p53 and Genetic Instability
Cancers arise from individual cells by a multi-step process involving accumulation of multiple genetic lesions within single cells. Cells with the appropriate constellation of genetic changes to provide optimal growth in a particular microenvironment are then selected for clonal outgrowth. The probability of each mutation being generated is minimized by coordinating the activities of proteins involved in DNA synthesis, repair, and chromosome segregation with additional proteins that comprise "cell cycle checkpoints" which insure the completion of one process prior to initiating a downstream event (Wahl et al. 2005). Mutations that interfere with any facet of this machinery will increase the probability that errors will accumulate, and consequently elevate the rate of cellular evolution and cancer development.
It is now clear that mutations affecting the fidelity of DNA replication and repair, or that abrogate cell cycle checkpoints, are highly selected during cancer progression. For example, one of the most frequently mutated genes in human cancer encodes the p53 protein. p53 is involved in several cell cycle checkpoints that affect chromosome stability and insure transmission of an identical set of chromosomes to each daughter cell at mitosis.
The pathway, controlled by the product of the p53 gene, is among the most frequently targeted for inactivation in human cancers. More than half of human tumors contain inactivating mutations within the p53 gene itself, while the rest encode wild type p53 that is functionally compromised. We are intensely studying the mechanisms (Figure 1) that control p53 activation in order to develop therapeutic strategies to activate p53 in tumors where it remains wild type. We are also interested in understanding how p53 controls additional biological processes including cell cycle arrest, senescence, apoptosis and autophagy.
Mdm2, an E3 ubiquitin ligase that binds to p53 and mediates its proteasomal degradation, controls p53 abundance. Interestingly, Mdm2 might also reduce p53 transcription function by binding directly to the p53 transactivation domain (Michael and Oren 2003). A second p53 negative regulator highly related to Mdm2 is Mdm4 (also called Mdmx) (Figure 2). Mdm4 has no measurable E3-ligase activity, but binds to the same site on p53 as does Mdm2, so it too can antagonize p53 transactivation. In the absence of genotoxic stress, Mdm2 and Mdm4 restrict the basal activity of p53. Genetic studies show that Mdm2 and Mdm4 are essential negative regulators of p53, and even slight decreases in the abundance of either can create a hair-trigger for p53 activation (Marine and Jochemsen 2005; Marine et al. 2006). Underscoring their importance in p53 regulation, overexpression of the genes encoding either Mdm2 or Mdm4 is observed in tumors retaining the wild type p53 allele. While most models for p53 regulation consider Mdm2 and Mdm4 independently, it is possible that Mdm2 and Mdm4 may also work together to modulate p53 activation (Singh et al. 2007). Therefore, the stoichiometry between p53 and its negative regulators is an important determinant of p53 activation. However, there have been no quantitative analyses to show how the levels or subcellular distributions of these proteins vary with p53 during a stress response.
We have taken a two-pronged approach to understanding how p53 stability and activity is regulated. First, based on the sensitivity of p53 to small changes in Mdm2 and Mdm4 abundance, we use mouse models to preserve the stoichiometric relationships between these proteins. For example, we use homologous recombination to generate defined changes in each protein and perform in vivo structure-function analyses (Krummel et al. 2005; Toledo et al. 2006; Toledo et al. 2007). Such models enable the study of p53 activation and downstream consequences in different cell types, and offer the potential of temporal and tissue-specific regulation. Second, we recently completed the first quantitative kinetic analysis of p53 activation to generate more accurate mathematical models, and to assess whether the data from the mouse genetic and in vitro quantitative approaches are concordant (Wang et al. 2007).
p53 activation has been proposed to involve changes in its abundance and stability that are controlled by Mdm2. A contemporary model is that Mdm2-mediated ubiquitylation of highly conserved lysine residues in the p53 C-terminus targets p53 for degradation by the proteasome. Activation occurs by phosphorylation of N-terminal serine residues, which engenders a conformational switch to reduce Mdm2 binding, and favors binding of positive co-activators such as the histone acetyl transferase p300/CBP. P300/CBP then acetylates the same C-terminal lysines targeted for ubiquitylation by Mdm2. Thus, displacement of an ubiquitin ligase from the N-terminus by p300, and consequent acetylation of the C-terminus stabilizes and activates p53 (Brooks and Gu 2003). We generated a mouse mutant in which the C-terminal lysines were converted to arginine (7KR) to test this model. We also introduced proline to alanine mutations in the domain implicated in the stress-induced conformational switch.
We were surprised to see that p53 stability was identical to wild type in the 7KR mice, although the activity of 7KR was slightly higher than wild type p53 (Krummel et al. 2005). The biological functions of wild type p53 and 7KR in MEFs and thymocytes were identical. Importantly, however, a significant fraction of the mutant mice have aberrant cardiac phenotypes and develop radiation-induced cancers not observed in wild type mice. The stability and activity of the proline-alanine p53 mutant was also similar to wild type p53 (Toledo et al. 2007). These data reveal that these highly conserved sites for post-translational modification function as tissue-specific fine-tuning mechanisms rather than as on-off switches.
What, then, constitutes the on-off switch that mediates p53 activation? Quantitative kinetic studies combined with additional mouse mutant analyses reveal that an on-off switch requires the regulated degradation of both of p53’s negative regulators, Mdm2 and Mdm4. For example, we showed that an early step in p53 activation involves a damage-kinase induced phosphorylation of Mdm2 that promotes Mdm2 auto-degradation (Stommel and Wahl 2004). This rapid degradation triggers p53 stabilization and activation. Mdm2 also targets Mdm4 for degradation, and we found that reducing the Mdm4 level was important to unleash the tumor suppressive powers of p53. Specifically, we made a mouse expressing a hypomorphic p53 mutant that enabled us to obtain mice that lacked Mdm4, but expressed Mdm2 (an otherwise lethal genotype if wild type p53 is present). Analysis of these mice showed that Mdm4 inhibits p53 transcription more effectively than Mdm2 (Toledo et al. 2006). While unable to prevent oncogenes from inducing tumors, the tumor suppression function of this mutant was restored by Mdm4 deletion. The lack of Mdm4 probably restores the ability of the mutant to induce genes involved in cell cycle arrest and apoptosis.
Our quantitative kinetic studies have also generated some surprises, and they point out the importance of such an approach for analysis of the p53 pathway (Wang et al. 2007). For example, we predicted that altering Mdm4 levels would profoundly affect p53 activation kinetics after DNA damage, but found that this was not the case. We inferred from this, that DNA damage reduces Mdm2 and Mdm4 stability so that they are no longer effective p53 antagonists, and that Mdm2 can still degrade Mdm4, even when the latter is overexpressed by 5-10 fold. Our analyses, as well as those of other labs, indicate that p53 activation results from accelerated Mdm2 auto-degradation, Mdm2-mediated degradation of Mdm4, and p53 stabilization. The increased p53 then transcriptionally activates the Mdm2 gene. This creates a positive feedback loop in which elevated Mdm2 levels (and increased E3 ligase activity) enhance Mdm4 degradation, thereby further increasing p53 transactivation. This provides an exquisite mechanism that couples p53 transcriptional output to the degree of stress sensed by the cell.
Current Directions and Future Prospects
We are pursuing multiple directions to understand p53 regulation further, and to translate our knowledge into potentially new chemotherapeutic agents.
An important goal is to understand the degree to which regulated Mdm4 degradation contributes to p53 activation. To this end, we are generating mice that express mutant Mdm4 protein that is resistant to DNA damage- or Mdm2-mediated destruction. We have already made one knock-in mouse expressing an Mdm4 mutant that cannot be phosphorylated by damage-activated kinases (Figure 3). Analysis of the properties of this mutant is underway. Other knock-in mutations designed to evaluate the role of Mdm2 and Mdm4 in p53 regulation are also being developed.
A second goal is to explore the mechanisms leading to life or death decisions by p53. It is well known that in response to the same stress, p53 can induce programs to elicit apoptosis or a senescent-like arrest. The decision depends on the cell type, and the stress. We have found that Mdm4 is at least one component involved in these decisions (Wade et al. 2006). Thus, in cells that do not undergo apoptosis in response to p53 activation by the chemical agonist Nutlin (which interferes with p53-Mdm2 interaction, see Figure 4), we can make the cells undergo apoptosis by reducing Mdm4 levels using siRNA.
The mechanism by which the switch in cell fate is made is unclear, but could be very important for the design of more effective p53 agonists. We have also observed that induction of Mdm2 and degradation of Mdm4 correlates with induction of apoptosis in some cell lines. By contrast, other cell lines treated similarly induce Mdm2, but fail to degrade Mdm4 and do not die, but undergo a senescent-like arrest. This indicates that the Mdm2 level per se is not the sole determinant of Mdm4 degradation. Indeed, this question exposes how little we understand about the factors involved in proteasomal degradation of Mdm4. We have therefore designed an siRNA screen to identify the kinases, proteasomal components, and other factors that participate in Mdm4 degradation.
The third area of investigation involves designing chemical screens for compounds to activate p53 by disrupting interactions with Mdm2 and Mdm4. While this presents significant challenges, it also comes with the exciting possibility of identifying potentially new therapeutic agents. To be able to evaluate the effectiveness of such agents in vivo, we are building mouse models to report p53 activation in different cell types. We are also developing mouse models of human cancers in which over-expression of Mdm2 or Mdm4 is suggested to have a critical role. This will enable us to test the in vivo efficacy of novel agents identified during our screening projects.
- Brooks, C. L. and Gu, W. 2003. Ubiquitination, phosphorylation and acetylation: the molecular basis for p53 regulation. Current Opinion in Cell Biology 15: 164.
- Krummel, K. A., Lee, C. J., Toledo, F. and Wahl, G. M. 2005. The C-terminal lysines fine-tune P53 stress responses in a mouse model but are not required for stability control or transactivation. Proc Natl Acad Sci U S A 102: 10188-93.
- Marine, J.-C., Francoz, S., Maetens, M., Wahl, G., Toledo, F. and Lozano, G. 2006. Keeping p53 in check: essential and synergistic functions of Mdm2 and Mdm4. Cell Death and Differentiation 13: 927–934.
- Marine, J. C. and Jochemsen, A. G. 2005. Mdmx as an essential regulator of p53 activity. Biochem Biophys Res Commun 331: 750-60.
- Michael, D. and Oren, M. 2003. The p53-Mdm2 module and the ubiquitin system. Semin Cancer Biol 13: 49-58.
- Singh, R. K., Iyappan, S. and Scheffner, M. 2007. Hetero-oligomerization with MdmX Rescues the Ubiquitin/Nedd8 Ligase Activity of RING Finger Mutants of Mdm2. J. Biol. Chem. 282: 10901-10907.
- Stommel, J. M. and Wahl, G. M. 2004. Accelerated MDM2 auto-degradation induced by DNA-damage kinases is required for p53 activation. Embo J 23: 1547-56.
- Toledo, F., Krummel, K. A., Lee, C. J., Liu, C.-W., Rodewald, L.-W., Tang, M. and Wahl, G. M. 2006. A mouse p53 mutant lacking the proline-rich domain rescues Mdm4 deficiency and provides insight into the Mdm2-Mdm4-p53 regulatory network. Cancer Cell 9: 273.
- Toledo, F., Lee, C. J., Krummel, K. A., Rodewald, L.-W., Liu, C.-W. and Wahl, G. M. 2007. Mouse Mutants Reveal that Putative Protein Interaction Sites in the p53 Proline-Rich Domain Are Dispensable for Tumor Suppression. Mol. Cell. Biol. 27: 1425-1432.
- Wade, M., Wong, E. T., Tang, M., Stommel, J. M. and Wahl, G. M. 2006. Hdmx Modulates the Outcome of P53 Activation in Human Tumor Cells. J. Biol. Chem. 281: 33036-33044.
- Wahl, G. M., Stommel, J. M., Krummel, K. A. and Wade, M. 2005. Gatekeepers of the guardian: p53 regulation by post-translational modification, mdm2 and mdmx. 25 Years of p53 Research. Edited by K. Wiman and Hainaut, P. Netherlands, Springer. pp73-113.
- Wang, Y. V., Wade, M., Wong, E., Li, Y.-C., Rodewald, L. W. and Wahl, G. M. 2007. Quantitative analyses reveal the importance of regulated Hdmx degradation for P53 activation. Proceedings of the National Academy of Sciences 104: 12365-12370.
The p53 mutation database
p53, Mdm2 and Mdmx-related information
Mouse models of human cancers consortium
American Association for Cancer Research
|Yao-Cheng (Leo) Li, Ph.D. graduated with a B. S. in Plant Pathology, 1992 from National Chung-Hsing University and a M.S. in Biochemistry, 1994 from the National Yang-Ming University, Taiwan. He graduated in 2002 with a Ph.D. in Pharmaclogy from the State University of New Jersey and Robert Wood Johnson Medical School, New Jersey. Leo joined the Wahl Lab in 2003.
Yao-Cheng (Leo) Li CV
|Mark Wade, Ph.D. Following a PhD working with Professor Martin Allday at Imperial College, London, Mark joined the Wahl lab to study the interplay between the c-myc oncogene and the p53 tumor suppressor. Subsequently, he has focused on the mechanisms by which the Mdm2/Mdmx ubiquitin ligase complex regulates p53 level and activity. Mark's broader interests are in the role of ubiquitylation in disease, and in the drug discovery process in general.
Mark Wade CV
|Yunyuan (Vivian) Wang, Ph.D. graduated from the Penn State University in 2005 with a Ph D. In the same year, she joined the Wahl Lab. Her research focuses on using mouse models to study the role of Mdmx in the p53 regulatory pathway.
Yunyuan (Vivian) Wang CV