The p63 gene encodes at least six different proteins with homology to the tumour suppressor protein p53 and the related p53 family member p73. So far, there have been limited data concerning the expression patterns of individual p63 proteins, due to a lack of reagents that distinguish between the different isoforms. Three antibodies have been produced specifically directed against the two N-terminal isoforms (TAp63 and DeltaNp63) and the C-terminal region of the p63alpha proteins. TAp63 proteins are located suprabasally in stratified epithelia compared with the N-terminal truncated forms, which are more abundantly expressed in the basal cell layer, indicating a switch in expression of p63 isoforms during normal cellular differentiation. Analysis of squamous cell carcinomas shows DeltaNp63alpha to be the most widely expressed isoform, compatible with a role for this protein in promoting neoplastic cell growth in these tissues. DeltaNp63 protein expression is also restricted to basal cells in breast and prostate, whilst TAp63 isoforms are more widely expressed in these tissues as well as in tumours at these sites. TAp63, but not DeltaNp63 or p63alpha, is detected in normal colon and in colon carcinoma. TAp63 proteins are also expressed in the nuclei of a sub-population of lymphoid cells and in most malignant lymphomas, whereas DeltaNp63 proteins are not expressed. Taken together, a hitherto unrecognized regulation of p63 isoform expression in vivo has been uncovered, with different p63 proteins expressed during differentiation and in different cell types. The data indicate roles for specific p63 isoforms not only in maintaining epithelial stem cell populations, but also in cellular differentiation and neoplasia.

p53 missense mutant proteins commonly show increased stability compared to wild-type p53, which is thought to depend largely on the inability of mutant p53 to induce the ubiquitin ligase MDM2. However, recent work using mouse models has shown that the accumulation of mutant p53 occurs only in tumour cells, indicating that stabilization requires additional factors. To clarify the stabilization of p53 mutants in tumours, we analysed factors that affect their folding and degradation. Although all missense mutants that we studied are more stable than wild-type p53, the levels correlate with individual structural characteristics, which may be reflected in different gainof-function properties. In the absence of Hsp90 activity, the less stable unfolded p53 mutants preferentially associate in a complex with Hsp70 and CHIP (carboxy terminus of Hsp70-interacting protein), and we show that CHIP is responsible for ubiquitination and degradation of these mutants. The demonstration of a complex interplay between Hsp90, Hsp70 and CHIP that regulate the stability of different p53 mutant proteins improves our understanding of the pro-tumorigenic effects of increased Hsp90 activity during multi-stage carcinogenesis. Understanding the roles of Hsp90, Hsp70 and CHIP in cancers may also provide an important avenue through which to target p53 to enhance treatment of human cancers.

P53 controls the growth and survival of cells by acting in response to a multitude of cellular stresses. It is, however, not yet fully understood how different p53 activation pathways result in either cell cycle arrest or apoptosis. We and others have described an N-terminally truncated p53 protein (p53/47) originating from a second translation initiation site in the p53 messenger RNA (mRNA), which can interact with p53 and impose altered stability and transactivation properties to p53 complexes. Here we show that cap-dependent and cap-independent mechanisms of initiation govern the translation of the p53 mRNA. Changes in synthesis of full-length p53 or p53/47 are regulated through distinct cell stress-induced pathways acting through separate regions of the p53 mRNA. We also show that some cytotoxic drugs require the presence of full-length p53 to induce apoptosis, whereas for others p53/47 is sufficient. This indicates that by harbouring alternative translation initiation sites, the p53 mRNA gives rise to different levels of the p53 isoforms which help to orchestrate the cell biological outcome of p53 activation in response to different types of cell stress. This sheds new light into the way p53 can integrate and differentiate a large multiplicity of changes in the cellular environment.

We have examined p53 protein levels in cell lines selected for resistance to the chemotherapeutic drug cis-diamminedichloroplatinum (II), cisplatin. The majority of the independent cisplatin-resistant clones isolated by a single selection with cisplatin from the ovarian tumour cell line A2780 showed increased levels of p53 protein compared to the parental cell line. Elevated p53 protein levels were also observed in cisplatin-resistant ovarian human tumour lines isolated after multiple exposures to cisplatin (A2780/cp70 and OVIP/DDP). Direct PCR sequencing of p53 cDNAs showed that both the A2780/cp70 and the parental A2780 cell lines had a wild-type p53 gene sequence. The OVIP and OVIP/DDP lines both had a heterozygous mutation at codon 126. Cell-cycle analysis after gamma-irradiation or cisplatin treatment showed evidence of a G1/S and G2/M cell-cycle checkpoint in both A2780/cp70 and the sensitive parental cell lines. However, the resistant cell line A2780/cp70 showed less inhibition of DNA synthesis after gamma-irradiation than the sensitive cell line. Transfection of a mutant p53 gene construct (containing a mutation at codon 143, val to ala) into the A2780/cp70 resistant cells conferred a significantly increased sensitivity to cisplatin, suggesting that p53 is a direct determinant of cisplatin resistance in these cells. However, expression of this mutant p53 in the A2780 cells did not affect sensitivity.

The ability to separate the isoforms of human tumour suppressor protein p53 expressed in insect cells using heparin-Sepharose correlates with differences in the isoelectric point of p53, demonstrating that p53 can be heterogeneously modified and providing support for the use of insect cells as a model system for identifying novel signalling pathways that target p53. One p53 isoform that was reduced in its binding to the monoclonal antibody DO-1 could be stimulated in its binding to DO-1 by prior incubation with protein phosphatases, suggesting the presence of a previously unidentified N-terminal phosphorylation site capable of masking the DO-1 epitope. A synthetic peptide from the N-terminal domain of p53 containing phosphate at Ser(20) inhibited DO-1 binding, thus identifying the phosphorylation site responsible for DO-1 epitope masking. Monoclonal antibodies overlapping the DO-1 epitope were developed that are specific for phospho-Thr(18) (adjacent to the DO-1 epitope) and phospho-Ser(20) (within the DO-1 epitope) to determine whether direct evidence could be obtained for novel phosphorylation sites in human p53. A monoclonal antibody highly specific for phospho-Ser(20) detected significant phosphorylation of human p53 expressed in insect cells, whereas the relative proportion of p53 modified at Thr(18) was substantially lower. The relevance of these two novel phosphorylation sites to p53 regulation in human cells was made evident by the extensive phosphorylation of human p53 at Thr(18) and Ser(20) in a panel of human breast cancers with a wild-type p53 status. Phospho-Ser(20) or phospho-Thr(18) containing p53 peptides are as effective as the phospho-Ser(15) peptide at reducing mdm2 (mouse double minute 2) protein binding, indicating that the functional effects of these phosphorylation events might be to regulate the binding of heterologous proteins to p53. These results provide evidence in vivo for two novel phosphorylation sites within p53 at Ser(20) and Thr(18) that can affect p53 protein-protein interactions and indicate that some human cancers might have amplified one or more Ser(20) and Thr(18) kinase signalling cascades to modulate p53 activity.