Immunopathology in schistosomiasis is controlled by antigen-specific regulatory T cells primed in the presence of TLR2

Authors


Abstract

Regulatory T cells (Treg) are vital in maintaining the homeostasis of immune reactions. In chronic infections, such as schistosomiasis, it remains unclear whether engagement of the TLR family is required to induce Treg activity. Thus, we performed in vivo studies using TLR2–/– mice infected with Schistosoma mansoni and found elevated immunopathology, decreased egg burden and extended antigen-specific Th1 responses. Simultaneously, the population of Treg failed to expand. To evaluate the role of Treg during infection, we functionally inactivated CD4+CD25+ T cells and observed that the resulting immunopathology mirrored that in TLR2–/– mice. Egg burden was also reduced in anti-CD25-treated mice, indicating that without Treg eggs are more efficiently destroyed. In addition, antigen-specific T cells from both TLR2–/– and anti-CD25-treated mice displayed an extended Th1 phase. Finally, adoptive transfer of schistosome-primed, but not naive CD4+CD25+ T cells was able to resolve the immunopathology in TLR2–/– recipients and interestingly, this amelioration was independent of TLR2 being present on the transferred Treg. We conclude that TLR2 is necessary for priming active Treg and their expansion during schistosomiasis.

Abbreviation:
SEA:

schistosoma-egg-antigen

Introduction

Human schistosomiasis results from a chronic helminth infection and remains a severe morbidity factor in endemic tropical areas such as Africa and South America. Morbidity mainly arises from the immunopathology in the host's liver and gut. This stems from granulomatous, immune-mediated responses to eggs that are released by fecund females residing in the venous system of those organs 1. Clinically, patients suffer from liver fibrosis, portal hypertension, and intestinal bleeding but generally, the inflammation remains finely balanced throughout the course of the disease, presumably to ensure parasite survival. The induction of immune responses against the parasite involves both the innate and adaptive immune system in which the latter primarily involves MHC class II-restricted CD4+ T cells specific for schistosomal antigens. Even though the chronic phase of infection reflects a state of immune suppression, the adaptive immune reaction is distinguished by an initial Th1 response that is subsequently replaced by a much stronger Th2 response.

The importance of Th2 cytokines has been demonstrated in S. mansoni-infected IL-10-deficient mice, since they were unable to dampen Th1 responses and suffered from increased mortality 2, 3. In correlation, the role of IL-10- or TGF-β-secreting Treg in maintaining the balance of inflammation has been shown by the adoptive transfer of CD4+CD25+ T cells into RAG–/– mice and retroviral Foxp3 gene transfer experiments 46. Other studies have also shown a functional requirement of Treg in maintaining the homeostasis of immune responses to several microbial infections 7. For example, pathogen-specific Treg have been shown to down-regulate the Th1 response towards Bordetella pertussis and Leishmania major, leading to either prevention or to retardation of pathogen eradication 8, 9.

However, the question remains whether or not the innate immune system is required for the induction of regulatory mechanisms during parasitic infections. We have recently shown that MyD88, a central adaptor molecule of the innate immune system acting downstream from most TLR, is indeed involved in schistosomal recognition: the lack of this molecule resulted in diminished immunopathology, loss of Th1 responses and elevated IL-10 secretion 10. Although the role of TLR in vivo remains unclear, in vitro evidence has demonstrated that these receptors can be triggered directly by schistosomal components. For example, DC can be triggered via TLR2 with phospatidylserine (lyso-PS) from S. mansoni eggs, and were shown to induce IL-10-producing T cells that had a regulatory phenotype 11, 12. Even though TLR signaling has always been associated with the induction of Th1 responses 13, other infectious systems are demonstrating that signaling through TLR2 may also induce Th2 responses 14, 15 or Treg 16. In correlation to this changing dogma, in vivo treatment with LP40, a TLR2 ligand, was shown to significantly decrease granuloma size in the S. mansoni-egg-induced lung model 17.

Because these studies implicated a connection between triggering of TLR2 and the induction of Treg, we began to investigate the pathological phenotype in vivo and immunological responses in TLR2–/– mice infected with schistosomiasis. After finding an enhanced immunopathology and reduced Treg cell population in these mice, we broadened our study to investigate how the depletion of CD4+CD25+ T cells altered the course of natural infection and whether adoptive transfer of schistosome-primed Treg rescued the immunopathology in TLR2-deficient mice.

Results

Schistosome-infected TLR2–/– mice display aggravated immunopathology

Histological preparations from the livers of infected TLR2–/– mice (Fig. 1B) revealed a more pronounced immunopathology than in wild-type C3H controls (Fig. 1A). Fig. 1E shows that in comparison to wild-type mice, the average granuloma size in infected TLR2–/– mice was significantly increased. Since the C3H strain is more susceptible to S. mansoni infection, mice were already in the chronic stage by the 8th week of infection. To evaluate the level of fibrosis we prepared histological sections using Masson's staining. Although granulomas in TLR2–/– mice (Fig. 1D) were much larger than in wild-type controls (Fig. 1C), the intensity of blue in the collagen fibers appeared less. We further verified this by quantitative measurement of soluble collagen in individual livers (Fig. 1F). In correlation with the reduced fibrosis, Th2 cytokine levels, including the major pro-fibrogenic cytokine IL-13, were also lower in liver homogenates (Fig. 1G) as was the percentage of eosinophils in granulomas of infected TLR2–/– mice (data not shown). Together, S. mansoni-infected TLR2–/– mice suffer from a greater immunopathology, which is linked to an increased Th1, as shown by IFN-γ, and decreased Th2 response.

Figure 1.

Lack of TLR2 aggravates schistosome-induced immunopathology. Liver sections from C3H-infected wild-type (A, C) or TLR2–/– mice (B, D) were stained with H&E (A, B) or Masson's Blue (C, D); magnification was x100. The average granuloma size was determined in each mouse (E). Collagen and cytokine levels were measured in individual liver samples using the Biocolor assay kit or ELISA, respectively (F and G). Bars represent the arithmetic mean and SD from two infections comprising eight mice/group. Asterisks indicate significant differences (*p <0.05; **p <0.01; ***p <0.001).

TLR2 mediates Th2 development and expansion of primed CD4+CD25+ T cells

To evaluate whether the decreased Th2 cytokine expression and enlarged granulomas in infected TLR2–/– mice resulted from the cytokine milieu of primed T cells, we measured Ag-specific T cell responses ex vivo. In brief, T cells from infected mice were co-incubated with SEA and naive APC from wild-type mice. Mirroring the cytokine activity in the liver, both IL-10 (Fig. 2B) and IL-13 (Fig. 2C) responses were significantly higher by T cells from infected wild-type mice whereas T cells from TLR2–/– mice produced more IFN-γ (Fig. 2A). In terms of antigen presentation, the ability of wild-type T cells to respond to SEA was reduced when naive APC from TLR2–/– mice were used (Fig. 2D). From the long lasting Th1 responses found within the effector T cell pool in TLR2–/– mice we hypothesized, and confirmed below using anti-CD25-treated mice, that this phenomenon was due to a lack of schistosomal-specific Treg expansion (Fig. 2E and 4).

Figure 2.

T cells from Infected-TLR2–/– mice retain their Th1 response and display reduced numbers of CD4+CD25+ T cells. T cells from infected mice were cultured in vitro with SEA and naive irradiated APC from wild-type mice. Culture supernatant was analyzed for IFN-γ (A), IL-10 (B) and IL-13 (C) by ELISA. SEA-specific T cells from wild-type mice were also stimulated using naive APC from either wild-type or TLR2–/– mice (D). Absolute Treg populations in the spleens of individually infected or non-infected mice were determined by flow cytometry (E and F). Bars show the arithmetic mean ± SD of two pooled experiments. Asterisks indicate significant differences (*p <0.05; **p <0.01; ***p <0.001).

We reasoned that the uncontrolled inflammation in infected TLR2–/– mice might be due to a regulatory cell deficit and indeed the populations of CD4+CD25+ Treg was reduced by 50% in the spleen (Fig. 2E) and MLN (data not shown) in TLR2–/– mice; as also reflected in the decreased CD4+Foxp3+ population. Moreover, this phenomenon was due to a failure of Treg to expand during infection, as in naive mice the initial populations were equal. Fluctuations in other Treg markers were not as prominent in infected TLR2–/– mice (Fig. 2F) and concerning CD103+ Treg cells, this correlates with recent findings that also showed only a modest increase in this population in the spleen during infection of wild-type mice 18. Thus, lack of TLR2 during the expansion phase of Treg results in an unbalanced homeostasis between effector and Treg and leads to uncontrolled granuloma development.

Inactivation of CD25+ T-cells results in uncontrolled immunopathology

To establish the role of primed CD4+CD25+ T cells during schistosomiasis we studied the ensuing immunopathology in mice that had functionally inactivated Treg 19. After only 5 weeks of infection, granuloma development had already begun in the anti-CD25-treated group and by the 8th week, granulomas were significantly larger (Fig. 3E), indicating the necessity of CD25+ T cells to control granuloma development. In an additional group of control mice that were not infected, treatment with anti-CD25 did not result in any cellular infiltration or pathological abnormalities (data not shown and 18).

Figure 3.

Lack of CD25 results in uncontrolled immunopathology. Liver sections from C3H-infected control (A, C) or anti-CD25-treated mice (B, D) were stained with H&E (A, B) or Masson's Blue (C, D); magnification was x100. The average granuloma size was determined in each mouse at the indicated time points (E). Collagen and cytokine levels were measured in individual liver samples using the Biocolor assay kit or ELISA (F and G). Real-time TaqMan PCR results are presented as copies per 106 β-actin copies (H). Bars represent the arithmetic mean and SD from four infections comprising ten mice/group. Asterisks indicate significant differences (*p <0.05; **p <0.01; ***p <0.001).

The severe immunopathology in anti-CD25-treated mice strongly resembled that in infected TLR2–/– mice (Fig. 1B). Both staining procedures show that the granulomas formed in the anti-CD25-treated mice (Fig. 3B and D) were significantly larger and showed signs of central necrosis. As observed in TLR2–/– mice (Fig. 1D), the appearance of viable eggs in the granulomas was also reduced and in the Masson-stained sections the complete destruction of the egg was apparent (Fig. 3D). Moreover, the visible collagen content in and around the granuloma was also less than that in controls (Fig. 3C), indicating a lower level of fibrosis. This was confirmed by measuring collagen levels in individual liver samples (Fig. 3F). In Fig. 3G we also illustrate that anti-CD25-treated mice express lower levels of Th2 cytokines, which corresponds to the reduced levels of collagen and increased Th1 responses in both homogenates and on the mRNA level in anti-CD25-treated mice (Fig. 3H). All these areas of analysis implicate that the lack of CD25+ cells or TLR2 produces similar deviated immune responses.

Active CD25+ T cells allow the development of subtle Th responses

Administration of anti-CD25 during infection prevented the usual expansion of CD4+CD25+ T cells in schistosomiasis (Fig. 4D and for comparison Fig. 2E). Interestingly, expression of Foxp3 and CD25 (Fig. 3H) was elevated in the livers of these mice indicating that the immune system was continually trying to induce regulatory mechanisms in functionally inactive cells as observed previously 19. We also evaluated the effect this treatment had on other Treg cell markers and both CD4+CD25+CD103+ and CD4+CD25+CD45RBlow populations were reduced (Fig. 4D). Since CD4+ T cells are vital for granuloma development, we investigated how effector CD4+ T cells from the spleen and MLN (data not shown) responded to SEA during the course of infection. Fig. 4A–C show the Ag-specific response of CD4+ T cells from the spleen after 3, 5, 7 and 8 weeks of infection. Surprisingly, after only 3 weeks CD4+ T cells from anti-CD25-treated mice were already producing IFN-γ (Fig. 4A). In the Th2 phase, CD4+ T cells from anti-CD25-treated mice were still producing IFN-γ and simultaneously responded more vigorously with IL-10 (Fig. 4B) and IL-13 (Fig. 4C). Thus, CD4+ T cells produce elevated IFN-γ and Th2 cytokines when exposed to S. mansoni and treated with anti-CD25 antibody. Interestingly, studies performed in the S. mansoni egg model showed increased frequencies of CD4+ T cells producing Th1 and Th2 cytokines after anti-CD25 treatment 19, 20. These data suggest that CD4+ T cells from control-infected mice respond less erratically than those from anti-CD25-treated mice, a conclusion clearly reflected in the granuloma development.

Figure 4.

Absence of CD25+ cells creates elevated Th responses. Responder CD4+ T cells from infected mice were cultured in vitro with SEA and naive APC. Thereafter, culture supernatant was analyzed for IFN-γ (A), IL-10 (B) and IL-13 (C) by ELISA. Graphs depict the changes in Th response at the indicated time points. Spleen cells from individual mice of either infected group were stained with combinations of anti-CD4 and Treg cell markers. After flow cytometric analysis, the absolute numbers of the depicted Treg cell populations was determined (D). Bars show the arithmetic mean ± SD. Asterisks indicate significant differences (**p <0.01; ***p <0.001).

Absence of Treg enhances egg destruction by the host

We verified the visible reduction of eggs in the histological sections of TLR2–/– and anti-CD25-treated mice by analyzing egg numbers in the liver, intestine and stool (Table 1). Since adult worm burden was comparable in all infected groups, eggs appear to be more efficiently cleared by the immune system of TLR2–/– and anti-CD25-treated animals.

Table 1. Parasite burden, organ weight increase and liver damage parameters
C3HC57BL/6C3HC57BLl/6
Cont.Anti-CD25Cont.Anti-CD25WTTLR2–/–WTTLR2–/–
Eggs/liver (x103)25.413.919.59.419.611.921.67.8
Eggs/g intestine (x103)24.06.716.78.38.54.9
Eggs/g stool11.65.611.85.749.512.4516.25.7
Worm burden37.932.420.918.2n.tn.t6.47.6

Priming of Treg in TLR2–/– mice benefits both host and parasite

Next, we attempted to re-attain the Treg homeostasis in infected TLR2–/– mice by adoptively transferring CD4+CD25+ T cells from wild-type donors primed with SEA. Fig. 5A and B show the purity of these cells after separation and the absolute numbers of Treg beforehand. Fig. 5C and D depict the ability of effector CD4+ T cells from SEA-primed mice, but not naive mice, to respond to SEA ex vivo and the anergic state of Treg from SEA-primed mice when cultured under the same conditions. Moreover, Treg from SEA-primed mice were able to suppress Ag-specific responses of CD4+CD25 T cells from infected mice whereas naive Treg failed to do so, clearly showing a level of antigen-specificity on the part of the former Treg.

Figure 5.

CD4+CD25+ T cells from SEA-primed mice suppress antigen-specific T cell responses. Purity of CD4+CD25+ T cells from SEA-primed mice prior to transfer (A). Treg populations in SEA-primed mice prior to purification (B). Responder CD4+CD25 T cells from infected or naive mice were cultured in vitro with SEA, naive irradiated APC and Treg from SEA-primed or naive mice. Proliferation (C) and IL-10 levels (D) were measured via incorporation of [3H]thymidine 10 and ELISA, respectively. Bars show the arithmetic mean ± SD. Asterisks indicate significant differences (***p <0.001).

Fig. 6A shows granuloma development; in wild-type groups no differences could be calculated. In contrast, TLR2–/– recipients that received an adoptive transfer of Treg from SEA-primed donors showed a dramatic reduction in granuloma size when compared to control-infected TLR2–/– mice and recipients that received naive Treg. Interestingly, whereas the transfer of either donor Treg restored the homeostasis of CD4+CD25+ T cells in TLR2–/– recipients (Fig. 6B), neither significantly altered the Treg populations in wild-type recipient groups, indicating that superfluous Treg might be removed in order to maintain homeostasis. More in-depth investigations would be required to unravel this interestingly hypothesis. Whereas priming with SEA was not required for expansion of Treg in TLR2–/– recipients it was essential for the effects on both parasite burden and Ag-specific Th1 responses. The secretion of IFN-γ by T cells from TLR2–/– recipients that received SEA-primed Treg was significantly reduced when compared to recipients that received no cell transfer (Fig. 6C). In fact, the production now equaled those secreted by T cells from wild-type mice. Histologically, the formation and composition of granulomas in TLR2–/– recipients that received a transfer of SEA-primed Treg resembled those found in wild-type mice (data not shown). This included significantly more viable eggs within the granuloma and correlates to egg numbers found in the livers of individual recipients (Fig. 6D). Again, no significant differences could be detected between the wild-type recipient groups. Since egg numbers did not increase in TLR2–/– recipients that received naive Treg, one could speculate that Treg, which develop during the course of infection, are also skewed towards the parasite since "helminth-friendly" Treg ensure viable eggs.

Figure 6.

Transfer of SEA-primed CD4+CD25+ T cells reduces immunopathology and reverts Th response in infected TLR2–/– mice. CD4+CD25+ T cells (3 × 105) from SEA-primed or naive C3H/HEN donors were adoptively transferred into TLR2–/– or wild-type recipients. Recipient mice were then infected with S. mansoni. After 8 weeks, the following parameters were determined in recipient mice: Microscopic determination of granuloma size (A). Absolute number of CD4+CD25+ T cells in the spleens of individual mice (B). IFN-γ secretion from responder T cells after incubation with wild-type naive APC and SEA (C). Total number of liver eggs in individually infected mice (D). Bars show the arithmetic mean ± SD from two pooled transfer experiments. Asterisks indicate significant differences (*p <0.05; **p <0.01; ***p <0.001).

Transferred Treg do not require TLR2 to control immunopathology

Finally, we tested the ability of CD4+CD25+ T cells from S. mansoni-infected mice in restoring granuloma formation in TLR2–/– mice and simultaneously addressed whether the phenomenon of enlarged granulomas in TLR2–/– mice was restricted to the C3H mouse strain. Thus, C57BL/6 and TLR2–/– donor mice were infected and then cured with praziquantel to ensure that recently activated CD25+ T cells (which only transiently up-regulate CD25) were not transferred. Donor Treg were then isolated and adoptively transferred into recipient groups that were subsequently infected. Granulomas in TLR2–/– recipients that received no cell transfer were significantly larger than those from wild-type recipients (Fig. 7A). After adoptive cell transfer of TLR2–/– or wild-type Treg, granuloma size in TLR2–/– recipients was significantly reduced when compared to control TLR2–/– recipients. All recipient groups that obtained a cell transfer had elevated numbers of CD4+CD25+ T cells in the spleen (Fig. 7B) and MLN (data not shown). Interestingly, the transfer of primed Treg from wild-type donor mice significantly increased the amount of Treg in both the wild-type and TLR2–/– recipients, indicating that in contrast to naive Treg (Fig. 6B), which were eliminated in the recipient, Treg with an activated or altered specificity are retained. The transfer of Treg from TLR2–/– donors significantly elevated the number of CD4+CD25+ T cells in TLR2–/– recipients, demonstrating the connection between Treg cell number and immunopathology.

Figure 7.

Adoptive transfer of infection-primed CD4+CD25+ T cells reduces immunopathology and reverts Th response in infected TLR2–/– mice. C57BL/6 wild-type and TLR2–/– donor mice were cured of schistosomiasis after repeated praziquantel treatment. Upon verification (stool analysis of eggs) that no schistosome infection remained, CD4+CD25+ T cells (3 × 105) from these mice were adoptively transferred into TLR2–/– or wild-type recipients. Recipient mice were then infected with S. mansoni. After 9 weeks, the following parameters were determined in recipient mice: Microscopic determination of granuloma size (A). Absolute number of CD4+CD25+ T cells in the spleens of individual mice (B). Total number of liver eggs in individually infected mice (C). Bars show the arithmetic mean ± SD from two pooled transfer experiments. Asterisks indicate significant differences (*p <0.05; ***p <0.001).

In parasitological terms, TLR2–/– mice with no cell transfer had fewer eggs (Fig. 7C), a finding that correlates with the infected C3H strain (Fig. 6D). No significant differences in the amount of eggs could be observed between wild-type recipient groups. In contrast, elevations in egg numbers could only be observed in TLR2–/– recipients that received a transfer of CD4+CD25+ T cells from wild-type donors. Thus, it would appear that Treg perform different tasks during schistosomiasis controlling both excessive granuloma formation and aiding the survival of parasite eggs.

Discussion

The first port of call during a parasite's entry into the host is the innate immune system. Even though members of the schistosome family have developed some rather deft cloaking devices, TLR do recognize certain schistosomal antigens 21, 22. It is becoming increasingly clear that these signals have been manipulated and tuned to favor the helminth's passage, development and fecundity. Our initial observations with an in vivo model of S. mansoni infection, demonstrated the importance of MyD88 in recognizing schistosomal antigens since its absence instigated a reduced immunopathology that was associated with elevated Th2 cytokine secretion by primed T cells 10. In this report, we primarily focused on the role played by TLR2 and discovered an almost contrary immunopathological phenotype, which constituted an overwhelming inflammation and enhanced Ag-specific Th1 response. Therefore, we analyzed regulatory T cell populations during infection and found that in TLR2–/– mice, CD4+CD25+ T cells failed to expand. We reasoned that this deficit could account for their altered immunopathology. To substantiate our conjecture we proceeded to study the nature and dynamics of induced Treg by silencing the CD25 marker in wild-type mice. By blocking the ability of CD4+CD25+ T cells to develop and function during infection, we induced the same immunopathology found in infected TLR2–/– mice. Furthermore, by adoptively transferring Treg that had been previously exposed to schistosomal antigens, we could clearly demonstrate that these cells regulate the immune responses in favor of both the host (influence on adaptive immune responses and granulomatous pathogenesis) and the parasite (egg survival).

This is the first study, which demonstrates that due to the lack of TLR2 stimulation CD4+CD25+ T cells fail to expand during S. mansoni infection, and in turn completely unbalances the immunopathology. This uncontrolled granuloma development was not initially expected because lack of MyD88–/– had resulted in diminished granuloma formation and indicates that other TLR, such as TLR4 23, are also involved in the recognition of schistosomal antigens. The expression of several TLR on Treg has been reported to influence both their function and expansion. For example, using bacteria lipoprotein, a TLR2 ligand, or Candida albicans it has been demonstrated that triggering TLR2 on intrinsic Treg promotes their proliferation 2426, a finding that correlates with the distinct lack of Treg expansion in schistosome-infected TLR2–/– mice (Fig. 2E). In addition to its dire effects on granuloma composition, the lack of TLR2 also extended the Th1 phase, which could be actively noted in the Ag-specific T cell responses (Fig. 2) and the reduced Th2 cytokines found in situ (Fig. 1). The composition and size of the granuloma are determined by the activity of Ag-specific CD4+ T cells and their cytokine milieu 27. Obviously, in the TLR2–/– mice infected here, the unregulated activity of IFN-γ-producing T cells encouraged the massive granuloma growth. In confirmation, infected IFN-γ-deficient mice show diminished granuloma size 28. Moreover, increased secretion of IFN-γ has anti-fibrotic actions 2 that strikingly correlate with our results. For the parasite, elevated IFN-γ also resulted in decreased egg burden and interestingly, studies in schistosomiasis have shown that a pronounced Th1 response induced increased NO-production by macrophages 1. It will be interesting to discriminate which mechanism is more efficient in reducing egg numbers.

Development, activation, cytokine stimulation, life span and antigen specificity are prime topics currently addressed in Treg cell research. During infection, we find that the ratio of effector CD4+CD25 and Treg remains homeostatic and a disturbance to this ratio elicit enhanced immunopathology. Within these studies, we also delved into how Ag-specific effector T cells respond during granuloma development. Under the "watchful eye" of CD4+CD25+ T cells, primed Th1 cells maintain their IFN-γ secretion until the 5th week of infection (Fig. 4A). At this time point eggs begin to travel into the intestine and there is a subtle switch from Th1 to Th2, a progression already shown to be under the regulation of Treg 5. When one compares the dynamics of the CD4+ T cells from normally infected mice to those without Treg control, it is obvious that Treg are necessary in maintaining the finely balanced process. The erratic nature of effector CD4+CD25 T cells and marked increases in Th1 and Th2 expand the findings observed in the S. mansoni-egg-induced model 18, 20 and indicate that Treg exert a feedback mechanism on both immune responses. Similar effects were noted in an OVA-immunization model in which the depletion of Treg in vivo elicited increased Ag-specific Th1 responses 29. In contrast, the reduced but still active Treg population in the infected TLR2–/– mice might provide sufficient regulatory action to dampen Th2 producing cells.

The presentation of schistosomal antigens by APC is critical for the development of granulomas since in the absence of MHC class II molecules there is only marginal immunopathology 30. We speculate that initial priming is also imperative for mechanisms in which the helminth controls the responses of the host's immune system. TLR2 stimulation on APC has already been shown to provide temporal suppression on Treg activities 26 and might provide an explanation for the residual Th1 responses we find during the chronic phase of both TLR2–/– and anti-CD25-treated mice. In vitro investigations have shown that when DC are matured with the schistosomal Ag lyso-PS, they can drive naive T cells into a regulatory phenotype, and interestingly this process was shown to be dependent on TLR2 being present on the APC 12. Moreover, using an alternatively structured form of lyso-PS, matured DC induced T cells with a non-regulatory Th2 phenotype 31. Speculation on the ability of schistosomal egg antigens to trigger TLR on DC has been hindered by the non-classical stimulation pattern of those cells (only slight up-regulation of MHC class II and no overt cytokine production) 32. This is incredibly interesting since transfection studies show enhanced stimulation of certain TLR (including TLR2) to schistsomal egg components 12, 21. Studies in our laboratory have confirmed these opposing stimulation properties (non-classical DC stimulation vs. transfection) and in addition, it appears that TLR2–/– DC are even less reactive than WT DC in terms of IL-6 secretion and up-regulation of MHC class II (data not shown). Mechanistically, these findings could be explained alongside the microarray studies performed on the interaction of DC and schistsomal antigens, which showed that suppression of TLR signaling could be altered 32. Thus, the studies performed here demonstrate that whereas TLR2 is necessary to induce Treg in vivo (see Fig. 2E) it is not required for the effector T cell response (Fig. 2A–C). Finally, in our ex vivo assays (Fig. 2D) we found a reduced Th cytokine secretion from wild-type T cells when incubated with naive APC from TLR2–/– mice, implying that TLR2 on APC is required for the ability of established Ag-specific effector T cells to respond to presented schistosomal antigens.

We demonstrate here that without correct priming Treg are unable to expand and control the immunopathological balance. Furthermore, we show that prior exposure of SEA to Treg aids both host and parasite since naive CD4+CD25+ T cells in TLR2–/– recipients failed to contain granuloma development and increase egg expulsion (Fig. 6). The complexity of inducing Treg, primed with schistosomal antigens, was shown using lyso-PS. This study found that by altering the number of acyl chains on the antigen they could no longer induce the cells with a regulatory phenotype 12. Thus, slight alterations in the biochemical components could severely alter the response of the host and in turn the course of immunopathology. Of course what actually comprises schistosomal egg antigens? They are a complex mixture of glycolipids and proteins and the activation of effector CD4+ T cells might be primarily established after the binding of self-proteins to schistosomal components. Chemically induced immune reactions have revealed that effector T cells can respond to altered self-peptides that arise after self-proteins covalently bind to or are structurally modified by chemicals 33, 34. Thus, CD4+ T cells responding to schistosoma-self-complexes could activate and recruit naturally occurring Treg cells that begin to suppress T cells responding to self-proteins. In fact, numerous autoantibodies have been detected in the sera of S. mansoni-infected patients 35. This hypothesis would explain why TLR2–/– recipients showed regulated granuloma development after the transfer of Treg from infected TLR2–/– donors (Fig. 7A). We believe that during infection, both schistosome-specific and self-reactive Treg are induced and importantly, in contrast to the former subset, the latter are activated in a TLR2-independent fashion and do not aid the parasite as shown by only marginal elevation of egg numbers in the TLR2–/– recipient (Fig. 7C).

From the literature that is accumulating on S. mansoni, it is becoming increasingly clear that small alterations to any part of the immune system, whether a missing receptor or cell population, leads to dramatic changes in the course of infection, some more disastrous to the host others a hindrance for the helminth. Thus, it is tempting to speculate that co-evolutionary development has allowed the development of Treg populations which are induced by both TLR2-independent and -dependent mechanisms and provide protection for host and parasite.

Materials and methods

Strains of mice

Female specific pathogen-free (spf) C3H/HEN and C57BL/6 mice were purchased from Harlan Winkelmann (Germany). Female TLR2–/– mice were from Amgen (South San Francisco, USA), in-house bred under spf conditions and backcrossed nine times onto the C3H or C57BL/6 background.

Experimental infection and parasite burden

Mice were infected with cercariae from a Brazilian strain of S. mansoni from Biomphalaria glabrata snails (Brazilian origin) and sacrificed after 8 to 9 weeks of infection. Parasite burden including eggs/organ or stool and adult worms were performed as previously described 10.

Antibodies and reagents

Microbeads, including the CD4+CD25+ isolation kit, and separation columns were from Miltenyi Biotech. All antibodies were purchased from BDbiosciences. Sterile schistosome-egg-antigen (SEA) was prepared according to standard protocols 10.

Histology

Paraffin embedded sections (3 µm) from the left liver lobe of each mouse were stained with H&E or Masson's Blue. Granuloma size was microscopically determined (Axioskop, Zeiss) after calculating the individual size of 40 granulomas/mouse. Each granuloma contained a single egg.

Collagen and cytokine levels in the liver

The collagen content of individual liver samples was colorimetrically determined using the SircolTM soluble collagen assay kit (Biocolor, Newtonabbey, Northern Ireland). Cytokine levels from homogenized liver samples were determined using ELISA kits (R&D Systems).

Th immune responses and suppression assay in vitro

T cell and APC separation

Splenic T cells from infected mice were enriched after incubation with either anti-CD90 (Thy1.2) or anti-CD4 microbeads. Naive APC were obtained with anti-MHC class II microbeads. The purity of both cell fractions was >97%.

T cell assay

Responder T cells (2 × 105) from infected groups were incubated for 72 h with 20 µg/mL SEA and 25-Gy irradiated naive APC (2 × 105).

Suppression assay

Responder CD4+CD25T cells (1 × 105) from infected mice or naïve mice were incubated for 96 h with 20 µg/mL SEA, naive irradiated APC and 1 × 105 Treg from SEA-primed or naive mice. Cytokine levels and proliferation were measured using ELISA kits or incorporation of [3H]thymidine, respectively. Experiments were performed in replicates of six and repeated at least twice to ensure reproducibility.

In vivo depletion of CD25+ T cells

Hybridoma PC61 was grown to over confluent levels in DMEM medium (PAA) supplemented with 10% Ultra low IgG FCS (PAN Biotech). Anti-CD25 antibody was then obtained using an IgG Sepharose column with the ÄKTAFPLCTM (Amersham Biosciences). Antibody concentration levels were determined by standard protein assays after concentrating the antibody using centripreps (Millipore GmbH, Schwalbach, Germany). Three days prior to infection, mice were injected with 1 mg of antibody i.p. A second mg of antibody was administered on day 35 to ensure that CD25+ cells remained silent during the entire length of infection.

Flow cytometry

Prior to staining, Fc receptors were blocked using anti-CD16/CD32. Thereafter, combination staining was performed with anti-CD4 mAb (PerCP-labeled), anti-CD25 mAb (APC- or PE-labeled), and PE-labeled anti-CD103, anti-CD152, anti-CD45RB mAb and intracellular PE-labeled anti-FoxP3 from eBioscience. Fluorescence was analyzed using a FACSCaliburTM flow cytometer and CELLQuestTM software (BDbiosciences).

Adoptive transfer of CD4+CD25+ T cells

SEA-primed

C3H/HEN donor mice were injected i.p. with 100 µg SEA or PBS. After 10 days CD4+CD25+ T cells from the spleen were isolated using the CD4+CD25+ isolation kit and injected i.v. (3 × 105) into TLR2–/– and wild-type recipients (six mice/group). Twenty-four hours later recipient mice were infected with S. mansoni.

Infection-primed

TLR2–/– and wild-type donor mice were cured of S .mansoni after repeated oral administration of praziquantel over 5 consecutive days via oral gavage (100 mg/kg body weight). Three weeks later CD4+CD25+ T cells from the spleens and MLN were isolated as described above and injected i.v. (3 × 105) into TLR2–/– and wild-type recipients (six mice/group). Twenty-four hours later recipient mice were infected with S. mansoni. From a >98% pure CD4+ T cell fraction, the purity of transferred CD4+CD25+ T cell fraction was 85%.

Real-time PCR and RT-PCR

Quantitative expression analysis from mouse tissue was performed by real-time quantitative TaqMan RT-PCR as described previously 36. In brief, RNA from homogenized liver tissue was isolated using peqGOLD TrifastTM (Peqlab Biotechnologies) reagent and transcribed into cDNA using Superscript II reverse transcriptase (Invitrogen). PCR was performed on the ABI PRISM 7700 Sequence Detection System (Perkin-Elmer). Cytokine mRNA copy numbers were normalized to β-actin copies. Primer and probe sequences are shown in Table 2. Probes were labeled with the reporter dye FAM at the 5′-, and the quencher dye TAMRA at the 3′-end.

Table 2. Sequences of primers and probes used for real-time quantitative RT-PCR
Forward primerReverse primerFluorogenic probe
β-Actin 5′-CGTGAAAAGATGACCCAGATCA-3′5′-CACAGCCTGGATGGCTACGT-3′5′-TTTGAGACCTTCAACACCCCAGCCA-3′
Foxp35′- TTCGAGGAGCCAGAAGAGTTTC –3′5′- GGGCCTTGCCTTTCTCATC –3′5′- CAAGCACTGCCAAGCAGATCATCTCCT –3′
IFN-γ5′- CAGCAACAGCAAGGCGAAA –3′5′- CTGGACCTGTGGGTTGTTGAC –3′5′- AGGATGCATTCATGAGTATTGCCAAGTTTGA –3′
CD255′- GAGTGAGACTTCCTGCCCCATA-3′5′- TCTCCGTCATTGCAGTTGTTTC-3′5′- CCACCACAGACTTCCCACAACCCAC-3′

Statistical analysis

Statistical differences were analyzed by either ANOVA or Student's t-test using GraphPad Prism software (San Diego, CA). Cytokine concentrations were determined using the standards incorporated into each assay and SigmaPlot (Systat Software).

Acknowledgements

The authors would like to thank Carsten Kirschning for breeding TLR2–/– mice and Roland Lang for critical reading of the manuscript. This work was supported by a grant from the German Research Council (DFG CO-457/1–1).

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

Ancillary