Identifying Mechanisms of Enhanced NK Cell Cytotoxicity Using Permanent NK Cell Lines.

Although NK cells are promising candidates for adoptive immunotherapy and at least one permanent cell line is in clinical trials, further studies evaluating efficacy and mechanisms of action are warranted. As a first step towards identifying the most potent effector cells, we investigated the molecular mechanisms of cytotoxicity of the three natural killer lines, KHYG-1, NK-92 and YT, and the NK-T cell line, SNT-8, under standardized culture conditions with human serum as the only serum source. We confirmed the previously established differential killing potential of the 4 cell lines against target K562 cells using a new method based on detecting Annexin V (+) target cells by flow cytometry. By labeling the NK cells with specific antibodies, the assay is designed to screen any target cell for NK cytotoxicity. In contrast to previous reports, we found KHYG-1 the most cytotoxic, followed by NK-92, SNT-8 and YT. Genotypic and transcriptional phenotypic analysis of the cell lines for killer cell Ig-like receptors (KIRs) by SSP-PCR showed that inhibitory KIRs outnumbered activating KIRs in all cases but did not explain the differential cytotoxicity. A correlation with cytotoxicity was found with expression of the activating type II C lectin-like receptor, NKG2D: KHYG-1, 99%+; NK-92, 91%+; SNT-8, 6%+ and YT, 2%+. Moreover, the ITAM-bearing adaptor molecule DAP12, involved in the alternative activation signaling pathway via NKG2C-CD94 and activating KIRs, was detected only for KHYG-1 by immunoblotting, These data suggest that the superior cytotoxicity of KHYG-1 may be due, in part, to the additional activation of this alternative pathway that is not triggered in the other lines. The downstream signaling molecules involved in NK cytotoxicity, including the tyrosine phosphatases SHP-1, SHP-2 and SHIP-1 (inhibitory), as well as SHIP-2, the tyrosine kinases ZAP-70, Syk, PI3K and the MAP kinase phospho-ERK-2 (activating) were compared among the lines by immunoblotting followed by densitometry normalized to b-actin or ERK-2 for phospho-ERK-2. We found that the activating kinase Syk was expressed only in NK-92 and KHYG-1 at even higher levels. Also, phospho-ERK-2, was hyperphosphorylated only in KHYG-1. Perforin, granzyme A and granzyme B, present in cytotoxic granules, were compared by RT-PCR and intracellular flow cytometry and/or immunoblotting. Perforin was found to be almost exclusively fully processed to the active 60 kD form only in KHYG-1, in contrast to the other lines, which displayed approximately half the levels of the active form. These data provide a further explanation for the superior cytotoxicity of KHYG-1 and demonstrate the value of comparing cell lines with diverse cytotoxic potential as a means of elucidating cell killing mechanisms. It is conceivable that targeted modifications to the signaling pathways for cytotoxicity in this model will lead to the generation of activated NK cells with even greater efficacy.