RAS G12 and Q61 codon mutations confer distinct disease characteristics in Acute Myeloid Leukemia and exhibit differential response to RAS(ON) inhibitor RMC-7977.

Abstract Background RAS pathway mutations are present in 10-20% of acute myeloid leukemia (AML) cases at diagnosis, and frequently occur as late, subclonal events in leukemogenesis. Bulk sequencing studies have demonstrated that NRAS mutations are most frequent, and that G12/G13 codon mutations predominate, followed by Q61 codon mutations. There is growing interest in understanding and targeting RAS mutations in AML, owing to studies indicating that RAS mutations are associated with relapse after multiple novel targeted therapy agents, including azacitidine/venetoclax, and FLT3 and IDH1/2 inhibitors. In addition, the advent of novel RAS(ON) inhibitors showing promise in solid cancers offers a therapeutic strategy that could be deployed to target these mutations. Our group has previously described the use of tamoxifen inducible, Cre-recombinant single- and double-mutant mouse models harboring Npm1c, NrasG12D, and NrasQ61R mutant alleles and shown that there are distinct codon specific phenotype differences in RAS-mutant AML. Here, we report the use of these models to further characterize differences in disease phenotype, as well as response to the RAS(ON) inhibitor, RMC-7977. Methods We utilized our inducible, single- and double-mutant mouse models harboring Npm1c, NrasG12D, and NrasQ61R mutant alleles to study differences in response to RAS(ON) inhibitor, RMC-7977. In vivo treatment with RMC-7977 or vehicle began upon the development of disease symptoms and continued for 28 days with regular monitoring via peripheral blood draws and flow cytometric analysis. Ex vivo studies were conducted in mutant hematopoietic stem/progenitor cells (HSPCs) using murine bone marrow endothelial co-culture (BMEC) methods, with endpoints including cell counts, flow cytometry, and Western blots. To complement these studies, we also treated a panel of RAS-mutant human AML cell lines that harbor G12/Q61 codon mutations with RMC-7977 and measured response through cell-titer glo, cell proliferation, cell cycle arrest and annexin V assays. Results We observed that double-mutant Npm1c/NrasQ61R (NRQ61) exhibited higher WBC counts and spleen weights, compared to Npm1c/NrasG12D (NRG12), with more severe anemia and thrombocytopenia at 12 weeks post-activation of mutant alleles. Flow cytometric analysis of HSPCs and mature lineages revealed significant differences between the models, specifically that NRG12 exhibited higher levels of HSPCs than NRQ61, while NRQ61 exhibited higher levels of monocytes, with increased skewing toward a more mature myelomonocytic phenotype. Moreover, we further observed that NRQ61 exhibited the most aggressive disease phenotype in vivo, with the significant clonal fitness in competitive transplant assays and the shortest overall survival compared to NRG12 and single-mutant cohorts. Utilizing these models to study differences in response to RMC-7977, we performed methocult colony growth assays and found that RMC-7977 significantly decreased the number of colonies of both leukemic models in a dose dependent manner. In addition, we found that RMC7977 significantly reduced the overall number and percentage of multiple HSPC populations after 1 week of treatment using our BMEC co-culture system, with a more drastic effect observed in the NRG12 model compared to NRQ61. Moreover, the percentage of lineage positive cells increased during treatment, indicating the drug has a differentiation effect. Finally, in vivo testing of RMC-7977 in NRG12 and NR61 mouse models found that the RAS(ON) inhibitor was able to improve survival in NRG12 mice, with more limited effects observed in NRQ61 mice. Analysis of on-treatment and off-treatment timepoint samples confirmed codon specific effects in response and RAS pathway signaling. To validate these findings, we performed characterization of RMC-7977 in RAS-mutated AML cell line models. RMC-7977 was shown to ablate MAPK signaling and have high potency across a range of RAS-mutated AML cell lines, with highest potency observed in THP1 (G12D, IC50 585pM), compared to OCI-AML3 (Q61L, IC50 5.85pM) and HL60 (Q61L, IC50 2.78nM). RMC-7977 induced a variable degree of apoptosis and cell cycle arrest, which varied with the specific codon mutation present. Conclusions Overall, our study demonstrates that there are codon-specific differences in disease latency and overall survival in models of AML with distinct Nras codon mutations, and that the responsiveness to RAS(ON) inhibition varies with codon mutation.