Cytotoxic T lymphocytes (CTL) and natural killer (NK) cells use perforin to deliver serine proteases called granzymes into the cytosol of virally infected or tumor target cells to induce apoptosis. Mice deficient in perforin are severely immunocompromised as are humans with familial hemophagocyticlymphohistiocytosis due to mutations in the perforin gene. The mechanism of perforin action is poorly understood and is the goal of this proposal. Perforin multimerizes in membranes in a calcium-dependent manner. The original model of perforin action by forming pores in the plasma membrane has been called into question by experiments that show that perforin-treated cells are impermeant to low molecular weight dyes and are too small to allow granzymes through. In preliminary data perforin was found to induce rapid endocytosis of granzymes to form large vesicles (""""""""gigantosomes"""""""") that lyse to release granzymes into the target cell cytosol. This proposal will test a model of perforin acting via triggering endocytosis and then endosomolysis. It will identify the cellular trafficking pathways involved in perforin delivery of granzymes and determine the biochemical requirements for perforin activity. Mutational studies will identify the critical perforin domains required for membrane insertion, multimerization and endosomolysis. Structural studies will focus on the perforin C2 domain, hypothesized to undergo a structural change in response to binding Ca++, which is likely to be the first step in membrane binding and insertion.
