CRAF R391W is a melanoma driver oncogene.

Approximately 75% of melanomas have known driver oncogenic mutations in BRAF, NRAS, GNA11 or GNAQ, while the mutations providing constitutive oncogenic signaling in the remaining melanomas are not known. We established a melanoma cell line from a tumor with none of the common driver mutations. This cell line demonstrated a signaling profile similar to BRAF-mutants, but lacked sensitivity to the BRAF inhibitor vemurafenib. RNA-seq mutation data implicated CRAF R391W as the alternative driver mutation of this melanoma. CRAF R391W was homozygous and over expressed. These melanoma cells were highly sensitive to CRAF, but not BRAF knockdown. In reconstitution experiments, CRAF R391W, but not CRAF WT, transformed NIH3T3 cells in soft-agar colony formation assays, increased kinase activity in vitro, induced MAP kinase signaling and conferred vemurafenib resistance. MAP kinase inducing activity was dependent on CRAF dimerization. Thus, CRAF is a bona fide alternative oncogene for BRAF/NRAS/GNAQ/GNA11 wild type melanomas.

JUN dependency in distinct early and late BRAF inhibition adaptation states of melanoma.

A prominent mechanism of acquired resistance to BRAF inhibitors in BRAF (V600) -mutant melanoma is associated with the upregulation of receptor tyrosine kinases. Evidences suggested that this resistance mechanism is part of a more complex cellular adaptation process. Using an integrative strategy, we found this mechanism to invoke extensive transcriptomic, (phospho-) proteomic and phenotypic alterations that accompany a cellular transition to a de-differentiated, mesenchymal and invasive state. Even short-term BRAF-inhibitor exposure leads to an early adaptive, differentiation state change-characterized by a slow-cycling, persistent state. The early persistent state is distinct from the late proliferative, resistant state. However, both differentiation states share common signaling alterations including JUN upregulation. Motivated by the similarities, we found that co-targeting of BRAF and JUN is synergistic in killing fully resistant cells; and when used up-front, co-targeting substantially impairs the formation of the persistent subpopulation. We confirmed that JUN upregulation is a common response to BRAF inhibitor treatment in clinically treated patient tumors. Our findings demonstrate that events shared between early- and late-adaptation states provide candidate up-front co-treatment targets.