Heat shock protein peptides reactive in patients with Behçet's disease are uveitogenic in Lewis rats.

Mycobacterial and homologous human heat shock protein T cell peptide epitopes specific for T lymphocytes in Behçet's disease were investigated for their pathogenicity in Lewis rats. The potential pathogenicity of eight peptides and two controls was assessed by administering the peptides in enriched Freund's adjuvant into the footpads of male Lewis rats. Anterior uveitis which is a major manifestation of Behçet's disease was induced with two out of the four mycobacterial and all four homologous human peptides. The most effective peptides inducing iridocyclitis in 64-75% of rats were peptides with amino acids 336-351 and 136-150, derived from the sequence of the human 60-kD heat shock protein. A few of the rats also showed evidence of focal loss of photoreceptors. These results suggest that selected peptides within heat shock protein 60 kD which function as T cell epitopes in Behçet's disease are capable of inducing uveitis in rats. This supports the view that the peptide T cell determinants may be involved in the pathogenesis of Behçet's disease.

Cross-reactive antigens in the pathogenesis of onchocerciasis.

The ocular disease associated with infection with Onchocerca volvulus is unique in that there is a wealth of epidemiological evidence to support the casual nature of the association but there is little known about the pathogenic mechanisms involved. We have identified a 44,000 M(r) component of ocular tissues that shows immunological cross-reactivity with an O. volvulus antigen. This immunological cross-reactivity between parasite and a component of host tissues may underlie the development of ocular disease in onchocerciasis. Preliminary experiments indicate that it is possible to initiate ocular disease in susceptible rats using the recombinant parasite antigen. This should allow the development of a laboratory model of ocular onchocerciasis and further our understanding of the mechanisms by which an infective organism can produce an auto-immune-like disease in the host.

Experimental autoimmune uveoretinitis in the RCS rat: the influence of photoreceptor degeneration on disease expression.

S-antigen induced experimental autoimmune uveoretinitis (EAU) was produced in the Royal College of Surgeons (RCS) strain of rat which develops a photoreceptor dystrophy within 2 weeks of birth. Animals were sensitised at 60, 90, and 105 days of age: all animals developed disease, but onset was significantly delayed in older (105 day) animals compared with those aged 60 days at sensitisation (p 0.003). Disease was characterised by the early development of complete serous retinal detachment which resolved in a few days: the prevalence of retinal detachment was increased to 80% in dystrophic animals compared with 10% in the congenic, non-dystrophic controls (p < 0.001). Anterior uveitis was seen in 17/30 control strain eyes, but in none of 30 dystrophic eyes (p < 0.001). Genetically determined photoreceptor and retinal pigment epithelium dysfunction in the RCS rat, which may involve the local accumulation of altered S-antigen, predisposes the dystrophic strain to display an acute retinal detachment in the early stages of EAU. This phenomenon illustrates how biochemical dysfunction of a target organ may influence susceptibility, form, and severity of an experimental autoimmune disease.

Antibody affinity to retinal S-antigen in patients with retinal vasculitis.

Using a modified enzyme-linked immunosorbent antibody method that included dissociation of antigen antibody complexes with sodium thiocyanate, we examined the functional affinity of antibody to retinal S-antigen in 48 patients with retinal vasculitis and 46 age-matched healthy control subjects. Antibody affinity was markedly lower in patients with retinal vasculitis than in healthy subjects. Low-affinity antibody was more prevalent in acute retinal vasculitis and in patients with normal levels of circulating immune complexes. We found distinct differences between the antiretinal antibodies found in patients with retinal vasculitis and those in control subjects. The association of low-affinity antibody with normal levels of circulating immune complexes may suggest defective regulation of antiretinal autoimmunity and have important pathogenic implications.

Circulating immune complexes may play a regulatory and pathogenic role in experimental autoimmune uveoretinitis.

We compared the time course of changes in serum levels of circulating immune complexes (CICs) and of IgG antibody after sensitization of albino Lewis and pigmented Lister strain rats with uveitogenic (retinal S-antigen) and non-uveitogenic (ovalbumin) protein antigens of comparable molecular weight. Normal levels of CICs were far lower in Lewis rats in which experimental autoimmune uveoretinitis (EAU) takes the form of a severe panuveitis, than in Lister rats, in which the disease is mild, focal, confined to the posterior segment, and of lower incidence. After sensitization with either S-antigen or ovalbumin, polyethylene-glycol-precipitable CIC (PEG-CIC) peaked and fell as IgG antibody levels rose in both rat strains. However, peak levels of PEG-CIC were lower and subsequent IgG antibody levels were higher in the Lewis strain than in the less susceptible Lister strain. In both strains of rat these linked PEG-CIC/IgG antibody responses occurred earlier after sensitization with uveitogenic (S-) antigen than with ovalbumin, whether or not individual S-antigen-sensitized Lister rats developed EAU. In contrast, complement-binding CIC rose substantially only in those rats of both strains displaying EAU in response to S-antigen and not in response to ovalbumin. We suggest that immune complex (idiotypic) regulation of IgG antibody responses may be more readily perturbed by a pathogenic autoantigen (S-antigen) than by a bland antigen (ovalbumin). We also suggest that differences between the balance of regulatory and pathogenic CIC responses to uveitogenic retinal antigen may underlie or reflect strain differences in susceptibility to and severity of EAU.

Induction of experimental autoimmune uveoretinitis in Lewis rats with purified recombinant human retinal S-antigen fusion protein.

Full-length human retinal cDNA for S antigen (S-ag) and for the alpha subunit of transducin (alpha-Td) were subcloned into a bacterial expression plasmid vector to generate recombinant fusion proteins with glutathione-S-transferase (GST). The recombinant GST-S-ag and rGST-alpha-Td fusion proteins were purified from bacterial extracts by continuous flow preparative gel electrophoresis under denaturing conditions, and were assessed for their ability to induce experimental autoimmune uveoretinitis (EAU). Immunization of Lewis rats with single doses of 10 micrograms-100 micrograms rGST-S-ag in Freund's complete adjuvant supplemented with Bordetella pertussis readily induced clinical signs of EAU. Immunization with GST alone did not induce EAU indicating that disease activity was ascribable to the S-ag residues in the fusion protein. Although the alpha-Td shares limited sequence homology with S-ag, the rGST-alpha-Td fusion protein was also not uveitogenic in Lewis rats. The clinical severity of EAU in Lewis rats sensitized with rGST-S-ag was found to be milder than that induced with native S-ag preparations purified from human retina. However, humoral antibody responses to sensitization with the recombinant S-ag fusion protein were of a higher magnitude than with native S-ag. The availability of recombinant preparations of human S-ag protein will be of value in studying its processing and presentation to T cells derived from patients with autoimmune retinal vasculitis.

Pathogenicity and immunogenicity of recombinant human retinal S-antigen fusion protein.

A full-length cDNA clone to human S-antigen (HS-ag) was isolated from lambda gt 10 human retinal library and expressed as a fusion protein with glutathione S-transferase (GST) in E. Coli. Uveitogenicity and immunogenicity of recombinant GST-HS-ag fusion protein and native HS-ag were compared in EAU-susceptible Lewis rats. Recombinant HS-ag was found less uveitogenic than native HS-ag. Animals inoculated with recombinant HS-ag developed EAU on day 17, three days later than those inoculated with native HS-ag, the incidence of the disease was reduced from 80% to 58% and the score of clinical severity reduced from 2.2 to 1.3 points respectively. In contrast, rGST-HS-ag was more immunogenic than native HS-ag as it elicited four times higher levels of antibodies which reacted specifically with both antigens.

Passive administration of antibody against retinal S-antigen induces electroretinographic supernormality.

Electroretinographic supernormality, affecting both the a- and b-waves of the electroretinogram (ERG), occurs consistently before the onset of experimental autoimmune uveoretinitis (EAU) in rabbits and rats. To investigate the possible role of antibody to S-antigen (S-ab) in this phenomenon, affinity-purified polyclonal rat S-ab was injected intravenously into normal rats and administered to isolated rat eyecup preparations by bolus perfusion. In both situations, ERG supernormality was observed. The effect in vivo peaked 90 min after injection, and ERG changes in vitro were observed within 15 sec. The ERG response in vivo and in vitro was dose dependent and was abolished in vivo by pretreatment with cyproheptadine (a serotonin antagonist). The ERG was not affected in either system by a control rat antibody (antiovalbumin) or by murine monoclonal or rabbit polyclonal antibodies to S-antigen. The induction of ERG supernormality in vivo and in vitro by homologous S-ab indicates the operation of species-specific mechanisms both involving and bypassing the blood-retinal barrier and highlights a significant role for humoral autoimmunity in the pathogenesis of S-antigen-induced EAU in the rat.

Idiotypic characterisation of monoclonal antibodies with restricted epitope specificity for retinal S-antigen.

This paper describes the idiotypic specificities of eight murine monoclonal antibodies directed to three independent epitopes on retinal S-antigen. The antigenic sites recognised by these monoclonal antibodies have previously been localised to a small region near the C-terminal of bovine S-antigen. Xenogeneic, site-related anti-idiotypes prepared against each of the monoclonal antibodies recognised common idiotypes only amongst those monoclonal antibodies which reacted with the same epitope on S-antigen. Two of the three idiotypes were detected in the sera of BALB/c mice but not in two strains of rat immunised with xenogeneic S-antigen and none could be detected in the sera of patients with anti-photoreceptor autoantibodies. Our results demonstrate that the idiotypes of murine monoclonal antibodies to retinal S-antigen exhibit restricted epitope specificity but are species-restricted and imply that the S-antigen lacks a dominant antigenic epitope.

Precipitation of experimental autoallergic uveoretinitis by cyclosporin A withdrawal: an experimental model of uveitis relapse.

This study set out to determine whether withdrawal of cyclosporin A (CyA) in Lewis rats sensitized to retinal S antigen would precipitate experimental autoallergic uveoretinitis (EAU), and whether challenge of such animals with S antigen or an unrelated stimulus would accelerate EAU onset after drug withdrawal. Rats were sensitized with 50 micrograms S antigen in Freund's complete adjuvant (FCA) and EAU onset was suppressed by 18 days of treatment with CyA at doses ranging from 3 to 10 mg/kg daily. Without challenge, seven out of 11 animals developed EAU with a median onset of 78 days. This was reduced to 68 days in rats challenged on day 32 with FCA alone, to 48 days with 10 micrograms S antigen in FCA, and to 41 days with 50 micrograms S antigen in FCA. The incidence, onset and severity of anterior uveitis and extent of photoreceptor destruction were related to both CyA dose and nature of challenge. The extent of photoreceptor destruction ran parallel with severity of anterior uveitis; and delayed-type hypersensitivity reactivity on day 43 was related to both severity of anterior uveitis (P less than 0.001) and photoreceptor damage (P less than 0.002). At the highest dose, CyA also delayed the appearance of antibody to S antigen; however, subsequent antibody levels were unrelated to EAU severity or to nature of challenge. The results indicate that CyA-induced suppression of the immunological response to S antigen can recover spontaneously after drug withdrawal, that challenge with either S antigen or FCA alone can accelerate the subsequent onset of EAU, and that these phenomena may provide a basis for investigating mechanisms underlying relapse of human uveoretinitis.

A point prevalence study of 150 patients with idiopathic retinal vasculitis: 1. Diagnostic value of ophthalmological features.

This paper describes the ophthalmological features of 150 patients with idiopathic retinal vasculitis, 67 of whom had isolated retinal vasculitis (RV) and 83 had RV associated with systemic inflammatory disease (RV + SID). The diagnosis of retinal vasculitis was made by ophthalmoscopy and fluorescein angiography, and patients with any identifiable cause (infection, ischaemia, or malignancy) were excluded from the study. Patients with isolated RV tended to have peripheral vascular sheathing, macular oedema, and diffuse capillary leakage. Those with RV accompanying Behçet's disease often had branch vein retinal occlusions and retinal infiltrates together with macular oedema and diffuse capillary leakage; the retinal infiltrates were pathognomonic for Behçet's disease. In sarcoidosis the retina typically showed features of periphlebitis associated with focal vascular leakage. Patients with uveomeningitis, multiple sclerosis, arthritis, or systemic vasculitis showed diffuse retinal capillary leakage associated with a mixture of the other features. Poor visual function was particularly associated with macular oedema and branch vein retinal occlusion, while the retina appeared to 'withstand' the impact of vascular sheathing, periphlebitis, or neovascularisation alone. Within the limitations of a point prevalence study it was concluded that different patterns of retinal vasculitis occur in different systemic inflammatory diseases, and that in isolated retinal vasculitis there is a particular association between peripheral vascular sheathing, macular oedema, and diffuse capillary leakage. In Part 2 we describe the results of examining the sera of these patients for the presence of antiretinal antibodies and circulating immune complexes.

A point prevalence study of 150 patients with idiopathic retinal vasculitis: 2. Clinical relevance of antiretinal autoimmunity and circulating immune complexes.

This study describes the occurrence of antiretinal antibodies and circulating immune complexes in the sera of a large series of patients with idiopathic retinal vasculitis whose ophthalmological and clinical features are presented in Part 1. Antiretinal antibodies were measured by indirect immunofluorescence and passive haemagglutination, and circulating immune complexes were measured by polyethylene glycol precipitation and Clq binding. The occurrence of antiretinal antibodies and that of circulating immune complexes were analysed in relation to each other, to severity of retinal disease, to the type of associated systemic inflammatory disease, and to the presence of individual features of retinal inflammation. In patients with retinal vasculitis together with systemic inflammatory disease circulating immune complexes were usually accompanied by antiretinal antibodies. However, those patients with antiretinal antibodies in the absence of circulating immune complexes tended to have more severe retinal vasculitis, a feature particularly evident in Behçet's disease (p = 0.028). In patients with isolated retinal vasculitis, severity of disease was associated with antiretinal antibody (p = 0.013), as well as with the occurrence of both antiretinal antibody and circulating immune complexes together (p = 0.010). In the series as a whole there was a tendency for individual features of retinal vasculitis to be associated with antiretinal antibodies unaccompanied by circulating immune complexes; especially in macular oedema (p = 0.028). In isolated retinal vasculitis there was also an additive effect of antiretinal antibodies and circulating immune complexes in relation to disease severity; in contrast, in patients with systemic inflammatory disease, the coexistence of antiretinal antibodies and concluded that both antiretinal autoimmunity and circulating immune complexes may act as immunopathogenetic factors in idiopathic retinal vasculitis but that, in certain patients, circulating immune complex formation seems to protect against the more severe forms of autoimmune retinal inflammatory disease.

Morphometric analysis of T lymphocyte compartmentation in experimental autoimmune uveoretinitis.

Experimental autoimmune uveoretinitis (EAU) in the Lewis rat is characterized by extensive infiltration of inflammatory cells into all compartments of the eye, only some of which become irreversibly damaged. The apparent differences in the pathogenic impact of inflammatory cells within different ocular compartments may suggest that different mechanisms underlie cellular infiltration and selective tissue destruction. In order to investigate the importance of T lymphocyte infiltration, we carried out a precise topographical and temporal analysis of T cell infiltration into five compartments of the eye using an improved method for the fixation of ocular tissue. Our study showed that T cell infiltration began in the ciliary body and was most numerous and sustained in this area during EAU. The peak of T cell infiltration into the retina was comparatively delayed and was of lesser magnitude. Analysis of T cell subsets revealed a tendency for the helper phenotype to predominant during the course of disease in all ocular compartments except the retina where both helper and cytotoxic/suppressor T cells were equally represented at the height of inflammation. We suggest that the pathogenetic impact of autoreactive lymphocytes in EAU depends on the accessibility of relevant tissue antigen and on local microenvironmental features of lymphocytic traffic within different ocular compartments.

Epitope mapping of bovine retinal S-antigen with monoclonal antibodies.

We examined the binding of seven murine monoclonal antibodies raised to S-antigen, an immunopathogenic, 404 residue photoreceptor cell autoantigen which induces experimental autoimmune uveoretinitis. S-antigen has also been identified as arrestin, a protein involved in the regulation of phototransduction. One additional monoclonal antibody (C10C10), raised to a synthetic peptide (peptide N) corresponding to residues 281 to 302 in bovine S-antigen, was also studied. In preliminary studies we examined the specificity of the antibody response to bovine S-antigen in sera from Balb/c mice. Western blots of mouse sera on the cyanogen bromide digest of bovine S-antigen demonstrated that all animals produced antibody which recognized epitopes within the C-terminal cyanogen bromide peptide designated CB46. Mice of the H-2d haplotype, including the Balb/c strain often used to produce monoclonal antibodies, showed little activity to cyanogen bromide peptides other than CB46. Also, all seven of the monoclonals raised to S-antigen are specific for epitopes in the CB46 peptide. The epitopes recognized by the monoclonal antibodies could be grouped into four distinct sites defined by peptides AE-1 (A2G5), peptide AA (PDS-1), peptide 19-OV (A9C6), and peptide 199 (BDS-1,2,3 and 4). The mono-clonal antibody, C10C10, raised to peptide N recognizes an epitope in the N peptide and binds to a larger cyanogen bromide peptide designated CB123 as well as intact S-antigen. Fine mapping of these epitopes was done with various subpeptides. None of the antibodies bound the known immunopathogenic peptide, peptide M, which resides in CB123 although the C10C10 antibody binds a peptide adjacent to peptide M.

A longitudinal study of clinical and immunological findings in 52 patients with relapsing retinal vasculitis.

Fifty-two patients with retinal vasculitis--26 with idiopathic disease and 26 with associated systemic inflammatory disease--were followed up for periods ranging from six months to 12 years. The aim of the study was to determine the relationship between relapse of uveitis, visual outcome, and the occurrence of circulating immune complexes (CIC) and antiretinal antibodies. In a total of 69 relapses, CIC were increased in one-third of patients and antiretinal antibodies in one-half. In those 34 patients who expressed antiretinal antibodies 27 (79%) of the relapses were characterised by antiretinal antibodies in the absence of raised CIC levels (p less than 0.01). These findings support our previous hypothesis that CIC may have a protective role in autoimmune retinal vasculitis and that antiretinal autoimmunity is of pathogenetic importance in relapse. In individual patients the visual outcome was not related to the number of relapses or to the CIC-autoantibody pattern, suggesting the operation of additional features which merit identification.

Analysis of antigenic determinants of retinal S-antigen with monoclonal antibodies.

Eight murine monoclonal antibodies to retinal S-antigen were found to recognize three antigenic clusters in competitive binding experiments. Variable patterns of immunohistochemical staining of retinal sections supported the competitive binding studies. Monoclonal antibodies to two of these three antigenic clusters on S-antigen reacted by Western immunoblotting with both intact S-antigen and two cyanogen bromide cleaved peptide fragments of bovine S-antigen. The binding of the monoclonal antibodies was localized to epitopes contained on two peptide fragments of 26,000 and 18,000 molecular weight. The same peptide fragments also react with polyclonal rat or rabbit antisera to bovine S-antigen. Monoclonal antibodies to the third antigenic cluster did not react with either intact or fragmented S-antigen in immunoblotting studies.

Differential effect of Bordetella pertussis on experimental posterior uveitis in the black-hooded Lister rat.

The effect of an additional adjuvant, Bordetella pertussis, on the clinical and histopathologic features of experimental autoimmune uveitis in black-hooded Lister rats was investigated. Disease was induced by a single footpad injection of purified retinal S-antigen in Freund's complete adjuvant. In those animals that did not receive B Pertussis the clinical features were those of a retinal vasculitis with disc edema, periphlebitis, and deep retinal infiltrates. In contrast, animals that received B pertussis developed lesions in the pigment epithelium and choroid. Histopathologic studies disclosed focal photoreceptor necrosis associated with mononuclear cell infiltration in both groups of animals. However, in the group that did not receive B pertussis the disease was predominantly a retinitis associated with perivascular infiltration of retinal vessels, whereas in the group that did receive B pertussis the main feature was a focal choroiditis, with superficial retinal lesions being rarely observed. Retinal photoreceptors were the target tissue in both groups of rats, but the route by which they were damaged was altered from predominantly retinal to choroidal by the addition of Bordetella pertussis as an adjuvant. This change may be ascribed to the ability of B pertussis toxin to sensitize vascular endothelium to local mast cell products, these cells being plentiful around choroidal vessels but absent in the retinal circulation.

Experimental posterior uveitis. I: A clinical, angiographic, and pathological study.

The clinical, angiographic, and histopathological features of experimental posterior uveitis in the black hooded Lister rat are described. This mild form of experimental allergic uveoretinitis (EAU) is induced by sensitisation with retinal S antigen in Freund's complete adjuvant, and the inflammation produced is confined to the posterior segment of the eye. This allows for the first time precise photographic and angiographic documentation of the evolution of clinical signs, because there is minimal clouding of the vitreous by inflammatory cells. Clinically the disease is characterised by the appearance of disc oedema and periphlebitis, followed by focal infiltrates in the deep retinal layers, with eventual atrophy of the pigment epithelium. Histologically, retinal vasculitis is associated with focal mononuclear cell infiltration and necrosis of the photoreceptor layers. This model closely resembles the clinical features of idiopathic retinal vasculitis seen in man.

Antigenicity and uveitogenicity of partially purified peptides of a retinal autoantigen, S-antigen.

S-antigen, a potent retinal autoantigen involved in human inflammatory eye disease, has been chemically digested with cyanogen bromide to generate various peptide fragments. Cleavage of bovine S-antigen at methionyl residues generates seven major polypeptide fragments of apparent molecular weight 26,000, 22,000, 19,000, 18,000, 12,500, 8000 and 3000, respectively. Immunoblotting following SDS-polyacrylamide gel electrophoresis either with monoclonal antibodies known to be directed to two separate antigenic determinants on S-antigen or with various polyclonal antisera identified two peptide fragments of 26,000 and 18,000 MW. The extreme insolubility of the larger peptide fragments in aqueous or organic buffers makes the purification of the polypeptides by biochemical procedures difficult. However partial purification of the remaining soluble peptides by gel filtration in urea containing buffers made it possible to ascertain that the 18,000 MW peptide is an important constituent that carries a uveitogenic determinant of this autoantigen.

An improved method for the purification of retinal S-antigen using selective hydrophobic adsorption chromatography.

This paper describes the use of phenyl-Sepharose CL-4B as a solid-phase hydrophobic adsorbent in the purification of S-antigen from protein extracts of bovine, porcine and human retina. Chromatographic conditions were ascertained whereby the majority of contaminating proteins were bound to the adsorbent leaving S-antigen in the liquid phase. In combination with size fractionation on Ultrogel AcA, the method conveniently yielded porcine and bovine S-antigen preparations up to 100% purity. Immunogenicity of purified S-antigens was verified by induction of experimental autoimmune uveoretinitis in albino Lewis rats. The method is preparative in scale, fast in performance and yields S-antigen in high purity and antigenic potency.