Functional And Molecular Basis Of Thrombopoietin Receptor Activation By Mutant Calreticulin

The thrombopoietin receptor MPL is required for oncogenic transformation by mutant CALR (CALRMUT). We and others have previously shown that CALRMUT binds to the thromobopoietin receptor, MPL, to activate the MPL-JAK-STAT signaling axis, and that the positive charge of the CALRMUT C-terminus is required to mediate this interaction. To provide additional understanding of the mechanism by which CALRMUT is oncogenic, we sought to further resolve which amino acids within the CALRMUT C-terminus are required for CALRMUT to bind MPL, activate MPL signaling, and transform hematopoietic cells. To do this, we generated serial truncations of the mutant C-terminal tail. We found that even the most severely truncated form of the CALRMUT C-terminus retained the ability to bind MPL, but failed to activate JAK-STAT signaling and transform Ba/F3-MPL cells to IL-3 independence. This provides the first evidence that mutant CALR binding to MPL does not uniformly correlate with activation of JAK-STAT signaling and cellular transformation, and suggests a model whereby different thresholds of positive charge in the CALRMUT C-terminus are required to enable binding of CALRMUT to MPL and to activate MPL signaling. We next sought to elucidate the components of MPL that are required to bind CALRMUT and to support oncogenic transformation. We observed that CALRMUT bound to the extracellular domain of MPL, but not the intracellular domain, but that neither domain alone was capable of supporting CALRMUT-associated transformation. We therefore hypothesized that CALRMUT interaction with the MPL extracellular domain still required the intracellular domain of MPL to transmit canonical MPL-associated downstream signaling. We focused on three tyrosine residues within the carboxy-terminal of MPL (Tyr-591, Tyr-626 and Tyr-631), and generated MPL variants where these residues were systematically mutated individually, in tandem, or altogether. We observed no changes in the ability of any of the MPL variants to physically bind to CALRMUT. In contrast, MPL variants harboring mutation of all 3 tyrosine residues were unable to support cytokine-independent growth. Specifically, substitution of Tyr-626 significantly impaired the ability of MPL to support CALRMUT-associated transformation, whereas substitution of Tyr-591/Tyr-31 only had minor effects. Finally, we sought to gain functional insight into how CALRMUT interacts with MPL and how this interaction drives cellular transformation. Wild-type calreticulin is an ER-resident chaperone that interacts with newly synthesized glycoproteins as well as with the polypeptide chain to enable proper protein folding. We created variants of CALRMUT harboring mutations in critical residues implicated in three key functionalities of wild-type CALR - (i) lectin activity, (ii) polypeptide binding and (iii) chaperone activity - and tested their capacity to bind to MPL and confer cytokine-independence. We observed that lectin-deficient CALRMUT were unable to bind to MPL or confer cytokine independence. In contrast, polypeptide binding-deficient variants of CALRMUT retained MPL binding ability and conferred cytokine-independent growth. Finally, we observed that CALRMUT harboring mutations in residues implicated in wild-type CALR chaperone activity retained the ability to physically interact with MPL and confer cytokine-independent growth. Consistent with this, an in vitro binding assay demonstrated that CALRWT exhibits strong, direct binding to the co-chaperone molecule ERp57, whereas CALRMUT does not. Cumulatively, these data suggest that the interaction of CALRMUT with MPL is dependent on the CALRMUT lectin activity but not on its chaperone functionality. In conclusion, our data provide additional insights into the molecular mechanism by which CALRMUT interacts with MPL to induce MPN. Specifically, we uncouple the binding of CALRMUT to MPL from MPL activation and define the key properties of MPL required for CALRMUT binding and for its activation. Together, our findings consolidate the critical role of the positive electrostatic charge of the CALRMUT C-terminus in the oncogenic activity of CALRMUT and suggest a model of transformation whereby CALRMUT acquires a neomorphic function involving pathogenic binding to MPL and possibly to other proteins. Disclosures No relevant conflicts of interest to declare.