CRAF (RAF1) is one of three RAF-family kinases that initiate MAP kinase signaling in response to activated RAS and is essential for oncogenic signaling from mutant KRAS. Like BRAF, CRAF is regulated by 14-3-3 engagement and by intramolecular autoinhibitory interactions of its N-terminal regulatory region. Unlike BRAF, it is thought to require tyrosine phosphorylation in its N-terminal acidic (NtA) motif for full catalytic activation. Here we describe cryo-EM reconstructions of full-length CRAF in complex with MEK1 and a 14-3-3 dimer. These structures reveal a fully autoinhibited conformation analogous to that observed for BRAF and two “open monomer” states in which the inhibitory interactions of the CRD and 14-3-3 dimer are released or rearranged, but the kinase domain remains inactive. Structure-function studies of the NtA motif indicate that phosphorylation or acidic mutations in this segment increase catalytic activity by destabilizing the inactive conformation of the kinase domain. Collectively, these studies provide a structural foundation for understanding the shared and unique regulatory features of CRAF and will inform efforts to selectively block CRAF signaling in cancer.
The RAS-RAF-MEK-ERK signaling pathway controls cellular growth and proliferation, and mutational activation of this pathway is a frequent cause of cancer. Most prominently, the V600E mutation in BRAF causes malignant melanoma, papillary thyroid cancer, and other malignancies. Rare but recurrent activating mutations in the other two RAF isoforms, ARAF and CRAF, have also been identified in diverse cancers. Distinct classes of RAF inhibitors have been developed, particularly for BRAFV600E, but their potencies against the three RAF isoforms have not been systematically compared. Here, we biochemically characterize monomeric and dimeric preparations of ARAF, BRAF, and CRAF kinases and measure the potencies of a panel of thirteen type I, type I.5, and type II RAF inhibitors against each active RAF preparation. Type I inhibitor SB590885 is roughly equipotent across RAF isoforms and, as expected, type I.5 inhibitors are typically most potent against BRAFV600E. Despite their reputation as pan-RAF inhibitors, type II inhibitors as a class are potent inhibitors of CRAF but exhibit relative sparing of ARAF and intermediate potencies against BRAF. Type II compounds inhibit BRAF and CRAF with marked positive cooperativity, and their apparent potencies are insensitive to ATP concentrations. Crystal structures of CRAF in complex with type I.5 inhibitor PLX4720 reveal an asymmetric CRAF dimer with one CRAF subunit bound in the inactive state and the second bound in an αC-helix-in, active conformation with an altered inhibitor pose. Our findings have important implications for understanding the pharmacology of current RAF inhibitors and will inform development of new agents with distinct isoform selectivity.