Umbilical cord mesenchymal stem cells increase expansion of cord blood natural killer cells.

Natural killer (NK) cell-mediated cytotoxicity can control leukemia relapse while protecting patients from graft-versus-host disease (GVHD) after allogeneic stem cell transplant. Cord blood (CB) is rich in NK cell progenitors with similar properties of proliferation and cytotoxicity as adult blood NK cells. Hence, it is attractive to expand and potentially utilize these cells for adoptive immunotherapy. In this study, CB mononuclear cells were CD3-depleted by immunomagnetic microbead selection to remove T cells. This CD3(dep) CB-MNC fraction was then plated for ex vivo expansion, with or without a feeder layer of irradiated umbilical cord mesenchymal stem cells (UC-MSC), with or without cytokines that have been shown to be critical for NK expansion: IL-2, IL-15, IL-3, and FLT-3L. At an average of 2 weeks of culture, there was significantly higher expansion (64.7 +/- 8.4-fold) of CD56(+)/CD3(-) NK cells in the presence of the UC-MSC feeder layer and cytokines compared to controls (no increase with feeder layer only and 6.4 +/- 1.5-fold increase with cytokines only, P < .05). Contact between CD3(dep) CB-MNC cells and UC-MSC augmented NK expansion. The combination of all 4 cytokines was superior to IL-2 alone or 2 cytokines combinations: mean 64.7 +/- 8.4-fold expansion with 4 cytokines combination versus IL-2 alone, IL-2 + FLT-3L, IL-2 + IL-15 or IL-2 + IL-3 (12.2 +/- 2.0, 14.4 +/- 2.4, 10.4 +/- 4.1, 25.2 +/- 8.1 respectively). We also observed that only fresh CD3(dep) CB-MNC preparations could be expanded reliably, whereas frozen and thawed CD3(dep) CB-MNC cells did not expand consistently (mean fold increase 6.5 +/- 3.2). Cytotoxicity of expanded NK cells was compared with NK cells from fresh and overnight IL-2 activated CD3(dep) CB-MNC. Whereas fresh cells displayed no discernible killing, strong cytotoxicity against K562, Raji, REH, and SUP-B15 cells lines was noted after overnight activation in IL-2. Cytotoxicity of expanded NK cells against Raji, REH, and SUP-B15 was lower, which, however, correlated with a predominant expansion of CD56(+)/CD16(-) cells known to have less cytolytic activity than CD56(+)/CD16(+). To test the transfection efficiency in NK cells, fresh or expanded CD3(dep) CB-MNC cells were electroporated with either DNA or mRNA constructs for GFP. DNA had a low transfection efficiency (<10%), whereas the one for mRNA reached 52%, but at the cost of significant cell death. Our results suggest that CB NK cell progenitors can be expanded to obtain large numbers by using an irradiated feeder of UC-MSC. They maintain an elevated cytotoxic profile, and may be genetically manipulated-all characteristics that make them suitable for cellular therapies.