The MEK/ERK Inhibitor Trametinib Reduces Fibrosis in a Transduction-Transplantation Model of Mutated Calreticulin

Insertion or deletion mutations in calreticulin (CALR) are present in the majority of JAK2^V617F-negative MPN patients. We utilized a murine retroviral transduction-transplantation model to express the 52bp CALR deletion mutation (CALR^DEL) in both BALB/c and C57B/6 backgrounds. As described previously (Marty et al., Blood 2016;127:1317), recipients of CALR^DEL-transduced marrow developed persistent thrombocytosis without leukocytosis or erythrocytosis by two months post-transplant. Mice were euthanized at six and nine months post-transplant to evaluate the tempo of disease progression. At six months CALR^DEL mice had impressive expansion of megakaryocytes expressing the CALR^DEL mutant protein in the bone marrow (BM) without fibrosis or significant splenomegaly. By nine months BM fibrosis and splenomegaly were present. Both whole BM and spleen cells were able to serially transplant the MPN phenotype into secondary recipients.

When cultured in collagen-based media supplemented with thrombopoietin, CALR^DEL BM cells produced an increased number of megakaryocyte colonies as compared to empty vector. The increased colony formation potential of CALR^DEL bone marrow cells was limited to megakaryocytes, we found no increase in colony formation from CALR^DEL hematopoietic stem and progenitor cells in methylcellulose with cytokines supporting erythroid and GM colony formation. However, CALR^DEL enhanced the serial replating ability of LKS (lineage^neg, c-kit⁺ Sca-1⁺) cells.

Both pSTAT5 and pERK were increased in whole spleen lysates from CALR^DEL mice as compared to wild-type BALB/c mice. Therefore, we tested the impact of ruxolitinib, a JAK1/2 inhibitor, and trametinib, a MAPK/ERK inhibitor, on the MPN phenotype of CALR^DEL mice. At six months post-transplant mice were treated with either ruxolitinib (90mg/kg PO BID), trametinib (3mg/kg PO daily), or vehicle for 40 days. Ruxolitinib reduced pSTAT5 but caused a paradoxical increase in pERK in whole spleen lysates, while trametinib reduced pERK but not pSTAT5. Trametinib caused a transient increase in platelets and white cells. In spite of pharmacodynamic evidence of effective dosing, ruxolitinib had no significant effect on platelet or leukocyte count but did reduce hemoglobin slightly. Both ruxolitinib and trametinib reduced spleen weight. Ruxolitinib reduced the fraction of the mutant CALR^DEL allele (inferred from percentage of GFP⁺ cells) in the spleen but not the bone marrow, while trametinib had no impact on disease allele burden in any organ. Neither ruxolitinib nor trametinib reduced the expansion of megakaryocytes in the bone marrow but trametinib significantly reduced marrow fibrosis (average score MF-2.5 for vehicle, MF-1.75 for ruxolitinib, MF-1 for trametinib). To assess the role of STAT5 in the pathogenesis of the ET-like MPN induced by the CALR^DEL mutant, we transduced BM from syngeneic Balb/c donors carrying a floxed Stat5ab allele in combination with a Stat5ab null allele (Mx-Cre;Stat5ab^fl/-; Walz et al., Blood 2012;119:3550). Haploinsufficiency for Stat5ab significantly delayed the development of ET-like MPN and attenuated thrombocytosis, implicating JAK2-STAT5 signaling directly in the pathogenesis of this disease.

In summary, this CALR^DELmouse model results in an MPN phenotype resembling essential thrombocythemia followed by myelofibrosis. CALR^DELresults in expansion of megakaryocytes and platelets without expansion of other myeloid cell types. Both pSTAT5 and pERK are increased in our CALR^DEL model and pharmacologic inhibition of pERK results in reduction of fibrosis without reducing megakaryocytes. These studies implicate pERK as a potential anti-fibrosis therapeutic target in MPN.