Understanding why the same genetic mutation causes cancer in some tissues but not others is one of oncology's most vexing puzzles — and cracking it could reshape how targeted therapies are designed and matched to patients. KRAS mutations are among the most common oncogenic drivers in human cancer, yet they cluster overwhelmingly in pancreatic, colorectal, and lung tissues. This selective vulnerability has long lacked a molecular explanation.
Published in PNAS, this study mapped transcriptomic changes induced by mutant KRAS expression across multiple distinct tissue lineages, then systematically identified synthetic lethal dependencies — genes whose loss becomes lethal specifically in the context of KRAS mutation. The findings reveal that KRAS does not simply flip a universal oncogenic switch; instead, the downstream transcriptional reprogramming and the resulting vulnerabilities differ substantially depending on the tissue environment in which the mutation occurs. In other words, a KRAS mutation in a pancreatic cell recruits a meaningfully different molecular support network than the same mutation in a lung or colorectal cell.
This work adds important mechanistic depth to a long-standing clinical observation. Drugs targeting KRAS — such as sotorasib and adagrasib, approved for KRAS G12C-mutant lung cancer — have shown far less efficacy in colorectal settings, a disparity that previously lacked a clear molecular rationale. Tissue-specific synthetic lethal maps could explain these differential responses and point toward combination therapy targets uniquely suited to each tumor type. The limitation here is that transcriptomic profiling and synthetic lethality screens, while powerful, are largely conducted in cell-line or preclinical models, and the translation to patient tumors remains to be validated. Still, this represents a conceptually significant advance: it reframes KRAS-mutant cancers not as a single disease class but as a family of tissue-conditioned vulnerabilities, each requiring its own therapeutic strategy. For drug development, that distinction is potentially paradigm-shifting.