Unveiling KRAS Mutant Structures with AlphaFold 3: New Avenues for Targeted Cancer Therapy

Abstract

Despite recent advances in targeting the "undruggable" KRAS (Kirsten rat sarcoma viral oncogene homologue) protein, much of its structural information remains elusive. Utilizing a "branch-pruning" strategy, this study systematically mutated the Switch I and II regions of KRAS, predicting mutants structures with AlphaFold 3 and analyzing their structure differences with KRAS wildtype using PyMOL. The analysis of 494 mutants showed that over 64% had root-mean-square deviation (RMSD) scores below 0.200, indicating minor structural changes compared to wildtype KRAS. Notably, Switch II exhibited greater conformational variability, suggesting its potential role in oncogenic signaling. This work found that Q61 mutants showed limited structural changes, but nearby sites like E63 and M67 were more dynamic. Highly dynamic sites D38 (on Switch I) and E76/G60 (on Switch II) were identified, potentially affecting KRAS dimerization and post-translational modifications in oncogenesis. Mutants like D31I, R73C, and T74Q induced conformational changes, yet their biological functions are still unknown, possibly impacting effector binding and allosteric communication. These findings could reveal cryptic pockets for drug targeting, deepening our understanding of KRASs role in cancer.