The regulation of the tumour suppressor and transcription factor p53 is highly complex and involves different types of post-translational modification, including phosphorylation and ubiquitylation. Recent findings reported in Nature now add to the complexity, as Danny Reinberg and colleagues have shown that p53 can be specifically methylated, and that this regulates its stability and the expression of target genes.

The histone lysine methyltransferase Set9 is known to methylate Lys4 of histone H3 in vitro. While searching for other substrates for Set9, Reinberg and co-workers identified p53 as an in vitro substrate — other proteins that were tested did not serve as Set9 substrates, and alternative protein methyltransferases couldn't methylate p53. Importantly, however, other polypeptides that are present in HeLa-cell-derived extracts also served as substrates for Set9, and a recent report indicates that the transcription-initiation factor TAF10 can also be methylated by Set9. The p53 methylation site was mapped to a single Lys residue (Lys372) in the regulatory C-terminal region of p53, which is subjected to multiple post-translational modifications.

The affinity of Set9 for a p53 peptide was, in fact, higher than its affinity for an equivalent histone-H3 peptide. This led the authors to determine the structure of Set9 in complex with a monomethylated p53 peptide, and compare it with the previously solved complex of Set9 with a monomethylated histone-H3 peptide. Overall, the structures are very similar, and the Set9 residues that interact with the three amino-acid residues N-terminal to the methylated Lys residue are the same for the p53 peptide and the H3 peptide, despite the difference in peptide sequence. Reinberg and colleagues speculate that more distant residues in p53 might contribute to the substrate specificity.

To study the significance of Set9-methylation of p53 in vivo, the authors stably transfected cells that expressed endogenous p53 with wild-type Set9 or with a methylation-defective Set9 mutant. Using an antibody that is specific for Lys372-methylated p53, methylated p53 could be detected in extracts from the former cells but not the latter. When untransfected cells were treated with an anti-cancer drug that induces DNA damage and therefore a p53-responsive pathway, the amount of methylated p53 was increased compared with untreated cells. However, by introducing Set9 small interfering (si)RNA into these cells, the level of methylated p53 decreased after chemical treatment. So, Set9 methylates p53 in vivo and Set9-specific methylation of p53 occurs in response to DNA damage.

Next, the authors found that methylated p53 is exclusively nuclear, which drew their attention to the transcriptional activity of p53. Focusing on a well-characterized transcriptional target of p53 — p21 — they showed that the increased expression of p21 correlated with elevated levels of methylated p53. In addition, the induction of DNA damage further increased the level of p21 expression. By contrast, overexpression of catalytically inactive mutant Set9, or the siRNA-mediated depletion of Set9, impaired p21 expression and decreased the total level of p53.

This latter finding suggested to Reinberg and colleagues that methylation of p53 might affect its stability. Indeed, the stability of p53 increased in cells that expressed wild-type Set9, but not in those that expressed the methylation-defective mutant. Consistent with this 'hyper-stabilization' of p53, cells that overexpressed Set9 also showed higher apoptotic staining, which is indicative of a stronger p53-dependent apoptotic response.

The in vivo role of methylation in p53 regulation implies complex regulatory mechanisms, the coordination of which remains unclear. In addition, it raises the question of how the methylation of a single residue might interfere with the ubiquitylation and subsequent degradation of p53.