Beyond KRAS(G12C):\nBiochemical and Computational Characterization\nof Sotorasib and Adagrasib Binding Specificity and the Critical Role\nof H95 and Y96

Mutated KRAS proteins are frequently expressed in some\nof the most\nlethal human cancers and thus have been a target of intensive drug\ndiscovery efforts for decades. Lately, KRAS(G12C) switch-II pocket\n(SII–P)-targeting covalent small molecule inhibitors have finally\nreached clinical practice. Sotorasib (AMG-510) was the first FDA-approved\ncovalent inhibitor to treat KRAS(G12C)-positive nonsmall cell lung\ncancer (NSCLC), followed soon by adagrasib (MRTX849). Both drugs target\nthe GDP-bound state of KRAS(G12C), exploiting the strong nucleophilicity\nof acquired cysteine. Here, we evaluate the similarities and differences\nbetween sotorasib and adagrasib in their RAS SII–P binding\nby applying biochemical, cellular, and computational methods. Exact\nknowledge of SII–P engagement can enable targeting this site\nby reversible inhibitors for KRAS mutants beyond G12C. We show that\nadagrasib is strictly KRAS- but not KRAS(G12C)-specific due to its\nstrong and unreplaceable interaction with H95. Unlike adagrasib, sotorasib\nis less dependent on H95 for its binding, making it a RAS isoform-agnostic\ncompound, having a similar functionality also with NRAS and HRAS G12C\nmutants. Our results emphasize the accessibility of SII–P beyond\noncogenic G12C and aid in understanding the molecular mechanism behind\nthe clinically observed drug resistance, associated especially with\nsecondary mutations on KRAS H95 and Y96.