Defining the requirements for the pathogenic interaction between mutant calreticulin and MPL in MPN.

Mutations in calreticulin (CALR) are phenotypic drivers in the pathogenesis of myeloproliferative neoplasms. Mechanistic studies have demonstrated that mutant CALR binds to the thrombopoietin receptor MPL, and that the positive electrostatic charge of the mutant CALR C terminus is required for mutant CALR-mediated activation of JAK-STAT signaling. Here we demonstrate that although binding between mutant CALR and MPL is required for mutant CALR to transform hematopoietic cells; binding alone is insufficient for cytokine independent growth. We further show that the threshold of positive charge in the mutant CALR C terminus influences both binding of mutant CALR to MPL and activation of MPL signaling. We find that mutant CALR binds to the extracellular domain of MPL and that 3 tyrosine residues within the intracellular domain of MPL are required to activate signaling. With respect to mutant CALR function, we show that its lectin-dependent function is required for binding to MPL and for cytokine independent growth, whereas its chaperone and polypeptide-binding functionalities are dispensable. Together, our findings provide additional insights into the mechanism of the pathogenic mutant CALR-MPL interaction in myeloproliferative neoplasms.

Zinc-dependent multimerization of mutant calreticulin is required for MPL binding and MPN pathogenesis.

Calreticulin (CALR) is mutated in the majority of JAK2/MPL-unmutated myeloproliferative neoplasms (MPNs). Mutant CALR (CALRdel52) exerts its effect by binding to the thrombopoietin receptor MPL to cause constitutive activation of JAK-STAT signaling. In this study, we performed an extensive mutagenesis screen of the CALR globular N-domain and revealed 2 motifs critical for CALRdel52 oncogenic activity: (1) the glycan-binding lectin motif and (2) the zinc-binding domain. Further analysis demonstrated that the zinc-binding domain was essential for formation of CALRdel52 multimers, which was a co-requisite for MPL binding. CALRdel52 variants incapable of binding zinc were unable to homomultimerize, form CALRdel52-MPL heteromeric complexes, or stimulate JAK-STAT signaling. Finally, treatment with zinc chelation disrupted CALRdel52-MPL complexes in hematopoietic cells in conjunction with preferential eradication of cells expressing CALRdel52 relative to cells expressing other MPN oncogenes. In addition, zinc chelators exhibited a therapeutic effect in preferentially impairing growth of CALRdel52-mutant erythroblasts relative to unmutated erythroblasts in primary cultures of MPN patients. Together, our data implicate zinc as an essential cofactor for CALRdel52 oncogenic activity by enabling CALRdel52 multimerization and interaction with MPL, and suggests that perturbation of intracellular zinc levels may represent a new approach to abrogate the oncogenic activity of CALRdel52 in the treatment of MPNs.