Immune privilege within the eye is due in large part to Ag-specific, systemic down-regulation of Th1 immune responses, a phenomenon termed anterior chamber-associated immune deviation (ACAID). Since the cytokine milieu influences Th cell differentiation, we hypothesized that TGF-β, an immunosuppressive cytokine secreted by ocular cells, determines the nature of the immune response to Ags introduced into the anterior chamber. Accordingly, an in vitro model of the eye was used to determine the cytokine profile of ocular APC. TGF-β preferentially induced APC to secrete a Th2-type cytokine, IL-10, and concomitantly suppressed the production of the Th1-inducing cytokine, IL-12. APC incubated with TGF-β and anti-IL-10 Ab lost their ability to induce ACAID. In the absence of TGF-β, Ag-pulsed APC preferentially secreted IL-12 and elicited Ag-specific Th1 responses (i.e., delayed-type hypersensitivity (DTH)). However, APC pulsed with Ag and exogenous IL-10 behaved in a manner similar to ocular APC and induced Ag-specific suppression of DTH. The role of IL-10 in ACAID was confirmed in IL-10 knockout mice. Anterior chamber injection of OVA into IL-10 knockout mice elicited normal DTH responses rather than ACAID. Moreover, Ag-pulsed APC from IL-10 knockout mice were unable to induce ACAID following in vitro treatment with TGF-β. Thus, TGF-β predisposes ocular APC to secrete IL-10 during Ag processing. This, in turn, directs the immune response away from a Th1 pathway and toward a Th2-like response in which DTH is suppressed.

Primary exposure of naive CD4+ T cells to Ag results in differentiation to a defined helper type. The primary helper types are Th1 and Th2 and are characterized by distinct patterns of cytokine secretion. Th1 cells secrete IL-2, TNF-β, and IFN-γ and are involved in cell-mediated inflammatory responses (1). In direct contrast, Th2 cells secrete IL-4, IL-5, IL-6, IL-9, IL-10, and IL-13 and are efficient promoters of Ab responses (1). There is much evidence that the cytokine products of each individual Th cell subset inhibit both the differentiation and effector functions of the other. For example, IFN-γ has been shown to prevent Th2 cell proliferation, whereas IL-10 profoundly inhibits the synthesis of Th1 cytokines (2, 3). Therefore, the emergence of a Th2-type response typically results in the inhibition of Th1 differentiation and the down-regulation of Th1-mediated immune responses.

The preferential development of a particular Th cell subset has been correlated directly with either susceptibility or resistance to certain disease states. In the case of infection with Leishmaniamajor, it has been shown that C57BL/6 mice develop a Th1-dominated response and recover from disease, whereas BALB/c mice develop a Th2-type response and die from the infection (4, 5, 6, 7). Susceptibility to disease in BALB/c mice has been associated directly with the down-regulation of a protective, Th1-mediated DTH3 response. Although Th1 responses are crucial for recovery from many infectious diseases, DTH can inflict significant collateral damage to normal host tissues. Th1-mediated lesions can be tolerated in many organs; however, other tissues lack regenerative capacities and cannot retain their physiologic function as a consequence of unbridled DTH (8). This is particularly true of the eye, in which most of the crucial tissues necessary for vision lack the capacity to regenerate. The expression of DTH-mediated immunity in the eye often culminates in blindness (9, 10, 11). However, Ags introduced into the anterior chamber of the eye characteristically evoke a spectrum of immune responses in which the induction and expression of DTH are profoundly suppressed (12). This unique immunoregulatory phenomenon has been termed anterior chamber-associated immune deviation (ACAID) and is one of several important mechanisms for sustaining immune privilege in the anterior chamber of the eye (13, 14, 15). Recent studies have suggested that Ags delivered into the anterior chamber of the eye are processed by resident APC that preferentially migrate to the spleen, where they induce the generation of T cells that down-regulate DTH (16). The down-regulation of Th1 immunity appears to be the consequence of Th2-mediated cross-regulation of Th1 responses (17, 18). CD4+ T cells from the spleens of mice manifesting ACAID display a Th2-like cytokine profile in which IL-10 production is increased, while IFN-γ secretion is strongly inhibited (17, 18).

A small, but significant, category of Ags will not induce ACAID, and instead provoke strong DTH following anterior chamber presentation (11, 19, 20, 21, 22, 23, 24, 25, 26). The cytokine pattern in these mice is characteristic of a Th1-dominated response in which splenic CD4+ T cells produce increased amounts of IL-2 and IFN-γ, but insignificant amounts of IL-10 and no detectable IL-4 (18). Thus, Ags delivered into the anterior chamber of the eye can elicit either Th1- or Th2-like systemic immune responses.

Wilbanks and co-workers (27, 28) have suggested that TGF-β, present in the aqueous humor, alters the behavior and Ag-presenting characteristics of APC within the eye. Peritoneal macrophages exposed to TGF-β in vitro, at the same concentration that is present in the aqueous humor, present Ag to T cells in vivo in a manner that leads to the down-regulation of DTH and the development of an immunologic phenotype that is characteristic of ACAID.

In the present study, we wished to utilize the in vitro model of the eye to test the hypothesis that under the influence of TGF-β, APC can elaborate a Th2-type cytokine, IL-10, which prevents the induction of Th1 immunity. A corollary of this hypothesis predicts that Ags that do not induce ACAID prevent the secretion of IL-10 by TGF-β-modified APC and allow for the production of cytokines, such as IL-12, that promote the emergence of Th1 immunity.

Six- to ten-week old C57BL/6 and BALB/c mice were obtained from the mouse colony at University of Texas Southwestern Medical Center at Dallas. C57BL/6 IL-10 knockout mice (C57BL/6-IL-10tm1Cgn) and C57BL/6 IL-4 knockout mice (C57BL/6J-IL-4tm1Cgn) were purchased from The Jackson Laboratory (Bar Harbor, ME).

UV5C25 is a highly immunogenic, UV light-induced fibrosarcoma tumor cell line originally derived in BALB/c mice (kindly provided by Dr. Margaret Kripke, M. D. Anderson Tumor Institute, Houston, TX). The immunologic characteristics of UV5C25 tumor cells have been described (20). P815 is a mastocytoma tumor line derived in DBA/2 mice. UV5C25 tumor cells were maintained in complete MEM medium (JRH Biosciences, Lenexa, KS) containing 10% heat-inactivated FBS (HyClone Laboratories, Logan, UT), 2 mM l-glutamine (JRH Biosciences), 1 mM sodium pyruvate (JRH Biosciences), 2 mM MEM vitamins (JRH Biosciences), and 1% penicillin-streptomycin-fungizone solution (BioWhittaker, Walkersville, MD). P815 tumor cells were maintained in complete DMEM medium (JRH Biosciences) containing 10% heat-inactivated FBS (HyClone Laboratories), 2 mM l-glutamine (JRH Biosciences), 1 mM sodium pyruvate (JRH Biosciences), 2 mM MEM vitamins (JRH Biosciences), and 1% penicillin-streptomycin-fungizone solution (BioWhittaker). UV5C25 tumor cells in complete MEM were sonicated (1 min) with a sonic dismembrator (Artek Systems Corp., Farmingdale, NY).

Mice were anesthetized with 0.66 mg of ketamine hydrochloride (Vetalar; Park-Davis and Co., Detroit, MI) given i.p. A glass micropipette (approximately 80 μm diameter) was fitted onto a sterile infant-feeding tube (no. 5 French; Professional Medical Products, Greenwood, SC) and mounted onto a 0.1-ml Hamilton syringe (Hamilton Co., Whittier, CA). A Hamilton automatic dispensing apparatus was used to dispense 5 μl of a 20 mg/ml OVA (Sigma Chemical Co., St. Louis, MO) solution in sterile PBS (= 100 μg OVA) or 2 × 107 UV5C25 cells/ml in complete MEM medium (= 1 × 105 cells) or 2 × 107 P815 cells/ml in complete DMEM medium (= 1 × 105 cells) into the anterior chamber via the glass micropipette.

Peritoneal exudate cells (PEC) were collected by peritoneal lavage of normal mice that were injected i.p. with 1 to 1.5 ml of 3% thioglycolate medium 5 days earlier.

In a previously described model of the anterior chamber of the eye, TGF-β-treated PEC were found to mimick the function of ACAID APC (27). PEC from normal mice were collected and suspended in complete RPMI 1640 (JRH Biosciences) containing 10% FBS (HyClone Laboratories), 2 mM l-glutamine (JRH Biosciences), 10 mM HEPES buffer solution (JRH Biosciences), 1 mM sodium pyruvate solution (JRH Biosciences), 1% nonessential amino acids solution (BioWhittaker), and 1% penicillin-streptomycin-fungizone solution (BioWhittaker). PEC were incubated on plastic tissue culture dishes (Falcon 3803; Becton Dickinson Labware, Lincoln Park, NJ) at 37°C in 5% CO2 for 2.5 h. Nonadherent cells were washed off with HBSS (BioWhittaker). Adherent cells, the vast majority of which were macrophages, were collected by incubating the plates at 4°C for 15 min and gently dislodging the cells with a Nitex filter swab (Tetko, Briarcliff Manor, NY). Adherent PEC were resuspended in complete RPMI medium and aliquoted at 1 to 1.5 × 106 cells/well in a 24-well plastic tissue culture plate (Falcon 3047; Becton Dickinson Labware). In some experiments, 5 mg/ml OVA (Sigma Chemical Co.) was added to each well. In other experiments, 1 × 105 sonicated UV5C25 cells in complete MEM medium were added to each well. Human rTGF-β2 (Genzyme Diagnostics, Cambridge, MA) in complete RPMI was added to some wells (at 2 ng/ml). This treatment caused these cells to subsequently function like ocular ACAID APC. Control wells received an identical aliquot (100 μl) of complete RPMI alone. Cell cultures were incubated at 37°C in 5% CO2 overnight. The next day, all cells were collected, washed with HBSS, resuspended in complete RPMI, and infused i.v. (1–3 × 105 cells in 150 μl) into naive, syngeneic recipients.

In some experiments, mice were immunized by s.c. injection of OVA (125 μg) emulsified 1:1 in CFA (0.5 mg Mycobacterium/ml; Behring Diagnostics, La Jolla, CA) in a total volume of 50 to 100 μl. In other experiments, mice were immunized by s.c. injection of 0.5 to 2 × 106 UV5C25 cells suspended in 0.1 to 0.3 ml complete MEM or 3 × 106 P815 cells suspended in 0.2 ml complete DMEM.

Seven days after s.c. immunization, both ear pinnae of experimental and control animals were measured with a Mitutoyo engineer’s micrometer immediately before challenge. In experiments with OVA, 400 μg OVA in 20 μl sterile PBS (pH 7.4) was injected s.c. into the left ear pinnae. The right ear pinnae received 20 μl sterile PBS alone (negative control). In other experiments, X-irradiated (3000 cGy) UV5C25 cells suspended in 20 μl complete MEM (= 8 × 105 cells) were injected s.c. into the left ear pinnae. The right ear pinnae received 20 μl complete MEM alone (negative control). In other experiments, X-irradiated (3000 cGy) P815 cells suspended in 20 μl complete DMEM (= 8 × 105 cells) were injected s.c. into the left ear pinnae. The right ear pinnae received 20 μl of complete DMEM alone (negative control). Both ear pinnae were measured 24 h later, and the difference in ear pinnae size was used as a measure of DTH. Results were expressed as: specific ear swelling = (24-h measurement − 0-h measurement) for experimental ear − (24-h measurement − 0-h measurement) for control ear.

Adherent PEC (3 × 106 cells/well) were cultured for 18 to 24 h. Supernatants (1 ml) were collected and tested for the presence of IL-10 and IL-12 by double mAb sandwich ELISA. Primary and secondary mAb for each mouse cytokine were: IL-10, JES5.2A511 (rat IgG1) and SXC-1 (rat IgM); IL-12, Red-T (hamster IgG)/G297-289 (rat IgG2a) and C17.8 (rat IgG2a). Activity in culture supernatants was compared with purified murine (m) IL-10 and rmIL-12 standards. The hybridoma producing anti-IL-10 mAb (JES5.2A511) was kindly provided by Dr. J. Abrams (DNAX, Palo Alto, CA). Biotinylated anti-IL-10 mAb (SXC-1), anti-IL-12 mAb mixture (Red-T/G297-289), rmIL-12, and biotinylated anti-IL-12 mAb (C17.8) were purchased from PharMingen (San Diego, CA). D10 culture supernatant containing mIL-10 was kindly provided by Dr. N. Street (Department of Microbiology, University of Texas Southwestern Medical Center, Dallas, TX). Fifty microliters of each primary mouse mAb, anti-IL-10 at 3 μg/ml and anti-IL-12 at 8 μg/ml, were coated on sterile ELISA microplates (Corning Glass Works, Corning, NY). The plates were covered and incubated at 4°C overnight. Each well was blocked with 200 μl of PBS containing 1% BSA (Sigma Chemical Co.). Plates were incubated for 3 h at room temperature. Standards and samples (100–200 μl/well) were incubated at 4°C overnight. All samples were run in triplicate. Biotinylated rat anti-mouse secondary mAb (1–2 μg/ml) was added at 100 μl/well and incubated at room temperature for 60 min. Peroxidase-conjugated streptavidin (0.25–0.5 μg/ml) (Jackson ImmunoResearch Laboratories, West Grove, PA) was added (100 μl/well) and incubated at room temperature for 45 min. Plates were washed with 0.05% Tween-20 (Sigma Chemical Co.) in PBS between each step. Finally, 2,2-Azino-bis (3-ethylbenz-thiazoline-6-sulfonic acid; ABTS; 1 mg/ml; Sigma Chemical Co.) substrate with 0.003% H2O2 was added at 100 μl/well and allowed to develop at room temperature for 10 to 90 min. Plates were read with an ELISA Microplate Reader (Molecular Devices, ThermoMax, Menlo Park, CA) at an OD of 405 nm.

Adherent PEC (1–3 × 106 cells/well) were incubated overnight, as described. In some experimemts, anti-IL-10 mAb (20 μg/ml; JES5.2A511) was added to the cultures. Control wells received an isotype-matched mAb (20 μg/ml; 11B11, rat IgG1; American Type Culture Collection, Rockville, MD). In other experiments, anti-IL-12 mAb (10 μg/ml; Red-T/G297-289) was added to the well. Control wells received an irrelevant control mAb (10 μg/ml; 11B11). The next day, all cells were collected, washed, resuspended in complete RPMI, and infused i.v. (1.5–3 × 105 cells in 150 μl) into naive, syngeneic recipients.

Adherent PEC (1–3 × 106 cells/well) were incubated overnight with or without mIL-10 (5 or 10 U mIL-10/ml). Other cultures received 10 ng/ml rmIL-10 (PharMingen, San Diego, CA). The next day, all cells were collected, washed, resuspended in complete RPMI, and infused i.v. (1.5–4 × 105 cells in 150 μl) into naive, syngeneic recipients.

Unless otherwise indicated, all experimental and control groups contained five animals (n = 5). Differences between groups were analyzed by Student’s t test. p values < 0.05 were considered significant.

Experiments were performed to confirm that the in vitro model of ACAID mimicked the in vivo counterpart. Accordingly, PEC (APC) from BALB/c mice were pulsed in vitro for 24 h with OVA in the presence or absence of the ocular cytokine, TGF-β. Ag-pulsed APC were injected i.v. into normal, syngeneic BALB/c mice. APC recipients, as well as normal untreated mice, were immunized s.c. with 125 μg OVA 7 days later. DTH responses were assessed 7 days after immunization. The results demonstrate that APC pulsed with OVA in the presence of TGF-β induced down-regulation of OVA-specific DTH and prevented the hosts from responding to the normally immunogenic s.c. injection of OVA (Fig. 1,A). The suppression of OVA-specific DTH was comparable with that found in mice primed in the anterior chamber with OVA before s.c. immunization (Fig. 1 B). Thus, the in vitro model of ACAID produced down-regulation of OVA-specific DTH that was of the same magnitude as that produced by anterior chamber presentation of Ag.

FIGURE 1.

Suppression (ACAID) to OVA in vitro and in vivo. A, Adherent BALB/c PEC were incubated overnight with soluble OVA (5 mg/ml), either with (+) or without (−) 2 ng/ml TGF-β. The next day, 1 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. B, Soluble OVA (100 μg) was injected intracamerally into naive BALB/c mice (OVA, i.c.). Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p > 0.05 for TGF-β (A) or OVA, i.c. (B) vs negative controls. ▴, p < 0.0003 for medium control and positive control vs negative control; **p < 0.006 for positive control vs OVA, i.c. and negative control.

FIGURE 1.

Suppression (ACAID) to OVA in vitro and in vivo. A, Adherent BALB/c PEC were incubated overnight with soluble OVA (5 mg/ml), either with (+) or without (−) 2 ng/ml TGF-β. The next day, 1 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. B, Soluble OVA (100 μg) was injected intracamerally into naive BALB/c mice (OVA, i.c.). Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p > 0.05 for TGF-β (A) or OVA, i.c. (B) vs negative controls. ▴, p < 0.0003 for medium control and positive control vs negative control; **p < 0.006 for positive control vs OVA, i.c. and negative control.

Close modal

As mentioned earlier, some Ags do not induce ACAID (non-ACAID) and instead provoke vigorous DTH following anterior chamber presentation. It was important to determine whether the in vitro model of ACAID reproduced the in vivo phenotype with such Ags. Experiments similar to those described above were performed using the highly immunogenic, BALB/c UV5C25 tumor cell line. BALB/c APC were pulsed in vitro for 24 h with 1 × 105 sonicated UV5C25 tumor cells, either in the presence or absence of TGF-β. APC were then injected i.v. into normal BALB/c mice. Mice were immunized with 2 × 106 viable UV5C25 7 days later. DTH responses to UV5C25 tumor Ags were assessed 7 days after immunization. Unlike the results with OVA, APC pulsed with UV5C25 tumor Ag in the presence of TGF-β failed to induce down-regulation of DTH responses (Fig. 2,A). APC recipients mounted DTH responses that were indistinguishable from those in mice primed s.c. or in the anterior chamber of the eye (Fig. 2 B). Thus, the in vitro model of ACAID recapitulates the in vivo counterpart using both ACAID- and non-ACAID-inducing Ags.

FIGURE 2.

UV5C25 tumor cells fail to induce suppression of Th1 immunity in vitro and in vivo. A, Adherent BALB/c PEC were incubated overnight with 1 × 105 sonicated UV5C25 cells, either with (+) or without (−) 2 ng/ml TGF-β. The next day, 2 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 2 × 106 UV5C25 cells. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 2 × 106 UV5C25 cells. Negative control animals (Neg. Control) were not immunized. B, UV5C25 tumor cells (1 × 105 cells) were injected intracamerally into naive BALB/c mice. Seven days later, mice were immunized s.c. with 5 × 105 UV5C25 cells. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 5 × 105 UV5C25 cells. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. ▴p < 0.03 for TGF-β and medium control vs negative control. **p > 0.05 for TGF-β and medium control (A) or UV5C25, i.c. (B) vs positive control. *p < 0.001 for UV5C25, i.c. vs negative control.

FIGURE 2.

UV5C25 tumor cells fail to induce suppression of Th1 immunity in vitro and in vivo. A, Adherent BALB/c PEC were incubated overnight with 1 × 105 sonicated UV5C25 cells, either with (+) or without (−) 2 ng/ml TGF-β. The next day, 2 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 2 × 106 UV5C25 cells. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 2 × 106 UV5C25 cells. Negative control animals (Neg. Control) were not immunized. B, UV5C25 tumor cells (1 × 105 cells) were injected intracamerally into naive BALB/c mice. Seven days later, mice were immunized s.c. with 5 × 105 UV5C25 cells. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 5 × 105 UV5C25 cells. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. ▴p < 0.03 for TGF-β and medium control vs negative control. **p > 0.05 for TGF-β and medium control (A) or UV5C25, i.c. (B) vs positive control. *p < 0.001 for UV5C25, i.c. vs negative control.

Close modal

Previous studies have demonstrated that cytokines elaborated by APC can profoundly influence the elicitation of a Th1 vs Th2 response (29, 30). The next experiments examined the effect of TGF-β on the cytokines secreted by Ag-pulsed APC in vitro. We began by testing the hypothesis that TGF-β renders the APC tolerogenic by enhancing their production of the Th2-inducing cytokine, IL-10. BALB/c APC were pulsed in vitro for 24 h with OVA or sonicated UV5C25 Ag in the presence or absence of TGF-β. Culture supernatants were collected and tested for the presence of IL-10 and IL-12 by sandwich ELISA. In some experiments, the APC were subsequently injected i.v. into normal BALB/c mice to verify function. As shown in Figure 3, APC exposed to OVA and TGF-β secreted significantly higher levels of IL-10 than APC exposed to OVA in the absence of TGF-β. As previously shown in Figure 1,A, these APC induced suppression of OVA-specific DTH when injected i.v. into normal BALB/c mice. APC exposed to UV5C25 tumor Ag and TGF-β, however, did not produce significantly higher levels of IL-10 than UV5C25 Ag alone (Fig. 3). In addition, the total quantity of IL-10 produced by APC exposed to UV5C25 Ag was significantly lower than that produced by APC exposed to OVA in the presence or absence of TGF-β. These results indicate that TGF-β stimulates the production of IL-10 by APC, but this effect is highly dependent on the nature of the Ag encountered. Ags that do not induce ACAID in vivo (i.e., UV5C25 tumor Ags) inhibit the production of IL-10 by APC in vitro.

FIGURE 3.

An Ag that induces ACAID stimulates APC secretion of IL-10. Adherent BALB/c PEC were incubated overnight in the presence of 5 mg/ml OVA or 1 × 105 sonicated UV5C25 cells, either with (+) or without (−) 2 ng/ml TGF-β. In this representative experiment, culture supernatants (1 ml) were collected and analyzed for IL-10 by sandwich ELISA. Results are expressed as mean units of IL-10/ml supernatant. Differences between OVA groups were significant (p = 0.01). Differences between UV5C25 groups were not significant (p > 0.05).

FIGURE 3.

An Ag that induces ACAID stimulates APC secretion of IL-10. Adherent BALB/c PEC were incubated overnight in the presence of 5 mg/ml OVA or 1 × 105 sonicated UV5C25 cells, either with (+) or without (−) 2 ng/ml TGF-β. In this representative experiment, culture supernatants (1 ml) were collected and analyzed for IL-10 by sandwich ELISA. Results are expressed as mean units of IL-10/ml supernatant. Differences between OVA groups were significant (p = 0.01). Differences between UV5C25 groups were not significant (p > 0.05).

Close modal

In other experiments, we tested the hypothesis that TGF-β renders APC tolerogenic by inhibiting their production of a Th1-promoting cytokine, IL-12. APC pulsed with OVA in the presence of TGF-β produced significantly lower amounts of IL-12 than APC exposed to OVA alone (Fig. 4). Although TGF-β impaired the secretion of IL-12 by APC pulsed with UV5C25 tumor Ags, the amount of IL-12 produced was comparable with that produced by APC pulsed with OVA in the absence of TGF-β (Fig. 4). It is noteworthy that in both cases, these APC promoted the development of Ag-specific DTH following adoptive transfer (Figs. 1 A and 2A). Thus, it appears that Ag presentation may require a certain threshold of IL-12 secretion to promote the induction of Th1 immunity. Likewise, the data suggest that a minimal level of IL-10 secretion is associated with the down-regulation of Th1 responses and/or preferential induction of Th2 responses. Although TGF-β, and presumably the intraocular environment, influences the behavior of APC in promoting a Th2-type systemic immune response, certain Ags (e.g., UV5C25 tumor) can redirect APC cytokine secretion to favor the development and emergence of Th1 immunity. We next wanted to determine whether the secretion of either cytokine had a direct effect on the acquired phenotype of the APC.

FIGURE 4.

Ags that do not induce ACAID stimulate APC secretion of IL-12. Adherent BALB/c PEC were incubated overnight in the presence of 5 mg/ml OVA or 1 × 105 sonicated UV5C25 cells, either with (+) or without (−) 2 ng/ml TGF-β. In this representative experiment, culture supernatants (1 ml) were collected and analyzed for IL-12 by sandwich ELISA. Results are expressed as mean picograms of IL-12/ml supernatant. Differences between OVA groups and between UV5C25 groups were significant (p < 0.05).

FIGURE 4.

Ags that do not induce ACAID stimulate APC secretion of IL-12. Adherent BALB/c PEC were incubated overnight in the presence of 5 mg/ml OVA or 1 × 105 sonicated UV5C25 cells, either with (+) or without (−) 2 ng/ml TGF-β. In this representative experiment, culture supernatants (1 ml) were collected and analyzed for IL-12 by sandwich ELISA. Results are expressed as mean picograms of IL-12/ml supernatant. Differences between OVA groups and between UV5C25 groups were significant (p < 0.05).

Close modal

To test the possibility that the induction of IL-10 secretion by TGF-β is involved in suppressive, Th2-inducing Ag presentation, we removed IL-10 by incubating APC in the presence of a neutralizing concentration of anti-IL-10 mAb. BALB/c APC were incubated for 24 h with OVA and with or without TGF-β. Wells received either anti-IL-10 mAb (20 μg/ml) or an isotype-matched Ab control. Ag-pulsed APC were injected i.v. into normal, BALB/c mice. As expected, APC incubated with OVA, TGF-β, and control Ab induced suppression of DTH (Fig. 5). By contrast, anti-IL-10 mAb treatment reversed the effects of TGF-β on APC function; APC incubated with OVA, TGF-β, and anti-IL-10 mAb induced positive DTH to subsequent challenge (Fig. 5). Therefore, TGF-β-induced IL-10 secretion by the APC is necessary for that APC to acquire a suppressive, Th2-inducing phenotype. Removal of IL-10 during the early stages of Ag processing blocks this transformation in vitro.

FIGURE 5.

Removal of IL-10 prevents the generation of ACAID APC in vitro. Adherent BALB/c PEC were incubated overnight in the presence of OVA (5 mg/ml) and the following: 2 ng/ml TGF-β and 20 μg/ml irrelevant mAb (TGF-β); 20 μg/ml irrelevant mAb alone (Medium Ctl.); or 2 ng/ml TGF-β and 20 μg/ml anti-IL-10 mAb (anti-IL-10). The next day, 1.5 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p < 0.0002 for anti-IL-10 vs negative control. **p > 0.05 for anti-IL-10 vs positive control.

FIGURE 5.

Removal of IL-10 prevents the generation of ACAID APC in vitro. Adherent BALB/c PEC were incubated overnight in the presence of OVA (5 mg/ml) and the following: 2 ng/ml TGF-β and 20 μg/ml irrelevant mAb (TGF-β); 20 μg/ml irrelevant mAb alone (Medium Ctl.); or 2 ng/ml TGF-β and 20 μg/ml anti-IL-10 mAb (anti-IL-10). The next day, 1.5 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p < 0.0002 for anti-IL-10 vs negative control. **p > 0.05 for anti-IL-10 vs positive control.

Close modal

Because the removal of IL-10 from cultures containing TGF-β blocked the generation of suppressive, Th2-inducing APC (ACAID APC), it appeared that the effect of TGF-β was mediated directly through the secretion of IL-10. To confirm the role of IL-10, we added mIL-10 to APC cultures in vitro. We hypothesized that, in the absence of TGF-β, the addition of mIL-10 to the cultures would still generate ACAID APC. We began our studies by adding a biologically relevant 10 U mIL-10/ml culture supernatant. As before, APC incubated with OVA and TGF-β induced suppression of DTH (Fig. 6,A). APC incubated with OVA and mIL-10 in the absence of TGF-β were able to induce suppression of OVA-specific DTH in vivo (Fig. 6,A). This was confirmed in additional studies using 5 U mIL-10/ml (data not shown). Because the crude supernatant containing mIL-10 also contained nominal levels of IL-3, IL-4, and IL-5, we wanted to eliminate any possible role for these cytokines. Accordingly, purified rmIL-10 was used in subsequent assays. APC cultured with OVA and 10 ng/ml rmIL-10 also suppressed OVA-specific DTH (Fig. 6,B). Interestingly, APC pulsed with UV5C25 Ag in the presence of TGF-β and mIL-10 or mIL-10 alone were able to induce suppression of UV5C25-specific DTH (Fig. 6 C). This strongly suggests that the failure of UV5C25 tumor Ags to induce IL-10 secretion by APC in vitro is a key factor in the elicitation of Th1 immunity. In summary, these results show that the effects of TGF-β are mediated through IL-10 and that a nominal level of exogenous IL-10 is sufficient to confer a suppressive phenotype on APC in vitro.

FIGURE 6.

IL-10 is sufficient to generate ACAID APC in vitro. Adherent BALB/c PEC were incubated overnight in the presence of OVA (5 mg/ml), either with (+) or without (−) 2 ng/ml TGF-β. Some test wells (mIL-10) contained crude mIL-10 (D10 supernatant) at 10 U/ml (A and B). Other test wells received rmIL-10 at 10 ng/ml (B). The next day, 1.5 to 2.5 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. C, Adherent BALB/c PEC were incubated overnight in the presence of 1 × 105 sonicated UV5C25 cells/ml, either with (+) or without (−) 2 ng/ml TGF-β and with (+) or without (−) 10 U/ml crude mIL-10 (D10 supernatant). The next day, 4 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 5 × 105 UV5C25 cells. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 5 × 105 UV5C25 cells. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p > 0.05 for mIL-10 groups (A and B) and rmIL-10 (B) vs negative control. **p < 0.0004 for mIL-10 groups (A and B) and rmIL-10 (B) vs positive control. ▴, p > 0.05 for mIL-10 and TGF-β + mIL-10 vs negative control. ▪, p < 0.0004 for mIL-10 and TGF-β + mIL-10 vs positive control.

FIGURE 6.

IL-10 is sufficient to generate ACAID APC in vitro. Adherent BALB/c PEC were incubated overnight in the presence of OVA (5 mg/ml), either with (+) or without (−) 2 ng/ml TGF-β. Some test wells (mIL-10) contained crude mIL-10 (D10 supernatant) at 10 U/ml (A and B). Other test wells received rmIL-10 at 10 ng/ml (B). The next day, 1.5 to 2.5 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. C, Adherent BALB/c PEC were incubated overnight in the presence of 1 × 105 sonicated UV5C25 cells/ml, either with (+) or without (−) 2 ng/ml TGF-β and with (+) or without (−) 10 U/ml crude mIL-10 (D10 supernatant). The next day, 4 × 105 cells were injected i.v. into naive BALB/c mice. Seven days later, mice were immunized s.c. with 5 × 105 UV5C25 cells. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 5 × 105 UV5C25 cells. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p > 0.05 for mIL-10 groups (A and B) and rmIL-10 (B) vs negative control. **p < 0.0004 for mIL-10 groups (A and B) and rmIL-10 (B) vs positive control. ▴, p > 0.05 for mIL-10 and TGF-β + mIL-10 vs negative control. ▪, p < 0.0004 for mIL-10 and TGF-β + mIL-10 vs positive control.

Close modal

As mentioned, previous studies in normal mice had shown that ACAID correlates with increased production of IL-10 in the spleen (18). Curiously, these studies were unable to detect the presence of IL-4 protein or mRNA in these animals. To confirm both our present findings and this earlier work, we utilized mice that had a homozygous mutation in the gene for IL-4 or IL-10 (31, 32). OVA was injected into the anterior chamber of IL-4- and IL-10-deficient mice. Seven days later, mice were s.c. immunized with OVA in CFA. DTH was assessed 7 days after s.c. immunization. As shown in Figure 7,A, deletion of the IL-4 gene did not affect the induction of ACAID. In direct contrast, deletion of the gene for IL-10 prevented the induction of ACAID to OVA (Fig. 7,B). Additional experiments were performed to confirm that these results were not limited to soluble Ags, but would also apply to particulate Ags known to induce ACAID. P815 mastocytoma induces the down-regulation of DTH responses to DBA/2 alloantigens when inoculated into the anterior chamber of allogeneic mice (33). The results shown in Figure 7,C indicate that, like the soluble Ag OVA, the particulate Ag P815 (DBA/2; H-2d) failed to induce ACAID in IL-10 knockout mice (C57BL/6; H-2b). Therefore, the production of IL-10 is necessary for the ACAID state in vivo. We next wanted to confirm the role of IL-10 in the generation of suppressive, Th2-inducing ACAID APC in vitro. IL-10-deficient APC were incubated with OVA, either in the presence or absence of TGF-β, and injected i.v. into normal, syngeneic C57BL/6 mice. As shown in Figure 8,A, IL-10-deficient APC were unable to induce ACAID. These results confirm the requirement for IL-10 production by the APC for the induction of the suppressive state. We next wanted to determine whether the secretion of IL-10 was necessary solely for the APC to obtain a suppressive phenotype (i.e., autocrine effects) or whether IL-10 secretion was also necessary for the APC to induce a suppressive, Th2-type response in the spleen (i.e., autocrine and paracrine effects). Accordingly, IL-10-deficient APC were incubated for 24 h with exogenous mIL-10 and injected i.v. into normal, syngeneic C57BL/6 mice. The results show that IL-10-deficient APC incubated with OVA, TGF-β, and mIL-10 were able to induce suppression in normal, syngeneic C57BL/6 mice (Fig. 8 B). Moreover, exposure to exogenous mIL-10 in the absence of TGF-β also conferred a suppressive phenotype on IL-10-deficient APC in vitro. Thus, the in vitro data suggest that the requirement for IL-10 in the down-regulation of Th1 immunity may be local and limited to initial ocular Ag processing and may not be necessary during subsequent presentation of peptide fragments by APC to T cells in lymphoid tissues.

FIGURE 7.

IL-10, but not IL-4, knockout mice fail to support the induction of ACAID in vivo. Soluble OVA (100 μg) was injected intracamerally into naive wild-type C57BL/6 mice (B6 w.t.) and C57BL/6 IL-4 (A) or IL-10 (B) knockout mice (IL-4 k.o. and IL-10 k.o., respectively). Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days after s.c. immunization. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. C, P815 tumor cells (8 × 105 cells) were injected intracamerally into naive wild-type C57BL/6 mice (B6 w.t.) and C57BL/6 IL-10 knockout mice (IL-10 k.o.). Seven days later, mice were immunized s.c. with 3 × 106 P815 cells. DTH was assessed 7 days after s.c. immunization. Positive control animals (Pos. Control) were immunized s.c. with 3 × 106 P815 cells. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p > 0.05 for IL-4 k.o. vs negative control. **p < 0.002 for IL-4 k.o. vs positive control. ▴, p < 0.002 for IL-10 k.o. (B and C) vs negative control. ▪, p > 0.05 for IL-10 k.o. (B and C) vs positive control.

FIGURE 7.

IL-10, but not IL-4, knockout mice fail to support the induction of ACAID in vivo. Soluble OVA (100 μg) was injected intracamerally into naive wild-type C57BL/6 mice (B6 w.t.) and C57BL/6 IL-4 (A) or IL-10 (B) knockout mice (IL-4 k.o. and IL-10 k.o., respectively). Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days after s.c. immunization. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. C, P815 tumor cells (8 × 105 cells) were injected intracamerally into naive wild-type C57BL/6 mice (B6 w.t.) and C57BL/6 IL-10 knockout mice (IL-10 k.o.). Seven days later, mice were immunized s.c. with 3 × 106 P815 cells. DTH was assessed 7 days after s.c. immunization. Positive control animals (Pos. Control) were immunized s.c. with 3 × 106 P815 cells. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p > 0.05 for IL-4 k.o. vs negative control. **p < 0.002 for IL-4 k.o. vs positive control. ▴, p < 0.002 for IL-10 k.o. (B and C) vs negative control. ▪, p > 0.05 for IL-10 k.o. (B and C) vs positive control.

Close modal
FIGURE 8.

IL-10-deficient APC fail to support the induction of ACAID in vitro. A, Adherent, IL-10-deficient PEC were incubated overnight in the presence of 5 mg/ml OVA and 2 ng/ml TGF-β alone (TGF-β), or medium alone (Medium Ctl.). The next day, 3 × 105 cells were injected i.v. into naive, wild-type C57BL/6 mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. B, mIL-10 restores the ability of IL-10-deficient APC to induce suppression in wild-type mice. Adherent, IL-10-deficient PEC were incubated overnight in the presence of 5 mg/ml OVA and the following: 2 ng/ml TGF-β plus 10 U/ml mIL-10 (TGF-β + mIL-10); 10 U/ml mIL-10 alone (mIL-10); or medium alone (Medium Ctl.). As a control, wild-type PEC were incubated with OVA and TGF-β alone (TGF-β, w.t.). The next day, 5 × 105 cells were injected i.v. into naive, wild-type mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p < 0.0002 for TGF-β and medium control vs negative control. **p > 0.05 for TGF-β and medium control vs positive control. ▴, p > 0.05 for TGF-β + mIL-10 and mIL-10 vs negative control. ▪, p < 0.0002 for TGF-β + mIL-10 and mIL-10 vs positive control.

FIGURE 8.

IL-10-deficient APC fail to support the induction of ACAID in vitro. A, Adherent, IL-10-deficient PEC were incubated overnight in the presence of 5 mg/ml OVA and 2 ng/ml TGF-β alone (TGF-β), or medium alone (Medium Ctl.). The next day, 3 × 105 cells were injected i.v. into naive, wild-type C57BL/6 mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. B, mIL-10 restores the ability of IL-10-deficient APC to induce suppression in wild-type mice. Adherent, IL-10-deficient PEC were incubated overnight in the presence of 5 mg/ml OVA and the following: 2 ng/ml TGF-β plus 10 U/ml mIL-10 (TGF-β + mIL-10); 10 U/ml mIL-10 alone (mIL-10); or medium alone (Medium Ctl.). As a control, wild-type PEC were incubated with OVA and TGF-β alone (TGF-β, w.t.). The next day, 5 × 105 cells were injected i.v. into naive, wild-type mice. Seven days later, mice were immunized s.c. with 125 μg OVA in CFA. DTH was assessed 7 days later. Positive control animals (Pos. Control) were immunized s.c. with 125 μg OVA in CFA. Negative control animals (Neg. Control) were not immunized. All results are expressed as mean swelling ± SEM. *p < 0.0002 for TGF-β and medium control vs negative control. **p > 0.05 for TGF-β and medium control vs positive control. ▴, p > 0.05 for TGF-β + mIL-10 and mIL-10 vs negative control. ▪, p < 0.0002 for TGF-β + mIL-10 and mIL-10 vs positive control.

Close modal

In the present study, we have attempted to correlate the production of cytokines with the state of Ag presentation in an immune privileged site. Immune privileged sites such as the eye have evolved a number of mechanisms that protect against potentially deleterious immune responses. ACAID protects the eye from the damaging bystander effects of DTH responses and can be elicited by exposure to certain Ags in the ocular compartment. Other Ags, however, elicit a systemic, positive immunization when injected into the anterior chamber of the eye. Previous data have shown that the ACAID state correlates with the development of a systemic Th2-type response, whereas the non-ACAID state correlates with a systemic Th1 response. We hypothesized that a key factor influencing both the ocular APC and the resultant immune response was the initial cytokine repertoire elicited by Ag exposure.

Various parameters have been shown to influence the differentiation of a particular Th response, including Ag type, Ag dose, the type of APC, and the presence of cytokines (34, 35). For example, Schistosoma mansoni eggs preferentially elicit a Th2 response (36), whereas Listeria monocytogenes stimulates a Th1 response (37). The type of APC also has a profound effect on the resultant immune response. Langerhans cells and dendritic cells normally present Ags such that Th1 immunity is induced (38, 39), whereas B cells tend to induce a Th2 response (40). A major factor in the disparity in APC function is the elaboration of particular cytokines upon Ag encounter. One of the most important factors in Th differentiation is the cytokine milieu. The most potent cytokine inducer of Th2 cells is IL-4. Several types of immune cells can produce IL-4 in the initial immune response, including NK1.1+CD4+ T cells, γδ T cells, CD8+ T cells, and/or mast cells (35, 41). Because our previous data had shown that IL-4 was not induced in ACAID animals and our present data indicate that ACAID can be induced in IL-4 knockout mice, we predicted that another cytokine capable of inducing Th2 development was involved. As previously mentioned, IL-10 is known to inhibit the induction of Th1 immunity and allows for the elaboration of a Th2 response. Since IL-10 can be produced by certain APC (42, 43), we hypothesized that under the influence of TGF-β, the preferential production of IL-10 by the APC leads to the development of Th2 immunity in ACAID. In addition, we hypothesized that the autoregulatory effects of IL-10 on APC function might be a critical factor in the ability of an APC to present Ag in a suppressive manner (3).

In direct contrast to IL-4 or IL-10, IL-12 strongly induces the development of Th1 cells (44). Activation of monocytes/macrophages usually results in the elaboration of proinflammatory, IL-12 cytokine (30). The effect of IL-12 is twofold. One effect is the direct stimulation of Th1 differentiation (41). Another effect is the stimulation of IFN-γ production by T and NK cells. IFN-γ is a potent inducer of Th1 immunity both by the additional stimulation of IL-12 secretion and the inhibition of IL-4 function (29, 30). On the other hand, IL-10 and IL-4 are powerful inhibitors of IL-12 production (45, 46). Because previous experiments had measured increased levels of IFN-γ in the immunizing or non-ACAID state, we predicted that the preferential secretion of IL-12 by the APC was responsible for the resultant Th1 immune response.

Our data show that at least one factor in the anterior chamber of the eye, TGF-β, has a profound effect on the cytokines produced by Ag-stimulated APC. We have shown that the presence of TGF-β can significantly enhance the production of IL-10, as measured by sandwich ELISA. In direct contrast, TGF-β was also found to significantly suppress the production of the proinflammatory cytokine, IL-12 (p70). These cytokine data were confirmed with C57BL/6 APC (data not shown). These results suggest that TGF-β contributes to the maintenance of immune privilege by predisposing an APC to the preferential production of IL-10 cytokine, which in turn might down-regulate or alter certain Ag-presenting functions leading to the generation of a suppressive, Th2-type response. The simultaneous inhibition of IL-12 secretion contributes to this response. In addition to the effect of TGF-β alone, it appears that the Ag itself can have a significant impact on the APC cytokine profile. We have found that soluble OVA induces a twofold increase in the production of IL-10 as compared with UV5C25 tumor Ags. In direct contrast, UV5C25 tumor Ags stimulated a twofold increase in the production of IL-12 as compared with OVA. In addition, UV5C25 tumor Ags appear to completely inhibit the TGF-β-induced production of IL-10. UV5C25 Ags did not completely override the suppressive effects of TGF-β on IL-12 secretion, however. We conclude then that the immune privileged eye is endowed with at least one factor, TGF-β, that has a profound effect on local APC cytokine production. This effect is the concomitant enhancement of IL-10 and suppression of IL-12 secretions. However, the effect of TGF-β can be inhibited or circumvented depending on the specific Ag encountered.

The systemic suppression of DTH to OVA and the systemic immunization to UV5C25 Ags can be directly correlated with the initial APC cytokine profile. We have shown that the suppressive, ACAID state correlates with increased IL-10 and decreased IL-12 secretion in vitro. Conversely, we have shown that the immunizing, Th1 state correlates with decreased IL-10 and increased IL-12 secretion in vitro. Our data also indicate that there may be a minimal threshold of cytokine secretion necessary for an APC to be Th1 or Th2 inducing in nature. The results in Figure 3 suggested that IL-10 secretion ≥5 U/ml culture supernatant in vitro may be necessary for eventual suppressive, Th2-inducing Ag presentation in vivo. Subsequent experiments in which APC were pulsed with OVA and 3 or 5 U/ml mIL-10 in vitro have confirmed that 5 U/ml, but not 3 U/ml, mIL-10 is sufficient for the induction of ACAID in vivo (data not shown). IL-12 secretion ≥500 pg/ml is indicative of eventual Th1 immunity in vivo (Fig. 4).

We had hypothesized that the effect of TGF-β on APC function was mediated through IL-10 and that the autoregulatory effects of IL-10 were a critical component of ACAID Ag presentation. To test this hypothesis, we removed IL-10 from the APC cultures. Even in the presence of TGF-β, treatment with neutralizing anti-IL-10 Ab in vitro completely reversed subsequent ACAID Ag presentation in vivo. These data indicated that the suppressive effect of TGF-β on macrophage APC was mediated directly through the enhanced secretion of IL-10. If this were true, then the addition of IL-10 in the absence of TGF-β would induce a suppressive phenotype. We have shown that the addition of 10 U mIL-10/ml of culture supernatant in vitro completely substituted for the effect of TGF-β alone. Because the D10 supernatant used in these studies also contained nominal levels of IL-3, IL-4, and IL-5 in addition to IL-10, we wanted to rule out any effect of these ancillary cytokines. The addition of 10 ng rmIL-10/ml culture supernatant also effectively substituted for D10 supernatant or TGF-β alone. In total, these data show that the presence of IL-10 in the APC microenvironment is critical for eventual ACAID Ag presentation. These data demonstrate that the effect of TGF-β on macrophage APC can be directly mediated through the induced secretion of IL-10. Interestingly, the addition of 10 U/ml mIL-10 to UV5C25-pulsed APC induced ACAID to UV5C25 tumor Ags in vivo. This indicates that a lack of suppression to certain Ags, such as UV5C25, in the immune privileged eye is the direct result of insufficient secretion of IL-10. It therefore appears that the secretion of IL-10 serves as a primary determinant in the initial decision to respond in a Th1- vs Th2-mediated manner. Teleologically, the eye may be generally protected against harmful immune responses to innocuous agents, such as soluble proteins. However, if a more threatening agent is encountered, such as a growing tumor or a pathogen, circumvention of ACAID would be preferred over death. Unfortunately, weakly immunogenic tumors capable of inducing ACAID can escape immune surveillance and grow progressively, eventually destroying the eye and killing the host.

Although IL-12 is not reported to have a direct effect on macrophages, we wanted to determine whether neutralizing Ab to IL-12 affected the immunizing phenotype of ocular APC pulsed with UV5C25 Ag. As expected, the neutralization of IL-12 in the culture supernatant did not have any direct effect on APC function in vivo (data not shown). This finding contrasts with the suggested role for IL-10, since IL-10 is shown to have a direct effect on APC in the microenvironment of the in vitro eye. It is probable that the enhanced production of IL-12 in the non-ACAID state primarily serves as a stimulant to Th1 differentiation in the spleen.

A previous report had shown that systemic anti-IL-10 Ab treatment in vivo abrogates ACAID (18). We have confirmed those results in mice deficient in the gene for IL-10. We were unable to induce ACAID to OVA or P815 tumor line injected into the anterior chamber of IL-10 knockout mice. Furthermore, our in vitro studies showed that in the presence or absence of TGF-β, APC from IL-10 knockout mice could not generate ACAID to OVA. In addition, we have confirmed that the production of IL-10, but not IL-12, is impaired in IL-10 knockout mice (data not shown). These results demonstrate that the ACAID APC must have the capability to produce IL-10 for the generation of the suppressive state in vivo. Our in vitro data indicate that the production of IL-10 is necessary in the local, ocular environment. We propose that the effect of IL-10 in the local environment is to feedback on the APC in an autoregulatory loop and thereby alter normal Ag presentation function. We have also tested whether the production of IL-10 by the APC is necessary once that cell has migrated from the eye to lymphoid organs such as the spleen. The addition of exogenous IL-10 to the IL-10-knockout APC cultures restored the ability of these cells to generate suppression in wild-type mice. We conclude that, although IL-10 is requisite in the local environment of the eye, the production of IL-10 by the APC is not necessary once that cell migrates to lymphoid tissues such as the spleen.

Many cell types have been reported to produce IL-10 cytokine, including keratinocytes (47), T cells (3), B cells (42), mast cells (48), and monocytes (43). Since its discovery, countless studies have shown that IL-10 can have a profound effect on an immune response. It has been shown that IL-10 generally inhibits cell-mediated immune responses such as contact hypersensitivity and DTH, while promoting the development of Ab responses (29, 49). Apart from the direct effect on T cells, these studies have concluded that the alteration of macrophage function by IL-10 plays a key role in the preferential development of Th2-type responses. One study with experimental allergic encephalomyelitis in SJL mice has demonstrated that depletion of IL-4 or IL-10 in nonresponder mice leads to the expansion of previously nonexistent, responding Th1 cells (50). This indicates that the presence of certain cytokines before Ag exposure can regulate Ag presentation and eventual Th development. We have shown definitively that a minimal level of IL-10 cytokine present at the time of Ag exposure can induce APC to subsequently present Ag to a population of T cells that suppress DTH.

TGF-β is expressed in multiple immune privileged sites, including the anterior chamber of the eye (51), the vitreous humor (52), the central nervous system (53), the testes (54), and the adrenal cortex (55). Apart from its other immunomodulatory effects, we predict that the presence of TGF-β contributes to immune priviege in each of these sites by the described alteration of APC cytokine secretion and function. The mechanism whereby TGF-β induces IL-10 secretion remains a mystery. However, it is noteworthy that in other models of ACAID, Ag-bearing cells introduced into the anterior chamber undergo apoptosis following exposure to Fas ligand expressed within the eye and elaborate IL-10 before succumbing to apoptotic death (56). We have preliminary evidence that either aqueous humor or TGF-β also induces apoptosis of APC in vitro. Thus, it is possible that within the anterior chamber, Ag-bearing cells are exposed to at least two conditions that could culminate in apoptosis and the elaboration of IL-10. The presence of duplicate pathways for inducing IL-10 production seems quite plausible considering the redundancy of the immune system. The presence of an IL-10-enriched milieu would favor the development of a Th2 response. The suppression of IL-12 and probably other proinflammatory cytokines also contributes to this effect by discouraging the differentiation and activation of Th1 cells.

The reagents and intellect of Dr. Nancy Street are greatly appreciated. We thank Dr. Michael Bennet and Dr. Christopher Lu for their advice.

1

This work was supported by National Institutes of Health Grant EY05631 and an unrestricted grant from Research to Prevent Blindness, New York, NY.

3

Abbreviations used in this paper: DTH, delayed-type hypersensitivity; ACAID, anterior chamber-associated immune deviation; m, murine; PEC, peritoneal exudate cell(s).

1
Mosmann, T. R., S. Sad.
1996
. The expanding universe of T-cell subsets: Th1, Th2 and more.
Immunol. Today
17
:
138
2
Gajewski, T. F., F. W. Fitch.
1988
. Anti-proliferative effect of IFN-gamma in immune regulation. I. IFN-gamma inhibits the proliferation of Th2 but not Th1 murine helper T lymphocyte clones.
J. Immunol.
140
:
4245
3
Fiorentino, D. F., M. W. Bond, T. R. Mosmann.
1989
. Two types of mouse T helper cell. IV. Th2 clones secrete a factor that inhibits cytokine production by Th1 clones.
J. Exp. Med.
170
:
2081
4
Sadick, M. D., R. M. Locksley, C. Tubbs, H. V. Raff.
1986
. Murine cutaneous leishmaniasis: resistance correlates with the capacity to generate interferon-γ in response to Leishmania antigens in vitro.
J. Immunol.
136
:
655
5
Scott, P., P. Natovitz, R. L. Coffman, E. Pearce, A. Sher.
1988
. Immunoregulation of cutaneous leishmaniasis: T cell lines that transfer protective immunity or exacerbation belong to different T helper subsets and respond to distinct parasite antigens.
J. Exp. Med.
168
:
1675
6
Locksley, R. M., F. P. Heinzel, M. D. Sadick, B. J. Holaday, K. D. Gardner, Jr.
1987
. Murine cutaneous leishmaniasis: susceptibility correlates with differential expansion of helper T-cell subsets.
Ann. Inst. Pasteur Immunol.
138
:
744
7
Coffman, R. L., R. Chatelain, L. M. Leal, K. Varkila.
1991
. Leishmania major infection in mice: a model system for the study of CD4+ T-cell subset differentiation.
Res. Immunol.
142
:
36
8
Streilein, J. W..
1993
. Immune privilege as the result of local tissue barriers and immunosuppressive microenvironments.
Curr. Opin. Immunol.
5
:
428
9
Meyers, R. L., T. H. Pettit.
1975
. Pathogenesis of experimental allergic uveitis induced by retinal rod outer segments and pigment epithelium.
J. Immunol.
114
:
1269
10
Atherton, S. S., G. A. Pesicka, J. W. Streilein.
1987
. Retinitis and deviant immune responses following intravitreal inoculation of HSV-1.
Invest. Ophthalmol. Vis. Sci.
28
:
859
11
Niederkorn, J. Y., P. C. Meunier.
1985
. Spontaneous immune rejection of intraocular tumors in mice.
Invest. Ophthalmol. Vis. Sci.
26
:
877
12
Niederkorn, J. Y., J. W. Streilein.
1983
. Alloantigens placed into the anterior chamber of the eye induce specific suppression of delayed-type hypersensitivity but normal cytotoxic T lymphocyte and helper T lymphocyte responses.
J. Immunol.
131
:
2670
13
Niederkorn, J. Y..
1990
. Immune privilege and immune regulation in the eye.
Adv. Immunol.
48
:
191
14
Streilein, J. W., B. R. Ksander, A. W. Taylor.
1997
. Immune deviation in relation to ocular immune privilege.
J. Immunol.
158
:
3557
15
Ferguson, T. A., T. S. Griffith.
1997
. A vision of cell death: insights into immune privilege.
Immunol. Rev.
156
:
167
16
Streilein, J. W., G. A. Wilbanks, S. W. Cousins.
1992
. Immunoregulatory mechanisms of the eye.
J. Neuroimmunol.
39
:
185
17
Griffith, T. S., X. Yu, J. M. Herndon, D. R. Green, T. A. Ferguson.
1996
. CD95-induced apoptosis of lymphocytes in an immune privileged site induces immunological tolerance.
Immunity
5
:
7
18
Li, X. Y., T. J. D’Orazio, J. Y. Niederkorn.
1996
. Role of Th1 and Th2 cells in anterior chamber-associated immune deviation.
Immunology
89
:
34
19
Benson, J. L., J. Y. Niederkorn.
1991
. In situ suppression of delayed-type hypersensitivity: another mechanism for sustaining the immune privilege of the anterior chamber.
Immunology
74
:
153
20
Knisely, T. L., M. W. Luckenbach, B. J. Fischer, J. Y. Niederkorn.
1987
. Destructive and nondestructive patterns of immune rejection of syngeneic intraocular tumors.
J. Immunol.
138
:
4515
21
Niederkorn, J. Y., T. L. Knisely.
1988
. Immunological analysis of a destructive pattern of intraocular tumor resolution.
Curr. Eye Res.
7
:
515
22
Knisely, T. L., J. Y. Niederkorn.
1990
. Immunologic evaluation of spontaneous regression of an intraocular murine melanoma.
Invest. Ophthalmol. Vis. Sci.
31
:
247
23
Niederkorn, J. Y..
1991
. The immunopathology of intraocular tumour rejection.
Eye
5
:
186
24
Niederkorn, J. Y., J. L. Benson, E. Mayhew.
1990
. Efferent blockade of delayed-type hypersensitivity responses in the anterior chamber of the eye.
Reg. Immunol.
3
:
349
25
Ma, D., S. Comerford, D. Bellingham, J. Sambrook, M. J. Gething, H. Alizadeh, R. Anand, J. Mellon, J. Y. Niederkorn.
1994
. Capacity of simian virus 40 T antigen to induce self-tolerance but not immunological privilege in the anterior chamber of the eye.
Transplantation
57
:
718
26
Kielty, D., S. W. Cousins, S. S. Atherton.
1987
. HSV-1 retinitis and delayed hypersensitivity in DBA/2 and C57BL/6 mice.
Invest. Ophthalmol. Vis. Sci.
28
:
1994
27
Wilbanks, G. A., M. Mammolenti, J. W. Streilein.
1992
. Studies on the induction of anterior chamber-associated immune deviation (ACAID). III. Induction of ACAID depends upon intraocular transforming growth factor-beta.
Eur. J. Immunol.
22
:
165
28
Wilbanks, G. A., J. W. Streilein.
1992
. Fluids from immune privileged sites endow macrophages with the capacity to induce antigen-specific immune deviation via a mechanism involving transforming growth factor-beta.
Eur. J. Immunol.
22
:
1031
29
Mosmann, T. R., K. W. Moore.
1991
. The role of IL-10 in crossregulation of TH1 and TH2 responses.
Immunol. Today
12
:
A49
30
Trinchieri, G..
1993
. Interleukin-12 and its role in the generation of TH1 cells.
Immunol. Today
14
:
335
31
Kuhn, R., K. Rajewsky, W. Muller.
1991
. Generation and analysis of interleukin-4 deficient mice.
Science
254
:
707
32
Kuhn, R., J. Lohler, D. Rennick, K. Rajewsky, W. Muller.
1993
. Interleukin-10-deficient mice develop chronic enterocolitis.
Cell
75
:
263
33
Niederkorn, J., J. W. Streilein, J. A. Shadduck.
1981
. Deviant immune responses to allogeneic tumors injected intracamerally and subcutaneously in mice.
Invest. Ophthalmol. Vis. Sci.
20
:
355
34
Seder, R. A., W. E. Paul.
1994
. Acquisition of lymphokine-producing phenotype by CD4+ T cells.
Annu. Rev. Immunol.
12
:
635
35
Thompson, C. B..
1995
. Distinct roles for the costimulatory ligands B7-1 and B7-2 in T helper cell differentiation?.
Cell
81
:
979
36
Sabin, E. A., E. J. Pearce.
1995
. Early IL-4 production by non-CD4+ cells at the site of antigen deposition predicts the development of a T helper 2 cell response to Schistosoma mansoni eggs.
J. Immunol.
155
:
4844
37
Hsieh, C. S., S. E. Macatonia, C. S. Tripp, S. F. Wolf, A. O’Garra, K. M. Murphy.
1993
. Development of TH1 CD4+ T cells through IL-12 produced by Listeria-induced macrophages.
Science
260
:
547
38
Hauser, C..
1990
. Cultured epidermal Langerhans cells activate effector T cells for contact sensitivity.
J. Invest. Dermatol.
95
:
436
39
Macatonia, S. E., N. A. Hosken, M. Litton, P. Vieira, C. S. Hsieh, J. A. Culpepper, M. Wysocka, G. Trinchieri, K. M. Murphy, A. O’Garra.
1995
. Dendritic cells produce IL-12 and direct the development of Th1 cells from naive CD4+ T cells.
J. Immunol.
154
:
5071
40
Macaulay, A. E., R. H. DeKruyff, C. C. Goodnow, D. T. Umetsu.
1997
. Antigen-specific B cells preferentially induce CD4+ T cells to produce IL-4.
J. Immunol.
158
:
4171
41
Reiner, S. L., R. A. Seder.
1995
. T helper cell differentiation in immune response.
Curr. Opin. Immunol.
7
:
360
42
Go, N. F., B. E. Castle, R. Barrett, R. Kastelein, W. Dang, T. R. Mosmann, K. W. Moore, M. Howard.
1990
. Interleukin 10, a novel B cell stimulatory factor: unresponsiveness of X chromosome-linked immunodeficiency B cells.
J. Exp. Med.
172
:
1625
43
De Waal Malefyt, R., J. Abrams, B. Bennett, C. G. Figdor, J. E. de Vries.
1991
. Interleukin 10 (IL-10) inhibits cytokine synthesis by human monocytes: an autoregulatory role of IL-10 produced by monocytes.
J. Exp. Med.
174
:
1209
44
Trinchieri, G., P. Scott.
1994
. The role of interleukin 12 in the immune response, disease and therapy.
Immunol. Today
15
:
460
45
DAndrea, A., M. Aste-Amezaga, N. M. Valiante, X. Ma, M. Kubin, G. Trinchieri.
1993
. Interleukin 10 (IL-10) inhibits human lymphocyte interferon gamma-production by suppressing natural killer cell stimulatory factor/IL-12 synthesis in accessory cells.
J. Exp. Med.
178
:
1041
46
DAndrea, A., X. Ma, M. Aste-Amezaga, C. Paganin, G. Trinchieri.
1995
. Stimulatory and inhibitory effects of interleukin (IL)-4 and IL-13 on the production of cytokines by human peripheral blood mononuclear cells: priming for IL-12 and tumor necrosis factor alpha production.
J. Exp. Med.
181
:
537
47
Enk, A. H., S. I. Katz.
1992
. Identification and induction of keratinocyte-derived IL-10.
J. Immunol.
149
:
92
48
Thompson-Snipes, L., V. Dhar, M. W. Bond, T. R. Mosmann, K. W. Moore, D. M. Rennick.
1991
. Interleukin 10: a novel stimulatory factor for mast cells and their progenitors.
J. Exp. Med.
173
:
507
49
Ferguson, T. A., P. Dube, T. S. Griffith.
1994
. Regulation of contact hypersensitivity by interleukin 10.
J. Exp. Med.
179
:
1597
50
Cua, D. J., R. L. Coffman, S. A. Stohlman.
1996
. Exposure to T helper 2 cytokines in vivo before encounter with antigen selects for T helper subsets via alterations in antigen-presenting cell function.
J. Immunol.
157
:
2830
51
Granstein, R. D., R. Staszewski, T. L. Knisely, E. Zeira, R. Nazareno, M. Latina, and D. M. Albert. 1990. Aqueous humor contains transforming growth factor-β and a small (less than 3500 daltons) inhibitor of thymocyte proliferation. [Published erratum appears in 1991 J. Immunol. 146:3687.] J. Immunol. 144:3021.
52
Jiang, L. Q., J. W. Streilein.
1991
. Immune privilege extended to allogeneic tumor cells in the vitreous cavity.
Invest. Ophthalmol. Vis. Sci.
32
:
224
53
Fabry, Z., D. J. Topham, D. Fee, J. Herlein, J. A. Carlino, M. N. Hart, S. Sriram.
1995
. TGF-β2 decreases migration of lymphocytes in vitro and homing of cells into the central nervous system in vivo.
J. Immunol.
155
:
325
54
Pollanen, P., M. von Euler, K. Jahnukainen, T. Saari, M. Parvinen, S. Sainio-Pollanen, O. Soder.
1993
. Role of transforming growth factor beta in testicular immunosuppression.
J. Reprod. Immunol.
24
:
123
55
Chambaz, E. M., S. Souchelnitskiy, S. Pellerin, G. Defaye, C. Cochet, J. J. Feige.
1996
. Transforming growth factors-beta s: a multifunctional cytokine family: implication in the regulation of adrenocortical cell endocrine functions.
Horm. Res.
45
:
222
56
Ferguson, T. A., Y. Gao, J. M. Herndon, T. S. Griffith.
1997
. Apoptotic cell death and IL-10 production in the eye are essential for the induction of immune deviation.
Invest. Ophthalmol. Vis. Sci.
38
:
S492