DIRAS3 mimetic peptide achieves dual MAPK and PI3K pathway suppression across KRAS hotspot mutations

Activating mutations in RAS occur in approximately 30% of human cancers, with hotspot alterations at codons G12, G13, and Q61 driving persistent downstream signaling through the MAPK and PI3K pathways. Although recent tricomplex RAS inhibitors have shown promise, their efficacy varies among different RAS mutants. Accumulating evidence indicates that the biochemical and signaling properties of G12X and Q61X mutants differ substantially, influencing their response to upstream and downstream modulation. G12X mutants depend on receptor tyrosine kinase activity for both MAPK and PI3K pathway activation, whereas Q61X mutants can independently activate MAPK signaling and maintain PI3K output even when RAS is inhibited. Consequently, Q61X tumors display intrinsic resistance to current RAS-targeted therapies, underscoring the need for agents capable of simultaneously suppressing both effector arms. DIRAS3 has emerged as a unique endogenous RAS inhibitor that disrupts RAS nanoclustering and effector engagement, while also modulating the PI3K/AKT axis in ovarian and pancreatic cancers, suggesting broader regulatory potential. To systematically define DIRAS3's inhibitory spectrum, we utilized RASless mouse embryonic fibroblasts (MEFs) reconstituted with individual KRAS hotspot alleles. DIRAS3 expression markedly reduced RAS-dependent clonogenic growth, inhibited MAPK and PI3K/AKT signaling in cells expressing oncogenic KRAS variants, with minimal effect on wild-type KRAS unless stimulated with EGF. Functional analysis demonstrated that both N- and C-terminal domains of DIRAS3 are essential for membrane localization and full inhibitory activity. Importantly, DIRAS3 disrupted KRAS nanoclustering in both G12X and Q61X mutant backgrounds, supporting its capacity to interfere directly with the KRAS-KRAS interface. To recapitulate DIRAS3's inhibitory effects, we designed a conformationally constrained cyclic DIRAS3 mimetic peptide based on the sequence spanning parts of alpha5 and the hypervariable regions critical for DIRAS3-KRAS binding. For translational comparison, we evaluated the tricomplex RAS inhibitor RMC-7977 alongside this DIRAS3-derived peptide. In G12X cells, both agents suppressed MAPK and PI3K outputs. However, in Q61X cells, RMC-7977 inhibited MAPK signaling but failed to suppress PI3K/AKT activity, whereas the DIRAS3 peptide effectively suppressed both pathways. Reduced PI3K activity correlated with enhanced sensitivity to the DIRAS3 peptide, and cells harboring Q61X mutations, which are less responsive to RMC-7977, remained equally responsive to the DIRAS3 peptide. Collectively, these results indicate that DIRAS3 provides a dual-pathway blockade of MAPK and PI3K signaling across diverse KRAS mutations, including Q61X variants that are refractory to existing agents, supporting its development as a broad-spectrum therapeutic for RAS-driven cancers.