Transient Deficiency of Dendritic Cells Results in Lack of a Merozoite Surface Protein 1-Specific CD4 T Cell Response during Peak Plasmodium chabaudi Blood-Stage Infection

Abstract

Splenic dendritic cells are crucial for controlling the immune response to malaria by initiating a CD4 gamma interferon (IFN-γ) response early in a blood-stage infection, which contributes to parasite clearance as well as to acute-stage immunopathology. CD8⁻ CD11c^high dendritic cells have been described previously to be important antigen-presenting cells for induction of these CD4 T cell responses in the spleens of Plasmodium chabaudi-infected mice. However, when isolated during the period of maximum parasitemia and shortly thereafter, the dendritic cells transiently lose their ability to stimulate T cells, recovering only as the parasitemia is controlled. This loss of a CD4 T cell response is also observed in vivo during this part of the infection. CD4 T cells from a T cell receptor-transgenic mouse recognizing a peptide of merozoite surface protein 1 (MSP1) injected into BALB/c mice during peak parasitemia proliferate poorly, and very few cells produce IFN-γ and interleukin-2 (IL-2), compared with transgenic T cells injected earlier in the blood-stage infection. CD8⁻ dendritic cells at day 10 can process and present peptides on major histocompatibility complex (MHC) class II with an efficiency similar to that of dendritic cells from earlier in infection. The failure of the day 10 dendritic cells to activate MSP1-specific CD4 T cells fully in vitro is associated with reduced expression of CD86 and lower production of IL-12 rather than with induction of inhibitory DC receptors or production of IL-10.

INTRODUCTION

An immune response to the malaria parasite, Plasmodium, needs to be tightly regulated to enable pathogen clearance without inducing excessive pathology. There are many reports that CD4 T cell responses both contribute to and control immunopathology and are essential for protective immunity in acute blood-stage malaria in both experimental models and humans (reviewed in reference 34). The types of CD4 T cells induced are themselves controlled by several factors, such as antigen dose, type of antigen-presenting cell (APC), physiological state of the APC, other innate cells such as NK cells, and the general cytokine environment (2).

After infection or exposure to pathogen products, dendritic cells (DCs) differentiate and show an increased ability to capture an antigen or pathogen and upregulate the antigen-processing machinery and expression of major histocompatibility complex (MHC) molecules and costimulatory molecules ready for antigen presentation (reviewed in reference 36). Subsequently, when they have matured sufficiently to present peptides and activate T cells, they no longer take up new antigen. It is thought that this regulation is a normal process to ensure that the T cell response is both focused and appropriately limited (45, 46). However, for acute and chronic inflammatory infections such as malaria, it is still debated whether appropriate regulation of the DC response can take place or whether, or how far, inflammation and other factors occurring during the acute infection affect the normal functioning of DCs and T cell activation.

In the first few days of experimental blood-stage malaria infection in mice, DCs appear to be able to activate and allow differentiation of T cells into effector cells (3, 15, 26, 35). In addition, DC from naïve mice coincubated with erythrocytes infected with Plasmodium chabaudi or P. yoelii (iRBC) are stimulated to produce tumor necrosis factor (TNF), interleukin-6 (IL-6), and low levels of IL-12 (31) and can activate T cells (11). However, there is general agreement that at some point in an acute blood-stage malaria infection in mouse models, DCs are no longer able to activate either CD4 or CD8 T cells (15, 23, 35). It has been suggested that infection of mice with P. berghei may prematurely accelerate DC maturation such that uptake and processing of antigen cannot take place, resulting in impairment of cross-presentation to CD8 T cells or MHC class II-dependent activation of CD4 T cells (16, 17). Others, however, suggest that lack of T cell responsiveness is because of inhibition of DC maturation by the parasite, an observation found for human DCs and P. falciparum as well as in mouse models (23, 40). The inhibition of DC maturation has been attributed to several factors: binding of the iRBC to the DC via CD36 may initiate an inhibitory signal (41), or parasites or parasite products such as hemozoin (20) or uric acid (25) may affect antigen-processing within endosomal compartments for either cross presentation on MHC class I (29) or presentation on MHC class II (44). It has also been proposed that TNF (48) or Toll-like receptor (TLR) stimulation induces tolerance in the DCs (27). A change in the relative production of IL-10 and IL-12, low expression of costimulatory molecules, and a shortening of the interaction time of DC with T cells may also impair the T cell response (24, 27).

Most studies have been carried out on third-party model antigens such as ovalbumin (OVA) or, in one case, with a model antigen expressed in a transfected parasite (16, 17). It is important to know how far these alterations of DC function affect T cell responses to naturally processed Plasmodium antigens, as this would provide direct information about whether malaria immunity might be impaired by the parasite itself. In addition, only very few studies looked at both in vitro and in vivo responses (16, 21, 46), leaving open the possibility that the inhibition of some T cell responses in vivo is due not to the DC itself but to the environment in which the T cell/DC interaction is taking place (36). Similarly, the relevance of bone marrow (BM)-derived DCs, which are only a subpopulation of the heterogeneous DCs found in spleen or other organs, to in vivo T cell activation by other populations of DCs has not been demonstrated, and it is possible that other APCs in other organs can substitute for the DCs in this capacity.

Here we have used a combination of in vivo and in vitro approaches to investigate the changes in the ability of DCs to activate malaria-specific CD4 T cell responses during an acute P. chabaudi infection in mice. For these studies we have made use of a CD4 T cell receptor (TCR)-transgenic (Tg) mouse carrying a TCR specific for a previously described peptide within merozoite surface protein 1 (MSP1) (37). We show here that CD8⁻ DCs isolated on days 10 and 13 of infection are no longer able to activate T cells. Our data suggest that loss of CD4 T cell responsiveness at this time of infection is not due to accelerated loss of the capacity of the CD8⁻ DCs to phagocytose, process, and present this MSP1 peptide but rather is the result of a downregulation of costimulatory molecules and IL-12p70, which control Th1 T cell polarization, IL-2 production, and T cell proliferation. As the CD4 T cell response in vivo later recovers but with a different T cell cytokine-inducing potential, we argue that this loss of DC presentation is rather a mechanism of immune regulation to prevent an overwhelming inflammatory Th1 response and excessive immunopathology.

MATERIALS AND METHODS

Mice.

BALB/cJ mice were originally obtained from the Jackson Laboratory and have been maintained as a colony at the National Institute for Medical Research (NIMR) for the past 20 years. BALB/c (CD90.1) congenic mice (N15, BALB/c) were a kind gift of David Tough, (Jenner Institute, Compton, United Kingdom) and were further backcrossed four generations to BALB/c mice bred at NIMR for adoptive transfers, B5 T cell receptor-transgenic mice (37) were bred in the specific-pathogen-free (SPF) unit at the National Institute for Medical Research and were used at 6 to 12 weeks of age. Female mice were conventionally housed with sterile bedding, food, and water for experimental purposes and were used at 6 to 12 weeks of age.

Parasites.

Plasmodium chabaudi chabaudi clone AS was routinely injected from frozen stocks. Infections were then initiated by intraperitoneal (i.p.) injection of 10⁵ iRBC obtained from infected mice before the peak of parasitemia. Parasitemia was monitored by analysis of thin blood films as described previously (19).

Cell lines, peptides, and media.

CTLL-2 cells and the B5 CD4 T cell hybridoma (28) were cultured in Iscove's modified Dulbecco's medium (IMDM) (Sigma, Dorset, Poole, United Kingdom) supplemented with 10% fetal calf serum (FCS), 1 mM l-glutamine, 10 mM HEPES, 5 × 10⁻⁵ M 2-mercaptoethanol, 100 μg/ml penicillin, 100 U/ml streptomycin, and 1 mM sodium pyruvate (complete IMDM). CTLL-2 cells were cultured with human recombinant IL-2 (10 U/ml). The peptide ISVLKSRLLKRKKYI, recognized by the B5 transgenic CD4 T cells and CD4 T cell hybridoma (28), was synthesized by Jerini AG, Berlin, Germany.

Antibodies for cell sorting and flow cytometry.

Monoclonal antibodies (MAb) used for cell sorting and fluorescence-activated cell sorter (FACS) analysis were anti-NK (DX5), CD3, Ter119, CD19-biotin, anti-CD11c-allophycocyanin-Cy7, CD11c-allophycocyanin, CD8α-fluorescein isothiocyanate (FITC), class II H-2 A^d-FITC, class II H-2 A^d-allophycocyanin-Cy7, CD4-allophycocyanin-Cy7, CD4-Pacific Blue, PD1-phycoerythrin (PE), CD40-PE, CD86-PE, CD25-PE, CD25-allophycocyanin, CD44-allophycocyanin, B and T lymphocyte attenuator protein (BTLA)-PE, CTLA-4-biotin, CD62L-Pacific Blue, Ly6G-PE, CD273-biotin, CD274-PE, annexin V-Pacific Blue, IFN-γ-FITC, IFN-γ-PE-Cy7, IL-10-allophycocyanin, and IL-2-PE (BioLegend) as well as streptavidin-Alexa 610 (Invitrogen). Isotype control Ab (BioLegend) were included in each staining, as well as 7-aminoactinomycin D (7AAD) (Sigma). The Foxp3 kit for intracellular staining was used (eBioscience). Rat anti-mouse Fc receptor (2.4G2) (FcR block) (39) was purified from hybridoma supernatants.

Cell sorting.

Spleens were dissected from infected mice at different times postinfection and treated for 30 min at 37°C with 0.4 mg/ml of Liberase CI (Roche, Basel, Switzerland) in serum-free IMDM. After washing and centrifugation at 300 × g for 10 min, cell pellets were resuspended for 5 min in 8.3 g/liter NH₄Cl in 0.01 M Tris-HCl (pH 7.5) to lyse RBC. Spleen cells were washed and incubated with FcR block, followed by anti-CD11c magnetic beads and enrichment on magnetic columns (Miltenyi Biotec, Bergisch Gladbach, Germany). Enriched cells were stained with biotinylated DX5, CD19, Ter119, and CD3 antibodies and streptavidin-Alexa 610, CD11c-allophycocyanin-Cy7, CD8-FITC, Ly6G-PE, as well as 1 μg/ml 7AAD. Enriched cells were stained with the antibodies described above (1 μg/10⁶ cells for 20 min on ice), and live (7AAD negative), lineage-negative (including Ly6G-negative), CD11c^high CD8⁻ positive cells were sorted on an Aria II FACS (BD Biosciences) (see Fig. S3A in the supplemental material). The purity of the sorted cells was more than 95%. TCR-transgenic CD4⁺ T cells were isolated from spleens of uninfected B5 mice using anti-CD4 beads (Miltenyi Biotec). APCs were sorted into DCs, macrophages, and B cells after Liberase digestion and RBC lysis. The cells were positively selected with CD11c or CD11b magnetic beads (Miltenyi Biotec) and stained with CD11c, CD11b, and CD19 antibodies. DC sort gates were set on CD11c^high CD19⁻ cells. Macrophages were sorted as CD11c^low/− CD11b⁺ CD19⁻ cells. B cells were sorted as CD11c⁻ CD11b⁻ CD19⁺ cells.

CFSE labeling and TCR Tg CD4 T cell transfer.

A total of 2 × 10⁷/ml MACS-purified CD4 T cells were incubated with 1 μM carboxyfluorescein succinimidyl ester (CFSE) (Molecular Probes) in phosphate-buffered saline (PBS) for 5 min at room temperature (RT) and then washed three times in complete IMDM.

CD90.1 congenic BALB/c mice were injected intravenously (i.v.) with 1 × 10⁶ to 2 × 10⁶ purified CFSE-labeled B5 TCR Tg CD4⁺ T cells. Cells remained in vivo for 65 h before the mice were killed, and the spleen cells were prepared into a single-cell suspension, followed by RBC lysis, and then stained for FACS analysis as described below.

Antigen presentation assay using the B5 CD4 T cell hybridoma.

Different numbers of sorted CD8⁻ DCs were incubated with 2 × 10⁴ T cell hybridoma cells in 200 μl complete IMDM in round-bottom 96-well microtiter plates in a 7% CO₂ incubator at 37°C for 24 h. As positive controls, similar numbers of DCs were incubated with 1 μM peptide. The supernatants were then collected, frozen, thawed, and added to 5,000 CTLL-2 cells per well in complete IMDM in 96-well plates. After 24 h of incubation, the cells were pulsed with 1 μCi [³H]thymidine (Perkin-Elmer, United Kingdom) for 16 h and proliferation determined as previously described (28).

Flow cytometry.

Single-cell suspensions were resuspended in IMDM containing 12 mM HEPES and 5% (vol/vol) FCS. The splenocytes were centrifuged for 10 min at 4°C and 300 × g, and then the erythrocytes were lysed by 10 min of incubation with 0.16 M NH₄Cl₂-Tris buffer, pH 7.2. The cells were washed again in complete IMDM and resuspended in FACS buffer (0.5% [wt/vol] bovine serum albumin [BSA], 2 mM EDTA, and 0.05% sodium azide in PBS). A total of 5 × 10⁵ cells were preincubated at room temperature for 10 min with FcR block. All other antibody incubation steps were for 20 min on ice. Cultured cells were harvested, washed with FACS buffer, counted, and stained with antibodies as described above. Cells were stained with annexin V and 7AAD (1 μg/ml) after surface staining using the manufacturer's protocol and were acquired immediately on a BD LSR II flow cytometer (BD Biosciences) without fixing with paraformaldehyde. Intracellular cytokine staining was performed after incubating cells for 4 h at 37°C with 50 ng/ml phorbol myristate acetate (PMA) (Sigma) and 500 ng/ml ionomycin (Sigma); 5 μg/ml brefeldin A (BFA) (Sigma) was included for the last 2 h. Samples were then surface stained and fixed in 2% paraformaldehyde. The cells were permeabilized in permeabilization buffer (0.5% [wt/vol] saponin, PBS [pH 7.2], 0.5% BSA, 2 mM EDTA, 0.05% sodium azide) and stained for 40 min at 4°C with anticytokine antibodies before they were washed, acquired on a BD LSR II flow cytometer (BD Biosciences), and analyzed with FlowJo software (TriStar). Intracellular Foxp3 staining was performed after surface staining using a PE-anti-mouse Foxp3 staining kit (eBioscience).

Antigen-specific CD4 T cell proliferation.

CD4⁺ splenic T cells were isolated from spleens of B5 Tg mice using anti-CD4 beads (Miltenyi Biotech, Germany), treated with FcR block, and stained with anti-CD25PE, anti-CD44-allophycocyanin, and CD4-allophycocyanin-Cy7 as described above. CD4⁺ CD25⁻ CD44^low cells were sorted to greater than 98% purity on an Aria II FACS (BD Biosciences) (see Fig. S3B in the supplemental material). Sorted CD4 T cells were incubated with 1 μM CFSE (Molecular Probes) for 5 min at RT and washed three times before the T cells (10⁵) were cultured with 5 × 10⁴ sorted DCs in 200 μl in 96-well round-bottom plates. After 2, 3, 4, 5, and 6 days of culture, the cells were stained as described above with anti-CD4 after FcR block, acquired on a BD LSR II flow cytometer (BD Biosciences), and analyzed with FlowJo software (TriStar).

T cell cytokine production.

A total of 5 × 10⁴ CD8⁻ DCs purified by cell sorting from days 7, 10, and 13 infected mice were cultured in 200 μl in 96-well round-bottom plates for 6 days with 10⁵ CD4⁺ T cells purified from the spleens of Tg B5 mice as described above. The cells were removed from the wells, transferred to new plates, and cultured with 5 μg/ml cross-linked anti-CD3 and 2 μg/ml anti-CD28 antibody in 250 μl in flat-bottom 96-well plates as described previously (13) Supernatants were removed after 48 h of culture, and IFN-γ, IL-4, IL-2, and IL-10 were determined by enzyme-linked immunosorbent assay (ELISA) as described previously (31). Recombinant IFN-γ, IL-4, IL-2, and IL-10 standards were purchased from Invitrogen, United Kingdom.

Measurement of DC cytokines.

Dendritic cells (5 × 10⁴) isolated by cell sorting were cultured in 200 μl of complete IMDM in a 96-well plate (Costar) in the presence of 10 μg/ml CpG ODN 1668 (InvivoGen) for 24 h. Cytokine production was measured in the supernatant with a bead-based (multiplex) cytokine detection assay (eBioscience): IL-12p70 (BMS86004FF) (range, 27 to 20,000 pg/ml), IL-10 (BMS8614/2FF) (range, 13.3 to 20,000 pg/ml), IL-27 (BMS86024FF) (range, 68.6 to 50,000 pg/ml), IP-10 (BMS86018FF) (range, 41 to 30,000 pg/ml), IL-6 (BMS8603FF) (range, 27 to 20,000 pg/ml), IL-1β (BMS86002FF) (range, 69 to 50,000 pg/ml), and MIP-1α (BMS86013FF) (range, 27 to 20,000 pg/ml). The samples were processed using the manufacturer's protocol, and acquired on FACSCalibur, and analyzed using FlowCytomix Pro 2.4 software (eBioscience).

Statistical analysis.

Statistical analyses were performed using Student's t test (on transformed values) and the Mann-Whitney and Kruskal-Wallis tests.

RESULTS

Loss of antigen presentation capacity of splenic CD8⁻ dendritic cells at the peak of acute P. chabaudi parasitemia.

A blood-stage infection of BALB/c mice with Plasmodium chabaudi gives rise to an acute parasitemia with the peak occurring 8 days after injection of 10⁵ iRBC, which is resolved to low levels of less than 0.3% by 16 days postinfection (see Fig. S1A in the supplemental material).

Using CD4 T cells from a TCR-transgenic (Tg) mouse carrying T cells specific for a peptide of P. chabaudi MSP1 (MSP1 amino acids 1157 to 1171) (37), we have shown previously that splenic CD8⁺ and CD8⁻ CD11c⁺ dendritic cells have different abilities to present MSP1 peptides and activate Tg T cells during the acute infection (35). While CD8⁺ DC cells from uninfected mice were intrinsically able to process, present, and activate Tg T cells, we demonstrated that they lost this ability by day 7 in the infection and became apoptotic (35). By this time of infection, CD8⁻ CD11c^high DCs became the major DC population that presented the MSP1 peptide and activated T cells to proliferate, and they produced the cytokines IFN-γ, IL-2, IL-4, and IL-10 (35).

However, when we analyzed the presentation and activation capacities of the CD8⁻ DCs beyond the first 7 days of the infection, these DCs also lost their T cell activation capacity, similar to what was observed previously for CD8⁺ DCs, such that CD8⁻ DCs taken from infected mice at day 10 and day 13 induced less proliferation in vitro: only 33% (±3.5%) of Tg cells had divided when stimulated with CD8⁻ DCs from day 10 of infection, compared with 79.0% (±1.1%) of T cells stimulated with DCs from day 7 of infection. Furthermore, the majority of Tg T cells incubated with day 10 DCs had undergone fewer rounds of cell division (Fig. 1A), and less cell expansion was seen (Fig. 1B). Similarly, the cocultured Tg T cells produced significantly smaller amounts of the cytokines IFN-γ, IL-10, IL-2, and IL-4 when cultured with day 10 or day 13 CD8⁻ DCs than after culture with day 7 DCs (Fig. 1C).

Fig 1.

CD8⁻ DCs from day 7 infected mice induce more proliferation, expansion, and T cell cytokine production in malaria-specific TCR Tg T cells than DCs from day 10 and 13. (A) CD8⁻ DCs (5 × 10⁴) from uninfected (day 0) mice or mice infected with P. chabaudi for 7, 10, and 13 days, isolated as described in Materials and Methods, were cocultured with 10⁵ CFSE-labeled, purified Tg B5 CD4 T cells for 6 days. The percentages of divided or undivided cells are indicated. (B) The bar graphs show the percentage of B5 T cells that underwent more than two cycles of proliferation after coculture with day 7 or day 10 DCs (left) and the number of viable CD4 T cells recovered after the 6-day culture period (right). The histograms shown are the means and standard deviations (SDs) for replicate cultures of a representative experiment (of 3 performed). The P values were calculated using a Mann-Whitney or Kruskal-Wallis test. (C) CD8⁻ DCs (5 × 10⁴) from mice infected with P. chabaudi for 7, 10, and 13 days were cocultured with B5 Tg T cells for 6 days as described above and then further cultured at 10⁵ cells/well in plates with 5 μg/ml anti-CD3 and 2 μg/ml anti-CD28 as described in the Materials and Methods. IFN-γ, IL-4, and IL-10 in the supernatant were measured after 48 h; IL-2 was measured after 24 h. The histograms shown are the means and SDs for replicate cultures of a representative experiment (of 3 performed). The P values were calculated using a Kruskal-Wallis test or a Mann-Whitney test.

It was possible that at these later stages of infection, antigen presentation and T cell activation were taking place in other lymphoid organs or carried out by other splenic APCs. However, on day 10, there was only a minimal IL-2 response of the B5 CD4 T cell hybridoma when cultured in vitro with MHC class II⁺ cells from lymph nodes or blood (see Fig. S1B in the supplemental material), and no presentation of the MSP1 peptide by splenic macrophages or B cells was observed on day 7 or 10 p.i. (see Fig. S1C in the supplemental material). These data therefore suggest that significant presentation of the MSP1 peptide was not taking place in other organs at this time and that T cell activation by splenic DCs may be impaired or downregulated at the high antigen loads observed at peak parasitemia.

Transgenic CD4 T cell responses are impaired in vivo during the peak of infection.

Since other splenic APCs or APCs in other organs did not appear to present MSP1 peptides and CD8⁻ DCs were unable to activate Tg T cells at day 10, we investigated whether the Tg CD4 T cell response was also dampened in vivo during this time of infection. To do this, Tg T cells (CD90.2) were isolated, labeled with CFSE, and injected into infected congenic CD90.1 BALB/c mice at various times over the acute infection (Fig. 2A). Spleens were removed after 3 days and the Tg cells analyzed for cell division (Fig. 2B) and cytokine production (IFN-γ, IL-2, and IL-10) (Fig. 2C). Although not all the Tg cells divided or produced cytokines in this 3-day period (see Fig. S2B in the supplemental material), it can be clearly seen that there were reduced numbers of divided cells as well as fewer IL-2- and IFN-γ-producing cells after injection of Tg T cells at days 5 and 8 and recovery at days 8 and 11, respectively, compared with cells injected at day 2 (Fig. 2B). This reduced T cell response coincided with the period of maximum parasitemia. Greater numbers of divided cells were recovered again when cells were injected at days 12 and 17 and analyzed on days 15 and 20, coinciding with resolution of the acute parasitemia. The response of injected Tg T cells could not be measured after this time, as the amount of MSP1 peptide presented was no longer sufficient to activate Tg T cells either in vitro or in vivo (reference 35 and data not shown).

Fig 2.

Transgenic CD4 T cell responses are impaired in vivo during the peak of infection. (A) Course of a P. chabaudi infection in BALB/c mice (CD90.1) following i.p. injection of 10⁵ iRBC. B5 Tg CD4 T cells were injected into and retrieved from mice and analyzed by FACS analysis on the days indicated below the graph. (B and C) Total number of B5 Tg CD4⁺ T cells (CD90.2) recovered from spleen on each of the days indicated: divided cells (CFSE^low) (B) and IL-2-, IFN-γ-, and IL-10-producing B5 Tg CD4 T cells (C). The percentages of CD4, CD90.2, and CFSE^low cells were determined from the gating shown in Fig. S2 in the supplemental material. Intracellular cytokine staining and cell surface staining were carried out as described in Materials and Methods. Gates were set on isotype controls. Each experiment was performed with at least 3 mice per time point and repeated 3 times. The histograms represent the means and SDs. Significant differences were determined using the Kruskal-Wallis test.

In agreement with our previous observations, the numbers of IFN-γ-producing T cells were greatest early in the acute infection (32), and although more Tg T cells divided when injected at day 12, there was no concomitant recovery in the number of IFN-γ-producing cells. In contrast, IL-10-producing Tg CD4 T cells followed a different kinetic, whereby the greatest numbers were observed after the peak of parasitemia (Fig. 2B). Interestingly, a significant proportion of the IL-10⁺ Tg T cells (approximately half) also produced IFN-γ (see Fig. S2 in the supplemental material) suggesting that the responding CD4 T cells at this stage of infection may be able to contribute to the downregulation of acute stage pathology as we described recently (7).

In summary, the proliferative response of the transgenic CD4 T cells in vivo corroborates the in vitro experiments, in that CD4 T cells undergo little cell division during the peak of infection. After that period of nonresponsiveness, the pattern of cytokines produced by those responding T cells has changed.

CD8⁻ DCs from day 10 activate TCR Tg T cells but induce higher levels of inhibitory receptors and apoptosis.

The reduced T cell proliferation and cytokine production induced by CD8⁻ DCs at day 10 of infection was not due to lack of initial T cell activation. Expression of the low-affinity IL-2 receptor, CD25, was clearly upregulated on the B5 Tg T cells after culture with both day 7 and day 10 DCs. CD62L expression, was reduced to similar levels on the Tg T cells after culture with CD8⁻ DCs from both days, which is characteristic of T cell activation (Fig. 3).

Fig 3.

Splenic CD8⁻ DCs from day 10 of a P. chabaudi infection activate B5 Tg CD4 T cells in vitro but induce higher levels of CTLA-4 and BTLA expression. (A) Representative histograms (of 3 experiments) showing surface expression of CD25, CD62L, PD1, BTLA, and CTLA-4 on B5 Tg CD4 T cells after the indicated days of coculture of 10⁵ naïve CD25⁻ CD44^low CD4⁺ T cells with 5 × 10⁴ CD8⁻ DCs from day 7 (black lines) and day 10 (gray lines) of infection. Dotted lines show staining on the naïve TCR Tg CD4 T cells before culture. (B) Mean fluorescence intensity (MFI) of CTLA-4 and BTLA expression on the B5 Tg CD4 T cells described for panel A. The histograms represent the means and SDs of replicate values in a representative experiment of 3 performed. Significant differences were determined using the Mann-Whitney test.

The inhibitory receptors BTLA (43) and CTLA-4 (4) were upregulated on T cells stimulated with DCs isolated from both day 7 and day 10. However, there was a significantly greater degree of upregulation of these molecules on Tg T cells stimulated with day 10 DC than with day 7 DCs (Fig. 3). The responding Tg cells were not Foxp3⁺ regulatory T cells (see Fig. S4A in the supplemental material).

In addition to activation and induction of T cell cytokines, DCs can induce apoptosis and death in T cells not only as a result of lack of stimulation (33) but also by activation-induced cell death (8). We therefore asked whether more activation-induced cell death, determined by staining with 7AAD and annexin V, had occurred in the Tg T cells cultured with DCs at later time points. Indeed, there were more apoptotic Tg T cells after coculture with DCs from day 10 than after coculture with those from day 7 (Fig. 4). The apoptotic cells were found within both the undivided T cell population and those T cells that had undergone more limited cell division (see Fig. S4B in the supplemental material).

Fig 4.

CD8⁻ DCs from day 10 induce more apoptosis in B5 Tg CD4 T cells than DCs from day 7 of a P. chabaudi infection. (A) Two-parameter FACS plots showing apoptotic B5 T cells, as defined by annexin V⁺ and 7AAD ⁺, before culture (left panel) and after 6 days of culture with day 7 (middle panel) or day 10 (right panel) CD8⁻ DCs. Naïve B5 Tg T cells (1 × 10⁵) labeled with CFSE were cultured with 5 × 10⁴ CD8⁻ DCs from infected mice as described in Materials and Methods. (B) Bar graph showing the proportion of apoptotic cells recovered after 6 days of culture as described above. The graph show the means and SDs for replicate samples of one representative experiment of 2 performed. Significant differences were determined using the Kruskal-Wallis test.

In summary, CD8⁻ DCs from day 10 of infection induced a higher level of expression of inhibitory receptors and increased apoptosis in vitro, suggesting that induction of inhibitory receptors or activation-induced cell death may contribute to the reduced Tg cell response.

Inability to stimulate T cell proliferation by splenic CD8⁻ dendritic cells is associated with lower expression of the costimulatory molecule CD86 and reduced production of IL-12 but not lack of processing and presentation of the specific peptide.

Splenic dendritic cells from P. berghei-infected mice can lose their ability to phagocytose as the blood-stage infection progresses and therefore were no longer able to process and present antigen to T cells (46). However, CD8⁻ DCs at day 10 had clearly internalized parasites and/or parasite proteins in vivo and processed MSP1, as they were able to stimulate an IL-2 response from the MSP1-specific CD4 T cell hybridoma equivalent to that of day 7 DCs, indicating that similar levels of B5 peptide were presented on the surfaces of the two DC populations (Fig. 5A).

Fig 5.

Splenic CD8⁻ dendritic cells from day 10 of a P. chabaudi infection produce less IL-12p70 and have a reduced level of expression of CD86 compared with day 7 DCs but can process and present comparable amounts of MSP1 peptide to B5 Tg CD4 T cells. (A) IL-2 production by the MSP1-specific CD4 T cell hybridoma, B5, after in vitro culture with purified splenic CD8⁻ DCs from mice infected for 7 or 10 days with P. chabaudi in the absence (left panel) or presence (right panel) of 1 μM specific peptide. The amount of IL-2 in the culture supernatant was determined after 24 h using the CTLL-2 proliferation assay. The means and SDs for replicate samples of 1 representative experiment of 3 performed are shown. SDs of less than 10% of the mean are not shown. (B) Surface expression of MHC class II, CD86, CD40, and PDL1 on CD8⁻ DCs from uninfected BALB/c mice and from mice infected for 7 or 10 days with P. chabaudi. Top panels, gating strategy, showing lineage (CD3, CD19, DX5, Ter119, and Ly6G) and 7AAD (lin) versus forward scatter (FSC) FACS plot of spleen cells. Left, the gate was set on viable lineage-negative cells as well as side scatter (SSC) and FSC. Right, FACS plot shows CD11c and CD8 profiles of gated cells. Bottom panels, histograms of MFI of MHC class II (A^d), CD86, CD40, and PDL1 on CD11c^high CD8⁻ cells. Each experiment was performed with at least 3 mice per time point and repeated 3 times. Significant differences were determined using the Mann-Whitney test. (C) Production of cytokines by CD8⁻ DCs from day 7 and day 10 of a P. chabaudi infection after culture in vitro with 10 μg/ml CpG for 24 h. Cytokines in the supernatants of the cultures were measured with a bead-based cytokine detection assay and performed as described in the text. The values shown are the means and SDs of replicates. Significant differences were determined using Student's t test.

MHC class II, CD86, and CD40 were all upregulated on the surfaces of day 7 and day 10 DCs compared with DCs from uninfected mice. However, while MHC class II and CD40 expression was similar on day 7 and day 10 DCs, the mean fluorescence intensity (MFI) of CD86 was significantly lower on DCs isolated at day 10 (Fig. 5B). There was no difference in expression of PDL1, a ligand for PD1 on T cells (6), on DCs from day 7 or day 10 of infection, suggesting that differential regulation of this molecule was unlikely to explain the differences in stimulatory capacity of the DCs.

Day 10 DCs produced much smaller amounts of the inflammatory cytokines IP-10, IL-6, IL-1β, and MIP-1α after CpG stimulation (Fig. 5C) than day 7 DCs. There were no differences in the amounts of IL-27 secreted from DCs isolated ex vivo on day 7 and day 10, and significantly more IL-10 was produced by day 7 DCs than by day 10 DCs (Fig. 5C). Importantly, significantly less IL-12p70 was secreted from day 10 DCs than from day 7 DCs, which could have contributed to the lower T cell activity (18).

Together these data suggest that some impairment in costimulation and reduced production of the cytokine IL-12p70 rather than lack of uptake and processing of MSP1 has contributed to the inability of day 10 DCs to induce T cell proliferation and cytokine production.

DISCUSSION

It is a widely held view that blood-stage malaria induces suppression of the immune response and that dendritic cells are implicated in this process (reviewed in reference 15). Here we report that there is a period of time during and following the peak of an acute Plasmodium chabaudi infection when specific CD4 T cells do not proliferate or produce cytokines either in vivo or after in vitro coculture with splenic DCs. The reduced CD4 T cell responses were not due to defective antigen processing and presentation of peptide by the CD11c^high CD8⁻ splenic DCs at this stage of infection or to the induction of regulatory T cells. Initial activation of the Tg T cells did take place in vitro; however, the DCs appeared to be ineffective in inducing full T cell proliferation and production of cytokines. A combination of reduction in expression of the costimulatory molecule CD86 on DC, lower IL-12 production, and induction of inhibitory interactions via CTLA-4 and BTLA is the most likely explanation for this reduced or anergic CD4 T cell response.

Using the model antigen ovalbumin (OVA) expressed in P. berghei, it has been shown that splenic CD11c^high DCs isolated during a blood-stage infection are ineffective at activating OVA-specific CD4 or CD8 T cells. It was proposed that rapid maturation of DCs caused by the infection, resulting in an inability to phagocytose parasite material for processing and presentation, was the reason for the loss of T cell responses (16, 17, 46). Although we did not analyze directly the in vitro ability of day 10 DCs to phagocytose parasite material in this P. chabaudi infection, the DCs in our study had clearly taken up and processed MSP1 in vivo, as peptides were presented on MHC class II in sufficient amounts to induce a response from the MSP1-specific T cell hybridoma.

Low levels, or lack of expression, of costimulatory molecules on APCs can lead to inadequate stimulation of the TCR-associated CD28 complex on T cells and thus to incomplete or no T cell activation (30). This offers a possible explanation for the low T cell responses when day 10 DCs are used as APCs. High levels of coreceptor on APCs are required to provide sufficient avidity between the interaction between the TCR complex and the APC. T cells given an insufficient and short stimulation by low-avidity interactions will not proliferate or perform other effector functions (12). The expression of the costimulatory molecule CD86, although upregulated on the surface of day 10 DCs compared with DCs of uninfected mice (35), was lower than that on the day 7 DCs.

The MSP1-specific Tg T cells stimulated with day 10 DCs did undergo partial activation with the upregulation of CD25 and downregulation of CD62L, but this was accompanied by a more pronounced upregulation of the inhibitory molecule CTLA-4 in vitro. The amount of CTLA-4 transported to the T cell surface is important, given that minor differences in expression has been shown to have potentially major effects on T cell function and on the development of autoimmunity (30). The higher affinity of CTLA-4 for CD86 (reviewed in reference 5) and the lower levels of CD86 on day 10 DCs together may favor inhibitory CTLA-4/CD86 interactions rather than CD28/CD86 interactions, resulting in inhibition of ZAP 70 (10) and inhibition of IL-2 synthesis (42), giving rise to the smaller amounts of secreted IL-2 and proliferation of the Tg T cell observed after stimulation with day 10 DCs. Although CTLA-4 is required for the optimal function of Tregs (47), there was no significant induction of Foxp3 on the Tg cells either in vitro or in vivo.

One consequence of CTLA-4 engagement is induction of apoptosis (9). This was clearly more evident on the Tg T cells cultured with day 10 DCs but was more difficult to demonstrate in vivo, possibility because apoptotic cells are rapidly cleared by myeloid cells. B and T lymphocyte attenuator protein (BTLA) was also upregulated on Tg T cells cultured with day 10 DCs and was expressed at a higher level. BTLA has been shown to modulate cytokine responses of CD4⁺ T cells (14) and to deliver negative signals during innate and adaptive immune responses (22), and we have previously demonstrated that BTLA restricts the protective immune response during a P. yoelii infection of mice (1).

A dominant IFN-γ response characteristic of a Th1 response was observed in vivo only when cells were injected early (days 2 to 5) in the infection. In line with this, CD8⁻ DCs isolated from the spleen at day 10, unlike DCs from day 7 and from uninfected mice, could not produce IL-12p70 after CpG stimulation in vitro, suggesting that the low levels of IL-12p70, a cytokine necessary for the induction of IFN-γ in T cells and NK cells (38), led to poor induction of IFN-γ in the Tg CD4 T cells. The recovery of the T cell proliferative and IL-2 response, but not the IFN-γ response, of these cells suggests that the APCs later in the infection, while their T cell activation capacity and IL-2 production were restored, directed the T cells away from an IFN-γ response. Although the T cell response recovered in vivo, the phenotype of the responding cells had changed, with more IL-10-producing T cells, and interestingly, of those there was a significant proportion of IL-10/IFN-γ double-producing cells, which we have previously shown to be important in downregulating acute-stage pathology in P. chabaudi infections (7).

In conclusion, our data suggest that CD8⁻ DCs regulate a potentially damaging inflammatory Th1 response at the peak of a blood-stage P. chabaudi infection by the induction of transient T cell anergy with reduced expression of the costimulatory molecule CD86, lower IL-12 production, and induction of inhibitory interactions via CTLA-4 and BTLA. The transient nature of reduced T cell responses may explain why different laboratories report different effects of DC/T cell interactions, and we suggest that this downregulation of the response is a natural process in the control of exuberant T cell responses.