The p53 tumor suppressor is a homotetrameric, sequence-specific transcription factor that has crucial roles in apoptosis, cell cycle arrest, cellular senescence, and DNA repair. It is maintained at low levels in unstressed cells, but stabilized and activated following DNA damage through extensive post-translational modification (PTM). Our research has focused on identifying and exploring the biological roles of p53 PTMs to better understand how they modulate p53 function. Global effects of p53 PTMWe have used mouse models containing missense mutations at p53 PTM sites to investigate the complex effects of p53 PTMs in a physiological setting. Knock-in mice were generated containing mutation of Ser18 (Ser15 in humans) to alanine in both alleles of endogenous p53, thereby preventing phosphorylation of this site. Quantitative mass spectrometry analysis of wild type or p53S18A thymocytes was performed to investigate the role of this modification in the response of p53 to ionizing radiation (IR). The primary effect of the p53S18A mutation was a loss of wild type response to IR. Among those proteins that were differently affected were several pro-apoptotic proteins, which increased in protein level after IR in the wild type and were either unaffected by IR in the mutant or showed less increase than in the wild type. A second group of proteins that was differently affected by the mutation contains proteins with roles in energy and metabolism pathways. For example, two proteins that are critical for oxidative phosphorylation, a process that is promoted by p53, were both found to increase after IR in the wild type but were unaffected by IR in the mutant. The role of p53 in energy pathways is only recently becoming known, and these studies critically highlight new functions for p53 regulation by post-translational modification. We are currently initiating studies to further understand the modulation of p53-dependent effects on metabolism in these knock-in mice. Effects of p53 N-terminal phosphorylation on its protein-protein interactionsOne of the naturally expressed isoforms of p53, deltaNp53, lacks the first transactivation domain (TAD1) of p53 but does contain the second transactivation domain (TAD2). The expression and stability of the two proteins are affected differently by cell type, cell cycle phase and exposure to various stresses. p53 and deltaNp53 form heterotetramers and the relative abundance of deltaNp53 influences the transactivation activity and target gene specificity of p53. Our characterization of the binding of TAD1 and TAD2 of p53 to the Taz2 domain of the transcriptional coactivator p300 demonstrated that although the two domains bound to Taz2 with equal affinity, the binding of TAD1 was affected by p53 phosphorylations, whereas the binding of TAD2 was unaffected. To better understand the differences between the complexes of Taz2 with TAD1 and TAD2, we have determined the solution structure of a p53 TAD2 peptide in complex with Taz2. p53 TAD2 binds to a similar region on Taz2 and also forms a short alpha-helix upon binding, exposing hydrophobic residues to form the primary stabilizing interactions with Taz2. Additionally, mutagenesis experiments within p53 TAD2 suggest that, although we did not previously see an effect on the binding affinity when Thr55 was phosphorylated, mutation of this site to alanine did substantially alter the conformation of the complex. The differences in these results suggest that Thr55 may play a role in the kinetics of binding rather than the equilibrium affinity. Furthermore, comparison of the structures of the two complexes will shed light on how these two similar domains within p53 may function differently in co-activator recruitment after stress and suggest reasons for differences in transactivation between p53 and deltaNp53.Functional effects and interplay of p53 C-terminal modificationsThe C-terminus of p53 exhibits a diverse array of post-translational modifications, including phosphorylation, methylation, acetylation, ubiquitinylation, sumoylation, and neddylation, that are primarily localized to the terminal thirty residues. We are interested in understanding the specific effects of these site-specific modifications and the interplay between them. We have investigated the effects of mono- and dimethylation of p53 Lys382, a site that alternatively can be methylated, acetylated, or ubiquitinylated. Mono-methylation of p53 Lys382 results in repression of the activity of p53 as a transcription factor and we have continued to investigate the mechanism of repression. Dimethylation of p53 Lys382 is critical for the interaction of p53 with the tandem Tudor domain (TD) of the DNA damage response mediator 53BP1. We are currently exploring the role of additional modifications within the C-terminal regulatory domain that may modulate the binding of p53 to the TD domain. Since lysines can be acetylated or methylated, we synthesized a peptide containing acetylated lysine 381 and dimethylated lysine 382 and determined that the affinity for the TD domain increased. Moreover, in contrast to the crystal structure of the 53BP1 TD in complex with the p53 Lys382 dimethylated peptide, in which the p53 residues were not well resolved, the new structure shows binding of the p53 peptide in a single conformation with clear identification of several interacting residues. Further experiments will provide insight into the interactions of 53BP1 with p53 that facilitate repair of DNA damage.In addition to full-length p53 protein, the human TP53 gene encodes at least thirteen p53 protein isoforms. Among them, delta133p53, a p53 isoform that lacks the N-terminal 132 amino acid residues, inhibits the activity of full-length p53 and was downregulated during replicative senescence in human fibroblasts (Fujita et al. Nat Cell Biol 11, 1135-1142, 2009). Unlike full-length p53 protein, which is subject to proteasomal degradation, delta133p53 was not stabilized by treatment with a proteasome inhibitor. These findings prompted Drs. Fujita, and Harris to investigate whether the protein turnover of this p53 isoform is controlled by autophagy. Our data indicated that under conditions with activated autophagy, delta133p53 became ubiquitinated on lysines 381 and 382, the same lysine residues that are ubiquitinated in full-length p53. Since the N-terminal region of p53 that interacts with MDM2 is absent in delta133p53, loss of MDM2-mediated regulation of p53 may render ubiquitinated delta133p53 resistant to proteasomal degradation. Consistent with the absence of the MDM2 binding site in delta133p53, MDM2 was not detected in a mass spectrometry-based analysis of interacting proteins. These studies provide novel insight into the functional and regulatory connections between cellular senescence and autophagy.
Showing the most recent 10 out of 24 publications
