K-Ras is the most frequently activated oncogene in cancer, yet there are not drugs approved or in clinical trials which target K-Ras directly. Our lab recently discovered a new druggable pocket to inhibit the activity of the most common K-Ras mutant in lung cancer, K-Ras, G12C (Ostrem, JM, Peters, U, Sos, ML, Wells, JA, Shokat, KM. (2013) K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interactions. Nature. 503(7477):548-51). The challenge addressed in this grant application is to extend our understanding of this new pocket on to the mutants of K-Ras (G12V and G12D) which cause pancreatic and colon cancer.
The small GTPase K-Ras is the most frequently mutated oncogene in cancer. Direct inhibition of other oncogenes such as the fusion protein BCR-Abl, B-Raf V600E, and others, has resulted in breakthrough therapies for patients harboring the respective kinase target. Despite the prevalence of K-Ras mutations in cancer, direct inhibitors of this oncogene have been largely unavailable until several recent reports. We recently identified inhibitors of the most common K-Ras mutation in lung cancer, glycine-12 to cysteine (G12C), using a disulfide-based tethering screen. These inhibitors bind to a novel pocket behind switch-II, one of the two mobile domains of Ras. This pocket, which we have termed the switch-II pocket (S-IIP), can be exploited to allosterically control nucleotide affinity and effector interactions and lock Ras in an inactive state. Our current compounds depend on covalent attachment of the inhibitor to the mutant cysteine-12. However, across cancers of all tissues, non-cysteine substitutions account for a majority of K-Ras mutations. In order to develop inhibitors capable of targeting these mutants (including the most frequent mutants, G12D and G12V), we aim to develop small molecules that non-covalently target the K-Ras S-IIP and do not depend on the presence of a mutant cysteine at position-12. In the course of our covalent inhibitor studies we characterized a hydrophobic region within the S-IIP that accounts for a high proportion of inhibitor binding affinity, which we refer to as the high affinity sub-pocet. A survey of the original tethering screen library suggests that the majority of fragments were too short to reach this region. We propose to introduce unnatural cysteine residues in close proximity to the high affinity sub-pocket of the S-IIP to use as temporary covalent handles for screening an expanded library of tethering fragments. Using this approach to maximize the chemical space we scan, we aim to identify tight-binding tethering fragments that display high ligand efficiency (high affinity relative to their mass) to serve as starting points for the ultimae goal of developing non-covalent inhibitors of the S-IIP. Structural analysis of the S-IIP suggests that mutation of methionine-72 (M72) or valine-9 (V9) should afford optimal cysteine positioning. Preliminary screening of a library of disulfide-containing fragments against K-Ras M72C and K-Ras V9C using intact protein mass spectrometry uncovered several fragments that bind to M72C with high ligand efficiency. Initial chemical optimization of these reversible covalent hits i conjunction with structural characterization using X-ray crystallography in Aim 1 will be imperative for understanding the basis for binding in the S-IIP. These data will help guide the progression from fragments that require reversible covalent attachment through disulfide bonds (M72C) to lead compounds binding non-covalently to K-Ras G12D and G12V in Aim 2. Finally, we will evaluate the biochemical and cellular effects of these compounds in Aim 3.