Jak2V617F Bone Marrow Exposure Leads to Dysfunction of Normal NK Cells

Introduction: We have previously demonstrated a functional impairment of NK cells in the Jak2V617F murine model and peripheral blood cells from patients with Myeloproliferative Neoplasms, with emphasis on the increased frequency of immature NK cells and reduced cytotoxicity of the more mature CD27-CD11b+ fraction. It is not known whether MPN initiating cells are intrinsically resistant to tumor immunity or whether they can modify NK cells of the leukemic microenvironment, leading to dysfunction. Aims: To clarify these mechanisms, the present work aimed to model the leukemic exposure of normal NK cells and verify their modifications using a Jak2V617F chimeric transplant model. Methods: For transplantation, a non-myeloablative conditioning (single sublethal irradiation of 400cGy) was performed to partially preserve the recipient's hematopoiesis (50-60% of donor engraftment). 5 x 10 ⁶ total bone marrow (BM) cells from Jak2VF or Jak2WT control animals (both C57BL/6, CD45.2) were injected retro-orbitally into wild-type mice (C57BL/6, CD45.1, N=6 per group). Engraftment was evaluated by flow cytometry according to CD45.2/CD45.1 differential ratio. After engraftment was confirmed, frequency, receptors, and functional subtypes of exposed NK cells were evaluated. NK cytotoxicity against YAC-1 target cells was evaluated after culture with serum and supernatant of whole BM from chimeric animals. Results: The frequency of host NK cells (CD45.1) exposed to Jak2VF BM was lower than that of NK cells exposed to Jak2WT BM (Jak2WT 44.05% vs Jak2VF 19.76%, p<0.0001). Evaluation of functional subtypes showed that host NK cells had a lower frequency of immature secretory (Jak2WT 44.32% vs Jak2VF 30.70%, p=0,0407) and higher frequency of tolerant (Jak2WT 12.43% vs Jak2VF 24.34%, p=0.0016) and cytotoxic (Jak2WT 7.15% vs Jak2VF 19.30%, p<0.0001) NK cells when exposed to leukemia than when exposed to normal BM. In comparison, secretory immature NK cells from the leukemic BM (CD45.2) were low numbered as compared to Jak2WT NK cells (Jak2WT 73.30% vs Jak2VF 35.74%, p<0.0001) and cytotoxic NK cells were more frequent (Jak2WT 7.28% vs Jak2VF 19.16%, p=0.0003). In addition, host NK cells presented decreased expression of the activating receptor DNAM-1 when exposed to leukemia than when exposed to normal BM (Jak2WT 51.20% vs Jak2VF 46.10%, p=0.0006), whereas CD244 and NKG2D activating receptors were not differentially expressed. In contrast, the expression of TIGIT (but not NKG2A) inhibitory receptor was increased after leukemia exposure (Jak2WT 14.83% vs 46.10%, p=0.0003). Regarding the effect of the leukemic medium on NK cells, cytotoxicity was not altered when WT NK cells were exposed to serum or BM supernatant from chimeric transplanted animals. However, NK cells exposed to Jak2VF chimeric serum were activated according to CD69 expression (Jak2WT 0.060% vs Jak2VF 0.30%, p=0.0114) but presented lower expression of the CD107a degranulation marker (Jak2WT 91.92% vs Jak2VF 62.34%, p=0.0075) than NK cells exposed to Jak2WT cells. Conclusions: Our findings demonstrated that leukemia exposure led to altered NK cell frequency and maturation distribution as demonstrated by reduced immature and an increased tolerant and cytotoxic fraction. Changes in the expression profile of activating and inhibitory receptors were also observed. Therefore, we concluded that the leukemic environment leads to NK dysfunction and that strategies that aim to control NK maturation and the interaction between stem cell ligands and functional NK receptors can be explored to restore immunosurveillance in JAK2V671F Myeloproliferative Neoplasms.