LAUSR

TGF-β expression is elevated in most cases of myeloproliferative neoplasms (MPNs). However, the contribution of TGF-β to disease pathogenesis is not well understood. Prior studies have shown that TGF-β regulates hematopoietic stem cell (HSC) quiescence. There also is published evidence that increased TGF-β may contribute to myelofibrosis. However, this is controversial, as recent studies have implicated other inflammatory cytokines in the development of myelofibrosis. Here, we test two specific hypotheses. First, we hypothesize that increased TGF-β signaling in mesenchymal stromal cells (MSCs) is required for the development of myelofibrosis. Second, we hypothesize that Jak2 mutated HSCs are resistant to the growth suppressive effect of TGF-β, resulting in a competitive advantage that contributes to their clonal expansion in MPN and clonal hematopoiesis. To test the first hypothesis, we abrogated TGF-β signaling in mesenchymal stem/progenitor cells by deleting Tgfbr2 using a doxycycline-repressible Osterix-Cre transgene (Osx-Cre), which targets all mesenchymal stromal cells in the bone marrow. Osx-Cre was induced at birth (by removal of doxycycline), since we recently reported that the post-natal loss of TGF-β signaling in mesenchymal stromal cells has no discernible effect on basal hematopoiesis or the stem cell niche. We transplanted bone marrow cells from UBC-CreERT2; Jak2V617F mice or c-kit+ cells transduced with MPLW515L retrovirus into irradiated wildtype or Osx-Cre; Tgfbr2f/fmice. For the Jak2V617F model, mice were treated with tamoxifen 6 weeks post transplantation to induce mutant Jak2 expression. Of note, elevated Tgfb1 was present in both MPN models. MPLW515L induced a rapidly fatal MPN with reticulin fibrosis in the bone marrow. A similar hematopoietic phenotype was observed in Osx-Cre; Tgfbr2f/f recipients. However, myelofibrosis, as measured by reticulin staining and Collagen III (Col3a1) mRNA expression, was reduced, but not completely abrogated in Osx-Cre; Tgfbr2f/fmice. Likewise, in the Jak2V617F MPN model, the hematopoietic phenotype was similar in wildtype and Osx-Cre; Tgfbr2f/fmice. Although overt myelofibrosis was not observed in this MPN model, increased RNA and protein expression of Collagen III were detected in the bone marrow. Although still above baseline, Col3a1 expression was significantly reduced in Osx-Cre; Tgfbr2f/frecipient mice. To examine the role of canonical TGF-β signaling in the induction of myelofibrosis, we cultured bone marrow derived MSCs from wildtype and Osx-Cre; Smad4f/f mice, in which canonical TGF-β signaling is abrogated. Treatment of wildtype MSC cultures with TGF-β1 induced the expression of fibrosis associated genes, including Col1a1 and Acta2, and down-regulated expression of key niche factors, including Cxcl12, Scf, and Bglap. Surprisingly, TGF-β1-induced expression of Col1a1 was intact in Smad4-deleted MSC cultures. Conversely, treatment of wildtype MSC cultures with the JNK inhibitor SP600125 abrogated TGF-β1-induced expression of Col1a1. Collectively, these data suggest that TGF-β signaling in MSCs contributes to the development of myelofibrosis in MPN through activation of the non-canonical JNK pathway. To test the second hypothesis, we set up a competitive transplantation assays using wildtype and Jak2V617F hematopoietic cells in which Tgfbr2 had been deleted to abrogate TGF-β signaling in HSCs. Specifically, we transplanted UBC-CreERT2; Jak2V617F; Tgfbr2f/f and UBC-CreERT2; Tgfbr2f/f bone marrow cells at 1:5 ratio into lethally irradiated wild type recipients. For control, we transplanted a 1:5 ratio of UBC-CreERT2; Jak2V617F to wild type cells. Mice were treated with tamoxifen to activate Cre expression 6 weeks post transplantation. Complete blood counts and donor chimerism were measured 6, 10, 14 and 18 weeks after tamoxifen. As expected, in the control group, donor chimerism with Jak2 mutated cells increased over time in myeloid lineages. In contrast, in the Tgfbr2-deleted group, no expansion of Jak2 mutated cells was observed. These data support our hypothesis that Jak2 mutated HSCs are resistant to the growth suppressive effect of TGF-β, providing a selective advantage that contributes to their clonal expansion in MPN and possibly clonal hematopoiesis. No relevant conflicts of interest to declare.