TNFR2 Signaling Promotes Myeloproliferative Disease Progression: A Role for MAPK8?

The majority of patients diagnosed with Philadelphia negative (Ph^-) myeloproliferative neoplasms (MPNs) harbor somatic, gain-of-function mutations in JAK2, CALR or MPL. All of these mutations are associated with constitutive JAK/STAT signaling which confers a proliferative advantage to MPN cells and leads to malignant myeloid expansion at the expense of normal hematopoiesis. The bone marrow (BM) microenvironment in MPNs, particularly myelofibrosis (MF), is characterized by high concentrations of inflammatory cytokines, and we have previously shown that MF cells generate tumor necrosis factor alpha (TNF) in a JAK2-dependent manner. Elevated TNF promotes the survival of JAK2^V617F mutant cells over their JAK2^WT counterparts, creating a feedback loop in which the mutant cells enhance the inflammatory environment that supports their survival and expansion (Fleischman et al. Blood. 2011 Dec 8;118(24):6392-8).

To determine which hematopoietic lineages contribute to increased levels of TNF in MPN, we measured intracellular TNF expression in immunophenotypically defined white blood or BM cells from MF patients and normal controls. TNF expression was relatively low in unstimulated cells. However, lipopolysaccharide (LPS) treatment induced a 16-fold greater increase of TNF expression in hematopoietic stem cells (HSCs) from MF patients relative to normal controls (p<0.005), while expression in mature populations was similar. In HSCs isolated from mice with JAK2^V617F - induced MPN, TNF levels were 3-fold higher in JAK2^V617F-expressing HSCs (GFP⁺) compared to GFP^- controls (p<0.005), irrespective of LPS treatment.

TNF signaling is mediated by two receptors, TNF receptor 1 (TNFR1) and TNF receptor 2 (TNFR2). To evaluate the contribution of each receptor to the TNF-mediated survival of MPN cells, we cultured CD34⁺ cells from MF patients or normal controls for 72 hours ± TNFR1 or TNFR2 blocking antibodies (BA), then plated in clonogenic assays. While TNFR1 BA had no effect, TNFR2 BA reduced colony numbers in MF specimens by 26% (p<0.01) with no effect on normal cells. In a patient with residual JAK2^WT cells, TNFR2 BA reduced total colony counts by 32%, but increased JAK2^WT colonies by 16%. We also infected MF and normal CD34⁺ cells with inducible TNFR1 or TNFR2 shRNAs. Induction of TNFR1 shRNA had no effect on colony numbers, while the TNFR2 shRNA inhibited colony formation by 31% (p<0.005), without inhibiting normal controls. For additional validation, we purified lineage^- (Lin^-) Kit⁺ cells from mice with JAK2^V617F - induced MPN, followed by liquid culture and clonogenic assays as described for human cells. Colonies were genotyped for JAK2^V617F expression through GFP detection by flow cytometry. TNFR2 BA reduced the percentage of JAK2^V617F colonies by 89% (p<0.01), while TNFR1 BA had no effect. Altogether these data indicate that TNF promotes MPN in a TNFR2-dependent manner.

To identify TNF signaling networks that differentiate the TNF responsiveness of cells expressing JAK2^V617F compared to JAK2^WT, BM cells were harvested from JAK2^V617F MPN mice and cultured for 16 hours ± TNFR1 or TNFR2 BA. Lin^-Kit⁺ cells were sorted based on GFP (JAK2^V617F) expression and 3 independent experiments were tested with Affymetrix Mouse Genome 430 2.0 Arrays. We focused our analysis on genes with differential regulation between JAK2^V617F and JAK2^WT, whoseexpression was equalized with TNFR2 BA. We then prioritized genes with reported involvement in TNF signaling. The top candidate was Mapk8 (Jnk1) whose expression was strongly downregulated in untreated JAK2^V617F relative to JAK2^WT expressing cells (3.4-fold; p<0.005) and restored to the level of JAK2^WT with TNFR2 BA. Preliminary experiments show that MAPK8 mRNA expression is also lower (3.5-fold) in human CD34⁺ MF cells relative to normal controls. These data suggest an autoregulatory circuit where: (i) JAK2^V617F MPN cells produce TNF that activates TNFR2 to suppress MAPK8, (ii) TNF-induced apoptosis via MAPK8 activation limits JAK2^WT cells which maintain MAPK8 expression (Saadatzadeh et al. Blood. 2009 Mar 19;113(12):2655-60), (iii) JAK2^V617F cells exploit this survival advantage to favor their expansion and further MPN pathogenesis. Additional studies are ongoing to validate the role of MAPK8 downregulation as a survival mechanism employed by MPN cells. Our data implicate the TNF/TNFR2/MAPK8 network as a potential therapeutic target in MPN.