Synergistic effect of IL-12 and IL-18 induces TIM3 regulation of γδ T cell function and decreases the risk of clinical malaria in children living in Papua New Guinea

Abstract

Background: γδ T cells are important for both protective immunity and immunopathogenesis during malaria infection. However, the immunological processes determining beneficial or detrimental effects on disease outcome remain elusive. The aim of this study was to examine expression and regulatory effect of the inhibitory receptor T-cell immunoglobulin domain and mucin domain 3 (TIM3) on γδ T cells. While TIM3 expression and function on conventional αβ T cells have been clearly defined, the equivalent characterization on γδ T cells and associations with disease outcomes is limited. This study investigated the functional capacity of TIM3+ γδ T cells and the underlying mechanisms contributing to TIM3 upregulation and established an association with malaria disease outcomes.

Methods: We analyzed TIM3 expression on γδ T cells in 132 children aged 5-10 years living in malaria endemic areas of Papua New Guinea. TIM3 upregulation and effector functions of TIM3+ γδ T cells were assessed following in vitro stimulation with parasite-infected erythrocytes, phosphoantigen and/or cytokines. Associations between the proportion of TIM3-expressing cells and the molecular force of infection were tested using negative binomial regression and in a Cox proportional hazards model for time to first clinical episode. Multivariable analyses to determine the association of TIM3 and IL-18 levels were conducted using general linear models. Malaria infection mouse models were utilized to experimentally investigate the relationship between repeated exposure and TIM3 upregulation.

Results: This study demonstrates that even in the absence of an active malaria infection, children of malaria endemic areas have an atypical population of TIM3-expressing γδ T cells (mean frequency TIM3+ of total γδ T cells 15.2% ± 12). Crucial factors required for γδ T cell TIM3 upregulation include IL-12/IL-18, and plasma IL-18 was associated with TIM3 expression (P = 0.002). Additionally, we show a relationship between TIM3 expression and infection with distinct parasite clones during repeated exposure. TIM3+ γδ T cells were functionally impaired and were associated with asymptomatic malaria infection (hazard ratio 0.54, P = 0.032).

Conclusions: Collectively our data demonstrate a novel role for IL-12/IL-18 in shaping the innate immune response and provide fundamental insight into aspects of γδ T cell immunoregulation. Furthermore, we show that TIM3 represents an important γδ T cell regulatory component involved in minimizing malaria symptoms.

Keywords: IL-12; IL-18; Malaria; Plasmodium; TIM3; γδ T cells.

Fig. 1

TIM3 expression is maintained on γδ T cells after drug treatment and parasite clearance. C57BL/6 mice were infected with P. chabaudi and then drug-treated with chloroquine and pyrimethamine. (a) Liver lymphocytes and (b) splenocytes were stained for TIM3 expression at different time points following end of drug cure to assess the percentage of TIM3+ γδ T cells, (c) and (d) number of TIM3+ γδ T cells, and (e) and (f) the total number of γδ T cells in the liver and spleen. The data represent three mice per time point and shows mean ± standard deviation (SD). Chronically P. chabaudi-infected mice (n = 8) were assessed on day 98 post-infection for TIM3+ γδ T cells in the (g) liver and (h) spleen. The data represent two independent experiments. Statistical analysis was performed using (a-f) paired t tests with Holm-Sidak method or (g) and (h) Student’s t test. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 2

γδ T cell TIM3 expression in individuals from malaria endemic areas is increased compared to healthy controls (HC). PBMCs from individuals living in malaria endemic areas recently infected in the last 30 days, not recently infected, and healthy controls (HC) were surface stained for TIM3 expression. a Frequency of γδ T cells expressing TIM3 (recently infected, filled triangles, n = 92; not recently infected, open circles, n = 40; and healthy controls, filled circles, n = 20). b Percentage of γδ T cells expressing TIM3 in individuals recently infected with either P. vivax only (filled circles; n = 32), P. falciparum only (filled squares; n = 13), or co-infected with multiple species (P. vivax, P. falciparum, P. ovale, and P. malariae) (filled triangles; n = 47). c Frequency of TIM3+ γδ T cell subsets distinguished by expression of CD27 and CD45RA (TCM central memory T cell, TEM effector memory T cell, TEMRA terminally differentiated effector memory T cell). d Frequency of TIM3-expressing CD16+ (open circles) and CD16– (squares) TEMRA γδ T cells (mean ± SD). Statistical analysis was performed using (a–c) Kruskal-Wallis tests with Dunn’s post-test; c multiple comparison to naive, and (d) paired Student’s t tests.*P < 0.05, **P < 0.01, ***P < 0.001

Fig. 3

Infection and exposure to multiple malaria species are associated with upregulated TIM3 expression. C57BL/6 mice were infected with P. chabaudi and then drug-treated with chloroquine and pyrimethamine. (a) Liver lymphocytes and (b) splenocytes were stained for TIM3 expression on day 7 following completion of drug cure to assess the percentage of TIM3+ γδ T cells from mice which received either three sequential P. chabaudi infections (Multiple P. chabaudi, n = 5), single P. chabaudi infection (P. chabaudi Day 7, n = 9), or single P. chabaudi infection 98 days prior to assessment (P. chabaudi Day 98, n = 3). C57BL/6 mice were infected with either P. chabaudi or P. berghei and then drug-treated. (c) Liver lymphocytes and (d) splenocytes were assessed for TIM3+ γδ T cells on day 7 following end of drug cure from mice which received either P. chabaudi infection followed by P. berghei infection (P. chabaudi + P. berghei, n = 8), P. berghei infection followed by P. chabaudi infection (P. berghei + P. chabaudi, n = 9), or single P. berghei infection (P. berghei Day 7, n = 5). The data represent two independent experiments. Statistical analysis was performed using Kruskal-Wallis tests with Dunn’s post-test; comparison to naive mice. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 4

TIM3+ γδ T cells are functionally inactive following stimulation with iRBCs and IPP. PBMCs from individuals living in malaria endemic areas were stimulated with iRBCs and then surface stained for TIM3 expression. Frequency of TIM3+ γδ T cells in (a) and (e) cytokine responders and non-responders and (b) and (f) in individuals with or without cytotoxically active γδ T cells after stimulation with iRBC or IPP (responders; filled squares and non-responders; open circles). Comparison of TIM3 expression by (c) and (g) cytokine-producing γδ T cells and (d) and (h) cytotoxic γδ T cells (TIM3+; filled circles and TIM3–; filled squares) in iRBC- or IPP-responding individuals. Statistical analysis was performed using (a, b, e, and f) Mann-Whitney tests and (c, d, g, and h) Wilcoxon matched pairs signed-rank test. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 5

Children who experience a clinical episode during follow-up have decreased frequency of TIM3+ CD16+ TEMRA γδ T cells. Frequency of TIM3+ CD16+ TEMRA γδ T cells in children with a clinical episode (filled circles; n = 50) versus children with asymptomatic infection (open circles; n = 72) during follow-up. Statistical analysis was performed using Mann-Whitney tests. *P < 0.05

Fig. 6

TIM3 is upregulated by IL-12/18 and IPP. PBMCs from healthy individuals (n = 10) were stimulated with (a) iRBCs (3:1), LPS, IPP, or cytokines (IL-6, IFN-γ, TNF-α, IL-12/IL-18, IL-4, IL-1β) or (b) IL-12 and IL-18 for 24 h and then surface stained for TIM3 expression. The frequency of γδ T cells expressing TIM3 was compared to that of unstimulated cells. Correlation of IL-18 plasma levels and TIM3 expression in (c) all children (n = 132), (d) children with P. falciparum infection at enrollment (n = 55), and (e) children with no P. falciparum infection at enrollment (n = 77). Statistical analysis was performed using Kruskal-Wallis tests with Dunn’s post-test (a and b) and Spearman rank correlation (c and d). **P < 0.01, ***P < 0.001