|Home | About | Journals | Submit | Contact Us | Français|
Neutralization of macrophage migration inhibitory factor (MIF) increases anti-tumor cytotoxic T cell responses in vivo and IFN-γ responses in vitro, suggesting a plausible regulatory role for MIF in T cell activation. Considering that IFN-γ production by CD4+ T cells is pivotal to resolve murine malaria and that secretion of MIF is induced by Plasmodium chabaudi adami parasites, we investigated the effect of MIF deficiency on the infection with this pathogen. Infections with P.c. adami DK parasites were more efficiently controlled in MIF-neutralized and MIF-deficient (KO) BALB/c mice. The reduction in parasitemia was associated with reduced production of IL-4 by non-T/non-B cells throughout patent infection. At day 4 post-infection, higher numbers of activated CD4+ cells were measured in MIF KO mice, which secreted more IFN-γ, less IL-4 and less IL-10 than CD4+ T cells from WT mice. Enhanced IFN-γ and decreased IL-4 responses also were measured in MIF KO CD4+ T cells stimulated with or without IL-12 and anti-IL-4 blocking antibody to induce Th1 polarization. However, MIF KO CD4+ T cells efficiently acquired a Th2 phenotype when stimulated in the presence of IL-4 and anti-IL-12 antibody, indicating normal responsiveness to IL-4/STAT6 signaling. These results suggest that by promoting IL-4 responses in cells other than T/B cells during early P.c. adami infection, MIF decreases IFN-γ secretion in CD4+ T cells and in addition, has the intrinsic ability to render CD4+ T cells less capable of acquiring a robust Th1 phenotype when stimulated in the presence of IL-12.
Macrophage migration inhibitory factor (MIF) is a pleiotropic cytokine secreted by several cell types including activated T cells and macrophages (1–3) that plays a central role in inflammation, delayed type hypersensitivity (DTH) reactions (4, 5) and angiogenesis (6). MIF counter-regulates the anti-inflammatory effects of glucocorticoids (7), enhances phagocytosis and H2O2 production in macrophages, and synergizes with IFN-γ to up-regulate NO production (8). MIF-deficient macrophages show impaired pro-inflammatory responses, which has been attributed to lower expression of TLR-4, diminished NF-κB activation (9), and enhanced activation-induced apoptosis (10). MIF is secreted from pre-formed intracellular pools, and its expression is further induced by various stimuli including oxidative stress, cytokines and infection (11).
MIF was initially described as a soluble mediator secreted by activated T cells that inhibits the migration of macrophages and thus contributes to DTH reactions (4, 5). Although expressed constitutively by resting Th1 and Th2 cells, the Con A-stimulated release of MIF was greatest in Th2 clones (1). The immune neutralization of MIF markedly inhibits antigen-specific proliferation of splenic T cells as well as antibody responses (12), suggesting an involvement for MIF in the modulation of T cell immunity and T cell-dependent antibody responses. MIF is associated with Th2 inflammatory pathologies such as asthma and experimental allergic reactions (13, 14) and it has been implicated in the control of helminthic infections that rely on Th2 effector immunity (15). The in vivo neutralization of MIF enhances CTL responses in mice bearing OVA-transfected EG.7 tumours and further increases specific IFN-γ and CTL responses in vitro (12).
MIF has been recently involved in the pathology associated to malaria (16), a parasitic disease characterized by systemic inflammation, oxidative stress and the pathologic complications of cerebral disease and anaemia (17, 18). Martiney et al. measured enhanced release of MIF in mice with Plasmodium chabaudi chabaudi AS infection and reported inhibitory effects of this cytokine on the differentiation of erythroid progenitors (19). Although high plasma levels of MIF correlate with the severity of cerebral and placental malaria, its precise role in the pathologic progression of human malaria remains unknown (20–24).
The malarial pigment hemozoin, which is an insoluble heme polymer produced by parasite catabolism, of host haemoglobin, contributes to the inflammatory response by its ability to activate TLR9 (25) and the NALP3 inflammasome (Griffith JW, Sun T, McIntosh MT, Bucala R. 2009. Pure Hemozoin Is Inflammatory In Vivo and Activates the NALP3 Inflammasome via Release of Uric Acid. J Immunol. 183:5208-5220.) In vitro treatment of monocytes and macrophages with Plasmodium-infected RBC and with the malarial pigment hemozoin induces the robust secretion of MIF (19), and in our laboratory, in vivo administration of synthetic hemozoin significantly increased MIF levels in the serum of naïve BALB/c mice (unpublished results). Thus, through its ability to induce the release of MIF, hemozoin may contribute to the severity of malarial anaemia (26). Accordingly, milder anaemia and enhanced survival to lethal P. c. chabaudi AS infection were measured in MIF-deficient (MIF KO), BALB/c mice when compared to wild type susceptible mice (26). In this study, MIF deficiency did not alter IFN-γ nor TNF-α responses in P.c. chabaudi-infected mice but substantially improved bone marrow erythropoiesis and increased haemoglobin levels (26).
Blood stage murine malaria is characterized by the induction of Th1 and Th2 cell responses (27, 28) and in this context, the self-resolving P.c. adami DK model in BALB/c mice is particularly useful to study cell-mediated parasite killing since early IFN-γ production by CD4+ T cells is pivotal for the control of primary parasitemia (29). Indeed, administration of IFN-γ delays the onset of patent infection (30) and vaccines inducing IFN-γ responses significantly reduce peak and cumulative parasitemia(31). In addition, it has been proposed that the Th2 response that is progressively induced at peak infection is responsible for complete resolution of blood parasite burden (27), but the mechanism involved in Th1 to Th2 switch remains unknown.
Considering the regulatory effects of MIF on the CTL and IFN-γ responses in vivo and in vitro, we investigated the effects of MIF deficiency on the kinetics of infection with non-lethal P.c. adami DK parasites. Our data suggests a role for MIF in the down-regulation of IFN-γ and up-regulation of IL-4 responses at early infection, indicating that this versatile cytokine modulates adaptive responses in CD4+ T cells.
MIF KO mice were developed as described (32) and backcrossed for more than 10 generations to a BALB/c genetic background. Four to six weeks old female wild type (WT) (Charles River, Canada) and female MIF KO BALB/c mice (bred at the animal facility of the University of Quebec in Montreal) were infected with 105 P. c. adami DK (556 KA) parasitized RBC by the i.p. route. Parasitemia was measured daily in methanol fixed tail blood smears stained with a 10% Giemsa solution in PBS. The study was conducted in compliance with the regulations from the Animal Committee of the University of Quebec in Montreal.
Neutralizing anti-MIF monoclonal Ab (clone IIID9, IgG1 isotype) was prepared as previously described (12). An isotype control Ab (IgG1) was purified from culture supernatants of the 5D4-11 hybridoma, which secretes an irrelevant Ab specific for type 3 dengue virus (ATCC, USA). To study the contribution for secreted MIF in vivo, WT BALB/c mice received a first injection of anti-MIF mAb or isotype control IgG (0.5 mg i.p., 100 µl volume) 3 days prior to infection, and received further injections on the day of infection and at 3 days intervals until peak infection or until complete resolution of parasitemia.
Single-cell suspensions of splenocytes from naive and P. c. adami DK infected mice (days 4, 8 and 12 post-infection) were prepared, RBC were lysed in Red Blood cell Lysing Buffer Hybri-max (Sigma Aldrich, USA) and spleen WBC were cultured in complete media (RPMI-1640 supplemented with 10% FBS, 1% HEPES, penicillin and streptomycin, Invitrogen, USA) at a concentration of 4 × 106 cells/ml in 24 well plates (Saarstedt, Canada). In experiments assessing the contribution of CD90− lymphocytes in the cytokine responses, CD90+ cells were removed using the EasySep® Mouse CD90+ positive selection kit (StemCell Technologies, Canada). CD4+ T cells were enriched by negative selection using the EasySep® Mouse CD4+ T cell enrichment kit (StemCell Technologies, Canada). Purity levels were ≥95% as assessed by flow cytometry.
Splenic cells were recovered from WT and MIF KO mice at day 4 of infection and were labeled with anti-CD4- FITC and anti-CD69 PERCP monoclonal Ab (Biolegend®, USA) for estimation of absolute numbers of activated CD4+ T cells in the spleen. Concentrations of IFN-γ, TNF-α, IL-4 and IL-10 were measured in 48 h cultures of total splenocytes, CD4+ T cells and CD90− lymphocytes from naive and infected mice without further stimulation using specific MAXTM Set (Deluxe) ELISA kits (Biolegend®, USA). IFN-γ and IL-4 also were assessed in 48 h culture supernatants from CD4+ T cells (4 × 106 cells/ml) from WT and MIF KO mice, stimulated with bound anti-CD3 (1.5 µg/ml; Cedarlane Laboratories, Canada) and soluble anti-CD28 Ab (1.5 µg/ml; eBioscience, Canada).
Total RNA was isolated from WT and MIF/KO CD4+ T cells from naive and P. c. adami (DK) infected mice with TRIzol reagent following the manufacturer’s instructions. The cDNA was synthesized with 10 µl of total RNA and the extension was made with 10 mM of dNTP SuperScript II (Gibco BRL, Rockville, MD). PCR amplification of cDNA was made in a final volume of 50 µl with 0.5 µg of template DNA, 1 unit of Taq polymerase (Genscript, New Jersey, USA) and 20 µM of specific primers. The sequences of the specific primers used in this study were: T-bet sense 5’TGCCTGCAGTGCTTCTAACA 3’, and antisense 5’TGCCCCGCTTCCTCTCCAACCAA 3’, GATA-3 sense 5’ GGTTGAAGGAGCTGCTCTTG 3’ and antisense 5’CTGGAGGAGGAACGCTAATG 3’; GAPDH F: 5’ GGAGATTGTTGCCATCAACGA-3’; R: 5’-TGGGAGTTGCTGTTGAAGTCG-3’. Each primer was designed in distinct exons to ensure specific transcript amplifications. Amplifications were carried out for 40 cycles according to incubation of 1 min at 94°C, 30 sec at 58°C and 1 min at 72°C. Amplification products were resolved in 2% agarose gel with ethidium bromide revelation. Relative expression of T-bet and GATA3 was determined by densitometry.
CD4+ T cells from naïve WT and MIF KO mice were labeled with 10 µM CFSE (Molecular Probes, Invitrogen Detection Technologies, USA) for 8 min, after which an equal volume of cold FCS was added to block the reaction. Following 15 min of incubation on ice, the cells were washed exhaustively in cold RPMI and were immediately stimulated.
CFSE-labeled CD4+ T cells or unlabelled CD4+ T cells from WT and MIF KO naive mice were plated at a concentration of 4 × 106 cells (500 µl per well) in 48 well plates (Sarstedt, Canada) previously coated with anti-CD3 monoclonal Ab. An anti-CD28 Ab then was immediately added. The mitotic events of CFSE-labeled CD4+ T cells were analyzed 48 h later using a FACSscan and the ModFit LT Software. In experiments using unlabelled CD4+ T cells, supernatants from cell cultures were recovered 48 h later and replaced with complete media (500 µl). Twenty five microliters of the combined tetrazolium compound (MTS) and the electron coupling reagent, phenazine methosulfate (PMS) (Cell Titer 96 aqueous non-radioactive cell proliferation assay, Promega, USA) then were added and absorbance at 490 nm corresponding to the dehydrogenase enzyme activity found in metabolically active cells, was measured 4 h later in a BioRad Model 550 microplate reader.
For SDS-PAGE, 15% polyacrylamide gels were used. For the analysis of MIF in sera, 10 µl of non-diluted sera were deposited into individual wells. Recombinant MIF (R&D systems) was used as positive control. For the detection of MIF in Western blots, a 1:200 dilution of a rabbit anti-mouse MIF polyclonal Ab (Biosource) was incubated over night, after which the membranes were further incubated with a 1:3000 dilution of a bovine biotinylated anti Rabbit IgG-Ab. Following incubation with Streptavidin-peroxidase (1:2500 dilution), a chemiluminiscent substrate (Pierce) was used to develop the reaction.
T-bet and Gata3 protein expression was followed by intracellular staining and flow cytometry using the FIX&PERM® kit (Caltag) kit. One million resting CD4+ T cells from naive MIF KO and WT mice or stimulated for 48 h with plate bound anti-CD3 Ab and soluble anti-CD28 Ab, were washed twice in PBS, the pellets were resuspended in 100 µl of Reagent A (fixation medium) and incubated for 15 min at room temperature. Five ml of PBS were added, and the cells were centrifuged for 5 min at 300g. Following removal of the supernatant, 100 µl of Reagent B (permeabilization medium) and 1 µg of Alexa Fluor®488 Mouse anti-T bet (BD Pharmingen) or Gata3 (D13C9) XPTM Rabbit Mab (Cell signalling Technology®) were added. The cells were incubated for 30 min at room temperature, washed with PBS and resuspended in 500 µl of PBS for immediate analysis in a FACScan (Becton Dickinson, USA). In samples treated with Gata 3 antibody, an additional labelling step using anti-rabbit IgG (H+L), F(ab’)2 Fragment (Alexa Fluor® 647 Conjugate) (Cell signalling Technology®) preceded the analysis.
In four independent experiments, CD4+ T cells enriched by negative selection were plated at 0.5 × 106 cells/ml (1ml per well) in 24 well plates pre-coated overnight with anti-CD3 Ab (1.5 µg/ml; clone C363.29B, Cedarlane Laboratories, Canada), and soluble anti-CD28 Ab (1.5 µg/ml; eBioscience, Canada) was added as co-stimulation. For Th1 differentiation, CD4+ T cells were stimulated in the presence of recombinant IL-12 (Biolegend®, USA) at 30 ng/ml and anti-IL-4 Ab (10 µg/ml; Biolegend®, USA). For Th2 differentiation, the cells were stimulated in the presence of recombinant IL-4 (Biolegend) at 40 ng/ml, anti-IL-12 (Biolegend®, USA) and anti-IFN-γ (Biolegend®, USA) Ab (10 µg/ml). Following three days of stimulation, the cells were split 1:4, 250 µl of fresh medium supplemented with IL-2 (10 ng/ml; Biolegend®, USA) was added, and cells were cultured for 4 additional days. The cells were recovered, washed once in RPMI-1640 and further incubated without or with plate-bound anti-CD3 Ab (1.5 µg/ml) for 24 h, after which IFN-γ, IL-4 and IL-13 levels were assessed in culture supernatants by ELISA. The effect of anti-MIF neutralizing antibody (100 µg/ml) was assessed in parallel in WT CD4+ T cells cultured under Th1 of Th2 polarizing conditions.
Statistical analysis was performed using a non-parametric Student Test for the comparison of parasite burden and cytokine responses between WT and MIF KO cells or mice using the Prism Software.
As mentioned previously, the P.c. adami DK malaria model in BALB/c mice is suitable to study the interactions between macrophages and T cells involved in parasite elimination and their modulation by MIF, which is readily induced during infection (Fig. 1A). The progression of parasitemia was monitored in wild type and MIF KO mice (Fig. 1B, C). The infection of MIF KO mice was characterized by significant reduction in peak parasitemia (Fig. 1C), and cumulative parasitemia (Fig. 1D), and a drop in blood parasite burden was evident at day 4 post-infection (Fig. 1E).
Considering the enhanced control in patent infection measured in MIF KO mice, we also assessed the effect of MIF neutralization with Ab on the progression of infection (Fig. 2A). As in the case of infected MIF KO mice, peak parasitemia was significantly lower in MIF-neutralized mice (Fig 2B), and the cumulative parasitemia also was significantly reduced in these mice (Fig. 2C). As for MIF KO mice, these effects were significant at day 4 post-infection (Fig. 2D).
As cumulative parasitemia was significantly reduced in MIF KO mice at early in infection (day 4) in two independent experiments, we evaluated IFN-γ, TNF-α, and IL-10 responses at day 4 post-infection in two additional experiments with MIF KO and WT mice. Total splenic cells were harvested and CD4+ T cells were further purified by negative selection. To estimate the contribution of non-T cells in these responses, CD90− cells were enriched in parallel by magnetic removal of CD90+ T cells. Although comparable levels of IFN-γ were measured in culture supernatants from WT and MIF KO splenic cells, IFN-γ levels were significantly increased in CD4+ T cell cultures from MIF KO mice (Fig. 3A; p<0.05). The IFN-γ content of culture supernatants from CD90− cells was relatively low and comparable in MIF KO and WT mice, suggesting that CD4+ T cells are the major source for IFN-γ at this time point of infection. In contrast to IFN-γ responses, TNF-α concentrations were relatively similar in total spleen, CD4+ and CD90− cell cultures from WT and MIF KO mice (Fig. 3B) but the levels of IL-10 were significantly lower both in total splenic cells and purified CD4+ T cell cultures from MIF KO mice (Fig. 3C).
For each of these two experiments the infection was allowed to progress in a remaining group of mice until day 8 (corresponding to peak parasitemia) and at this time point, IFN-γ levels were comparable in total splenic cells, CD4+ T cells and CD90− cells from WT and MIF KO mice (Fig. 3D), whereas TNF-α levels were significantly higher in culture supernatants from MIF KO CD4+ T cells (Fig. 3E). IL-10 responses remained relatively low in MIF KO mice at peak infection, but the contribution for CD4 + T cells and CD90− cells from WT or MIF KO mice on the IL-10 response was comparable when these cells were cultured separately (Fig. 3F).
In contrast to the enhanced IFN-γ response measured in CD4+ T cells from MIF KO mice at early infection, IL-4 levels were significantly low both in culture supernatants from total splenic cells and purified CD4+ T cells (Fig. 4A), and interestingly, cells others than CD4+ T cells (Fig. 4A), CD90+ cells or B cells (data not shown) were the major source for IL-4. Interleukin-4 responses remained relatively low in MIF KO mice at day 8 of infection, and associated primarily with CD4− cells (Fig. 4B). Four days following resolution of patent infection, IL-4 was still detected in culture supernatants from total splenic cells from WT and MIF KO and this response was persistently low in MIF KO cells (Fig. 4C); at this time point of infection IL-4 was not detected in culture supernatants from purified CD4+ T cells. Compared to WT mice, increased percentages and absolute numbers of splenic CD4+ T cells expressing CD69 were measured in MIF KO mice at day 4 post-infection (Fig. 5 A, B, respectively) and interestingly, although higher numbers of splenic CD4+ T cells were measured in naive MIF KO mice compared to naive WT mice, these differences were not noticed at day 4 post-infection (Fig. 5C). An additional interesting difference was observed with respect to T-bet/GAPDH gene expression (Fig. 6A) and T-bet/Gata3 mRNA ratios were relatively high in naive MIF KO CD4+ T cells, although these differences disappeared in following 4 days of infection (Fig. 6 B, C).
Taken together, our experiments revealed higher numbers of CD4+ T cells in uninfected MIF KO mice, which also expressed higher ratios of T-bet/GAPDH and T-bet/Gata3 (Fig. 5, ,6).6). We therefore evaluated whether MIF deficiency altered the capacity of CD4+ T cells to respond to TCR stimulation. For this analysis, CD4+ T cells from naive WT and MIF KO mice were purified by negative selection and were stimulated with anti-CD3 and anti-CD28 Ab for 48 h. The proliferation of CD4+ T cells was followed in cells labelled with CFSE prior to stimulation for 48 h or by estimating the metabolic activity of non-labelled cells with MTS reagent for the last 4 h of stimulation. As shown in Fig. 7A–C, higher stimulation indexes were measured in MIF KO CD4+ T cells when compared to WT cells in three independent experiments, and comparable effects were measured with the MTS assay (Fig. 7 D).
The enhanced proliferation measured in MIF KO CD4+ T cells in response to in vitro stimulation was associated with significantly higher secretion of IFN-γ (Fig. 8A), suggesting that in absence of MIF CD4+ T cells proliferate more efficiently and secrete more IFN-γ. Importantly, under these conditions, CD4+ MIF KO-stimulated cells secreted lower amounts of IL-4 (Fig. 8 B). Analysis of T-bet protein expression by flow cytometry confirmed its relatively higher expression in MIF KO naïve CD4+ T cells when compared to WT cells (Fig. 8 C), whereas Gata3 levels remained comparable in resting cells (Fig 6E). As for infected mice, no significant differences were revealed for T-bet and Gata3 levels in WT and MIF KO CD4+ T cells 48 h after stimulation.
Considering the striking differences measured in MIF KO and WT CD4+ T cells after short-term stimulation, we also assessed the capacity of these cells to differentiate into Th1 or Th2 cells in vitro. When cultured in the presence of recombinant IL-12 and anti-IL-4 antibody, MIF KO CD4+ T cells acquired a more robust Th1 phenotype since relative to WT cells, higher IFN-γ levels (Fig. 9A) and lower IL-4 and IL-13 levels (Fig. 9B,C) were measured in 24 h culture supernatants from MIF KO CD4+ T cells when stimulated with anti-CD3 Ab. However, MIF KO CD4+ T cells responded to Th2 differentiation in a manner comparable to WT cells, secreting significantly less IFN-γ Fig. 9A), more IL-4 (Fig. 9B) and IL-13 (Fig. 9C) than CD4+ T cells cultured under Th1 polarizing conditions. Neutralization of MIF with a specific Ab enhanced the secretion of IFN-γ Fig. 10A) and reduced the secretion of IL-4 (Fig. 10B) in WT CD4+ T cells cultured under Th1 polarizing conditions, indirectly corroborating that MIF attenuates the Th1 phenotype in CD4+ T cells.
The progression of blood stage P.c. adami DK infection and cytokine responses were studied in MIF deficient mice in the BALB/c background. Secretion of MIF was corroborated in P.c.adami DK-infected mice and its neutralization or deficiency decreased peak and cumulative parasitemia. As early IFN-γ responses by CD4+ T cells are enhanced in these MIF deficient mice, we propose that the lower parasite burden measured in these mice is associated with a higher level of IFN-γ. In parallel, a reduction in the expression of IL-10 was measured at early and peak infection in MIF deficient mice; these effects may be concurrent with milder inflammation considering the pro-inflammatory effects associated to MIF. Relatively lower IL-4 responses were measured in MIF KO mice through-out patent infection and interestingly, cells other than T cells or B cells were major sources for IL-4. In this context, Helmby et al. (33) described a population of non-B non-T cells, suggested to be mast cells or basophils, as the major source for IL-4 during P. chabaudi AS infection (33) and our results indicate that non-B, non-T cells are also the major source for IL-4 at peak infection with heterologous P.c. adami DK parasites. Taken together, our data indirectly suggests that MIF is required for the induction of IL-4 responses in mast cells and basophils as well as in CD4+ T cells at early malaria infection. The potential regulatory role for MIF on IL-4 responses in granulocytes and on Th2 sketching in malaria is interesting as these cells, centrally involved in the acute inflammatory reactions to allergens, are now being considered general inducers of Th2 responses (34). Increased IFN-γ responses also have been reported in MIF KO BALB/c mice infected with the helminth Taenia crassiceps (35) and although in this study IL-4 responses remained comparable in MIF KO and WT-infected mice, MIF deficient mice were more susceptible to infection. The contrasting effects of MIF deficiency in lethal P.c. chabaudi AS infection, for which no effects on parasitemia but increased survival were measured in MIF KO mice (26) and non-lethal P.c. adami DK infection indicates that factors additional to Th1 versus Th2 tuning are involved in parasite virulence. In this context, relative to non-lethal P.c. adami DK infections, the release of MIF is more robust in mice infected with the aggressive P.c. adami DS strain (unpublished data). As for MIF KO mice infected with non-virulent DK parasites, milder anaemia and decreased cumulative parasitemia were measured in MIF KO mice infected with lethal DS parasites and the death of these mice was only delayed for 24 h (data not shown). It is plausible that the stimulatory effects associated to MIF on IL-4 responses may rely on the Th2-prone nature of BALB/c mice. Indeed, decreased levels of Th2 cytokines and minor pulmonary inflammation were measured in OVA-primed MIF deficient mice of the BALB/c background compared to WT mice (14). However, the lethality of P.c. adami DS in respect to DK infection was shown to associate to a relatively weaker Th1 response and a stronger Th2 response in the Th1-prone NIH mouse strain, suggesting that the DS strain favours the induction of IL-4 responses (36).
Although the outcome of Th cell immunity is controlled by cytokines, with IL-12 promoting Th1 development through signal transducer and activator of transcription (STAT)4 and IL-4 favouring Th2 differentiation through STAT6, it is still unclear whether these messengers instruct cells to adopt specific Th phenotypes or whether they act as growth signals to sustain pre-determined Th fates. Ectopic expression of T-bet, the master transcriptional factor for Th1 development seems sufficient in inducing IFN-γ signalling, and enhancement of IFN-γ by IL-12 seems only a secondary event (37). In our study, higher basal ratios of T-bet/GAPDH and T-bet/Gata3 mRNA ratios were measured in naive MIF KO CD4+ T cells but these differences were attenuated during infection and a comparable phenotype was measured for CD4+ T cells stimulated in vitro. The fact that MIF KO CD4+ T cells develop a more robust Th1 response when stimulated in the presence of IL-12 and anti-IL-4 Abs than WT cells, which secrete relatively higher amounts of IL-4, suggests a plausible role for T-bet expression on the reinforced Th1 phenotype measured in MIF KO CD4+ T cells. However, IFN-γ blockade with a neutralizing antibody did not improve the IL-4 response in MIF KO CD4+ T cells stimulated in the presence of IL-12; this excludes a possible involvement of a T-bet/IFN-γ autocrine loop (38) in the impaired IL-4 responses measured (data not shown). As enhanced IFN-γ responses and compromised IL-4 responses were found in MIF KO CD4+ T cells both in vitro and in vivo, it would appear that these cells have an “intrinsic” ability to differentiate into Th1 cells when cultured in the presence of IL-12 and in an IL-4 poor environment.
In BALB/c mice, primary infections with the P.c. adami DK strain are self-controlled and characterized by patent parasitemia from day 3 to day 12 post-infection. Activation of macrophages is considered a first protective response to P. chabaudi infection and associates with by TNF-α and IL-12 production, which are known to stimulate NK cells (29). IFN-γ production by NK cells has been shown to be crucial for protection against P. c. chabaudi AS blood-stage infection in C57BL/6 mice (reviewed in (39) and contributes to control the primary wave of infection, allowing in parallel the development of an adaptive Th1 immune response (40, 41). The contribution for NK cells in the resolution of P. c. adami infection in BALB/c mice remains unclear as we failed in detecting IFN-γ secreting DX5+ cells at early or peak infection in wild type and MIF KO mice (data not shown).
Our interest in studying the effects of MIF neutralization/deficiency on P.c. adami infection was driven by several indications of a regulatory role for MIF on T cell activation. MIF is constitutively expressed in T cells and is secreted in response to mitogens or antigens (1), and its neutralization enhances CTL responses and IFN-γ secretion in a murine model of cancer (12). In addition, MIF appears essential for the host response to Schistosoma japonicum infection, which is controlled by Th2 responses (15). We therefore hypothesized that MIF should modulate the balance between Th1 and Th2 effector responses required for efficient control of P.c. adami DK infection. In this context, Th1 responses are pivotal to control primary infection whereas Th2 responses sustain antibody-dependent control of secondary infections (28). Accordingly, an early IFN-γ response is induced by P.c. adami DK parasites in resistant mice (29), and vaccines triggering IFN-γ responses also significantly inhibit peak and cumulative parasitemia with this parasite strain (31). Furthermore, administration of recombinant IFN-γ delays the onset of P.c. adami DK parasitemia in CBA mice, suggesting indeed that the early presence of this cytokine is protective (30). In contrast, Th2 responses are accentuated at the early phase of infection in susceptible mouse strains (27, 42). It also should be noted that in agreement with these experimental studies, lower parasitemas have been noted falciparum malaria patients with low expression, MIF alleles (Zhong et al. Nucleic Acids Research, 2005, Vol. 33, No. 13).
Protective immunity to malaria relies in part on cytophilic IgGs and antibody-dependent cell mediated inhibition (ADCI), a process by which parasites opsonised with cytophilic antibodies bind to specific Fc receptors expressed in monocytes and induce the release of soluble inhibitors of parasite growth (43). IFN-γ is essential for this response as it promotes the synthesis of cytophilic IgG isotypes and blocks the induction of non cytophilic IgG/IgE by IL-4 (44). Interestingly, it has been suggested that the delayed acquisition of immunity to malaria in endemic regions may be concurrent to a IL-4-driven switch from cytophilic IgG to non-cytophilic IgE induced in response to the current co-infections with helminths (worms), specially hookworms (45). From the functional perspective, we speculate that the relatively higher IL-4 levels found in mice expressing MIF during infection, besides mitigating IFN-γ responses in CD4+ T cells, may also modulate Ab isotype switching. Apart from ADCI, opsonising antibodies enhance the phagocytosis and elimination of infected red blood cells by macrophages in vitro and are induced by vaccines protecting BALB/c mice against P.c. adami infection (46, 47). The potential regulatory role for MIF on IL-4 responses in granulocytes, on Th2 sketching and on protection in malaria is interesting as these cells, centrally involved in the acute inflammatory reactions to allergens, are now being considered general inducers of Th2 responses (34). Two studies in C57BL/6 mice associated protective effects to TNF-α production by IgE-activated mast cells against P. berghei blood stage infection (48, 49). Peroxiredoxin, an enzyme involved in the detoxification of hydrogen peroxide and nitric oxide was identified as a target for IgEs, and these antibodies were shown to be required for TNF-α secretion by mast cells through a TLR4-dependent mechanism (49). The results with the P. berghei/C57BKL/6 model are in disagreement with our data using the P.c. adami/BALB/c model. Although early TNF-α production by cells others than CD4+ T cells was shown to be reduced in P. chabaudi-infected MIF KO mice, these mice controlled more efficiently primary parasitemia.
The molecular events modulating CD4+ T cell commitment in P.c. adami infections are unknown and may differ from those conferred by mere in vitro stimulation with anti-CD3 and anti-CD28 Ab. It is assumed that the intensity and duration of the primary stimulus through the TCR may profoundly influence type 1/type 2 differentiations. Th2 cell development requires longer periods of TCR engagement and the polarizing effect of IL-4 is induced only if this cytokine is present during the initial events in TCR triggering (50). As significantly lower IL-4 responses are induced in the BALB/c MIF KO mice at day 4 post-infection and as these responses remained relatively low throughout patent infection, we suggest that an unfavourable environment for Th2 cell differentiation is a major cause for enhanced IFN-γ responses in MIF KO CD4+ T cells from mice at early infection.
In opposition to improved resolution of P.c. adami infection in MIF deficient hosts, intracellular infections with Trypanosoma cruzi, Toxoplasma gondii and Leishmania major, essentially controlled by Th1 cells are exacerbated in MIF deficient hosts (51–53). It is tempting to suggest that the contrasting outcomes of MIF deficiency on P.c. adami versus T. gondii, T. cruzi and L. major infections rely on the probable involvement for MIF in IFN-γ-independent innate responses that are protective in these infections but deleterious in malaria. This assumption is supported by the fact that purified MIF has been shown to directly mediate intracellular killing of L. major through an IFN-γ-independent but TNF-α-dependent mechanism (54). In addition, GPI-anchored proteins from T. gondii trigger the release of TNF-α in macrophages through TLR-2 and TLR-4 signalling, which is important for protection as TLR-2/TLR-4 double-deficient mice, albeit producing IL-12 and IFN-γ, are more susceptible to Toxoplasmosis (55). In this context, the lower TLR expression reported in MIF KO macrophages (9), also corroborated in our laboratory, may hamper protective T. gondii GPI-dependent TNF-α responses in MIF-deficient mice. In contrast, MyD88 and TLR are important factors in malarial pathology (56). Indeed, MyD88 deficient mice display decreased TNF-α and IFN-γ responses and minor morbidity when infected with non-lethal P.c. adami DK parasites and interestingly, this deficiency does not affect their capacity to rapidly resolve secondary infections (57).
Taken together, our data suggests that MIF is not only a contributor to inflammation, but also attenuates the development of Th1 responses in malaria by enhancing IL-4 responses, most probably by basophils and mast cells. The role for MIF on the down-regulation of T-bet expression in resting CD4+ T cells and its eventual impact on their abilities to secrete IFN-γ and to respond to IL-12 stimulation remains to be investigated.
This work was supported by the Natural Sciences and Engineering Research Council of Canada discovery grant 311589 (T. Scorza) and the National Institutes of Health Grants A1042310 and AI051306 (R. Bucala).