SEARCH

SEARCH BY CITATION

Keywords:

  • Foxp3;
  • Nematode;
  • Pathology;
  • Th2 response;
  • Treg

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References

Here, we show that Treg limit intestinal pathology during nematode infection and that they control the onset and magnitude of the anti-parasitic Th Th2 response. Using mice expressing the diptheria toxin receptor under the control of the foxp3 locus, we removed Foxp3+ Treg during the early phase of infection with Heligmosomoides polygyrus bakeri. Depletion of Treg in infected animals did not affect adult worm burden, but led to increased pathology at the site of infection. Infected, depleted mice displayed higher frequencies of activated CD4+ T cells and increased levels of the Th2 cytokines IL-4 and IL-13. The stronger parasite-specific Th2 response was accompanied by higher levels of IL-10. Only a moderate change in Th1 (IFN-γ) reactivity was detected in worm-infected, Treg-depleted mice. Furthermore, we detected an accelerated onset of parasite-specific Th2 and IL-10 responses in the transient absence of Foxp3+ Treg. However, adult worm burdens were not affected by the increased Th2-reactivity in Treg-depleted mice. Hence, our data show that Treg restrict the onset and strength of Th2 responses during intestinal worm infection, while increasing primary Th2 responses does not necessarily lead to killing of larvae or accelerated expulsion of adult worms.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References

Parasitic worms are the most potent inducers of highly polarized Th cell type 2 (Th2) responses. Helminth infections tend to be long lasting and are often associated with insufficient immunity to re-infection 1, 2. Chronic infections favor immunosuppression, the so-called modulated Th2-response, associated with poor parasite-specific reactivity alongside with the strong production of anti-inflammatory cytokines, such as IL-10 and TGF-β 1. This helminth-induced immunosuppression may spill over to bystander antigens 3, downmodulate reactivity against other pathogens 4 and impair vaccination efficacy 5. However, such effects can be beneficial by downregulating inflammatory reactions to allergens and inflammatory disorders of the intestinal tract. Studies in animal models showed that helminths can suppress airway hyperreactivity and colitis 6–8 and nematodes have been used to efficiently treat human inflammatory bowel disease in clinical trials 9, 10.

In animal models, the beneficial effects of nematode infections on inflammatory disorders were associated with the presence of Treg populations 6, 11, 12. However, it is not yet clear whether such Treg suppress anti-parasite immune reactivity and thus contribute to survival and well-being of the worms, or whether they predominantly dampen immunopathology. Several studies investigated the role of Treg in infections with helminths or other parasites and yielded in diverging conclusions (reviewed in 1 and 13). There is evidence that certain parasites depend on Treg action for their persistence 14–17. Other species profit more indirectly from Treg circuits, which control infection-induced pathology 18–20.

In a previous study we observed quantitative and qualitative changes in Treg subsets in mice infected with Heligmosomoides polygyrus bakeri. Treg derived from infected mice displayed an activated phenotype, parasite-specificity and an extraordinary strong suppressive activity on conventional T cells 21. Therefore, we hypothesized that the potentially harmful Th2 reactivity is controlled by Treg in infections with H. p. bakeri and aimed to clarify whether the removal of Treg has an effect on worm establishment. We addressed these questions via the specific depletion of Treg in H. p. bakeri-infected mice. To date, Treg depletion or inactivation in parasitic infections was achieved by application of Ab against the surface marker CD25 18, 19, 22–24, but there is evidence for an incomplete depletion of Treg by anti-CD25 Ab treatment 22, 25. Furthermore, not only Treg, but also activated T cells transiently express CD25 and may thereby be affected. This has recently been shown in acute Toxoplasma gondii infection, where anti-CD25 Ab treatment abrogated effector T cell responses 26. Here, we analyzed the role of Treg by using a transgenic mouse model permitting the selective depletion of Foxp3+ cells with preserving effector T cells. DEREG (depletion of Treg) mice express a fusion protein of enhanced GFP and the receptor for diphtheria toxin (DT) under the control of the foxp3 locus, allowing both detection and depletion of total Foxp3+ cells (CD25+ and CD25) without negatively affecting CD25+ effector cells 27. Application of DT during the early phase of infection with H. p. bakeri led to the complete, transient depletion of Treg, entailing increased frequencies of activated T cells and stronger Th2 cytokine responses. We show that Treg depletion during H. p. bakeri infection had no effect on the establishment of the worms, but led to dysregulated inflammatory reactions in the parasitized intestine. This suggests that the predominant role of Treg during intestinal nematode infection is the control of excessive pathology.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References

Analysis of Treg after infection with H. p. bakeri

We have previously analyzed changes in Treg numbers in gut-draining MLN and the intestines of H. p. bakeri-infected BALB/c mice 21. As the current study was performed with animals of the C57BL/6 background, and inbred mouse strains differ in susceptibility and immune responses to H. p. bakeri28, we first analyzed the composition of Treg subsets in these animals. We found that C57BL/6 mice infected with H. p. bakeri had elevated frequencies of CD4+CD25+Foxp3+ Treg cells with an activated/effector-memory phenotype in MLN (Fig. 1A). These effector/memory-like Treg (e/mTreg) are characterized by the expression of the integrin αE(CD103)β729. The increase of these e/mTreg was first detectable in MLN in the early acute phase at day 6 post infection (p.i.) (Table 1), while the larvae undergo a histotrophic phase in the small intestinal tissue. We detected a disproportionately high increase in absolute numbers of naïve-like (CD103) and e/m-like (CD103+) CD25+Foxp3+ Treg in the MLN at days 6 and 14 post H. p. bakeri infection (Table 1). In addition to the MLN, the small intestine was analyzed for changes in the numbers of Foxp3+ Treg (Fig. 1B). Similar to previous data from BALB/c mice 21, C57BL/ 6 mice showed a transient increase in small intestinal Treg numbers, with a peak around day 6 p.i. (Fig. 1B). The maximum is in accordance with the development of an inflammatory infiltrate around developing larvae (Fig. 3A). In the chronic phase, small intestinal Treg numbers decreased (Fig. 1B). Hence, infection with H. p. bakeri led to local quantitative and qualitative differences within the Treg population. We therefore hypothesized that depletion of Treg early during infection might interfere with regulatory circuits and lead to changes with respect to parasite establishment and/or the hosts' immune response

thumbnail image

Figure 1. Treg analysis in MLN and small intestine of H. p. bakeri colonized C57BL/6 mice. (A) CD4+ T cells from MLN were analyzed for the expression of CD25 and CD103. Plots are representative for groups of five naïve and acutely H.p.b.-infected (14 d.p.i.) mice. The proportion of CD4+CD25+CD103+ effector/memory-like (e/m) Treg is depicted in the upper right quadrant, naïve-like CD4+CD25+CD103 Treg are gated in the upper left quadrant. Histograms show Foxp3 expression in CD4+CD25+CD103+ (black line) and CD4+CD25CD103 (filled gray) T cells at day 14 p.i. The percentage of Foxp3+ cells in the CD25+CD103+ population is depicted. (B) Counts of Foxp3+ cells in duodenal cross sections. Mean±SEM of five mice per group is shown. Data are representative for two to three independent experiments with four to seven mice per group. *p<0.05, compared with naïve controls, Mann–Whitney test.

Download figure to PowerPoint

thumbnail image

Figure 2. Treg depletion in peripheral blood and MLN. (A) Peripheral blood lymphocytes were gated according to forward and side scatter properties and analyzed for CD4+ T cells (left panel), Foxp3 and GFP expression by CD4+ cells (mid panel) and CD25 and GFP expression by CD4+ cells (right panel). Representative plots from infected PBS treated (upper row) or DT-treated mice are shown. (B) Representative histograms for CD25 expression by GFP+ and GFPFoxp3+ cells in peripheral blood derived from an infected DEREG animal after four rounds of depletion (day 11 p.i., black line) and a PBS-treated, infected DEREG control (filled gray). CD25 MFI±SEM of six mice per group is indicated. (C) CD4+ MLNC stained for Foxp3 and analyzed for GFP expression at day 7 or 14. Plots are representative for four to six mice per group. (D) Percentages of Foxp3+ cells in MLN at days 7 and 14. Mean+SEM of four to six mice per group is shown. *p<0.05, **p<0.005, ns: no significant difference; Mann–Whitney test. One representative result from two independent experiments is shown. H.p.b.: H. p. bakeri-infected, WT: WT littermates.

Download figure to PowerPoint

thumbnail image

Figure 3. Treg depletion at the site of infection, intestinal pathology and adult worm burdens. (A) Foxp3+ cells were stained in duodenal cross sections on day 7. Representative pictures for an infected WT control (left) and a Treg-depleted, infected DEREG animal (right) are shown. Magnification 200×(upper panel) and 400×(lower panel). Red arrows: developing H.p.b. larvae. Black arrows: Foxp3+ cells. Mean+SEM of Foxp3+ cell numbers in high power fields (HPF, 400×magnification) of four mice per group are depicted as bar graph. (B) Foxp3+ numbers in HPF of mice at day 14 (five to six mice per group). (C) H&E stained duodenal cross sections used for histological scoring of intestinal inflammation. Red arrows: H.p.b. adults. Black arrows: severe inflammatory cell infiltration. Note severely disturbed tissue architecture and blunted appearance of villi as well as altered ratio of villus length to crypt depth in the depleted, H.p.b.-infected DEREG animal. (D) Mean+SEM of inflammatory scores of mice at day 14 (five to six mice per group). (E) Counts of adult worms recovered from the intestinal lumen 2 wk p.i. Individual counts and mean±SEM are shown. One representative result from two independent experiments is shown. *p<0.05, **p<0.005, ns: no significant difference; Mann–Whitney test.

Download figure to PowerPoint

Table 1. Kinetic of total, CD4+ and Treg cell numbers in MLN during H. p. bakeri infectiona)
 Naïve6 d.p.i.14 d.p.i.28 d.p.i.
  • a)

    a) *p<0.05, **p<0.01 compared with naïve; bold numbers indicate fold increase over naïve. nTreg: naïve-like Treg.

MLN total cell number9.36×106±1.4518.91×106±1.25**24.86×106±3.92*15.78×106±3.68
  2.02±0.132.66±0.421.69±0.39
Total CD4+2.7×106±0.415.98×106±0.45**6.79×106±1.22*3.79×106±0.92
  2.16±0.162.45±0.441.37±0.33
Total CD4+CD25+CD103-Foxp3+1.06×105±0.143.47×105±0.41**5.24×105±1.12**2.62×105±0.65
(nTreg) 3.27±0.384.95±1.0692.47±0.62
Total CD4+CD25+CD103+Foxp3+0.38×105±0.050.99×105±0.1**1.84×105±0.35**1.08×105±0.29
(e/mTreg) 2.64±0.264.89±0.932.87±0.77

Foxp3+ Treg depletion is highly efficient but transient in peripheral blood and lymphatic organs

Transgenic DEREG mice expressing GFP and the receptor for DT under the control of the foxp3 locus 27 were infected with H. p. bakeri and subjected to DT treatment according to the scheme described in the Material and methods. We have chosen day 4 p.i. to begin with Treg depletion, as this precedes the peak in intestinal Treg numbers and the phenotypic changes of Treg in MLN (Fig. 1). Naïve DEREG mice and infected or naïve WT littermates served as controls.

To survey efficiency of depletion, mice were bled on days 5, 9 and 11 p.i. Peripheral blood cells from PBS-treated mice served as controls. Figure 2A shows that virtually all GFP+Foxp3+ Treg were removed from peripheral blood on day 5 p.i. (after the first round of depletion). As untreated DEREG animals generally show a small proportion of GFPDTR Foxp3+ Treg, cells were always counterstained for Foxp3+ protein at all time points. DT-treated animals showed a rapid replacement of depleted GFP+DTR+Foxp3+ cells by GFPDTRFoxp3+ cells detected on day 9 p.i. (after third depletion) due to the outgrowth of the pre-existing small GFPDTR Treg pool (Fig. 2A). Still, the reappearing Treg clearly expressed lower levels of the IL-2 receptor alpha chain (CD25) as measured by cellular MFI, possibly arguing for an immature and not fully functional phenotype of these cells (Fig. 2B). Similarly, MLN cells (MLNC) from DT-treated animals showed drastically reduced Treg frequencies at day 7 p.i. (after second depletion), with a partial replacement by GFPDTRFoxp3+ cells (Fig. 2C and D). Animals dissected at day 14 p.i. had partially restored their MLN Treg pool with GFPDTRFoxp3+ cells to levels still remarkably lower than undepleted controls (Fig. 2C and D). Similar data were obtained for spleen cells (data not shown).

Hence, the removal of GFP+DTR+Foxp3+ Treg was highly efficient with an almost complete depletion of Foxp3+ cells during the early phase of infection. As DEREG animals are able to replace the removed cells by expanding a small proportion of GFPDTRFoxp3+ Treg, the model is not applicable for prolonged Treg removal.

Treg depletion aggravates parasite-associated pathology and has no influence on worm establishment

As we were interested in the involvement of Treg in controlling intestinal pathology in response to infection with H. p. bakeri, we analyzed the efficiency of Treg depletion at the infected site. To this end, intestinal sections were analyzed for the numbers of Foxp3+ cells at two time points after infection. Depleted infected and depleted naïve DEREG mice had Foxp3+ cell numbers reduced by 97% compared with WT infected control animals at day 7 p.i. (Fig. 3A). At day 14 p.i., uninfected depleted DEREG mice had replaced Treg to comparable levels as seen in infected WT littermates (Fig. 3B). Mice infected with H. p. bakeri still showed a significant reduction in intestinal Treg numbers (Fig. 3B).

When analyzing signs of intestinal pathology by scoring changes in the architecture of the small intestinal epithelium and inflammatory infiltrates, we found that infected, depleted DEREG animals developed significantly stronger pathological scores than uninfected, depleted DEREG mice or infected WT littermate animals (Fig. 3C and D). Pathology occurred as villous blunting and atrophy, crypt hyperplasia and formation of cellular infiltrates in the lamina propria (Fig. 3C). Uninfected, depleted DEREG controls had only low signs of inflammation, showing that the transient Treg depletion in naïve animals did not provoke major intestinal reactions.

To investigate whether removal of Treg had an influence on parasite development or fitness, we analyzed adult worm burdens and could demonstrate that the transient depletion of Treg had no effect on the establishment of infection (Fig. 3E). All groups showed similar adult worm burdens at day 14 p.i. Worm fitness was surveyed by quantifying the egg production by female worms kept in culture after dissection. No differences were detectable (data not shown). Comparable worm burdens and fecundity in mice depleted of Treg strengthens the view that these cells are predominantly controlling pathology, while not being essentially required for permitting successful establishment of worm infection.

Treg depletion does not interfere with physiological changes at the site of infection

We also investigated changes in the gut epithelium in response to infection and Treg depletion by detection of cells undergoing mitosis and assessing goblet cell numbers. Thereto, small intestinal sections were treated with an Ab against Ki-67, a marker for proliferating cells (Fig. 4A). We found a trend of more epithelial cells undergoing mitosis in small intestinal crypts of infected WT mice compared with naïve WT controls and a significant increase in Ki-67+ cells was detected in infected, depleted DEREG mice compared with uninfected DEREG controls. However, the counts of Ki-67+ cells in infected WT controls and infected, depleted DEREG mice were similar (Fig. 4A, B). Similarly, we detected a significant increase in small intestinal mucus-producing goblet cells in response to H. p. bakeri infection compared with naïve controls and this was not altered by Treg-depletion (Fig. 4C).

thumbnail image

Figure 4. Epithelial cell proliferation and goblet cell counts in the small intestine (A) Ki-67 stain (red) in cross sections for the detection of cells undergoing mitosis. Magnification 100×(upper panel) and 400×(lower panel). (B) Mean+SEM of Ki-67+ cells in small intestinal crypts (five to six mice per group) at day 14. (C) Mean+SEM of goblet cells per small intestinal crypt. One representative result from two independent experiments is shown. *p<0.05, **p<0.005, ns: no significant difference; Mann–Whitney test.

Download figure to PowerPoint

Taken together, these data clearly show that the transient depletion of Treg during infection with an intestinal nematode leads to increased inflammation and pathology at the site of infection, while physiological responses to the worm infection (epithelial cell hyperproliferation and goblet cell hyperplasia) were not altered by Treg depletion.

Depletion of Foxp3+ Treg leads to stronger cytokine responses and more activated CD4+ T cells

Having shown that the selective depletion of Treg led to an increased pathology and did not change worm burdens, we next examined whether the removal of Treg had an impact on the cytokine response. MLNC and splenocytes were stimulated with mitogen or adult worm soluble antigens to measure the general immune status and the parasite-specific immune response, respectively. We found a consistent pattern of significantly higher levels of Th2 (IL-4, IL-13) responses by MLNC and splenocytes from infected Treg-depleted mice in response to mitogen stimulation as compared with untreated, infected controls (Fig. 5A). Interestingly, also IL-10 levels were significantly increased in Treg-depleted mice. Importantly, this also held true for spleen cells (and, partially, cells from MLN) stimulated with parasite extracts (Fig. 5A). Cells from infected, undepleted DEREG mice were stimulated as additional controls and we detected no significant differences with respect to cytokine production as compared with infected WT mice. Thereby, we could formally exclude that the transgenic mice show an inherent stronger Th2 response (data not shown).

thumbnail image

Figure 5. Local and systemic cytokine responses. MLN and spleen cells isolated at day 14 were stimulated with the mitogen concanavalin A or H.p.b. Ag extract and the levels of (A) IL-4, IL-13 and IL-10 and (B) IFN-γ and IL-17 were detected in culture supernatants by ELISA. Mean+SEM of five to six mice per group is shown. Mean cytokine levels in supernatants of unstimulated cells (when above detection threshold) are indicated by black arrowheads. One representative result from two independent experiments is shown. *p<0.05, **p<0.005, ns: no significant difference; Mann–Whitney test.

Download figure to PowerPoint

We also assessed the impact of Treg depletion on Th1-like and Th17 responses (Fig. 5B). MLNC from Treg-depleted mice showed trends for stronger IFN-γ and IL-17 responses after mitogen stimulation irrespective of the absence or presence of worms (Fig. 5B). Interestingly, uninfected, Treg-depleted DEREG mice showed a strong systemic IFN-γ response significantly higher than in infected, Treg-depleted mice, whereas systemic IL-17 responses were relatively low (Fig. 5B). In response to the parasite extracts, MLNC from Treg-depleted, infected animals reacted with a moderate, but significantly increased IFN-γ production (Fig. 5B); IL-17 was not detected in response to worm antigens (data not depicted).

Furthermore, we found an earlier onset of the parasite-specific cytokine responses following depletion of Treg, as observed after restimulation of cells from animals at day 7 p.i. (after the second round of depletion) (Fig. 6). While cells from infected WT animals did not react with secretion of marked amounts of any of the tested cytokines at this early time point of infection, depleted mice readily produced relatively high amounts of IL-4 and IL-13. IL-10 was also secreted in higher amounts by cells from infected, depleted mice as compared with infected WT controls. The detected IFN-γ levels in response to parasite extracts were not significantly different from baseline levels secreted by unstimulated cells (Fig. 6). These data indicate an early control of the anti-nematode Th2 response by Treg.

thumbnail image

Figure 6. Accelerated parasite-specific cytokine response after Treg depletion. MLNC were isolated at day 7 and stimulated with H.p.b. Ag extracts. Cytokines were measured in supernatants by ELISA. Mean+SEM of four mice per group is shown. Mean cytokine levels in supernatants of unstimulated cells (when above detection threshold) are indicated by black arrowheads. One representative result from two independent experiments is shown. p<0.05, **p<0.005, ns: no significant difference; Mann–Whitney test.

Download figure to PowerPoint

To gather more information on whether Treg might also influence the activation status of conventional CD4+ T cells, we determined the frequency of CD62 Llow and CD69+ activated CD4+ T cells by flow cytometry. We detected drastic changes in the expression of both markers after Treg depletion (Fig. 7). While uninfected, Treg-depleted DEREG mice showed a significant increase of CD62 Llow effector/memory CD4+ T cells in comparison to non-depleted naïve and infected WT controls, this effect was even more pronounced in MLN and spleens of infected, Treg-depleted DEREG mice (Fig. 7 A, B and E). The frequencies of recently activated CD4+CD69+ cells followed a similar pattern (Fig. 7C–E). These data clearly show that an earlier and strengthened Th2 cytokine response occurs after Treg depletion in worm-infected DEREG mice, a finding coinciding with more activated CD4+ T effector cells. However, the increased Th2 reactivity did not facilitate worm expulsion.

thumbnail image

Figure 7. Frequencies of activated CD4+ T cells. MLNC and splenocytes were stained ex vivo for the expression of L-selectin (CD62 L) and CD69 and analyzed by flow cytometry. (A and B) Frequencies of effector/memory cells (CD62 Llow/neg) in CD4+ T cells. Mean+SEM for groups of five to six mice is shown. (C and D) Frequencies of recently activated CD4+CD69+ T cells. Mean+SEM for groups of five to six mice is shown. (E) Exemplary plots for CD62 L/CD69 expression by CD4+ T cells derived from spleens. Frequencies of gated CD69+ cells are depicted. One representative result from two independent experiments is shown. p<0.05, **p<0.005, ns: no significant difference; Mann–Whitney test.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References

In this study we used a novel transgenic mouse model permitting the selective depletion of Foxp3+ cells in order to study the role of Foxp3+ Treg in an intestinal nematode infection. Mice that were depleted of Foxp3+ Treg during the early acute phase of a H. p. bakeri infection developed severe inflammation of the small intestine and reacted with an accelerated and increased parasite-specific Th2 reactivity.

Our data suggest the following scopes of Treg action during intestinal worm infection: (i) Treg are crucial for the control of immunopathology at the sites of infection; (ii) Treg control the onset and magnitude of parasite-specific Th2 responses as their depletion leads to accelerated and amplified effector cytokine production; (iii) the transient removal of Treg has no effect on worm establishment or fecundity. Therefore, our data clearly argue for a primary role of Treg in curtailing pathological immune reactions as well as dampening host protective responses precociously.

These data are particularly interesting with regard to recent studies investigating the influence of nematodes on inflammatory diseases such as colitis or airway hyperreactivity. It has been demonstrated that nematode infections may prevent and control undesired immune reactions 6–12, 30. Several studies showed that the beneficial effects of the infections could be mimicked by transfer of Treg derived from nematode-infected donors to naïve recipients rendering them refractory to the inflammatory disorders such as asthma and colitis 6, 8, 11, 12. These data indicate that circuits of Treg action during nematode infection may also modulate unrelated bystander immune reactions and supports the view of Treg as important immunoregulators that can be modulated by parasites. Here, we directly addressed the function of Treg with respect to control of the anti-nematode immune response and worm-induced pathology.

Strikingly, the transient removal of Treg led to an increased local and systemic parasite-specific Th2 response seen as stronger production of IL-4 and IL-13. As cells from naïve controls did not react with a relevant Th2 cytokine production to parasite extracts, we conclude that the detected cytokines mainly derive from parasite-specific Th cells and were not produced by innate immune cells, e.g. in response to TLR stimuli present in parasite extracts. Furthermore, we clearly detected increased frequencies of activated effector/memory CD4+ T cells, arguing for an increased Th2 effector pool after Treg depletion. In contrast, a moderate increase of parasite-induced IFN-γ production was only detected locally in gut-draining MLN. These data are in line with a study demonstrating a more pronounced Th2 response in mice infected with the intestinal nematode Trichinella spiralis after removal of Treg by treatment with anti-CD25 Ab 19. Studies using anti-CD25 Ab during filarial and schistosome infections also report on increased cytokine responses 15, 16, 18, 24, 31. On the contrary, it has been shown recently that anti-CD25 Ab treatment during acute T. gondii infection impairs effector T-cell responses 26. These contradictory results emphasize the advantage of the DEREG model, enabling the specific depletion of Foxp3+ Treg. Furthermore, the DEREG model allows Treg depletion irrespective of CD25 expression, which is important as especially Treg in the gut often lack CD25 expression.

In our model, Foxp3+ Treg depletion in naïve mice provoked an increased pro-inflammatory local and systemic Th1 response after mitogen stimulation, seen as higher magnitudes of IFN-γ. Interestingly, this systemic effect was counterbalanced in depleted mice in the presence of H. p. bakeri, while MLNC from infected, depleted mice reacted with similar IFN-γ and IL-17 production as cells from naïve depleted controls. The trends for an increased pro-inflammatory reactivity of MLN cells from infected, depleted mice might be due to an increased availability of components from the intestinal microflora in the MLN due to the heavily inflamed small intestine. This shows that Treg are essential in controlling pro-inflammatory local responses that may emerge in response to nematode-driven intestinal pathology. Hence, our data indicate that interfering with modes of Treg action during an intestinal nematode infection strengthened the typical Th2 response against the parasite, while Treg depletion in naïve mice led to a modest increase in Th2 reactivity and supported systemic Th1 responses.

Moreover, we observed an accelerated development of a parasite-specific response in the gut-draining MLN after removal of Foxp3+ Treg early during infection, pointing out that the early commitment and outgrowth of Th2 effector cells is subject to the control by Treg. Importantly, H. p. bakeri-infected mice transiently depleted of Foxp3+ Treg displayed a strong anti-inflammatory IL-10 response alongside with an earlier and stronger Th2 response. We and others have shown that Treg produce IL-10 in a parasite-specific manner and thereby contribute to the anti-inflammatory cytokine response seen as a hallmark of nematode infections 21, 31, 32. However, studies with Schistosoma mansoni-infected mice clearly showed that effector T cells are the dominant source of IL-10 during infection 18, 31–33. Our current study suggests that non-Treg are the dominant source of IL-10 during intestinal nematode infections. However, future studies will have to show whether the removal of Foxp3+ Treg leads to the induction of adaptive Foxp3IL-10+ (Tr1-like) regulatory populations 34 or whether IL-10-producing Th2 cells or innate cells compensate for the loss of Treg.

Our data clearly show that Treg are essential for controlling the parasite-induced pathology in the small intestine. These data are in line with a recent study showing that the application of anti-CD25 Ab leads to an increased pathology in mice infected with Trichuris muris23. By examining the colon of H. p. bakeri-infected mice as an intestinal control site, we showed that the stronger pathology after Treg removal was restricted to the small intestine and excluded that such inflammatory reactions were a general consequence of the complete (short-term) and partial (long-term) removal of Foxp3+ Treg (data not shown). Intestinal pathology in our model infection appeared as increased cellular infiltration of the lamina propria and a disturbed architecture of the intestinal epithelium with villus atrophy and crypt hyperplasia. For infections with T. spiralis it has been shown that IL-13 is responsible for villus atrophy as well as crypt and goblet cell hyperplasia 35. Hence, we speculate that the increased IL-13 response in infected, depleted DEREG mice is one factor responsible for the disturbed architecture of the mucosal tissue.

It is interesting to note that an increased intestinal pathology as well as Th2 reactivity occurred despite the transience of Treg depletion in our model. We could show that Treg were efficiently removed from peripheral blood, lymphatic organs and the intestine for a relatively short time window and were then partially replaced by GFPDTRFoxp3+ Treg. To date it is not clear whether, compared with the depleted population, the reappearing Treg have the full spectrum of the T-cell receptor repertoire and whether the GFPDTRFoxp3+ Treg are fully functional. Unlike protozoan infections, where Treg were shown to play a beneficial and essential role for the survival of the parasite 13, 14, we found that the interference with Treg activity had no influence on the worm burden. The removal of Foxp3+ cells did not interfere with worm development and fecundity and did not lead to an early expulsion of adult worms. Similarly, D'Elia and colleagues showed that applying anti-CD25 Ab during T. muris had no effect on worm establishment or expulsion 23. However, the release of a broad spectrum of Th2 cytokines was impaired after anti-CD25 Ab treatment, arguing for the side-by-side depletion of Th2 effector cells 23. In contrast, our model showed that the selective Treg depletion strongly promoted the Th2 response to H. p. bakeri antigen.

The general view suggests that expulsion of adult intestinal nematodes is critically dependent on CD4+ T cells secreting the Th2 effector cytokines IL-4 and IL-13 36–38 and that these cytokines lead to an inhospitable environment, possibly enabling the expulsion of the parasites in the chronic phase of infection 37, 38. It has also been shown that the application of high doses of recombinant IL-4 abrogates primary infections with H. p. bakeri39. However, our study shows that although Th2 responses were accelerated and pronounced in infected mice, the worm burden was not affected. Furthermore, the physiological changes provoked by the worm infection, such as epithelial cell proliferation and goblet cell hyperplasia, were similar in Treg-depleted and control mice. Thus, it seems that H. p. bakeri, which generally leads to chronic infections, tolerates a broad spectrum of Th2-mediated immune and physiological reactions compared with other intestinal nematodes such as Nippostrongylus brasiliensis, T. muris or T. spiralis, which are more efficiently expelled in acute infections 37, 38.

In conclusion, our data show that Treg are crucial for the control of helminth-induced pathological changes in the intestine and for limiting Th2 responses. Further studies will have to show whether Treg are involved in mediating the general immunosuppression seen in chronic intestinal nematode infections. The DEREG model harboring GFP+Foxp+ Treg allows to dissect the role of Treg while preserving effector T cells and may thereby help to further delineate the role of nematode-induced Treg in disease models, such as colitis and asthma.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References

Mice and parasites

DEREG mice 27 were bred under specific pathogen-free conditions at the German Arthritis Research Center, Berlin. Male mice were infected with 200 third stage larvae of H. p bakeri. WT C57BL/6 mice (purchased from Bundesinstitut für Risikobewertung, Berlin, Germany) and WT littermates of DEREG mice were used as controls. All experiments were approved by the German ethics committee for the protection of animals.

Experimental design

Groups of infected mice received 1 μg of DT (Merck, Darmstadt, Germany) dissolved in PBS, pH 7.4, intraperitoneally on days 4, 6, 8 and 10 after infection. Naïve DEREG mice, infected WT C57BL/6 and non-transgenic DEREG-littermates served as controls. Success of Treg depletion was surveyed by flow cytometry analyses of peripheral blood samples on days 5, 9 and 11. Mice were killed on day 7 or 14 p.i. by CO2 inhalation.

Immunohistochemistry

Intestines were removed and the duodenum was opened longitudinally to collect adult worms. In brief, sections from formalin-fixed paraffin-embedded samples were subjected to H&E staining or to a heat-induced epitope retrieval step before incubation with primary Ab for 30 min. For detection of Foxp3, the rat Ab clone FJK-16s (eBioscience, San Diego, CA, USA) was applied at a dilution of 1:100 and slides were blocked using a commercial peroxidase-blocking reagent (Dako, Glostrup, Denmark) followed by a secondary rabbit anti-rat Ab (Dako) and the EnVision peroxidase kit against rabbit Ab (K4003, Dako). For Ki-67 (TEC-3, Dako, 1:500) labeling biotinylated rabbit anti-rat (Dako) secondary Ab were used followed by the streptavidin alkaline phosphatase kit (K5005, Dako). Alkaline phosphatase was revealed by Fast Red as chromogen and peroxidase was developed with a highly sensitive diaminobenzidine chromogenic substrate for approximately 10 min. Negative controls were performed by omitting the primary Ab. Numbers of Foxp3+ cells were counted in ten HPF (400× magnification) per sample. Ki-67+ and goblet cells were counted in ten crypts per sample. Histological scoring was performed according to a two-parameter scale (i/ii, each from 0 to 4): (i) appearance of intestinal architecture (0, normal; 1, slight blunting of villi; 2, moderate blunting of villi, slight crypt hyperplasia; 3, strong blunting of villi and crypt hyperplasia; 4, strong villus atrophy and crypt hyperplasia); (ii) signs of inflammation (0, no signs of inflammation; 1, mild leukocyte infiltration; 2, moderate leukocyte infiltration; 3, marked leukocyte infiltration, thickening of bowel wall; 4, transmural infiltration, mucin depletion, strong bowel wall thickening, ulcerations).

In vitro cell culture

Spleen and MLN were removed aseptically and passed through cell strainers in PBS, pH 7.4, containing 0.2 % BSA. After erythrocyte lysis, splenocytes and MLNC were counted using a CASY counter (Innovatis, Germany), washed and plated at 3.5×105 cells/well in RPMI 1640 containing 100 U/mL penicillin, 100 μg/mL streptomycin, 20 mM L-glutamine (all from PAA, Austria) and 10% fetal calf serum (Biochrom, Germany) on 96-well plates in 200 μL. Cells were stimulated for 48 h with concanavalin A (2 μg/mL) or for 72 h with adult worm soluble extract (12 μg/mL, for extract preparation see 21). Supernatants were removed and stored at −80°C. Media were replaced, followed by a pulse with 1 μCi of 3H-thymidine (MP Biomedicals, Germany) for 20 h. The proliferative response was analyzed using a beta-counter (PerkinElmer, Germany). All cultures were performed in quadruplicates.

Cytokine analysis

ELISA-kits for the detection of IL-4, IL-10 and IFN-γ were purchased from BD Biosciences. Duo sets for the detection of IL-13 and IL-17 were from R&D Systems. Assays were performed as recommended by the manufacturers.

Flow cytometric analysis

The following monoclonal Ab and kits were purchased from eBioscience: Pacific blue anti-CD4 (clone RM4-5), PE-Cy7 anti-CD69 (clone H1.2F3), Alexa 700 anti-CD25 (clone PC61), PE anti-Foxp3 (staining kit, clone FJK-16s). Biotinylated anti-CD103 (clone M290) and anti-CD62 L (clone MEL-14) were a kind gift from Alexander Scheffold from the German Arthritis Research Center. PE-Cy7 coupled to streptavidin (BD Biosciences) was used as secondary conjugate. Peripheral blood samples were cleared of erythrocytes using the BD lysis kit. Cells were stained in PBS containing 0.2% BSA. Samples were acquired using a LSR II flow cytometer (BD Biosciences) and analyzed with FloJo software (Treestar, USA).

Statistic analysis

Each experiment was performed with four to six mice/group and the data shown are representative of at least two independent experiments.

Statistical analysis was performed with GraphPad Prism software (San Diego, USA) using the two-tailed Mann–Whitney U test. The significance of differences between multiple groups was analyzed with the Kruskal–Wallis test. Data are represented as means±SEM. Values of p<0.05 were considered to be statistically significant.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References

The authors thank Simone Spiekermann, Bettina Sonnenburg and Marion Müller for excellent technical assistance. This research was supported by German Research Council Grant SFB 650 (to S.H., A.H. and R.L.).

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

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References