UC Berkeley

Though many proteins have been identified as relevant to cancer pathogenicity, very few are now targeted clinically due to a lack of both a mechanistic understanding of the protein(s) and the identification of small molecule modifiers. Even highly-studied oncogenes are rarely therapeutically utilized, since solvent-accessible amino acid residues to which small molecule drugs can bind have often not yet been identified. In fact, it is currently thought that such small molecule binding pockets have only been identified on 10-15% of human genes, and less than 5% are currently exploited therapeutically.(1) In order to develop targeted cancer therapies, it is imperative to both identify and effectively pharmacologically manipulate small molecule binding pockets on disease-relevant protein targets.

Many existing targeted cancer therapies utilize or build upon naturally-occurring compounds that exert anti-cancer activity. Many of these natural products contain electrophilic moieties, reacting covalently with nucleophilic amino acids. While the use of natural products themselves as drugs seems like a promising idea, such complex compounds often interact widely in the proteome, resulting in non-therapeutic interactions that could potentially result in unwanted physiological side effects. A more effective route toward natural product-based drug development, therefore, requires identifying the particular covalent modification responsible for the anti-cancer activity and developing a compound that binds to the same residue more selectively. A variety of approaches to identifying targets of natural products have arisen, but most rely on piecemeal methods, such as derivatizing the natural product or assessing its reactivity one protein at a time.

Chemoproteomics have arisen as strategies to address these two challenges, allowing for proteome-wide assessment of a compound’s reactivity directly in biological systems. The use of chemoproteomic technologies has contributed to a deeper understanding of natural products’ protein interactions while also allowing for the identification of more selective, covalently-acting small molecules, whose smaller size also renders them more synthetically accessible.

In this work I provide a comprehensive discussion of the current understanding of metabolic pathways implicated in cancer and the associated therapies in development and use. In addition, I demonstrate the efficacy of chemoproteomic technologies in a study that identifies a novel protein target and druggable site targeted by an anti-cancer natural product, followed by the discovery and development of small molecules that bind more selectively to the protein target.