KRAS G12R: the atypical PDAC mutant that signals differently

Most discussion of oncogenic KRAS treats the codon-12 variants as interchangeable drivers that differ mainly in which warhead can reach them. G12R breaks that assumption. It is the third most common KRAS mutation in pancreatic ductal adenocarcinoma (PDAC), accounting for roughly 20% of cases, yet it is nearly absent from lung and colorectal cancer. More importantly, it does not signal the way G12D and G12V do, which has real consequences for how you would try to drug it.

Why arginine at position 12 is different

Substituting a bulky, positively charged arginine at residue 12 does the usual thing of impairing GTP hydrolysis and locking KRAS in the active state. But the side chain also reaches into switch II and perturbs the surface that the lipid kinase PI3K docks against. Hobbs et al. (Cancer Discovery, 2020) showed that KRAS G12R is defective for interaction with the p110a catalytic subunit of PI3K. As a result, G12R-mutant PDAC does not drive macropinocytosis through the canonical RAS-to-PI3K route that feeds G12D and G12V tumors their nutrients; the cells lean instead on a RAS-independent, PI3K-gamma-supported pathway.

That is not a trivia point. Macropinocytosis is how PDAC scavenges protein from a nutrient-poor stroma, and the PI3K axis is a major survival input. A mutant that has rewired both of those is a different therapeutic target than its codon-12 siblings, even though the activating lesion sits on the same residue.

Why the existing KRAS drugs miss it

G12R falls into the gap between the two main drugging playbooks.

• Covalent G12C inhibitors (sotorasib, adagrasib) are irrelevant: there is no cysteine 12 to form a Michael adduct with. Arginine offers no comparable nucleophile.
• Non-covalent pan-KRAS inhibitors have struggled here too. BI-2865, the pan-KRAS tool compound from Kim et al. (Nature, 2023), suppresses a wide spectrum of KRAS mutants but specifically excludes G12R and the Q61 variants, because the arginine reshapes the pocket those compounds rely on.
• MEK inhibition was the early rationale: if G12R cannot engage PI3K, the tumor should lean harder on the MAPK arm. In practice a phase 2 study of selumetinib in G12R-mutant PDAC produced no radiographic responses, with a median progression-free survival around 3 months. Loss of PI3K input did not translate into MEK dependence in the clinic.

Where the hope is now

The most credible path to G12R is the RAS(ON) multi-selective class. Daraxonrasib (RMC-6236) glues active RAS-GTP to cyclophilin A and blocks effector engagement regardless of which residue is mutated, so it covers G12R alongside G12D, G12V and G13D rather than depending on a residue-specific handle. In the phase 1 program it produced a 36% objective response rate in second-line KRAS G12X-mutant metastatic PDAC at the 300 mg dose, with a median progression-free survival of 8.8 months, and earned FDA breakthrough therapy designation. The phase 3 RASolute 302 study explicitly enrolls G12X, G13X and Q61X disease, so G12R patients finally sit inside a registrational trial rather than outside every targeted option.

Try the docking yourself

The G12R pocket is the interesting part: the arginine guanidinium is exactly what reshapes switch II and defeats pocket-binders tuned for the smaller codon-12 mutants. Open Studio and pick KRAS from the target catalog with the G12R mutation chip, then run the same ligand against G12R, G12D and wild-type in parallel. The molecular docking scores should show a non-covalent binder that fits G12D losing ground against G12R as the arginine crowds the site, which is the structure-level version of why BI-2865 excludes this variant. The point of the comparison is the ΔΔ across mutants, not the absolute score on any one.

Liganx is molecular docking online and free in the browser, built for this kind of mutant-versus-wildtype triage. If you want to run molecular docking on KRAS G12R without a local install, it is the fastest way to see the pocket difference for yourself.

Primary sources

• Hobbs GA, et al. Atypical KRASG12R mutant is impaired in PI3K signaling and macropinocytosis in pancreatic cancer. Cancer Discov 10, 104-123 (2020). doi:10.1158/2159-8290.CD-19-1006
• Kim D, et al. Pan-KRAS inhibitor disables oncogenic signalling and tumour growth. Nature 619, 160-166 (2023). doi:10.1038/s41586-023-06123-3
• Wolpin BM, et al. Daraxonrasib in previously treated advanced RAS-mutated pancreatic cancer. N Engl J Med (2026). doi:10.1056/NEJMoa2505783