Abstract
The somatic mutation MPL W515L constitutively activates signaling pathways including JAK/STATs, MAPK, and PI3K/AKT, leading to myeloproliferative neoplasm (MPN). How each of these signaling pathways contributes to MPL W515L-induced MPN has not been well defined. Moreover, the requirement for the tyrosine residues 599 and 604, which have been shown to be critical for normal MPL signaling, in MPL W515L-induced disease remains uncharacterized. Thus, we created W515L/Y599F double mutant (abbreviated as MPL 515/599), W515L/Y604F double mutant (MPL 515/604), and the MPL W515L/Y599F/Y604F triple mutant (MPL 515/599/604) and tested their activities in the TPO-dependent G1ME cell line. Interestingly, MPL 515/604 supported TPO-independent proliferation similar to MPL W515L whereas MPL 515/599 exhibited significantly impaired ability to do so. In contrast, the triple mutant completely lost the ability to support TPO-independent proliferation. Consistent with their effect on cell proliferation, MPL W515L and MPL 515/604 significantly increased CFU-Mk in murine bone marrow while MPL 515/599 and the triple mutant did not. Moreover, MPL 515/599 exhibited impaired phosphorylation of Shc, STAT3, STAT5, AKT, and ERK while MPL 515/604 barely had any effect on their phosphorylation. Notably, MPL 515/599/604 further reduced STAT5 phosphorylation although it retained STAT3 phosphorylation as compared to MPL 515/599. Together, these results reveal that Y599 is critical in mediating MPL W515L-induced megakaryocyte hyperproliferation and signaling.
To further test which downstream signaling pathway is critical for MPL W515L-induced megakaryocyte hyperproliferation, we restored each signaling in MPL 515/599/604-transduced G1ME cells by overexpressing constitutively active forms of downstream signaling molecules and GFP. Cells overexpressing constitutively active STAT5 or AKT showed an increased GFP percentage over time with TPO depletion. However, constitutively active STAT3 or Ras did not affect cell growth. We verified the requirement for STAT5 in MPL W515L-induced megakaryocyte hyperproliferation by overexpressing MPL W515L in murine bone marrow progenitors from STAT5-null and littermate control mice. Deletion of STAT5 significantly reduced CFU-Mk and impaired CD41 expression. Notably, MPL W515L-induced expansion of CFU-Mk observed in littermate control cells was dramatically impaired in STAT5-null cells. These findings suggest that STAT5 is critical in MPL W515L-induced megakaryocyte hyperproliferation.
To further test the in vivo role of Y599 and Y604 in MPL W515L-induced MPN, we performed bone marrow transplantation. Consistent with the in vitro phenotype, MPL W515L and MPL 515/604 caused MPN characterized by hypercellularity of bone marrow, leukocytosis, thrombocytosis, splenomegaly, hepatomegaly, and infiltration of blood cells to liver and lung. Interestingly, the disease burden of MPL 515/604 was more severe, with higher WBC and platelet counts than that seen with MPL W515L. Surprisingly, MPL 515/599 and MPL 515/599/604 caused hypercellularity of bone marrow, although they did not cause any other significant features of MPN. MPL 515/599 also caused splenomegaly without other symptoms. Unexpectedly, all MPL mutant alleles supported expansion of LSK cells in bone marrow. Despite the finding that MPL W515L and MPL 515/604 caused thrombocytosis, they did not increase MEP numbers. In fact, all mutant MPL alleles except the triple mutant increased HPC, CMP, and GMP cells. In addition, mature myeloid and megakaryocyte lineages were expanded, possibly at the expense of erythrocytes. In the spleen, we also observed expansion of megakaryocytes in the double mutant alleles and a huge expansion of myeloid lineage cells, but a decrease of erythrocytes in MPL 515/604. Thus our study has revealed distinct features of Y599 and Y604 in mediating MPL W515L-induced signaling, megakaryocyte hyperproliferation, and MPN. Y599 plays a critical role and cooperates with Y604. Our study also suggests that MPL cytosolic phosphorylated Y599 and flanking sequences could become targets for pharmacologic inhibition in MPNs.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.