Interleukin-10-secreting Peyer's patch cells are responsible for active suppression in low-dose oral tolerance

Abstract

We demonstrate the induction of antigen-specific interleukin-10 (IL-10)-secreting cells in murine Peyer's patches (PPs) after low-dose β-lactoglobulin (BLG) feeding. In addition, we show that PP cells can inhibit the T-cell proliferative response in vitro as well as T-cell-mediated inflammation in vivo. The active suppression mediated by these regulatory cells was seen only within a narrow range of antigen dosage (feeding), with the most prominent effect at 5 × 1 mg BLG. On either side of this range, T-helper 1-like cytokine responses were observed when PP cells were stimulated with antigen in vitro. This result correlated with reduced production of regulatory cytokines as well as reduced activity of bystander suppression. We found that changes in IL-10 production correlated inversely with changes in interferon-γ production. Inhibitory effects mediated by CD4⁺ PP cells were partially neutralized by antibodies to IL-10 and transforming growth factor-β. Interestingly, the generation of such regulatory cells after low-dose BLG feeding exhibited organ dependence. Among spleen, lymph node and PP cells derived from orally tolerized mice, PP cells were the most effective in promoting bystander suppression in the presence of BLG, indicating the significance of PPs as an inductive site for antigen-specific regulatory cells upon induction of low-dose oral tolerance. Moreover, PP cells from mice fed 5 × 1 mg BLG were shown to suppress a BLG-specific delayed-type hypersensitivity response induced in footpads, suggesting that IL-10-secreting PP cells regulate systemic inflammation.

Introduction

Oral tolerance is defined as a state of immunological unresponsiveness to a specific antigen after exposure to that antigen by the enteric route.^1,2 Given that this approach could be applied to the intervention of diseases associated with adverse immune responses, such as allergy and autoimmune diseases, the cellular mechanisms of oral tolerance have been extensively studied.^3,4

During the development of oral tolerance, the dose of antigen fed to animals is critical. Depending on the antigen dose, antigen-specific cells are stimulated in different ways, resulting in different states of tolerance. Previous studies have shown that multiple administration of antigen at low doses induces cell populations that actively suppress bystander T-cell proliferation, whereas high doses induce a state of anergy as well as deletion of antigen-specific T cells.^5–7 Further, low-dose feeding induces prominent secretion of interleukin-4 (IL-4), IL-10, and transforming growth factor-β (TGF-β) whereas high-dose oral tolerance elicits only nominal secretion of these cytokines.⁸ These findings indicate that inhibitory cell types previously described by characterization of T-cell clones, such as TGF-β-secreting T-helper (Th)3 cells and/or IL-10-secreting T-regulatory (Tr)1 cells, may play roles as active suppressors in peripheral tolerance after low-dose antigen feeding.^9–11

The existence and location of regulatory cells that affect T cells in vivo (as well as the cytokines that are secreted at different stages of oral tolerance) remain to be addressed in physiological terms. In order to help resolve these issues and to obtain information about effective doses for applying the approach to the prevention of milk allergy in the future, we addressed the conditions and cellular mechanisms of low-dose oral tolerance against β-lactoglobulin (BLG), a potent milk allergen. Our results demonstrate that BLG-specific/IL-10-secreting cells in Peyer's patches (PPs) are induced following BLG feeding, and that these cells inhibit the T-cell proliferative response in vitro and T-cell-mediated inflammation in vivo.

Materials and Methods

Mice

BALB/c mice were purchased from Charles River Japan (Tokyo, Japan) before weaning (2 weeks of age) and were fed a commercial diet (MM3; Funabashi Farm, Chiba, Japan) not containing bovine β-lactoglobulin (BLG). They were 6–8 weeks old at the start of the experiments. All animals received humane care as outlined in the Guide for the Care and Use of Experimental Animals (National Institute of Animal Industry Animal Care Committee).

Antigen

BLG (genetic variant A) was prepared from fresh milk by the method of Aschaffenburg and Drewry and purified by ion-exchange chromatography.¹² Purity was confirmed by gel electrophoresis. Ovalbumin (OVA) was purchased from Sigma (St Louis, MO).

Induction of oral tolerance and immunization

BLG (0·1, 0·5, 1, 5, or 10 mg) in saline (or saline alone) was fed to mice five times in 10 days via gastric intubation with an animal-feeding needle. Seven days after the last feeding, the animals were killed for the preparation of lymph node (LN), spleen (SP) and PP cells. For T-cell proliferation assays, mice were immunized in the hind footpads and base of the tail with 40 µg of BLG in 50 µl of complete Freund's adjuvant containing Mycobacterium tuberculosis H37Ra (Difco, Detroit, MI) 7 days after the last feeding. LN cells were harvested 7 days after immunization.

Preparation of cells

Preparation of LN, SP, and PP cells

LN, SP and PP cells were harvested from mice given different doses of BLG, or from control mice. Single-cell suspensions of LN cells from periaortic, popliteal and inguinal LNs were prepared by pressing the isolated nodes through a 200 mesh polyester screen with the plunger from a 5-ml polypropylene syringe. Cells were washed three times with phosphate-buffered saline (PBS). After each wash, cells were centrifuged at 300 g for 10 min at 4°. Single-cell suspensions of SP cells were prepared in the same manner, and erythrocytes were depleted using a red blood cell lysing solution (Sigma) followed by a wash with RPMI-1640 containing 10% fetal calf serum (FCS). PBS containing 1% FCS was used for the following two washes. PPs were dissected from the small intestine and were washed in PBS containing 1% FCS. Without further treatment, PPs were pressed through a 200 mesh polyester screen with the plunger from a polypropylene syringe to make a single-cell suspension. PP cells were washed four times with PBS containing 1% FCS and kept on ice until use on the same day.

Preparation of CD4+ T cells

To remove adherent cells, PP cultures were incubated on plastic culture dishes for 60 min at 37° in a 5% CO₂ atmosphere. From non-adherent cells, CD4⁺ T cells were negatively selected with anti-Ia (M5/114, Pharmingen, San Diego, CA) and anti-CD8 (53-6.7, Pharmingen) antibodies, and subsequently removed from solution using anti-rat immunoglobulin G (IgG)-conjugated magnetic beads and a magnetic separator (BioMag; PerSeptive Biosystems, Framingham, MA). Briefly, cells were incubated with antibody-conjugated magnetic beads on ice for 30 min in a 25-cm² culture flask (Falcon, Becton Dickinson, Franklin Lakes, NJ), and then set on a magnetic separator for another 15 min on ice. Magnetic separation was then repeated. Selected CD4⁺ T cells were washed three times with PBS prior to use for in vitro assays.

T-cell proliferation assay

Seven days after immunization, a single cell suspension was prepared from periaortic, popliteal and inguinal LN as described above. Pooled LN cells from five mice (3 × 10⁵ cells in 0·2 ml) were seeded into round-bottom 96-well plates and stimulated with different concentrations of BLG. The culture medium was RPMI-1640 supplemented with 1% normal BALB/c mouse serum, 100 U/ml penicillin, 100 µg/ml streptomycin, 10 mm HEPES, and 50 µm 2-mercaptoethanol (2-ME) (Gibco, Life Technologies, Rockville, MD). After 72 hr of incubation at 37° in a 5% CO₂ atmosphere, the cells were pulsed with 1 µCi [³H]thymidine per well. After 16–20 hr, the cells were harvested, and the radioactivity was quantified in a liquid scintillation counter (Tri-Carb 1600TR, Packard, Meriden, CT).

In vitro assay for assessing active suppression

LN, SP and PP cells were prepared as described above from mice given different doses of BLG (without systemic immunization). BLG-specific CD4⁺ T-cell clones H1.1 or 2.11G, derived from LNs of BLG-immunized BALB/c mice,¹³ were activated in the presence of primary culture cells.

Freshly isolated primary culture cells (5 × 10⁶ cells/well) and T-cell clones (H1.1 or 2.11G) were co-cultured either directly in a 24-well culture plate (Falcon, Becton Dickinson, NJ) or separately using culture inserts (with 0·45 µm filters) for 24-well plates (Intercup; Sanko, Tokyo, Japan). The culture medium was RPMI-1640 supplemented with 10% FCS (Gibco), 100 U/ml penicillin, 100 µg/ml streptomycin, 10 mm HEPES, and 50 µm 2-ME. In both co-cultures, T-cell clones (1 × 10⁵ cells/well) were placed in each well in the presence of 50 µg/ml of BLG together with 5 × 10⁶ spleen cells treated with mitomycin C (Wako, Osaka, Japan) as antigen-presenting cells. Freshly isolated LN, SP, or PP cells were either placed directly in the well or seeded in culture inserts.

After 48 hr of incubation at 37° in a 5% CO₂ atmosphere, the cells from the lower chamber were pelleted by centrifugation and resuspended in 400 µl fresh culture medium containing 10% FCS. One hundred microlitres of each culture was transferred to 96-well plates and pulsed with 1 µCi [³H]thymidine per well for additional overnight culture. After 16 hr, the cells were harvested, and the radioactivity was quantified in a liquid scintillation counter (Tri-Carb 1600TR, Packard, CT). The net proliferation of T-cell clones (H1.1 or 2.11G) or of a mixture of T-cell clones and primary culture cells for the separate co-culture system and direct co-culture system, respectively, was thus analysed.

For the neutralization assay, CD4⁺ PP T cells from mice fed BLG (1 mg × 5) and 2.11G were cultured separately in the presence of neutralizing antibodies. Neutralizing antibodies against IL-10 (Pharmingen), TGF-β (Genzyme, Cambridge, MA), and isotype-matched control antibodies (Pharmingen) were used at a concentration of 30 µg/ml.

Cytokine analysis

PP cells (5 × 10⁶ cells) were cultured in the presence of BLG (50 µg/ml) for 4 days in a 24-well plate, and the culture supernatants were collected. The culture medium was RPMI-1640 supplemented with 1% normal BALB/c mouse serum, 100 U/ml penicillin, 100 µg/ml streptomycin, 10 mm HEPES, and 50 µm 2-ME. To detect the cytokines in the supernatant of the primary culture cells, we used commercial kits for TGF-β (R & D, Minneapolis, MN), and specific antibody sets for IL-4, IL-10 and interferon-γ (IFN-γ) (Pharmingen). The concentration of TGF-β was analysed according to the manufacturer's instructions. As a customized analysis for IL-4, IL-10 and IFN-γ, titre plates (MaxiSorp F96; Nunc, Roskilde, Denmark) were coated with 3 µg/ml capture antibodies in 0·1 m sodium carbonate buffer (pH 9·5) for anti-IL-4, or 0·2 m sodium phosphate (pH 6·5) for anti-IL-10, and incubated overnight at 4°. Titre plates were then washed and blocked with 1% bovine serum albumin for 2 hr at room temperature. After washing, cell culture supernatants were added to wells for another overnight incubation at 4°. The wells were washed and treated sequentially with 1 µg/ml of biotinylated detecting antibodies in PBS containing 0.05% Tween 20 (PBST) (nacalai tesque, Kyoto, Japan) for 2 hr and in avidin–peroxidase conjugate (Southern Biotechnology Associates, Birmingham, AL) in PBST for 30 min at room temperature. A solution of 3,3′,5,5′-tetramethylbenzidine (TMB; KPL, Gaithersburg, MD) was used as substrate, and 0·18 m H₂SO₄ was used to stop the enzymatic reaction. The resulting colour was measured at 450 nm on a microplate reader (BioRad 3550, Hercules, CA).

Delayed-type hypersensitivity (DTH) analysis

BALB/c mice were injected intravenously (i.v.) with PP cells in 100 µl of saline (1 × 10⁷ cells/animal), derived from mice given BLG (1 mg × 5) or from control mice who did not receive BLG. Control mice were injected with saline alone. Within 30 min, 2.11G (1 × 10⁶ cells/foot) and BLG (50 µg/foot) were injected into these mice in the left hind footpads to elicit a DTH reaction.¹⁴ Footpad swelling was measured with a dial calliper (Ozaki Co., Ltd, Tokyo, Japan) 52 hr after the injection. The right hind footpads were injected with saline as controls.

Statistics

Statistical evaluation was performed using the two-tailed Student's t-test.

Results

Effective feeding dosage for induction of oral tolerance

The induction of oral tolerance is both antigen- and dose-dependent. BALB/c mice were fed different doses of BLG antigen, and LN cells from these mice after systemic immunization were tested for their proliferative response to BLG (Fig. 1). BLG-specific oral tolerance was induced at doses of 5 × 1 mg or higher. Feeding at levels below 1 mg BLG resulted in partial inhibition of the T-cell proliferative response.

Figure 1.

T-cell proliferative response to BLG after oral administration of different doses of BLG. BLG was administered orally to BALB/c mice at various doses (0 or 5 times; 0·1 mg, 0·5 mg, 1 mg, 5 mg, 10 mg, or 100 mg BLG). Mice were immunized in the footpads and the tail base with BLG in complete Freund's adjuvant. On day 7, inguinal, popliteal and periaortic lymph nodes were removed, and single cell suspensions were prepared as described in the Materials and methods with the indicated concentration of BLG, or 50 µg/ml of control proteins (OVA, ovalbumin; LYS, lyzosyme; BCN, β-casein). The proliferative response was analysed by measuring [³H]thymidine incorporation. Values are the arithmetic mean of triplicate cultures (SEs were < 10% of means). Representative data from different experiments are shown. Feeding conditions are indicated next to each symbol in the graph.

Active suppressors are induced in PP after low-dose feeding of BLG

As an indicator of active bystander suppression, we analysed the proliferative response of H1.1 (a BLG-specific T-cell clone) co-cultured with either SP, LN, or PP cells from orally tolerized mice. This procedure allowed us to assess whether the low response of T-cell proliferation in vitro resulted from the induction of regulatory cells, and to identify the location of those regulatory cells in vivo. As shown in Fig. 2, PP cells were the most effective in suppressing H1.1 proliferation. The induction of active suppression occurred within a narrow dosage range, with the most effective dose being 5 × 1 mg BLG. The feeding dose of 5 × 0·5 mg BLG was also effective when PP cells were cultured separately in culture inserts. Concurrent with previous findings,⁵ these cell populations were not effectively induced when mice were fed higher doses of BLG. These results suggest that the low proliferative response with LN cells at high feeding doses (Fig. 1) may be due to T-cell anergy and/or T-cell clonal deletion. Interestingly, PP cells from control mice (not fed BLG) also showed moderate suppression, indicating that inhibitory cytokines are secreted constitutively by PP cells in normal mice. This non-specific suppression was also observed in co-cultures of H1.1 with SP cells, but not in co-cultures with LN cells.

Figure 2.

Suppression of the proliferation of a T-cell clone by primary culture cells from orally tolerized mice. SP cells, LN cells and PP cells from mice fed different doses of BLG (0 or 5 times 0·1 mg, 0·5 mg, 1 mg, 5 mg, or 10 mg BLG) were harvested 7 days after the last feeding. In the presence of BLG (50 µg/ml), the cells were co-cultured either directly or separately in culture inserts as described in the Materials and methods. The proliferative response of H1.1 was measured by analysing [³H]thymidine incorporation in the harvested cells. The proliferative response of control culture (Cont) was determined by the BLG-stimulated culture of H1.1 without primary culture cells. Representative data from three independent experiments are shown.

PP cells from mice fed 5 × 1 mg of BLG regulate antigen-specific DTH response

To test whether PP cells with suppressive effects on T-cell proliferation in vitro are also able to down-regulate the inflammatory response in vivo, we injected BALB/c mice i.v. with PP cells from mice fed BLG (5 × 1 mg) or from control mice, followed by injection of 2.11G and antigen. The protocol is a model of ex vivo induction of BLG-specific DTH.¹⁴ As shown in Fig. 3, footpad swelling was significantly decreased by transfer of PP cells from mice fed BLG, whereas PP cells from saline-fed mice had no effect on the localized inflammation.

Figure 3.

Suppression of DTH response by PP cells from orally tolerized mice. PP cells from orally tolerized mice that received 5 × 1 mg of BLG or saline were transferred i.v. to BALB/c mice (five mice/group). Concomitantly, mice were treated with 2.11G to induce a BLG-specific DTH response in footpads. Thickness of footpads was measured 52 hr after elicitation of a DTH response. Representative data from three independent experiments are shown.

IL-10 is produced in abundance by PP cells from mice fed 5 × 1 mg BLG

Suppression in vitro may be ascribed to soluble factors, since cell–cell contacts were not required to achieve suppression. Thus, we analysed regulatory cytokines produced by PP cells following in vitro culture in the presence of BLG. Figure 4 shows the levels of the inhibitory cytokines IL-4, IL-10 and TGF-β found in culture supernatants. A large amount of IL-10 was produced in the culture of PP cells derived from mice fed 5 × 1 mg BLG, suggesting that these BLG-specific cells may be responsible for the active suppression of the T-cell response observed in vitro and in vivo. TGF-β and IL-4 levels at 5 × 1 mg BLG did not differ significantly from the levels measured in the absence of BLG feeding. However, there was a strong correlation between the increase in IL-10 and the decrease in IFN-γ at all doses of BLG.

Figure 4.

Regulatory cytokines produced in vitro by PP cells from mice fed with BLG. PP cells were taken from mice fed different doses of BLG at 7 days after the last administration, and cultured in BLG as described in the Materials and methods. After 4 days of incubation, the concentrations of IL-4, IL-10, TGF-β and IFN-γ in culture supernatants were analysed by ELISA. Concentrations of IL-4, TGF-β and IFN-γ are shown in pg/ml (left ordinate), and IL-10 values are shown in ng/ml (right ordinate). Values are arithmetic means of triplicate assays (SEs were < 10% of means). Representative data from three independent experiments are shown.

Antibodies to IL-10 and TGF-β partially neutralize the inhibitory effect of PP cells on T-cell proliferation

To test whether regulatory cytokines detected in the PP cell culture are responsible for the observed active suppression of the T-cell response, we added neutralizing antibodies directed against IL-10 and TGF-β to the culture media of in vitro active suppression assays. We focused on the contribution of CD4⁺ T cells in these assays, as CD8⁺ T cells and adherent cells were present in small numbers only. Figure 5 shows that both anti-IL-10 and anti-TGF-β partially neutralized the suppressive effect of CD4⁺ PP cells, suggesting that both cytokines may be responsible, at least in part, for the inhibitory effect of PP cells on the T-cell response.

Figure 5.

Neutralization of CD4⁺ PP cell-mediated suppression by anti-IL-10 and anti-TGF-βCD4⁺ PP cells from orally tolerized mice fed 5 × 1 mg of BLG were taken 7 days after the last oral administration of BLG. In the presence of 10% FCS, BLG (50 µg/ml) and neutralizing antibodies (anti-IL-10, anti-TGF-β, or control antibodies), cells (5 × 10⁶ cells/well) were seeded in culture inserts and cocultured with a BLG-specific T-cell clone (2.11G; 1 × 10⁵ cells/well) separately in 24-well plates. The proliferative response of 2.11G was measured by analysing [³H]thymidine incorporation in harvested cells. Significant values at P < 0·01 are indicated by asterisks. Representative data from three independent experiments are shown.

Antigen-specific regulatory cells are also induced in PPs from mice fed 5 × 1 mg OVA

In order to test whether the generation of IL-10-secreting cells is antigen-specific during bystander-T-cell suppression, we fed BALB/c mice with OVA at doses similar to those previously used for BLG. PP cells from these OVA-fed mice were then tested using the in vitro active suppression assay and the BLG-specific T-cell clone H1.1. Figure 6 shows that administration of a low dosage of OVA (5 × 1 mg) resulted in the generation of active suppressor cells. These results are very similar to those obtained when mice were fed BLG.

Figure 6.

Suppression of T-cell clone proliferation by PP cells from OVA-fed mice. PP cells from mice fed different doses of OVA (0 or 5 times 0·1 mg, 1 mg, or 10 mg of OVA) were harvested 7 days after the last feed. In the presence of OVA (50 µg/ml) and BLG (50 µg/ml), cells (5 × 10⁶ cells/well) were co-cultured in 24-well plates with a BLG-specific T-cell clone (H1.1; 1 × 10⁵ cells/well) separately in culture inserts. The proliferative response of 2.11G was measured by analysis of [³H]thymidine incorporation in harvested cells. The proliferative response of control cultures (Cont) was determined by the BLG-stimulated culture of H1.1. Representative data from three independent experiments are shown.

Discussion

Although the importance of PPs in the induction of oral tolerance has been previously suggested, it is not yet clear where and what kind of immunoregulatory cells mediate this response. Our present study demonstrates that PP cells promote active suppression of bystander T cells in vitro, and that PP cells predominate in this aspect (relative to SP and LN cells; Fig. 2). These results suggest that, in physiological terms, the regulatory cells involved in active suppression are primarily induced in PPs.

The regulatory functions of PP cells observed in the present system seem to depend largely on their ability to secrete IL-10. A high level of IL-10 was produced by PP cells after oral administration of BLG (at a dose effective for inhibition of T-cell proliferation; Fig. 4). Furthermore, an antibody to IL-10 effectively neutralized the inhibitory activity of CD4⁺ cells from PPs (Fig. 5). We cannot exclude the possibility that other cell populations, such as macrophages, also secrete IL-10. However, because mice were primed orally with BLG and cytokine levels in supernatants were analysed after in vitro re-stimulation with BLG, antigen-specific events may well play a role.

An inverse correlation observed in the production of IL-10 and IFN-γ suggests that an immune deviation, or a switching of T-cell populations from Th1-like subsets to Th2-like subsets takes place among antigen-specific T-cell populations (Fig. 4). The data show that switching of the population for the induction of IL-10-secreting regulatory cells is tightly regulated by the feeding dose. Interestingly, the production in IFN-γ recovered with higher feeding doses of BLG (Fig. 4). We have also observed that T-cell clones derived from PPs of mice given high doses of BLG produce a large amount of IFN-γ (N. M. Tsuji et al., in preparation).

In studies of dose dependence, it is of interest to determine whether additional antigens produce the same (or opposite) result. The fact that OVA and BLG gave similar response profiles suggests that regulatory cells are induced in a similar range of antigen-feeding doses (Fig. 6). The effective range of the antigen feeding dose we measured is consistent with that observed in a previous study on TCR (specific to OVA) transgenic mice, for detection of IL-10 secretion by splenocytes after five feedings of OVA.⁷ Moreover, switching the cell population from Th1-like subsets to Th2-like subsets at 1–10 µm of antigen concentration, followed by a switching back to Th1-like cell populations at higher antigen concentrations was also observed in an in vitro study on TCR (specific to OVA) transgenic mice (T. Yoshida et al., personal communication). In that study, T cells from naive donors were stimulated in the presence of different doses of OVA and antigen-presenting cells from PPs of naive wild-type mice. It is interesting to see if the in vitro antigen concentration reflects the feeding dose window described in the present study. With respect to T cells that become IL-10-secreting cells after multiple low-dose antigen feedings, we assume that the observed changes reflect an important and perhaps transitional stage of T cells toward inhibitory subsets and/or anergic cells. Thus, we are vigorously investigating possible T cell-specific and/or antigen-specific-cell-specific co-stimulatory factors that are necessary for distinct T-cell differentiation.

Also worthy of consideration is the fact that PPs are constantly exposed to various types of foreign antigens at different doses. It is probable that IL-10-secreting/antigen-specific regulatory cells in PPs are consistently induced in response to various antigens, making possible an immunosuppressive environment with abundant IL-10. As often discussed, the cytokine environment is one of the most influential factors that determines the differentiation of naive T cells to effector Th1/Th2 subsets.¹⁵ Here, we showed that immunological priming via mucosal surfaces in vivo results in a prominent Th2-like response following multiple low-doses of BLG (5 × 1 mg), while lesser or greater doses were associated with a Th1-like cytokine response. We note that in mice fed > 1 mg BLG, BLG-specific cells did not show the proliferative response in LNs after systemic immunization (so-called anergy at high-dose oral tolerance; Fig. 1). Immunization with soluble protein antigens is apt to induce a Th2 response at low doses and a Th1 response at high doses.¹⁵ However, oral administration of BLG elicited a unique response in PP cells that was distinct from that seen in other studies. This result may be a consequence of the different cytokine composition in the microenvironment.

It was recently proposed that certain types of immunoregulatory cells (T-regulatory cells 1, or Tr1) represent distinct cell populations that are induced by chronic exposure to IL-10 in vitro, and secrete large amounts of IL-10.^10,11 Such cells show autocrine characteristics in that IL-10 is required as a growth factor. In addition to having direct effects on T cells, IL-10 may also affect APCs by prompting them to induce inhibitory T cells.¹⁶ Importantly, IL-10-secreting regulatory cells in the present study were primary-culture cells that were physiologically induced in PPs. Because PPs have the potential to produce high levels of IL-10, as discussed above, it is possible that the PP is an organ involved in the effective generation of Tr1-like cells.

Moreover, IL-10 down-regulates a broad spectrum of pro-inflammatory and inflammatory cytokines in macrophages and inflammatory T cells.^17,18 Recently, direct suppressive effects were shown even on eosinophils and mast cells.¹⁹ BLG is a potent milk allergen that can cause both type I and type IV allergic inflammation. The immunoregulatory function of IL-10 and IL-10-producing cells suggests a plausible means by which an effective therapy for allergic inflammation could be developed.

Th3 is another type of CD4⁺ regulatory T cell previously described, whose characteristics suggest their role in oral tolerance.⁹ Cloned Th3 T cells produce high levels of TGF-β, which mediates their inhibitory activities.^9,20 In our studies with BLG, a slight increase in TGF-β was evident under conditions in which active suppression was observed, although IL-10 seemed to be a more potent mediator of T-cell suppression in the neutralization assay (Fig. 5). According to Groux et al.,¹⁰ Tr1 cells consistently produce low levels of active TGF-β and high levels of IL-10, and both cytokines contribute to the suppressive activity of these cells.¹⁰ Again, it is likely that Tr1-like cells play a dominant role in the active suppression of low-dose oral tolerance.

A constant level of TGF-β was produced in PPs even without BLG stimulation. TGF-β may be responsible for the moderate spontaneous suppressive effect of PP cells observed with control naive mice. Moreover, antibodies to TGF-β neutralized this effect (data not shown). Histochemical analysis of PPs in OVA-TCR transgenic mice shows that IL-4-, IL-10- and TGF-β-secreting cells are present over a period of hours after feeding OVA.²¹ In the present study, we present evidence for the existence of abundant antigen-specific IL-10-secreting cells in PPs from non-transgenic animals after BLG feeding and re-stimulation in vitro. We assume that, even after 1 week from the last feeding, IL-4-, IL-10- and TGF-β-secreting cells that Gonnella et al.²¹ observed in PPs retain the ability to respond to antigens and/or to secrete regulatory cytokines. Although little is known about the function of these cytokines in regulatory cell induction, the combination of IL-4, IL-10 and TGF-β seems to affect the microenvironment of PPs, thus making this organ unique with respect to generating immunoregulatory cells.

Finally, we demonstrated that PP regulatory cells are capable of reducing inflammation in mouse footpads induced by a BLG-specific T-cell clone (2.11G). We assume that circulating PP cells, in addition to their ability to home into PPs, may freely migrate into these immune sites in order to regulate antigen-specific inflammation. Our results suggest the possibility that inducing regulatory cells via PPs can be an effective strategy for regulating systemic inflammation.

In summary, in normal mice, PPs seem to be inductive sites for IL-10-secreting regulatory cells during low-dose oral tolerance induced by BLG. It may be relevant that gnotobiotic mice, which have only small PPs, are difficult to induce oral tolerance in.²² We also found that the range of the effective antigen feeding dose was quite narrow, and therefore critical for inducing efficient active suppressor in T-cell-related immune-responsive cells, in accordance with previous studies.^5,6 This fact underlies the importance of defining the optimal dose when applying oral tolerance methods to disease. Clearly, there is a need to further extend our knowledge of how and why regulatory cells in PPs are induced over a specific range of antigen doses. When we consider the increasing prevalence of allergy and immunological disorders throughout the world, the quest for an effective strategy for inducing regulatory cells via diet would certainly be of great benefit.

Abbreviations

BLG

β-lactoglobulin

DTH

delayed-type hypersensitivity

FCS

fetal calf serum

LN

lymph node

OVA

ovalbumin

PP

Peyer's patch

SP

spleen

Th

T-helper

Tr

T-regulatory