Autoimmune uveitis induced by molecular mimicry of peptides from rotavirus, bovine casein and retinal S-antigen

2.1 Sequence homologies of mimicry peptides

Amino acid (aa) alignment of peptides PDSAg, Rota and Cas are shown in Table 1. Peptide Cas has five, Rota seven continuous aa homologies with the C-terminal part of peptide PDSAg, which represents the major T cell epitope 4.

Table 1. Amino acid alignment of peptides PDSAg, Rota and Cas^a)

• a) aa homologies of the peptides are marked by gray shades. PDSAg: retinal S-Ag (aa 341–354), Rota: rotavirus outer capsid protein vp4 (aa 591–601), Cas: bovine αs2-casein (aa 73–84).

2.2 Induction of uveitis in rats by subcutaneous immunization and adoptive T cell transfer

Immunization of 30 rats with peptide Rota resulted in uveitis (scores of single eyes ranging from 0.5 to 3) in 30% of rats (22% of eyes affected). Twenty-four percent of 17 rats immunized with Cas (12% of eyes) developed uveitis (scores from 1 to 4), as did two of four rats immunized with αs2-casein (scores from 1 to 3). No disease was elicited when rats were immunized with ovalbumin or control peptide B7PD. Sixty-six percent of animals injected intraperitoneally with T cells specific for Rota or Cas developed uveitis, as well as one of four rats injected with T cells specific for casein (Table 2, Fig. 1). All animals receiving PDSAg-specific T cells developed uveitis. Although the incidence of uveitis was lower for Rota, Cas and casein compared to PDSAg, the severity of affected eyes was similar following immunization with PDSAg or Cas. Adoptive transfer of T cell lines specific for peptide antigens caused similar disease (Fig. 1). Peptide Rota was less pathogenic by active immunization than PDSAg and Cas, and casein protein was less uveitogenic with respect to both, active immunization and adoptive transfer.

In all affected eyes infiltration of CD4⁺ T cells and destruction of the retina was observed, irrespective of the antigen used for immunization (Fig. 2) or the specificity of the adoptively transferred T cell line. No signs of arthritis, paralysis, skin alterations or diarrhea were observed in rats immunized s.c. or orally. Autopsy revealed no other than ocular pathology.

2.3 Cross-reactive rat T cell lines

Lymph node cells from rats immunized with PDSAg, Rota, Cas or casein were propagated in vitro with their respective antigen and tested for cross-reactivity in proliferation assays (Fig. 3). Primary cultures did not reveal stimulation indices (SI) higher than 2.0; specific proliferation was not observed before the second restimulation in vitro. PDSAg-specific T cell lines did not cross-react with peptides Rota or Cas (Fig. 3A, B). Rota-specific T cell lines retained their high corecognition of PDSAg without cross-reactivity to Cas (Fig. 3C, D), whereas Cas-specific T cell lines (Fig. 3E, F) remained cross-reactive to PDSAg but lost reactivity to Rota (Fig. 3C). Cas-specific T cells also proliferated in response to the whole casein protein as in vitro antigen, suggesting that the epitope represented by peptide Cas is easily processed and presented by APC also in culture. In contrast, PDSAg-specific T cell lines never proliferated to S-Ag in culture (unpublished observations). Although the casein-specific line did not show cross-reactivity to PDSAg in vitro (Fig. 3G, H), it was uveitogenic after adoptive transfer in rats, suggesting that only a minor portion of T cells, which can not be detected in our proliferation assays, is specific for the pathogenic mimicry epitope.

All T cell lines were CD4⁺TCRα/β⁺ (data not shown). The peptide-specific lines were RT1.B-restricted, as shown by inhibition of antigen presentation with antibodies specific for RT1.B (Ox 6) and RT1.D (Ox 17) (Fig. 4). Casein-specific T cells recognized epitopes presented on both, RT1.B and RT1.D (Fig. 4). Differences in MHC class II binding affinity between the peptides were determined by competition with peptide PDSAg (Fig. 5A, B) and inhibition of proliferation of a non-cross-reactive PDSAg-specific T cell line as read-out system (Fig. 5A). At a concentration of 50 μM (tenfold excess to PDSAg) peptide Cas showed more efficient competition with PDSAg for RT1.B binding than Rota (Fig. 5B), which suggests a better presentation of Cas than of peptide Rota.

2.4 Induction of oral tolerance

Only oral pretreatment of rats with peptide PDSAg resulted in significant reduction of PDSAg-induced disease, while feeding peptides Rota and Cas or casein protein had no significant ameliorating effect on uveitis compared to feeding with control peptide B7PD (Fig. 6A).

2.5 Oral induction of uveitis

Cofeeding rats with CTX and uveitogenic proteins/peptides was aimed at inducing eye disease due to oral immunization rather than oral tolerance induction. Only feeding casein protein (1 or 2 mg) with CTX resulted in uveitis in six of seven rats (9/14 eyes with scores between 0.5 and 2; Fig. 6B, C), whereas feeding 200 μg of casein or S-Ag or peptides at either concentration failed to induce disease. In the group fed with 1 mg casein, uveitis was clinically visible already 8 days after the first gavage. A second feeding did not increase incidence or intensity of clinical signs of disease. No difference in intensity of uveitis was observed after immunization with either 1 mg casein and 10 μg CTX applied twice or 2 mg casein and 15 μg CTX given once. Histology of rat eyes revealed no difference in uveitis induced by s.c. immunization with casein or oral immunization with casein and CTX (Fig. 6C).

2.6 ELISA with human sera

Sera from patients with anterior uveitis (iritis, iridocyclitis; n=50), posterior uveitis (n=48) and from healthy donors (n=36) were tested for binding to PDSAg, S-Ag, Cas, casein, Rota and tetanus toxoid as recall antigen control. Sera from iritis patients, but not sera from patients with posterior uveitis, were significantly (p<0.001) higher reactive to the tested antigens than sera from healthy donors. Only tetanus toxoid was significantly better recognized by sera from patients with posterior uveitis (iritis vs. healthy: p<0,016) than from iritis patients (posterior uveitis vs. healthy: not significant) or healthy controls (Fig. 7A).

We further determined the frequency of optical density (OD) for each antigen above the tolerance limit of control samples, which was set at mean OD 450 of healthy donors + 2 SD. This enabled to identify reactivities of individual sera to more than one antigen (Table 3). Only very few samples of the control sera (3%–8%) were positive for one or more of the tested antigens. Of 25 S-Ag-reactive sera from iritis patients, 3 sera exclusively recognized PDSAg or SAg and did not simultaneously react to either casein or Rota, whereas in the group of patients with posterior uveitis we found 7 of 15 sera recognizing S-Ag, but neither casein nor Rota. S-Ag and PDSAg were recognized by 50% of sera from iritis patients, followed by casein/Cas (46%) and Rota (38%). Thirty-eight percent of sera from iritis patients corecognized S-Ag and casein or S-Ag and Rota, and 32% all three antigens. More than 30% of sera from patients with posterior uveitis reacted with S-Ag, but only half of them showed also binding to casein.

Although most of our patients with posterior uveitis were under immunosuppressive therapy, reactivity to tetanus toxoid was higher than in healthy donors or patients suffering from iritis, which argues against the suppression of antibody responses in patients with posterior type of disease. Therefore, differences in treatment cannot be the reason for the reduced serological reactivity to retinal antigen, the mimicry peptides or casein. Compared to healthy donors, patients with anterior uveitis did not respond better to tetanus toxoid, whereas their reactivity to the tested retinal and mimicry antigens was significantly increased.

2.7 Lymphocyte proliferation of human T cells

The proliferation assay of frozen peripheral blood lymphocytes (PBL) revealed significantly (p≤0.03) higher mean SI with S-Ag, Cas, casein and Rota from iritis patients' lymphocytes compared to healthy controls. PBL from patients with posterior uveitis responded only to Rota significantly better than healthy donors (p≤0.03; Fig. 7B). Nevertheless, we found a trend to increased proliferation of cells from patients with iritis to the antigen PDSAg and with cells from patients with posterior uveitis to PDSAg, S-Ag and casein. There was no significant difference in reactivity of lymphocytes to tetanus toxoid between all groups. Determination of the frequencies of responses [exceeding the mean SI of controls + 2 SD] showed that PDSAg/S-Ag and Cas/casein was most frequently recognized by PBL from anterior uveitis patients, and Rota by lymphocytes from both, patients with anterior and posterior uveitis (Table 3). Concomitant recognition of S-Ag and casein was observed in proliferation assays from anterior uveitis patients. Healthy donors showed only rare and minor reactions to the tested autoantigen and mimicry peptides.

4.1 Antigens

Customary peptides were purchased from Biotrend (Cologne, Germany), bovine αs2-casein, ovalbumin and CTX from Sigma (Deisenhofen, Germany). Tetanus toxoid was a generous gift from Dr. Hungerer, Chiron (Marburg, Germany), and S-Ag was purified from bovine retinas as described 25.

The peptide sequences are: PDSAg: S-Ag aa 341–354, FLGELTSSEVATEV; Rota: human rotavirus MP409, outer capsid protein vp4, aa 591–601, WTEVSEVATEV (gi:4929212, Rao, C. D., Raman, S. and Jagannath, M. R.); Cas: bovine αS2-casein, aa 73–84, SEESAEVATEEV; PDSAgV4: S-Ag aa 344–354, ELTSSEVATEV; B27PD: HLA-B aa 125–138, ALNEDLSSWTAADT; B27V4: HLA-B aa 128–138, EDLSSWTAADT; B7PD: HLA-B7 aa 125–138, ALNEDLRSWTAADT.

4.2 Induction of uveitis

All animal experiments were approved by the Review Board of the Government of Oberbayern. Treatment of the animals conformed to the ARVO Statement on the Use of Animals in Ophthalmic and Vision Research. Male and female Lewis rats were obtained from RCC (Basel, Switzerland), Janvier (Le Genest St. Isle, France) or bred in our own colony, and used for experiments at the age of 6–8 weeks. Uveitis was induced by s.c. immunization into both hind legs with a total volume of 200 μl containing 100 μg peptide (except PDSAg: 15 μg) or casein protein or ovalbumin, respectively, emulsified in CFA and supplemented to a final concentration of 2.5 mg/ml with Mycobacterium tuberculosis strain H37Ra (BD, Heidelberg, Germany), or by gavage [1 mg or 200 μg peptide or protein together with 10 μg native CTX twice, every 2 weeks or as a single gavage of 2 mg antigen with 15 μg CTX (Sigma)]. Eyes were collected for histology 2–3 weeks after s.c. immunization or 1 week after the second gavage, respectively. Uveitis was graded clinically and histologically as described elsewhere 26.

Uveitis was induced by adoptive transfer of rat T cells lines that were restimulated in vitro one to three times with their respective antigen. Three days after the last antigen stimulation, 3×10⁶–7×10⁶ cells were intraperitoneally injected into naive animals. Eyes were collected when clinical signs of disease had subsided, and uveitis was graded as mentioned above.

4.3 Rat T cell lines

Single-cell suspensions of draining lymph node cells from immunized rats were cultivated in triplicates as described 4. Antigens were used at a concentration of 10 μg/ml, the cultures were pulsed with [³H]thymidine after 2–3 days. The results are presented as SI ± SE. To determine the presenting MHC class II molecule, increasing amounts of antibodies to RT1.B (Ox 6) or RT1.D (Ox 17) (Serotec/Biozol, Eching, Germany) were added to the cultures. For blocking of the binding groove with competitor peptides, APC (thymocytes) were pre-incubated with various concentrations of competitor peptide for 2 h before adding the PDSAg-specific T cell line and PDSAg (here: 2 μg/ml). Proliferation of cultures stimulated with 2 μg PDSAg only was defined as 100%.

4.4 Oral tolerance induction

Rats were fed with 200 μg of peptides (PDSAg, Cas, Rota or control peptide B7PD) or 1 mg casein protein in PBS three to four times every other day. Two days after the last gavage the animals were immunized with 15 μg peptide PDSAg to induce uveitis. Statistics were performed using t-test.

4.5 Proliferation assays with human blood lymphocytes

The collection and use of human peripheral blood was approved by the local ethical committee. The informed consent from uveitis patients and healthy donors was obtained. Lymphocytes from 32 uveitis patients and 14 healthy donors were separated by Ficoll density gradient centrifugation and cultivated for 6 days in triplicates in round-bottom microtiter plates (Renner, Dannstadt, Germany) at a density of 5×10⁵/ml as described 4. Peptide and protein antigens were used at concentrations of 10 μg/ml. After 5 days, 2 μCi [³H]thymidine/well was added and cells cultured for another 18 h before harvested. Proliferation is shown as SI, which was regarded as positive when it was equal to or exceeded the mean SI of healthy controls + 2 SD, as described elsewhere 27. Statistics were performed using Mann-Whitney test.

4.6 Peptide-specific ELISA

Immunoplates (Maxisorp, Nunc, Wiesbaden, Germany) were coated with 20 μg peptide or protein/ml 0,05 M carbonate buffer (pH 9,6) and blocked with 1% bovine serum albumin (BSA) in PBS. Humansera (from 98 uveitis patients and 36 healthy donors) were diluted 1:20 in PBS/BSA and 100 μl as duplicate wells incubated overnight. Binding was detected with biotinylated rabbit anti-human IgG and peroxidase-conjugated streptavidin (both from Dianova, Hamburg, Germany); tetramethylbenzidine was used as substrate. OD was read at 450 nm and regarded as positive response when it was higheras the mean OD of healthy controls + 2 SD 20. Statistics were performed using Mann-Whitney test.