Search

Although an autoimmune pathogenesis for non toxic goiter has been suggested, reports concerning circulating antibodies to TSH receptor structures have been conflicting. Intra thyroid lymphocytes, capable of secreting IgG, have been shown to be involved in the pathogenesis of Graves' and Hashimoto's diseases; therefore, the ability of conditioned media obtained from intra thyroid lymphocyte culture, and of IgG purified from these media, to stimulate cAMP accumulation and [3H]-Thymidine (TdR) uptake in FRTL-5 cells was investigated. The activity of IgG produced "in vitro" was compared with that of circulating IgG. Thyroid tissue samples were obtained at surgery from 21 patients with non toxic multinodular goiter (MNG), 5 patients with active Graves' disease (GD), and from 10 normal subjects, undergoing neck surgery for non-thyroidal pathology. IgG purified from media of GD lymphocyte cultures stimulated both cAMP accumulation and [3H]-TdR in 5 out of 5 cases: all of the IgG purified from control or MNG lymphocyte culture media was not active in either assay. Circulating IgG did not affect cAMP accumulation or [3H]-TdR in any of the non toxic MNG cases: controls showed no changed at all. However, both activities represented were increased by GD IgG. Conditioned media from intra thyroid lymphocyte cultures significantly inhibited basal cAMP accumulation in 7 out of the 21 non toxic MNG samples and totally abolished the response in all GD patients. [3H]-TdR was not affected by IgG of any of the controls, but it had an inhibitory effect on 8 out of 21 non toxic MNG patients, and significantly stimulated [3H]-TdR in all GD patients. In conclusion, present data demonstrate that intra thyroid lymphocytes from non toxic MNG do not produce antibodies capable of mimicking TSH actions through the adenylate cyclase cascade. Conversely, soluble factors interacting in TSH-mediated functions of FRTL-5 cells are present in conditioned media of intra thyroid lymphocytes of GD and MNG thyroid lymphocytes of GD and MNG thyroid cultures.

Immunoglobulin G (IgG) preparations derived from the sera of patients with hyperthyroidism due to Graves' disease (TSAb) as well as a monoclonal IgG derived from heterohybridoma fusions of Graves' lymphocytes augmented cAMP levels in a continuous strain of functioning rat thyroid cells (clone FRTL-5) in culture. Optimal stimulation was the same for both types of IgG preparations when measured after 2 h of incubation with 5 X 10(4) cells/well and using cells maintained in a nongrowth, TSH-deficient medium for 7 days. At low IgG concentrations, the stimulatory activities of both preparations exhibited a linear dependence on concentration and similar Ka values (approximately 4 X 10(-8) M) despite the fact that 65% of the Graves' serum IgG preparations had a significantly better ability to inhibit TSH binding to membrane preparations. The Ka value for TSH in the same assay was about 5 X 10(-12) M. Using this cell assay, 90% of a series of hyperthyroid Graves' IgG preparations exhibited stimulating activity, a value comparable to the frequency of positive results found by ourselves and others using human thyroid cell and slice systems. In contrast, only 10% of patients who were euthyroid 1 yr after antithyroid drug withdrawal (n = 21) exhibited stimulating activity, and no stimulating activity was detected in patients with nontoxic nodular goiter (n = 11), toxic adenoma (n = 5), or thyroid carcinoma (n = 6). The studies suggest that an optimized rat FRTL-5 thyroid cell system is a clinically useful and convenient alternative to human thyroid cell and slice systems for detecting TSAbs.

In order to verify the existence of a "short-loop" negative feedback between iodothyronines and adenylate cyclase system of human thyroid, we have studied the effect of preincubation with iodothyronines, iodotyrosines, iodothyronine analogues and iodide on TSH-induced cAMP cellular accumulation in normal human thyroid cells in primary culture. Iodide did not produce an inhibitory effect on TSH-dependent adenylate cyclase system both in normal human thyroid plasma membranes and cultured cells. Iodothyronines at a 30-40 microM concentration did not inhibit the TSH-dependent adenylate cyclase activity of human thyroid plasma membranes; however at a 1 microM concentration they were able to inhibit the TSH-dependent cAMP accumulation by cultured cells. Preincubation with iodotyrosines and iodothyronine analogues failed to inhibit the TSH-responsive cAMP accumulation in human thyroid cultured cells.

Human thyroid cells in primary culture were used for studies of thyroid cell surface antibodies in patients with thyroid autoimmune disorders. Radioiodinated IgG preparations containing thyroid microsomal antibody (TMAb), thyroid stimulating antibody (TSAb) and/or thyroglobulin antibody (TgAb) were tested for binding to thyroid cells. Binding was observed with radioiodinated IgG from patients with Graves' disease, Hashimoto's thyroiditis and idiopathic myxoedema containing TMAb, irrespective of the presence of TSAb and TgAb, while negative results were obtained with normal IgG. A dose-dependent inhibition of binding to thyroid cells was produced by the addition of the corresponding unlabelled IgG preparations. Evidence for tissue specificity was provided by the absence of binding to human skin fibroblasts used as controls. Preabsorption with human thyroid microsomes completely abolished the binding to thyroid cells of a radioiodinated TMAb positive IgG preparation, while only incomplete removal of the reactivity to thyroid microsomes was produced by preabsorption with thyroid cells. These data suggest that some but not all microsomal antigenic determinants are expressed on the thyroid cell surface. Binding to thyroid cells was also observed with purified TgAb, indicating that thyroglobulin antigenic determinants are present on the surface of thyroid cells. No evidence of binding was obtained with a TSAb positive Graves' IgG preparation with undetectable TMAb and TgAb. Unlabelled IgG preparations containing TMAb from patients with either Hashimoto's thyroiditis or idiopathic myxoedema were shown to inhibit the binding to thyroid cells of radioiodinated TMAb positive Graves' IgG and vice versa. These data indicate that antibodies present in these thyroid autoimmune disorders share common thyroid cell surface antigens. However, the binding of radioiodinated IgG from a patient with idiopathic myxoedema was only partially inhibited by Graves' or Hashimoto's IgG, suggesting that some of the thyroid cell surface antibodies of idiopathic myxoedema may not be detectable in other thyroid autoimmune disorders.

As the interactions of iodothyronines on adrenergic and vipergic receptors are not clear, the effect of exogenous T3 and T4 on catecholamine- and VIP-induced cAMP accumulation in human normal thyroid cells after eight days of primary culture has been investigated. To evaluate the effect of endogenous iodothyronines, the response of the adenylate cyclase system to isoprenaline, adrenaline, VIP, and TSH was studied during a 10 d period. T3 and T4 were unable to modify the catecholamine- and VIP-induced cAMP accumulation in human normal thyroid cells after 6-8 days of culture, while the response to TSH was significantly inhibited. In cells cultured from thyrotoxic tissue, the response of the adenylate cyclase system to catecholamines and VIP, during a 10 d primary culture, showed a behaviour similar to controls. TSH responsiveness was negligible up to the fourth day of culture, while in normal cells a response to all the agonists was present from the beginning. In view of the lack of effect of iodothyronines on catecholamine- and VIP-induced cAMP accumulation, and of the superimposable behaviour of the response to catecholamines and VIP in normal and hyperthyroid cells during the first days of culture, we can conclude that iodothyronines do not directly modify the response of the adenylate cyclase system to adrenergic and vipergic stimulation in human thyroid follicular cells. The lack of responsiveness to TSH of cells obtained from hyperthyroid tissue during the first 4 d of culture, associated with normal responsiveness to catecholamines and VIP, points to a possible involvement of biogenic amines and neuropeptides in sustaining such hyperthyroid states.

We used a strain of differentiated rat-thyroid cells in continuous culture (the FRTL-5 strain) to detect the presence of growth-promoting antibodies in serum samples from patients with autoimmune thyroid disease. We found that IgG preparations from 17 of 20 patients (85 per cent) with active Graves' disease and two of five patients (40 per cent) with Hashimoto's thyroiditis could augment thyroid-cell growth. In parallel with IgG-induced elevations in intracellular cyclic AMP levels in the same cell line, all 20 of the patients with active Graves' disease had thyroid-stimulatory antibodies. Patients' IgG preparations fell into three subclasses: those with both potent cyclic AMP stimulation and potent growth-promoting activity; those with potent cyclic AMP stimulation but low-level growth promotion; and those with potent growth promotion and low-level cyclic AMP action. Growth-promoting antibodies were not detected in patients with Graves' disease in remission (seven patients), nodular goiter (seven), subacute thyroiditis (five), or atrophic thyroiditis (one). Simultaneous assays of growth promotion and cyclic AMP stimulation may be useful in the care of patients with autoimmune thyroid disease.

Forskolin, at 10(-11) M, stimulates guanylate cyclase activity in primary human thyroid cell cultures, but does not modify cAMP accumulation. At a 10-fold higher concentration it still stimulates guanylate cyclase activity and becomes an inhibitor of cAMP production. Above 10(-9) M, forskolin stimulation of cGMP decreases, while it also becomes a stimulator of cAMP production. There is an additive effect of TSH and forskolin on cAMP production at concentrations of the diterpene which are stimulatory. Concentrations of forskolin which are inhibitory for cAMP, but stimulatory for cGMP, are inhibitory for TSH stimulation of cAMP. The addition of 8-bromo-cGMP duplicates the forskolin effect at low concentrations.

Even though adrenergic nerve terminals between and around thyroid follicles and catecholamine stimulation of thyroid adenylate cyclase have been reported, there is no uniform concept on catecholamine interaction with thyrotrophin (TSH) receptors. Therefore, the effect of catecholamines on TSH-stimulated cyclic AMP (cAMP) accumulation in human follicular thyroid cells has been investigated, to thus eliminating the extrathyroidal actions of catecholamines. Epinephrine, norepinephrine and isoproterenol appeared to be rapid and potent stimulators of intracellular cAMP accumulation, the half maximum increase doses being 4 X 10(-7)M, 1 X 10(-5)M and 5 X 10(-7)M, respectively. While propranolol (1 X 10(-5)M) prevented the stimulatory effect of catecholamines and failed to inhibit the effect of bovine TSH, phentolamine (1 X 10(-5)M) enhanced the potency of norepinephrine and bovine TSH, leaving that of epinephrine unchanged. The effects of epinephrine (2 X 10(-8)M) and isoproterenol (2 X 10(-8)M) were additive to that of bovine TSH (0.5 mU/ml), but the effect of simultaneous stimulation with norepinephrine (5 X 10(-7)M) and bovine TSH (0.5 mU/ml) was lower than expected. Prenalterol, a selective beta 1-agonist, did not stimulate cAMP accumulation, while terbutaline, a selective beta 2-agonist, exerted a potent stimulation. Metoprolol, a selective beta 1-adrenergic blocker, did not affect the response of thyroid follicular cells to isoproterenol. These results demonstrate the existence of beta-adrenergic receptors in human thyroid follicular cells, mainly of the type beta 2, apparently not correlated with TSH receptor. The existence of alpha-adrenergic receptors which counter-regulate TSH functional responses in human thyroid follicular cells is suggested.

TSH-induced cAMP accumulation in cells obtained from normal and pathological thyroid tissue was studied during the first 12 days of primary culture. In normal thyroid tissue cultures (N = 7), the response of cAMP to TSH was present from the second day of culture and reached its maximum after 8 days. A similar behaviour was observed in cultures obtained from euthyroid sporadic goitres (N = 8), even if the rate of response was slightly lower than that of normal tissue. Similarly, cultured cells from euthyroid 'autonomous' nodules (N = 8) appeared to be responsive to TSH during the period of study, but the rate of response was also lower than in the controls. On the contrary, in cultures obtained from toxic adenomas (N = 5) and from diffuse toxic goitres (N = 5) the response to TSH was absent during the first 4 days of culture. The cells became sensitive to TSH from 6 and 6 day onwards, with the rate of response increasing progressively and reaching its maximum on day 12. Finally, in cultured cells obtained from different areas of multinodular toxic goitres (N = 4), the response to TSH was similar to that of euthyroid goitres in cells prepared from 'cold' areas, and to that of toxic adenomas in cells obtained from 'hot' areas. The present data demonstrate the existence of an inhibitory action of unknown factors, possibly iodothyronines or thyroglobulin, on the TSH effect in short-term cultures obtained from thyrotoxic tissues. A normal TSH responsiveness can be restored when the culture is prolonged.

To clarify the role of insulin, IGF-I and TSH in thyroid cell regulation, their effects on amino acid transport were studied separately. These effects were noted and compared using both the Wistar rat thyroid cell line and human thyroid cell cultures. Insulin, IGF-I and TSH were able independently to induce the radiolabelled alpha-aminoisobutyric acid transport within the rat thyroid cells: TSH stimulated the amino acid transport in rat thyroid cells in a dose-dependent way from 1 pmol/l to 10 nmol/l. Similarly, insulin increased amino acid transport significantly from 0.17 nmol/l up to 0.17 mumol/l and IGF-I from 0.13 pmol/l up to 0.13 mumol/l. The combined effects of insulin and TSH on amino acid transport were equal to the theoretical sum of the activities, whereas those of IGF-I and TSH were greater than the theoretical one. When human thyroid cell cultures were used, a significant increase of labelled amino acid transport was induced by TSH, i.e. from 0.1 pmol/l to 10 pmol/l; IGF-I stimulated amino acid transport in a range from 0.13 pmol/l to 13 pmol/l, under the same conditions. Conversely, only large doses of insulin, i.e. 1.7 nmol/l, were able weakly to stimulate amino acid transport. When submaximal TSH and IGF-I doses were co-incubated in human thyroid cells, an additive effect on amino acid transport was observed.

A basic reason for undertaking these studies was to further our knowledge of the structure and function of the TSH receptor as well as its interaction with other receptors on thyroid cells. The multiplicity of observations suggests the approach is bearing fruit, does not provide a simple answer, and can have pitfalls. We hope they may also contribute to understanding the structure and function of autoantigens in Graves' disease and glycoprotein hormone receptors in general. The authors are grateful to their collaborators in the National Dental Institute, particularly Drs. Bellur Prabhakar, Edward Oates, and Abner Notkins, in the National Cancer Institute, Drs. W. O. McBride and M. Lerman for their contributions to the cloning studies.

When individually tested, vasoactive intestinal peptide (VIP), cholera toxin, and monoclonal thyroid-stimulating antibodies (TSAbs), but not TSH binding-inhibiting antibodies (TBIAbs), elevate cAMP levels in cultured human thyroid cells. In this study, we tested the effect on cAMP levels in human thyroid cells of the simultaneous presence of VIP and the other ligands noted. Monoclonal TBIAbs (11E8 and 122G3), which interact with a membrane glycoprotein containing the high affinity binding site of the TSH receptor, did not alter VIP-induced cAMP accumulation in the thyroid cells. These data indicate that VIP and TSH bind to distinct sites on the cell membrane. In contrast, monoclonal TSAbs 208F7 and 307H6, which interact with a portion of the TSH receptor other than its high affinity binding site, not only did not have their usual agonist activity, but, instead, caused a marked inhibition of human thyroid cAMP accumulation when VIP was also present. Mutual antagonism by two agonists, i.e. inhibition evident when TSAbs and VIP were mixed, was also found when cholera toxin was coincubated with VIP. The inhibitory effect of mixing cholera toxin and VIP was nearly immediate and was duplicated with mixtures of the beta-subunit of cholera toxin and VIP. The inhibition evident in mixing VIP and the TSAbs or cholera toxin was not duplicated in mixtures of isoproterenol and VIP or in mixtures of forskolin and VIP. These results suggest that the mutual inhibition of VIP and either TSAbs or cholera toxin is at a step that couples the TSH and VIP high affinity receptor-binding sites to the adenylate cyclase complex.

The present report describes a sensitive and quantitative binding radioassay for measurement of thyroid cell surface antibodies (TCSAb). Enzyme-dispersed thyroid cells from surgical specimens of human normal thyroid tissue were used after 7 days of culture. 125I-labelled Graves' IgG was shown to bind to cultured thyroid cells. The binding was time-and temperature-dependent and increased linearly with the number of thyroid cells. Evidence for specificity was provided by the lack of binding of radioiodinated Graves' IgG to human fibroblasts and by the negligible binding of 125I-labelled normal IgG to thyroid cells. A dose-dependent inhibition of binding of 125I-labelled Graves' igG to thyroid cells was produced by the addition of graded amounts of the unlabelled original Graves' IgG preparation, but not by normal IgG. Assays for TCSAb were performed on IgG preparations from patients with and without thyroid autoimmune disorders using the original Graves' IgG preparation as reference standard. Results were expressed in terms of arbitrary units/100 microgram IgG, 1 unit corresponding to the minimum amount of the standard IgG producing a significant inhibition of binding. Negative tests were found in most normal subjects (15/18) while low TCSAb levels (less than or equal to 1.8 U/100 microgram IgG) were detected in 3 cases. Increased TCSAb levels were found in the majority of the patients wit Graves' disease (14/21), in most of the patients with idiopathic myxedema (9/10) and in all of those with Hashimoto's thyroiditis (10/10).

In order to investigate the presence of alpha-adrenergic receptors in human thyroid, we have studied the effect of alpha-adrenergic agonists and antagonists on cGMP cellular content of human thyroid cells in primary culture. Epinephrine as well as TSH were not able to modify the cGMP cellular levels, while norepinephrine significantly increased cGMP accumulation already at 10 nM, a dose inactive on cAMP accumulation. A non selective alpha-adrenergic antagonist, phentolamine, significantly inhibited cGMP accumulation induced by norepinephrine. Norepinephrine-induced cGMP accumulation was unaffected by prazosin, an alpha 1-adrenergic antagonist, but was abolished by yohimbine, an alpha 2-adrenergic antagonist. Phenylephrine, an alpha-adrenergic agonist, produced an increase of cellular cGMP levels without modifying cAMP content. In the presence of TSH, the cGMP response to norepinephrine was not modified; however, the increase of cAMP levels was inhibited by norepinephrine at doses inactive on cAMP accumulation, but active on cGMP levels. The present results demonstrate the existence in human thyroid cells of alpha 2-adrenergic receptors, regulating the guanylate cyclase system. It may be postulated that the counter-regulation exerted by alpha-adrenergic agonists on the response to TSH operates on the TSH-dependent adenylate cyclase.

Insulin-like growth factor-I (IGF-I) stimulates the growth of thyroid cells of different animal species. However, conflicting results have been reported as regards the function and mechanism of the action of IGF-I at the thyroid level. This study was designed to determine whether normal human thyroid cells have IGF-I receptors and whether IGF-I could act on such cells through an autocrine mechanism. Human thyroid follicular cells in primary culture, not contaminated by fibroblasts, were used. They had specific and saturable binding sites for IGF-I, as revealed by radiolabeled binding method, and displayed an average receptor number of 2000/cell. Under the same experimental conditions, insulin receptors were not detectable. Human thyroid follicular cells secreted IGF-I into the culture medium, as assessed by a specific chemiluminescent immunoassay. The IGF-I secretory process was detectable for at least 12 days of culture, but a high degree of variability has been found among individual samples. Acid-gel chromatography demonstrated that IGF-I and a higher mol wt IGF-I, most likely IGF-I bound to its binding protein, were secreted by human thyroid cells. TSH stimulated the secretion of the two molecules in normal human thyroid cells. The TSH effect on IGF-I secretion was concentration dependent between 0.1 nmol/L and 0.1 mumol/L. GH stimulated IGF-I synthesis by thyroid cells in a concentration range from 20-200 micrograms/mL. Since binding studies demonstrated the presence of IGF-I receptors on human thyroid cells, IGF-I probably regulates human thyroid function through an autocrine mechanism.

In the presence of Ro 20-1724, a selective inhibitor of cyclic nucleotide phosphodiesterase, carbamylcholine increases cAMP and cGMP levels in human thyroid cells in primary culture. The increase of cAMP exhibited at concentrations of carbamylcholine between 10 fM and 10 pM, is dose- and time-dependent, it is maximum after 30 min and is abolished after 60 min. At higher carbamylcholine concentration (10 microM), cAMP increases rapidly, becoming maximum after 15 min, but returns to unstimulated values after 30 min. The increase of cGMP is also dose-dependent (0.1 nM-10 microM); it reaches the maximum after 30 min and returns to unstimulated values after 120 min. A significant increase of phosphodiesterase activity is observed at 10 microM carbamylcholine. Atropine, a muscarinic receptor antagonist, blocks carbamylcholine effects on both cAMP and cGMP production without affecting the thyrotropin-induced cAMP accumulation. Hexamethonium, a nicotinic receptor antagonist does not affect the cholinergic effects. In the presence of Ro 20-1724, 10 microM carbamylcholine significantly inhibits the effect of thyrotropin on cAMP production, while the combined addition of low doses of carbamylcholine and thyrotropin (0.1 nM and 10 pM, respectively) results in an additive effect on cAMP levels. Inhibition of thyrotropin activity on cAMP production, similar to that exerted by 10 microM carbamylcholine is produced by increasing free intracellular calcium; this inhibition is relieved by using a calmodulin-sensitive phosphodiesterase inhibitor, M and B 22948 at 50 microM dose. High concentrations (10 microM) of carbamylcholine increase the adenylate cyclase activity, without any significant effect on the thyrotropin-induced activation of the enzyme.(ABSTRACT TRUNCATED AT 250 WORDS)

It has already been shown that Wistar rat thyroid (WRT) cells in low concentrations of calf serum (0.5%) are under the influence of both thyrotropin (TSH) and insulin as regards growth. The present data show that epidermal growth factor (EGF), in concentrations up to 10 micrograms/ml, is not able to modify DNA synthesis in WRT cells. On the other hand, insulin-like growth factor I (IGF-I) stimulates DNA synthesis from a dose which is 10-fold lower than that of insulin alone. Combined stimulation of EGF and TSH in WRT cells is equal to that of TSH alone in relation to DNA synthesis, while the combined presence of TSH and IGF-I, or TSH and insulin, in the same medium results in an effect which is greatly superior to the theoretical sum of activities. Repetition of the same experiments using the original clone of WRT cells, but in high concentrations of calf serum (5%), shows that EGF stimulates DNA synthesis in a dose-dependent way from 0.1 to 100 ng/ml. Under these conditions, combined stimulation of EGF with TSH shows that DNA synthesis is equal to the predicted theoretical sum. No other differences in WRT cell sensitivity to either IGF-I or insulin, or IGF-I and TSH and insulin and TSH, can be noted. This finding is confirmed by the demonstration of specific and sensitive binding sites for EGF on WRT cells cultured in 5% calf serum; these binding sites are not present on WRT cells adapted to grow in 0.5% calf serum. Present data support the hypothesis that EGF and serum growth actions are mediated through the same analogous pathway, which is, however, different from those of TSH and/or IGF-I and/or insulin.

Mutant cells varying in the pathways of their responses to hormonal stimulation are useful in defining the subcellular steps in the mechanisms of hormone action. FRTL5, a strain of normal and differentiated cells originally derived from adult rat thyroids, which depends on TSH for growth in vitro, was used to produce five TSH-independent mutants, after chemical mutagenesis and selection in medium lacking TSH. Their characterization and comparison with wild type cells demonstrated full retention of differentiated thyroid function markers such as thyroglobulin production and active iodide transport, and a slower growth rate. Characterization of cAMP metabolism in mutants revealed levels of basal cAMP and adenylate cyclase and phosphodiesterase activities similar to those of wild type cells kept in a nonproliferative state in medium lacking TSH. Adenylate cyclase responsiveness to very low doses of TSH (10(-12) M) was fully retained in all mutant clones, but the TSH-dependent cAMP elevation, although comparable to that reported in wild type cells, was not followed by significant growth stimulation in mutants. These findings demonstrate that the persistence of functional TSH receptors in these cells and that of growth regulation in them is independent of cAMP elevation.

Human monoclonal antibodies have been generated from heterohybridomas obtained by fusing mouse myeloma cells with peripheral lymphocytes from patients with active Graves disease. This report characterizes four antibodies as presumptive thyrotropin receptor antibodies because they specifically inhibit thyrotropin binding and competitively inhibit thyrotropin-induced cAMP levels in human thyroid cells. Two of these antibodies, 208F7 and 206H3, are representative of autoimmune stimulators in Graves disease sera because they stimulate thyroid function in all assays, including the mouse bioassay; their ability to inhibit thyrotropin-induced cAMP increases in thyroid cells competitively is complemented by more than additive agonism at low (10 pM) thyrotropin concentrations. These stimulating antibodies interact more potently with human thyroid ganglioside preparations than with bovine thyroid or brain gangliosides; in contrast, they are poor inhibitors of 125I-labeled thyrotropin binding to liposomes containing the glycoprotein component of the human thyrotropin receptor. Antibodies 129H8 and 122G3 appear to be representative of inhibiting or "blocking" antibodies in Graves disease sera. Thus they have no intrinsic stimulatory action in assays of thyroid function but rather inhibit thyrotropin activity in the assays tested. These two antibodies do not react with human thyroid gangliosides but are strong inhibitors of thyrotropin binding to liposomes containing the high-affinity glycoprotein component from human, bovine, and rat thyroid membranes. The data unequivocally establish the pluritopic nature of the immunoglobulins in Graves disease and relate individual components or determinants of the thyrotropin receptor structure with specific autoimmune immunoglobulins.

Thyrotropin causes a rapid and significant increase in the fluorescence polarization of DPH when this hydrophobic probe is incorporated into a strain of functioning rat thyroid cells (FRTL5). This increase is ligand-specific and is not related to cAMP production. The phenomenon seems to reflect the interaction of thyrotropin with the glycoprotein component of its membrane receptor, as suggested by experiments in which thyrotropin causes increases in DPH fluorescence polarization in liposomes embedded with this receptor component but not with gangliosides. A strain of nonfunctioning rat thyroid cells (FRT), exhibiting no reactivity with monoclonal antibodies to the glycoprotein component of the thyrotropin receptor, requires two orders of magnitude higher concentrations of thyrotropin to exhibit a comparable phenomenon.

Homologous species specificity is demonstrated with bovine and human thyroglobulin in which the two terminal sugars of the B carbohydrate chain, sialic acid and galactose have been removed by enzymatic hydrolysis. The species specificity is demonstrated by measuring the ability of the deglycosylated thyroglobulin derivatives to inhibit thyrotropin-induced increases in cAMP in human, rat and bovine thyroid cells in culture. Thus human-human or bovine-bovine interactions have higher activity coefficients by at least an order of magnitude than their heterologous counterparts. The homologous interactions are confirmed in binding studies and shown to be associated with negligible degradation of the bound ligand over a 24 hour period.

A new strain, named WRT cells, has been generated from primary cultures of rat thyroids. The primary culture was grown in Coon's modified Ham's F12 medium with 5% calf serum, insulin, hydrocortisone, transferrin, somatostatin, glycyl-L-histidyl-L-lysine and thyrotropin (TSH). On the basis of the following facts, the WRT cell strain, cloned from the primary culture, was considered 'normal': the cells are euploid, not carcinogenic, not able to grow in soft agar, and show contact inhibition. Their differentiated functions consist of the ability to synthesize thyroglobulin and to take up iodide, and they have a TSH-dependent adenylate cyclase system. TSH increases cellular adenosine 3',5'-cyclic monophosphate (cAMP) levels and [3H]thymidine incorporation in WRT cells from a concentration similar to that active on another clonal rat cell line (FRTL-5), even though the cell replication appears to be differently regulated in the two cell strains. In fact, the WRT cell doubling time is 42 h and they are also able to grow in the absence of TSH, though more slowly. In the same conditions, FRTL-5 cells have a population doubling time of 38 h, but they are not able to grow in the absence of TSH. When the effect of the other growth factors of the medium was studied, insulin appears to be a growth stimulus by itself, while it is only a facilitative step for TSH action in FRTL-5 cells. WRT cells, unlike FRTL-5 cells, can grow with a population doubling time of 80 h, when cultured for prolonged periods in a medium with a low serum concentration (0.5%), but containing insulin plus TSH. In conclusion, the WRT cell strain is a new and interesting experimental model for studying growth factors at the level of the thyroid, especially for their mechanism of action on the TSH receptor.

FRTL-5 rat thyroid cells have receptors for both insulin and IGF-I which can be distinguished in binding studies. The ability of TSH to regulate each in an antiparallel manner is atypical. If these receptors are shown to have independent as well as coordinate activities, studies of the mechanisms of their receptor cross-talk in these cells will be relevant to understanding IGF-I and insulin receptors in other tissues.

Since Prostacyclin (PGI2) is a major product of arachidonic acid metabolism in the human thyroid, we have studied the effects of PGI2 on cAMP accumulation in human thyroid slices and cultured thyrocytes. In both systems, PGI2 caused a dose- and time-dependent increase of cAMP accumulation with higher potency and efficacy than PGE2. Two optically active isomers of 5,6-dihydro-PGI2, i.e. stable synthetic analogs of PGI2, had qualitatively similar effects to PGI2. The relative potency ratio between the alpha- and beta- isomer as well as their potency compared to PGI2 were substantially similar to their potency in inhibiting human platelet aggregation. In thyroid slices, PGI2 and its stable analogs had a greater effect than TSH in causing cAMP accumulation; however, in contrast to TSH, this effect was not associated with increased iodothyronine release except at maximal PGI2 concentrations. TSH had no detectable effect on thyroidal PGI2 synthesis and release. In cultured thyrocytes the effects of PGI2 and its stable analogs were considerably less than those obtained with TSH and required higher concentrations. Such a discrepancy was not found in the case of PGE2. These findings suggest the existence of a specific PGI2-responsive adenylate cyclase system in human thyroid cells other than thyrocytes, of possible physiologic significance.