Susceptibility of human melanoma cells to autologous natural killer (NK) cell killing: HLA-related effector mechanisms and role of unlicensed NK cells.

BACKGROUND: Despite Natural Killer (NK) cells were originally defined as effectors of spontaneous cytotoxicity against tumors, extremely limited information is so far available in humans on their capability of killing cancer cells in an autologous setting. METHODOLOGY/PRINCIPAL FINDINGS: We have established a series of primary melanoma cell lines from surgically resected specimens and here showed that human melanoma cells were highly susceptible to lysis by activated autologous NK cells. A variety of NK cell activating receptors were involved in killing: particularly, DNAM-1 and NKp46 were the most frequently involved. Since self HLA class I molecules normally play a protective role from NK cell-mediated attack, we analyzed HLA class I expression on melanomas in comparison to autologous lymphocytes. We found that melanoma cells presented specific allelic losses in 50% of the patients analyzed. In addition, CD107a degranulation assays applied to NK cells expressing a single inhibitory receptor, revealed that, even when expressed, specific HLA class I molecules are present on melanoma cell surface in amount often insufficient to inhibit NK cell cytotoxicity. Remarkably, upon activation, also the so called "unlicensed" NK cells, i.e. NK cells not expressing inhibitory receptor specific for self HLA class I molecules, acquired the capability of efficiently killing autologous melanoma cells, thus additionally contributing to the lysis by a mechanism independent of HLA class I expression on melanoma cells. CONCLUSIONS/SIGNIFICANCE: We have investigated in details the mechanisms controlling the recognition and lysis of melanoma cells by autologous NK cells. In these autologous settings, we demonstrated an efficient in vitro killing upon NK cell activation by mechanisms that may be related or not to abnormalities of HLA class I expression on melanoma cells. These findings should be taken into account in the design of novel immunotherapy approaches against melanoma