Measurement of NK activity by the microcytotoxicity assay (MCA): a new application for an old assay

Natural killer (NK) cells are spontaneously cytotoxic immune effector cells with the ability to selectively destroy tumor cells without harming normal cells. To perform this function, NK cells utilize two main cytotoxicity pathways, the well known perforin/granzyme-mediated secretory/necrotic killing and the recently defined TNF family ligand-mediated non-secretory/apoptotic killing. The former mechanism is manifested mainly against a few cultured leukemia cell targets, while the latter mediates killing against a large variety of tumor cell targets. Therefore, the biological role and significance of these mechanisms might be different. The NK cell-mediated necrotic killing has been reliably and selectively measured in humans by the standard 4-h 51Cr release assay (CRA) against K562 myeloid leukemia cell targets. However, no standardized high throughput assay is available for testing the NK cell-mediated apoptotic killing. Here, we introduce the modified MCA as a convenient method for measuring perforin/granzyme-independent NK cell-mediated apoptotic killing. The assay is performed in microwells of Terasaki tissue culture microtest plates, using adherent tumor cell targets, which are selectively susceptible to non-secretory/apoptotic killing and resistant to secretory/necrotic killing mediated by NK cells. Target cells are plated in microwells and incubated overnight to adhere to the plastic surface and to regenerate cell surface-bound TNF family receptors. Following this adherence, target cells are co-incubated with freshly isolated human peripheral blood mononuclear leukocytes (PBMNL) or purified subpopulations of immune cells for 24 h in various effector/target (E/T) ratios. During this incubation, dead target cells become non-adherent and are removed by washing the wells. Remaining adherent (viable) target cells are fixed, stained and optically counted. A notable dose-dependent (peak at 200:1 E/T ratio), time-dependent (peak at 24 h of incubation) and donor-dependent killing of tumor cells was consistently and reproducibly induced by PBMNL of normal donors. Using purified subpopulations of immune cells, it was demonstrated that among PBMNL, CD3(-)CD56(+)CD16(+) mature NK cells are the only mediators of tumor cell killing in MCA, as well as in CRA. Comparative studies of NK activity detected by MCA and CRA, performed with PBMNL from normal individuals and breast cancer patients, showed no significant correlation between the cytotoxicities measured in the two assays. In addition, while NK activity measured in CRA was normal in most breast cancer patients, NK activity assessed in MCA was decreased in a large majority of the patients. Thus, MCA is a sensitive NK assay, which is biologically different from CRA, and may be clinically relevant. MCA has also a higher throughput, and is more practical and economical than CRA.