Production, isolation, and characterization of rabbit anti-idiotypic antibodies directed against human antithyrotrophin receptor antibodies.

Previous studies have shown that anti-idiotypic antibodies can be developed in vivo through animal immunization with idiotype, and that these antibodies can be isolated from other anti-immunoglobulin antibodies by affinity purification. These techniques have relied on large amounts of idiotype, which were produced either by hyperimmunization or by monoclonal antibodies, to serve as the affinity adsorbent. In the present study, we produced anti-idiotypic antibodies to human anti-thyroid-stimulating hormone (TSH) receptor antibodies by first injecting rabbits with (TSH receptor purified) IgG from Graves' patients. The resulting antiserum was then adsorbed with Sepharose-coupled TSH in an attempt to specifically bind and isolate the anti-idiotype. The antibody obtained from this process was shown to bind specifically to TSH receptor-binding antibodies from Graves' patients, and this binding could be inhibited by 56% with the addition of 10(-4) M TSH but not by HCG (10(-2) M). The anti-idiotype also bound to TSH, and this binding could be specifically inhibited by receptor-purified Graves' IgG (60% inhibition at 10 micrograms/ml IgG), but not by IgG from normal subjects (no inhibition at 50 micrograms/ml IgG). In a TSH receptor binding assay, the anti-idiotype could inhibit TSH receptor binding in Graves' sera at a 1,000-fold lower concentration than could anti-kappa/lambda antiserum; the anti-idiotypic antiserum also inhibited in vitro TSH-mediated adenylate cyclase stimulation at an IgG concentration of 5 micrograms/ml, while heterologous anti-TSH antisera and normal IgG at similar concentrations had no effect. Finally, despite being generated against a single patient's TSH receptor binding antibody, the anti-idiotype was able to block TSH receptor binding in the serum of six other Graves' patients, thus suggesting that there may be conformational conservation in the antigen that is recognized by different individuals' TSH receptor-binding immunoglobulins.

Partial characterization and clinical correlation of circulating human immunoglobulins directed against thyrotrophin binding sites in guinea pig fat cell membranes. Development of a direct enzyme immunoassay.

To obviate several problems inherent in indirect thyroid-stimulating hormone (TSH) receptor antibody assays, we developed an enzyme-linked immunosorbent assay (ELISA) that measures antibodies binding to guinea pig fat cell membrane, which contain high concentrations of TSH receptors. Solubilized guinea pig fat cell membranes were adsorbed to plastic microtiter plates and served as the solid-phase antigen. Test sera and affinity-purified alkaline phosphatase-conjugated anti-human IgG were co-incubated with membranes, after which p-nitrophenyl phosphate was added. Results were read when a positive control reached a standard color change (OD405nm). Specificity of this assay was demonstrated by the inability of albumin, insulin, TSH subunits, propranolol, or dexamethasone to block binding 30. normal subjects had a mean OD value of 0.080 +/- 0.050 (SD). 23 of 25 untreated Graves' patients had OD values at least 2 SD above the normal mean (Grave's mean +/- SD; 0.46 +/- 0.33, P less than 0.001) and in each case 10(-6) M TSH inhibited the binding by at least 60%, suggesting that the immunoglobulins were directed at the TSH receptor. Seven of 25 serum samples from patients with Hashimoto's disease, seven of 23 serum samples from patients with transient hyperthyroidism (subacute thyroiditis or painless thyrotoxic thyroiditis), and two of 10 samples from patients with thyroid carcinoma had significant elevations in the titers of membrane-directed immunoglobulins. Graves' patients who were treated with ablative therapy at least 6 mo earlier and who were euthyroid when restudied continued to have abnormally elevated membrane-directed immunoglobulins in six of eight samples studied. Further studies involved the substitution of affinity-purified alkaline phosphatase anti-IgM antisera for the anti-IgG antisera routinely used. Seven of 12 serum samples from patients with Graves' disease had significant elevations in binding which in every instance was inhibited by greater than 60% by 10(-6) M TSH. In sum, the present results indicate that (a) we have developed a sensitive, specific, reproducible, convenient ELISA for the measurement both of the total amount of circulating membrane-directed antibodies and of TSH-displaceable membrane-directed immunoglobulins. (b) This ELISA detected significant elevations in TSH-displaceable guinea pig membrane binding in 23 of 25 untreated Graves' patients as well as in approximately 30% of patients with Hashimoto's thyroiditis and subacute thyroiditis. (c) Elevated membrane directed antibodies may continue to be present many months or years after restoration of the euthyroid state. (d) Circulating membrane binding IgM immunoglobulins have been detected in patients with Graves' disease. Further studies using this ELISA should prove useful in a variety of investigative and clinical studies.

Autoantibodies to insulin are present in sera of patients with autoimmune thyroid disease.

It has been clinically suspected that patients with autoimmune thyroid disease are at an increased risk of developing other autoimmune diseases later in life. To determine the presence and potential importance of a more generalized deregulation of immune response in patients with Grave's disease and Hashimoto's disease, sera from 33 patients with Graves' disease and 16 patients with Hashimoto's disease were screened for the presence of anti-insulin antibodies and anti-insulin-receptor antibodies. An enzyme-linked immunosorbent assay was used to identify the presence of IgG against human insulin. The optical density indicating the presence of IgG against insulin in sera from patients with Graves' disease averaged .172 +/- .024 (mean +/- SE; range .010-.802), compared to the mean normal value of .098 +/- .0009 (range .012-.238) in 33 control subjects. Ten of 33 patients with Graves' disease had values greater than .200, whereas control sera values were less than .200 in all but one case (P less than .005, Graves' sera vs. controls). The sera from patients with Hashimoto's disease had a mean optical density of .110 +/- .016, with 15 of 16 values between .010 and .200. These values were not significantly different from controls with an insulin-binding inhibition assay. Anti-insulin-receptor antibodies were not detected in any of 33 patients with Graves' disease, and cytoplasmic islet cell antibodies were not detected in sera from seven patients with Graves' disease who had insulin-binding antibodies. These data support the hypothesis that the immunologic response in autoimmune thyroid disease may be more heterogeneous and polyclonal than previously believed.

Reduction in hepatic triiodothyronine binding capacity induced by fasting.

Triiodothyronine (T3) receptor kinetics were determined in liver nuclei isolated from fasting and fed rats. The results indicate that although affinity equilibrium constants (Ka) did not differ in the two groups, mean (+/- SE) maximal binding capacity (MBC) was reduced significantly to .30 +/- .05 nM/mg DNA in fasting compared to .46 +/- .07 nM/mg DNA (p less than .01) in the fed state. This observed decrease in MBC during fasting apparently could not be accounted for by a differential rates of loss of either DNA or of the receptor during the period of incubation.

Uremia decreases nuclear 3,5,3'-triiodothyronine receptors in rats.

The maximal binding capacity (MBC) of hepatic T3 nuclear receptors was decreased in uremic rats (132 +/- 37 fmol/mg DNA) compared to sham-operated controls (212 +/- 44 fmol/mg DNA; P < 0.025), while the equilibrium affinity constants (Ka) remained unaltered (1.8 +/- 0.4 and 1.5 +/- 0.3 X 10(9) M-1 in the uremic and control rats, respectively, P = NS). There was also a reduction in the MBC of the kidney T3 receptors, from 73 +/- 14 fmol/mg DNA in the control animals to 32 +/- 7 fmol/mg DNA in the uremic rats (P < 0.10), while the Ka values were identical in both groups (1.9 +/- 0.5 X 10(9) M-1). In addition, there were significant reductions in serum T4 (1.5 +/- 0.7 microgram/dl) and T3 (92 +/- 10 ng/dl) in the uremic rats compared to control rats, whose T4 levels averaged 4.4 +/- 0.1 microgram/dl (P < 0.005) and whose T3 levels averaged 140 +/- 13 ng/dl (P < 0.005). Further, insulin levels averaged 83 +/- 21 microU/ml in uremic rats and 38 +/- 7 microU/ml in control rats (P < 0.025), while glucagon levels averaged 457 +/- 114 pg/ml in the uremic rats and 101 +/- 30 pg/ml in the control animals (P < 0.0125). These data suggest that 1) in addition to starvation and hepatectomy, uremia is another pathological condition associated with the modification of the number of T3 receptors, 2) the reduction in MBC observed may be generalized rather than organ specific for hepatic nuclear receptors, and 3) elevated glucagon levels are associated with reduced MBC in uremia, but it is indeterminate whether hyperglucagonemia is the etiology of the decrease.

The ret/PTC mutations are common in sporadic papillary thyroid carcinoma of children and young adults.

The ret/PTC rearrangements (PTC-1, PTC-2, and PTC-3) are characteristic of papillary thyroid cancer (PTC). In adults, PTC-1 is common and may be associated with an aggressive clinical course. The incidence and significance of ret/PTC mutations are less well understood in children. We examined spontaneous PTC from 33 patients (23 females and 10 males) with a median age of 18 yr (range, 6-21 yr) and a median follow-up of 3.5 yr (range, 0-13.4 yr). The ret/PTC mutations were identified in 15 tumors (45%), including 8 PTC-1 (8 of 15, 53%), 2 PTC-2 (2 of 15, 13%), 2 PTC-3 (2 of 15, 13%), and 3 (3 of 15, 20%) combined PTC mutations (PTC-1 and PTC-2). This distribution is significantly different (P = 0.001, by chi2 analysis) from that reported for children with radiation-induced PTC. There was no correlation between the presence or type of ret/PTC mutation and patient age, tumor size, focality, extent of disease at diagnosis, or recurrence. We conclude that ret/PTC mutations are 1) common in sporadic childhood PTC, 2) predominantly PTC-1, 3) frequently multiple, and 4) of different distribution than that reported for children with radiation-induced PTC.

Antithyrotropin antibodies in the sera of Graves' disease patients.

To determine the presence and potential importance of antiidiotypic antibodies (anti-id) in the immune regulation of Graves' disease, sera from 57 patients with Graves' disease were screened before or during antithyroid therapy by enzyme-linked immunoabsorbent assay (ELISA) for presumptive anti-id, as defined by the presence of immunoglobulins (Igs) directed against TSH. The mean optical density, indicating the presence of TSH-binding antibodies, was 0.34 +/- 0.28 (+/- SD) in the sera of Graves' disease patients and 0.19 +/- 0.12 in the sera of 24 normal subjects (P less than 0.004). Control antigens (hCG and albumin) did not bind significant amounts of serum Igs. In 8 Graves' patients whose sera bound TSH, 40-80% inhibition was obtained with the addition of TSH receptor-purified IgG (approximately 1 microgram/ml) derived from a single Graves' patient's serum; no inhibition was found with normal IgG (approximately 10 micrograms/ml). Presumptive anti-id was isolated from sera of 6 Graves' patients by affinity purification with a TSH affinity column; the resultant IgG blocked immunoglobulin binding to the TSH receptor when added to the serum of the same patient from whom it had been isolated. The presence of anti-id correlated inversely with the presence of TSH receptor antibodies (r = -0.76; P less than 0.01). These studies demonstrate that 1) significant TSH binding is present in sera from Graves' disease patients, and 2) this TSH binding is specifically inhibitable by Graves' IgG, but not by normal IgG. These data support the hypothesis that TSH-binding immunoglobulins may represent anti-id that are present in Graves' disease as part of the immunological response to TSH receptor or TSH receptor antibodies. Such anti-id could modulate the expression of disease activity in Graves' disease by altering TSH receptor antibody action or production.

Solubilized nuclear thyroid hormone receptors in circulating human mononuclear cells.

In order to assess iodothyronine receptor interactions in man, we have developed a receptor assay for T3 and T4 in solubilized nuclear extracts from circulating mononuclear cells. This assay utilizes the technique of salt solubilization to isolate nuclear receptors and employs standard saturation analysis for T3 and T4 to determine maximal binding capacity (MBC) and equilibrium dissociation constants (Kd). We have determined that 11 normal subjects had a MBC for T3 of 1.20 +/- 0.20 pmol/mg DNA (+/- SE) and a Kd of 3.4 +/- 0.2 X 10(-10) M; the T4 MBC was 8.44 +/- 1.22 pmol/mg DNA and the Kd was 2.7 +/- 0.3 X 10(-10) M. Hypothyroid patients had a mean T3 MBC of 7.32 +/- 2.28 pmol/mg DNA and a mean T4 MBC of 40.04 +/- 21.36 pmol/mg DNA (P less than 0.05 compared to normal). Obese subjects (n = 12) had a basal fed MBC that was 0.66 +/- 0.13 pmol/mg DNA for T3 (P less than 0.05 compared to normal) and was 3.58 +/- 0.56 pmol/mg DNA for T4 (P less than 0.01 compared to normal). During fasting, the average T3 MBC increased to 1.43 +/- 0.31 pmol/mg DNA and the average T4 MBC increased to 9.63 +/- 2.46 pmol/mg DNA, values that are both significantly higher than those in the fed period; the dissociation constants were unaltered in obese subjects (compared to normals) in fed and fasting states. Gel filtration with 0.5 M agarose was employed to ascertain if the physicochemical properties of the solubilized mononuclear human cell receptor were similar to those previously observed in rat and human liver and kidney receptors. The elution profile obtained was similar to that reported earlier. The major binding activity has an estimated Stokes radius of 35 A and a molecular weight ratio of approximately 50,000 daltons. These studies indicate that: 1) high affinity T3 and T4 receptors exist in human mononuclear cells and have properties similar to those for T3 and T4 described previously in rat liver; 2) T3 and T4 receptor number tends to increase in hypothyroid subjects and tend to be lower in obese patients than in normal weight control subjects; 3) fasting is associated with an increase in both T3 and T4 MBC; and 4) despite their apparent physicochemical similarity, T3 receptors in rat liver and human mononuclear cells may be regulated differently, at least during fasting since hepatic T3 receptors decrease in the fasted rat. Collectively, these observations support the concept that human white cell T3 nuclear receptor binding is capable of rapid fluctuations, suggesting a mechanism for homeostatic regulation of T3 action.

Analysis of human TSH receptor gene and RNA transcripts in patients with thyroid disorders.

Human TSH receptor (hTSH-R) gene and RNA transcripts were analyzed by Southern and Northern blots in patients with various thyroid disorders, and in tissue cell lines. A 1.4 Kb cDNA encoding the extracellular human TSH-R domain was used as a probe. Southern analysis revealed two constant bands of 11.0 and 5.0 Kb (hTSH-R) in the thyroid and human white cell samples studied, regardless of the disease process. Northern analysis showed a predominant band at about 4.4 Kb in the thyroid tissues but not in non-thyroid tissue or cell lines tested. There were no gene rearrangements or abnormal transcripts in Graves' disease or multinodular goiter samples. In contrast, the labelled cDNA TSH-R probe did not bind to RNA isolated from 1 of 2 papillary cancer samples. A portion of the unique area of the h-TSH receptor (approximately nucleotides 1100-1230) was directly sequenced in thyroid glands from patients with Graves' disease, multinodular goiter, and differentiated thyroid cancer. No mutations or polymorphisms were identified in these samples, as compared to normal thyroid or control placenta, although further definition of sequence variation in other areas of the TSH receptor, as well as in more samples, needs to be performed. The present study indicates the normal patterns of DNA and RNA hybridization in a variety of thyroid tissues and disease states, and demonstrates that pathologic thyroid samples, with the possible exception of thyroid cancer, were not associated with specific nucleotide abnormalities in the unique area of the TSH receptor that was studied.

Lymphocytes express thyrotropin receptor-specific mRNA as detected by the PCR technique.

Human lymphocytes are known to play a critical role in autoimmune diseases both by producing antibodies and by participating in lymphokine-cellular interactions. TSH, a classic pituitary hormone, may be secreted by human lymphocytes, and controversy has existed whether a specific, authentic TSH receptor also was present on the surface of these cells. The objective of our study was to identify TSH receptor transcripts after designing specific oligonucleotides that would recognize a unique putative TSH binding area of the thyroidal TSH receptor. The existence of TSH receptor transcripts was probed by employing these primers in a PCR reaction with cDNA derived from normal peripheral human lymphocytes and human thyroid tissue, as well as with cDNA from a medullary cancer cell line and rat liver. Human lymphocytes and thyroid tissue, but not medullary cancer cells or rat liver, demonstrated specific TSH receptor amplification product both by ethidium bromide staining and by Southern blot hybridization with labeled TSH receptor cDNA. The lymphocyte cDNA was partially sequenced and found to be identical to the thyroid-derived cDNA. These findings indicate that normal, nonactivated, human lymphocytes produce transcript for a TSH receptor that appears identical to that in thyroid tissue. Future studies should focus on the regulation of this transcript, as well as on the role TSH and TSH receptor may play in modulating local lymphokine activation of T and B cells, both in normal conditions and in autoimmune thyroid disease.

Hepatitis C virus E2 envelope protein induces dendritic cell maturation.

Maturation is a critical process for dendritic cells (DC) to gain or enhance their functions in antigen presentation and T-cell activation. In this study, we investigated the effect of hepatitis C virus (HCV) envelope protein E2 on DC maturation and related functions. We show that binding of E2 protein to DC leads to a change from immature to mature phenotype as detected by an increased expression of cell surface molecules including CD83, CD80, CD86, CD11c and MHC class II. The E2-matured DC showed higher capacity to stimulate T-cell proliferation and interferon-gamma production and displayed higher levels of interleukin-12 production when compared with immature DC. The induction of DC maturation by E2 is both time- and dose-dependent and can be inhibited by anti-E2 antibodies. In addition, DC matured by E2 showed decreased uptake of bovine serum albumin and latex beads, indicating their decreased activities of endocytosis and phagocytosis upon maturation. Taken together, our results demonstrated that E2 protein is able to induce dendritic cell maturation and suggested that E2 protein may play an important role in regulation of immune responses during HCV infection.

Ipodate and 8-anilino-1-naphthalene sulfonic acid block receptor binding of T3 in rat liver.

We have determined that 8-anilino-1-naphthalene sulfonic acid (ANS) and ipodate are effective inhibitor in vitro of 125I-T3 binding to rat hepatic nuclei receptors. Both of these agents are estimated to have a Kd for the T3 receptor of about 1--2 x 10(-4) M. Indirect preliminary studies suggest that ANS is a non-competitive inhibitor and ipodate is a competitive inhibitor of T3 binding. Compounds such as tyropanoate and diatrizoate and iodide had no effect on T3 receptor binding. Further in vivo studies with ipodate suggested that T3 receptor binding inhibition also occurred when ipodate was given intravenously to rats.

Microplate solid-phase radioimmunoassay for rat prolactin.

A rat prolactin solid-phase radioimmunoassay has been developed that uses 96-well microtiter plates with removable wells to which the antibody is firmly adsorbed, resulting in a solid-phase antibody. Antigen as either reference or unknown competes with radioactivity labeled antigen for binding sites on the solid-phase antibody. After immunoreaction, free antigen is removed by washing the wells with phosphosaline solution. The solid-phase antibody-antigen complex is counted for quantitation with data reduction methods currently used in routine radioimmunoassay procedures. This microplate solid-phase radioimmunoassay has several advantages over conventional methods without sacrificing specificity, sensitivity, or accuracy. This method is rapid, compact, economical, easily automated, and could be readily established in other laboratories.

Nucleotide and amino acid homology between the human thyrotropin receptor and the HIV-1 Nef protein: identification and functional analysis.

A comparison of the nucleotide sequence of the human thyrotropin receptor (hTSH-R) with that of HIV-1 revealed 61% homology between a 161 base pair region encoding a unique portion of the hTSH-R and an immunogenic HIV-1 regulatory protein, nef. Amino acid analysis of this region shows 27% homology, including a segment in which 7/10 consecutive amino acids are identical. Sera from rabbits successfully immunized with a 16 amino acid portion of the hTSH-R (352-367, p1) was assessed for reactivity against a partially homologous nef peptide (nef-1) by ELISA, with a finding five-fold higher post-immunization values compared to pre-immune sera. The specificity of this response was verified with Western blot, using recombinant nef protein. An ELISA using nef-1 gave 64% higher values with sera from Graves' disease patients than with normal controls. This homology and immunologic cross-reactivity suggests an avenue through which a shared immune response against an HIV-1 related retrovirus could play a role in the pathogenesis of Graves' disease.

Graves' IgG recognizes linear epitopes in the human thyrotropin receptor.

Twenty-nine peptides covering the full extracellular domain of the human thyrotropin receptor have been synthesized and tested for reactivity with Graves' patients' and normal sera in ELISA. Two peptides, amino acids 331-350 and the second extracellular loop of the transmembrane segment, bound IgG-s from 5 and 4 of 10 Graves' disease patients' sera, respectively. Neither of these two peptides showed enhanced binding to normal IgG. There were no apparent differences between the Graves' disease and normal group with respect to the other 27 peptides. We conclude that peptide 331-350 and the second extracellular loop carry important linear epitopes which may contribute to the disease process in selected Graves' patients.

Ras mutations are uncommon in sporadic thyroid cancer in children and young adults.

Mutations in the ras genes (H-ras, K-ras, and N-ras) occur in 10-15% of all human cancers, and commonly arise from single base substitutions at codons 12, 13, or 61. Although ras mutations have been found in adult thyroid cancers, they were absent from the two studies which examined childhood thyroid cancers. Both studies included only children with radiation induced thyroid cancer, and it remains unclear if ras mutations occur in children without radiation exposure. To answer this question, we examined archival tissue blocks from 31 children with papillary thyroid cancer (PTC) 4 with follicular thyroid cancer (FTC), 2 with medullary thyroid cancer (MTC), and 1 with lymphoma (LYM). Only 1 patient with PTC had previous radiation exposure. Genomic DNA was extracted and used for PCR amplification of the ras genes. The PCR products were analyzed by oligospecific hybridization for mutations at codons 12, 13, and 61. Two of the PTCs (6.5%) contained ras mutations. Both patients had class II disease and no history of previous radiation exposure. One patient subsequently developed bone and lung metastases. The patient with lymphoma also had a ras mutation (N-61), but ras mutations were absent from all FTC and MTC. These results suggest that ras mutations are uncommon in spontaneous childhood thyroid cancer, but occur with a frequency similar to that found in previous reports of adult differentiated thyroid cancers. The number of subjects was too small to determine if ras mutations are more common in patients with aggressive papillary thyroid cancer.

Binding of thyrotropin to selected Mycoplasma species: detection of serum antibodies against a specific Mycoplasma membrane antigen in patients with autoimmune thyroid disease.

Radiolabeled human (hTSH) and bovine (bTSH) thyroid stimulating hormone was shown to bind to five species of Mycoplasma, the wall-less prokaryotes. The maximum binding capacity of 125I-bTSH to these five species was about 7.9 x 10(-13) moles-1.4 x 10(-12) moles for 50-100 micrograms protein with dissociation constants of approximately 1.7 to 2.2 x 10(-7)M. Approximately 50% of the 125I-bTSH binding was displaced by excess, unlabeled bTSH or hTSH, but labeled bTSH was not effectively displaced by growth hormone, LH, FSH, prolactin, or the beta subunit of hTSH, FSH and LH. Antisera prepared against Mycoplasma gallisepticum and Mycoplasma pneumoniae bound to human thyroid membranes and guinea pig fat cells, suggesting that receptors on human thyroid tissues and on Mycoplasma cells may have similarities in antigenicity. These findings were substantiated by the occurrence of TSH binding to Mycoplasma antisera. Further, sera from three of six patients with Graves' disease containing antibodies to thyroid tissues also reacted to a 108 Kd polypeptide of Mycoplasma gallisepticum.

Immunogenicity of a unique region of the human thyrotropin receptor.

Clarifying the role of the TSH receptor protein in the autoimmune process may be the key to understanding the development of Graves' disease. In the present study we used a 16 amino acid peptide of the human TSH receptor (hTSHR) to immunize rabbits. A comparable, but theoretically less immunogenic, peptide was injected into other rabbits. The antibody response against these and other peptides, as well as against solubilized human thyroid membrane (TM) and guinea pig fat cell membrane (GPF) proteins, was tested using ELISA and Western blots. The GPF and TM binding pattern of rabbits' sera was compared to that of Graves' patients' sera. We have identified an area of antigenic cross-reactivity between GPF and TM; a 63 kD protein was present in both GPF and TM, and this protein uniformly bound IgG-s of the rabbits' postimmunization sera and one of eight Graves' patient's serum. We have shown that i) a theoretically immunogenic 16 amino acid peptide was indeed highly immunogenic in rabbits, ii) antibodies binding to GPF and TM were detected after immunization, and iii) the peak of thyroid stimulating immunoglobulin activity of sera was followed by a transient elevation of serum triiodothyronine levels. Further studies investigating the immunogenic epitopes of the hTSHR as well as characterizing the 63 kD protein are indicated.

The influence of triiodothyronine, thyroxine, thyrotropin, and methimazole on thyroid cell MHC class II antigen expression.

We have studied the influence of triiodothyronine (T3), thyroxine (T4), thyrotropin (TSH), and methimazole (MMI) on the expression of major histocompatibility (MHC) Class II antigen expression in human thyroid cells. T3, T4, TSH, and MMI in various combinations were added together with interferon-gamma (IFN-gamma) to human thyrocytes or to cultured FRTL-5 cells. Neither T3 nor T4, alone, caused inhibition of the IFN-gamma stimulation of thyrocyte HLA-DR expression. Moreover, the combination of both drugs at various concentrations did not inhibit this expression except only in low ranges (T3 at 0.3 nmol/liter and T4 at 12.9 nmol/liter). MMI only at a concentration of 3.0 mmol/liter caused significant inhibition of IFN-gamma-induced HLA-DR expression. However, the addition of T3 (range, 0.3-9.2 nmol/liter) or T4 (12.9-129.0 nmol/liter) prevented the MMI-induced inhibition. This phenomenon may be explained by the action of MMI on inhibiting the synthesis of T3 and T4. At a concentration of 100 microU/ml, TSH enhanced IFN-gamma-induced HLA-DR expression. It is possible that TSH induced the expression of large numbers of IFN-gamma receptors, thereby enhancing the production of HLA-DR in response to IFN-gamma. Our studies suggest that MMI does not alter thyrocyte HLA-DR expression in vitro, especially when combined with T3 or T4; however, MMI may still induce or perpetuate immune effects in vivo secondary to its influence on thyroid hormone production or thyroid antigen presentation.