Abstract
Human natural killer (NK) cells in peripheral blood spontaneously recognize and kill a wide variety of target cells. It has been suggested that ion channels are involved in the killing process because there is a Ca-dependent stage and because killings by presensitized cytotoxic T lymphocytes, which in many respects resembles NK killing, is associated with changes in K and Na transport in the target cell. However, no direct evidence exists for ion channels in NK cells or in their target cells. Using the whole-cell variation of the patch-clamp technique, we found a voltage-dependent potassium (K+) current in NK cells. The K+ current was reduced in a dose-dependent manner by the K-channel blockers 4-aminopyridine and quinidine and by the traditional Ca-channel blockers verapamil and Cd2+. We tested the effects of ion-channel blockers on killing of two commonly used target cell lines: K562, which is derived from a human myeloid leukemia, and U937, which is derived from a human histiocytic leukemia. Killing of K562 target cells, determined in a standard 51Cr-release assay, was inhibited in a dose-dependent manner by verapamil, quinidine, Cd2+, and 4-aminopyridine at concentrations comparable to those that blocked the K+ current in NK cells. In K562 target cells only a voltage-dependent Na+ current was found and it was blocked by concentrations of tetrodotoxin that had no effect on killing. Killing of U937 target cells was also inhibited by the two ion-channel blockers tested, quinidine and verapamil. In this cell line only a small K+ current was found that was similar to the one in NK cells. We could not find any evidence of a Ca2+ current in target cells or in NK cells; therefore, our results cannot explain the Ca dependence of killing. Our findings show that there are K channels in NK cells and that these channels play a necessary role in the killing process. In contrast, the endogenous channel type in the target cell is probably not a factor in determining target cell sensitivity to natural killing.