Cord Blood NK Cytotoxic Activity Can Be Selelctively Modulated by Accessory.

Introduction The expansion of cord blood (CB) derived natural killer (NK) cells is a promising cell therapy approach. CB NK cell expansion and activation has proven to be feasible and effective. However, the selectivity of NK killing in an allograft setting raises concern. Accessory cells, in particular dendritic cells and mesenchymal stem cells (MSCs), have been shown to interact with and modulate NK cell proliferation and function. We explored the ability of accessory cells to modulate NK activity in the setting of clinically-relevant CB NK expansion protocols. Methods Human MSCs were isolated from normal bone marrow donors. 5 × 106 CD56-selected NK cells or 5 × 107 unselected or CD3-depleted CB mononuclear cells (MNC) (averaging 10% CD56+ NK cells) were layered on top of the MSCs in the presence of 200u/mL IL-2. CB NK cells were tested for cytolytic activity by 51Cr release assay using MSCs and K562 AML cells. IFN-g secretion was detected by Elispot assay. Results Cord blood-derived NK cells activated by exposure to IL-2 efficiently kill allogeneic MSCs in vitro. A 4 hour chromium release assay revealed that CD56-selected CB NK cells lysed 60% of allogeneic MSCs at E:T ratio 10:1. MSC co-culture with CB NK induced IFN-g secretion as determined by IFN-γ Elispot assay. In contrast, unselected or CD3-depleted CB MNCs with equal numbers of CD56+ NK cells were unable to lyse MSCs despite IL-2 activation. This suggests that CB accessory cells can modulate the activation or cytolytic activity of CB-derived NK cells. To address these two possibilities, we tested the ability of these CB cultures to kill a classic NK target, K562 human myeloid leukemia cells. Surprisingly, in all three CB cultures NK efficiently lysed K562 cells, 95% of lysis at 10:1 E:T ratio by chromium release assay suggesting that these accessory cells did not prevent IL-2 activation of NK cells, but selectively-modulated the cytolytic activity. Thus, preserving anti-leukemia killing while protecting non-malignant MSCs. Conclusions Our data provide evidence for selective modulation of cord blood-derived NK cell cytolytic activity by CB accessory cells. Indeed, activation of cord-derived NK cells in the presence on CD3-depleted accessory cells yields NK cells capable of killing K562 target cells but not allogeneic MSCs. Although the underlying mechanism is unknown, this selective modulation of NK cells may be exploited in the clinical expansion of cord blood NK cells.