The role of thyroid hormone autoantibodies in serum transport.

Eleven sera known to contain thyroid hormone autoantibodies were analysed by reverse-flow electrophoresis for the equilibrium distribution of thyroid hormones between these autoantibodies and the three normal binding proteins found in serum. The binding properties of the autoantibodies determined in vitro did not necessarily predict their contribution to transport in serum of T1 and T3. Some could both bind in vitro and transport in serum. Others were able to bind both hormones but transported only one. However, some autoantibodies could be specific, binding and transporting one hormone only. In some sera, the autoantibody was the dominant transport protein having drawn hormone from thyroxine-binding globulin which is normally the most important. The autoantibodies were not saturated even in euthyroid individuals, indicating that they bind hormone reversibly and are a part of an equilibrium system.

Thyrotrophin (TSH)-binding proteins in bacteria and their cross-reaction with autoantibodies against the human TSH receptor.

A screen of a range of bacteria normally found in gut flora identified eight with the ability to bind TSH specifically. These included the previously reported Yersinia enterocolitica, Gram-positive, Gram-negative, pathogenic and commensal organisms. Eleven preparations of TSH-receptor autoantibodies strongly able to displace 125I-labelled TSH from the mammalian TSH receptor differed in their ability to displace the tracer from binding to bacterial extracts. None could displace the tracer from E. coli 06-1, four displaced 125I-labelled TSH from E. coli V21/1 and five displaced the tracer from Y. enterocolitica. Of those immunoglobulin preparations which did react with the bacterial protein, their apparent potency compared with that of TSH in displacing tracer from bacterial binders was an order of magnitude greater than with the mammalian receptor. This is consistent with the autoantibodies having a relatively better fit with the bacterial antigen than with the receptor when compared with TSH. The bacterial-binding activity and mammalian receptor-binding activities in each of two samples co-chromatographed on a Remazol yellow GGL-Sepharose affinity column strongly indicated that the same immunoglobulin species reacts with both antigens. These results are consistent with the proposal that a bacterial protein is the primary immunogen for the TSH-receptor antibodies in at least some patients with Graves' disease.

Thyroxine-binding globulin deficiency re-examined.

Seven sera, previously categorised as completely deficient in thyroxine-binding globulin (TBG) by an immunoelectrophoretic technique, were re-examined with a sensitive ELISA method. Only one of the sera was confirmed completely deficient by ELISA. The remaining six contained 0.08-0.19 mg/l. The protein was immunologically identical with 'normal' TBG, could bind to thyroxine and had the correct mobility in reverse-flow electrophoresis. Complete TBG deficiency may therefore be rare.

Subfractionation of human immunoglobulin G with dye-Sepharose adsorbents.

Seventeen reactive dyes were separately covalently attached to Sepharose-4B and examined for their ability to subfractionate the immunoglobulin G found in human serum. Each dye-Sepharose-4B adsorbent appeared to bind a characteristic proportion of the protein when tested with "pure" preparations. It is suggested that this is due to stereospecific interactions between each dye and various interaction sites on the immunoglobulin G molecules. These sites of interaction are not those which define the four major subclasses of the H-chain nor those which distinguish the kappa and lambda light chains. The adsorption may be influenced by the presence of other proteins when serum samples are tested due to competition for binding to the immobilised ligands but sequential use of the adsorbents can progressively enrich a particular immunological activity in human serum. This method offers a convenient and gentle way to subfractionate immunoglobulin G especially when the antigen is not known or unavailable for conventional affinity chromatography.

Thyroid function and immunological activity during and after medical treatment of Graves' disease.

The variable clinical course of Graves' disease has been followed in 27 patients each studied for 2 years from the time of diagnosis. Thyroid hormone synthesis was blocked with large doses of antithyroid drugs for the first 12 months while euthyroidism was maintained with triiodothyronine. The latter was given alone from 12 to 18 months, and for the last 6 months the patients received no treatment. The activity of the disease was determined by repeated measurements of thyroid uptake of pertechnetate and by assay of thyrotrophin receptor antibodies (TSH binding inhibitory immunoglobulins). Retrospectively there were no features on presentation which singly or in combination indicated the clinical outcome: 16 patients remained in remission (Group 1) whilst in 11 hyperthyroidism had recurred before the end of the study (Group 2). Both measures of disease activity (thyroid uptake and antibody levels) fell during the first 12 months in patients of both groups. Recurrence of Graves' disease could be predicted in some but not all patients of Group 2 at 12 months by higher thyroid uptakes and levels of thyrotrophin receptor antibodies. There was, however, evidence of abnormal thyroid function, from which we infer continuing activity of the disease, 12 to 18 months after diagnosis in all patients of Group 1, even though these patients had normal TRH tests during the last phase of the study. The difference in the course of Graves' disease 12 to 24 months after diagnosis between those patients who remained in remission and those who did not was relative: in no patient was completely normal physiological control of thyroid function re-established. Clinical remission from hyperthyroidism at this time is a level of disease activity at which the normal physiological output of thyroid hormones is not exceeded.

Consistency and variability in the character of thyrotrophin receptor antibodies in Graves' disease.

At diagnosis there was no correlation between the uptake of pertechnetate by the thyroid and thyrotrophin receptor antibodies (TRAb) measured as TSH binding inhibitory immunoglobulins in a series of 27 patients with Graves' disease. TRAb were detectable initially in 19 patients, in 11 of these there was a significant positive correlation (P less than 0.05) between serial measurements of pertechnetate uptake and TRAb made during 2 years following diagnosis. In five patients pertechnetate uptake fell with time whilst TRAb levels were maintained or fluctuated. In the remaining three of the 19 patients both measurements were low and did not change during treatment. We conclude that TRAb in any individual patient are a mixture of immunoglobulins of variable effectiveness as thyroid stimulators. In a majority of patients the composition of this mixture remains constant during the course of the illness and the clinical state reflects the levels of TRAb in the blood. In a minority, however, the character of these antibodies may alter with time or there is a change in the responsiveness of the thyroid gland. The general lack of correlation between measurements of thyroid stimulating activity and TSH binding inhibitory immunoglobulins in groups of patients is due to differences between patients in the composition of TRAb.

Studies on the antigenicity of human thyroglobulin.

Human iodothyronine-binding auto-antisera were used to define a thyroxyl containing antigenic site on thyroglobulin. The peptide sequences flanking the thyroxyl residue are important antigenic determinants of this region and only one such thyroxyl seems to be exposed on the thyroglobulin molecule. Apparent specificity for binding of free thyroxine or triiodothyronine by these auto-antibodies does not necessarily imply the presence of either one or the other in the immunogen.

Restricted antigenicity of thyroxyls in human thyroglobulin.

Interactions between two human iodothyronine-binding autoantisera and three preparations of human thyroglobulin (Tg) were not proportional to the latter's thyroxyl residue content. Probably only one of the several thyroxyl-containing sites in Tg reacted with the immunoglobulins from both antisera. In the case of one of the antisera, which was thyroxine (T4)-specific, the thyroxyl residue was the immunodominant feature of the antigenic site. The other antiserum, which had a specificity for 3,5,3'-tri-iodothyronine (T3), recognized different determinants around the same thyroxyl residue, but this residue was not itself an important element of the binding site. Thus, despite the specificity for T3 free in solution, the presence of T4 in the complete antigenic site was tolerated, since other structures supplied the bulk of the binding energy. 'Specificity' of this antiserum for T3 in solution is therefore coincidental and need not be ascribed to the presence of T3 in the original immunogen. Some results obtained in these studies may be interpreted as supporting the possibility that a modified Tg was the immunogen for the generation of these naturally occurring human antisera.

Graves' disease: thyroid function and immunologic activity.

Patients with Graves' disease were studied for two years during and after a twelve-month course of treatment. Disease activity was determined by repeated measurements of thyroidal uptake of [99mTc]pertechnetate during tri-iodothyronine administration. These in-vivo measurements of thyroid stimulation were compared with the results of in-vitro assays of Graves, immunoglobulin (TSH binding inhibitory activity--TBIA). There was no correlation between the thyroid uptake and TBIA on diagnosis. Pertechnetate uptake and TBIA both declined during the twelve months of antithyroid therapy. TBIA was detectable in sera from 19 of the 27 patients at diagnosis; in 11 of these 19 patients there was a good correlation (p less than 0.05) throughout the course of their disease between the laboratory assay of the Graves, immunoglobulin and the thyroid uptake. Probability of recurrence can be assessed but sustained remission of Graves' disease after treatment cannot be predicted from either measurement alone or in combination.

A comparative study of the binding of Graves' immunoglobulins by the patients own and other thyroid membranes.

Thyroid membrane preparations from six patients with active Graves' disease were tested in an assay which detects the thyroid interactive immunoglobulins of Graves' disease by their inhibition of binding of [125I]-thyroid stimulating hormone (TSH). With all preparations inhibition of binding of 125I-TSH by excess TSH could be demonstrated (specific binding). The patients' own immunoglobulins were assayed against their own thyroid membranes and against each other's under exactly comparable conditions. Inhibition of binding by IgGs from the patients varied between membrane preparations: with one preparation 5/6 IgGs were inhibitory but with another none were effective. Of the six patients, their own IgG inhibited binding of 125I-TSH to their own thyroid membrane preparation in only four instances, and when interaction did occur this did not reliably predict that the membrane preparation would interact with IgGs from other patients with Graves' disease. The selection of a membrane preparation for this assay cannot be made solely on ability to specifically bind TSH but the measure of the specific interaction with a Graves' IgG of proven potency must also be considered. Moreover, because of the variability between different membrane preparations, sequential clinical studies on individual patients, of the changes in concentration of Graves' IgG, must be performed using the same selected thyroid membrane preparation. We infer from these observations that the membrane structure in the vicinity of the TSH binding site is an important determinant of the interaction of Graves' IgGs with the TSH receptor, and that the configuration of this area is variable between individuals of the same species. The distinction between 'human-specific' and 'non-species-specific' thyroid stimulating antibodies is therefore probably not valid. The observation that the patient's own IgG was not often the most potent IgG inhibitor of binding of TSH also suggests that the Graves' IgG binding site is not identical or restricted to the TSH binding site; alternative explanations are discussed.