JAK2 V617F Dimerizes Homodimeric Cytokine Receptors Cytosolic Domains By Requiring Pseudokinase Domain Residues That Promote JAK2 Dimerization and Oncogenic Activity

JAK2 plays roles in several signaling pathways by binding to multiple distinct cytokine receptors. Dysfunction of JAK2 signaling is associated with hematopoietic malignancies. The acquired mutation JAK2 V617F, which constitutively activates JAK2 kinase, is the most common molecular event associated with myeloproliferative disorders.

Using a combination of bioluminescence energy transfer and protein fragment complementation assays we show that V617F located in the pseudokinase domain of JAK2 plays an important role on JAK2 dimerization via the C-terminal kinase domains. Using NanoBRET we show that JAK2 dimerization via the C-terminal domains is reduced by single mutations (E596R ¹, F595A ², A598F and F537A) that in the context of JAK2 V617F revert the mutant to wild type phenotype with respect to signaling by homodimeric erythropoietin and thrombopoietin receptors (EpoR and TpoR). Thus, residues that are required for V617F activation strengthen JAK2 dimerization via C-terminal kinase domains, implicating the pseudokinase domain in global JAK2 dimerization and suggesting that part of the V617F activation mechanism relies on dimerization of mutated pseudokinase domains.

Using NanoBiT protein fragment complementation assay, we explored the effects of JAK2 V617F on dimerization of the cytosolic ends of cytokine receptors. We show that JAK2 V617F promotes strong enhancement of dimerization of EpoR and TpoR cytosolic domains, when compared to wild type JAK2. This result is unexpected for EpoR that was shown to exist in a preformed dimer configuration at a significantly higher degree than TpoR^3-4. Nevertheless our results indicate that JAK2 V617F enhances close apposition of cytosolic domains of EpoR and TpoR, which could occur within preformed complexes or by a switch from monomer to dimer configuration induced by JAK2 V671F. This enhancement of dimerization is prevented when JAK2 V617F carries any of the reverting mutations (E596R, F595A, A598F, F537A) that prevent V617F activation. We also detected enhanced EpoR dimerization for exon 12 JAK2 mutation and JAK2 T875N, but not for other JAK2 mutants that are involved in lymphoid malignancies. The signals we detect are not simply the result of signaling and clustering in vesicles or internalization, as inhibition of signaling via Ruxolitinib does not prevent the dimerization effects on receptors. Our results indicate that rather than preformed dimeric receptors aiding in JAK2 V617F activation⁵, JAK2 V617F promotes close apposition of cytosolic domains of dimeric receptors. This result is in agreement with a recent study that showed that JAK2 V617F requires a closer proximity of EpoR chains than signaling by wild type JAK2⁶. Such differences in dimeric requirements might open new therapeutic avenues for JAK2 V617F inhibition.

We extended these explorations to other cytokine receptors utilizing JAK2 or JAK1 belonging to type I and type II subfamilies. Results on JAK2 are transposable to JAK1 for homologous mutations. Most receptors tested, such as EpoR, TpoR, G-CSFR, IL2, IL9 and type I interferon respond to their cytokine when JAK2 (or JAK1 for those utilizing JAK1) carries mutations known to inhibit JAK2 V617F (or JAK1 V658F, homologous to JAK2 V617F). One prominent exception is interferon γ-receptor (IFNGR), which is significantly impaired in its response to IFNγ when JAK2 and JAK1 carry such mutations. We show that homodimeric interactions within the heterotetrameric IFNGR recruiting JAK2 and JAK1 explain this inhibitory effect.

Thus, it appears that a specific region around pseudokinase domain helix αC, which is required for JAK2 V617F and JAK1 V658F oncogenic activation, is also critical for relaying normal IFNγ signaling via its tetrameric receptor complex, which appears to require enhanced dimerization between JAK1 and JAK2 pseudokinase domains. Our results provide critical information for the design of strategies aiming to inhibit JAK2 V617F and JAK1 V658F, and also provide new avenues to select for compounds able to prevent excessive IFNγ signaling, which is present in certain autoimmune conditions.

1. Leroy, E.; et al. Biochem J2016,473 (11), 1579-91.

2. Dusa, A.; et al. PLoS One2010,5 (6), e11157.

3. Defour, J. P.; et al. Proc Natl Acad Sci U S A2013,110 (7), 2540-5.

4. Constantinescu, S. N.; et al. Proc Natl Acad Sci U S A2001,98 (8), 4379-84.

5. Lu, X.; et al. Proc Natl Acad Sci U S A2005,102 (52), 18962-7.

6. Moraga, I.; et al. Cell2015,160 (6), 1196-208.