Abstract
A fascinating feature of the intestinal mucosal immune system is its ability to guard against invasion by pathogens while avoiding a response to the many potential Ags present in food. The phenomenon of systemic tolerance after oral administration of protein Ags is well documented, but the cellular and molecular basis for the observed nonresponsiveness is not fully understood. To gain insight into the role of the mucosal microenvironment in the induction of orally induced nonresponsiveness, we attempted to induce tolerance to OVA in mice primed for a Th2-biased mucosal immune response by infection with the nematode parasite Heligmosomoides polygyrus. We found that oral tolerance for Th1-type responses to OVA is maintained when OVA is fed during the peak of the mucosal immune response to H. polygyrus. Tolerance for Th2 cytokine responses or a Th2-dependent isotype of IgG is not induced in this Th2-biased mucosal environment. Treatment of infected mice with rIL-12 to reverse the Th2 polarity of the parasite-specific immune response restores tolerance of both Th1 and Th2 responses to OVA. We conclude that the polarized Th2 response induced by this enteric infection plays a central role in determining whether or not systemic tolerance is induced. Our results imply that attempts to use oral administration of Ag to suppress systemic immune responses will be influenced strongly by the presence of mucosal infection.
The term oral tolerance refers to the induction of peripheral nonresponsiveness as a result of the oral administration of soluble protein Ags. The induction and maintenance of oral tolerance to food and other innocuous luminal Ags are central to the effective function of the intestinal mucosal immune system in guarding the epithelial barrier against pathogenic invasion. The immunologic mechanisms through which the mucosal immune system achieves these opposing functions are not fully understood, but are clearly of biologic relevance, as a failure of oral tolerance to food and other luminal (e.g., bacterial) Ags has been postulated to play a role in the induction of intestinal inflammation and inflammatory bowel diseases (1).
Systemic nonresponsiveness has been observed after feeding of a wide variety of Ags, including soluble proteins, contact sensitizing agents, and heterologous RBC (reviewed in 2 . The demonstration that oral administration of type II collagen suppresses the induction of collagen-induced arthritis (3, 4) and myelin basic protein suppresses the induction of experimental autoimmune encephalitis (5), and the subsequent generalization of these findings to many other experimental models of autoimmune disease (reviewed in 6 have sparked renewed interest in understanding the mechanism(s) by which oral Ag leads to peripheral nonresponsiveness. The suppression of both humoral and cell-mediated immune responses in the peripheral lymphoid tissue is often accompanied by a local mucosal secretory IgA response. All of the known mechanisms for the induction of peripheral tolerance, including clonal deletion, clonal anergy, and regulation by cytokines (as well as TGF-β-mediated active suppression) have been postulated to play a role in oral tolerance, the relative contributions of each varying with the dose of Ag fed (6). However, whether or not the microenvironment of the gut-associated lymphoid tissue (GALT)3 plays a unique role in the induction of oral tolerance has not been clear. It has been suggested that oral tolerance to high doses of soluble Ag results when small amounts of Ag gain access to the systemic circulation (2) and are then presented by nonprofessional APCs in the absence of appropriate costimulation (6, 7), implying little role for the GALT.
CD4+ T cells can be divided into two functional subsets, Th1 and Th2, which are defined by their pattern of cytokine secretion. Differentiation of Th cells into Th1 or Th2 subsets is determined by the presence of the priming cytokines IL-12 or IL-4, respectively, during a primary antigenic stimulus. Indeed, recent evidence has suggested that the immune response to a variety of infectious agents is accompanied by preferential expansion of one Th subset with a concomitant down-regulation of the other (8). To explore the role of the GALT, we examined oral tolerance to the model Ag OVA in the context of a mucosal infection. Heligmosomoides polygyrus is a natural murine parasite with a strictly enteric life cycle that induces a vigorous, polarized Th2-type immune response characterized by peripheral blood eosinophilia and marked increases in serum IgG1 and IgE, and accompanied by a marked induction of the costimulatory molecules B7.1 and B7.2 (9). To induce tolerance to OVA, mice were fed a single high dose of OVA before immunization with OVA in CFA. Draining popliteal lymph node (PLN) cells from OVA-fed, noninfected mice were impaired in their ability to produce both Th1 (IL-2 and IFN-γ) and Th2 (IL-5 and IL-10) cytokines upon restimulation with OVA in vitro, but responded as well as PBS-fed controls to stimulation with plate-bound anti-CD3. Both Th1-dependent (IgG2a) and Th2-dependent (IgG1) OVA-specific serum Ab responses were reduced in OVA-fed mice compared with controls. Intragastric administration of OVA to mice at the peak of the mucosal immune response to H. polygyrus at 8 days postinfection (p.i.) resulted in maintenance of Th1 cytokine tolerance upon stimulation with OVA in vitro, while Th2 responses were similar to or higher than those of PBS-fed controls. In keeping with these findings, serum OVA-specific IgG2a, but not IgG1, was reduced compared with PBS-fed controls. Treatment of OVA-fed, H. polygyrus-infected mice with rIL-12 during the first 8 days p.i. restored and potentiated tolerance of serum OVA-specific Ab responses of both Th1- and Th2-dependent isotypes as well as Th1 and Th2 cytokine responses to OVA in vitro. We favor the interpretation that these results point to a unique role for the cytokine microenvironment of the GALT in the induction and maintenance of oral tolerance.
Materials and Methods
Mice and parasitic infection
Female BALB/cByJ mice (8–10 wk of age) were purchased from The Jackson Laboratory (Bar Harbor, ME) and were maintained in a specific viral pathogen-free facility. Infective, ensheathed, third-stage larvae (L3) of H. polygyrus (generously provided by Dr. Marilyn Scott, Institute of Parasitology, McGill University, Montreal, Canada) were propagated as previously described and stored at 4°C until use (10). Mice were inoculated orally with 200 L3. Adult worms in the intestinal contents were determined at sacrifice (4 wk p.i.). Parasite-induced peripheral blood eosinophilia was measured using the Unopette test (Becton Dickinson, Rutherford, NJ).
Induction of tolerance to OVA
Eight days after infection with H. polygyrus, either 0.5 ml of PBS or OVA (25 mg in PBS; Sigma Chemical Co., St. Louis, MO) was administered intragastrically using a ball-tipped feeding needle to groups of three or four mice. Control mice received PBS or OVA without concomitant H. polygyrus infection. One group of OVA-fed, H. polygyrus-infected mice was treated with murine rIL-12 (R&D Systems, Minneapolis, MN) during the first 8 days p.i. The IL-12 was diluted in PBS containing 1% BALB/c serum and was injected i.p. in 0.1 ml. For the first IL-12 treatment experiment, the mice were injected seven times with 950 ng/mouse/injection. After seven doses, the mice showed some signs of toxicity and the dose was reduced to six injections of 833 ng/mouse in subsequent experiments. Seven days after feeding, mice were immunized in the hind footpads with 100 μg of OVA in CFA. Two weeks after immunization, cells from the draining PLN were harvested for restimulation in vitro and serum was collected for determination of OVA-specific Ab responses.
In vitro restimulation
Pooled PLN cells from each group of three or four mice (2 × 106/ml) were cultured in triplicate in flat-bottom microtiter plates (Costar, Cambridge, MA) in complete DMEM (Life Technologies, Grand Island, NY; complete DMEM contains 10% FCS (HyClone Laboratories, Logan, UT), 10 mM HEPES, 2 mM l-glutamine, 100 U penicillin/ml, 100 μg streptomycin/ml, 50 μM β-mercaptoethanol, 0.1 mM nonessential amino acids, and 1 mM sodium pyruvate) in 200 μl with or without varying concentrations of OVA. Some wells were coated with purified anti-CD3 (PharMingen, San Diego, CA; 10 μg/ml) as a positive control. To measure cellular proliferation, plates were pulsed at 48, 72, and 96 h with 1 μCi/well of [3H]TdR (DuPont NEN, Boston, MA) and harvested 16 h later. [3H]Thymidine incorporation was determined by liquid scintillation counting (Beckman LS1801, Fullerton, CA).
Measurement of cytokine production
At 24 h after the initiation of the cultures, 50 μl of supernatant was harvested for assessment of IL-2 production using the indicator cell line HT-2 (American Type Culture Collection (ATCC), Rockville, MD) in the presence of neutralizing concentrations of the IL-4-specific Ab 11B11 (also obtained from ATCC). HT-2 cells (5 × 103) were incubated with culture supernatants, in triplicate, for 24 h; 1 μCi/well of [3H]TdR was added for the last 6 h of culture. Murine rIL-2 (Genzyme Corp., Boston, MA) was used as the standard.
Additional cultures in complete DMEM were set up in 24-well plates at 4 to 5 × 106 cells/ml in 1 ml for the assessment of IL-4, IFN-γ, IL-5, and IL-10 production in 48- and 72-h culture supernatants by ELISA. Cultures were stimulated with 1, 10, or 100 μg/ml OVA, 5 μg/ml Con A, or plate-bound anti-CD3 (10 μg/ml). ELISA capture (BVD4-1D11, IL-4; R4-6A2, IFN-γ, TRFK-5, IL-5, and JESS-2A5, IL-10) and biotinylated second Abs (BVD6-24G2, IL-4; XMG1.2, IFN-γ; TRFK4, IL-5, and SXC-1, IL-10) were purchased from PharMingen. Immulon 2 ELISA plates (Dynatech Labs., Chantilly, VA) were coated with capture Abs in PBS overnight at 4°C before blocking in PBS/3% FCS for 2 h at room temperature. After washing with PBS/0.05% Tween, culture supernatants or standards were incubated, in triplicate, on coated plates overnight at 4°C. The plates were then washed and incubated with the biotinylated second Abs for 1 h at 37°C. The wells were washed with PBS/Tween and incubated with peroxidase-conjugated streptavidin (0.5 μg/ml; Zymed Labs., San Francisco, CA), developed with O-phenylenediamine (OPD; Zymed Labs.), stopped with 2 N H2SO4, and read at 492 nm. Standard curves were obtained using recombinant murine IFN-γ, IL-4 (Genzyme Corp.), IL-10 (R&D Systems), and IL-5 (PharMingen), and are expressed in pg/ml ± SEM. OD values were converted to pg/ml for each cytokine by linear regression with Delta Soft II (Biometallics, Princeton, NJ). The limits of detection of the ELISA assays are 5 pg/ml for IFN-γ and IL-4, 40 pg/ml for IL-10, and 8 pg/ml for IL-5.
Ab assays
Each mouse was bled at sacrifice, and individual sera were assayed for OVA-specific IgG1 and IgG2a by ELISA on OVA-coated Immulon 2 plates. Plates were blocked with PBS/3% BSA for 1 h at RT and washed in PBS/Tween-20 before addition of diluted serum samples, in triplicate. After 2.5 h at RT, isotype-specific Ab responses were detected using horseradish peroxidase-conjugated goat anti-mouse IgG1 and IgG2a (Southern Biotechnology, Birmingham, AL). The reaction was developed with OPD and read at 492 nm. OD values were converted to μg/ml of OVA-specific IgG1 or IgG2a by comparison with a standard curve of purified OVA-specific Ig (developed with either anti-mouse IgG1 or anti-mouse IgG2a) with Delta Soft linear regression analysis, and are expressed as the mean concentration for each group of mice ± SEM. OVA-specific Ig was purified from the pooled serum of OVA-immunized mice by affinity chromatography on an OVA-Sepharose 4B column (Pharmacia, Piscataway, NJ) prepared according to the manufacturer’s instructions.
Mice were also bled weekly to monitor the course of the parasitic infection. Total IgG1 and total IgE were detected by ELISA on plates coated with goat anti-mouse Ab to IgG1 (Southern Biotechnology) or rat anti-mouse Ab to IgE (PharMingen). Blocked, washed plates were incubated with diluted serum samples in triplicate for 1.5 h at RT. After washing with PBS/Tween, IgG1 was detected with HRP-conjugated goat anti-mouse IgG1 (Southern Biotechnology); IgE was detected with biotin-conjugated rat anti-mouse IgE (PharMingen) and peroxidase-conjugated streptavidin (Zymed Labs.). Both reactions were developed with OPD and read at 492 nm. OD values were converted to μg/ml of IgG1 and IgE by comparison with standard curves of purified IgG1 (Southern Biotechnology) or IgE (PharMingen) by linear regression analysis, and are expressed as the mean concentration for each group of mice ± SEM.
Statistical differences in serum Ab levels were determined using a two-tailed Student’s t test with StatView software (Abacus Concepts, Berkeley, CA). A p value < 0.05 was considered significant.
Results
In vitro proliferative responses to OVA are tolerized in both infected and noninfected mice
Previous studies have shown that the mucosal immune response to H. polygyrus peaks at 8 days p.i. Gause and colleagues demonstrated that B220+ B cells in the mesenteric lymph node (MLN) of infected mice exhibit a maximal increase in cell size, MHC class II, and B7-2 expression at this time point (11). Moreover, reverse-transcriptase PCR analysis showed that cells in the MLN and Peyer’s patch already exhibited a polarized Th2 cytokine response by day 8 (12). We were interested in determining whether a polarized, parasite-induced, Th2 mucosal cytokine response influenced subsequent systemic responses to orally administered Ag. We therefore used the oral tolerance protocol shown in Figure 1. Eight days after H. polygyrus infection, mice were fed a single high (25 mg) dose of OVA or PBS and immunized in the footpads 1 wk later with OVA/CFA. One group of infected mice was treated with rIL-12 six times during the first 8 days p.i., a protocol previously shown to inhibit Th2 cytokine production and enhance Th1 responses to primary nematode infection (13). Two weeks after immunization, cells from the draining PLN were restimulated in vitro with varying doses of OVA or plate-bound anti-CD3. Proliferative responses were determined by [3H]thymidine incorporation at various time points, and cytokines secreted into the culture supernatant were analyzed by ELISA. Three independent experiments produced comparable results. The results shown in Figures 2 through 4 are from the same representative experiment. Figure 2 A shows that T cells from both H. polygyrus-infected (closed symbols) and noninfected (open symbols) PBS-fed mice proliferate in response to stimulation with OVA in vitro. The response increased markedly over the time period examined. As expected from previous reports (14, 15), oral administration of OVA virtually abrogated the proliferative response to OVA in vitro. The proliferation of T cells from OVA-fed, infected mice is partially reduced. Treatment of OVA-fed infected mice with rIL-12 in vivo to block the parasite-induced Th2 response restored nonresponsiveness to OVA restimulation in vitro.
T cells from these five different treatment groups exhibited interesting differences in their response to cross-linking with anti-CD3 in vitro. Figure 2,B shows that all five cell populations responded similarly to anti-CD3 after 66 h of culture, indicating that all of the cell populations were viable and contained similar numbers of potentially responsive T cells, and emphasizing the specificity of the nonresponsiveness to OVA seen in Figure 2,A. T cells from noninfected mice decline or plateau after 66 h in culture. The anti-CD3 response from OVA-fed mice declined more rapidly than that of PBS-fed mice, perhaps reflecting an enhanced susceptibility of orally tolerized lymphocytes to cell death in vitro (16). The response of T cells from both PBS- and OVA-fed infected mice to plate-bound anti-CD3 continued to climb throughout the culture period examined, presumably driven by parasite-specific priming in vivo. The response to the parasite is diminished in IL-12-treated infected mice (see Fig. 4), and T cells from these mice proliferate less vigorously to anti-CD3 stimulation in vitro, but at levels at least comparable with those seen in noninfected mice. However, whereas the response to anti-CD3 declined over time in noninfected mice, the response of T cells from OVA-fed, IL-12-treated infected mice continued to increase from 66 to 114 h in culture in a manner similar to that observed for the other groups of parasite-infected mice.
Th2 cytokine responses, and Th2-dependent Ab responses are not tolerized in mice fed OVA 8 days after infection with H. polygyrus; Th2 nonresponsiveness is restored by treatment of infected mice with rIL-12
We next asked whether oral administration of OVA in the polarized, Th2-biased mucosal environment present in H. polygyrus-infected mice influenced the cytokine profile seen after Ag restimulation in vitro. Culture supernatants were analyzed for a panel of Th1 (IFN-γ) and Th2 (IL-4, IL-5, IL-10) cytokines at both 48 and 72 h after the initiation of the culture by ELISA. IL-2 was measured in 24-h supernatants by bioassay (see Materials and Methods). The data shown in Figure 3 represent the peak of the Ag-specific cytokine response. The Ag specificity of the cytokine responses is demonstrated by their dose dependence; in vitro restimulation with increasing concentrations of OVA elicited secretion of higher levels of cytokines for each cytokine assayed. We confirmed that secretion of both Th1 and Th2 cytokines was reduced markedly by feeding of OVA to noninfected mice (15). Figure 3 shows that production of the Th1 cytokines IL-2 (Fig. 3,A) and IFN-γ (Fig. 3,B) was tolerized by OVA feeding in both infected and noninfected mice. PLN from each group responded similarly to restimulation with anti-CD3 in vitro. Markedly different results were observed when the same culture supernatants were analyzed for the Th2 cytokines IL-5 and IL-10. Th2 cytokine production was tolerized by OVA feeding in noninfected mice. However, OVA-fed infected mice produced levels of IL-10 comparable with that seen in PBS-fed infected and noninfected controls (Fig. 3,C). Tolerance to OVA was maintained in PLN cells from OVA-fed, infected mice treated with rIL-12 in vivo. OVA-fed, infected mice produced more IL-10 in response to anti-CD3 stimulation than that observed for PBS controls; PLN cells from OVA-fed noninfected mice secreted less IL-10 in response to anti-CD3. Although observed repeatedly, the mechanism by which OVA feeding suppresses the Th2 cytokine response to stimulation with anti-CD3 in noninfected mice is not clear. Enhanced responsiveness to restimulation with anti-CD3 by cells from infected mice may reflect parasite-induced activation in vivo. In the experiment shown in Figure 3,D, OVA feeding markedly enhanced IL-5 production by PLN cells from infected mice in response to OVA restimulation in vitro; in other experiments, OVA- and PBS-fed infected mice produced similar amounts of IL-5 (data not shown). IL-12 treatment of OVA-fed infected mice restored tolerance of IL-5 responses to OVA. As noted for IL-10, feeding primed for enhanced production of IL-5 in response to stimulation with anti-CD3 in infected mice, while noninfected, OVA-fed mice produced less IL-5 in response to cross-linking with anti-CD3. H. polygyrus infection is known to induce an elevation of IL-4 mRNA in the gut-associated lymphoid tissue (12), and IL-4 is clearly involved in the protective immune response to the parasite (17). Although IL-4 was readily detectable in cultures of anti-CD3-stimulated splenic or PLN lymphocytes from infected mice, we could not reproducibly detect an OVA-specific IL-4 response in the culture supernatants of PLN lymphocytes restimulated with OVA in vitro. Accordingly, IL-4 is not included in the Th2 cytokine panel shown in Figure 3. Therefore, noninfected mice exhibit tolerance for Ag-specific Th1 and Th2 responses in vitro. Mice fed Ag during a Th2-biased mucosal immune response cannot be tolerized for Th2 cytokine restimulation in vitro.
While it is well documented that cell-mediated immune responses are tolerized much more readily than humoral responses by orally administered Ag (2), it is also possible to suppress both Th1-dependent and Th2-dependent OVA-specific serum Ab responses by oral administration of OVA (15). Table I shows that both OVA-specific IgG1 and IgG2a were reduced by OVA feeding in noninfected mice. However, the Th2-dependent (IgG1) Ab response was tolerized by OVA feeding in noninfected mice, but not in mice infected with H. polygyrus. A statistically significant reduction in the OVA-specific IgG1 response was restored by treatment of infected mice with rIL-12. Thus, oral administration of OVA during a Th2-biased mucosal immune response prevents induction of tolerance for a Th2-dependent Ab response.
Treatment . | OVA-Specific IgG2a (μg/ml ± SEM) . | OVA-Specific IgG1 (μg/ml ± SEM) . |
---|---|---|
Noninfected | ||
PBS fed (n = 10) | 557 ± 124 | 1484 ± 133 |
OVA fed (n = 10) | 351 ± 54 | 680 ± 95* |
Infected | ||
PBS fed (n = 10) | 492 ± 130 | 1229 ± 108 |
OVA fed (n = 10) | 124 ± 38* | 1036 ± 105 |
OVA fed+rIL12 (n = 6) | 146 ± 35* | 726 ± 104* |
Treatment . | OVA-Specific IgG2a (μg/ml ± SEM) . | OVA-Specific IgG1 (μg/ml ± SEM) . |
---|---|---|
Noninfected | ||
PBS fed (n = 10) | 557 ± 124 | 1484 ± 133 |
OVA fed (n = 10) | 351 ± 54 | 680 ± 95* |
Infected | ||
PBS fed (n = 10) | 492 ± 130 | 1229 ± 108 |
OVA fed (n = 10) | 124 ± 38* | 1036 ± 105 |
OVA fed+rIL12 (n = 6) | 146 ± 35* | 726 ± 104* |
Results shown are the mean OVA-specific IgG2a or IgG1 Ab concentration for each group of mice (n = 6–10) ± SEM. Sera were assayed individually in triplicate, and OVA-specific Ab concentrations were determined by comparison to a standard curve of affinity-purified anti-OVA Ig. The data shown are pooled from three independent experiments, one of which is shown in Figures 2 through 42–4. *p value < 0.05 in comparison to PBS-fed group using two-tailed Student’s t test.
IL-12 treatment during the first 8 days p.i. inhibits parasite-induced Th2 responses
Infection with H. polygyrus induces a marked increase in serum IgG1 and IgE (17). To monitor the effects of OVA feeding and rIL-12 treatment on the response to the parasite, we determined eosinophil counts and serum IgE and IgG1 levels weekly during the course of the oral tolerance experiments. rIL-12 treatment has been shown to inhibit IgE and eosinophil responses during a primary infection with the intestinal nematode parasite Nippostrongylus brasiliensis (13). Figure 4,A shows that the IgE response to H. polygyrus infection is also significantly inhibited by rIL-12 treatment during the first 8 days p.i. The parasite-induced IgE response is reduced by about 50% at day 8 p.i. (p < 0.05), the time point at which OVA is fed, and remains at about 75% of PBS control levels even at 30 days p.i. The eosinophil response to H. polygyrus infection at day 8 p.i. is abrogated completely by rIL-12 treatment (Fig. 4,B) and remains below levels seen in infected, untreated mice throughout the course of the experiment. The parasite-induced IgG1 response is not affected by rIL-12 treatment during the first 8 days of infection (Fig. 4,C). This was expected, since previous work has shown that while the polyclonal IgE response to H. polygyrus is inhibited by Abs to IL-4 and eosinophilia is inhibited by Abs to IL-5, neither Ab treatment inhibited total IgG1 (18). In addition, IgE responses are generally more readily inhibited by rIL-12 treatment than IgG responses (19). Finally, we determined the impact of rIL-12 treatment and OVA feeding on parasite survival. Figure 4 D shows that the number of parasites recovered from mice fed with OVA at 29 days p.i. did not differ from that in the control, PBS-fed mice. However, the number of parasites in rIL-12-treated mice was elevated, indicating that IL-12 treatment resulted in prolonged survival. Taken together, these results suggest that IL-12 treatment of infected mice resulted in a marked diminution in the parasite-induced Th2 cytokine response.
As further confirmation of the effectiveness of IL-12 treatment in reversing the cytokine profile of H. polygyrus-infected mice, we stimulated MLN and spleen cells at 8 days p.i. from noninfected, infected, and infected plus IL-12-treated mice in vitro with Con A or anti-CD3 and examined the secretion of the Th1 cytokine IFN-γ and the Th2 cytokines IL-4, IL-5, and IL-10 into 48-h culture supernatants by ELISA. Table II shows a clear Th2 bias in the response to Con A or anti-CD3 in cells derived from the MLN of infected mice. The MLN shifts from 30% B220+ and 70% TCR-αβ+ in uninfected mice to 70% B220+ and 30% TCR-αβ+ in infected mice at 8 days p.i. (data not shown). The polarized Th2 response in the MLN is clear even without normalizing the data for this marked downward shift in the proportion of T cells in infected mice. IL-12 treatment greatly reduces, but does not completely remove, the Th2 response, as would be expected from the data on the response to the parasite in Figure 4. Th2 cytokine production is enhanced (and Th1 cytokine production reduced) in Con A- or anti-CD3-stimulated spleen cells from infected mice. The splenic T cell response is not polarized to a Th2-type response as it is in the MLN, and we did not obtain evidence for any alteration in the proportion of T and B cells in the spleens of infected mice. The increased Th2 cytokine response in infected mice was eliminated by treatment with rIL-12 in vivo. These results are consistent with previous analyses of cytokine mRNA induced in response to nematode infection. When examined in the MLN and Peyer’s patch 8 days after primary infection with another nematode parasite, N. brasiliensis, IL-12 treatment inhibited the induction of mRNA for the Th2 cytokines IL-4 and IL-5, but enhanced the production of the Th1 cytokine IFN-γ as well as the Th2 cytokine IL-10 (13). We did not see any enhancement of IL-10 responses in infected mice by examining secretion of cytokines by T cells restimulated with Con A or anti-CD3. The up-regulated production of IL-10 mRNA noted in the previous study might be derived from non-T cells.
Stimulus . | Source of Cells . | Cytokine Production (ng/ml ± SEM) . | . | . | . | |||
---|---|---|---|---|---|---|---|---|
. | . | IFN-γ . | IL-5 . | IL-10 . | IL-4 . | |||
MLN | ||||||||
Con A | Noninfected | 5.01 ± 0.09 | 0.305 ± 0.0 | 0.08 ± 0.01 | 0.42 ± 0.09 | |||
Infected | 0.94 ± 0.01 | 24.39 ± 0.27 | 27.21 ± 0.09 | 19.86 ± 0.26 | ||||
Infected+ rIL-12 | 17.33 ± 0.02 | 3.63 ± 0.07 | 5.23 ± 0.05 | 1.93 ± 0.06 | ||||
Anti-CD3 | Noninfected | 11.45 ± 0.14 | 1.06 ± 0.16 | 0.59 ± 0.10 | 0.45 ± 0.02 | |||
Infected | 16.26 ± 0.17 | 87.33 ± 0.99 | 31.74 ± 0.17 | 41.9 ± 0.28 | ||||
Infected+ rIL-12 | 18.8 ± 0.20 | 21.36 ± 0.12 | 12.78 ± 0.18 | 11.85 ± 0.32 | ||||
Spleen | ||||||||
Con A | Noninfected | 12.59 ± 0.16 | 0.71 ± 0.08 | 1.22 ± 0.03 | 0.33 ± 0.04 | |||
Infected | 6.96 ± 0.03 | 6.81 ± 0.12 | 4.39 ± 0.07 | 3.00 ± 0.06 | ||||
Infected+ rIL-12 | 13.2 ± 0.06 | 2.29 ± 0.09 | 2.04 ± 0.01 | 0.48 ± 0.05 | ||||
Anti-CD3 | Noninfected | 16.51 ± 0.08 | 1.0 ± 0.07 | 4.76 ± 0.08 | 0.58 ± 0.20 | |||
Infected | 12.51 ± 0.16 | 14.96 ± 0.18 | 6.28 ± 0.19 | 13.77 ± 0.24 | ||||
Infected+ rIL-12 | 15.26 ± 0.19 | 8.45 ± 0.15 | 2.49 ± 0.09 | 5.56 ± 0.16 |
Stimulus . | Source of Cells . | Cytokine Production (ng/ml ± SEM) . | . | . | . | |||
---|---|---|---|---|---|---|---|---|
. | . | IFN-γ . | IL-5 . | IL-10 . | IL-4 . | |||
MLN | ||||||||
Con A | Noninfected | 5.01 ± 0.09 | 0.305 ± 0.0 | 0.08 ± 0.01 | 0.42 ± 0.09 | |||
Infected | 0.94 ± 0.01 | 24.39 ± 0.27 | 27.21 ± 0.09 | 19.86 ± 0.26 | ||||
Infected+ rIL-12 | 17.33 ± 0.02 | 3.63 ± 0.07 | 5.23 ± 0.05 | 1.93 ± 0.06 | ||||
Anti-CD3 | Noninfected | 11.45 ± 0.14 | 1.06 ± 0.16 | 0.59 ± 0.10 | 0.45 ± 0.02 | |||
Infected | 16.26 ± 0.17 | 87.33 ± 0.99 | 31.74 ± 0.17 | 41.9 ± 0.28 | ||||
Infected+ rIL-12 | 18.8 ± 0.20 | 21.36 ± 0.12 | 12.78 ± 0.18 | 11.85 ± 0.32 | ||||
Spleen | ||||||||
Con A | Noninfected | 12.59 ± 0.16 | 0.71 ± 0.08 | 1.22 ± 0.03 | 0.33 ± 0.04 | |||
Infected | 6.96 ± 0.03 | 6.81 ± 0.12 | 4.39 ± 0.07 | 3.00 ± 0.06 | ||||
Infected+ rIL-12 | 13.2 ± 0.06 | 2.29 ± 0.09 | 2.04 ± 0.01 | 0.48 ± 0.05 | ||||
Anti-CD3 | Noninfected | 16.51 ± 0.08 | 1.0 ± 0.07 | 4.76 ± 0.08 | 0.58 ± 0.20 | |||
Infected | 12.51 ± 0.16 | 14.96 ± 0.18 | 6.28 ± 0.19 | 13.77 ± 0.24 | ||||
Infected+ rIL-12 | 15.26 ± 0.19 | 8.45 ± 0.15 | 2.49 ± 0.09 | 5.56 ± 0.16 |
MLN or spleen cells pooled from three mice/group were collected at 8 days p.i. and cultured at 5 × 106/ml with 5 μg/ml of Con A or on plates coated with anti-CD3 (10 μg/ml). Cytokine production in 48-h culture supernatants was measured by ELISA. The results are displayed as the mean ± SEM of triplicate wells.
Discussion
The major point to emerge from this study is that intragastric administration of a soluble protein Ag in the context of a polarized, enteric, helminthic infection has a dramatic effect on the subsequent peripheral immune response to immunization with the same Ag. Tolerance for Th2 cytokine responses (Fig. 3) and a Th2-dependent isotype of IgG (Table I) was broken in H. polygyrus-infected mice. Differentiation of Th cells into Th1 or Th2 subsets is determined by the presence of the priming cytokines IL-12 or IL-4, respectively, during a primary antigenic stimulus (8). One interpretation of our results (and the one that we favor) is that the induction of large quantities of IL-4 in the GALT by parasitic infection primes for a Th2 response to orally administered OVA and prevents the development of tolerance. Work in other systems has documented the ability of a polarized Th2 response to one Ag to alter or deviate the cytokine profile of the response to another Ag (20, 21) and its reversal by treatment with rIL-12 (22, 23). Tolerance of Th2 responses to orally administered OVA is restored by treatment of infected mice with rIL-12, providing further support for our contention that parasite-induced Th2 cytokines are responsible for the break in Th2 tolerance to OVA. In another model system, an established T cell tolerance to i.v. injected staphylococcal enterotoxin B was broken by infection with the nematode parasite N. brasiliensis (24). Anergic, staphylococcal enterotoxin B-tolerized, CD4+ T cells proliferated and produced IL-4 following infection with N. brasiliensis, suggesting activation, by infection, of previously tolerant cells. We show in this study that tolerance for Th1-mediated responses to OVA can still be induced during an ongoing infection, and that it is the polarized Th2 response induced by the parasite that dictates whether the response to a normally tolerizing form of orally administered Ag will be immunity or tolerance. However, although H. polygyrus maintains a strictly enteric life cycle, it does lead to some Ag nonspecific stimulation in peripheral sites, as can be seen in the responses to in vitro restimulation of spleen and PLN cells from infected mice with anti-CD3 in Figure 2 and Table II. This suggests that either activated T cells or APC bearing H. polygyrus Ags can migrate out of the GALT to peripheral sites. Although this response does not exhibit the marked Th2 polarity seen in the mucosal (MLN) response to H. polygyrus, it suggests the alternative possibility that the induction of cytokine release in the periphery in response to this mucosal infection or cellular traffic to the PLN may play a role in influencing peripheral tolerance induction to orally administered Ag.
It has been suggested that oral administration of high doses of Ag leads to tolerance because small amounts of Ag gain access to the peripheral circulation and are then presented, in the periphery, in the absence of the costimulatory signals required to induce immunity. Although not addressed directly in this study, use of this mucosal infection model should provide insight into the role of the GALT in the processing and presentation of mucosal Ag. During H. polygyrus infection, the costimulatory molecules B7.1 and B7.2 are markedly up-regulated on B220+ cells (with concomitant down-regulation on TCR-αβ+ cells) in the MLN by 8 days p.i. (Ref. 9 and data not shown). Moreover, the induction of a protective immune response to H. polygyrus has been shown recently to be B7 dependent, demonstrating that the levels of costimulation induced are sufficient for the development of immunity to the parasite (25). H. polygyrus infection is restricted to the intestine, but some parasite-induced activation of B cells and macrophages occurs in the spleen (data not shown and Table II). These initial observations have led us to speculate on the role of costimulation in the development of tolerance to orally administered Ag. If Ag presentation in the absence of costimulation is responsible for the Th1 nonresponsiveness observed in our model, one would have to assume no role for presentation of orally administered Ag in the MLN and preferential Ag presentation in the spleen. Yet the marked Th2 cytokine bias present only in the MLN in response to this enteric parasite (Ref. 12 and Table II) clearly alters Th2 nonresponsiveness to OVA. The ability to suppress both Th1 and Th2 responses in IL-12-treated, infected mice in the absence of parasite-induced Th2 cytokines shows that both types of Th responses can be tolerized in the presence of parasite-induced up-regulation of costimulation, which is not altered substantially by IL-12 treatment (data not shown). We favor the interpretation that the induction of peripheral tolerance is preceded by transient T cell activation (26) and requires a B7-dependent signal (27, 28). The ability of an Ag to induce tolerance or immunity may be decided by whether it is recognized in the context of CD28 (which seems to induce immunity) or CTLA-4 (which induces tolerance). It will be of great interest to examine the role of these ligands in orally induced nonresponsiveness in H. polygyrus-infected mice.
Taken together, we think that our results favor the interpretation that the cytokine environment in the GALT plays a key role in regulating peripheral nonresponsiveness to orally administered Ag. Whether this involves a unique role for the GALT in Ag presentation, recruitment of cells from the periphery into the GALT, or the generation and/or migration of cells that secrete regulatory cytokines, will be the subject of future experiments. Since helminthic infections with polarized cytokine responses are endemic in many developing countries (incidence can reach 100% by age 10 in some areas (9)), our observations are clearly of clinical relevance. Mucosal infection is likely to have an impact on both the therapeutic use of orally administered Ag in the treatment of autoimmune disease (6) and the disregulation of mucosal cytokine responses that appears to predispose for intestinal inflammation and inflammatory bowel disease (1).
Acknowledgements
We thank Paul Alfaro for superb technical assistance, and Drs. Abul Abbas, Bobby Cherayil, Gerburg Spiekermann, and Jeanette Thorbecke for critical review of the manuscript.
Footnotes
This work was supported by grants from National Institutes of Health (PO1DK35506 and R29DK47017) and Massachusetts General Hospital Center for the Study of Inflammatory Bowel Disease (DK43551). I.D. was supported by predoctoral training fellowship T32AI07498.
Abbreviations used in this paper: GALT, gut-associated lymphoid tissue; MLN, mesenteric lymph node; OPD, O-phenylenediamine; p.i., postinfection; PLN, popliteal lymph node; RT, room temperature.