New Paradigms for the Mechanisms of Thrombopoietin Receptor Activation and Dysregulation By the JAK2V617F Mutation

Abstract

Janus kinase (JAK2)V617F is the most common mutation found in patients with Philadelphia chromosome negative myeloproliferative neoplasms (Ph- MPNs). The discovery of this mutation over 15 years ago revolutionised MPN diagnosis and inspired the development of JAK inhibitors as new therapeutic interventions. However, despite extensive structural and biophysical studies using JAK2 domains in isolation, the exact molecular mechanisms of JAK2V617F activation remains elusive. We have previously demonstrated that expression of the thrombopoietin (TPO) receptor, MPL, which interacts directly with JAK2, is essential for disease development in a mouse model of a JAK2V617F-positiveMPN (Blood 2014 124:3956-3963). Using total internal reflection fluorescence (TIRF) microscopy, we visualized MPL interaction dynamics in live cells on single molecule level. Effective cell surface MPL fluorescence labelling and dual-color imaging allowed us to determine the level of MPL dimerization under various experimental conditions. Using this assay, we clearly established that MPL is monomeric at physiologically relevant receptor densities. However, TPO stimulation results in significant dimerization of MPL (>50%) and an equilibrium between monomers and dimers. This counters the current dogma that MPL exists at the membrane as a pre-formed dimer. Strikingly, we found that JAK2V617F shifts this monomer-dimer equilibrium leading to significant TPO-independent MPL dimerization providing a novel mechanistic model of oncogenic JAK2 activation. To highlight the role of ligand-independent receptor dimerization in JAK2 activation, we compared three groups of autoactivating mutations in the PK domain covering the FERM-SH2 (FS2)-PK linker region (Group I), residues in the proximity of the αC helix (Group II) and at the autoinhibitory PK-TK interface (Group III). Consistent MPL dimerization was only observed for mutations in groups I and II. Mutations in these groups both localize to a potential homomeric PK/PK interface that has been implicated as a switch of JAK activation. Using MD simulations, we also found that the FERM domain of JAK2 strongly interacts with the inner leaflet of the lipid bilayer of the plasma membrane via a single hydrophobic residue (L224) surrounded by several positively charged residues that allows the region to act as a membrane anchor. This tight coupling to the membrane enforces an appropriate orientation between the JAKs within the receptor dimers required for optimal intermolecular PK/PK interaction that is critical for receptor dimerization. To interfere with membrane anchoring, we introduced a negative charge in this position (L224E). Strikingly, ligand-independent MPL dimerization and activation by JAK2V617F was dramatically reduced upon introducing L224E, supporting the vital importance of L224 for orienting JAK2 at the membrane to allow productive PK-PK interactions. Here, we demonstrate that JAK2V617F mutation acts by altering and strengthening the intermolecular interactions involving the PK/PK dimerization interface. In essence, these mutations drive cytoplasmic stabilization of receptor-JAK dimers, bypassing extracellular stabilization of dimers via cytokine binding. These results provide critical and entirely novel mechanistic insights into signal initiation in MPNs and readdress the roles of receptor-associated proteins. Disclosures Hubbard: Ajax Therapeutics, Inc.: Membership on an entity's Board of Directors or advisory committees, Other: Co-Founder.