The JAK2V617F mutation is a unifying feature in the majority of patients with myeloproliferative neoplasms (MPNs). The presence of JAK2V617 in a hematopoietic stem (HSC) or progenitor cell confers a proliferative advantage over native JAK2 counterparts. Several inflammatory cytokines are elevated in MPN patients and in murine models of JAK2V617F-driven MPN. In particular, expression of tumor necrosis factor alpha (TNF-α) is increased by constitutive JAK2 kinase activity and imparts a competitive advantage on JAK2V617Fexpressing cells.

To identify cell types responsible for increased levels of TNF-α we performed intracellular cytokine staining in leukocytes from myelofibrosis (MF) patients and normal controls. Using a panel of markers to define cellular subsets we found expression of TNF-α was uniformly low in unstimulated cells. However, when cells were treated with lipopolysaccharide (LPS), TNF-α expression was 16-fold higher in HSCs of MF patients compared to normal controls (p<0.005), while expression in mature populations was not different. In a murine model of JAK2V617F driven MPN, where JAK2V617F cells are identified by GFP, TNF-α expression was 3-fold higher in JAK2V617F-expressing HSCs (GFP+) compared to GFP-controls (p<0.005), irrespective of LPS stimulation.

To evaluate anti-TNF therapy in MPNs, we treated JAK2V617F mice with etanercept, a soluble TNF receptor fusion protein that binds and inactivates TNF-α. Etanercept did not significantly restrain JAK2V617F-associated WBC or hematocrit increases over a 10-week period, despite suppression of TNF-α activity in plasma. As TNF-α may activate pro-apoptotic (predominantly through TNF receptor 1, TNFR1) and pro-survival pathways (predominantly through TNF receptor 2, TNFR2), we reasoned that global blockade of TNF-α may not shift the balance in favor of normal hematopoiesis. To investigate this we sorted primitive (Lin-, cKit+) cells from mice with JAK2V617F-induced MPN by TNF receptor expression. TNFR2+ cells showed significantly increased colony formation compared to TNFR1+ cells, demonstrating that TNFR2 expression is associated with increased clonogenic potential. Additionally we treated these primitive cells with specific antibodies blocking TNFR1 or TNFR2. Colony assays performed after 3 days in liquid culture with TNFR-blocking antibodies confirmed that interrupting TNFR2-mediated survival signaling resulted in decreased colony formation (61% reduction) while blocking TNFR1 increased colony formation (5% increase). In addition we tested CD34+ cells from MF patient samples with TNFR-blocking antibodies. As seen with the murine MPN cells, colony formation was decreased with the TNFR2 blocking antibody (30% reduction) and increased with the TNFR1 blocking antibody (48% increase) consistent with a differential activation of survival signals by TNF-α in MPN cells. These studies were performed without addition of exogenous TNF-α, which confirms the ability and requirement for these primitive JAK2V617Fcells to utilize TNF-α production to actively support survival.

Our data suggest that TNF-α generated by primitive MPN cells promotes their survival through activation of TNFR2, while TNFR1 activation is suppressive. If selective inhibition of TNFR2 shifts the equilibrium toward normal hematopoietic cells, this would support the use of TNFR2 blockade to treat MF and other myeloproliferative neoplasms. Genotyping of human MF and murine MPN colonies cultured with TNF receptor blocking antibodies and experiments in normal human CD34+ cells are ongoing and will be presented.


Deininger:BMS, Novartis, Celgene, Genzyme, Gilead: Research Funding; BMA, ARIAD, Novartis, Incyte, Pfizer: Advisory Board, Advisory Board Other; BMS, ARIAD, Novartis, Incyte, Pfizer: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution