Allosteric mechanisms driving paradoxical activation of RAF kinases

ABSTRACT Mutations in the BRAF protein kinase are key drivers of melanoma, thyroid cancer, and colon cancer. Existing drugs, like vemurafenib, that target the most common BRAF mutant (V600E) are initially effective in patients, but clinical resistance invariably develops after a few months. V600E BRAF normally signals as a monomeric kinase,

which is readily inhibited by these first-generation drugs. Resistance to these drugs is mediated by BRAF

dimerization, which triggers a conformational change of the kinase from an “C-out to an “C-in” state and blocks drug binding. New inhibitors have been developed to recognize the C-in state and thereby target BRAF dimers, but these molecules exhibit a mysterious phenomenon called paradoxical activation in which they can increase

BRAF activity under some conditions, rather than inhibit it. Paradoxical activation is linked to induction of BRAF dimers by inhibitors, but the underlying molecular mechanism is not well understood, rendering it challenging to design new inhibitors that avoid it. Using an integrated biophysical approach, we built a model for paradoxical activation that describes the

allosteric coupling between inhibitor binding and BRAF dimerization in quantitative thermodynamic terms. In our approach, a dataset of FRET measurements that quantify inhibitor-driven BRAF dimerization are globally fit to our dimerization model, yielding thermodynamic parameters that describe the allosteric coupling mechanisms

underlying drug-driven BRAF dimerization. Remarkably, the results show that inhibitor-driven dimerization is asymmetrical, with the first drug binding event triggering dimerization strongly, and the second binding event contributing little additional dimerization affinity. This model accurately predicts the shape of BRAF kinase

activation curves measured as a function of inhibitor concentration, demonstrating that it provides a realistic physical framework for understanding paradoxical activation. We plan to use this approach to expand our understanding of paradoxical activation in RAF kinases. In Aim 1, we classify a large set of RAF inhibitors based on their allosteric coupling strengths and the conformation of

the C-helix they promote, unraveling the mechanistic connection between inhibitor binding, conformational change, and BRAF dimerization. In Aim 2, we apply this approach to oncogenic mutations of BRAF to understand how these mutations are coupled to inhibitor-driven dimerization and activation. In Aim 3, we extend our studies

to higher-order complexes of BRAF, including the scaffolding protein 14-3-3, which can either weaken or enhance BRAF dimerization depending on the mode of interaction, and BRAF/CRAF heterodimers, which are an important signaling species in cells with catalytically-inactive oncogenic BRAF mutations. Collectively this

work will reveal the molecular mechanisms underlying paradoxical activation of RAF complexes and aid the design of new inhibitors that avoid triggering paradoxical activation.