Differential effector pathways regulate memory CD8 T cell immunity against Plasmodium berghei versus P. yoelii sporozoites

Abstract

Malaria results in >1,000,000 deaths per year worldwide. Although no licensed vaccine exists, much effort is currently focused on subunit vaccines that elicit CD8 T cell responses directed against Plasmodium parasite liver stage Ags. Multiple immune-effector molecules play a role in antimicrobial immunity mediated by memory CD8 T cells, including IFN-gamma, perforin, TRAIL, Fas ligand, and TNF-alpha. However, it is not known which pathways are required for memory CD8 T cell-mediated immunity against liver stage Plasmodium infection. In this study, we used a novel immunization strategy to generate memory CD8 T cells in the BALB/c mouse model of P. berghei or P. yoelii sporozoite infection to examine the role of immune-effector molecules in resistance to the liver stage infection. Our studies reveal that endogenous memory CD8 T cell-mediated protection against both parasite species is, in part, dependent on IFN-gamma, whereas perforin was only critical in protection against P. yoelii. We further show that neutralization of TNF-alpha in immunized mice markedly reduces memory CD8 T cell-mediated protection against both parasite species. Thus, our studies identify IFN-gamma and TNF-alpha as important components of the noncytolytic pathways that underlie memory CD8 T cell-mediated immunity against liver stage Plasmodium infection. Our studies also show that the effector pathways that memory CD8 T cells use to eliminate liver stage infection are, in part, Plasmodium species specific.

Figure 1. Similar kinetics and magnitude of memory CD8 T cell responses following DC-CS_252-260-prime, LM-CS_252-260-boost immunization of WT and mutant BALB/c mice

(A) Experimental design. Parasite-specific memory CD8 T cell responses were generated in WT and mutant BALB/c mice as described in Materials and Methods. The magnitude, kinetics and phenotype of CS_252-260-specific memory CD8 T responses in the blood, liver or spleen were characterized greater than 60 days post immunization. (B) Representative H2-K^d/CS_252-260 tetramer staining used to detect memory CD8 T cell responses. (C) Kinetics of splenic CS_252-260-specific memory CD8 T cell responses in WT and mutant BALB/c mice. Data represent Mean +/− SD for 3 independent experiments with 3–5 mice/group. (D) Magnitude of memory CD8 T cell responses in the blood of WT and mutant BALB/c mice detected following ex vivo stimulation of PBMC with CS_252-260 peptide and ICS for TNF-α. (E) Magnitude of the CS_252-260-specific memory CD8 T cell responses in the spleen, liver and peripheral blood of WT and mutant BALB/c mice as detected by tetramer staining. Analyses were performed >60 days post DC-prime, LM-boost immunization. Data represent Mean +/− SD for 3 independent experiments with 3 mice/group.

Figure 2. Similar phenotypic and functional properties of CS_252-260-specific memory CD8 T cells from WT and mutant BALB/c mice following DC-prime, LM-boost immunization

(A) Gating strategy and cell surface staining histograms for representative WT and IFN-γ^−/− (GKO) mice are shown. (B) Summary of phenotypic analyses of CS_252-260-specific memory CD8 T cells recovered from spleen, liver and blood of WT and mutant BALB/c mice >80 days post-immunization. Data in B represent analyses of 6–9 individual mice/group from multiple independent experiments. (C) TCR Vβ chain utilization of total splenic CD8 T cells recovered from the spleens of naïve WT and mutant BALB/c mice. Data represent the Mean +/− SD for 3 mice/group. (D) TCR Vβ chain utilization of K^d/CS_252-260 tetramer-positive memory CD8 T cells recovered from the spleens of WT, GKO and PKO BALB/c mice 114 days following DC-prime, LM-boost immunization. Data represent the Mean +/− SD for 3 mice/group. (E) Functional avidity of CS_252-260- specific memory CD8 T cells recovered from the spleens of the WT, GKO and PKO mice. The concentration of CS_252-260 peptide required to elicit the ½ maximal response was determined as described in Materials and Methods. Data represent the Mean +/− SD for 3 mice/group. (F) Polyfunctionality of WT and mutant (IFN-γ-competent) CS_252-260-specific memory CD8 T cells. PBMC from the indicated mouse strains were stimulated ex vivo in the presence of CS_252-260 peptides. The proportion of single, double or triple positive cytokine-producing cells was determined by Boolean gating using FloJo software. Data represent Mean +/− SD for 7–10 individual mice/strain.

Figure 3. IFN-γ, but not perforin, TRAIL or FasL, contributes to memory CD8 T cell-mediated protection against liver stage P. berghei infection

(A) WT, GKO and PKO BALB/c mice were immunized as described in Figure 1. Eighty to 120 days following immunization the frequency of circulating CS_252-260-specific memory CD8 T cells in all mice was determined by K^d/CS_252-260 tetramer staining. Each bar represents an individual mouse and the horizontal line denotes the 1% numerical threshold required for sterilizing immunity in >95% of WT BALB/c mice (34). Two to 3 days following T cell analyses, mice were challenged i.v. with 1000 P. berghei sporozoites. The presence of blood stage parasitemia was evaluated with Giemsa stain 10 days later. Filled bars represent parasitized mice of the indicated genotype. Numbers indicate % protection (no. protected/no. challenged). Data in A are pooled results from 2 independent experiments (χ²=1.21, p=0.271 for WT vs PKO; χ²=11.62, p=0.00065 for WT vs GKO). (B-C) WT and TRAIL-deficient BALB/c mice (B) or WT and gld BALB/c mice (C) were immunized and T cell numbers were evaluated in individual mice as described above. Mice were challenged with 1000 P. berghei sporozoites and blood stage parasitemia was evaluated with Giemsa stain of thin blood smears 10 days later. Numbers indicate % protection (no. protected/no. challenged). Data in B are pooled results from 2 independent experiments (χ²=0.952, p=0.329 for WT vs TRAIL-KO). Data in C are pooled results from 2 independent experiments (χ²=1.56, p=0.211 for WT vs gld).

Figure 4. TNF-α contributes to memory CD8 T cell-mediated protective immunity against liver stage P. berghei infection

(A) Experimental design. Circulating CS_252-260-specific memory CD8 T cell frequencies in individual WT mice were determined using tetramer staining 67 days following immunization. 500μg of anti-TNF-α or rat IgG control Abs were administered one day before and one day following challenge of mice with 1000 P. berghei sporozoites. (B) Diminished memory CD8 T cell-mediated protection following TNF-α neutralization in WT BALB/c mice bearing >1% circulating CS₂₅₂-specific PBMC. Each bar represents an individual mouse. Filled bars indicate parasitized mice as determined by Giemsa stain of thin blood smears 10 days following challenge. Numbers indicate % protection (no. protected/no. challenged). Data are pooled results from 2 independent experiments (χ²=8.36, p=0.004 for rIgG vs anti-TNF-α).

Figure 5. IFN-γ and perforin, but not TRAIL or FasL, contribute to memory CD8 T cell-mediated protective immunity following P. yoelii sporozoite challenge

(A) WT and mutant BALB/c mice were primed with DC-CS_280-288 and boosted 7 days later with LM-CS_280-288. Greater than 70 days following immunization, the frequency of circulating CS_280-288-specific memory CD8 T cells in individual mice was determined by H2- K^d/CS_280-288 tetramer staining. Mice were subsequently challenged with 100 P. yoelii sporozoites and monitored for blood stage parasitemia on d10 p.i. Filled bars represent parasitized mice of the indicated genotype. Numbers indicate % protection (no. protected/no. challenged). Data in A are pooled results from 2 independent experiments (χ²=5.31, p=0.021 for WT vs GKO; χ²=7.94, p=0.005 for WT vs PKO). (B) WT, gld and TRAIL-deficient BALB/c mice were immunized as described in A. Greater than 60 days following immunization, mice were challenged with 100 P. yoelii sporozoites and protection was evaluated as described above. Data in B are pooled results from 2 independent experiments (χ²=0.004, p=0.949 for WT vs gld; χ²=0, p=1.0 for WT vs TRAIL-KO).

Figure 6. Diminished memory CD8 T cell-mediated protection against P. yoelii sporozoite challenge following TNF-α neutralization in WT BALB/c mice

(A) Experimental design. T cell analyses were performed 57 days following DC-CS_280-288- prime, LM-CS_280-288-boost immunization of WT BALB/c mice. 500μg of anti-TNF-α or rat IgG control Abs were administered one day before and one day following challenge of mice with 100 P. yoelii sporozoites. Protection was evaluated as described above. Numbers indicate % protection (no. protected/no. challenged). Data are pooled results from 2 independent experiments (χ²=7.50, p=0.006 for rIgG vs anti-TNF-a).

Figure 7. Generalized inflammation protects against P. berghei, but not P. yoelii, liver stage infection

Experimental overview (A) and tabulated results (B) of P. berghei and P. yoelii sporozoite challenge studies conducted with the TLR agonists, CpG, LPS and PolyI:C. Data in B are pooled results from 3 independent experiments. Numbers indicate % protection (no. protected/no. challenged). (C) Neutralization of TNF-α or IFN-γ abrogates CpG-induced protection against P. berghei sporozoite challenge. Neutralizing antibodies were administered to mice at the time of sporozoite challenge. Data in C are results from a single experiment. (D) P. yoelii is less susceptible to generalized inflammation induced by administration of CpG to naïve WT BALB/c mice. CpG was administered at the indicated time surrounding P. yoelii sporozoite challenge. Data in D are pooled results from 2 independent experiments.