Abstract 795

Elevated levels of pro-inflammatory cytokines such as TNF-alpha (TNF) have been implicated in the constitutional symptoms commonly seen in JAK2 V617F MPN. However, it is unclear to what extent they contribute to the disease process itself. We hypothesized that JAK2 V617F may confer a selective advantage on hematopoietic progenitors residing in the high cytokine milieu uniquely present MPN patients, focusing on the pro-inflammatory cytokine TNF.

We used ELISA and qPCR to measure TNF in plasma and neutrophils of MPN patients and normal controls. All MPN subtypes had increased plasma concentrations of TNF, with primary myelofibrosis, polycythemia vera and essential thrombocythemia displaying 6, 4 and 2 fold higher mean concentrations of plasma TNF as compared to normal controls, respectively. Median neutrophil TNF mRNA was 2 fold higher in PV patients as compared to normal controls (p<0.001). We then tested the effect of TNF on CFU-GM and BFU-E formation from MPN vs normal mononuclear cells. At low concentrations (1ng/ml) of TNF we observed a decrease in BFU-E from normal controls to 82% of that without TNF but an increase in MPN BFU-E to 120% (p<0.05) of that without TNF. We also observed a blunting of TNF induced CFU-GM suppression in MPN versus normal. We genotyped colonies and found that in the presence of TNF colony formation was almost exclusively JAK2 V617F, demonstrating that TNF selects for JAK2 V617F colony growth. To distinguish between direct and indirect TNF effects (i.e. effects mediated through bystander cells) as well as to potentially augment the differential effects of TNF on normal and MPN by the use of a more homogenous cell population we performed equivalent colony formation experiments on CD34+ cells. The most notable difference observed was the effect 10ng/ml TNF on CFU-GM. TNF reduced CFU-GM colony formation by normal CD34+ cells to 50%, but increased CFU-GM colony formation by MPN cells to 120% (p<0.001). The suppressive effect of 10ng/ml on BFU-E was also blunted in MPN versus normal CD34+ cells (49% vs 30% of that without TNF) (p=0.054). Genotyping of MPN colonies allowed us to compare the effect of TNF on colony formation on three distinct populations: (1) JAK2WT MPN CD34+ cells, (2) JAK2V617F MPN CD34+ cells and (3) normal CD34+ cells. As compared to normal controls (59% decrease), JAK2 non-mutated MPN CD34+ cells are hypersensitive to myeloid colony suppression (78% decrease) whereas TNF paradoxically enhances colony formation by JAK2 V617F CD34+ cells (180% increase). This suggests that “pre-JAK2” CD34+ cells are functionally disadvantaged when compared to those from normal controls and implicates TNF in the selection and clonal expansion of the JAK2 V617F clone. To confirm the latter we tested whether ectopic expression of JAK2 V617F in murine hematopoietic progenitors blunts TNF-induced inhibition of myeloid colony formation. In the presence of TNF, colony formation of hematopoietic progenitors expressing empty vector or BCR-ABL (included as an additional control) was decreased to 60%, while TNF had no effect on progenitors with ectopic expression of JAK2 V617F. This suggests that JAK2 V617F directly imparts TNF resistance. To interrogate the role of TNF for the development of MPN in vivo, we infected TNF+/+ and TNF−/− bone marrow with JAK2 V617F-GFP retrovirus and transplanted the cells into genotypically matched recipients. Two strains of mice were used in independent experiments (C57BL6 and B6129S6). Equal numbers of TNF+/+ and TNF−/− GFP+ cells per mouse were injected. Although peripheral blood counts were comparable, JAK2 V617F burden (assessed by GFP%) was significantly lower in recipients of TNF−/− knockout bone marrow compared to recipients of TNF+/+ marrow. This difference in JAK2 V617F burden between the two groups was apparent in peripheral blood (47% vs 3.7%, p <0.05) and bone marrow (52% vs 2%, p<0.05) and the spleens of the TNF−/− were smaller (0.46g vs 0.11g, p=0.078).

Our data suggest that JAK2 V617F-induced TNF resistance may drive the expansion of the mutant clone by conferring a growth advantage to JAK2 V617F cells over their JAK2 wild-type counterparts in the high cytokine microenvironment of MPN patients. Further, JAK2WT MPN cells exhibit intrinsic TNF hypersensitivity compared to normal cells, which greatly enhances this selective advantage. This JAK2V617F-negative state may play a role in the predisposition to acquire MPN.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution