HeLa. Expression of the gadd-45 mRNA is constitutive in the revertants while waf-1 mRNA is inducible by irradiation. p53-specific DNA binding activity in EMSA assays is also inducible by y-irradiation and actinomycin D in the revertants and not in HeLa cells. Taken together, studies of the functional properties of endogenous p53 in revertants and HeLa correlate the transcriptional activity of p53 with the reversion from cervical carcinoma and offer valuable evidence that the transcriptional properties of the wild-type p53 contribute to its role in growth regulation. An in vitro ubiquitination-degradation system was used to investigate the causes of increased p53 protein in the revertants. It was shown that differential targeting of the p53 protein for proteolysis occurs between HeLa and revertants since ubiquitination-dependent degradation of in vitro translated p53 protein is faster upon incubation with HeLa than with revertant cell lysate. This differential degradation of p53 in revertant cell extracts resulted from incomplete ubiquitination of the p53 protein. Protein-protein interaction studies in HeLa and revertants showed that p53-E6 binding in HeLa but not ine either revertant suggesting that the differential degradation of p53 in he revertants was caused by loss of E6 function. Furthermore p53 in HeLa and HA was associated with the cell cycle regulator p34cdc2, while in the HF cell line this association was undetectable. p34cdc2 has been previously characterized as a modulator of p53 phosphorylation both in vitro and in vivo and phosphorylation of p53 at the p34cdc2 site has been found necessary for the ubiquitin-dependent degradation of the p53 protein. Thus, in the HF cell line, differential degradation of endogenous p53 may result from the failure of p53 to associate with the cell cycle regulator p34cdc2 and be phosphorylated by it. Further characterization of the HeLa revertants included protein-protein interaction studies in the Rb pathway that showed differential binding of Rb to the transcription factor E2FI. Cell-cycle dependent binding and inhibition of this transcription factor is a known growth suppressory function of Rb. This result suggests that a common regulator of the p53 and Rb pathway may have been the target of the reversion event in this cell line. In order to exclude that reversion had resulted from E6 and/or E7 mutation we expressed ectopically by retroviral integration HPV-16 E6, E7 or a combination of the two and observed no effect on the revertant phenotype suggesting that the reversion event has bypassed the E6 and E7 functional pathways. Ectopic expression of SV40 Large T antigen, however, increased the clonigenicity in soft agar of the HA but not the HF cell line by -fold. Hence SV40 large T possesses a biochemical function that can override the cause of reversion in the HA cell line. Taken together these results clearly establish that functional activation of p53 in two independent HeLa revertants is the result of change in two different pathways. They also show the great promise of the HeLa revertant system as a tool for the study of cervical transformation and for the involvement of the p53 pathway in this type of transformation. The process of identification of the molecular events that led to p53 and Rb activation in the revertants will provide valuable insight into the interplay of factors important for cell growth and cell growth suppression, and may identify potential targets for cancer therapy.