Lymphocytes, especially CD4+ T cells, are essential for clearance of the yeast-like organism Cryptococcus neoformans from the infected host. The mechanism(s) by which the lymphocytes facilitate elimination of cryptococci has not been elucidated. It is generally thought, however, that lymphocytes reactive with C. neoformans indirectly function by production of lymphokines to enhance clearance of the organism by natural effector cells such as macrophages. In the present study, we assessed the ability of freshly isolated human lymphocytes to interact directly with C. neoformans and to limit the growth ofthe organism in vitro. We found that large granular lymphocytes (LGL) as well as T cells bound to cryptococcal cells when the lymphocytes were mixed with the cryptococcal cells at a 2:1 ratio. The physical binding interactions of the two lymphocyte populations were different. LGL attached to the cryptococcal cells by many microvilli; T lymphocytes associated with the yeast through broad areas of membrane attached to the cryptococcal cell surface. The two types of lymphocyte interactions did not result in phagocytosis but resulted in direct inhibition of cryptococcal growth, making these lymphocyte interactions with cryptococci distinctly different from interactions of monocytes with cryptococci. With the human natural killer (NK) cell line, NK 3.3, we confirmed that NK cells that were present in the LGL population were capable of limiting the growth of C. neoformans. Through immunoelectron microscopy, human CD3+ lymphocytes were seen attached to cryptococcal cells and by mass cytolysis, human CD3+ lymphocytes were shown to be responsible for inhibition of C. neoformans growth. The direct inhibitory interactions of NK cells and T lymphocytes with cryptococcal cells may be important means of host defense against this ubiquitous organism that frequently causes lifethreatening disease in AIDS patients. (J. Clin. Invest. 1993.

Murine natural killer (NK) cells have been shown to bind to and inhibit the growth of Cryptococcus neoformnans in vitro and to contribute to clearance of the organism in vivo. However, it is unclear whether NK cells actually kill cryptococci or simply inhibit proliferation of the fungal target. Therefore, the studies presented here were designed to determine whether NK cells are fungicidal to C. neoformans targets. C.

Earlier investigations have shown that murine natural killer (NK) cells bind to and inhibit the growth of the fungal pathogen Cryptococcus neoformans in vitro and in vivo. To define the stages of NK cell-mediated inhibition of C. neoformans growth and the requirements for the completion of these stages, the events which lead to cryptococcal growth inhibition were compared with those previously elucidated for NK cell-mediated tumor cell lysis. Our data indicate that NK cell-cryptococci binding is a distinct event that precedes inhibition; is temperature independent, although it is slowed at 4°C; and is Mg2' dependent. In contrast to binding, NK cell-mediated cryptococcal growth inhibition is temperature, Mg2+, and Ca2+ dependent. The removal of Ca2+ by EDTA addition within 3 h after maximal NK cell-cryptococci binding significantly reduced cryptococcal growth inhibition, indicating that Ca2+ is required either late in the NK cell trigger stage or early in the inhibitory stage. These stages and requirements are similar to those previously demonstrated for the model of NK cell-mediated tumor cell lysis; however, the NK cell-cryptococci interactions are somewhat slower than the interactions which culminate in the lysis of the YAC-1 tumor cell targets. These results suggest that C. neoformans cells, although structurally distinct from the standard tumor cell targets, are capable of similar cell-to-cell interactions with NK effector cells as the tumor cell targets.

Murine natural killer (NK) cell-mediated inhibition of growth of a yeast-like target cell, Cryptococcus neoformans, was completely abrogated by blocking the effector cell secretory process with monensin. Therefore, further studies were performed to determine the ability of various cytoplasmic fractions of NK cells to mediate inhibition of cryptococcal growth. Percoll-fractionated homogenates of rat LGL tumor cells demonstrated that the granule-containing fractions plus three additional sets of less dense cytoplasmic fractions displayed anti-cryptococcal activity; whereas only the cytoplasmic granule-containing fractions had cytotoxic activity against YAC-1 tumor cell and sheep erythrocyte targets. Maximal cryptococcal growth inhibition induced by LGL granules occurred after a 1 h incubation, required the presence of Ca2+ (1.0 mM) or Mg2+ (0.5 mM or 5.0 mM), and was completely abrogated in the presence of rabbit anti-LGL granule IgG. Cytolysin, the granule component which mediates tumor cell and sheep erythrocyte lysis, effectively limited the growth of cryptococci. Since Percoll gradient fractionation of the LGL homogenates demonstrated three separate peaks of anti-cryptococcal activity other than the granule peak, it is possible that the cytolysin-containing granules are not the only subcellular component of NK cells playing a role in inhibition of C. neoformans growth.

Previous reports demonstrate that natural killer (NK) cells inhibit the growth of Cryptococcus neoformans in vitro, but conclusive evidence supporting the effectiveness of NK cells in host resistance to cryptococci is not available. The objective of these studies was to assess the ability of NK cells to clear C. neoformans from the lungs, livers, and spleens of infected mice. CBA/J mice were depleted of NK cells, as well as other natural effector cells, by an intraperitoneal injection of cyclophosphamide (Cy), 240 mg/kg of body weight. One day later, 7.5 x 107 nylon wool-nonadherent (NWN) spleen cells, either untreated or treated with anti-asialo GM1 and complement to remove NK cells, were adoptively transferred to Cy-pretreated mice. On day 2 after Cy treatment, the mice were injected intravenously with 2 x 104 cryptococci. At 4 and 6 days after Cy treatment, tissues were assayed for NK reactivity, using a 4-h 51Cr-release assay, and for in vivo clearance of cryptococci as reflected by mean log1o CFU per organ. We observed that Cy treatment depleted NK activity against YAC-1 targets and reduced in vivo clearance of C. neoformans from the tissues of infected mice. Additionally, Cy treatment depleted the total lung and spleen cellularity and the total number of peripheral blood lymphocytes when compared with those in normal untreated control mice. Also, spleen weights were significantly decreased in comparison with those of untreated animals 4 days after Cy treatment. Adoptive transfer of untreated NWN spleen cells into Cy-depressed mice restored the NK cell activity which correlated with enhanced clearance of cryptococci from lungs, livers, and spleens. In contrast, treatment of NWN spleen cells with anti-asialo GM1 and complement before adoptive transfer abrogated the ability of these cells to restore NK activity or reduce the numbers of cryptococci present in tissues of infected mice. Taken together, these data indicate that NK cells are the cells effective in diminishing the numbers of cryptococci in tissues of infected mice. Consequently, NK cells may play a role in first-line host resistance against C. neoformans.