Inhibition of adjuvant-induced arthritis by interleukin-10-driven regulatory cells induced via nasal administration of a peptide analog of an arthritis-related heat-shock protein 60 T cell epitope

Abstract

Objective

To prevent and treat experimental arthritis via nasal administration of an altered peptide ligand (APL) from the major arthritogenic epitope in adjuvant-induced arthritis (AIA) and to explore the mechanisms involved.

Methods

Peptides were administered nasally before and after induction of arthritis. Splenocytes and lymph node cells draining both the site of inflammation and the site of tolerance induction were used for cell transfer and were studied for antigen-specific T cell characteristics. In addition, attempts were made to stop T cell tolerance in vitro, using anticytokine antibodies.

Results

Nasal administration of a modulatory APL of the heat-shock protein 60 (Hsp60) 180–188 T cell epitope, alanine 183, had a suppressive effect in AIA that far exceeded that of the wild-type epitope. In addition to its effectiveness in preventing AIA, alanine 183 may be effective in the treatment of ongoing AIA. The protective effect of alanine 183 can be passively transferred using activated splenocytes. Nasal administration of alanine 183 did not lead to detectable T cell proliferation or interleukin-2 (IL-2) production in mandibular lymph node cells, while transforming growth factor β (TGFβ), IL-10, and IL-4 were readily produced. Likewise, after nasally induced tolerance, followed by induction of arthritis, inguinal lymph node cells produced IL-4, TGFβ, and IL-10. After neutralizing in vitro the individual cytokines with anticytokine antibodies, only blocking of IL-10 production led to reversal of tolerance, at the site of tolerance induction and the site of inflammation.

Conclusion

Nasal administration of an APL of Hsp60 180–188 induces highly effective protection against AIA through generation of regulatory cells that produce IL-4, TGFβ, and IL-10, whereas the induced tolerance is driven mainly by production of IL-10.

Antigen-specific peripheral T cell tolerance can be induced through oral or nasal administration of a relevant antigen (1, 2). In several experimental autoimmune models, oral administration of the disease-triggering autoantigen led to considerable suppression of disease activity (3–5). In comparison with orally induced tolerance, nasally induced tolerance has been shown to be equally or, in the case of a peptide antigen, even more effective in suppressing experimental autoimmune diseases (1, 6–9).

Previously, we described nasally induced tolerance for an arthritis-related heat-shock protein 60 (Hsp60) T cell epitope in adjuvant-induced arthritis (AIA) (10), an experimental arthritis model with close histopathologic resemblance to rheumatoid arthritis. AIA can be passively transferred by a single T cell clone, obtained from AIA rats, recognizing the 180–188 amino acid sequence of the mycobacterial Hsp60 (11, 12). Nasal administration of the 15-mer mycobacterial Hsp60 176–190 peptide (which contains the core 180–188 sequence) prior to induction of AIA delayed the onset and reduced the severity of arthritis (10). The arthritis-suppressive effect of Hsp60 176–190 was not complete, however, and a minority of the rats developed arthritis despite treatment.

We questioned whether use of a peptide analog of the wild-type Hsp60 180–188 peptide could enhance the arthritis-suppressive effect. We previously defined such a modulatory altered peptide ligand of Hsp60 180–188, peptide 180–188_183L→A (alanine 183) (13). Alanine 183 can effectively inhibit the proliferative A2b response in vitro and prevents AIA upon immunization in vivo. In addition to the modulatory effect on A2b, the leucine-to-alanine substitution at position 183 resulted in a higher major histocompatibility complex (MHC)-binding affinity for the rat MHC class II RT1.Bl molecule compared with that of the original peptide (14).

We now demonstrate that nasally administered alanine 183 suppresses AIA much more efficiently than does the wild-type epitope. Importantly, nasal administration of alanine 183 was effective not only in the prevention of AIA but also in the treatment of ongoing arthritis. This protective effect can be passively transferred with activated spleen cells, which indicates induction of an active T cell regulatory mechanism rather than clonal deletion of arthritogenic T cells. Analysis of cytokine production by T cells generated after nasally induced tolerance with alanine 183 showed up-regulation of interleukin-10 (IL-10), IL-4, and transforming growth factor β (TGFβ) production. However, we provide evidence that the tolerance is driven mainly by IL-10 production, suggesting that nasal administration of alanine 183 induces induction in vivo of so-called type 1 T regulatory cells (Tr1) (15).

MATERIALS AND METHODS

Animals

Male inbred Lewis rats (RT1.Bl) were obtained from the University of Limburg (Maastricht, The Netherlands) and from Harlan (Indianapolis, IN). Rats were 6–9 weeks old at the start of each experiment.

Antigens and adjuvants.

Heat-killed Mycobacterium tuberculosis (strain H37Ra) was obtained from Difco (Detroit, MI). The purified recombinant Hsp60 of Mycobacterium bovis bacillus Calmette-Guérin (identical to M tuberculosis Hsp60) was purified as described previously (12). Freund's incomplete adjuvant (IFA; Difco) was used as adjuvant. The peptides used in this study were prepared in large quantities by standard solid-phase 9-fluorenylmethoxycarbonyl chemistry. Peptides were obtained as C-terminal amides. All peptides were analyzed and purified by reverse-phase high-performance liquid chromatography and checked by fast atom-bombardment mass spectrometry.

The following peptides were used: M tuberculosis Hsp60 180–188, 176–190, alanine 183, and ovalbumin (OVA) 323–339. Hsp60 180–188 contains the M tuberculosis Hsp60 sequence 180–188 (TFGLQLELT), which is also included in M tuberculosis Hsp60 sequence 176–190. Hsp60 180–188 is the core sequence recognized by the arthritogenic T cell clone A2b. Hsp60 176–190 is a dominant T cell epitope found after AIA and after immunization with mycobacterial Hsp60 (16). Alanine 183 is a modulatory peptide analog of Hsp60 180–188, containing an alanine residue at position 183 instead of the leucine residue present in the native sequence (13). OVA 323–339 (ISQAVHAAHAEINEAGR) contains an RT1.Bl-binding T cell epitope recognized by the OVA-specific T cell clone 1/C11.P7 (17). Peptide myelin basic protein 72–85 (QKSQRSQDENPV) was biotinylated during synthesis and used as marker peptide in the MHC-peptide binding assays (17).

Tissue culture reagents.

Iscove's modified Dulbecco's medium supplemented with 5% fetal calf serum (FCS), 2 mML-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin (all from Gibco BRL, Grand Island, NY), and 5 × 10⁻⁵M 2-mercaptoethanol was used as culture medium.

MHC-peptide binding assay

Lewis rat MHC class II molecules were purified from the Z1a T cell clone through affinity chromatography using the OX-6 monoclonal antibody, as described previously (17). The MHC binding affinity of the peptides was determined by using a competition assay on the isolated rat MHC class II RT1.B1 molecules, essentially as described previously (17). Briefly, detergent-solubilized MHC class II molecules (3 μM) were incubated with 100 nM of biotinylated marker peptide (72–85) and nonbiotinylated competitor peptides (dose range 0–128 μM) for 48 hours at room temperature in the presence of a protease inhibitor mix. The MHC-peptide complexes were analyzed using nonreducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Following Western blotting (ECL GST Western blotting detection kit; Amersham, Arlington Heights, IL), the biotinylated peptides were visualized on preflashed Hyperfilm (Amersham) through enhanced chemiluminescence (Amersham).

T cell proliferation assay.

Primed lymph node cells and splenocytes were cultured in triplicate (200 μl per well) in 96-well, round-bottomed plates (Costar, Cambridge, MA) at 2 × 10⁵ cells per well, with or without antigen. Lymph node cells and splenocytes were tested for proliferation to individual peptides at 2, 10, and 20 μg/ml. Concanavalin A (Con A; 2.5 μg/ml) was used as a positive control for T cell proliferation. Cultures were incubated for 96 hours at 37°C in a humidified atmosphere of 5% CO₂ and pulsed for the final 16 hours with ³H-thymidine (³H-TdR; Amersham), 0.4 μCi/well, specific activity 1 Ci/mmole. TdR uptake was measured using a liquid scintillation beta counter. The magnitude of the proliferative response was expressed as stimulation index (SI): the mean counts per minute of cells cultured with antigen, divided by the mean cpm of cells cultured with medium alone.

Cytokine neutralization assays.

Lymph node cells were cultured as described above, in the presence of the following anticytokine antibodies: anti-pan-TGFβ (polyclonal rabbit; R&D Systems, Minneapolis, MN), anti-IL-4 (polyclonal rabbit; BioSource, Fleurus, Belgium), anti-IL-10 (polyclonal rabbit; BioSource), or anti-interferon-γ (anti-IFNγ) (polyclonal rabbit; BioSource). All neutralizing antibodies were used at a concentration of 5 μg/ml according to the manufacturers' instructions (R&D Systems; BioSource). Proliferation was measured as described above.

Determination of IL-2 production.

IL-2 production in culture supernatants was determined by the ability to stimulate proliferation of the IL-2-dependent CTLL-16 line. Briefly, supernatants were collected after culturing lymph node cells or splenocytes for 24 hours with or without antigen. Supernatants were serially diluted in 96-well, round-bottomed plates in culture medium supplemented with 10% FCS. A total of 2 × 10⁴ CTLL-16 cells were added to each well, and the plates were incubated for 24 hours at 37°C. Subsequently, cultures were pulsed for 16 hours with ³H-TdR, and TdR incorporation was measured.

Determination of TGFβ production.

TGFβ was measured in culture supernatants of lymph node cells or splenocytes after 48 hours of culture with or without antigen, using a capture enzyme-linked immunosorbent assay (ELISA) as previously described (18), with minor modifications. Briefly, 96-well, flat-bottomed ELISA microtiter plates (Costar) were coated for 1 hour at room temperature with 100 μl monoclonal mouse anti-human TGFβ_1,2,3 antibody (Genzyme, Cambridge, MA). After coating, plates were washed 3 times with wash buffer (phosphate buffered saline [PBS] containing 0.005% Tween 20), blocked for 2 hours at room temperature with 100 μl/well PBS containing 0.05% Tween 20 and 1% bovine serum albumin (BSA; Sigma), and washed again 3 times. Culture supernatants (100 μl) were added, serially diluted in blocking buffer (PBS with 0.05% Tween and 1% BSA), and incubated overnight at 4°C. The plates were washed 3 times and incubated for 2 hours at room temperature with 100 μl of biotinylated chicken anti-human TGFβ1 (2 μg/ml; R&D Systems). The plates were washed 4 times, and 1:40,000 diluted streptavidin-horseradish peroxidase (HRP) in blocking buffer (100 μl/well) was added for 1 hour at room temperature. The plates were washed 4 times with tap water and incubated at room temperature with 100 μl/well of HRP substrate solution (Genzyme) before absorbency was read at 450 nm on an ELISA reader.

Intracellular cytokine staining

Lymph node cells were cultured for 24 or 48 hours with medium or antigen. During the last 4 hours of culture, 1 μM monensin (GolgiStop; PharMingen, San Diego, CA) was added to the cultures. Viable cells were harvested, incubated for 20 minutes on ice in blocking buffer (PBS with 10% normal rat serum and 0.02% 1M sodium azide), and subsequently stained for 20 minutes on ice with phycoerythrin (PE) or fluorescein isothyocyanate (FITC)-conjugated anti-rat CD4 (clone OX-35, mouse IgG2a,κ; Pharmingen). The cells were washed twice in staining buffer (PBS containing 2% FCS and 0.02% 1M sodium azide) and resuspended in 100 μl of fixation buffer (Cytofix/Cytoperm; PharMingen) for 20 minutes on ice. The fixed cells were washed twice in permeabilization buffer (Perm/Wash, PharMingen), resuspended in 100 μl of permeabilization buffer, and stained with the following conjugated monoclonal antibodies: PE-conjugated anti-rat IL-4 (clone OX-81, mouse IgG1κ; PharMingen), PE-conjugated anti-rat IL-10 (clone A5-4, anti-mouse IgG2b; PharMingen), and with FITC-conjugated anti-rat IFNγ (clone DB.1, mouse IgG1; Harlan).

Finally, the cells were washed twice and resuspended in staining buffer. PE- or FITC-conjugated anti-mouse IgG1 (clone MOPC-21; PharMingen) and PE-conjugated anti-mouse IgG2b (Caltag, South San Francisco, CA) were used as isotype controls. As a second specificity control, fixed and permeabilized cells were incubated for 30 minutes with nonconjugated anticytokine antibody (the same clone as the conjugated anticytokine antibody) prior to the actual staining with the conjugated anticytokine antibody. This preincubation procedure blocked the staining to background level. Stained cells were analyzed on a FACScan cytometer (Becton Dickinson, Mountain View, CA). At least 5,000 events were acquired from each sample and subsequently analyzed with Lysis II and CellQuest software (Becton Dickinson).

Peptide immunotherapy protocols

Pretreatment (vaccination) protocol. Rats were lightly anesthetized using ether, and 100 μg of peptide dissolved in PBS was administered nasally by micropipette, in a total volume of 10 μl (5 μl per nostril; peptide concentration 10 mg/ml). Peptide was administered on days −15, −11, −7, and −3 preceding induction of arthritis.

Treatment protocol. As treatment, the same dose, 100 μg of peptide dissolved in 10 μl of PBS, was administered nasally. Treatment was started when >50% of the rats demonstrated weight loss (i.e., at the time of onset of clinical arthritis). Nasal administration of peptide was repeated 3 times, at 3-day intervals.

Induction and clinical assessment of AIA.

AIA was induced by administering 0.5 mg M tuberculosis suspended in 100 μl of IFA as a single intradermal injection in the base of the tail. Rats were examined daily, in a blinded manner, for clinical signs of arthritis. The severity of arthritis was determined by assessing weight loss and by scoring each paw (0–4 scale) based on the degree of swelling, erythema, and deformation of the joints (maximum arthritis score 16) (19).

Adoptive transfer of protection

Donor rats were treated nasally with alanine 183 or OVA 323–339, 100 μg of peptide dissolved in 10 μl of PBS, 4 times at 3–4-day intervals. Splenocytes were harvested 3 days after administration of the last dose, and lymphocytes were isolated using a Ficoll gradient. In order to activate T cells, the cells (5 × 10⁶/ml) were cultured for 48 hours in medium with 2% normal rat serum and 2 μg/ml Con A (according to a method adapted from Miller et al [20]). After 48 hours, viable cells were harvested using a Ficoll gradient, and 1 × 10⁸ cells were injected intravenously into naive recipient rats. Simultaneously, recipient rats were immunized with 0.5 mg M tuberculosis in IFA to induce AIA.

Statistical analysis.

The Mann-Whitney U test was used to compare arthritis scores and weight loss between the different groups.

RESULTS

Effects of nasal administration of Hsp60 peptides 176–190 and 180–188, and the peptide analog alanine 183 in the prevention of AIA.

In a previous study we showed that nasal administration of Hsp60 peptide 176–190 delays the onset and moderately suppresses the severity of arthritis in AIA and in avridine arthritis (10). The first goal of the current study was to attempt to increase the suppressive effect in AIA using alanine 183, a modulatory peptide analog of the Hsp60 180–188 core epitope recognized by the arthritogenic T cell clone A2b. Figure 1 shows the MHC class II (RT1.B1) binding affinities of the peptides used in this study. Peptides Hsp60 176–190 and OVA 323–339 are strong RT1.B1 binders, while the Hsp60 180–188 core epitope is an intermediate binder. The alanine-to-leucine substitution in Hsp60 180–188, peptide alanine 183, resulted in increased MHC binding affinity.

Figure 1. Binding of heat-shock protein 60 (Hsp60) 180–188, alanine 183, Hsp60 176–190, and ovalbumin (OVA) 323-339 to rat major histocompatibility complex (MHC) class II (RT1.B1) molecules. Detergent-solubilized MHC class II molecules were incubated with the biotinylated marker peptide myelin basic protein 72–85 and unlabeled competitor peptides (dose range 0–128 μM). Peptide binding was analyzed by Western blotting as described in Materials and Methods.

Peptides Hsp60 176–190, Hsp60 180–188, alanine 183, and OVA 323–339 were administered nasally on days −15, −11, −7, and −3 prior to induction of AIA. On day 0, rats were immunized with M tuberculosis to induce arthritis. The results of a representative experiment are shown in Figure 2. Rats that were nasally exposed to Hsp60 180–188 showed a delayed onset of arthritis (mean time of onset in Hsp60 180–188-treated rats 14 days after immunization compared with 12 days after immunization in OVA 323–339-treated rats), and had a lower arthritis score (mean ± SEM maximum score 5.0 ± 2.0 for Hsp60 180–188 treated rats versus 8.2 ± 2.1 for control rats treated with OVA 323–339). A similar, modest arthritis-suppressive effect was demonstrated with Hsp60 176–190 (mean time of onset day 14; mean ± SEM maximum arthritis score 5.2 ± 1.8). The differences in mean maximum arthritis scores between the 3 groups (Hsp60 180–188, Hsp60 176–190, and OVA 323–339) did not reach statistical significance in this experiment.

Figure 2. Modulation of adjuvant-induced arthritis (AIA) development, as assessed by arthritis score and body weight change, after nasal administration of mycobacterial Hsp60 peptides 176–190, 180–188, alanine 183, or OVA 323–339. Rats were treated on days −15, −11, −7, and −3. On day 0, rats were immunized with 0.5 mg Mycobacterium tuberculosis in 100 μl Freund's incomplete adjuvant to induce AIA. There were 5 rats in each treatment group. Values are the mean and SEM. See Figure 1 for other definitions.

The effect of nasal administration of alanine 183, however, was much more pronounced (mean time of onset day 17; mean ± SEM maximum arthritis score 1.0 ± 0.7). The difference in the mean maximum arthritis score compared with that of each of the other groups was significant (P < 0.05). Weight curves (a sensitive measure of physical well-being) are shown in Figure 2. The mean ± SEM weight gain on day 21 (the day of the maximum arthritis score) compared with that on day 10 (before the onset of arthritis; see above) was 18 ± 9 grams for rats treated with alanine 183, compared with mean ± SEM weight losses of 13 ± 15 grams in rats treated with Hsp60 180–188, 18 ± 19 grams in rats treated with Hsp60 176–190, and 48 ± 11 grams in rats treated with OVA 323–339 (P < 0.05). This experiment was repeated 3 times with similar results.

Effects of nasally administered alanine 183 in the treatment of AIA.

Most successful immune interventions in AIA have used preventive designs. Successful treatment after the onset of AIA is much more difficult. We investigated whether nasally administered alanine 183 could also be used to treat ongoing arthritis. Rats were immunized with M tuberculosis to induce AIA and were weighed daily, beginning on day 8. Treatment was started on day 12 after arthritis induction (i.e., at the time of onset of clinical arthritis), when >50% of the rats had lost weight. Treatment was also administered on days 15, 18, and 22. The results of a representative experiment (1 of 3) are shown in Figure 3.

Figure 3. Effects of treatment of ongoing adjuvant-induced arthritis (AIA) with nasally administered alanine 183 or ovalbumin (OVA) 323–339, as assessed by arthritis score and body weight change. On day 0, rats were immunized with 0.5 mg Mycobacterium tuberculosis in Freund's incomplete adjuvant to induce AIA. Starting on day 8, rats were weighed daily. Treatment was started when >50% of the rats had lost weight (i.e., onset of clinical arthritis). At that time, rats were randomized into 3 treatment groups: no treatment, nasal OVA 323–339, or nasal alanine 183 (n = 5 in each treatment group). Treatment was repeated 3 times at 3-day intervals. Values are the mean and SEM. PBS = phosphate buffered saline.

Treatment with alanine 183 had a significant (P < 0.05) arthritis-suppressive effect (mean ± SEM maximum arthritis score 2.0 ± 1.5, compared with 7.3 ± 2.2 in controls treated with PBS) (Figure 3). Treatment with OVA 323–339 did not influence the course of disease (mean ± SEM maximum arthritis score 6.6 ± 1.6). There was no difference in the incidence of arthritis between treatment groups (4 of 5 alanine 183-treated rats; 4 of 5 PBS-treated rats; 5 of 5 OVA-treated rats). Figure 3 shows the weight curves of rats in the same experiment. After treatment was started, rats receiving alanine 183 regained the weight they previously lost, whereas control rats receiving OVA 323–339 or PBS kept losing weight until day 27. Again, the differences in weight change between the alanine 183 group and the 2 control groups reached statistical significance (P < 0.05). These results show that nasal administration of alanine 183 can interfere in an ongoing arthritis process.

Passive transfer of protection induced by nasally administered alanine 183.

To determine whether the protective effect of nasally administered alanine 183 was mediated via active immune regulation, we investigated whether the protection was passively transferable. In this experiment, donor rats were treated nasally with either alanine 183 or OVA 323–339, as described above. Three days after the final dose was administered, splenocytes were harvested and activated in vitro with Con A. After 48 hours, viable cells were injected intravenously into naive recipient rats. Simultaneously, the recipient rats were immunized with M tuberculosis in IFA to induce AIA. The results of 1 of 3 representative experiments are shown in Figure 4.

Figure 4. Adoptive transfer of protection induced by nasal administration of alanine 183, as assessed by arthritis score and body weight change. Donor rats were given 4 nasal treatments with alanine 183 or ovalbumin (OVA) 323–339 (n = 5 in each group). Values are the mean and SEM. AIA = adjuvant-induced arthritis.

After passive transfer, rats that received activated splenocytes from alanine 183-treated donor rats had a significantly lower maximum arthritis score compared with recipient rats receiving activated splenocytes from OVA 323–-339-treated donor rats (mean ± SEM 1.1 ± 0.7 versus 8.2 ± 2.1; P < 0.05). Likewise, the alanine 183-treated recipient rats had a more favorable weight curve than did the OVA 323–339-treated recipient rats (Figure 4). These findings indicate that the protection found after nasal administration of alanine 183 is cell-mediated, ruling out clonal deletion of the arthritogenic A2b-like cells as the main mechanism of protection.

Lymphocyte proliferation and IL-2 production after nasal administration of peptides.

To analyze the T cell response induced by nasal administration of peptide, we first measured antigen-specific T cell proliferation and IL-2 production. Rats were treated nasally with either alanine 183 or OVA 323–339. Three days after the final dose was administered, splenocytes and mandibular lymph node cells were harvested and restimulated with peptide. IL-2 production in supernatants of cells was measured after 24 hours of culture, and proliferation was measured after 96 hours of culture. The results of a representative experiment (1 of 4) are shown in Table 1. After nasal administration of alanine 183, no antigen-specific proliferation (SI <2) or IL-2 production (SI <2) could be detected in mandibular lymph node cells and splenocytes.

Table 1. Lymphocyte proliferation and interleukin-2 production after nasal administration of peptides*

Cell source	In vivo treatment and in vitro stimulation	Proliferation, SI	Interleukin-2, SI
* Rats were treated on days −15, −11, −7, and −3 prior to the experiment. Cells from mandibular lymph nodes (MLNs; 3 rats) and splenocytes (3 rats) were cultured with or without antigen. OVA = ovalbumin; stimulation index (SI) = the mean cpm of cells cultured with antigen, divided by the mean cpm of cells cultured with medium alone.
Spleen	Alanine 183	1.2	1.1
MLN	Alanine 183	1.4	<1
Spleen	OVA 323–339	1.1	1.8
MLN	OVA 323–339	2.8	1.3

Antigen-induced cytokine production after nasally induced tolerance.

Next, we analyzed the antigen-specific production of TGFβ, IL-4, IL-10, and INFγ induced by nasal immunization with peptide. Rats were treated nasally with either alanine 183 or OVA 323–339, as described above. Three days after the final dose was administered, mandibular lymph node cells were harvested and cultured with alanine 183 or OVA 323–339. In a second set of experiments, nasal induction of tolerance was followed by induction of arthritis using M tuberculosis. In these experiments, both mandibular and inguinal lymph node cells were harvested 10 days after induction of arthritis and cultured with Hsp60 176–190, alanine 183, OVA 323–339, or medium alone.

TGFβ production. After 48 hours of culture, supernatants were collected, and TGFβ production was determined using a capture ELISA. The results of a representative experiment (1 of 4) are shown in Table 2. After nasal administration of both alanine 183 and OVA 323–339, mandibular lymph node cells produced TGFβ in an antigen-specific manner. However, if nasally induced tolerance was followed by induction of arthritis, inguinal lymph node cells from rats treated with OVA 323–339 produced no detectable levels of TGFβ. In contrast, inguinal lymph node cells from rats treated with alanine 183 did produce TGFβ after in vitro stimulation with both alanine 183 and the wild-type epitope hsp60 180–188. In all studies, Con A was used as a positive control for cytokine production. No significant differences in cytokine production between the 2 treatment groups were found following Con A activation.

Table 2. Production of transforming growth factor β (TGFβ) after nasal administration of peptides*

Cell source	In vivo treatment	Induction of AIA	In vitro stimulation	Production of TGFβ, pg/ml
* Rats were treated on days −15, −11, −7, and −3 prior to the experiment. On day 0, mandibular lymph node cells (MLNs; 3 rats) were isolated and cultured with or without antigen. In a second set of expeirments, nasally induced tolerance was followed by induction of adjuvant-induced arthritis (AIA). Ten days after induction of AIA, inguinal lymph node cells (ILNs; 3 rats) were isolated and cultured with or without antigen. TGFβ = transforming growth factor β; OVA = ovalbumin. † Below the assay's limit of detection.
MLN	Alanine 183	No	Alanine 183	276
MLN	OVA 323–339	No	OVA 323–339	3,732
ILN	Alanine 183	Yes	Alanine 183	72
ILN	OVA 323–339	Yes	OVA 323–339	Below†
ILN	Alanine 183	Yes	180–188	103
ILN	OVA	Yes	180–188	Below†

Intracellular detection of IL-4, IL-10, and INFγ. To measure intracellular cytokine production of IL-4, IL-10, and INFγ, mandibular lymph node cells were isolated after nasal induction of tolerance and cultured with either alanine 183 or medium alone. After 48 hours, viable cells were harvested and stained for surface markers and intracellular cytokines. The results are shown in Figure 5. Antigen-specific intracellular cytokine production was expressed as the percentage of double-positive cells (both CD4+ and cytokines) in samples cultured with antigen minus the percentage of double-positive cells in samples cultured with medium alone. The results are shown in Figure 5A. After nasal induction of tolerance with alanine 183 and in vitro activation with alanine 183, a significant number of CD4+ mandibular lymph node cells produced IL-10 (mean ± SEM 9.9 ± 4.7%; P < 0.05 compared with OVA 323–339-restimulated cells) and IL-4 (mean ± SEM 2.5 ± 1.1%; P < 0.05 compared with OVA 323–339-stimulated cells) but did not produce INFγ. As an additional control we used mandibular lymph node cells from rats treated nasally with OVA 323–339. In vitro activation of those cells with alanine 183 did not lead to antigen-specific cytokine production (Figure 5A).

Figure 5. Intracellular cytokine production of interleukin-4 (IL-4), IL-10, and interferon-γ (IFN) after nasal induction of tolerance with alanine 183. A, Experiments with mandibular lymph node cells (MLNs). MLNs were isolated from alanine 183-treated rats and cultured with alanine 183 or medium alone. After 48 hours, viable cells were harvested and stained for surface markers and intracellular cytokines. As a control, MLNs were obtained from ovalbumin (OVA) 323–339-treated rats and cultured in a similar manner, with alanine 183 or medium alone. B, Experiments with inguinal lymph node cells (ILNs). Nasal induction of tolerance with either alanine 183 or OVA 323–339 was followed by induction of arthritis with Mycobacterium tuberculosis. Ten days after immunization, ILNs were isolated and cultured with heat-shock protein 60 176–190 or medium. After 48 hours, viable cells were harvested and stained for surface markers and intracellular cytokines. Antigen-specific intracellular cytokine production was expressed as the percentage of double-positive cells for both CD4 and cytokine in cells cultured with antigen, subtracted by the percentage of double-positive cells in cell samples cultured with medium alone. The gates for analysis were set on the isotype controls of the different antibodies.

When nasal induction of tolerance was followed by induction of arthritis, the same pattern of cytokine production was observed in inguinal lymph node cells (ILNs), not only upon in vitro restimulation with alanine 183 but also upon in vitro stimulation with Hsp60 176–190 (Figure 5B). Again, ILN cells from rats treated with OVA 323–339 before induction of arthritis were used as controls and did not show significant cytokine production following in vitro activation with either alanine 183 or Hsp60 176–190 (Figure 5B).Thus, nasally induced tolerance using alanine 183 leads to an antigen-specific increase of IL-10, IL-4, and TGFβ production in draining lymph nodes both at the site of tolerance induction and at the site of inflammation.

Effects of cytokine neutralization on T cell responsiveness after tolerance induction.

To unravel the roles of the different cytokines in the regulation of tolerance, we blocked the effect of cytokines in vitro using anticytokine antibodies. As a readout system, we used antigen-specific lymphocyte proliferation toward Hsp60 176–190. At the site of tolerance induction (mandibular lymph node cells), we tested the effect of anticytokine antibodies on the specific lymphocyte proliferation for Hsp60 176–190 after nasal induction of tolerance followed by induction of arthritis (Figure 6). Both anti-IL-10 and anti-TGFβ led to restoration of the proliferative response to Hsp60 176–190 in mandibular lymph node cells from rats treated with alanine 183.

Figure 6. Effects of anticytokine antibodies on the proliferative response to 176–190 in mandibular lymph node cells (MLNs) and inguinal lymph node cells (ILNs) after nasal induction of tolerance with alanine 183 followed by induction of arthritis with Mycobacterium tuberculosis. Anti-IL-10 = anti-interleukin-10; anti-TGF = anti-transforming growth factor. Stimulation index (SI) = the mean cpm of cells cultured with antigen, divided by the mean cpm of cells cultured with medium alone.

Similarly, we tested the effect of anticytokine antibodies on the specific lymphocyte proliferation for peptide Hsp60 176–190 at the site of inflammation, the inguinal lymph node cells (Figure 6). Again, anti-IL-10 significantly enhanced the response to Hsp60 176–190 in rats treated with alanine 183, but other anticytokines did not influence the proliferative response. In contrast, in rats treated with OVA 323–339, none of the anticytokines influenced the proliferative response to Hsp60 176–190 in ILN cells.

DISCUSSION

We previously reported that in the experimental model of AIA, nasal administration of mycobacterial Hsp60 peptide 176–190 induces antigen-specific T cell tolerance and moderate suppression of arthritis (10). In the present study, we show that the peptide analog alanine 183 has an arthritis-suppressive effect that is superior to that of the wild-type epitope, and that this analog peptide can even be used to treat ongoing arthritis.

Other studies have shown that T cell epitope analogs can selectively change the phenotype of the T cell response and thereby can interfere in T cell-mediated autoimmune diseases (13, 21, 22). Alanine 183 is a previously defined modulatory analog peptide of the Hsp60 180–188 epitope recognized by the arthritogenic T cell clone A2b, and its affinity for rat MHC class II is higher than that of the wild-type epitope (13, 14). The strong arthritis-suppressive effect of alanine 183 seen after nasal administration is not, however, the consequence of its increased MHC-binding affinity, because the longer wild-type epitope Hsp60 176–190, which has an even higher MHC-binding affinity for rat MHC class II than does alanine 183, is less effective in preventing arthritis (10, 23).

Although results of previous studies suggested otherwise (24), the finding that the tolerance-inducing capacity of a given peptide cannot be explained solely by its MHC binding affinity is in line with results of a more recent study in experimental autoimmune encephalomyelitis (EAE) (25). This indicates that apart from its increased MHC-binding affinity, alanine 183 must have additional characteristics that contribute to the strong arthritis-suppressive effect. In EAE, it has been shown that immunization with T cell epitope analogs can lead to activation of a different population of cells capable of inducing bystander suppression (26, 27). This could indicate that nasal administration of alanine 183 does not induce a phenotypical and functional change in A2b-like cells but, instead, induces a new population of alanine 183-specific cells.

In the second part of the study, we analyzed the mechanism of nasal induction of tolerance using alanine 183. First, we were able to passively transfer protection from animals treated nasally with alanine 183 to untreated naive animals via activated spleen cells. This result indicates that the protection is an active cell-mediated process and rules out the possibility of clonal deletion of the arthritogenic A2b-like cells as the main mechanism of nasal induction of tolerance. In support of this evidence, we found no indications for increased apoptosis as measured by TUNEL assay and annexin V staining (data not shown).

Finally, we showed that after nasal administration of peptide and in vitro antigen-specific restimulation, mandibular lymph node cells neither proliferated nor produced significant levels of IL-2 and IFNγ, while TGFβ, IL-4, and, at a high level, IL-10 could be detected. When nasal induction of tolerance was followed by induction of arthritis, a similar cytokine pattern was demonstrated in lymph nodes draining the site of disease induction. Interestingly, this cytokine pattern was found not only upon in vitro restimulation with alanine 183 but also upon restimulation with the wild-type epitope, Hsp60 176–190. Neutralization of the effects of these cytokines with anticytokine antibodies in vitro showed that both anti-IL-10 and anti-TGFβ could reverse the suppressed Hsp60 176–190 response in mandibular lymph node cells. However, at the site of inflammation (the inguinal lymph node cells), only anti-IL-10 could reverse this state of tolerance. These findings suggest that although TGFβ is certainly involved, IL-10 is of decisive importance for the maintenance of tolerance. This does not necessarily mean that the in vivo maintenance of tolerance is completely dependent on IL-10. In a similar model of nasal induction of tolerance in BALB/c mice, both local and systemic administration of anti-IL-10 failed to abrogate nasal tolerance (28).

Similar to oral tolerance, 3 distinct mechanisms (e.g., clonal deletion, clonal anergy, and active suppression) of nasal tolerance have been proposed (6, 10, 25, 29–32). Previous studies attributed active suppression to a switch to Th2 or Th3 cytokines (33, 34), but other studies failed to reveal such a change (25, 35) or found the switch to be independent of tolerance induction (28). The cytokine profile of T cells generated after nasal induction of tolerance with alanine 183 also showed Th2 and Th3 characteristics, namely production of IL-10, IL-4, and TGFβ. However, we demonstrated that only IL-10 plays an important role in the maintenance of tolerance at the site of inflammation.

Previously, Groux et al showed in a transgenic mouse system that chronic activation of CD4+ T cells in the presence of IL-10 resulted in the generation of Tr1 cells (15). Such in vitro-induced Tr1 cells are characterized by a low proliferative capacity and secretion of IL-10 and TGFβ. The low proliferative response and the cytokine profile, with high levels of IL-10 found after nasal induction of tolerance with alanine 183, showed a remarkable resemblance to the pattern of such Tr1 cells, suggesting the induction of Tr1-like cells in vivo. The critical importance of IL-10 in the regulation of autoimmune diseases was also shown in studies of EAE and collagen-induced arthritis (CIA), in which IL-10-deficient mice developed more severe disease (36) than did IL-4 knockout or wild-type mice, whereas IL-10 transgenic mice appeared to be resistant to CIA (37).

These findings and our present data suggest that IL-10 plays a dominant role in the regulation of peripheral tolerance and in nasally induced tolerance. However, the strong protective effects of the regulatory T cells in vivo should not be attributed solely to IL-10. Other soluble mediators, such as TGFβ and IL-4, or direct cell-cell contact may ultimately determine the clinical efficacy.

Acknowledgements

We thank Chris Broeren and Wietse Kuis for their support and advice, Ruurd van der Zee for peptide synthesis, and Ms Erica Roks and Mrs. Patricia Lacor for expert secretarial support.