Oral tolerance induction by mucosal administration of cholera toxin B-coupled antigen involves T-cell proliferation in vivo and is not affected by depletion of CD25+ T cells

Authors


Dr A. George Chandy, Göteborg University Vaccine Research Institute (GUVAX) and Department of Medical Microbiology and Immunology, Göteborg University, Göteborg, Sweden.
Email: annie.george.chandy@microbio.gu.se
Senior author: Annie George Chandy,
email: annie.george.chandy@microbio.gu.se

Summary

Oral administration of antigens coupled to the B subunit of the cholera toxin (CTB) can dramatically reduce the amount of antigen needed for tolerance induction and has been used in several animal models to suppress conditions where the immune system overreacts to foreign and self-antigens. In this study, the cellular events following oral administration of CTB-coupled antigen was investigated. As a model system, limited numbers of CSFE-labelled cells from influenza haemagglutinin peptide (HApep) T-cell transgenic mice were transferred to wild type mice and the mice were then given CTB-coupled HApep orally. The inductive events of CTB-induced tolerance was characterized by extensive proliferation of HApep-specific T cells in the mesenteric lymph nodes (MLNs) and in the spleen. The proliferating cells up-regulated the gut homing molecule α4β7 and down-regulated the high endothelial venule binding molecule l-selectin. Addition of the whole cholera toxin (CT) to CTB-HApep showed a similar pattern as CTB-HApep feeding, with antigen-specific proliferation in the MLN and spleen and expression of α4β7 on the proliferating cells. However, addition of CT to CTB-HApep, produced a stronger and faster proliferative response and abrogated CTB-HA mediated oral tolerance. Feeding of CTB-HApep expanded CD25+ cells in the MLNs. CTB-induced oral tolerance could, however, not be explained by CD25+ dependent regulatory activity, as oral administration of CTB-HApep to mice depleted of CD25+ cells still gave rise to systemic tolerance. Thus, several mechanisms might co-orchestrate the systemic tolerance seen in response to feeding with CTB-coupled antigen.

Introduction

The generation of systemic tolerance after oral antigen exposure is well recognized, and generally requires feeding of high antigen doses. Coupling of antigens to cholera toxin B subunit (CTB) can dramatically decrease the amount of antigen needed for tolerance induction. The ability of minute amounts of antigen coupled to CTB to evoke immunological tolerance and thereby prevent untoward immune reactions has been demonstrated in animal models of T-cell mediated autoimmune conditions,1–3 immunoglobulin E-mediated allergic reactions4–6 and infection-induced pathological inflammatory conditions.7,8 Although the ability of CTB-conjugated antigens to induce immunological tolerance is well documented, the mechanisms governing the induction of CTB-mediated tolerance remain less understood. The induction of T-cell tolerance following oral administration of antigen may occur by a variety of mechanisms including clonal deletion, anergy and immunosuppression. The specific mechanism of tolerance has been suggested to depend upon the tolerizing regime employed. Thus, clonal deletion and/or anergy may result from feeding high doses of antigen, while some form of active regulation may be a feature of low dose/repeated feeding regimes.9 However, arguing against this simple concept, oral feeding of high doses of antigen may also result in transforming growth factor-β (TGF-β) production and bystander suppression in certain experimental situations – a hallmark of immunosuppression.10,11 Immunosuppression is an intrinsic property of the immune system and is partially mediated by T cells. The best defined T-cell population with immunosuppressive activity is enriched in the naturally activated subset of CD4+ T cells, which constitutively express the interleukin (IL)-2R α chain CD25. These thymus-derived suppressor cells are present in normal unmanipulated individuals and contribute to the maintenance of self tolerance and protect from a variety of autoimmune diseases.12 In addition to these naturally occurring regulatory T cells (Tregs), Treg cells can be induced in naïve peripheral CD4+ T cells by antigenic exposure.13 These induced Tregs are separated into subsets according to the cytokines they produce. Thus, Tr1s secrete IL-10 and only small amounts of IL-2, IL-4 and TGF-β14 in contrast to T helper 3 (Th3) regulatory cells, which secrete high levels of TGF-β but only small amounts of IL-10.15

While CTB-coupled antigens efficiently induce oral tolerance, coadministration of CT to antigen instead functions as an adjuvant to induce active immunity upon oral adminstration.16

The purpose of this study was to examine unique features of CTB-induced T-cell tolerance following antigen exposure via the oral route and to compare these to a T-cell priming protocol. We utilized an approach described by Kearny et al. in which small numbers of antigen-specific T cells from T cell receptor (TCR) transgenic mice were transferred into conventional mice.17 We used the haemagglutinin peptide (HApep) TCR transgenic system18 to examine the fate of antigen-specific T cells following oral antigen exposure. A single feeding with a low dose of antigen was shown to induce oral tolerance when the HApep was coupled to CTB. Addition of cholera toxin (CT) to CTB-HApep instead abrogated oral tolerance. Examining the initial cognate T-cell responses in response to feeding, we found that the systemic tolerance induced by oral administration of CTB-coupled antigen was preceded by activation in the regional draining lymph nodes. Addition of CT to CTB-HApep induced stronger and faster activation of HApep-specific T cells. Identifying the inductive events of CTB-mediated tolerance also allowed us to investigate the phenotype of T cells proliferating in response to CTB-HApep. CTB-HApep feeding promoted the emergence of an antigen-specific population with gut-homing properties and also expanded CD25+ T cells. Even though CD25+ cells expanded in response to CTB-HApep feeding, these cells did not appear to be crucial for CTB-HApep specific oral tolerance as depletion of CD25+ cells in vivo did not affect oral tolerance induction.

Materials and methods

Mice and adoptive transfer protocols

Wild type BALB/c mice were purchased from Taconic M & B (Lille Skensved, Denmark). Mice used throughout this study were male mice, 6–8 weeks of age. BALB/c mice expressing a TCR α/β for peptide 111–119 of the influenza virus H1 haemagglutinin in the context of I-Ed were a kind gift from Dr H. von Boehmer. (Harvard University, Boston, MA) and were bred in our animal facility. Mice were confirmed to be HApep-transgenic by analysing peripheral blood cells for the presence of the transgenic TCR using a clonotype specific monoclonal antibody (6.5) and TCR Vβ8. Purified politeal lymph node (PLN) and mesenteric lymph node (MLN) T cells from HApep-Tg mice (5–10 × 106 cells) were adoptively transferred by intravenous (i.v.) injection into normal BALB/c mice.

Antigens

HApep.  A synthetic peptide corresponding to amino acid residues 108–119 (SVSSFERFEIFPKC) of the influenza virus H1 subtype was purchased from Neosystem (Strasbourg, France).

CTB-HApep.  CTB-HApep genetic fusion protein was produced in house. The construction and purification of CTB-HApep has been described in detail elsewhere.19 In the CTB-HApep gene fusion protein, HA residues 108–119 replaced residues 56–63 in the CTB structure. The fusion protein was found to have GM1 receptor binding activity by means of G−1 enzyme-linked immunosorbent assay (ELISA). Assembly of the CTB-molecules to pentameric structure was detected in the G−1 ELISA by the CTB pentamer-specific antbody LT39.20

CT.  CT was purchased from List Biological laboratories Inc. (Madison, NJ).

Feedings and immunizations

For feedings of antigens, mice were given 5 or 30 µg HApep, (a) per se, (b) genetically linked to CTB, or (c) genetically linked to CTB and mixed with 5 µg CT. All feedings were done in 3% (w/v) NaHCO3 with 1 mg soybean trypsin inhibitor (Sigma) by intragastric gavage (20 gauge feeding needles).

For the measurement of recall responses to HApep after feeding, mice were immunized in the left footpad with 50 µg HApep emulsified in Freund's complete adjuvant (Difco Laboratories, Detroit, MI). Proliferative responses were measured two weeks after challenge on cells isolated from the draining PLN.

Cell preparations

T cells were obtained by purification on mouse T-cell enrichment columns (R & D, Minneapolis, MN). To separate CD25+ and CD25 T cells, the enriched T cells were magnetically labelled by using the CD25+ Microbead kit according to the manufacturer's instructions followed by separation on an MS column (Miltenyi Biotech, Auburn, CA). The CD25 negative population contained less than 0·2% CD25+ cells.

Antibodies, cell labelling and flow cytometry

For the detection of proliferating HApep-specific cells in response to feeding, T cells were purified on T-cell enrichment columns and labelled before cell transfer with 5,6-carboxy-succinimidyl-fluoresceine-ester (CSFE, Molecular Probes, Eugene, OR) as previously described.21 Briefly, cells were resuspendend in phosphate-buffered saline (PBS) at 5 × 107 cells/ml and incubated with CFSE at a final concentration of 10 µm for 15 min at 37°. The cells were then washed and again resuspended in PBS and incubated for 30 min at 37°, washed and finally injected intravenously. Proliferating CSFE+ HApep-specific cells were detected by costaining with phycoerythrin (PE)-labelled anti-Vβ8 (F.23.1, BD Pharmingen, San Diego, CA). For the detection of homing receptors on proliferating cells in response to HApep-feeding, MLN cells were either labelled with anti-Vβ8 biotin followed by avidin-allophyco cyanin (APC) (BD Biosciences, Piscataway, NJ), together with anti-L-selectin (PE) or with anti-Vβ8–PE together with anti-α4β7 followed by biotinylated anti-rat immunoglobulin G2a (IgG2a; BD Pharmingen) in 50% mouse serum, followed by avidin-APC. For the detection of proliferating CD25+ cells MLN cells were costained with anti-Vβ8 PE and αCD25 APC (clone PC61, BD Pharmingen). Data were acquired on a FACScan (Becton Dickinson, Mountain View, CA) and analysed with CellQuest software from Becton Dickinson. For the in vivo depletion of CD25+ cells, mice were injected with 0·7 mg of the CD25 depleting antibody PC61 (American Type Culture Collection, Manassas, VA) produced in house. The mice were injected once, 2 days after the adoptive transfer of T cells from HApep transgenic mice. To determine the efficiency of CD25+ cell depletion, blood lymphocytes were labelled with αCD25, clone 7D4 (BD Pharmingen).

Proliferation assays and cytokine analyses

For the measurement of proliferative responses after CTB-HApep feeding, draining PLN cells were prepared 2 weeks after priming with HApep and cultured in 96-well flat bottomed plates at 2 × 105 cells/well in the presence of HApep at 10 µg/ml concentration in Iscove's complete medium. Plates were cultured for 4 days including 12 hr of [3H]TdR incorporation at 1 µCi/well (Amersham, Frankfurt, Germany). Suppressive properties of CD25+ cells from the different treatment groups were examined by isolating CD25+ cells from spleens and MLN 2 weeks after oral antigen administration. The CD25+ T cells were added in graded numbers to cultures containing naïve peripheral lymph node cells from HApep transgenic mice at 105 cells/well in flat bottomed 96-well plates. Cells were cultured for 3 days including 12 hr of [3H]TdR incorporation at 1 µCi/well. Results from the proliferation assays are expressed as stimulation indexes (SI), defined as the ratio between [3H]TdR-incorporation of cells with antigen and [3H]TdR incorporation of cells without antigen.

Statistical analyses

Students' t-test with Bonferroni correction was used to compare mean values of different groups. One-tailed P-values < 0·05 were considered significant.

Results

CTB-conjugated antigen induces oral tolerance more efficiently than uncoupled antigen

Feeding of antigen coupled to CTB has previously been reported to induce systemic tolerance.22 We set up an adoptive transfer system to confirm this finding in a model using HApep conjugated to CTB as the tolerizing antigen. For this purpose mice were adoptively transferred with naïve HApep-transgenic T cells. Two days after the adoptive transfer, mice were given HApep or corresponding doses of CTB genetically linked to CTB (CTB-HApep) orally. The mice were then primed in the footpad with HApep emulsified in CFA, 2 days after feeding. Recall proliferative responses to HApep were tested in vitro on draining PLN cells 2 weeks after the priming (Fig. 1). Feeding of three doses of 30 µg CTB-HApep 2 days apart, efficiently induced systemic tolerance (P = 0·02, Fig. 1a), seen as reduced proliferation in CTB-HApep fed mice compared to control fed mice.

Figure 1.

 Oral administration of CTB-HApep efficiently induces systemic tolerance, which is abrogated by CT. Mice were adoptively transferred with T cells from HA transgenic mice. Four days after the adoptive transfer, mice were fed HA or CTB-HApep. Two days after feeding, the mice were primed in the footpad with 50 µg HApep in emulsified Freund's complete adjuvant. T-cell proliferation in response to HApep was measured on draining PLN cells 2 weeks after priming. (a) Feeding of 15 µg HApep conjugated to CTB three times 2 days apart. Data are means of three separate experiments with a total of nine mice per bar. Results are expressed as mean + SEM. P-value was calculated by one-tailed t-test. (b) Feeding of a single dose of 5 or 30 µg HApep. Dotted bars HApep, hatched bars CTB-HApep. Data shown as means of three mice/group. (c) Mice were fed a single dose of 30 µg CTB-HApep + 5 µg of CT. Data shown are means of three mice/group.

Also, a single feeding of antigen was tested for tolerance induction (Fig. 1b). While feeding of 5 µg HApep coupled to CTB was able to induce oral tolerance, the corresponding dose of HApep alone failed to induce oral tolerance. Upon feeding of a higher antigen dose (30 µg HApep), a minor suppression of proliferative responses could be observed. However, CTB-HApep was substantially more efficient at inducing oral tolerance at 30 µg of feeding dose than HApep per se (SI 15·0 versus 2·5). The effect of the oral adjuvant CT was tested in this system, and was able to reverse CTB-HApep induced oral tolerance (Fig. 1c). Addition of CT to CTB-HApep abrogated oral tolerance in this system.

Feeding of CTB-coupled antigen induces T-cell proliferation in MLN and spleen

We have previously reported that coupling of HApep to CTB could decrease the dose required for T cell activation by professional APCs >10 000-fold, monitored as proliferation of antigen-specific T cells and cytokine production in vitro.19 We investigated whether oral administration of CTB-coupled antigen could give rise to T-cell proliferation also in vivo, and in such cases where CTB-induced activation and subsequent tolerance were initiated. For this purpose, mice were adoptively transferred with CSFE-labelled T cells from HApep-transgenic animals. Spleens, MLNs and PPs were examined for antigen-specific dividing T cells by staining for Vβ8+ CSFE+ cells (Fig. 2) 2 and 4 days after feeding. While feeding of 30 µg HApep was unable to induce proliferation at any of these time points, feeding of CTB-HApep had induced proliferation in the MLNs and to a lesser degree in the spleen at 4 days postfeeding (Fig. 2a, b). Addition of the oral adjuvant CT induced stronger activation than CTB-HApep, seen as a higher degree of proliferation of Vβ8+ CSFElow cells than CTB-HApep per se. No expansion of CSFE+ cells could be observed after feeding of CT alone (results not shown). Interestingly, feeding of CTB-HApep + CT showed a higher degree of proliferation of both Vβ8+ as well as Vβ8 cells, indicating a higher degree of bystander activation with CTB-HApep + CT as compared to CTB-HApep.

Figure 2.

 Oral administration of CTB-HApep causes antigen proliferation in vivo. Mice were adoptively transferred with CSFE-labelled T cells from HA transgenic mice. The mice were given antigen orally 2 days after cell transfer. Spleens, MLN and PP were analysed for Vβ8+ CSFE labelled cells at various time-points after feeding. (a) Spleen, MLN and PP cells were analysed 4 days after feeding for Vβ8+CSFE labelled cells. One representative dot plot is shown. The percentage of CSFE+ Vβ8+ cells are shown in the upper right corner of each dot plot. (b) means of Vβ8+ CSFElow cells and Vβ8 CSFElow cells in the indicated organs (two mice/group) are shown. (c) MLN cells were analysed 2 days after feeding. Percentage of CSFE-labelled Vβ8+ cells are given in each figure. (d) means of Vβ8+ CSFE+ cells for two animals/group are shown.

While a small population of non-proliferating CSFE-labelled cells could be detected in the Peyer's patches (PPs) of non-fed and HApep-fed mice, a minor population of CSFElow cells could be detected in the PP of mice fed with CTB-HApep and CTB-HA + CT which had already undergone several divisions in response to feeding, suggesting that CTB-HA and CTB-HA + CT feeding induce proliferation earlier in the PP than in the MLN and spleen. Thus, oral administration of CTB coupled antigen leads to T-cell proliferation in MLN and to a lesser degree in the spleen. Examining proliferative responses at an earlier time-point, i.e. 2 days post-antigen administration, we observed that the kinetics of the proliferative response after CTB-HApep feeding was slower compared to the administration of CTB-HApep + CT. At 2 days after feeding, only low levels of proliferation could be detected mainly in MLNs (Fig. 2c, d). This suggests that the kinetics of CTB-induced proliferation changes with the addition of the oral adjuvant CT to activate T cells more rapidly than with feeding of CTB-HApep alone.

Proliferating HApep-specific cells down-regulate l-selectin and up-regulate α4β7 in response to CTB-HApep and CTB-HApep + CT feeding

To determine if HApep-specific cells were destined for the mucosa or for lymphoid organs after activation and proliferation in the MLN, adoptively transferred CSFE-labelled cells from HApep transgenic animals were stained for the gut homing molecule α4β7 and the high endothelial binding molecule l-selectin. We monitored the initial proliferation following antigen administration, i.e. 2 days after CTB-HApep + CT feeding and 4 days after CTB-HApep feeding (Fig. 3). Dividing cells at 2 and 4 days post-feeding expressed α4β7 and had down-regulated their expression of l-selectin. This indicates that they will have limited ability to access lymph nodes and PPs, but rather home to the lamina propria.

Figure 3.

 Proliferating T cells in response to CTB-HApep feeding up-regulate α4β7 and down-regulate l-selectin in the MLN. CSFE-labelled T cells from HA transgenic animals were adoptively transferred to wild type mice. Two days after transfer, mice were fed antigen. MLN cells were analysed at 2 days after CTB-HApep + CT feeding and four days after CTB-HApep feeding. Cells were gated on Vβ8+ lymphocytes and analysed for l-selectin or α4β7.

CTB-induced oral tolerance is associated with the expansion of CD25+ T cells

The induction of oral tolerance by administration of low doses of antigen is associated with changes in CD25 CD4+ or CD25+ CD4+ T cells.23,24 We asked whether oral administration or CTB-coupled antigen was also associated with an expansion of CD25+ T cells. To address this question, wild type mice were adoptively transferred with CSFE labelled T cells from HApep transgenic mice. The mice were then fed with CTB-HApep and killed at day 4 after feeding. Feeding of CTB-HApep was associated with an expansion of CD25+ cells in the MLNs, Addition of CT did not significantly change the proportion of HApep specific CD25+ dividing T cells (Fig. 4).

Figure 4.

 CTB-HApep feeding expands CD25+ cells in vivo. CSFE-labelled T cells from HA transgenic animals were adoptively transferred to wild type mice. Two days after transfer, mice were fed antigen. MLN cells were analysed at 4 days after feeding. Cells were gated on Vβ8+ lymphocytes and analysed for CD25. (a) One representative plot is shown. (b) Analyses of proliferating CD25+ cells in response to feeding (two mice/group) are shown.

CTB induced oral tolerance is independent of CD25+ cells

CD25+ regulatory T cells have earlier been implicated in oral tolerance. Consistent with this finding, we observed that purified CD25+ cells from adoptively transferred CTB-HApep fed mice were able to suppress in vitro proliferation of total PLN cells from HApep transgenic animals more efficiently than CD25+ cells from adoptively transferred PBS fed mice (results not shown). We therefore asked whether the induction of CD25+ regulatory T cells could explain tolerance induction by CTB-coupled antigen. To address this question, wild type mice were adoptively transferred with HApep-transgenic cells. Two days after the transfer, the recipients were injected with 0·7 mg of the CD25-depleting antibody PC61 i.v. The mice were fed CTB-HApep, and challenged with HApep peptide in the foot pads. Proliferative responses to HApep peptide was then tested on draining lymph node cells (Fig. 5, upper panel) 2 weeks after challenge. Depletion of CD25 positive cells were confirmed throughout the course of the experiment by flow cytometry using a monoclonal antibody specific for a distinct epitope (7D4), which showed that there was a reduction of total CD25+ cells (R2, Fig. 5, lower panel) that could be explained mainly by the preferential depletion of CD25bright cells (R3, 2·8% in PBS-injected controls versus 0·6 in PC61-injected mouse) at the time of evaluation of proliferative responses. The depletion of CD25+ cells in vivo gave similar levels of CTB-HApep-induced tolerance as seen in untreated mice (SI 4·7 versus 1·27 in CTB-HApep-fed, CD25-depleted mice compared to SI 4·7 versus 2·55 in CTB-HApep-fed mice). Of note, the possibility that the lack of proliferation in CTB-HA-fed, CD25-depleted mice was caused by the elimination of CD25 effector cells activated by the feeding regimen, was precluded by the finding that non-fed, HApep-primed mice were not affected by the PC-61 treatment (Fig. 5).

Figure 5.

 CTB-induced tolerance is intact in the absence of CD25+ cells. Mice were adoptively transferred with T cells from HApep transgenic mice. In the upper panel, mice injected with the CD25 depleting antibody PC61 or control mice were fed with CTB-HApep. Four days later, the mice were fed CTB-HApep and primed later in the footpads with HA. Proliferative responses against HA was measured 2 weeks after priming. In the lower panel, a representative plot for CD25 expression by blood lymphocytes 18 days after PC61 antibody injection is shown. R2 represents all CD25+ cells, R3 represents CD25bright cells.

To confirm this finding in yet another system, we isolated T cells from HApep transgenic mice, depleted the CD25+ cells and adoptively transferred the CD25 population to naïve recipient mice. The mice were then fed CTB-HApep and challenged with HApep in the footpad. Two weeks after challenge, proliferative responses to HApep were tested on draining lymph node cells. Transfer of CD25 cells gave the same level of tolerance as transfer of whole T cells, showing again that CTB-induced oral tolerance is intact in the absence of CD25+ Tregs (Fig. 6).

Figure 6.

 Adoptive transfer of CD25 cells leads to oral tolerance in response to CTB-HApep feeding. T cells from HA transgenic mice were depleted of CD25+ cells and adoptively transferred to wild type mice. The mice were then fed CTB-HApep and primed in the foot-pads with HApep. HApep-specific proliferative responses of draining PLN cells was measured 2 weeks after priming.

Discussion

We have in the present study examined the initial T-cell activating events that take place following oral exposure to CTB-coupled antigen. After feeding of CTB-coupled antigen, we found initial T-cell proliferation of adoptively transferred antigen specific T cells in the MLN and to a lesser degree in the spleen. Thus, antigen directed differentiation occurs as a part of CTB-mediated T-cell tolerance. We have previously reported that the coupling of antigen to CTB dramatically increases antigen presentation by dendritic cells (DC) and other professional antigen-presenting cells as well as IL-2 and interferon-γ production by cognate T cells.19 Our previous results are compatible with this study considering the following. First, epithelial cells would be the first cells with antigen-presenting potential that encounter antigen upon peroral administration. Antigen presentation by epithelial cells preferentially induce tolerance because of their inherent inability to provide costimulation in the absence of an inflammatory signal25 and their tendency to produce tolerogenic exosomes.26 CTB will effectively facilitate the binding and uptake of the linked antigen by intestinal epithelial cells and thus make it more available for immune recognition. DCs in the underlying lamina propria will thus gain access to the antigen carried by the CTB and will be able to present the antigen more efficiently to T cells in the draining lymph nodes. Second, the gut milieu provides an inherently immunosuppressive environment because of the production of immunosuppressive cytokines, which would condition the professional antigen presenting DCs in the lamina propria to preferentially induce regulatory T cells in the absence of a strong adjuvant.27 Thus, CTB-antigen conjugate-induced tolerance may resemble tolerance induced by feeding a much higher dose of the pure antigen, which will initially induce a wave of proliferation followed by the differentiation of regulatory T cells.28,29 In the light of our previous findings19 it seems likely that an increase in antigen uptake and presentation by gut APC may explain the ability of CTB to enhance oral tolerance. CTB-coupling very likely decreases the dose required for antigen presentation by APCs upon feeding by potentiating uptake of antigen through its receptor G−1 present on DCs and all other nucleated cells present in the gut mucosa.

Following oral administration of CTB-coupled antigen, we monitored T-cell proliferation both locally in the MLN and systemically in the spleen. In accordance with our own results, Kobets et al. have shown that orally administered antigen in the context of major histocompatibility complex II, can be detected locally, as well as systemically very soon after feeding, suggesting that antigen presentation occurs in both locations.30 In our system, the PP displayed a much lower frequency of labelled transgenic cells which made it difficult to monitor proliferation at this site. However as seen in Fig. 2, there was a population of T cells in the PP that had undergone several divisions in the CTB-HApep fed mice. It is clear from other studies that the activation of T cells in the PP occurs within 24 hr in response to fed antigens,31 which may suggest that the 4-day time point used in the present study is less optimal for studying the T-cell dynamics in the PP. The dogma that PP is the inductive site for mucosal immune responses and oral tolerance have been challenged by recent findings showing that in mice having a selective absence of PP32,33 oral tolerance can still be induced. Our results are consistent with findings suggesting at limiting antigen doses, T-cell activation in the gut-associated lymphoid tissues occurs preferentially in the MLN34 and that MLN have a crucial role in the induction of mucosal immunity and tolerance.

In the present study the T cells that proliferated in response to CTB-HApep feeding gradually lost their expression of l-selectin – a feature of activated/memory cells and they up-regulated the gut-homing molecule α4β7. Addition of CT did not alter these properties, which implies that both feeding regimens differentiate T cells with the propensity to home to the intestine. Our results are consistent with studies suggesting that the site of antigen encounter determines the homing phenotype of memory and effector T cells in general.35

We have also shown that feeding of doses as low as 30 µg of HApep coupled to CTB induces the differentiation of CD25+ antigen-specific T cells. However, in vivo depletion of CD25+ cells in our system showed also that in the absence of these cells, oral tolerance in response to CTB-coupled antigen was intact. Two possibilities might explain these findings. Either CTB-HApep feeding give rise to CD25 regulatory T cells that are sufficient to maintain the tolerant state. Indeed, in mice regulatory T cells with both CD25+ and CD25 phenotype have been described36 distinguished by their expression of the E-cadherin binding integrin αΕβ7.37 The other possibility is that various degrees of anergy and/or deletion are concomitant with the presence of the CD25+ regulatory T cells. Feeding of CTB-coupled antigen gave proliferation in draining lymph nodes as previously mentioned and gave rise to regulatory T cells with CD25 phenotype. These are features associated with low-dose tolerance, which is associated with the emergence of regulatory T cells rather than with clonal deletion. Recent studies have however, shown that CD4+ CD25+ cells were generated by high doses of oral antigen, suggesting that clonal anergy or deletion may not be an exclusive mechanism behind high-dose oral tolerance.24,38 Studies on the importance of CD25+ regulatory cells for oral tolerance induction and maintenance have generated conflicting results. Dubois et al. showed that the depletion of CD25+ cells abrogated oral tolerance against a hapten.39 They, however, studied hapten-induced contact-sensitivity, which is a CD8 dependent response. This could partly explain the differences seen between our study and theirs. However, in accordance with our own study, Chung et al. have demonstrated that induction of oral tolerance induced by a high dose feeding regimen was unaffected by depletion of CD25+ cells in vivo, and could only be reversed when both CD25+ cells were depleted and TGF-β was neutralized.40

The transcription factor Foxp3 is crucial in the development and function of natural CD25+ CD4+ Treg cells41–43 and is currently the most specific marker available to identify Tregs. However, as it is a nuclear protein it would not be possible to functionally establish a role for Foxp3+ Treg cells in CTB-induced tolerance; this is why we have not performed any Foxp3 analyses. Neither did we analyse Foxp3 expression after CD25 depletion in vivo to evaluate the efficiency of Treg depletion, because there exists a population of Foxp3+cells that are negative for CD25, which are contained in the CD45RBlow population.42 This population would potentially be unaffected by the CD25 depletion.

In summary, we have found that feeding of CTB-conjugated antigen causes cell proliferation in the gut draining lymph nodes; the MLN, and that the emerging T cells have a propensity to home to the mucosa rather than to the periphery. Addition of CT showed a faster and stronger activation compared to CTB-HApep alone, which was the only obvious difference we could monitor between CTB-HApep and CTB-HApep + CT fed mice, in accordance with previous studies comparing oral tolerance versus oral priming.44,45 Most importantly, the induction of tolerance was totally abrogated by the cofeeding of CT. We found that CTB-HApep expands a population of CD25+ T cells. However, in the absence of CD25+ cells, tolerance was still present, suggesting that either other regulatory cells than CD25+ cells or other mechanisms of oral tolerance, e.g. deletion or anergy might co-orchestrate the oral tolerance seen in response to CTB-coupled antigen. The above findings suggest that CTB-coupled antigens potentiates uptake and immune handling in the gut mucosa, thereby offering an efficient strategy to dramatically lower the dosage used for tolerance induction. We conclude that CTB-induced oral tolerance mechanistically display features attributed both to high and to low dose oral tolerance.

Acknowledgements

The critical reading of this manuscript by Dr Kristina Eriksson is gratefully acknowledged. This work was supported by The Knut and Alice Wallenberg Foundation through its support of the Göteborg University Vaccine Research Institute (GUVAX).

Ancillary