A role for B cells in the development of T cell helper function in a malaria infection in mice

Abstract

B cell knockout mice are unable to clear a primary erythrocytic infection of Plasmodium chabaudi chabaudi. However, the early acute infection is controlled to some extent, giving rise to a chronic relapsing parasitemia that can be reduced either by drug treatment or by adoptive transfer of B cells. Similar to mice rendered B-cell deficient by lifelong treatment with anti-μ antibodies, B cell knockout mice (μMT) retain a predominant CD4⁺ Th1-like response to malarial antigens throughout a primary infection. This contrasts with the response seen in control C57BL/6 mice in which the CD4⁺ T-cell response has switched to that characteristic of Th2 cells at the later stages of infection, manifesting efficient help for specific antibodies in vitro and interleukin 4 production. Both chloroquine and adoptive transfer of immune B cells reduced parasite load. However, the adoptive transfer of B cells resulted in a Th2 response in recipient μMT mice, as indicated by a relative increase in the precursor frequency of helper cells for antibody production. These data support the idea that B cells play a role in the regulation of CD4⁺ T subset responses.

The role of antigen-presenting cells (APC) in determining the outcome of a CD4⁺ T-cell response is still controversial. Depending on the experimental system, APC, cytokine environment, and specific major histocompatibility complex/peptide density on the surface of the presenting cell have all been shown to play a role (1–4). Infectious disease systems where strong CD4⁺ Th1 or Th2 responses are generated may provide good models in which to investigate important regulatory components. In an experimental malaria infection in mice, Plasmodium chabaudi chabaudi, the CD4⁺ T-cell response changes during the course of a primary erythrocytic stage infection. The early acute stage is typically accompanied by a Th1 response characterized by the production of interferon γ (IFN-γ). At later stages, after reduction of parasitemia, the response is predominantly of Th2 type with efficient in vitro help for malaria-specific antibodies and production of cytokines such as interleukin 4 (IL-4) and IL-10 (5–7).

The switch of CD4⁺ T cell from a Th1 to a Th2 response during infection suggests that the conditions under which CD4⁺ T cells undergo activation differ at different times after infection. Earlier studies in mice rendered B-cell deficient by treatment with anti-μ antibodies suggest that B cells might be important in the switch to Th2 cells in this system (8–10). In support of this, P. chabaudi-infected μ-suppressed mice given B cells had increased IL-4 and IL-10 responses (11). However, it is possible that the long-term treatment with anti-μ antibodies might have had adverse effects on APC that make it difficult to assess these effects in an unequivocal manner.

In the experiments described here, we show that the CD4⁺ T cell responses induced during a P. chabaudi infection in mice rendered B-cell deficient by targeted gene disruption of membrane IgM [μMT mice (12)] are of similar magnitude to those of control mice. However, there is a strong bias toward a Th1 response in contrast to the response of control mice. Reduction of parasitemia by adoptive transfer of B cells but not by treatment with antimalarial drugs allow a Th2 response to develop. These data suggest that B cells are not required for T cell priming in malaria infections but play an essential role in inducing a Th2 response and shutting down a Th1 response.

MATERIALS AND METHODS

Mice.

Female mice homozygous for a targeted mutation of the transmembrane exon of the IgM μ chain [μMT (ref. 12); a kind gift of Klaus Rajewsky and Daisuke Kitamura, Cologne, Germany] were backcrossed for 7 to 10 generations onto the C57BL/6 background. We bred the mice in isolators at the Max-Planck-Institut für Immunbiologie (Freiburg) and in the animal facilities of Imperial College (London). For some experiments, we obtained μMT and wild-type (WT) littermate controls by intercrossing animals heterozygous for the mutation. We tested the resultant offspring for the presence of IgM as described previously (12). In addition to WT littermates, we also used 6- to 10-week-old female C57BL/6 mice (Harlan) as control mice.

Parasites.

We infected mice with the erythrocytic stages of P. chabaudi chabaudi (AS) as described previously (5) by intraperitoneal injection of 10⁵ infected erythrocytes. We monitored the course of infection by examination of Giemsa-stained blood films throughout the experiment.

Drug Treatment to Eliminate P. chabaudi Infection.

We gave mice three doses of chloroquine orally (chloroquine diphosphate, 25 mg/kg in 100 μl water) at 48-hr intervals. Simultaneously, we gave the mice pyrimethamine (5 mg/kg as a suspension in sterile saline) intraperitoneally. We monitored the efficacy of the drug treatment by analysis of blood films.

Adoptive Transfer of B Cells.

We carried out the enrichment of splenic B cells by specific removal of T cells and macrophages from cell suspensions of spleens taken from C57BL/6 mice that had been infected with P. chabaudi 2 to 3 months previously (immune mice) as described (13). We incubated single-celled suspensions with a mixture of monoclonal antibodies specific for Thy-1, CD8, CD4, CD5, and Mac-1 (hybridomas J1j, YTS169, YTS191, C3PO, and Mac-1, respectively). We removed cells that had bound antibody (Ab) either by further incubation with magnetic beads coated with anti-Ig (Miltenyi Biotec, Bergisch Gladbach, Germany) and depletion on a magnet as described previously (13) or by lysis after the addition of complement (rabbit complement, Cedar Lane, Ontario). Cells treated in this manner contained less than 5% CD4 or CD8⁺ T cells and the population of Mac-1 cells was undetectable by flow cytometric analysis.

Flow Cytometric Analysis.

We performed two and three color staining using fluorescein (Fl)-, phycoerythrin (PE)-, and biotin-labeled antibodies. The second step reagents were streptavidin (str) red670 (GIBCO), Str-Fl (PharMingen), or Str-PE (PharMingen) appropriately. Purity of B cells for adoptive transfer and CD4⁺ T cells for limiting dilution and RT-PCR was evaluated using biotin-labeled anti-B220 (RA33A1.1; ref. 14) and Str-Red670, PE-anti-CD8, and FL anti-CD4 (PharMingen).

Limiting Dilution Analysis of T Cell Responses to Malarial Antigens.

CD4⁺ T cells were purified from spleens of infected mice by passage over a column of nylon wool and enrichment using the MACS system as described previously (8). In general, the purity of CD4⁺ T cells was greater than 80% as assessed by flow cytomety. Limiting dilution assays to measure the precursor frequencies of CD4⁺ T cells reponding to antigens of P. chabaudi have been described previously (5, 8, 15). The assays allow the simultaneous measurement of T cell proliferation, help for Ab production, and cytokine production. In the experiments described here, serial twofold dilutions (from 64,000 or 50,000 cells per culture) of CD4⁺ T cells were cocultured with immune T cell-depleted spleen cells (3 × 10⁴ per culture) as a source of APCs in 200 μl of Iscove’s medium containing 10% fetal calf serum, 100 units/ml of penicillin, 100 μg/ml of streptomycin, 1 mM of l-glutamine, 12 mM of Hepes, 5 × 10⁻⁵ M of 2-mercaptoethanol, and 0.5 mM of sodium pyruvate. A 0.1% suspension of P. chabaudi-infected erythrocytes was used as a source of antigen (Ag). We established control cultures using uninfected red cells as Ag. We measured malaria-specific Ig and cytokines by ELISA assays as described previously (5, 8, 15) We determined Ig in the culture supernatants after 7 days of culture and cytokines and measured proliferative responses after the cultures had been incubated for a further 2 days with irradiated normal Bl/6 spleen cells and Ag. We determined precursor frequencies from the zero order term of the Poisson distribution using regression analysis. We considered cultures to be positive when proliferation, Ab response, or cytokine production exceeded the background response (without T cells) by more than 3 SD. We compared precursor frequencies by Student’s t test on geometric means.

RESULTS

Reduction in Parasitemia in μMT Mice After Drug Treatment or Adoptive Transfer of B Cells.

As described previously (13), erythrocytic infection in female μMT mice is characterized by chronic relapsing parasitemia. This contrasted with the infection seen in WT or C57BL/6 mice, where parasites were cleared to subpatent levels after 30 to 40 days of infection. As no difference in the course of infection was seen between the WT and C57BL/6 mice, we used both interchangeably as controls. The primary infection is illustrated for three female μMT and WT mice in Fig. 1 A and B. We did not observe any mortality during this experiment. Treatment of μMT mice with chloroquine and pyrimethimine beginning at 20 days of infection reduced parasitemias to subpatent levels within 10 days of treatment (Fig. 1D). In agreement with earlier experiments, transfer of immune B cells to μMT mice also resulted in subpatent parasitemias within 10 to 12 days after transfer of the cells (Fig. 1C).

Figure 1.

Course of a primary P. chabaudi infection in female WT (A) and μMT mice (B) and in μMT receiving immune B cells (C) or chloroquine and pyrimethamine (D). Each line represents the infection in an individual mouse. The arrow indicates the time of adminstration of immune B cells (2 × 10⁷ given intravenously) or chloroquine and pyrimethamine (as described in text).

CD4⁺ T Cell Responses to Malarial Antigens in Infected μMT Mice.

To ascertain that a P. chabaudi infection could induce a CD4⁺ T cell response in μMT mice with similar magnitude to that described previously for the CD4⁺ T cells of C57BL/6 mice, we carried out limiting dilution assays to determine precursor frequencies of CD4⁺ T cells able to proliferate in response to malarial Ags (Fig. 2). After 8 days of infection, the frequencies were similar (two examples are shown in Fig. 2 Upper) and were in the ranges of those obtained previously for C57BL/6 mice at this stage of infection (5, 8, 15). Similarly, frequencies after 40 days of infection were not significantly different between the two groups of mice (Fig. 2 Lower, P > 0.1 for n = 6, Student’s t test), demonstrating that the absence of B cells had no effect on the overall magnitude of the CD4⁺ T cell response. As controls, limiting dilution assays included CD4⁺ T cells from uninfected mice as well as uninfected erythrocytes as Ag. The frequencies obtained from these cultures were less than 10 in 10⁶ cells in all experiments (data not shown).

Figure 2.

Precursor frequency analysis of CD4⁺ T cells from P. chabaudi-infected WT (□) and μMT (▪) mice able to proliferate in response to malarial Ags in vitro. The data are expressed as precursor frequencies, calculated as described in text. (Upper) At 8 days of infection. The bars show the responses in 2 experiments, each performed with a pool of cells from 2 mice. (Lower) At 30 days of infection. The bars represent the mean precursor frequencies obtained in 6 independent experiments. The standard errors of the mean are shown.

We have previously shown that during the first 2 weeks of infection of C57BL/6 mice with P. chabaudi, the predominant CD4⁺ T cell response is characterized by production of IFN-γ, whereas later in the infection the CD4⁺ T cell response exhibits Th2 functions such as efficient B cell help and IL-4 production (5, 8, 15). We therefore compared the CD4⁺ T cell response of μMT and WT mice to malarial Ags.

After 10 days of infection, both μMT and WT mice showed a predominant IFN-γ response to malarial Ags similar to previous findings with C57BL/6 mice (8, 15). The majority of responding microcultures in each case produced IFN-γ and only a minority were positive for malaria-specific Ab (T_helper activity) or IL-4 (data not shown). After 40 days of infection, we compared the CD4⁺ T cell response of μMT and control mice. Precursor frequencies between individual mice can be very variable (5, 8, 15). Nevertheless, the mean precursor frequency of T helper cells from WT mice was significantly greater than that of IFN-γ precursors (Fig. 3A, P < 0.05). In μMT mice, the reverse applied with a mean IFN-γ precursor frequency that was greater than that of either T helper cells or IL-4 precursors (Fig. 3B, P < 0.001 in both cases). Precursor frequencies of some individual animals are also shown in Fig. 4 a and c. Thus, CD4⁺ T cells of the μMT mice have retained a predominant IFN-γ or Th1-like phenotype at the later stages of infection.

Figure 3.

Comparison of precursor frequencies of CD4⁺ T cells producing IFN-γ (▪) or Il-4 (░⃞) or providing help for B cells (□) from WT (A) and μMT (B) mice obtained 40 days after infection with P. chabaudi. The histograms represent the mean precursor frequencies derived from the frequencies calculated from five, six, and four separate experiments for IFN-γ, Ab Help, and IL-4, respectively. The standard errors of the means are shown.

Figure 4.

Comparison of precursor frequencies of CD4⁺ T cells taken from individual WT and μMT mice after treatment with chloroquine or after adoptive transfer of immune B cells. CD4⁺ T cells were taken at 50 days of infection from WT and μMT mice (a and c, respectively), WT and μMT mice treated with Chloroquine at 30 days of infection (b and d, respectively), and μMT mice that received 2 × 10⁷ immune B cells i.v. at 30 days of infection (e). The percentages of B220⁺ cells in the spleens of the reconstituted mice were: (1) 6.75%; (2) 6.95%; (3) 6.36%, and (4) 1%. Precursor frequencies of IFN-γ (▪) and T helper cells for Ab production (□) are shown for each mouse.

Adoptive Transfer of B Cells but Not Treatment with Chloroquine Restores a Th2 Response in Infected μMT Mice.

As described, treatment of μMT mice with both chloroquine and adoptive transfer of immune B cells enabled the mice to clear their blood stage infection. We also examined whether the reduction in parasitemia allowed the switch of CD4⁺ T cells to a predominantly T helper cell response as seen in WT mice.

We performed limiting dilution assays with enriched CD4⁺ T cells taken from μMT and control mice 20 to 25 days after drug treatment or adoptive transfer (Fig. 4). Splenic CD4⁺ T cells of WT mice, as expected at this stage of infection, usually had higher precursor frequencies of Ab helper cells than IFN-γ producing cells (4 out of 5 mice in Fig. 4a). By contrast, both μMT mice and drug-treated μMT mice retained strong IFN-γ responses (all mice in Fig. 4 c and d had higher numbers of IFN-γ precursors than helper cells for malaria specific Ab production). As a control for the effects of drug treatment, we also gave WT mice chloroquine and pyrimethimine before we performed the limiting dilution assays. Similar to the untreated infected WT mice, these mice had a predominant T helper cell response (2 of 3 mice in Fig. 4b).

After adoptive transfer of B cells, limiting dilution analysis revealed that the CD4⁺ T cell response in the majority of reconstituted μMT had switched to a predominant Ab helper response with relatively lower frequencies of IFN-γ-producing cells (Fig. 4e) and higher frequencies of helper cells for malaria specific Ab. In 2 of 4 mice, this switch was very pronounced. In one mouse, IFN-γ was still the predominant response.

As described previously (13), although adoptive transfer of immune B cells into μMT mice resulted in the production of malaria-specific IgG antibodies detectable in the plasma of recipient μMT mice (data not shown), the level of B cell reconstitution of the spleen was low and rather variable (see Fig. 4 legend). In the single mouse where there was no switch to a T helper cell response, only 1% of spleen cells were B220⁺ 20 days after transfer of B cells.

DISCUSSION

Infection of mice with the blood stages of P. chabaudi chabaudi leads to the activation of subsets of CD4⁺ Th1 and Th2 cells (5–7). Here we show using μMT mice that B cells are important in the switch from Th1 cells producing IFN-γ to Ab-helper cells in this malaria infection.

Priming and activation of P. chabaudi-specific-CD4⁺ T cells occurred equally well during infection in the absence of B cells and Ab. There were no significant differences in the precursor frequencies of CD4⁺ T cells proliferating in response to malarial Ags from μMT or WT mice either early in the acute infection or during the later chronic infection. Earlier studies with B cell-depleted mice demonstrated a crucial role for B cells in activation of T cells (16, 17), while other studies showed that B cells were either unable to prime T cells or induced tolerance (18–21). Our data are in agreement with more recent findings using μMT mice showing that the absence of B cells has little or no influence on the overall magnitude of T cell responses (22–25).

μMT mice are unable to eliminate a P chabaudi infection, confirming the requirement for B cells in the clearance of blood stage parasites (8, 13). CD4⁺ cells taken from these mice at any time during the primary infection and stimulated with malarial Ag under limiting dilution conditions produced IFN-γ and were poor helper cells for Ab production, in contrast to CD4⁺ T cells from WT mice where late in infection the predominant response had switched to one characterized by a higher frequency of T helper cells for B cells and IL-4 production. These data agree with our earlier studies (8, 13) and those of others (10) where it was shown that CD4⁺ T cells from infected mice treated with anti-IgM antibodies to deplete B cells retained a strong IFN-γ response response throughout infection. In those experiments, it could have been argued that alteration of the CD4⁺ T cell reponse was a consequence of regularly injecting large amounts of anti-IgM Abs to deplete B cells. Confirmation of these data in μMT mice rules out Ig and/or Ig complexes as contributing factors in the sustained Th1 response observed. Despite the sustained Th1 response in μMT mice, they were unable to eliminate their parasites, thus lending weight to the view that the final effective mechanism of clearance of blood stage parasites is B cell- and Ab-mediated.

Because μMT mice could not clear their infection, there was a possibilty that parasite load rather than the absence of B cells was responsible for the late IFN-γ response. P. chabaudi infection induces an early IFN-γ burst (26), possibly from non-T cells. In addition, malaria toxins released at schizont rupture can activate macrophages directly to produce tumor necrosis factor-α and other cytokines (27, 28). It is possible that the continuous presence of parasite material might maintain high levels of IFN-γ and IL-12, thus sustaining a Th1 response. However, this is unlikely because transfer of B cells affected T helper cell development, whereas reduction of parasites and malarial toxins did not. Evidently, the increase in T helper cell responses required a signal dependent on the B cell. Similar findings have been obtained in infected anti-μ-treated mice with a restored B cell compartment (11). These data therefore give strong support for a role of B cells in the regulation of T cell help for Ab production in this infection and are in agreement with several recent pieces of evidence showing that B cells can skew CD4⁺ T cells toward a Th2 response in mice both in vivo and in vitro (22, 29–31).

In contrast, immunization studies in μMT mice using Ags such as KLH, HGG, and Schistosome eggs and adoptive transfer experiments into severe combined immunodeficient mice have shown that both IFN-γ and IL-4 are produced (23, 25, 32), suggesting that the absence of B cells has little or no influence on the production of Th2 cytokines. Discrepancies between the various studies may be because there are several routes by which cytokines necessary for T cell help can be induced (33, 34). Although cytokines, ligand density, co-stimulatory molecules, different signaling pathways, and APC have all been shown to influence CD4⁺ T cell differentiation (3, 4), a strong and possibly overriding influence appears to be the cytokine environment. The presence of IL-12 and IFN-γ (35, 36) during T cell activation will promote Th1 development whatever the APC. The interaction of CD40 with CD40 ligand (CD40L) on APC such as dendritic cells results in the up-regulation of IL-12 (37, 38), thus promoting a Th1 response. However, since B cells do not produce IL-12 (39), CD40/CD40L interaction might have different downstream effects. For Th2 cells, the major subset responsible for effective B cell help, IL-4 is a critical growth factor (1). In vivo studies by van Essen et al (31) and in vitro experiments of Stockinger et al (30) provide evidence that the interaction of T and B cells via CD40/CD40L interaction delivers signals to the T cells that enable them to become effective T helper cells for B cells inducing the expression of IL-4 (30). IL-4 can be also provided by cell types other than a Th2 cell itself such as NK1.1⁺ CD4⁺ T cells (40), mast cells, and eosinophils (41–45). It is possible that the signal for IL-4 release from these cells is provided by Ab (alone or as Ab/Ag complexes) from the transferred B cells binding FcR- or -C′ receptors. In this regard, we are currently investigating whether Ig from P. chabaudi-immune mice can replace B cells in switching the Th1 response in infected μMT mice. Induction of IL-4 in all these cells can also be achieved by different means. Cytokines such as IL-6 (46) will promote IL-4 production and distinct costimulatory signals may be also important. B cells themselves can deliver a signal to T cells to produce IL-4 and thus promote their own differentiation and expansion (31).

Effective clearance of erythrocytic stages of P. chabaudi is dependent on the presence or B cells and/or Ab (13). If similar Ab-mediated mechanisms of immunity are important for human infections, it may be of importance for vaccine development to determine the relative roles of anatomical location, Ags, and signals necessary to induce Th2 response under different circumstances.

ABBREVIATIONS

APC

antigen-presenting cells

IFN-γ

interferon-γ

IL

interleukin

WT

wild type

Ag

antigen

Ab

antibody