SummaryMalaria infection crisis, at which the parasitemia drops precipitously and the parasite loses infectivity to the mosquito vector, occurs in many natural malaria systems, and has not been explained. We demonstrate that in a simian malaria parasite (Plasmodium cynomolgi in its natural host, the toque monkey), 'the loss of infectivity during crisis is due to the death of circulating intraerythrocytic gametocytes mediated by crisis serum . These parasite-killing effects in crisis serum are due to the presence in the serum of cytokines tumor necrosis factor and interferon y, which are produced by the host as a result of the malaria infection . The killing activity of each cytokine is absolutely dependent upon the presence of additional, as yet unidentified factor(s) in the crisis serum . nfectivity of malarial infections to mosquitoes is due to the presence of circulating gametocytes, the sexual stages that fertilize in the midgut o£ a blood-fed mosquito. In several malaria host-parasite systems, it has been noted that peaks of parasitemia are associated with "crisis" in the infection (1-3) and the simultaneous reduction or loss of infectivity of the parasite to mosquitoes (4, 5); neither of these phenomena has been explained . During a blood infection of Plasmodium cynomolgi in its natural host, the toque monkey, Macaca sinica, peak parasitemia is often accompanied by a crisis in the infection; this is most pronounced in splenectomized animals and is characterized by the appearance of morphologically abnormal intraerythrocytic parasites. At crisis, there is a sudden loss of infectivity of the parasites to mosquitoes, which persists for 5 to 7 days. We show here that the loss of infectivity at crisis is due to death of circulating intraerythrocytic gametocytes mediated by crisis serum . These killing effects in the crisis serum are due to the presence of the cytokines TNF and IFN-y ; the killing activity of each cytokine is absolutely dependent upon the presence of additional, as yet unidentified, factor(s) in the crisis serum . Materials and MethodsAnimals. The toque monkey, Macaca sinica, the natural host ofP. cynomolgi ceylonensis, was used in this study. Wild-caught adult animals of either sex, weighing between 1 and 4 kg, which were free of malaria infections as determined by the indirect immunofluorescence test, were used for experiments. Splenectomies were performed using standard sterile techniques as previously described (6) .

SummaryThe percentage of peripheral blood mononuclear cells (PBMC) bearing the CD3 § phenotype and the ot/~ and 3//8 T cell receptors (TCR) in PBMC were examined in Plasmodium vivax malaria patients and convalescents. The cells were labeled with monoclonal antibodies, stained with either fluorescene or phycoerythrin, and examined by ultraviolet (UV) microscopy. A highly significant increase in both the proportion and the absolute numbers of 3//8 T cells (p <0.005 and <0.001, respectively, Student's t test) was observed in nonimmune P. vivax patients during clinical paroxysms compared to nonmalarial controls. These T cells, which normally constitute not more than 3-5% of PBMC, constituted ~<30% of PBMC during paroxysms in these nonimmune patients in whom the clinical symptoms were severe. A less significant increase of 3//8 T cells were also observed in these nonimmune patients during infection, between paroxysms and during convalescence. In contrast, in an age-matched group of semi-immune patients resident in a malaria-endemic region of the country, in whom the clinical disease was comparatively mild, there was no increase in 3//8 T cells either during infection, even during paroxysms, or convalescence.The severity of disease symptoms in patients as measured by a clinical score correlated positively with the proportion of 3//8 T cells in peripheral blood (r = 0.53, p <0.01), the most significant correlation being found between the prevalence and severity of gastrointestinal symptoms, nausea, anorexia, and vomiting, and the proportion of 3//8 T cells (r = 0.49, p = 0.002). These findings suggest that 3//8 T cells have a role to play in the pathogenesis of malaria, possibly in the general constitutional disturbances and particularly in gastrointestinal pathology in malaria.T lymphocytes, which express the CD3 marker and lack CD4 and CD8 markers, have been shown to lack mature messenger (m)RNA for ot and/3 genes and express the y and 8 genes and proteins (1). Such lymphocytes expressing the TCR-3//6 constitute a minor subpopulation in the peripheral blood of adults, and many of them reside in the spleen, and to a lesser extent in the thymus, tonsils (2), and the epithelia of the large intestine (3) and lung (4). The biological role of the cells bearing the TCR-3,/8 is as yet unclear, but there are indications that some 3//8 T cells can mediate cytotoxicity (4). 3'/8 T cells behave like c~/B T cells in that they secrete the lymphokines IL-2, IFN-y, TNF-o~ and ~3, and express cytotoxic activity (5-7).Recent evidence has led to speculation about an involvement of 3//8 T cells in immunity and/or immunopathology in human malaria. The frequency of 3//8 T cells in the peripheral blood of acute Plasmodium falciparum malaria-infected individuals has been shown to increase, and remain elevated for several weeks during convalescence (8). In nonimmune individuals, a vast majority of T cells vigorously proliferating in response to freeze-thawed extracts of P. falciparum blood stage parasites in vitro have been shown to be 3//8 ...