Immunoglobulin class and subclass distribution of eye muscle and fibroblast antibodies in patients with thyroid-associated ophthalmopathy.

Abstract

The investigation of the antibody response in thyroid-associated ophthalmopathy (TAO) using different antigens and assays has given inconsistent results. We have analysed antibodies against eye muscle and control antigens in a large group of TAO patients to assess whether specific eye muscle antibodies exist in TAO. We have also evaluated the presence of IgA and IgM class antibodies and examined IgG subclass distribution. Sera were obtained from all patients (TAO, Graves' disease without ophthalmopathy and Hashimoto's thyroiditis) within one year of diagnosis. Sera were also collected from healthy controls, with no family history of autoimmune thyroid disease. Thirty-eight patients had Graves' disease with Grade III or greater TAO; 15 patients had Graves' disease without ophthalmopathy and nine had Hashimoto's thyroiditis without any eye signs. The control group consisted of 14 subjects. Antibodies against porcine eye and skeletal muscle, human eye (membrane and soluble antigen) and skeletal muscle, human thyroid microsomal and thyroglobulin antigens and dermal and orbital fibroblast antigens were assessed using ELISA. Antibody isotypes and IgG subclasses were studied for porcine and human eye muscle antibodies. Eye muscle (porcine and human) and orbital fibroblast antibodies were further analysed by immunoblotting. There were no significant differences in the ability of either IgG or IgA in sera from the different groups to bind porcine and human eye muscle antigens. There was a significant correlation (P < 0.0001) between the binding to porcine eye muscle and skeletal muscle antigens (for both IgG and IgA). There was no difference between sera from TAO patients and control subjects in their binding to eye muscle fibroblasts for both IgG and IgA antibodies. However, IgA antibody activity against dermal fibroblasts differed significantly between TAO patients and controls (P < 0.05). By immunoblotting, the frequency of IgA antibodies recognizing 21 kDa (40% of patients) and 62 kDa (52%) bands in porcine eye muscle blots and 20, 24 and 38 kDa bands in blots of human eye muscle (soluble) antigen differed significantly between patients with TAO and controls (P < 0.05 in all cases). IgG antibodies recognizing 80 and 92 kDa bands in blots of the subcellular membrane antigen prepared from orbital fibroblasts were found more frequently in patients with TAO compared with controls (P < 0.05 in both cases). We found no evidence that eye muscle membrane or fibroblast antibodies are present in a significant proportion of TAO patients, using ELISAs based on antigens prepared from several sources. We have also failed to demonstrate the presence of previously described specific, TAO-associated antibodies, including those directed against a 64 kDa protein in eye muscle and a 23 kDa protein in fibroblasts. IgA class antibodies reactive with orbital components appeared to be more strongly associated with TAO than those of the IgG class, though even this relationship is weak. These results suggest that antibodies are of secondary importance in the pathogenesis of TAO, which is most likely a T cell-mediated disorder.