An absence of p53 protein binding to consensus sites on the p21 and GADD45 genes revealed by in vivo analyses on normal fibroblasts irradiated with ionizing radiation Free

1253

Introduction: The main pathway of p53 transcriptional activity is mediated through direct binding of the p53 protein to its DNA consensus sequence and the minor pathway is mediated by protein-protein interactions. It is believed that mutations in the DNA-binding domain of the p53 protein abolish its functions such as p21 activation. In contrast, we have demonstrated that a subset of p53 mutant proteins activates p21 expression. How can wild-type and mutated p53 proteins induce p21 expression without binding to its response element (RE)? Objective: We have investigated other mechanisms used by p53 to regulate the expression of two of its effector genes. Methods: Human fibroblasts were exposed to 20 Grays of gamma irradiation. At different time points (0, 1, 2, 4 and 8 h), cells were treated as follows: 1) With footprinting agents (DNaseI and UVC), DNA was purified, and ligation-mediated PCR (LMPCR) was carried out for the p21 and GADD45 genes; and 2) The Chromatin Immunoprecipitation (ChIP) technique, cells were fixed with 37% formaldehyde, chromatin was sheared by sonication, and the fragments associated with p53 were immunoprecipitated with p53 antibodies, and the fragments associated with Sp1 were immunoprecipitated with Sp1 antibodies. The presence of DNA fragments corresponding to the p53 RE on the p21 promoter, and the consensus sequence of Sp1 among co-immunoprecipitated DNA, were revealed by PCR using specific primers. Results: The p53 kinetic induction after UVB and Gamma irradiation show an increase of p53 expression. In vivo analysis using LMPCR and ChIP showed that there is no evidence in living cells of direct p53 binding to its consensus sequence of p21 and GADD45 genes in living cells. There is no direct binding of p53 to its first RE on the p21 promoter, a small interaction on its second RE, and a very strong binding of the p53/SP1 complex to the Sp1 consensus sequence. Also, with LMPCR technique, we showed that there is no footprint of p53 to its RE on the third GADD45 intron and a strong footprint of WT1 with its consensus sequence on the GADD45 promoter. Conclusion: There was no evidence for p53 binding to its consensus sequence under physiological conditions. All these data strongly suggest that p53 forms a complex with Sp1 or WT1, which are already bound to their consensus sequence on the p21 and GADD45 promoters, to induce their transcription. We propose that the main mechanism of p53 transcriptional regulation of p21 and GADD45 does not involve direct binding to the consensus sequence on these two effectors, but rather is mediated through protein-protein interactions with a transcription factor already bound to its consensus sequence.