CD4+ CD25+ Foxp3+ regulatory T cells, dendritic cells, and circulating cytokines in uncomplicated malaria: do different parasite species elicit similar host responses?

Abstract

Clearing blood-stage malaria parasites without inducing major host pathology requires a finely tuned balance between pro- and anti-inflammatory responses. The interplay between regulatory T (Treg) cells and dendritic cells (DCs) is one of the key determinants of this balance. Although experimental models have revealed various patterns of Treg cell expansion, DC maturation, and cytokine production according to the infecting malaria parasite species, no studies have compared all of these parameters in human infections with Plasmodium falciparum and P. vivax in the same setting of endemicity. Here we show that during uncomplicated acute malaria, both species induced a significant expansion of CD4(+) CD25(+) Foxp3(+) Treg cells expressing the key immunomodulatory molecule CTLA-4 and a significant increase in the proportion of DCs that were plasmacytoid (CD123(+)), with a decrease in the myeloid/plasmacytoid DC ratio. These changes were proportional to parasite loads but correlated neither with the intensity of clinical symptoms nor with circulating cytokine levels. One-third of P. vivax-infected patients, but no P. falciparum-infected subjects, showed impaired maturation of circulating DCs, with low surface expression of CD86. Although vivax malaria patients overall had a less inflammatory cytokine response, with a higher interleukin-10 (IL-10)/tumor necrosis factor alpha (TNF-α) ratio, this finding did not translate to milder clinical manifestations than those of falciparum malaria patients. We discuss the potential implications of these findings for species-specific pathogenesis and long-lasting protective immunity to malaria.

FIG. 1.

Correlation between parasitemia and the proportion (%) of CTLA-4-expressing Treg cells in P. vivax-infected patients (circles; n = 39), P. falciparum-infected patients (triangles; n = 14), and patients coinfected with both species (squares; n = 13). The gating strategy to characterize CTLA-4⁺ Treg cells is shown in the supplemental material. Parasitemia in P. vivax-infected patients was strongly correlated with the proportion of CTLA-4⁺ cells among Treg cells (r = 0.419; P = 0.008). When P. falciparum-infected patients were added to the Spearman correlation model, the coefficient of correlation remained nearly unchanged (r = 0.400; P = 0.003). However, the overall correlation was substantially attenuated (r = 0.277; P = 0.024) when patients coinfected with both species were added to the model. Therefore, parasitemia was strongly correlated with the proportion (%) of CTLA-4-expressing Treg cells in single-species infections (with either P. vivax or P. falciparum) but not in mixed-species infections. Figure S4 in the supplemental material shows data for P. vivax, P. falciparum, and mixed-species infections in three separate plots.

FIG. 2.

Correlation between parasitemia and proportion of DCs that were phenotypically characterized as plasmacytoid (CD123⁺) DCs in P. vivax-infected patients (circles; n = 38), P. falciparum-infected patients (triangles; n = 15), and patients coinfected with both species (squares; n = 13). The gating strategy to characterize CD123⁺ DCs is shown in the supplemental material. Parasitemia in P. vivax-infected patients was significantly correlated with the proportion of CD123⁺ cells among DCs (r = 0.371; P = 0.022). When P. falciparum-infected patients were added to the Spearman correlation model, the coefficient of correlation changed little (r = 0.304; P = 0.027). However, the overall correlation was substantially attenuated (r = 0.215; P = 0.084) when patients coinfected with both species were added to the model. Therefore, parasitemia was positively correlated with the proportion (%) of CD123⁺ DCs in single-species infections (with either P. vivax or P. falciparum) but not in mixed-species infections. Figure S5 in the supplemental material shows data for P. vivax, P. falciparum, and mixed-species infections in three separate plots.

FIG. 3.

Proportions of phenotypically mature (CD86-expressing) DCs in P. vivax-infected study participants and their association with the number of circulating CTLA-4-expressing T regulatory cells. In the upper panels, P. vivax-infected subjects (n = 38) are distributed according to the proportions of circulating myeloid (left) and plasmacytoid (right) DCs that express the costimulatory molecule CD86. The gating strategy to characterize CD86⁺ DCs is shown in Fig. S5 of the supplemental material. The x axis represents the proportion (%) of DCs that express high levels of CD86 on the surface, and the y axis shows the absolute number of patients. For example, 9 and 15 patients had <10% of their circulating myeloid and plasmacytoid DCs, respectively, expressing high levels of CD86 (first bar, from left to right, in the plots in the upper panels). Note that for both myeloid and plasmacytoid DCs, the distribution of subjects according to CD86 expression levels in their DCs was bimodal, with two subgroups of subjects clearly defined according to the maturation status of their peripheral blood DCs. In the lower panels, we compare the absolute numbers of circulating CTLA-4-expressing T regulatory cells in these subgroups of subjects with small and large proportions of circulating phenotypically mature myeloid (left) and plasmacytoid (right) DCs. The difference in CTLA-4-expressing T regulatory cell counts was highly significant in both comparisons (P < 0.0001; Mann-Whitney test).