Abstract

Mutations in the calreticulin (CALR) gene were recently discovered in patients with essential thrombocythemia and it has been turned out that CALR mutated patients have a lower risk of thrombosis than JAK2 V617F patients. However, the molecular mechanism for this differential risk remains obscure. It has been reported that CALR is a potential regulatory protein of integrin activation based on the interaction between CALR and a conserved sequence of GFFKR in the integrin α cytoplasmic tails. Recent studies suggest that calreticulin activates β1 integrin and modulates integrin-associated signaling. In this study, we examined if the mutant CALR proteins observed in patients with ET affect integrin αIIbβ3 activation which plays a crucial role on thrombus formation. We first identified mutations of JAK2, MPL, and CALR genes in 37 patients with WHO defined ET and explored clinical characteristics of patients with CALR mutation. The patients with JAK2 V617F were 22 (59%), MPL W515L was 1 (3%), and CALR mutations were 10 (27%). The two types of CALR mutations were found; deletion (52-bp deletion; c.1092_1143del) and insertion (5-bp insertion; c.1154_1155insTTGTC) mutations. The patients with CALR mutations had lower hemoglobin and leukocyte count compared with JAK2 V617F patients, but platelet count did not have a difference between the CALR and JAK2 mutation groups. Nine (41%) of 22 patients with JAK2V617F had a thrombotic event while 1 (10%) of 10 patients with CALR mutation did (p<0.05), suggesting that patients with CALR mutation had a lower risk of thrombosis than JAK2 V617F patients. Two patients with CALR mutations developed myelofibrosis while no patient with JAK2V617F did. One patient with CALR mutation developed acute myeloid leukemia, with persistence of the CALR mutation in his leukemic cells.

To see if the CALR mutation affects functional status of αIIbβ3, we examined the binding of PAC1, a monoclonal antibody recognizing the active conformation of αIIbβ3, to platelets from 5 patients with CALR mutation and 12 patients with JAK2V617F in the presence or absence of ADP. Platelets from all the 5 patients with CALR mutation showed the same level of PAC1 binding as platelets from healthy subjects. Overexpression of recombinant CALR proteins in Chinese Hamster Ovary (CHO) cells expressing αIIbβ3 by transfection of a protein-expression vector containing wild-type, deletion, or insertion mutant CALR had no effect on PAC1 binding. We further examined adhesive function of CHO cells stably expressing αIIbβ3 and mutant or wild-type CALR to various concentrations of immobilized fibrinogen. Expression of wild-type or mutant CALR had no effect on αIIbβ3-mediated cell adhesion to fibrinogen. Moreover, each cell adherent to fibrinogen showed apparently the similar extent of spreading.

On the other hand, platelets from 4 of 12 patients with JAK2V617F had an increase in PAC1 binding in the presence and absence of ADP compared with platelets from healthy subjects. All 4 (100%) of 4 patients with increased PAC1 binding had a thrombotic event while 4 (30%) of 13 patients with normal PAC1 binding did.

Our study suggests no functional activation of integrin αIIbβ3 by CALR mutation, which is contrary to a recent finding that CALR activates β1 integrin. Nonetheless, our finding is rather in line with a clinical finding of a low risk for thrombosis in patients with CALR mutation and may provide the molecular basis for the differential thrombotic risk between the patient with CALR and JAK2 mutations.

Disclosures

Fujiwara:Celgene: Honoraria, Other: Travel, Acomodations, Expenses.

Author notes

*

Asterisk with author names denotes non-ASH members.