NONINHIBITORY BINDING OF HUMAN INTERLEUKIN-2-ACTIVATED NATURAL-KILLER-CELLS TO THE GERM TUBE FORMS OF CANDIDA-ALBICANS

During incubation in vitro with yeast or germ tube forms of Candida al bicans, only 2 to 6% of freshly isolated human natural killer (NK) cel ls (>85% CD16(+), CD56(+), CD3(-); <15% CD3(+); cytolytic for the NK-s usceptible target K562 but not for the NK-resistant target DAUDI), wer e seen to interact with the fungal cells. As seen under the electron m icroscope, the contact area had a limited extent and was narrow, and n either the surface nor the intracytoplasmic organization of the NK cel l was altered, In contrast, more than 30% of interleukin-2-activated N K (LAK) cells (>96% CD16(+), CD56(+), CD3(-); 1.5% CD3(+); cytolytic f or both K562 and DAUDI targets) interacted closely with the fungus. Th is interaction was particularly extensive with the surface of the fung al germ tube that was intimately enveloped by villous protrusions from the lymphocyte surface. The fungus-interacting LAK cell also showed a remarkable redistribution of surface microvilli and polarization of c ytoplasmic organelles, such as the Golgi apparatus, centrioles, and gr anules, toward the area of fungal contact. Together with the elevated cytolytic potential against the K562 and DAUDI targets, all the morpho logical data suggested the presence of a potentially active lytic mach inery in the fungus-interacting LAK cell. Nonetheless, two independent assays for anticandidal activity did not show consistent killing or f ungal growth inhibition by either fresh NK or LAK cells. While offerin g direct evidence of the strong interaction between human LAK cells an d the germ tubes, precursors of tissue-invasive hyphal forms of C. alb icans, our observations also suggest that this interaction may not be sufficient to kill the fungus or arrest its growth.