M. Salvi

Restricted tissue reactivity of autoantibodies to a 64-kDa eye muscle membrane antigen in thyroid-associated ophthalmopathy☆

We studied the tissue specificity of eye muscle (EM) membrane-reactive autoantibodies detected in the serum of patients with thyroid-associated ophthalmopathy (TAO). In preliminary studies, such antibodies were shown to react, in sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting, with human thyroid (THY) and other human skeletal muscle (HSM) membrane antigens. We carried out absorption with human EM (HEM), THY, and HSM membranes of sera from patients with TAO and autoimmune thyroid disease without ophthalmopathy which reacted with one or more of 55-, 64-, and 95-kDa antigens in pig eye muscle (PEM) membrane in immunoblotting, the majority of which were also cytotoxic to HEM cells in an antibody-dependent cell-mediated cytotoxicity assay. In Western blotting, serum antibodies reactive with PEM membrane antigens of 55, 64, and 95 kDa were cross-absorbed by HEM, THY, and HSM but not by spleen or brain membranes and showed some species specificity, being absorbed by pig and human, but not bovine, EM membranes. When incubated with cultured HEM, THY, and HSM cells in vitro, autoantibodies in TAO sera immunoprecipitated a 64-kDa antigen from the first two tissues, but not from HSM, suggesting a specific binding to autoantigenic epitopes in HEM and THY. Sera from patients with TAO as well as those from patients with thyroid autoimmunity without ophthalmopathy immunoprecipitated a approximately 66-kDa protein, shown to be distinct from the 64-kDa antigen. The restricted immunological cross-reactivity of antibodies to a THY and HEM 64-kDa membrane antigen is discussed in the context of the association of ophthalmopathy with thyroid autoimmunity. Further experiments are needed to show whether autoantibodies to the 64-kDa eye muscle and thyroid shared antigen are cytotoxic, and thus likely to play a major role in the pathogenesis of the eye disease, or just markers of the orbital autoimmune process.

Eye Muscle Membrane Reactive Antibodies are not Detected in the Serum or Immunoglobulin Fraction of Patients with Thyroid-Associated Ophthalmopathy Using an Elisa and Crude Membranes

We tested sera and purified immunoglobulin (Ig) fractions from patients with autoimmune thyroid disorders (AITD), with and without ophthalmopathy, and normal subjects, for the presence of antibodies reactive with eye muscle membrane antigens in an optimized enzyme-linked immunosorbent assay (ELISA). We found no correlation between ELISA results and the presence or severity of ophthalmopathy in patients with AITD for either serum or Ig, and there were no significant differences between the mean values (+/- SE) for the three groups (AITD with ophthalmopathy, AITD without ophthalmopathy and normals) for either serum or Ig. In contrast Ig from 8 of 19 (45%) patients with thyroid-associated ophthalmopathy reacted with a 64 kDa eye muscle membrane antigen in SDS-polyacrylamide gel electrophoresis and Western blotting, while tests were positive in only one of the 8 patients with AITD without eye disease and in none of the 8 normal subjects. The presence of antibodies to a 64 kDa antigen in immunoblotting did not correlate with the levels of antibodies measured in ELISA. We conclude that the ELISA, incorporating a crude membrane fractions as antigen, is not useful as a clinical test for eye muscle autoantibodies.

Antibodies reactive with an intracellular epitope of a recombinant 64 kDa thyroid and eye muscle protein in patients with thyroid autoimmunity and ophthalmopathy.

We have developed an enzyme-linked immunosorbent assay (ELISA) for the measurement of antibodies reactive with a 98 amino acid fragment, called D1, of a recombinant thyroid and eye muscle membrane protein corresponding to a MW of 64 kDa (called 1D) in the serum of patients with thyroid autoimmunity with and without ophthalmopathy. Antibodies against the D1 fragment expressed as a fusion protein with β galactosidase, were detected in 29% of patients with thyroid-associated ophthalmopathy (TAO) of <1 yr duration, in 33% of those with disease of >3 yr duration, in 40% of patients with Graves’ hyperthyroidism (GH) without evident eye disease, in 31% of patients with lid lag and retraction but no other signs of progressive ophthalmopathy, in 25% of patients with euthyroid Graves’ disease and in 43% of patients with untreated Hashimoto’s thyroiditis (HT), but in none of 14 patients with other (non-immunological) thyroid disorders. Although tests were positive in 6 out of the 15 patients with ophthalmopathy and no overt thyroid autoimmunity overall there was no close association of the antibodies with clinical features of the eye disease or its course. In those sera in which Western blotting for antibodies reactive with a 64 kDa eye muscle membrane protein and ELISA were both carried out there was no close correlation between the two tests. In order to better understand the significance of antibodies reactive with 64 kDa eye muscle and thyroid membrane autoantigens and their possible relationship to ophthalmopathy it will be necessary to perform prospective studies of patients with untreated Graves’ hyperthyroidism beginning before the onset of eye signs, using antibody assays incorporating full length recombinant proteins.

Muscle and species reactivity of mouse monoclonal antibodies to human eye muscle membrane antigens

We studied the tissue and species reactivity of mouse monoclonal antibodies (MCAB) produced by immunizing mice with a 100,000g ultracentrifuged preparation of human eye muscle (HEM) membranes. Twenty-three MCABs, 20 of which reacted in an enzyme-linked immunosorbent assay (ELISA) with HEM membrane, 2 with human thyroid membrane, and 1 nonreactive negative control, were selected for the study. The muscle and species specificity of 6 of the most reactive and more restrictively reactive MCAB were studied in more detail. All reacted in ELISA with human skeletal muscle membrane and, to a lesser extent, with human cardiac muscle membrane, but not with human brain membrane. The 6 MCAB cross-reacted with eye muscle membrane prepared from pig but not rat, although reactivity with human tissue was greatest for all MCAB tested. When tested in immunoblotting with HEM and thyroid membranes, 3 of 6 MCAB reacted with a 64-kDa protein in HEM, 2 of which also reacted with an antigen of the same molecular weight in thyroid membrane. In a complement-mediated antibody-dependent cytotoxicity assay, 5 of 19 MCAB lysed HEM cells, 6 of 21 lysed human skeletal muscle cells, and 10 of 22 lysed human thyroid cells. These findings support results from earlier clinical studies which showed that eye muscle membrane reactive autoantibodies in the serum of patients with thyroid-associated ophthalmopathy cross-react with membrane prepared from other striated muscle. The significance of eye muscle, skeletal muscle, and thyroid cross-reactivity of MCAB is discussed in the context of autoimmune thyroid disease and ophthalmopathy.

Nature and significance of orbital autoantigens and their corresponding autoantibodies in thyroid-associated ophthalmopathy.

There is now considerable evidence that the pathogenesis of thyroid-associated ophthalmopathy is closely linked to the presence of a shared autoantigen(s) in the thyroid and the eye muscle, against which cytotoxic mechanisms are directed. Although the orbital connective tissue is certainly involved in the orbital inflammatory process, a 64 kDa membrane protein expressed by both the eye muscle and the thyroid and recognized consistently by antibodies in the sera of TAO patients, seems to be the most likely target candidate. While its presence in non ocular skeletal muscle is not as well established, more recent data tend to suggest the existence of a 64 kDa molecule in the three tissues. The availability of a cDNA encoding a 572 amino acid protein corresponding to a MW of 63-64 kDa, which may be the same molecule, will allow us to determine more clearly the structural characteristics of the different molecules proposed as targets. The role of the corresponding autoantibodies in the pathogenesis of the eye disease is far less well defined. Whether they play a role in the induction of the ophthalmopathy or only represent helpful markers remains to be clarified.

CELL-MEDIATED IMMUNITY TO ORBITAL TISSUE ANTIGENS IN THYROID-ASSOCIATED OPHTHALMOPATHY DETERMINED USING THE LEUKOCYTE PROCOAGULANT ACTIVITY ASSAY

While it is generally accepted that thyroid-associated ophthalmopathy (TAO), a progressive inflammatory disorder of the extraocular muscle and orbital connective tissue (OCT), is immunologically mediated the nature of the underlying abnormalities is poorly understood. Although there is considerable evidence for antibody-mediated immunity against both eye muscle (EM) and OCT antigens in TAO a role of cell-mediated immunity (CMI) has not been studied in detail. We have used a new sensitive test for CMI, the leukocyte procoagulant activity (LPCA) assay and tested blood leukocytes from patients with TAO and autoimmune thyroid disease (ATD) without evident ophthalmopathy for reactivity against pig eye muscle (PEM) and human thyroid and OCT membrane and soluble fractions. In some cases human EM fractions were also tested. Preparations of PEM membrane (PEMM) and human thyroid membrane induced a significant LPCA response in both groups of patients. PEM cytosol, human OCT membrane and cytosol and human spleen membrane did not evoke a significant response in either group of patients. There were significant positive correlations in patients with TAO between (i) LPCA in response to PEMM and that to human thyroid membrane and (ii) LPCA in response to human thyroid membrane and that to human EM membrane. In patients with TAO there were no significant associations between LPCA response to PEMM and the detection of serum antibodies to a 64 kDa EM membrane protein in immunoblotting, or between LPCA response to PEMM and the duration or severity of the ophthalmopathy or clinical evidence for eye muscle involvement. These findings confirm a role of cell-mediated immunity against eye muscle antigens in thyroid-associated ophthalmopathy.(ABSTRACT TRUNCATED AT 250 WORDS)

Cellular Mechanism for Autoimmune Damage in Thyroid-Associated Ophthalmopathy

Idiopathic (thyroid-associated) ophthalmopathy is an autoimmune disorder of the extraocular muscles and orbital connective tissue (1-3). It is associated with Graves 1 hyperthyroidism in 80% of cases and Hashimoto’s thyroiditis in 20%. Although a variety of antibodies against eye muscle (EM) and orbital connective tissue (OCT) antigens have been demonstrated (4-6) their role in the tissue damage of ophthalmopathy is unclear. We have recently identified antibodies which are cytotoxic to EM cells in antibody dependent cell-mediated cytotoxicity (ADCC), some of which cross-reacted with surface antigens on thyroid cells and orbital fibroblasts (7). The role of cellular immune mechanisms in the autoimmune reactions against eye muscle and OCT has not been extensively studied (reviewed in 8).

Two dimensional gel electrophoresis identifies minor differences in immunologically cross-reactive 64 KDA autoantigens in the thyroid and eye muscle

Among the candidate eye muscle autoantigens proposed as being relevant to the pathogenesis of thyroid-associated ophthalmopathy (TAO), a 64 kDa membrane autoantigen appears to be most closely associated with the eye disorder. We have examined the tissue localization and some of the physicochemical properties of this molecule in 3 human tissues, namely thyroid (THY), eye muscle (EM) and skeletal muscle (SKE), and in pig eye muscle (PEM), by two-dimensional (2-D) [isoelectric focusing (IEF)/sodium dodecyl Polyacrylamide gel electrophoresis (SDS-PAGE)] gel electrophoresis followed by Western blotting. Antibody probes used were whole sera from patients with TAO and antibodies affinity purified from TAO sera by binding to, and elution from, a sepharose-4B column conjugated with D1, a 98 amino acid peptide fragment of a recombinant 64 kDa thyroid autoantigen. Soluble membrane proteins eluted from a slice of SDSPAGE gel containing 60–70 kDa material was prepared from the four tissues and used as antigen for 2-D gel separation. The presence of a 64 kDa antigen in THY and EM recognized by sera from patients with TAO, but only rarely by those from normal individuals, was confirmed. Pretreatment of the eluted 60–70 kDa material with N-Glycosidase F to eliminate charge heterogeneity resulting from glycosylation differences, changed the pl and MW of molecules recognized by TAO sera, in THY and EM. This suggests that the 64 kDa molecule(s) in EM and THY targeted by sera from patients with TAO are glycoproteins and that they are different in the two tissues. On the other hand, the molecule recognized in SKE appeared to have a different MW, perhaps representing the 66 kDa protein identified in previous immunoprecipitation studies, while that recognized in PEM closely resembled the molecule identified in EM and THY, but with a more basic pl range.