Peripheral immune tolerance following i.v. administration of Ag has been shown to occur in the absence of B cells. Because different mechanisms have been identified for i.v. vs low dose oral tolerance and B cells are a predominant component of the gut-associated lymphoid tissue (GALT) they may play a role in tolerance induction following oral Ag. To examine the role of B cells in oral tolerance we fed low doses of OVA or myelin oligodendrocyte glycoprotein to B cell-deficient (μMT) and wild-type C57BL/6 mice. Results showed that the GALT of naive wild-type and μMT mice was characterized by major differences in the cytokine microenvironment. Feeding low doses of 0.5 mg OVA or 250 μg myelin oligodendrocyte glycoprotein resulted in up-regulation of IL-4, IL-10, and TGF-β in the GALT of wild-type but not μMT mice. Upon stimulation of popliteal node cells, in vitro induction of regulatory cytokines TGF-β and IL-10 was observed in wild-type but not μMT mice. Greater protection against experimental autoimmune encephalomyelitis was found in wild-type mice. Oral tolerance in μMT and wild-type mice was found to proceed by different mechanisms. Anergy was observed from 0.5 mg to 250 ng in μMT mice but not in wild-type mice. Increased Ag was detected in the lymph of μMT mice. No cytokine-mediated suppression was found following lower doses from 100 ng to 500 pg in either group. These results demonstrate the importance of the B cell for the induction of cytokine-mediated suppression associated with low doses of Ag.
Several studies have demonstrated that B cells can serve as APCs in the induction of Ag-specific tolerance in naive CD4+ and CD8+ T cells (1, 2) and in Ag-specific T cell clones (3). Thus, one of the primary functions of a B cell has been hypothesized to be the induction of tolerance in naive T cells. It has been demonstrated that B cells are not necessary as APCs for induction of peripheral T cell tolerance when Ag was administered by i.v. injection in B cell-deficient (μMT)3 mice (4, 5, 6). T cell tolerance as measured by diminished T cell proliferation and IFN-γ production was observed after both a low (5) and high (4) dose of Ag. These results indicated that B cells were not the only tolerogenic APC in vivo. Indeed, Ag targeted to dendritic cells has been shown to induce a tolerogenic T cell response (7). Also, peripheral T cells have been shown to be tolerized by Ag presented by nontraditional APCs in the absence of appropriate costimulation (8, 9, 10).
Several factors have been shown to affect the generation of tolerance. Differences in the route of Ag administration can determine the nature of the immune response (11). Intravenous administration induced tolerance in an Ag-specific manner as evidenced by suppression of disease, and no increase in Th2 cells was found (4, 5, 6, 11). Low doses of orally administered Ag resulted in induction of tolerance characterized by an up-regulation in Th2 cells and TGF-β (12). These results indicated that tolerance by i.v. administration of Ag and low dose oral tolerance may proceed by different mechanisms (12, 13).
Dose of Ag can determine the nature of the immune response. Intravenous tolerance and high dose oral tolerance were found to proceed by the mechanism of anergy, which was characterized by a suppression of proliferation that was reversible following incubation of cells in IL-2 (14, 15). In contrast, low dose oral tolerance was found to proceed by the mechanism of active suppression and was characterized by the production of anti-inflammatory cytokines and suppression of proliferation, which was not reversible following incubation in IL-2.
Another factor that has been shown to determine the nature of the immune response is the cytokine microenvironment. The critical role of cytokines in the induction of tolerance was evidenced by abrogation of oral tolerance following injection of IFN-γ (16).
Peyer’s patches are a major source of IgA producing B cells and have also been shown to be a site where regulatory cells that mediate the active suppression component of oral tolerance are generated (13, 17, 18). Thus it is possible that B cells may play an important role in the generation of oral tolerance. To evaluate the role of B cells in the induction of oral tolerization, we studied μMT mice, which were generated by the introduction of a nonsense mutation into the transmembrane exon of the IgM heavy chain resulting in a total deletion of B cells (19). T cell tolerance was measured by T cell proliferation and cytokine responses. Anergy was measured by reversal of suppression of proliferation by IL-2, and the cytokine microenvironment of the gut-associated lymphoid tissue (GALT) was determined by immunohistochemical analysis. Oral tolerance in μMT mice was found to proceed by the mechanism of anergy. μMT mice had profound differences in the GALT cytokine microenvironment and a defect in the induction of cytokine-mediated suppression with low doses of Ag.
Materials and Methods
μMT and wild-type C57BL/6 mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All animals were housed at the Harvard Institutes of Medicine under virus-free conditions.
Abs, Ags, and recombinant cytokines
Abs used for immunohistochemical analysis were: rat anti-mouse IL-2 clone S4B6 and rat anti-mouse IL-10 clone JESJ-2A5 (PharMingen, San Diego, CA); rabbit anti-porcine TGF-β (R&D Systems, Minneapolis, MN); rat anti-mouse IL-4, BVD6.24G (provided by A. Lichtman, Harvard Medical School, Boston, MA); and hamster anti-mouse IFN-γ clone H22.1 and rat anti-mouse IL-6 clone M-10 (Genzyme, Cambridge, MA). Abs were used at the following concentrations: IL-10, IL-6, IL-2, and IFN-γ (5 μg/ml), IL-4 (hybridoma supernatant), and TGF-β (2 μg/ml).
Abs used for ELISA were purified rat anti-mouse IL-2 (clone JES6-5H4), IL-4 (clone BVD4-1D11), IL-10 (clone JES5-2A5), and IFN-γ (clone R4-6A2) mAb; biotinylated rat anti-mouse IL-2 (clone JES6-5H4), IL-4 (clone BVD4-24G2), IL-10 (clone SXC-1), and IFN-γ (clone XMG1.2) mAb (PharMingen); and polyclonal chicken anti-TGF-β1 (R&D Systems) and monoclonal mouse anti-TGF-β (clone 1D11.16; Celtrix Pharmaceuticals, Santa Clara, CA). Recombinant cytokines were mouse IL-2, IL-4, IL-10, IFN-γ (PharMingen), and purified bovine TGF-β1 (Celtrix Pharmaceuticals).
Oral administration of Ag
Mice were fed various doses (25 mg, 0.5 mg, 250 μg, 250 ng, 100 ng, 10 ng, 1 ng, and 500 pg) of OVA dissolved in PBS by gastric intubation with an 18-gauge stainless steel feeding needle (Thomas Scientific, Swedesboro, NJ). In some experiments mice were fed 250 μg myelin oligodendrocyte glycoprotein (MOG35–55). Ag was administered every other day for a total of five times. All fed mice were immunized and compared with unfed immunized mice.
Immunization of OVA-fed mice
Mice fed OVA were immunized in the foot pad with 200 μg OVA in 0.1 ml PBS emulsified in an equal volume of CFA containing 2 mg/ml of mycobacterium tuberculosis H37 RA (Difco, Detroit, MI).
Induction and clinical evaluation of experimental autoimmune encephalomyelitis (EAE)
Mice were immunized with a s.c. injection in the flank of 50 μg MOG in 0.1 ml of dH2O:PBS (1:9) emulsified in an equal volume of CFA containing 4 mg/ml of mycobacterium tuberculosis H37 RA (Difco) and an i.v. injection of 150 ng pertussis toxin (List Biological Laboratories, Campbell, CA) in 0.1 ml of PBS. Then, animals received a second injection of pertussis toxin 48 h later. Animals were scored for EAE as follows: 0, no disease; 1, limp tail; 2, hindlimb weakness; 3, hindlimb paralysis; 4, hindlimb plus forelimb paralysis; 5 moribund.
Cell culture and proliferation assay
Cells were cultured at 0.5 × 106 or 1 × 106 per well in 0.2 ml of serum-free medium X-Vivo 20 (BioWhittaker, Walkersville, MD) supplemented with 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 10 mM HEPES, 2 mM l-glutamine, 100 U/ml penicillin, and 100 U/ml of streptomycin and containing various concentrations of Ag. For cytokine assays, culture supernatants were collected at 48 h for IL-2, IL-4, IL-10, and IFN-γ and at 72 h for TGF-β. For proliferation assays, 1 μCi of [3H]thymidine (1 Ci = 376β) was added to each culture at 72 h. Cells were harvested and radioactivity incorporated was counted 16 h later using a flatbed beta counter (Wallac, Gaithersburg, MD).
Detection of anergy
Anergy was demonstrated according to the method of Gaur et al. (19) as follows: cells were cultured in DMEM supplemented with10% FCS in a 96-well plate at 0.5 × 106 cells/well. Pooled cells from each group of mice were cultured in the absence of Ag (medium alone) and in the presence of varying doses of Ag. In addition, cells from fed immunized mice were cultured in the presence or absence of recombinant IL-2 (rIL-2, 5 U/ml) and compared with cells from unfed immunized mice that were cultured in the absence of rIL-2. IL-2 was added at the time lymph node or spleen cells were placed in culture. Cells were cultured for 4 days, and [3H]thymidine (1 μCi) was added for the last 18 h. Results are presented as the mean ± SD of triplicate cultures after subtraction of background radioactivity. Background counts were obtained by culturing cells in the absence of Ag. Experiments were repeated three times. Results of individual experiments are shown.
Collection of lymph
Lymph was collected according to the method of Korngold and Bennink (20) with the following modifications: mice were fed 0.5 mg OVA and 30 min later fed 0.3 cc corn oil for visualization of the thoracic duct. Because it was not necessary to collect lymph over a prolonged period of time, the animal was not suspended over a wheel. Lymph was collected by cannulation from 1 to 2 and from 2 to 3 h after feeding.
ELISA for cytokines
Quantitative ELISAs for IL-2, IL-4, IL-10, and IFN-γ were performed using paired mAbs specific for corresponding cytokines per manufacturer’s recommendations (PharMingen). TGF-β ELISA was performed as previously described (21). Results are expressed as mean values from triplicate cultures. Data presented are representative of two or three experiments.
ELISA for OVA
Quantitative ELISA for OVA was performed as follows: 96-well microtiter plates (Dynatech Laboratories, Chantilly, VA) were coated overnight at 4°C with 1 μg/ml sheep anti-OVA polyclonal Abs (Cortex, San Leandro, CA) in 100 μl carbonate buffer, ph 8.2. The plates were then washed three times with PBS containing 0.5% Tween 20, blocked with 1% BSA in PBS, washed, and incubated with lymph or OVA standard overnight at 4°C. The plates were washed again and incubated with 1 μg/ml biotinylated rabbit anti-OVA (Rockland, Gilbertsville, PA) for 1 h at room temperature. Avidin-peroxidase was added, and color was developed with a one-component tetramethylbenzidine reagent (Kirkegaard & Perry Laboratories, Gaithersburg, MD).
Immunostaining of light microscopic sections
Tissues were excised from anesthetized animals embedded in OCT Tissue-TeK (Miles, Elkhart, IL) snap frozen in liquid nitrogen-cooled isopentane. Cryosections were fixed in acetone and stored at −20°C. Frozen sections were thawed and fixed in acetone for 2 min. Endogenous peroxidase activity was quenched by incubation in periodic acid (0.005 M in water) for 10 min followed by washing in PBS and immersion in 0.003 M sodium borohydride for 30 min. After washing, tissues were incubated for 20 min in diluted normal blocking serum that was prepared from the species from which the secondary Ab was made. Endogenous biotin was blocked by incubation in avidin D for 1 h followed by incubation in biotin for 1 h. After washing in PBS, sections were incubated overnight in primary Ab diluted in blocking serum. Sections were rinsed in PBS and incubated in diluted biotinylated secondary Ab for 1 h. After washing, sections were incubated in Vectastain Elite ABC reagent, washed again, and incubated in peroxidase substrate solutions containing 3,3′-diaminobenzidine (DAB Substrate Kit; Vector Laboratories) for 5–7 min. Sections were washed for 5 min in tap water, counterstained in hematoxylin, cleared, and mounted.
Up-regulation of cytokine staining was determined by comparing fed immunized with unfed immunized mice. Quantification of data was performed by computerized image analysis (IP Labs Systems; Scanalytics, Fairfax, VA). Three mice per group were studied, and experiments were repeated three times. Nine sections (three per animal) were randomly selected and quantitated for each group. The region of interest (ROI) represents a defined area (116 μm × 44 μm) at ×100 magnification. The percent staining within the ROI was measured in 20 randomly selected regions. This analysis was conducted for the dome, corona, interfollicular, and germinal center of the Peyer’s patch and for the lamina propria. Data are expressed as the mean percent staining per ROI ± SD and were analyzed by Student’s t test.
Differences in oral tolerance in wild-type and μMT mice
We examined wild-type and μMT C57BL/6 mice for induction of oral tolerance following feeding of 0.5 mg OVA five times followed by immunization in the foot pad. Popliteal node cells were harvested 7 days after immunization. Upon stimulation with Ag in vitro both wild-type and μMT mice showed induction of oral tolerance evidenced by suppression of proliferation and IFN-γ production (Fig. 1). The increased level of proliferation seen in μMT mice compared with littermate controls may be due to a larger number of T cells in μMT cultures due to the absence of B cells. Although both wild-type and μMT mice showed induction of oral tolerance, wild-type mice showed up-regulation of IL-10 and TGF-β that was not seen in μMT mice. Small amounts of IL-2 (50 pg) were detected in wild-type mice that were suppressed with feeding to 10 pg. No IL-2 was detected in μMT, and no IL-4 was detected in either group. Similar results were found in spleen cells at 30 days after immunization (data not shown).
Immunohistochemistry of the GALT of wild-type and μMT mice
As shown in Figs. 2 and 3, there were dramatic differences in cytokine expression both in the Peyer’s patch and lamina propria of fed immunized wild-type and μMT mice fed 0.5 mg OVA. In Fig. 2, IL-4 was present throughout the GALT of wild-type mice (Fig. 2,A) but was absent in μMT animals (Fig. 2,B). Although IL-10 was present in the epithelium of both groups it was only up-regulated in the dome (Fig. 2, C and D) and lamina propria of wild-type mice. As shown in Fig. 3, there was up-regulation of TGF-β in wild-type mice (Fig. 3,a), whereas TGF-β was not detected in μMT mice (Fig. 3,b). IFN-γ was expressed in the lamina propria of wild-type (Fig. 3,c) mice but not in μMT mice (Fig. 3,d). IL-2 was expressed in wild-type mice (Fig. 3,e) and only sparsely distributed in the lamina propria of μMT mice (Fig. 3,f). IL-6 was expressed in both wild-type and μMT mice (Fig. 3, g and h). In animals fed 25 mg five times, up-regulation of TGF-β and an increasing trend for IL-10 were observed in the GALT of wild-type but not in μMT mice (data not shown). No detectable changes were found with feeding alone.
Having observed these differences in the GALT of wild-type verses μMT fed immunized mice, we then examined the GALT of naive (unfed, nonimmunized) wild-type and μMT mice and also found a marked difference in the cytokine microenvironment. IL-4, TGF-β, and IFN-γ were present in the GALT of wild-type mice but not in μMT mice. Small amounts of IL-10 were seen in the lamina propria of wild-type mice (0.28% ± 0.27) with a decreasing trend (0.02% ± 0.04) seen in μMT mice. In addition, similar amounts of IL-2 were detected in the corona of the Peyer’s patch of wild-type and μMT mice (0.28 ± 0.44 and 0.29 ± 0.3%, respectively). Thus, the cells of the GALT of μMT mice lack the Th2 cytokines and TGF-β found in wild-type mice indicating a different microenvironment for the presentation of Ag.
When MOG was fed as Ag, similar results to those seen with OVA were found in the GALT of wild-type and μMT mice. Mice were examined at 30 days after immunization. In wild-type mice IL-4 was increased in the dome, corona, and interfollicular region of the Peyer’s patch (Fig. 4,A). IL-10 was increased in the dome of the patch and in the lamina propria of the villi (Fig. 4 B). TGF-β was increased in the lamina propria (3.5 ± 2.7 fed vs 0.0 ± 0.0% unfed). In contrast, the GALT of both fed and unfed μMT mice was characterized by the absence of IL-4, TGF-β, and IFN-γ. Although IL-10 was present in the epithelium, only small amounts were found in the lamina propria, and no up-regulation with feeding was seen (data not shown).
Immunohistochemistry of the popliteal node of wild-type and μMT mice
Given the results in the GALT we examined in situ cytokine expression in the popliteal node. The cortex of the popliteal node of fed immunized wild-type mice fed 0.5 mg OVA showed decreased staining for IFN-γ and increased TGF-β with no significant increase in IL-4 or IL-10 when compared with unfed immunized mice (Fig. 5). In contrast, fed immunized μMT mice showed no detectable IFN-γ and no increase in TGF-β. Unfed immunized μMT mice expressed much more IL-10 than fed immunized mice with a large SD observed. μMT nodes were also characterized by the absence of IL-4. Thus, up-regulation of TGF-β, IL-4, or IL-10 was not seen in the popliteal node of fed immunized μMT mice. After feeding a high dose of 25 mg five times no up-regulation of TGF-β, IL-4, and IL-10 was seen in either group (data not shown).
Induction of anergy
Because up-regulation of IL-4, IL-10, and TGF-β associated with cytokine-mediated active suppression in wild-type mice was not seen in μMT mice, we examined these mice for the induction of anergy. Popliteal node cells from OVA-fed mice (0.5 mg) were incubated in 5 U recombinant IL-2 (rIL-2) and compared with cells from fed mice not incubated in rIL-2 and cells from unfed immunized animals. Results shown in Fig. 6 indicate that cell proliferation remained suppressed in fed wild-type mice after incubation in rIL-2. In contrast, μMT mice showed reversal of suppression following rIL-2 treatment indicating anergy. Next we tested for anergy in spleen cells from μMT and wild-type mice at 30 days after immunization. Similar to the popliteal node, reversal of suppression after incubation in rIL-2 was observed in μMT but not in wild-type animals (Fig. 7 A).
We then asked whether a similar effect was observed with a self Ag, MOG given to suppress EAE. Animals were fed 250 μg MOG five times and tested 30 days after immunization with CFA and pertussis. Tolerance was characterized by suppression of both cellular proliferation and production of IL-2 and IFN-γ. Analogous to what was seen with OVA, incubation in rIL-2 reversed the suppression in μMT mice but not in wild-type animals (Fig. 7,B). Furthermore, there was less suppression of EAE in μMT vs wild type (Fig. 8).
Analysis of thoracic duct lymphatic fluid from wild-type and μMT mice
Because we observed anergy in the popliteal node and spleen of μMT mice, we examined thoracic duct lymphatic fluid to determine whether there were differences in absorption of Ag in the two groups of mice. Lymphatic fluid from wild-type and μMT mice fed 0.5 mg OVA was collected by cannulation of the thoracic duct according to the method of Korngold and Bennink and assayed for detection of OVA by ELISA. Lymph was collected from 1 to 2 and from 2 to 3 h after feeding. As shown in Fig. 9 increased concentration of OVA was found in lymph of μMT mice compared with wild-type mice.
Cytokine-mediated suppression does not occur in μMT mice at doses that result in loss of anergy
We then asked whether reversal of suppression with IL-2 was also present at a lower dose of Ag administration or whether we could identify a low dose in μMT mice that induced cytokine-mediated suppression. Popliteal node cells from mice fed a 250 ng dose of OVA five times showed suppression of proliferation and IFN-γ production in both μMT and wild-type mice; however, suppression of proliferation was reversed in μMT mice but not in wild-type mice following incubation in rIL-2 (Fig. 10).
We then further titrated the dose of Ag fed from 100 ng to 500 pg but no suppression of proliferation (Fig. 11) or up-regulation of IL-4, IL-10, or TGF-β was observed in either wild-type or μMT mice.
In this study, we have found that the B cell is critical for maintaining the normal cytokine microenvironment of the gut that consists of IL-4, IL-10, and TGF-β. This was not only seen in animals orally administered Ag, but similar patterns were observed in the GALT of naive μMT mice. The different cytokine microenvironment appears to have the following effect. First, we did not observe the up-regulation of IL-4, IL-10, and TGF-β following orally administered Ags in μMT mice. These cytokines are associated with the induction of cytokine-mediated active suppression when low doses of Ag are administered orally. However, we found that the oral tolerance induced in μMT mice related to anergy as measured by reversal of suppression by rIL-2 at all doses that suppression was observed. Consistent with this we found increased amount of Ag entered the lymph in μMT animals than wild type, and we observed no induction of cytokines even at extremely low doses.
The absence of B cells reduces the Th2/Th3 microenvironment and enhances Ag absorption. The actual mechanisms underlying these observations are not known and could be mediated through other local cell populations. The reason that more Ag appears in the lymph in μMT mice could relate to either an altered structure of the GALT allowing greater epithelial permeability (absorptive or paracellular) and removal of Ag into the lymph, greater availability of Ag due to absence of luminal IgA or the binding of Ag by low-affinity Ab that is normally picked up by B cells but does not occur in μMT mice. This would alter the amount and form of Ag presented to macrophages and dendritic cells both in the GALT, and the draining lymphatic structures. Furthermore, if there is paracellular entry of Ag, our findings may represent access of Ag to conventional APCs in the lamina propria. The nature of the OVA in the thoracic duct was not examined. Previous studies of macromolecular uptake in the rat have found that some OVA absorbed from the gastrointestinal tract can be detected intact in the plasma and lymph (22). Both intact or processed OVA that retains antigenicity would be detected by our assay.
The mechanism of anergy is normally associated with high doses of oral Ag in wild-type mice or with Ag administered by the i.v. route in wild-type or μMT mice. The ability of μMT mice to tolerize following i.v. Ag is well established (4, 5, 6), and high dose (25 mg) oral tolerance has been reported (23). This study demonstrates that oral tolerance following low doses of Ag proceeds by the mechanism of anergy in μMT mice, whereas a nonanergic, cytokine-mediated mechanism was found in wild-type mice.
In this study we performed a titration of doses to determine whether a low dose could be identified in which oral Ag induced tolerance due to a cytokine-mediated suppression mechanism in the μMT mice. However, doses as low as 250 ng resulted in suppression of cellular proliferation in μMT, which was reversible following incubation in rIL-2 indicating anergy with no IL-4, IL-10, and TGF-β. In contrast, suppression in wild-type mice was not reversed by IL-2. Oral doses of 100 ng to 500 pg resulted in no suppression of cellular proliferation in either wild-type or μMT mice. The reason that titration to a low dose that induces active suppression in μMT mice is not achievable is not known but may be due to vast differences in the cytokine microenvironment in the GALT of wild-type and μMT mice. Our immunohistochemical analysis shows that unfed wild-type mice have IL-4, IL10, TGF-β, and IFN-γ that are reduced or absent in the GALT of μMT mice. Both wild-type and μMT mice have similar amounts of IL-2 in the corona of the Peyer’s patch, and IL-6 was found in the Peyer’s patch and lamina propria of both groups. We found small amounts of IL-12 in wild-type animals but none in μMT (data not shown). Thus the data suggest that Ag presentation takes place in a different environment in μMT mice. Previous studies have shown that injection of Th1 cytokines can circumvent the induction of oral tolerance (16). Taken together the data indicate that the B cell is necessary for the generation of the Th2 environment in the GALT and is essential for the induction of a low dose active suppression mechanism of tolerance.
In μMT mice Peyer’s patches are not visible to the eye as pronounced elevations along the intestinal wall as seen in wild-type mice, rather they appear as small opaque areas. Also the patches in μMT mice are reduced in size due to absence of B cells and are fewer in number than in wild-type mice. Thus some investigators may not have detected patches, whereas other investigators have called them lymphoid aggregates due to their reduced size. Peyer’s patches are characterized by the presence of follicle-associated epithelium (FAE), which contain M cells scattered among the columnar epithelial cells. The FAE covers the dome region of the patch, which is free of intestinal villi. Golovkina et al. (24) have published pictures of M cells in the FAE of Peyer’s patches of μMT mice and have reported that 1.2–6.6 M cells are present in μMT mice per every 1000 M cells in normal intestine.
Our immunohistochemical analysis of the GALT shows that in response to feeding low doses of OVA or MOG the up-regulation of IL-4, IL-10, and TGF-β occurs in wild-type but not μMT mice. In the popliteal node in situ we found that low dose feeding resulted in the up-regulation of TGF-β in wild-type but not μMT mice. Thus similar to results in vitro we observed cytokine up-regulation only in fed immunized wild-type mice. In addition, we observed that in the popliteal node of μMT mice much more IL-10 was present in unfed immunized than fed immunized mice. The reason for this is not known but may be due to apoptosis of Th2 cells following oral administration of an anergy-inducing dose of Ag. Previously we have shown that oral Ag can delete Ag-reactive T cells by apoptosis. Deletion was dependent on dosage and frequency of feeding (25). At lower doses apoptosis was not observed; however, at higher doses both Th1 and Th2 cells were deleted. In the case of μMT mice we have shown increased Ag in the lymph and the demonstration of anergy in the popliteal node.
Our titration experiments showed that in addition to the absence of tolerance, the feeding of a 500-pg dose resulted in a priming response in both wild-type and μMT mice. Previous studies have reported the phenomenon of priming in response to varying low doses of Ag (26, 27). Lamont et al. found the priming of systemic and local delayed-type hypersensitivity responses by feeding low doses of 5 μg of OVA to BALB/c mice (26). Furthermore, Stokes et al. reported that sensitization of systemic DTH following oral Ag occurred only in one of two mouse strains examined (27). Taken together, the data suggest that under certain circumstances in specific mouse strains priming can occur in response to very low doses of Ag. The reason for this is unknown, but we hypothesize that this may relate to differential Ag presentation and strength of the signal to the T cell that occurs at very low doses; we are presently investigating this. Furthermore, Lamont et al. have also speculated that it reflects different effects of fed Ag on distinct subsets of Th1 and Th2 cells (26). Our studies suggest that the cytokine microenvironment also may play a role because we observed evidence of priming in μMT mice fed nanogram doses of OVA after cells were challenged with 1000 μg in vitro (Fig. 11). Thus the priming at very low doses appears to be a real biological event whose mechanism remains unexplained at this time.
Evidence from several studies suggests that B cells acting as APC skew the differentiation of helper T cells toward a Th2 response (28, 29, 30, 31). The generation of Th2- and TGF-β-producing cells characterizes the active suppression mechanism of oral tolerance (12). Because T cells producing IL-4 and IL-10, and TGF-β have been shown to regulate Th1 cells in EAE (32), the absence of active suppression in μMT mice may explain the reduced protection against MOG-induced EAE in our μMT mice compared with wild type. The importance of the B cell and its possible role in the generation of a Th2 response was also suggested by studies of EAE that have shown that in the absence of the B cell there is a failure of spontaneous recovery in B10.PL μMT mice compared with B10.PL mice (33).
In conclusion, the data demonstrate that the B cell is essential for maintenance of a predominately Th2 cytokine environment in the GALT and the induction of the low dose cytokine-mediated suppression mechanism of oral tolerance in C57BL/6 mice. Furthermore, the data suggest that the cytokine microenvironment in the GALT and lymph nodes is important for the generation of regulatory cells indicative of this mechanism.
We thank Dr. Robert Korngold, Dr. Robert Sakstein, and Geoffrey Grove for helpful discussions.
This work was supported by a grant from the National Multiple Sclerosis Society (to P.A.G.) and National Institutes of Health Grant AI43458 (to H.L.W.).
Abbreviations used in this paper: μMT, B cell-deficient; GALT, gut-associated lymphoid tissue; MOG, myelin oligodendrocyte glycoprotein; EAE, experimental autoimmune encephalomyelitis; ROI, region of interest.