Expression of Jak2 R1063H Leads to Altered Epor/Tpor Signaling, Changes in HSC Fitness and Represents Increased Risk of Thrombosis

Non-canonical weakly activating germline JAK2 variants may predispose progenitor cells to acquire JAK2 V617F during the evolution of myeloproliferative neoplasms (MPN) (Lanikova et al. Blood 2016;128(18):2266) and represent risk factors for MPN transformation into AML (Benton et al. Cancer 2019;125:1855). We have previously described and functionally characterized the JAK2 R1063H germline variant, cooperating with the JAK2 E846D variant in a case of hereditary MPN with erythrocytosis and megakaryocytic atypia (Kapralova et al. Blood 2016;128:1418) or with the acquired JAK2 V617F causing increased JAK2 signaling and higher disease severity in patients with essential thrombocythemia (Mambet et al. Blood 2018;132:2695).

The mouse model bearing Jak2 R1063H mutation was created using microinjection of synthesized ssODNs containing the R1063H mutation, sgRNA mRNA and Cas9 protein into C57BL6/N-derived zygotes. In order to intensify a possible effect of the mutation, we bred and analyzed the knock-in Jak2 R1063H homozygous mice. The mice are born in normal (expected) frequency and display increased platelets (PLT) counts when compared to wild-type (wt) mice during life-span. Despite increased PLT counts, the plasma thrombopoietin level is normal whereas erythropoietin levels are increased without marked erythrocytosis. The preliminary co-immunoprecipitation experiments with human erythropoietin and thrombopoietin receptors (EPOR/TPOR) and JAK2 variants suggest increased TPOR and decreased EPOR binding with JAK2 R1063H kinase, possibly explaining the hematological phenotype. Also experiments with humanized mouse model of EPOR (Divoky et al. Proc Natl Acad Sci USA 2001;98:986) crossed with Jak2 R1063H mice supported these findings, where anemia due to hypoactivation of wt human EPOR signaling is more severe in Jak2 R1063H mice than in a control group and mild thrombocytosis is still present in the double mutant mice.

The Jak2 R1063H model displays increased number of long-term hematopoietic stem cells (HSC) suggesting increased cycling in this population at young age (3 months) and shift to increased short-term HSC and multipotent progenitors with aging (12 months). The immunophenotypically-defined myeloid progenitor cells (LK, Lin^-/Sca1^-/cKit^high) are statistically increased in Jak2 R1063H mice due to disproportionately increased megakaryocytic/erythroid progenitors (MEP) over other myeloid progenitors at 3 months of age. The LK compartment is compromised in old Jak2 R1063H mice due to substantial myeloid bias by disproportionately increased CMP cells and marked decreased MEP cells, explaining the more pronounced anemia of aging in these mice. Some of these features are reminiscent of aging-associated myeloid and megakaryocytic bias, increased LKs, and relative exhaustion of progenitors with an erythroid lineage-priming program in aged bone marrow (BM) of a mouse model of primary familial and congenital polycythemia (PFCP) due to gain-of-function EPOR (Kralova et al, Am J Hematol 2022; doi: 10.1002/ajh.26658). However, the gain-of-function EPOR-introduced changes into the BM microenvironment are not associated with augmented hypoxia signaling and systemic inflammation and do not represent permanent challenges that would contribute to an increased incidence of myeloid malignancies in PFCP. In this regard, it will be interesting to compare these characteristics with the possible predisposing potential of Jak2 R1063H; further analysis of proliferative history of Jak2 R1063H HSC populations and expression profiling are ongoing.

In the cohort of Jak2 R1063H mice, we also observed a relatively high frequency of sudden death in mice with median overall survival of 215 days (5/15) in comparison with the control group (0/15). We therefore measured the levels of D-dimers as a marker of thrombosis and found it statistically increased in the Jak2 R1063H group. Interestingly, in a WES study of patients with venous thromboembolism, the JAK2 R1063H was identified in one case, being considered a probable disease-causing variant (Lee at el. Blood Adv 2017;1:1224).

In conclusion, our results confirm a direct effect of germline Jak2 R1063H mutation on EpoR/TpoR signaling but also on HSC biology, suggesting that the variant may represent a risk factor for BM remodeling in the process of myeloid cell transformation.

Grant support: Czech Health Research Council NU21/03/00338.

No relevant conflicts of interest to declare.