Sublingual Tolerance Induction with Antigen Conjugated to Cholera Toxin B Subunit Induces Foxp3+CD25+CD4+ Regulatory T Cells and Suppresses Delayed-Type Hypersensitivity Reactions

Authors

  • J.-B. Sun,

    1. Institute of Biomedicine, Department of Microbiology and Immunology and Göteborg University Vaccine Research Institute (GUVAX), Sahlgrenska Academy at Göteborg University, Göteborg, Sweden
    Search for more papers by this author
  • N. Cuburu,

    1. INSERM U721, Nice, France
    Search for more papers by this author
  • M. Blomquist,

    1. Institute of Biomedicine, Department of Microbiology and Immunology and Göteborg University Vaccine Research Institute (GUVAX), Sahlgrenska Academy at Göteborg University, Göteborg, Sweden
    Search for more papers by this author
  • B.-L. Li,

    1. Institute of Biomedicine, Department of Microbiology and Immunology and Göteborg University Vaccine Research Institute (GUVAX), Sahlgrenska Academy at Göteborg University, Göteborg, Sweden
    Search for more papers by this author
  • C. Czerkinsky,

    1. INSERM U721, Nice, France
    2. International Vaccine Institute, Seoul, South Korea
    Search for more papers by this author
  • J. Holmgren

    1. Institute of Biomedicine, Department of Microbiology and Immunology and Göteborg University Vaccine Research Institute (GUVAX), Sahlgrenska Academy at Göteborg University, Göteborg, Sweden
    Search for more papers by this author

Dr J.-B. Sun, Department of Microbiology and Immunology, Göteborg University, Box 435, SE-405 30 Göteborg, Sweden. E-mail: jia-bin.sun@microbio.gu.se

Abstract

Although sublingual (s.l.) immunotherapy with selected allergens is safe and often effective for treating patients with allergies, knowledge of the immunological mechanisms involved remains limited. Can s.l. administration of antigen (Ag) induce peripheral immunological tolerance and also suppress delayed-type hypersensitivity (DTH) responses? To what extent can s.l.-induced tolerance be explained by the generation of Foxp3+CD25+CD4+ regulatory T cells (Treg)? This study addressed these questions in mice and compared the relative efficacy of administering ovalbumin (OVA) conjugated to cholera toxin B (CTB) subunit with administration of the same Ag alone. We found that s.l. administration of a single or even more efficiently three repeated 40-μg doses of OVA/CTB conjugate suppressed T-cell proliferative responses to OVA by cervical lymph node (CLN), mesenteric lymph node (MLN) and spleen cells and concurrently strongly increased the frequency of Ag-specific Treg in CLN, MLN and spleen and also transforming growth factor-β (TGF-β) levels in serum. The CLN and splenic cells from OVA/CTB-treated BALB/c mice efficiently suppressed OVA-specific T-cell receptor (TCR) transgenic (DO11.10) CD25CD4+ effector T-cell proliferation in vitro. Further, s.l. treatment with OVA/CTB completely suppressed OVA-specific DTH responses in vivo and T-cell proliferative responses in mice immunized subcutaneously with OVA in Freund's complete adjuvant. The intracellular expression of Foxp3 was strongly increased in OVA-specific (KJ1-26+) CD4+ T cells from OVA/CTB-treated mice. Thus, s.l. administration of CTB-conjugated Ag can efficiently induce peripheral T-cell tolerance associated with strong increases in serum TGF-β levels and in Ag-specific Foxp3+CD25+CD4+ Treg cells.

Introduction

Mucosal administration of soluble proteins can induce peripheral immunological tolerance, which plays a role in preventing or controlling unwanted immune responses to both self- and non-self-antigens (Ag). This strategy has attracted recent attention as a promising way to prevent or treat allergic, autoimmune or infection-induced immunopathological reactions [1, 2]. Most studies of mucosally induced tolerance have used oral or intragastric (i.g.) Ag administration, and hence the peripheral tolerance induced is commonly referred to as ‘oral tolerance.’ However, similar tolerance has also been achieved by administering Ag by nasal or rectal routes [1, 2]. Studies of mechanisms of mucosally induced tolerance in animal models have shown that oral or i.g. administration of excessive Ag doses can induce Ag-specific T-cell anergy or even depletion. When, on the other hand, lesser Ag dosages are given, oral tolerance is mediated largely by the mucosal induction of regulatory CD4+ T cells, which produce immunomodulating cytokines such as interleukin (IL)-10 and tumour growth factor (TGF)-β [1–4]. Several types of regulatory CD4+ T cells develop after oral administration of Ag [5–8]: Type 1 (Tr1) regulatory T cells secreting large amounts of IL-10 (and TGF-β), Th3 cells secreting predominantly TGF-β (and IL-10) and perhaps of most importance CD25+CD4+ regulatory T cells (Treg) expressing the transcription factor Foxp3 [1, 2]. The latter ‘induced Treg’ appear functionally to be the non-self-Ag-induced counterpart of the much more extensively studied ‘natural Treg’ cells, which comprise 5–10% of peripheral CD4+ T cells in healthy mice and humans and seem to have a crucial role in maintaining immune tolerance to self-Ag and in preventing the development of autoimmune disease [9–12]. Among several characteristics that have been attributed to the CD25+ Treg population, expression of Foxp3 appears to be critical for the generation and suppressive function of both induced and natural CD25+ Treg cells [13–15]. TGF-β plays an important role in the generation of natural as well as induced Treg cells [1, 2]. Recent evidence indicates that TGF-β is also required for generation of mucosal Treg in orally tolerized mice. Oral tolerance induction by i.g. administration of an effective Ag formulation leads to a markedly elevated level of TGF-β in serum, and conversely, depletion of TGF-β by specific monoclonal antibody (mAb) treatment completely suppresses the mucosal generation of CD25+ Treg in response to the i.g. Ag treatment [16].

We and others have shown that oral or nasal administration of relevant Ag conjugated to the non-toxic B subunit protein of cholera toxin (CTB) can in experimental models induce peripheral tolerance and suppress autoimmune diseases with an efficiency that greatly exceeds that achieved by treatment with Ag alone [17–22]. The tolerance induced by Ag conjugated to CTB is associated with increased IL-10 and TGF-β1 mRNA expression and cytokine production by mucosal regulatory T cells [19, 22]. Most strikingly, effective oral tolerization by Ag/CTB conjugate treatment has resulted in pronounced increases in Ag-specific Foxp3-expressing CD25+CD4+ Treg in draining and peripheral lymph nodes and in increased TGF-β levels in serum [16].

Recently, a specific form of mucosal tolerance induction by sublingual (s.l.) administration of allergen Ag has emerged as an especially promising form of non-invasive Ag-specific immunotherapy for the treatment of patients with allergic rhinitis. It is both efficacious and much safer than subcutaneous (s.c.) immunotherapy [23, 24]. A recent meta-analysis that evaluated s.l. immunotherapy (SLIT) in 22 clinical studies of 979 patients with allergic rhinitis to house dust mite, pollens and cat dander concluded that SLIT significantly reduced both symptoms and medication requirements [23] provided that high doses of allergen (i.e. 50- to 100-fold the s.c. dose) are administered. Whilst mainly used in allergic patients as a safer and more practical replacement of injection hyposensitization immunotherapy for generating blocking IgG and possibly IgA antibodies to allergens and possibly by also inducing a Th2 to Th1 cytokine shift, it is not known whether tolerance induction and specifically the generation of CD25+ regulatory T cells is associated with SLIT. It also is not known whether the s.l. route of mucosal Ag administration can suppress delayed-type hypersensitivity (DTH) inflammation and thus be potentially useful for immunotherapy of DTH-mediated allergic, autoimmune or infection-induced immunopathological reactions.

Therefore, in this study, we investigated the development of Ag-specific tolerance and Treg in response to s.l.-administered ovalbumin (OVA) given either conjugated to CTB (OVA/CTB) or alone to BALB/c mice that had adoptively received OVA T-cell receptor (TCR) transgenic (Tg) DO11.10 T cells as a means of increasing the frequency of Ag-specific T cells. We found that s.l. OVA/CTB treatment was much more efficient than OVA alone and even more efficient than oral OVA/CTB treatment in inducing peripheral tolerance. This treatment effectively suppressed Ag-specific DTH inflammation and was closely correlated with strongly increased levels of Ag-specific Foxp3+CD25+ Treg cells in both draining and peripheral lymph nodes and to increased levels of TGF-β in serum induced by the treatment.

Materials and methods

Mice.  For the experiments we used two strains of 6- to 8-week-old female mice: (i) BALB/c (B & K Universal AB, Stockholm, Sweden); and (ii) DO11.10 OVA TCR Tg mice on BALB/c background (Jackson Laboratory, Bar Harbor, ME, USA), a clone with nearly 50% of the CD4+ T cells expressing a TCR specific for the peptide323−339 fragment of OVA. The mice were kept in ventilated cages under specific pathogen-free conditions at the Department of Experimental Biomedicine, Göteborg University. Studies were approved by the Göteborg University Ethical Committee for Animal Experimentation.

Antigens and conjugation of OVA to CTB subunit.  Ovalbumin protein (grade VII) was purchased from Sigma (St Louis, MO, USA) and OVA peptide323−339 (ISQAVHAAHAEINEAGR) of >90% purity obtained from TAG Copenhagen A/S (Klampenborg, Denmark). Highly purified recombinant CTB was kindly provided by SBL Vaccin AB (Stockholm, Sweden). OVA protein was chemically coupled to CTB using N-succininmidyl (3-[2-pyridyl]-dithio) propionate (Pierce Biotechnology, Inc., Rockford, IL, USA) as a bifunctional coupling reagent as described [18]. Coupled OVA/CTB was purified by FPLC gel filtration [Superdex 200 16/60 column (Pharmacia Biotech) using the Biologic Workstation FPLC system (Bio-Rad Laboratories, Richmond, CA, USA)]. The purified conjugate was analysed by a GM1-ganglioside enzyme-linked immunosorbent assay (GM1-ELISA) [17] and shown to have retained GM1-binding activity and strong reactivity with antibodies to both OVA and CTB [16]. The conjugate also had strong capacity to induce OVA-specific T-cell proliferation when tested on DO11.10 splenocytes; further, in the latter assay, the activity of the conjugate was not significantly inhibited by pre-incubation and co-culture with polymyxin (10 μg/ml; Sigma) but was completely inhibited by pre-incubation and co-culture with highly purified GM1-ganglioside (10 nmol/ml; donated by late Prof. Lars Svennerholm).

S.l. or i.g. Ag administration and s.c. immunization.  BALB/c mice were first adoptively transferred with 1.5 × 107 DO11.10 T cells given intravenously (i.v.). Three or 4 days later, the mice were given either 40 μg of OVA or OVA/CTB conjugate S.l. either as a single dose or three times at 2-day intervals; controls received phosphate-buffered saline (PBS) s.l. For comparisons, other groups of mice received either a single 120-μg dose of OVA/CTB or three i.g. doses of either 40 or 200 μg of OVA/CTB. The s.l. administrations of the 40-μg doses were given in 10-μl volumes through a Finn pipette whereas the 120-μg dose was divided into three 10-μl aliquots and given at 10-min intervals. The i.g. administrations were given as described through a baby-feeding catheter with the Ag mixed with 0.3 ml of 6% (m/v) sodium bicarbonate.

In some experiments, mice were also immunized s.c. at the tail base 1 day after the last s.l. administration with 50 μg of OVA emulsified 1:1 in complete Freund's adjuvant (CFA) containing 100 μg of Mycobacterium tuberculosis H37Ra (Difco, Detroit, MI, USA). At indicated days after s.l. or i.g. treatment or s.c. immunization, the mice were sacrificed and draining cervical lymph nodes (CLN), mesenteric lymph nodes (MLN) or popliteal/inguinal nodes (PLN) and spleen were harvested for further in vitro studies and blood serum was prepared for TGF-β testing. Alternatively, the mice were challenged in vivo by an s.c. injection in one foot pad with OVA for DTH testing. In addition, in some cases, the sublingual mucosal tissue was excised and snap-frozen in liquid nitrogen for immunohistochemical studies.

T-cell proliferation and cytokine assays and adoptive transfer.  Spleen cells or cells from CLN or PLN were prepared as single-cell suspensions at indicated days after s.l. or s.c. immunization and erythrocytes were removed by lysis. For studies of proliferation of splenic and LN cells, 3 × 105 or 5 × 105 cells/well (in a 200-μl volume of Iscove medium supplememented with 10% fetal calf serum, 1%l-glutamine, 1% gentamicine and 50 μm mercaptoethanol) were cultured for 3 days in 96-well plates with or without 0.2–2 μg/ml of OVA323−339 peptide or 10 μg/ml OVA protein. [3H]-thymidine was added (1 μCi/well) for the last 16 h of culture and cells were harvested and [3H] incorporation was measured as described [21]. To test the regulatory function of CD4+ T cells from mice after different treatments, these cells were first isolated and then co-cultured in different numbers with 1 × 105 naïve DO11.10 CD25CD4+ effector T (Teff) cells together with 1 μg/ml OVA323−339 peptide as described [16]. These tests were performed using 96-well round-bottom plates (Nunc, Roskilde, Denmark) for 3 days with [3H]-thymidine added during the last 16 h before cell harvest and analysis of [3H] incorporation. Interferon-γ production by cultured cells was assayed in 3-day culture supernatants by ELISA. TGF-β levels in serum were also measured by ELISA (Duo Set kit; R&D Systems, Abingdon, UK). For adoptive transfer, a single T-cell suspension was prepared from all lymph nodes and spleen cells in naive DO11.10 mice by using a T-cell column (R & D Systems). We then i.v. injected 1.5 × 107 cells into the tail vein of naive BALB/c recipient mice.

Flow cytometry analyses.  Freshly isolated spleen or lymph node cells were incubated with fluorescein isothiocyanate- or phycoerythrin- or Allophycocyanin (APC)-labelled mAb to mouse CD4, CD25 or DO11.10 clonotypic TCR (KJ1-26) (BD Biosciences PharMingen, San Diego, CA, USA). For analysis of intracellular Foxp3, stained freshly isolated cells were fixed and permeabilized with Cytofix/Cytoperm solution (BD PharMingen, Franklin Lakes, NJ, USA) according to the manufacturer's suggested protocol and then incubated with APC-conjugated anti-Foxp3 FLK-16 mAb (Nordic Biosite, Taby, Sweden) (0.5–1 μg/106 cells) at 4 °C for 30 min in the dark. Cells were then washed and analysed by flow cytometry (Calibur FACS machine; BD PharMingen).

DTH.  BALB/c mice that were first adoptively transferred with OVA TCR Tg T cells, then s.l. treated three times with OVA/CTB, OVA or PBS, and 2 days after the last s.l. treatment immunized s.c. in the tail with 50 μg of OVA in 50 μl of CFA containing 100 μg of M. tuberculosis were 7 days later given an s.c. ‘challenge’ injection of 20 μg of OVA in PBS in the right footpad. Footpad thickness was measured before and 4 and 24 h after challenge in a blinded fashion using a calliper meter (Mitutoyo, Osaka, Japan).

Immunohistochemistry.  Frozen s.l. mucosal tissues were cut into 8-μm sections on a cryostat, fixed on glass slides with 100% acetone and stored at −70 °C until use. The slides were blocked for 30 min with 5% horse serum and then incubated with biotinylated anti-mouse CD11c (1/20) (BD PharMingen) for 1 h at room temperature before treatment with avidin–biotin–horseradish peroxidase complex (Vectastain ABC-HP kit; Vector Laboratories, Burlingame, CA, USA) for 30 min, and then overlaid with the chromogen substrate 3,3′-diaminobenzidine (Vector Laboratories). The sections were then washed with distilled water, counterstained with Mayerås haematoxylin, dehydrated and mounted with Mountex (Histolab, Göteborg, Sweden) before being examined by light microscopy.

Results

Sublingual administration of a single low dose of OVA/CTB conjugate suppresses T-cell responses and increases TGF-β

To examine if s.l. administration of Ag, alone or conjugated to CTB, could induce Ag-specific immunological tolerance, we first adoptively transferred 1.5 × 107 TCR Tg DO11.10 T cells to BALB/c mice to increase the frequency of OVA323−339 peptide-specific CD4+ T cells. We then treated the mice with a single s.l. dose of either 40 μg of OVA/CTB conjugate or 40 μg of OVA or PBS; 3 days later we tested the T-cell proliferative response in CLN and spleen cells to OVA323−339 peptide and also the ability of isolated CD4+ T cells to suppress Ag-specific proliferation of Teff cells (CD4+CD25 DO11.10 cells) in co-culture. We found that the OVA-specific T-cell proliferative response was almost completely suppressed in CLN (Fig. 1A) and partially suppressed in splenic cells (data not shown) in the mice given OVA/CTB s.l. compared with controls given PBS. T cell responses in OVA-alone treated mice were less suppressed than those in mice treated with OVA/CTB conjugate (Fig. 1A). Also, CLN CD4+ T cells from the OVA/CTB-treated mice were effectively suppressive when tested in co-culture with Teff cells stimulated with OVA323−339 peptide (Fig. 1B). In contrast, CLN CD4+ T cells from unconjugated OVA-treated mice were only partially suppressive and PBS-treated CLN cells were not suppressive at all when tested in a 1:1 ratio with Teff cells (Fig. 1B).

Figure 1.

 Sublingual (s.l.) administration of a single 40-μg dose of ovalbumin/cholera toxin B subunit (OVA/CTB) conjugate induces antigen (Ag)-specific suppressor activity in cervical lymph node (CLN) cells and increases transforming growth factor (TGF)-β levels in serum. Groups of BALB/c mice were adoptively transferred with 1.5 × 107 purified DO11.10 Tg T cells and 4 days later given a single s.l. 40-μg dose of OVA/CTB, OVA or phosphate-buffered saline (PBS). Three days later, CLN cells were isolated from the groups and their proliferative responses to stimulation with different concentrations of OVA323−339 peptide were tested (A) as was the ability of their purified CD4+ T cells to suppress the proliferative response to stimulation by the OVA323−339 peptide of co-cultured DO11.10 CD25 Teff cells (B). The levels of TGF-β in serum 1 day after the s.l. treatments were measured by ELISA (C). Results are shown as mean and SD for cells and sera from six mice per group, cells being pooled pairwise (three pairs per groups) and then tested in triplicate cultures. Results shown are from one of three experiments showing similar results. *, ** and *** denote P < 0.05, P < 0.01 and P < 0.001 by ANOVA (A) or Student's t-test (B, C) for indicated treatment groups in comparison with the PBS-treated group.

The levels of TGF-β in serum were also significantly increased in mice treated s.l. with OVA/CTB conjugate compared with mice treated with OVA alone or with PBS (Fig. 1C).

Sublingual administration of OVA/CTB conjugate increases total and OVA-specific Foxp3+CD25+CD4+ T cells in CLN and spleen

To determine if s.l. administration of Ag/CTB conjugate induces an Ag-specific Treg cell response, BALB/c mice were given DO11.10 T cells i.v. followed by s.l. treatment with a single 40-μg dose of OVA/CTB conjugate or PBS. Foxp3 is a key marker for identification and is critical for generation and maintenance of a functional CD4+CD25+ T-cell population [13–15]. One day after the s.l. treatment, we used fluorescence-activated cell sorter (FACS) analysis to assess the expression of total and Ag(OVA323−334 peptide)-specific (KJ1-26+) Foxp3 cells among CD4+ T cells from CLN and spleen from mice treated s.l. with OVA/CTB. As shown in Fig. 2, the expression of Foxp3 was significantly increased among the CD4+ CLN cells (Fig. 2A) and also among the OVA-specific (KJ1-26+) CD4+ CLN T cells (Fig. 2B) from mice given OVA/CTB compared with treatment with PBS alone. An increase in Foxp3+ total and OVA-specific CD4+ T cells was also seen in spleen cells (data not shown).

Figure 2.

 Increase in frequency of Foxp3+ cells among CD4+ T cells and KJ1-26+CD4+ cells 1 day after sublingual (s.l.) treatment with ovalbumin/cholera toxin B subunit (OVA/CTB). Groups of BALB/c mice were adoptively transferred with 1.5 × 107 purified DO11.10 Tg T cells and 4 days later given a single s.l. 40-μg dose of OVA/CTB or phosphate-buffered saline (PBS). One day later, the mice were killed. Cervical lymph node (CLN) cells were prepared and stained with monoclonal anitbodies (mAb) against different surface markers, followed by intracellular staining of Foxp3 and flow cytometry analysis. Results show proportions of Foxp3+ cells among gated CD4+ cells (left panels) and proportions of Foxp3+ KJ1-26+ cells among gated CD4+ KJ1-26+ cells (right panels). Values are percentages of indicated cell populations.

We next used FACS staining to determine total and OVA323−339 peptide-specific (KJ1-26+) CD25+CD4+ T cells. Both of these CD4+CD25+ T-cell populations were increased by more than twofold in CLN cells from mice treated s.l. with OVA/CTB 1 day earlier when compared with PBS-treated mice (data not shown).

In other groups of mice treated with a single 120-μg dose of OVA/CTB, we examined the frequencies of Treg in CLN and spleen cells at different times after treatment. The frequency of OVA-specific Foxp3+CD4+ T cells was strongly increased among lymphocytes from both CLN (Fig. 3A) and spleen (Fig. 3B) 1 day after treatment with OVA/CTB conjugate (compared with PBS treatment) and increased further at 3 and 7 days after treatment. On day 3, results showed that treatment with OVA/CTB was superior to treatment with OVA alone in inducing Foxp3+ Treg in both CLN (Fig. 3A) and spleen (Fig. 3B).

Figure 3.

 Sublingual (s.l.) administration of ovalbumin/cholera toxin B (OVA/CTB) conjugate efficiently increases the frequency of OVA-specific Foxp3+CD4+ T cells in cervical lymph node (CLN) and spleen. Groups of BALB/c mice were adoptively transferred with 1.5 × 107 purified DO11.10 Tg T cells and 4 days later given a single s.l. 120-μg dose of OVA/CTB or OVA alone or phosphate-buffered saline (PBS). Mice were killed 1, 3 or 7 days later and CLN and spleen cells were prepared and stained with monoclonal antibody (mAb) against different surface markers, followed by intracellular staining of Foxp3 and analysis by flow cytometry (cells from OVA-treated mice were only examined after 3 days; middle bars). Results show percentages (mean and SD values for three mice at each time point) of Foxp3+ KJ1-26+ cells after gating on CD4+ CLN (A) or CD4+ spleen lymphocytes (B).

Sublingual administration of three low doses of OVA/CTB conjugate further suppresses T-cell responses and increases the frequencies of total and OVA-specific Foxp3+CD4+ T cells

We next sought to determine whether repeated administration of Ag/CTB would be more efficient than single-dose administration and if s.l. administration would compare favourably to the more commonly used i.g. route for inducing tolerance. To this end, BALB/c mice transferred with DO11.10 T cells were given three 40-μg OVA/CTB doses every second day either s.l. or i.g.; other groups were given three 40-μg doses of OVA s.l., three 200-μg doses of OVA/CTB i.g. or PBS s.l. Three days after the last treatment, CLN and MLN cells were isolated and examined for T-cell proliferation in response to OVA stimulation in vitro and also analysed for expression of Treg cells by FACS staining. As shown in Fig. 4A, MLN cells from mice given either three 40-μg doses of OVA/CTB s.l. or three 200-μg doses i.g. completely suppressed the proliferative response of an equal number of DO.11.10 Teff cells stimulated with OVA in co-culture in vitro. Three i.g. treatments of 40-μg of OVA/CTB partially suppressed the T-cell response, but to a lesser extent than the corresponding low-dose OVA/CTB s.l. treatment. In contrast, cells from OVA-treated mice were not detectably suppressive when tested in a 1:1 ratio with Teff cells (Fig. 4A).

Figure 4.

 Sublingual (s.l.) administration of three ovalbumin/cholera toxin B (OVA/CTB) conjugate doses suppresses T-cell response and increases total and OVA-specific Foxp3+CD4+ T cells. Groups of BALB/c mice received 1.5 × 107 DO11.10 T cells i.v. before s.l. treatment with three 40-μg doses of OVA/CTB conjugate, OVA or phosphate-buffered saline (PBS) every second day or, for comparison, three i.g. doses of 40 or 200 μg of OVA/CTB at corresponding times. Three days after the last treatment, mesenteric lymph node (MLN) cells were examined for their ability in co-culture with an equal number of DO11.10 T cells to suppress T-cell proliferation in response to OVA323−339 peptide stimulation in vitro (A). Some of the s.l.-treated mice were also analysed for expression of Foxp3 among gated KJ1-26+ and KJ1-26 CD4+ T cells from spleen (B). Results in (A) show mean and SD values for cells from six mice per treatment group, cells being pooled pairwise (three pairs per group) and then tested in triplicate cultures; * and *** denote P < 0.05 and P < 0.001 by Student's t-test for indicated treatment groups in comparison with the PBS-treated group. (B) Distributions of stained cells. Values are percentages of indicated cell populations.

To further test if the suppression of Teff responses in co-culture was associated with induction of Treg cells, we examined the expression of Foxp3 in T cells KJ1-26+ and KJ1-26 CD4+ from CLN and MLN of mice treated with OVA/CTB, OVA or PBS. We found that the expression of Foxp3 was markedly increased in both KJ1-26+ and KJ1-26 CD4+ T cells of CLN and spleen in response to three s.l. 40-μg treatments with OVA/CTB compared with treatment with OVA or PBS alone as illustrated for spleen cells in Fig. 4B.

Sublingual administration of OVA/CTB conjugate suppresses in vivo OVA-specific DTH reactivity and T-cell proliferative responses in primed animals

Peripheral immunological tolerance after mucosal administration of Ag is characterized by a decreased ability to mount a systemic immune response to a parenteral immunization with the same Ag [1, 2]. We tested the effect of s.l. treatment with OVA/CTB on in vivo immunological responses, including DTH responses in immunized mice. Four days after adoptive transfer with OVA Tg T cells, BALB/c mice were given three 40-μg doses s.l. of OVA/CTB, OVA alone or PBS every second day; 2 days after the last treatment they were injected s.c. at the base with 50 μg of OVA in CFA. Seven days after the s.c. immunization, half of the mice were challenged by an s.c. injection of 20 μg of OVA in 20 μl of PBS in the right footpad. Footpad thickness was measured before and 4 and 24 h after challenge and differences were calculated (six mice per group). From the other half of the mice, CLN and PLN cells were collected and examined for T-cell proliferation in response to OVA stimulation in vitro. The results are shown in Fig. 5. There was partial suppression of DTH at both 4 h (data not shown) and at 24 h in mice treated s.l. with OVA alone and complete suppression in mice treated s.l. with OVA/CTB conjugate (Fig. 5A). There was also significant suppression of the T-cell proliferative response to OVA323−339 peptide in vitro in cultures of PLN cells from mice treated s.l. with OVA/CTB (Fig. 5B).

Figure 5.

 Sublingual (s.l.) administration of ovalbumin/cholera toxin B subunit (OVA/CTB) conjugate efficiently suppresses in vivo T-cell responses to s.c. immunization with OVA. (A) Delayed-type hypersensitivity (DTH) response. Groups of BALB/c mice were i.v. injected with 1.5 × 107 purified DO11.10 Tg T cells and starting 4 days later given three s.l. 40-μg doses of either OVA/CTB, OVA or phosphate-buffered saline (PBS). The next day, these mice were immunized s.c. with 50 μg of OVA in complete Freund's adjuvant (CFA) at the tail base; 7 days later they were challenged by an s.c. injection of 20 μg of OVA in 20 μl of PBS in the right footpad. Footpad thickness was measured before and at 4 h (not shown) and 24 h after challenge and differences were calculated (six mice per group). (B) T-cell proliferative response. Popliteal/inguinal nodes (PLN) cells from half of the mice treated, as in (A) but not challenged and tested for DTH, were isolated and cultured at 3 × 105 cells/well with 10 μg/ml of OVA. The cells were incubated for 72 h; [3H]-thymidine was added to the cell cultures 16 h before the cells were harvested and examined for [3H] incorporation in DNA by scintillation. Results in (A) and (B) are mean and SD of individual mice (six mice per group). Data are from one of two independent experiments giving similar results. * denotes P < 0.05 and ***P < 0.001 for indicated treatment groups in comparison with mice treated with PBS.

Sublingual administration of OVA/CTB generates strong local infiltration of CD11c+ dendritic cells in the s.l. mucosa

It is likely that the uptake of Ag (especially Ag conjugated to CTB) across the epithelial barrier and the subsequent uptake and processing by local dendritic cells or other APC are critical events in the induction of an immune response, whether tolerizing or immunizing. Thus, we used immunohistochemical methods to compare the effect of s.l. administration of OVA/CTB with OVA alone or PBS on the cellular composition of the s.l. mucosa. To this end, we treated mice s.l., after the standard adoptive transfer of DO11.10 Tg T cells, with 120 μg of OVA/CTB or OVA or PBS. One day later we collected and froze the s.l. mucosal tissue. Frozen tissues were sectioned and examined by immunohistochemical staining for CD11c+ DC, T cells and B cells.

In all treatment groups, the s.l. mucosa contained very few lymphocytes, and after administration of OVA or PBS, very few CD11c+ DC; however, the latter cell type was abundant in the s.l. mucosa of OVA/CTB-treated mice (Fig. 6).

Figure 6.

 Sublingual (s.l.) administration of ovalbumin/cholera toxin B subunit (OVA/CTB) conjugate generates strong local infiltration of CD11c+ DC in the s.l. mucosa. Groups of BALB/c mice were adoptively transferred with DO11.10 Tg T cells as described in Fig. 1. Four days later they were given a single s.l. 120-μg dose of PBS (A), OVA (B) or OVA/CTB (C). One day later, mice were killed and s.l. mucosal tissue was collected. Frozen tissue was sectioned and examined by immunohistochemical staining for CD11c+ DC (brown cells).

Discussion

To our knowledge, our study is the first to show that s.l. administration of a tolerizing formulation of a protein Ag effectively induces Ag-specific regulatory T cells in both draining and peripheral lymph nodes, predominantly typical Foxp3+CD25+CD4+ Treg. Similar to our previous findings in other systems in which we used oral or nasal administration [16–22], s.l. tolerance induction with Ag/CTB conjugate was superior to that induced with free Ag both in increasing the efficacy of tolerization and in dramatically reducing the minimal effective amount of Ag. Thus, the s.l. route adds to the mucosal routes that can be used to induce peripheral immunosuppression associated with and probably largely mediated by Foxp3-expressing regulatory T cells. However, it is notable that in comparison with the best-studied route (oral/i.g.) for inducing oral tolerance the same amount of OVA/CTB conjugate given s.l. was more efficient in suppressing Ag-specific T-cell responses and in inducing Treg development.

The peripheral immunological tolerance induced by s.l. Ag/CTB was associated with marked increases in TGF-β levels in serum. This finding is consistent with several reports indicating that TGF-β is important in the induction and expression of oral tolerance by T cells [1, 2]. TGF-β has also been identified as a critical cytokine for both the induction and suppressive function of natural and induced Treg [13, 25, 26]. It was preliminarily reported by Cuburu et al. [27] that s.l. administration of OVA linked to or even admixed with CTB can induce systemic tolerance and TGF-β secretion. In our study, the tolerance induction by s.l. administration of OVA/CTB and the increased levels of TGF-β1 in serum correlated closely with the increase in Treg cells in draining local (CLN) and systemic (spleen and MLN) lymphonodes. This correlation is in line with our recent finding that in vivo depletion of TGF-β by specific mAb treatment completely suppressed the generation of CD25+ Treg cells in response to i.g. Ag/CTB treatment [16].

We propose that the efficient induction of Treg by s.l. administration of Ag conjugated to CTB, unlike administration of Ag alone, is probably largely explained by the combined effects of enhanced receptor-mediated Ag uptake by the CTB molecule, first across the epithelial barrier and then into subepithelial APC [28]. This explanation would imply that the s.l.-administered Ag–CTB conjugate mainly targets and/or recruits local APC directly underneath the s.l. epithelium. In support of this, we found that 1 day after the s.l. administration of OVA/CTB, the s.l. mucosa contained a large number of CD11c+ cells, most likely DC, in contrast to the scarcity of such cells seen after PBS or OVA-alone treatment. No such cells were found in the s.l. mucosa after i.g. administration of OVA/CTB. These findings and the lesser effect of i.g. treatment with the same dose of OVA/CTB rule out the possibility that the tolerizing effect after s.l. Ag/CTB administration can be explained by an intestinal effect induced by a small amount of swallowed conjugate in mice receiving s.l. treatment. Our previous work showed that conjugation of Ag to CTB greatly facilitates Ag uptake and MHC class II-restricted Ag presentation by CD11c+ DC [29]. One possibility is that mucosally administered Ag/CTB conjugate preferentially binds to and acts on tolerogenic subsets of mucosal DC or other APC. Anjuere et al. [30] reported a selective increase in CD11c+CD8α+ B220+ DC in MLN following CTB feeding as well as the ability of this DC population from CTB-fed mice to support the differentiation of CD4+ Ag-specific regulatory T cells producing TGF-β and IL-10. We have further noted that the increase in frequency of Treg cells in Peyer's patches and MLN correlates closely with an increase in this DC subset 2 h to 2 days after i.g. treatment with OVA/CTB [16]. Work is in progress to determine the presence and role of this DC subset and other APC in the s.l. mucosa in response to s.l. Ag/CTB administration.

Our study also shows that s.l. administration of OVA/CTB, and to a lesser extent OVA alone, efficiently suppresses peripheral T-cell responses and DTH inflammation in vivo. Although SLIT has to date only been used to treat patients with IgE-mediated allergies [23, 24], our findings suggest that the s.l. route can be used for administering tolerance-inducing immunotherapy against DTH-induced inflammatory disorders that are causally associated with many autoimmune diseases or infection-induced immunopathological reactions. There is also evidence that the s.l. route can efficiently generate an Ag-specific protective immune response rather than tolerance, provided that the administered Ag is given with an immunostimulating agent such as cholera toxin [27]. This means that in addition to the already established use of s.l. immunotherapy in selected type-I allergic conditions, the s.l. route should be further examined as a promising means for inducing oral tolerance immunotherapy against DTH-mediated allergic, autoimmune or infection-related disorders or with different Ag and/or formulations for inducing vaccine protection against pathogens.

In conclusion, we demonstrated efficient Ag-specific suppression in vitro and in vivo of Th1/DTH T-cell responses by s.l. administration of Ag conjugated to CTB. Such s.l. tolerance induction is associated with rapid local accumulation of CD11c+ DC in the s.l. mucosa and with induction of Ag-specific Treg cells in both draining and peripheral lymph nodes and spleen and with increased TGF-β levels in serum. Our findings may lead to development of novel immunotherapeutic strategies for preventing Th1-mediated inflammation and may help explain the beneficial effects of SLIT in the treatment of patients with allergies.

Acknowledgment

This work was supported by research funding from the Swedish Science Council (Medicine) (project K2000-06X-03382) and by The Knut and Alice Wallenberg Foundation through its support of The Gothenborg University Vaccine Research Institute (GUVAX). The authors thank Marianne Lindblad for preparing and characterizing the OVA/CTB conjugate used in the studies.

Ancillary