Jak2V617F Reversible Activation Shows an Essential Requirement for Jak2V617F in Myeloproliferative Neoplasms (MPNs)

JAK inhibitors (JAKi) improve symptoms and outcomes in MPNs; however, mutant allele burden does not substantively change with chronic therapy. Alternative MAPK signaling and/or cooperating mutations are thought to contribute to MPN cell persistence; however, we hypothesized current JAKi fail to effectively abrogate mutant JAK2 signaling. To that end, we developed a conditionally inducible mouse model allowing for sequential activation, then deletion, of Jak2^VF from its endogenous locus using a Dre-rox/Cre-lox (Jak2^RL) dual recombinase system (Figure 1A).

Dre mRNA electroporation of lineage-negative UbcCreER-Jak2^RL bone marrow (BM) was used to induce Jak2^VF ex vivo, followed by transplant into lethally-irradiated recipient mice in competition with Cd45.1 BM. Transplanted mice developed a highly penetrant MPN characterized by leukocytosis, polycythemia, hepatosplenomegaly, and BM megakaryocytic hyperplasia. Tamoxifen administration 12 weeks post-transplant to delete Jak2^VF resulted in abrupt normalization of blood cell parameters, spleen weights, and serum pro-inflammatory cytokines, with a marked prolongation in overall survival (OS) compared to MPN control mice. Reduced peripheral blood (PB) and BM Cd45.2 mutant fraction was also observed, including within hematopoietic stem/progenitor cell (HSPC) compartments and to a greater degree than ruxolitinib (RUX) therapy. Notably, treatment with CHZ868, a type II JAKi, showed improvements in regard to hematocrit and spleen weight reduction on par with deletion, as well as significant reductions in BM HSPC mutant fractions, albeit to a lesser degree than with Jak2^VF reversion (Figure 1B). Transplant of Jak2^VF deleted BM failed to form phenotypic disease in secondary recipient mice consistent with loss of disease-propagating MPN stem cells. RNA-Seq on sorted LSKs following Jak2^VF deletion showed loss of STAT5, MAPK, and MTORC1 target gene expression, as well as negative enrichment in TGFb and TNFa/NFkB pathways consistent with reduced pro-inflammatory and proliferative output. The alterations in MAPK signaling were supported by Ybx1 RT-PCR of BM cKit+ cells and BM pERK IHC, both of which were increased with RUX but potently suppressed with deletion suggesting improved JAK2 targeting can overcome MAPK-mediated persistence.

The initial proximity of Jak2^RL lox sites grants Cre-mediated recombination prior to Dre-mediated Jak2^VF knock-in allowing for assessment of Jak2^VF oncogenic dependency alone or in the context of co-occurring mutation. Given the known effects of Tet2 loss on enhanced MPN HSC fitness, we evaluated Tet2 loss on sustaining MPN with Jak2^VF reversion. UbcCreER-Jak2^RL/Tet2^f/f mice were pulsed with TAM to knock-out Tet2 followed by sacrifice and Dre Jak2^RL knock-in 6 weeks later. Mice transplanted with UbcCreER-Jak2^VF/Tet2^-/- cells exhibited enhanced leukocytosis, spleen weights and PB/BM chimerism compared to Jak2^VF mice; however, Jak2^VF reversion resulted in similar reductions in MPN disease parameters, including Cd45.2 fraction, suggesting Tet2 loss does not dramatically alter dependence on mutant JAK/STAT signaling. Notably, while Jak2^VF/Tet2^-/- mutant cells displayed increased serial replating capacity, this was abrogated with Jak2^VF reversion suggesting despite the enhanced fitness of Tet2^­-/- HSCs, Jak2^VF creates a unique dependency that renders double-mutant cells susceptible to eradication with Jak2^VF reversion.

In sum, Jak2^VF deletion abrogates MPN features, extends OS, and reduces mutant cell fraction with depletion of disease-sustaining MPN HSCs, either alone or in the setting of cooperating Tet2 loss. These data suggest JAK2^VF mutant-selective inhibition offers greater therapeutic potential than current JAKi and a potential curative option for the treatment of MPN.