Cord blood transplantation (CBT) for high-risk hematologic cancers is limited by the low numbers of immune cells in a single CB unit, leading to diminished graft-versus-leukemia effect. Although natural killer (NK) cells can mediate potent graft-versus-leukemia effect, and are the first reconstituting lymphocytes after transplantation, the receptor-ligand interactions mediating their cytotoxicity are not well understood. I first studied killer-cell immunoglobulin-like receptor (KIR) and HLA genotypes, NK phenotype and function for 110 CBT recipients to identify specific patterns of KIR-HLA interaction that might predict for CBT effectiveness. I found that the donor genotype of HLA-C1-KIR2DL2/3 combined with KIR2DS2/3 to be an important predictor of disease control after CBT in patients with an HLA-C1/C1 or HLA-C1/C2 background. These findings suggest means to improve the clinical efficacy of NK cells in HLA-defined patient subgroups, especially those with HLA-C2 homozygosity. I then extended my studies to investigate the role of NK cells in the control of CMV reactivation, as cord blood grafts are known to be more susceptible to latent virus infection from lack of transfer of adaptive subsets unlike other graft sources. I found that CB grafts expressing a NKG2C deletion allele possessed higher risk of CMV reactivation post CBT, with the risk significantly reduced with the presence of the wild type allele. Results collectively suggested that the susceptibility of CBT recipients to CMV reactivation is determined by the NKG2C content of the infused CB units. Based on the current understanding of NK licensing/education, the most critical developmental process required for functional competence, I attempted to identify markers that distinguish between 'licensed/educated' and 'unlicensed/hyporesponsive' NK cells. I discovered differential adhesive properties between the two functionally distinct subsets, and also report the possible contribution of the adaptor protein CrkL in NK licensing/education.
