TNFR1 Blockade Rather then TNFR2 Blockade Reduces Chronic Inflammation in JAK2^+/V617F Mice

Introduction: Chronic myeloproliferative neoplasms (CMN) are characterized by expansion of myeloid cells, splenomegaly and thrombosis. Recent studies have demonstrated that chronic inflammation promotes the disease development as mutated hematopoietic cells are stimulated by pro-inflammatory cytokines. The most common driver mutation in BCR-ABL negative CMN is the activating point mutation JAK2-V617F. The constitutively activated JAK2 signaling leads to strongly elevated TNF levels in polycythemia vera, essential thrombocythemia and primary myelofibrosis (Fleischman et al, Blood, 2011). Recently, it has been reported that defective negative regulation of Toll-like receptor signals leads to excessive TNF in monocytes from CMN. Interestingly, this is not directly driven by JAK2-V617F in a cell-intrinsic manner (Lai et al., Blood, 2019). Thus, blockade of TNF signaling may represent a valuable anti-inflammatory pharmacological target in CMN. Therefore, in the Vav1-Cre x JAK2^+/V617F (JAK2^VF+) mouse model, we investigated the effects of TNFR1 or TNFR2 blockade using neutralizing antibodies (Ab) and we applied genetic disruption of the TNFRs.

Methods: To study the genetic disruption of TNFR1 or TNFR2, TNFR1^-/- or TNFR2^-/- mice were crossed with the JAK2^VF+ mouse model. For αTNFR1 Ab study, JAK2^VF+ were crossed with a mouse (huTNFR1) expressing a chimeric TNFR1 consisting of the human extracellular domain and the murine transmembrane and intracellular domains, allowing treatment with an anti-human TNFR1 Ab. Over 3 weeks, JAK2^VF+ mice received i.p. 3 x per week αTNFR1 (H398; 20 mg/kg) or 2 x per week αTNFR2 (TR75-54.7; 5 mg/kg) Abs or specific IgG controls. Blood was collected weekly for blood count measurements. After treatment, cell composition of blood, bone marrow and spleen was analyzed by flow cytometry. Serum cytokines were measured using a bead-based assay. HSCs and progenitor cells (MPP) were isolated from untreated huTNFR1 x JAK2^VF+ mice and plated in MethoCult ± αTNFR1 or αTNFR2 (10µg/l). Colonies were analyzed after 7 days.

Results: Blood count and spleen size of TNFR1^-/- or TNFR2^-/- x JAK2^VF+ mice showed no differences as compared to TNFR^WT x JAK2^VF+ mice. In addition, the immune cell composition was similar to TNFR^WT x JAK2^VF+ mice. These results indicate that the disruption of a single cytokine pathway has no influence in this model, as other cytokines likely will compensate the signal during initiation of the disease.In contrary, αTNFR1 Ab treatment (n = 6) reduced the mean HCT after 3 weeks from 72.5 to 60.0%, which was not observed in the control group (n = 5). This was partly a result of a reduced MCV in αTNFR1 treated mice (34.9 versus 30.9 fl) as there was no reduction of the RBC. WBC number increased upon αTNFR1 treatment from 14.3 to 17.9 x 10⁹ cells/l. Remarkably, αTNFR1 Ab treatment resulted in a major decrease of total cytokines as TNF, IL-1β, IL-10 and others when compared to the control group. Thus, αTNFR1 treatment is able to reduce chronic inflammation in the JAK2^VF+ mouse but has minor impact on the HCT and WBC. Similar experiments were performed using αTNFR2 Ab (n = 4). This treatment showed only a minimal change on the complete blood counts of αTNFR2 treated mice in comparison to the IgG treated mice (n = 4). Thus, the HCT of αTNFR2 treated mice declined from 79.3 to 71.3 %, MCV and RBC remained stable during treatment phase. Of note, αTNFR2 treatment had no major impact on serum cytokine levels. Further, no differences in spleen size or immune cell composition were detected. In total, αTNFR2 therapy failed to reduce inflammation in the JAK2^VF+model. To test whether αTNFR1/2 Ab treatment would impact on the stem cell compartment, a clonogenic assay was performed employing HSCs and MPPs ± αTNFR1 or αTNFR2. Almost no (αTNFR1) or minor (αTNFR2) reduction in colony number was measured; indicating that blockade of TNFRs has minor impact on the stem cell compartment in the JAK2^VF+ mouse.

Conclusions: Our study employing the JAK2^VF+ mouse model revealed an involvement of TNF-TNFR1 in induction of chronic inflammation rather than TNF-TNFR2. αTNFR1 treatment strongly reduces cytokine levels but only slightly reduces HCT. Therefore targeting TNFR1 in CMN may have beneficial effects on chronic inflammation. In our view, a clinical trial investigating combination therapy with JAK1/2 inhibitors such as Ruxolitinib in patients with persisting symptoms of chronic inflammation is warranted.