Cutting Edge: Plasmodium falciparum Induces Trained Innate Immunity

Abstract

Malarial infection in naive individuals induces a robust innate immune response. In the recently described model of innate immune memory, an initial stimulus primes the innate immune system to either hyperrespond (termed training) or hyporespond (tolerance) to subsequent immune challenge. Previous work in both mice and humans demonstrated that infection with malaria can both serve as a priming stimulus and promote tolerance to subsequent infection. In this study, we demonstrate that initial stimulation with Plasmodium falciparum-infected RBCs or the malaria crystal hemozoin induced human adherent PBMCs to hyperrespond to subsequent ligation of TLR2. This hyperresponsiveness correlated with increased H3K4me3 at important immunometabolic promoters, and these epigenetic modifications were also seen in Kenyan children naturally infected with malaria. However, the use of epigenetic and metabolic inhibitors indicated that the induction of trained immunity by malaria and its ligands may occur via a previously unrecognized mechanism(s).

Figure 1

P. falciparum iRBCs and Hz induce trained immunity. (A) Schematic of trained immunity assay. Adherent PBMCs were treated with media (RPMI+), iRBCs, uRBCs, or Hz for 24 hr and then washed. After 3 d rest, cells were harvested for ChIP or challenged with Pam3CSK4 (10 μg/mL) for 4-24 hr. (B and C). TNFα, IL-6, and IL-10 ELISA measurements of supernatants from adherent PBMCs trained with iRBCs (B) or Hz (C), rested for 3 d, then challenged with Pam3CSK4 for 24 hr. Bars represent mean ± SEM for three (IL-10) or eight (TNFα, IL-6) donors (all comparisons to RPMI+ trained, *p<0.05, **p<0.01, ***p<0.001, paired t-test).

Figure 2

Malaria-induced training has wide-range effects on the transcriptional response to subsequent challenge. (A and B) mRNA was harvested from adherent PBMCs trained with iRBCs or Hz, rested for 3 d, and challenged with Pam3CSK4 for 4 or 12 hr. Levels of TNF (A) and IL6 (B) relative to RPMI+ trained cells are shown. Bars represent mean ± SEM for eight (4 hr Pam3CSK4) or three (12 hr Pam3CSK4) donors (**p<0.01, paired t-test). (C) Gene expression from cells treated as in (A) and (B) was quantified using NanoString technology. Expression counts are represented as log2 fold change compared to control-trained cells using hierarchical clustering. Each row represents one gene. A subset of the total genes measured is listed after each cohort. Values shown are the means of two independent experiments from one of the donors in (A) and (B).

Figure 3

Malaria-induced training is not inhibited by co-treatment with anti-IFNγ neutralizing mAb. TNFα and IL-6 ELISA measurements of supernatants from adherent PBMCs trained with iRBCs or Hz for 24 hr, rested for 3 d, and challenged with Pam3CSK4 for 24 hr. Cells were treated with control IgG or neutralizing antibody during the 24 hr training. Bars represent mean ± SEM for two donors (all comparisons to control IgG conditions were not significant, paired t-test).

Figure 4

Epigenetic and metabolic regulation of malaria-induced training. (A) H3K4me3 ChIP was performed on monocytes from Kenyan children with acute malaria (AM), the same children 6 weeks after therapy (6wk), or adult North American controls (NAM). Bars represent mean ± SEM for five (AM), three (6wk), or six (NAM) donors (*p<0.05, Kruskal-Wallis test). (B) H3K4me3 ChIP was performed on adherent PBMCs trained for 24 hr and rested for 3 d. Bars represent mean ± SEM for five (RPMI+, iRBC, uRBC) or three (Hz) donors (all comparisons to RPMI+ trained, *p<0.05, **p<0.01, paired t-test). (C and D) Adherent PBMCs were co-treated with MTA (C), rapamycin (D) or relevant vehicle controls during the 24 hr training stimulation. Cells were rested for 3 d and challenged with Pam3CSK4 for 24 hr. Cytokine concentrations in supernatants were assessed by ELISA. Bars represent mean ± SEM for three or four donors (all comparisons to no inhibitor conditions and were not significant unless indicated, paired t-test).