CD4+Foxp3+ regulatory T cells mediate Toxoplasma gondii-induced T-cell suppression through an IL-2-related mechanism but independently of IL-10

Authors

  • Eda P. Tenorio,

    1. Departamento de Inmunología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Mexico City, Mexico
    Search for more papers by this author
  • Jacquelina Fernández,

    1. Departamento de Inmunología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Mexico City, Mexico
    Search for more papers by this author
  • Carlos Castellanos,

    1. Departamento de Inmunología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Mexico City, Mexico
    Search for more papers by this author
  • Jonadab E. Olguín,

    1. Departamento de Inmunología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Mexico City, Mexico
    Search for more papers by this author
  • Rafael Saavedra

    Corresponding author
    1. Departamento de Inmunología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Mexico City, Mexico
    • Departamento de Inmunología, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Ap. Postal 70228, Ciudad Universitaria, CP 04510, Mexico DF, Mexico Fax: +52-5555503893
    Search for more papers by this author

Abstract

Acute Toxoplasma gondii infection comprises an immunosuppression stage, characterized by a reduction in T-cell proliferation in vitro. Treg cells maintain the homeostasis of the immune system, but their role in T. gondii-induced suppression has not been addressed. We show herein that immunosuppression, affecting both CD4+ and CD8+ T-cell proliferation, concurs with a reduction in Treg-cell number. The residual Treg cells, however, are activated and display an increased suppressive capacity. We show that selective elimination of Treg cells using Foxp3EGFP mice leads to a full recovery of CD4+ and CD8+ T-cell proliferation. After Treg-cell removal, a reduced production of IL-10 was observed, but IL-2 levels were unchanged. The numbers of IL-10-producing Treg cells also increased during infection, although the in vitro neutralization of this cytokine did not modify T-cell proliferation, suggesting that IL-10 does not mediate the Treg-mediated suppression. However, addition of rIL-2 in vitro fully restored T-cell proliferation from infected animals. Thus, we show that Treg cells mediate the T-cell suppression observed during acute T. gondii infection through an IL-2-dependent mechanism. Our results provide novel insights into the regulation of the immune response against T. gondii.

Introduction

Toxoplasma gondii is a worldwide distributed intracellular protozoan parasite that infects approximately one-third of the human population. Toxoplasmosis is usually clinically asymptomatic in healthy individuals, but it can cause severe complications in pregnant women and immunocompromised patients. In the latter, chronic infection can reactivate leading to disseminated toxoplasmosis and/or encephalitis that are often lethal. Primary infection during pregnancy may lead to abortion, neonatal malformations or defects that appear during child development 1.

Infection with T. gondii activates DCs to produce large amounts of IL-12 2, 3 which in turn activates NK cells and T lymphocytes to produce IFN-γ 4, 5 leading to macrophage activation and parasite control 6, 7. A TH1 immune response and cooperation between CD4+ and CD8+ T cells are crucial for infection control 5, 8, 9. Downregulation of the extremely strong TH1 immune response caused by infection is mediated by IL-10, lipoxinA4 and IL-27 10–12.

During acute infection with T. gondii a transient reduction in the proliferative response of T cells to mitogens or antigens is observed in humans and mice 13–17. Analysis of cells and molecules involved in the immunosuppression observed during T. gondii infection has shown that IFN-γ-dependent reactive nitrogen intermediates (RNIs) produced by macrophages and IL-10 are implicated in this process 16–21. However, neutralization of these molecules restores only partially T-cell proliferation capacity. Furthermore, it has been demonstrated that splenocytes from infected IL-10−/− and IRF-1−/− mice are also suppressed 19, 22, thus indicating that additional mediators are involved in immunosuppression. A recent report suggested that Treg cells could be involved in the suppression observed during other parasitic infection 23.

Treg cells are CD4+ lymphocytes that constitutively express CD25 24, CTLA-4 25 and the Treg cell-specific transcription factor Foxp3 26, 27, which is required for the development and the suppressive capacity of these cells. Treg cells are involved in control of autoimmunity, immune response against tumors, tissue transplants and infectious agents 28, 29. Several roles for Treg cells during infectious diseases have been described, including control of the pathology, maintenance of immunity against reinfection and favouring pathogen establishment or persistence 29. We previously showed that Treg cells play an important role in the protective response against T. gondii, since removal of Treg cells led to an increased mortality rate in the resistant BALB/c mouse strain 30. Moreover, treatment of T. gondii-infected susceptible C57BL/6J mice with IL-2-anti-IL-2 complexes resulted in an increased Treg-cell frequency and survival, which correlated with reduced morbidity 31. Additionally, adoptive transfer of Treg cells has been reported to reduce the abortion rate in pregnant mice injected with excretory–secretory antigens from the parasite 32. These studies demonstrate that Treg cells are important mediators of the immune response during T. gondii infection.

The aim of this study was to determine whether Treg cells are involved in the immunosuppression observed during acute infection with T. gondii. We studied the suppression induced in C57BL/6J mice infected with the ME49 strain of T. gondii. We analysed the different cell subsets suppressed and characterized the Treg-cell population, including their suppressive capacity and expression of activation molecules. We evaluated the role of Treg cells in immunosuppression by selective elimination of these cells using Foxp3EGFP mice and explored some possible mechanisms for Treg cell-induced suppression during T. gondii infection.

Results

Infection with T. gondii induces suppression of CD4equation image and CD8equation image T-cell proliferation

In order to evaluate the suppression of different cell types during acute T. gondii infection, we analysed the mitogen-induced proliferation of splenocytes from C57BL/6J mice using CFSE. A representative FACS analysis (Fig. 1A) showed that proliferation of ungated splenocytes at 7 d post infection (dpi) was slightly reduced when compared with cells from uninfected mice, but at 14 dpi the reduction was stronger. Cell proliferation, however, was completely restored at 21 dpi. The same proliferation pattern was observed in CD4+ T cells. The proliferation of CD8+ T cells at 7 dpi was comparable to that of cells from uninfected animals, but was dramatically reduced at 14 dpi, and was restored at 21 dpi, while LPS-induced B-cell proliferation was not affected. Accordingly, data from different experiments showed that the percentage of divided cells from the ungated population (Fig. 1B) is significantly reduced at 7 and 14 dpi. The percentage of CD4+ divided cells was halved at 7 and 14 dpi, while in the CD8+ subset it was only significantly reduced at 14 dpi. The percentage of CD19+ divided cells, however, increased approximately 30% and remained significantly higher during the period analysed. These data demonstrate that T. gondii-induced immunosuppression in Con A-stimulated splenocytes and in isolated CD4+ T cells observed by 3H-thymidine incorporation 15, 33 is also detected using CFSE dilution. Furthermore, we show that CD4+ and CD8+ T cells have different suppression patterns while CD19+ cells display an increased proliferation.

Figure 1.

Infection with T. gondii induces immunosuppression in CD4+ and CD8+ T cells but not in B cells. Splenocytes from WT mice obtained at the indicated times post infection, were stained with CFSE and stimulated with Con A or LPS for 72 h. Cells were stained with anti-CD4, anti-CD8 or anti-CD19 mAbs and 7-AAD and immediately analysed by FACS. The lymphocyte gate was defined by FSC and SSC characteristics within living cells (7-AAD, Ungated), further subgatings of the CD4+, CD8+ and CD19+ cells were carried out and 10 000 events of each subpopulation were captured. (A) Representative analysis of proliferation by CFSE dilution for each gate. (B) Results of two independent experiments with three mice per group per time point expressed as percentage of divided cells. Data obtained with cells from infected mice were compared with data obtained with cells from uninfected mice (day 0) using One-Way ANOVA and Dunnett's multiple comparison test, *p<0.05, **p<0.01.

Treg-cell number decreases but residual Treg cells activate and increase their suppressive capacity

Given that Treg cells suppress the proliferation of different cell types 34, it was tempting to speculate that the immunosuppression observed during T. gondii infection could be mediated by this cell population. However, as can be observed in a representative FACS analysis (Fig. 2A), the percentage of CD4+Foxp3+ cells decreased at 7 dpi, and markedly dropped at 14 dpi. Results from several experiments showed that Treg-cell percentage decreased by 16.3% at 7 dpi and by 50.4% at 14 dpi (Fig. 2B) when compared with control animals. A similar reduction in the absolute number of Foxp3+ cells was also detected (Fig. 2C), demonstrating that the decline in Treg-cell percentage is not consequence of a disparity in the proportion of other cell subsets. Further analysis of the residual Treg cells showed that at 7 dpi the percentage of natural Treg cells (Helios+) and induced Treg cells (Helios) is comparable to that observed in uninfected animals, whereas at 14 dpi a slight reduction in the proportion of natural Treg cells was observed (Fig. 2D and E).

Figure 2.

Infection with T. gondii induces a reduction in Treg-cell number. Splenocytes from either WT or Foxp3EGFP mice were obtained at 7 and 14 days post infection (dpi). (A, B, D, E) Cells from WT mice were incubated with anti-CD4, anti-Foxp3 and anti-Helios mAbs and analysed by FACS. The lymphocyte region was first identified and gated according to its FSC and SSC characteristics, the CD4+ region was subgated and 10 000 CD4+ cells were captured. (A) Representative results of Treg-cell detection; percentages of CD4+Foxp3+ cells in the CD4+ gate are displayed. (B) Results of five independent experiments with at least two mice per group expressed as percentage of Foxp3+ cells in the CD4+ gate. (C) Cells from Foxp3EGFP mice were mixed with CountBright Absolute Counting Beads and immediately analysed by FACS. Results from two independent experiments with at least two mice per group are expressed as absolute numbers of Foxp3EGFP cells per spleen. (D) Representative results of Helios expression detection; percentages of Foxp3+ Helios+ cells in the CD4+ Foxp3+ gate are shown. (E) Results of two independent experiments with at least three mice per group are expressed as percentage of Helios+ cells in the CD4+ Foxp3+ gate. In (B, C, E), results were analysed using One-Way ANOVA and Bonferroni's multiple comparison test, *p<0.05, **p<0.01, ***p<0.001.

The above results indicate that T. gondii-induced suppression concurs with a reduction in Treg cell number. In order to explain this apparent contradiction, we analysed the expression of activation markers in the residual Treg-cells. We focused on cells from mice at 7 dpi because at this time point immunosuppression was already detected and the number of Treg cells still allowed a proper analysis. Expression of CD25, CTLA-4 and GITR rose up in Foxp3+ cells from infected mice (2.5-, 3- and 0.5-fold, respectively); the proportion of Treg cells expressing these molecules was also slightly increased (Fig. 3). Analysis of additional activation molecules showed that the percentage of CD69+ and CD62L cells increased 1.9- and 1.3-fold, respectively. Modulation of these molecules has already been reported after Treg-cell activation 25, 35–37. A significantly enhanced expression of CD69 was also detected; expression of CD62L and CD103 remained unchanged. Thus, although infection leads to a reduction in Treg-cell number, the residual cells display an activated phenotype.

Figure 3.

Treg cells display an activated phenotype in acute T. gondii infection. Splenocytes from WT mice obtained at 7 dpi were stained with anti-CD25, anti-CTLA-4, anti-CD69, anti-CD62L, anti-CD103 or anti-GITR, and analysed by FACS. The lymphocyte region was first identified and gated according to its FSC and SSC characteristics, the CD4+Foxp3+ region was subgated and 5000 events of this population were captured. Representative histograms are shown for each of the studied activation marker; dotted line indicates the fluorescence threshold. Mean fluorescence intensity was normalized and graphed as percentage expression. Percentage of cells expressing each molecule is also shown. Data were obtained from two independent experiments, including at least two mice per group. Data were analysed using Student t test, *p<0.05, **p<0.01, ***p<0.001. NI: Uninfected. Inf: Infected.

Treg-cell activation observed after infection suggested that these cells might also increase their suppressive capacity. We thus compared the suppression capacity of Treg cells from infected and uninfected mice against target cells from uninfected animals. We initially carried out suppression assays using CD4+CD25+ cells as Treg cells and CD4+CD25 cells as target cells, and found a slight increase in the suppression capacity of CD4+CD25+ cells obtained from infected mice (data not shown). Although this separation protocol is the most commonly used, an increase in the CD4+Foxp3CD25+ cell population, corresponding to activated T cells, is observed in infected mice (Fig. 4A, 1.3 versus 17.5%). Therefore, the CD4+CD25+ fraction used in that system was enriched with activated T cells, and the suppression capacity of Treg cells from infected animals cannot be addressed. In order to overcome this issue, we used Foxp3EGFP mice to sort Foxp3+ Treg cells and CD4+Foxp3CD25 cells as target cells to carry out the suppression assay (Table 1). Results obtained from three independent experiments showed that although Treg cells from uninfected animals are able to suppress proliferation at various degrees (36.1–85.7%), Treg cells from infected mice induced a significantly higher suppression of target cells proliferation (84.3–97.4%); as expected, Treg cells alone were unable to proliferate under these conditions. These results demonstrate that during infection, the residual activated Treg cells display an increased suppressive capacity.

Figure 4.

Treg cells mediate immunosuppression during acute T. gondii infection. Foxp3+ cells from uninfected or infected Foxp3EGFP mice splenocytes were eliminated by cell sorting. The Foxp3 cells were stained with CFSE and stimulated with Con A for 72 h. (A) Representative analysis of the purity of the sorted populations according to CD25 and Foxp3 expression. (B) Proliferation was evaluated as described in Fig. 1, in the ungated, CD4+, and CD8+ populations. Representative histograms are shown. (C, D) Results of two independent experiments with two mice per group per time point expressed as percentage of divided cells (C) and percentage of proliferating cells (D). Results were analysed using ANOVA and Bonferroni's multiple comparison test, *p<0.05, **p<0.01. NI: Uninfected. Inf: Infected.

Table 1. Con A-induced proliferation of target cells is highly suppressed by Treg cells from T. gondii-infected mice
Target cellsa)Treg cellsa)[3H] Thymidine uptake (cpm±SD)
  Experiment 1Experiment 2Experiment 3
  • a

    Data obtained from each co-culture were compared with that obtained with target cells alone according to One-Way ANOVA and Dunnett's multiple comparison test. *p<0.05, **p<0.01, ***p<0.001.

  • a)

    a) Target cells (CD4+Foxp3CD25) and Treg cells (CD4+Foxp3+) from Foxp3EGFP mice were separated by cell sorting. Cells were seeded at a ratio 2:1 (Target:Treg cells).

  • b)

    b) Percentage of suppression was calculated according to the following formula: [(cpm0−cpmmix)/cpm0]×100, where cpm0 is the [3H]-Thymidine uptake by target cells alone and cpmmix is the uptake by the co-culture of targets and Treg cells.

Uninfected6600.6±268.96858.1±1431.38963.7±2882.3
UninfectedUninfected4217.6±1426.2* (36.1%)b)3707.9±865.1* (46.0%)1278.9±554.4* (85.7%)
UninfectedInfected1035.0±162.4*** (84.3%)1897.1±282.2** (72.3%)228.8±69.7* (97.5%)
Uninfected254.6±78.1292.1±53.41213.8±1151.5
Infected202.9±26.9112.8±14.967.3±2.5

Treg cells mediate T. gondii-induced suppression

The activated phenotype and the increased suppression capacity of the residual Treg cells could explain the apparent discrepancy between the immunosuppression and the reduced proportion of Treg cells observed during infection. In a first attempt to evaluate the role of Treg cells in the observed immunosuppression, we injected animals with anti-CD25 mAb and examined whether proliferation was recovered. However, as we previously reported, treatment of C57BL/6J mice with anti-CD25 mAb before infection eliminates mainly activated cells, and thus the role of Treg cells is impossible to elucidate using this approach 38. Thus, we used Foxp3EGFP mice to directly assess whether Treg cells mediate immunosuppression.

Foxp3+ cells were eliminated by cell sorting (Fig. 4A) and proliferation of Foxp3 cells was analysed (Fig. 4B). As expected, proliferation of ungated, CD4+ and CD8+ lymphocytes was suppressed when unsorted splenocytes were assayed. These results are indistinguishable from those shown in Fig. 1, demonstrating that the EGFP+ phenotype does not alter the immunosuppression pattern of T. gondii-infected mice. When Foxp3+ cells were eliminated from infected mice splenocytes, a proliferation recovery was clearly observed in the ungated population. CD4+ cells showed a strong proliferation, similar to that observed in cells from uninfected mice. CD8+ cells from infected animals also recovered their proliferative response. Elimination of Foxp3+ cells from uninfected mice did not alter proliferation of CD4+ nor CD8+ cells.

Statistical analysis of the data collected from two independent experiments confirmed that after Treg-cell removal the percentage of divided CD4+ cells from infected mice was significantly enhanced and was similar to that of cells from uninfected animals (Fig. 4C); a non-significant increase in the percentage of divided cells from the ungated and CD8+ subsets was observed. Since the percentage of divided cells only represents the proportion of the original population that responded by dividing 39 we also calculated the percentage of proliferating cells (cells found in any round of division). Figure 4D shows that when Treg cells are eliminated, the percentages of proliferating CD4+ and CD8+ cells are similar for uninfected and infected animals. These data thus demonstrate that Treg cells mediate the immunosuppression observed during acute T. gondii infection.

Treg cells mediate immunosuppression through an IL-2-related mechanism but independently of IL-10

We analysed some possible mechanisms that could explain the Treg cell-mediated immunosuppression described above. Since it was previously reported that during T. gondii-induced suppression, IL-2, RNIs and IL-10 are involved 16, 17, 20, 21, 40, we evaluated the effect of Treg-cell removal on the production of these mediators in vitro. NO2 production was similar in cells from uninfected and infected animals and Treg-cell elimination had no effect in the production of this molecule (Fig. 5), demonstrating that in our system RNIs are not involved in Treg cell-mediated suppression.

Figure 5.

Selective elimination of Treg cells reduces IL-10 production, but IL-2 and NO2 levels are unaffected. Foxp3 splenocytes were sorted as described in Fig. 4. Culture supernatants were collected after a 24 h stimulation period with Con A, and IL-2, IL-10 and NO2 production was determined. Data were analysed using ANOVA and Bonferroni's multiple comparison test, ***p<0.001. NI: Uninfected. Inf: Infected.

The role played by IL-10 in T. gondii-induced suppression has been controversial 17, 19–22. However, since it has been described as a suppressive mechanism of Treg cells, we analysed IL-10 production. As can be observed in Fig. 5, no IL-10 could be detected in culture supernatant of cells from uninfected mice, while cells from infected animals produced highly significant levels of IL-10. Moreover, elimination of Treg cells led to a drastic reduction of the cytokine level. Because this reduction in IL-10 levels correlated with a recovery of T-cell proliferation after Treg-cell removal, we hypothesized that IL-10 produced by Treg cells could be a key molecule involved in the suppression. We thus first analysed IL-10 production by Foxp3+ and Foxp3 cells from infected mice. As can be observed in Fig. 6, IL-10 was produced by both Foxp3+ and Foxp3 cells, but after infection, a 3-fold increase in the proportion of IL-10-producing cells was observed in the Treg-cell population only, suggesting that these cells were the source of the increased amount of IL-10 found in the supernatant. We next carried out in vitro IL-10 neutralization in order to test if this cytokine was responsible of the Treg cell-mediated suppression. Addition of anti-IL-10 mAb did not alter the proliferation of the ungated, the CD4+ and CD8+ subsets from infected mice (Fig. 7A and B) demonstrating that IL-10 was not responsible for the Treg-cell suppressive effect on CD4+ and CD8+ T cells, despite the increased proportion of IL-10-producing Treg cells detected during infection.

Figure 6.

The numbers of IL-10-producing Treg cells increase during acute infection with T.gondii. Splenocytes from uninfected or infected Foxp3EGFP mice were obtained at 7 dpi. Cells were stimulated with PMA/ionomicin, stained with anti-CD4 and anti-IL-10 mAb and analysed by FACS. The lymphocyte gate was defined using its FSC and SSC characteristics within the CD4+ gate; further subgatings of the CD4+Foxp3+ and CD4+Foxp3 subsets were performed and these subsets were analysed for IL-10 expression. Five thousand CD4+ Foxp3+ events were captured. (A) Representative contour plots of IL-10-producing cells in the Foxp3+ and Foxp3 populations. (B) Data obtained from two independent experiments with at least two mice per group. Results were analysed using One-Way ANOVA and Bonferroni's multiple comparison test, ***p<0.001. NI: Uninfected. Inf: Infected.

Figure 7.

IL-10 neutralization does not modify T. gondii-induced suppression. Splenocytes from WT mice were obtained at 7 dpi, stained with CFSE and stimulated with Con A in the presence of anti-IL-10 (30 μg/mL) or isotype control mAbs. After 72 h, cells were stained and analysed as described in Fig. 1. (A) Representative analysis of cells obtained from one uninfected and one infected animal. Proliferation by CFSE dilution for each gate is shown. (B) Data obtained from two independent experiments including three mice per group, showing the percentage of divided cells. No statistical difference was found between groups using One-Way ANOVA and Bonferroni's multiple comparison test.

We finally explored the possibility that the observed suppression by Treg cells was IL-2-dependent. IL-2 levels in culture supernatants of stimulated splenocytes were drastically reduced in the supernatant of cells from infected animals when compared with uninfected animals (Fig. 5), as reported 17, 20, 21, 31, 33. Removal of Treg cells, however, led to a slight but non-significant reduction of IL-2 levels (Fig. 5), suggesting that Treg cells do not suppress IL-2 production. The absence of IL-2 accumulation also indicated that either this cytokine is not involved in Treg cell-mediated immunosuppression or that the Treg and conventional T (Tconv) cells could compete for the reduced IL-2 concentrations. Thus, to evaluate if the Treg-mediated suppression was related to a reduced availability of this cytokine, we added rIL-2 to cell cultures from infected mice, and found that proliferation of the ungated as well as the CD4+ and CD8+ cells was fully restored (Fig. 8A and B). Proliferation of T cells from uninfected mice, however, was unaffected by rIL-2 addition (Fig. 8A and B). All these results demonstrate that the Treg cell-mediated immunosuppression observed during acute T. gondii infection is consequence of a reduced IL-2 availability for T cells.

Figure 8.

Treg-cell-mediated immunosuppression during acute T. gondii infection is IL-2 dependent. Splenocytes from WT mice were obtained at 7 dpi, stained with CFSE and stimulated with Con A in the presence or absence of rIL-2 (20 U/mL). After 72 h, cells were stained and analysed as described in Fig. 1. (A) Representative analysis of cells obtained from one uninfected and one infected animal. Proliferation by CFSE dilution for each gate is shown. (B) Results of two independent experiments with three mice per group showing the percentage of divided cells. Data were analysed using ANOVA and Bonferroni's multiple comparison test, *p<0.05, **p<0.01, ***p<0.001.

Discussion

The aim of this work was to evaluate a possible role for Treg cells in the immunosuppression observed during the acute phase of T. gondii infection in C57BL/6J mice. This suppression has been described using different mitogens and the [3H]-thymidine incorporation assay. In order to determine the cell types affected by the parasite, we analysed proliferation of mouse splenocytes using CFSE. Our results confirm previous findings showing that T cells are unable to respond to mitogens during acute infection 15, 16, 33 and further show that only CD4+ and CD8+ T cells, but not B cells, were affected. Although suppression of CD4+ T cells has already been reported 33, this is the first report describing suppression of CD8+ T cells during T. gondii infection.

Treg cells suppress the proliferation and cytokine production of other cells 34 and have been shown to control immune response in several infection models 29. These properties suggested that these cells could mediate the immunosuppression observed during T. gondii infection. However, we found a reduction in the proportion and absolute numbers of Treg cells during the first two wks of infection, an observation which is in agreement with the recent reports 30–32. Oldenhove et al. recently reported a decrease in Treg cell number during T. gondii infection related to the inhibition of peripheral induction of Foxp3+ T cells in GALT 31 and suggested that an impaired Treg-cell conversion might be involved in this reduction. In order to further characterize the Treg-cell phenotype during infection, we examined the transcription factor Helios which has been recently described as a molecule that can be used to discriminate between natural and induced Treg cells 41, and it has already been employed as a marker in murine and human models 42–44. Analysis of this molecule in the residual Treg cells of T. gondii-infected mice showed that the proportion of natural and induced Treg cells was unchanged at 7 dpi, although a slight increase in Helios cells was observed at a later time point, suggesting that the amount of induced Treg cells is not impaired during the first wk of infection. However, a recent study demonstrated that Helios expression is more related to the method of activation of T cells than to the Treg-cell origin 45. Thus, given that the use of Helios as a definitive marker for natural Treg cells is still unclear, further studies are required to address this issue.

Treg-cell number reduction seemed contradictory with the observed immunosuppression. The paradox of a reduced number of Treg cells mediating suppression could be explained if the residual Treg cells were activated and displayed an increased suppressive capacity. The remaining Treg cells were indeed highly activated, as denoted by the increased expression of CD25, CTLA-4, CD69 and GITR, the loss of CD62L expression and their capacity to produce IL-10. Furthermore, suppression assays showed that Treg cells from infected animals display an increased suppressive capacity when compared with cells from uninfected mice. Since at the time point studied (7 dpi) a reduction of only 16.3% of Treg cells is observed, the activation and acquisition of a higher suppressive capacity of the remaining Treg cells could easily explain the ability of these cells to mediate immunosuppression. The activation of Treg cells described herein is consistent with data previously reported during other infectious diseases 46–50, and supports the idea that Treg-cell activation could be a natural response towards some pathogens. Whether Treg-cell activation depends on molecules derived from the parasite, on the proinflammatory environment, or both, remains to be established.

The increased suppressive capacity we observed in Treg cells from infected animals, however, contrasts with a recent report indicating that there is no difference between the suppression capacity of Treg cells from T. gondii-infected animals and that of uninfected mice 31. The discrepancy could be explained by differences in inoculum size, animal sex, T cell stimuli, source of T cells used in the assay and the methodology used for detection of proliferation.

Regardless of Treg-cell number reduction, the activation and increased suppressive function of the remaining Treg cells supported the hypothesis that these cells were involved in the immunosuppression. Full restoration of the proliferation pattern of CD4+ and CD8+ cells from infected mice splenocytes after selective elimination of Foxp3+ cells definitively demonstrated that Treg cells are the key cells mediating the suppression observed during acute T. gondii infection.

Since this is the first time that T. gondii-induced suppression is fully reversed, we studied some possible mechanisms to explain the Treg cell-mediated suppression. Earlier reports showed that RNIs produced by macrophages are important for induction of T. gondii-induced suppression 16, 17, 21, 22, 40. However, we did not find alterations in the in vitro NO2 concentration, neither after infection or after Treg-cell elimination, demonstrating that in our model NO2 is not involved in the suppression induced by Treg cells. Our results are supported by the data of Khan et al., who showed that Con A-stimulated splenocytes from T. gondii-infected IRF-1−/− mice remained suppressed even in the presence of the RNI inhibitor NG-monomethyl-L-arginine monoacetate 19.

Interestingly, other molecules previously studied as mediators of immunosuppression are also reported mechanisms of Treg-cell function, including the inhibitory cytokines TGF-β and IL-10 34. TGF-β does not seem to participate in T. gondii-induced suppression, since we did not detect membrane bound TGF-β in Treg cells from infected mice (data not shown), and previous reports showed that addition of anti-TGF-β antibodies to in vitro cultures of spleen cells from infected mice does not reverse immunosuppression 19, 20.

We thus analysed the possible role of IL-10 and found an increased level of this cytokine in cell culture supernatants from infected animals, as previously reported 17, 19–21, 33; Treg-cell removal led to a reduction in IL-10 levels, an observation that correlated with T-cell proliferation recovery. Additionally, we found an increased proportion of IL-10-producing Treg cells in infected animals, a result that reinforced the hypothesis that this cytokine could be responsible for the immunosuppression. This result was unexpected since it was previously reported that during infection with T. gondii most IL-10 is produced by Foxp3 TH1 cells 51. However, our results are supported by data previously published by Oldenhove et al. 31, who demonstrated that despite Treg-cell number reduction, these cells maintain their capacity to produce IL-10.

Analysis of CD4+ and CD8+ T-cell proliferation in the presence of anti-IL-10 mAb, however, revealed that this cytokine does not mediate immunosuppression. Our results agree with those obtained in T. gondii-infected IL-10−/− mice, where T-cell suppression is similar to that observed in WT mice 22, although earlier reports of IL-10 in vitro neutralization in splenocytes from infected animals showed a partial reversion of suppression 17, 19–21. Thus, despite an increase in IL-10-producing Treg cells in infected animals, and the concomitant reduction in IL-10 levels and T-cell proliferation recovery after Treg-cell removal, IL-10 is not involved in the Treg cell-mediated immunosuppression.

Given the lack of contribution of RNIs and IL-10 in Treg cell-mediated suppression, we evaluated a possible role of IL-2, since deprivation of this cytokine is a reported Treg-cell mechanism 52–55. We found reduced IL-2 levels in culture supernatants of cells from infected animals, as reported 17, 20, 21, 31, 33. Treg-cell removal did not restore IL-2 levels but fully reversed T-cell proliferation, suggesting that Treg cells do not inhibit IL-2 production. In contrast, when rIL-2 was added to cell cultures, complete restoration of T-cell proliferation occurred, even in the presence of Treg cells. Therefore, proliferation recovery was independently achieved either by removing Treg cells or by addition of rIL-2, showing that immunosuppression mediated by Treg cells during T. gondii infection is a consequence of a lack of IL-2 for Tconv cells.

The fact that T-cell proliferation from infected animals was fully restored in the absence of Treg cells (Fig. 5), even if IL-2 levels were low (Fig. 6), demonstrates that Tconv cells proliferate in a milieu of low IL-2 availability. These observations thus suggest that when limiting amounts of IL-2 exist, competition for this cytokine could take place between activated Treg and Tconv cells. Hence, Treg cells in our model might act by IL-2 deprivation. This hypothesis is supported by a recent mathematical model reported by Busse et al. 56 predicting that IL-2 deprivation by Treg cells occurs under conditions of limited IL-2 supply.

Clear evidence of IL-2 deprivation was recently provided by Pandiyan et al. 53, who demonstrated that Treg cells “imbibe” more IL-2 than Tconv cells, particularly after activation, and this IL-2 deprivation leads to apoptosis of Tconv cells. In our model, Treg cells are activated and express very high levels of CD25 and could thus become more efficient IL-2 consumers. Furthermore, we observed that addition of IL-2 also led to increased cell viability (data not shown). The results obtained in our work thus strongly suggest that Treg cells mediate immunosuppression by IL-2 deprivation. However, additional experiments are required to confirm this hypothesis.

IL-2 is a molecule essential for mice survival after T. gondii infection 31, 57 and our results highlight the importance of this cytokine. It has been demonstrated that the reduced number of Treg cells during acute infection is consequence of a reduced IL-2 availability 31, and is probably related to IL-27 58, which has been shown to cooperate with IL-12 to suppress IL-2 production during acute infection 59. Our results suggest that the reduced IL-2 levels favours the competition for this cytokine between activated Treg cells and Tconv cells and that IL-2 exhaustion by activated Treg cells leads to the immunosuppression of CD4+ and CD8+ cells, but not of B lymphocytes, that do not require IL-2 for proliferation 60. These events could thus contribute to the highly inflammatory immune response that is characteristic during T. gondii infection.

Analysis of Treg cells during T. gondii infection by several groups has shown a reduction of these cells in C57BL6/J, BALB/c and in pregnant mice 30–32. We have shown herein that regardless of their reduction, Treg cells display an activated phenotype and a higher suppressive capacity, leading these cells to mediate immunosuppression. Interestingly, IL-10 does not participate as a modulator of suppression, despite the increase of IL-10-producing Treg cells. Instead, our results suggest that IL-2 deprivation is the mechanism used by Treg cells to mediate T. gondii-induced suppression. The role of Treg cells we describe herein as the mechanism controlling immunosuppression opens a new insight in the immunoregulation previously described for T. gondi infection.

Materials and methods

Mice

Six–eight-wk-old C57BL/6J (WT), and Swiss-Webster mice were bred in our animal house and maintained in microisolator cages according to the institutional guidelines. Foxp3EGFP knock-in mice (B6.Cg-Foxp3tm2Tch/J), coexpressing Foxp3 and Enhanced GFP were obtained from The Jackson Laboratory and maintained under the same conditions in our animal house. All experiments were carried out with age and sex matched animals. Animal experimentation protocols were approved by the local Bioethics Committee for Animal Research.

Parasites and infection

The ME49 strain of T. gondii was maintained in Swiss-Webster mice as previously described 61. For parasite maintenance, Swiss mice were infected i.p. with ten cysts obtained from brains of infected animals. For peroral infection, mice weighing 18–20 g were anesthetized with Sevorane (Abbott) and infected by gavage with 25 cysts obtained from Swiss mice infected 2–4 months earlier.

Flow cytometry and mAbs

The following fluorochrome-conjugated mAbs were used: anti-CD3-FITC or -Cy5 (500A2); anti-CD4-TC, -PE or -APC (RM4-5); anti-CD8-FITC, -PE or -APC (5H10); anti-CD19-PE (6D5); anti-CD25-APC or -PE (PC61 5.3) from Caltag; anti-CD152-PE (CTLA-4, UC10-4B9); anti-Foxp3-Alexa Fluor 488 (FJK-16s) from eBioscience; anti-CD69-PE (H1.2F3), anti-CD62L-PE (MEL-14), anti-GITR-PE (DTA-1), anti-CD103-PE (2E7), anti-Helios-Alexa Fluor 647 (22F6) and anti-IL-10-PE (JES5-16E3) from Biolegend. Cell surface molecules were detected by incubating 106 cells with the indicated mAb in washing buffer (DPBS, 1% FCS, 0.1% NaN3) for 30 min (4°C, in the dark). Cells were washed twice, resuspended in DPBS and analysed by FACS. Foxp3, Helios and CTLA-4 were detected using the eBioscience Foxp3 detection kit following manufacturer's instructions. For viability determination, cells were stained with 1 μg/mL of 7-amino-actinomycin D (7-AAD, Molecular Probes), as previously described 62. Cells were acquired using a FACScan, FACScalibur or FACSAria cytometer (Becton Dickinson). Data were analysed using the FlowJo Software V.5.7.2 (Tree Star).

Determination of absolute Foxp3equation image cell numbers

Splenocytes from Foxp3EGFP mice were obtained by perfusion and red blood cells were lysed with hypotonic NH4Cl solution. Cells were washed and resuspended in 10 mL of DPBS. One hundred μL of the cell suspension were diluted 1:5 with DPBS and 50 μL of CountBright Absolute Counting Beads (Molecular Probes) were added. The diluted suspension was immediately analysed by FACS and the cell concentration was calculated following the manufacturer instructions. Total Foxp3EGFP cell number per spleen was calculated as described elsewhere 63.

Intracellular IL-10 detection

Ten million splenocytes from Foxp3EGFP mice were incubated with 20 ng/mL PMA, 1 μg/mL ionomycin and 2 μM monensin in 1 mL of complete RPMI medium (RPMI 1640 supplemented with 10% FCS, 2 mM L-glutamine, 10 mM non-essential aminoacids, 1 mM sodium pyruvate, 25 mM HEPES, 50 μM 2-ME and 50 IU/mL penicillin streptomycin [GIBCO]), in each well of a 24-well plate (Costar) for 5 h at 37°C in a humidified atmosphere containing 5% CO2 in air. Cells were harvested, stained with anti-CD4-TC and intracellular cytokine detection was performed as previously described 64. Briefly, cells were washed with washing buffer, resuspended in 1 mL of 4% paraformaldehyde and incubated (15 min, room temperature) with occasional stirring. Cells were washed once (1500×g, 4°C, 5 min) and resuspended in washing buffer. One million fixed cells were washed with 1 mL of DPBS-S (DPBS containing 10 mM HEPES, 1 mM CaCl2, 1 mM MgSO4, 0.1% saponin, 0.05% NaN3, 0.1% BSA) and incubated (30 min, 4°C) with 25 μL of DPBS-S/Milk (5% nonfat dry milk in DPBS-S cleared by centrifugation [15 000×g, 30 min]). Cells were centrifuged and incubated with anti-IL-10-PE mAb in DPBS-S/milk (30 min, 4°C), washed twice with DPBS-S, resuspended in DPBS and immediately analysed by FACS.

Cell sorting

Splenocytes from Foxp3EGFP mice were first enriched by positive selection using anti-CD4 Microbeads (Miltenyi Biotec) following manufacturer's instructions. The CD4 fraction from uninfected animals was irradiated (3000 rad) and used as feeder cells. The CD4+ fraction was stained with anti-CD4 and anti-CD25 mAbs. Treg and target cells were sorted using the CD4+Foxp3+ and CD4+Foxp3CD25 gates, respectively, and used immediately in suppression assays. Purity of each population was always ≥90%. For Treg-cell elimination, splenocytes from Foxp3EGFP mice were obtained and the EGFP population was sorted in a FACSAria and used immediately for proliferation assays. Purity of the EGFP population was always >99%.

Proliferation assay

CFSE staining was carried out as previously described with some modifications 62. Briefly, 2.5×107 cells/mL were stained with 2.5 μM CFSE (Molecular Probes) in DPBS (5 min, room temperature, in the dark) with occasional stirring. Staining was stopped with five volumes of DPBS containing 10% FCS; cells were centrifuged (5 min, 490×g), resuspended in complete RPMI medium and immediately used. CFSE-stained splenocytes (5×105 cells/mL) in 2 mL of complete medium were stimulated with 1 μg/mL Con A (Sigma) or 5 μg/mL LPS (Sigma) in each well of a 24-well plate (Costar). In some experiments, murine rIL-2 (20 U/mL, Roche) was added at the beginning of the culture. For IL-10 neutralization experiments, 30 μg/mL of anti-IL-10 (JES5-2A5, Biolegend) or control isotype mAbs (RTK2071, Biolegend) were added at the beginning of the culture and incubated for 30 min before stimulation. Seventy two hours later, cells were washed twice with buffer (1% FCS in DPBS) and stained with anti-CD4, anti-CD8 or anti-CD19 mAbs and 7-AAD.

Suppression assay

Fifty thousand target cells (CD4+Foxp3CD25) were seeded with 2.5×104 Treg cells (CD4+Foxp3+) and 2×105 feeder cells. Cells were stimulated with 1 μg/mL Con A in a final volume of 200 μL in triplicate wells of a 96-well flat bottom plate (Costar). Cells were pulsed with 0.5 μCi of [3H]-Thymidine (45 Ci/mmol, Amersham) for the last 18 h and were harvested onto glass-fiber filters using an automatic cell harvester. Radioactivity uptake was measured by scintillation spectroscopy on a LS6500 Multi-Purpose Scintillation Counter (Beckman) using Meltilex A solid scintillant (Wallac).

Determination of cytokines and equation image

IL-2 and IL-10 quantification was carried out by ELISA following manufacturer's instructions (Biolegend); detection limits were 3.9 ng/mL for IL-2 and 31.25 ng/mL for IL-10. NO2 determination was carried out by the Griess assay as described 40 with some modifications. Briefly, 100 μL of each sample was added to each well of a 96-well plate in duplicate, 50 μL of 1% sulfanilamide (Sigma) in 2.5% H3PO4 was added and incubated for 5 min, 50 μL of 0.1% naphtylenediamine dihydrochloride (Sigma) was added and incubated for 10 min (room temperature, in the dark); absorbance was read at 540 nm. Standard curves were prepared with sodium nitrite and the detection limit was 1.56 μM.

Statistical analysis

Statistical differences between groups were determined by the unpaired two-tailed Student t-test or One-Way ANOVA with Dunnett's or Bonferroni's Multiple Comparison tests using the PRISM software (GraphPad).

Acknowledgements

This work was supported by grants IN-200608 and IN-209111 from PAPIIT (DGAPA, UNAM, Mexico) and by grants 79775, 102399 and 102984 from CONACYT (Mexico). The authors are grateful to M. V. Z. Georgina Díaz and M. V. Z. Jorge Omar García for their expert advice and help in the care of the animals and Katharine A. Muirhead for helpful advices on cell tracking dyes. E. P. T. is recipient of a PhD fellowship from CONACYT (Registro 199991). This work was performed in partial fulfillment of the requirements for the PhD Program of Doctorado en Ciencias Biomédicas of E. P. T. at the Universidad Nacional Autónoma de México.

Conflict of interest: The authors have declared no financial or commercial conflict of interest.

Ancillary