Chromatin immunoprecipitation analysis fails to support the latency model for regulation of p53 DNA binding activity in vivo

p53 can adopt two forms in vitro, a latent form that binds naked DNA poorly and an active form that binds DNA well. Conversion of the latent form to the active form is thought to occur by an allosteric mechanism induced by phosphorylation and acetylation. Despite the large differences in affinity produced by regulatory modifications in vitro, mutation of putative regulatory sites has not produced correspondingly large effects on transcription of p53 target genes in vivo. To determine whether genotoxic stress regulates DNA binding by p53 in vivo, we have performed quantitative chromatin immunoprecipitation (ChIP) assays on tumor and normal cell lines containing wild-type p53. ChIP recovers several hundredfold more p21 and MDM2 promoter DNA from p53 wild-type than p53-null cells, indicating that the assay is specific for p53. Genotoxic stress induces much smaller increases in chromatin precipitation, which are matched by changes in the p53 protein level. Thus, in the experimental systems tested, allosteric regulation of DNA binding is not a major level of regulation of p53 activity. The p53 target genes tested can be divided into a group showing high promoter occupancy in vivo (p21, MDM2, and PUMA) and a group giving substantially weaker or background p53 binding (bax, AIP1, and PIG3). Neither group shows selective recruitment of p53 to the promoter in cells undergoing apoptosis, indicating that the decision to undergo apoptosis or cell cycle arrest depends on other changes in the cell.