The tumor suppressor p53 controls cell proliferation and protects against carcinogenesis through a combination of transcriptional activation, repression and direct activation of the intrinsic cell death pathway. p53 is lost or functionally inactivated in nearly of human tumors. Because of its crucial tumor suppressor activity, p53 is an appealing target for pharmacological intervention in cancer treatment. One of emerging approaches to modulate the tumor suppressor activity of p53 is to explore its tendency to form complexes with other transcription factors and proteins. Historically, targeting protein-protein interactions as potential intervention points for developing therapeutic agents for cancer therapy was perceived to be unattractive, but recent success in identifying of small molecule drug-like antagonists lends support for protein-protein interactions as potential targets for anti-cancer drug discovery.

The overriding goal of this thesis was to understand how protein-protein interactions affect p53 function in order to explore p53 and its complexes as targets for intervention in cancer chemotherapy. Two interacting partners of p53 were examined in order to evaluate: I) stabilization as an approach for functional rescue of mutant p53; and 2) antagonism to inhibit co-transcriptional activation of an oncogene.

Our first approach was to design a drug-like molecule from an interaction where the interfaces of two interacting proteins were known at atomic resolution. Rationally designing a drug based on known structure and function seemed reasonable, but our experience indicates that rational design can become irrational. Because of a lack of atomic resolution structure or accurate biophysical probes for the protein ligand interaction, it is hard to identify whether the ligand binds to the same place that the parental protein binds. However, this study led to the identification of a novel orphan binding site that may be useful as a potential target for drug discovery.

By studying the interaction interfaces of the oligomeric proteins p53 and TFAP2, we have confronted the conventional wisdom that it is difficult to antagonize protein-protein interactions with small molecules, because the affinity of protein-ligand interactions is often overwhelmed by the avidity of protein-protein interactions. However, our work shows convincingly that modulating post-translational modifications offers an alternative approach to inhibit even protein-protein interactions that exploit avidity to generate substantial binding affinity. This finding rationalizes the uses of pharmacological agents that modulate post-translational modifications in combination with current treatments for cancer patients, and suggests that protein-protein interactions remain viable targets for pharmacologic intervention in disease.