CROSS-TALK BETWEEN HUMAN NK CELLS AND MACROPHAGES: INFLUENCE OF THE TUMOR MICRO-ENVIRONMENT

Natural killer (NK) cells are important effectors of innate immune responses providing cellular immunity against tumor-transformed and virally-infected cells. The existence of cross-talks between NK cells and myeloid cells, in particular dendritic cells, is well established, but information on the cross-talk between NK cells and macrophages is scanty. These interactions have been analyzed using an in vitro reconstituted tumoral micro-environment, as a simplified model to define soluble factors involved and/or cell contact dependency. Autologous human NK cells and monocyte-derived macrophages were obtained from buffy coats of healthy donors after magnetic beads cell purification. Macrophages were polarized into M0, M1 and M2, using well described stimuli. First, the influence of human polarized macrophages on NK cell anti-tumoral activities was studied. The co-cultures between NK cells and macrophages were performed in direct contact or by treating NK cells with macrophage-conditioned media. Activating receptors expression and degranulation ability (CD107a assay) of NK cells were evaluated by flow cytometry. IFN-γ production by NK cells was quantified by RT-PCR and ELISA. Then, the effect of NK cell-derived IFN-γ on macrophage polarization was assessed. Gene expression of markers, cytokines and chemokines well described to characterized M1 or M2 polarization were evaluated by RT-PCR. In parallel, cytokine and chemokine secretion were detected by ELISA. M1 polarization was required to enhance IFN-γ production and degranulation by resting NK cells. M1 ability to activate NK cells was further confirmed by the upregulation of CD69 activation marker. Importantly, either soluble mediators and direct contact interactions were involved in this process. However, the level of expression of NKp44 and NKG2D resulted increased only when NK cells were treated with M1-conditioned medium (M1-primed NK cells). Higher NKp44 and NKG2D expression correlated with enhanced NK cell degranulation towards altered cells. Although both NK cell subsets upregulated both receptors, M1-secreted IL-1β was responsible for NKp44 induction on CD56dim population, whereas IFN-β released by M1 favored increased expression of NKG2D by the CD56bright counterpart. Importantly, M1 secretion of IFN-β triggered NK cell expression of IL-15 and IL-15Rα, inducing a mechanism of IL-15 cis-presentation. IL-15 cis-presentation strongly enhanced IFN-γ secretion, that was further sustained by 2B4-CD48 interactions during direct co-cultures. On the contrary, NKG2D upregulation was responsible for increased degranulation by M1-primed NK cells. In parallel, IL-15 trans-presentation mediated by M1, together with NKG2D and NKp30 engagement, were needed to trigger NK cell degranulation during direct contact interactions. On the other hand, IFN-γ secreted by M1-primed NK cells was sufficient not only to downmodulate CD206 and ALOX15 expression by alternatively-activated macrophages, but also to induce pro-inflammatory cytokine (IL-1β and IL-15) and chemokine (CCL-5, CXCL-9 and CXCL-10) production. Importantly, also CD80 and IL-15Rα, which expression is strictly associated to M1 phenotype, were upregulated. In conclusion, we demonstrate for the first time in a human model that IL-15/IL-15Rα complex plays a key role in the crosstalk between NK cells and M1 polarized macrophages. Both, cis and trans-presented IL-15 favors NK cell secretion of high amount of IFN-γ and enhances NK cell cytotoxic activity towards tumor cells. Furthermore, having determined a functional correlation between M1-derived IL-1β and NKp44 expression, we propose new effects of IL-1β on NK cell biology. Finally, we demonstrate that IFN-γ provided by activated NK cells is sufficient to partially revert the anti-inflammatory phenotype typical of alternatively-activated macrophages into a pro-inflammatory one. This confers to NK cells a potential involvement in TAMs re-education.