Evidence for genetic transmission of thyroid peroxidase autoantibody epitopic "fingerprints".

Autoimmune thyroid disease is characterized by the tendency to cluster in families and by IgG class autoantibodies to antigens such as thyroid peroxidase (TPO). The epitopes recognized by polyclonal serum autoantibodies can be quantitatively fingerprinted using four recombinant human TPO autoantibodies (expressed as Fab) that define A and B domain epitopes in an immunodominant region. To determine whether these fingerprints are genetically transmitted, we analyzed fingerprints of 63 members of 7 multiplex Old Order Amish families and 17 individuals from 4 Hashimoto thyroiditis families. Inhibition of serum autoantibody binding to [125I]TPO by the recombinant Fab was used to assess recognition of the TPO immunodominant region (4 Fab combined) and recognition of domain A or B (individual Fab). Complex segregation analysis was performed using a unified model (POINTER). For the 4 Fab combined inhibition phenotype, the no transmission model was rejected (chi2(4) = 20.67; P < 0.0032), and the most parsimonious model includes a major gene effect. More importantly, evidence for genetic transmission was obtained for the phenotype defined by the ratio of inhibition by subdomain Fab B1:B2. Thus, for this ratio (reflecting recognition of the B domain), the no transmission model was rejected chi2(4) = 63.59; P < 0.000008). Moreover, the polygenic hypothesis could be rejected, but not the major locus hypothesis, suggesting that major genes might be involved in familial transmission of this trait. In conclusion, our findings suggest that autoantibody recognition of the TPO immunodominant region and the TPO B domain is genetically transmitted. These data may open the way to the identification by candidate analysis or positional cloning of at least one gene responsible for the development of Hashimoto's thyroiditis.

Perforin expression by thyroid-infiltrating T cells in autoimmune thyroid disease.

Infiltration of the thyroid gland by lymphocytes is a hall-mark of autoimmune thyroid disease; it is particularly evident in Hashimoto's thyroiditis but is also seen in most patients with Graves' disease. Infiltrating cells are comprised primarily of T lymphocytes, of which only a minority appears to be activated. Their precise pathogenic role is largely unknown. Since perforin has been a marker for functionally activated cytotoxic T cells in situ we elected to assess the presence of perforin-containing cells in thyroid-infiltrating lymphocytes and establish their phenotype. Cells were isolated from seven subtotal thyroidectomy specimens, five from patients with Graves' disease and two with Hashimoto's thyroiditis. The novel findings were as follows: CD4+ perforin-containing T cells occurred only in Hashimoto's glands, suggesting a class II-restricted component of cytotoxicity; in Graves' disease, and to a lesser extent in Hashimoto's, perforin-expressing cells were primarily T cell receptor alpha beta- CD4-CD8- (double negative); double negative perforin-containing cells in peripheral blood of normal individuals were largely gamma delta + T cells. In Hashimoto's samples, the predominant population of T cells expressing perforin was CD8+. By comparison, in studies of the synovial fluid of knee joints from patients with rheumatoid arthritis only a minor population of the perforin-containing cells was double-negative. The data suggest significant differences in cytotoxic autoimmune mechanisms between the two autoimmune thyroid diseases. Functional characterization of double-negative T cells is necessary to define their role in autoimmunity.

Visualization of the thyrotropin receptor on the cell surface by potent autoantibodies.

In autoimmune thyroid disease there are various autoantibodies (Ab) to thyroid cell components. Among the best characterized are those to thyroid peroxidase (TPOAb) and to the thyrotropin receptor (TRAb). While TPOAb were successfully used to visualize TPO in human thyroid cells (HTC) and a rat thyroid cell line (FRTL5) by indirect immunofluorescence (IFL), similar attempts with TRAb and thyrotropin receptor (TSHR) failed. This could have been due to either relatively low serum levels of TRAb and/or low number of TSHR on thyroid cells. To test these hypotheses, we estimated the number of TSH binding sites on HTC, FRTL5 and Chinese hamster ovary (CHO) cells, transfected with cloned human TSHR (JP-26 cells), and for IFL staining employed 3 sera with the highest potency TRAb in our possession. A clear granular surface staining was detected on all 3 cell types with two sera; with the third, the least potent, no staining was seen. The density of staining paralleled the estimated number of TSHR per cell, i.e., JP-26 > FRTL5 > HTC. TSHR was also visualized on transiently transfected cells (COS-7-TSHR), facilitating quantitation of transfection. Our results suggest that the limiting factor in direct visualization of the TSHR is the TRAb concentration.

Characteristics and clinical correlates of a novel thyroid-stimulating autoantibody.

We reported a patient who gave birth to 3 children with transient neonatal hypothyroidism. She had 3 different antibodies (Ab) to the thyrotropin receptor (TSHR) in her serum, viz., TSH binding-inhibiting (TBIAb), thyroid-stimulating (TSAb) and an additional stimulating Ab (SAb). The SAb differed from TSAb in that its in vitro stimulating effect in human thyroid and FRTL5 cells was not inhibited by TBIAb [similar data now obtained with Chinese hamster ovary (CHO) cells transfected with cloned human TSHR]. Because of symptomatic goiter enlargement the patient underwent subtotal thyroidectomy. About 50% of the gland was infiltrated with lymphocytes; thyroid follicles had columnar epithelium, despite suppression of TSH by thyroxine and the presence of the potent TBIAb. Fifteen months later, when all 3 Ab showed a decline of approximately 3 fold, she gave birth to hypothyroid twins. These data support the following conclusions: 1) thyroidectomy and immunosuppression of pregnancy do not prevent neonatal thyroid disease if TSHR Ab (TRAb) are of high titer; 2) the thyroid is not a major site of TRAb production; 3) SAb is a thyroid stimulator, distinct from TSAb in that it does not share binding epitopes on the TSHR with either TSH or TBIAb; 4) SAb was the probable cause of thyroid growth in this patient.

Complex segregation analysis of antibodies to thyroid peroxidase in Old Order Amish families.

Autoimmune thyroid disease (AITD) may be characterized by the measurement in serum of antibodies to thyroid peroxidase. A population of Old Order Amish individuals and families was investigated to determine the prevalence of these antibodies and to examine hypotheses about the mode of transmission of thyroid antibodies. Complex segregation analyses were performed on 4 large multigenerational Old Order Amish families composed of 26 nuclear families containing 199 first degree relatives. Several alternative hypotheses of genetic transmission were examined. Hypotheses of no transmission, polygenic inheritance, single locus transmission, and mixed inheritance were compared. The analyses incorporated population prevalences obtained from a random sample of individuals. Results suggest that the pattern of transmission of thyroid antibodies in these families is consistent with a mixed model in which the major gene is transmitted in an autosomal dominant pattern. The mixed model postulates that there is a single gene of major effect as well as a polygenic component that can act separately and/or together to confer susceptibility for this phenotype. The parameter estimates for the major locus are: gene frequency (q), 0.16 +/- 0.01; maximum male penetrance, 0.35; and maximum female penetrance, 0.70. The heritability of the polygenic background is estimated at 0.41.

Evaluation of the biological significance of leukocyte infiltration of the thyroid in Graves' disease.

Graves' goiter size and gland function may be affected by interferon-gamma influencing actions of the thyroid-stimulating antibody. Goiter weight (n = 20), lymphocytic infiltration and class II antigen expression were assessed. The largest goiters, strikingly, has least infiltration but, overall, the looked-for negative correlation between goiter size and lymphocyte infiltration did not materialize. This was presumably due, in part, to inhibiting antibodies in many (8/18) patients' sera. In addition, our data do not support a major pathogenetic role for class II antigen in Graves' disease.

Fetal and neonatal hyperthyroidism and hypothyroidism due to maternal TSH receptor antibodies.

Autoimmune thyroid disease is a generic term that includes Graves' disease and Hashimoto's thyroiditis. In the former, there is overactivity of the thyroid due to the action of a thyroid-stimulating antibody (TSAb). Pathogenesis of Hashimoto's thyroiditis is largely cell-mediated immune destruction of the thyroid. Nonetheless, there may be either a goiter or an atrophic gland. There is evidence that in some patients the lack of goiter is associated with the presence in the blood of an antibody that inhibits the binding of TSH to its receptor. This TSH-binding inhibiting antibody (TBIAb), therefore, prevents TSH from stimulating the thyroid and constitutes an acceptable explanation for an agoitrous state. Collectively, TSAb and TBIAb, both of which are IgG, are known as TSH receptor antibodies (TRAb).

The site of the molecular defect in the thyroid gland of the hyt/hyt mouse: abnormalities in the TSH receptor-G protein complex.

The hyt/hyt mouse has a severe and pervasive primary inherited hypothyroidism with significantly depressed serum T4, elevated serum and pituitary TSH, and reduced thyroid gland iodide uptake. Previous ultrastructural and histologic analysis of the hyt/hyt thyroid gland along with these biochemical abnormalities support an inherited defect in TSH responsiveness of the hyt/hyt thyroid gland. In order to evaluate the potential site of the defect in the hyt/hyt mouse, we have studied the hyt/hyt gland and hyt/hyt TSH from a biochemical and molecular standpoint. Based on demonstrated bioactivity of hyt/hyt serum in the McKenzie bioassay, this reduced responsiveness to TSH in the hyt/hyt mouse is not due to reduced bioactivity of hyt/hyt TSH or a major structural abnormality in the hyt/hyt TSH molecule. In comparison to hyt/ + euthyroid littermates and +/+ BALB/cBY progenitor strain mice, the hyt/hyt mouse demonstrates a twofold reduction in thyroid gland basal cAMP and a markedly diminished response of adenylyl cyclase to exogenous TSH. However, hyt/hyt cAMP production is equivalent to the euthyroid mice after stimulation of thyroid glands by forskolin, cholera toxin, PGE1, and isoproterenol. These results support a defect in the TSH-G protein-adenylyl cyclase system in the hyt/hyt thyroid gland. Specifically, these findings suggest that the hyt/hyt mouse has a defect in TSH responsivity due to an inherited defect in the thyroid gland TSH receptor molecule. Since the hyt/hyt gland makes T3 and T4 but at diminished levels, the proposed defect in the TSH receptor would still impart partial function. Both hyt/hyt and euthyroid hyt/ + littermates make TSH receptor mRNAs of 5500 and 2400 base pairs. This suggests that the receptor defect does not represent a major structural abnormality of the gene. The receptor defect could represent a reduction in receptor number, receptor-TSH affinity, or TSH receptor-G protein coupling. The specificity of this effect on adenylyl cyclase-cAMP is shown by the reduction of TSH-cAMP regulated thyroid peroxidase (TPO) and thyroglobulin mRNAs in the hyt/hyt thyroid gland. Given the importance of TPO and thyroglobulin in normal thyroid hormone synthesis, the reductions in TPO and thyroglobulin mRNAs in the hyt/hyt thyroid gland may underlie the significant decrease in thyroid hormone production by the hyt/hyt mouse.

Comparison of FRTL-5 cell growth in vitro with that of xenotransplanted cells and the thyroid of the recipient mouse.

The present work was designed to compare in vitro cell growth kinetics with in vivo growth under conditions as similar as possible using labeling with [3H]thymidine. To this purpose, FRTL-5 cells were cultured as monolayers and as three-dimensional spheroids embedded in collagen gels and transplanted simultaneously into nude mice treated with perchlorate and a low iodine diet. The growth of the transplants was compared to that of the thyroids in host mice. In the intact thyroid, the fraction of [3H]thymidine-labeled follicular cells (FLC; 24-h labeling) increased sluggishly to a maximum of 10% after 3 weeks of goitrogen exposure, with a subsequent autoregulatory decrease to 3% at 7 weeks. A 4-fold higher FLC was found in six adenomas, indicating focal failure of growth-restraining mechanisms. In nonconfluent monolayer cultures the FLC was as high as 90%, even within large individual clusters where cells are in tight mutual contact. Solid, highly cellular grafts growing from transplanted monodispersed cells showed an average FLC of 20%, which is 5 times higher than the FLC in the identically stimulated mouse thyroid. In collagen-embedded cells, forming three-dimensional spheroids, the mean FLC decreased from 70% at 1 week in vitro (40% in vivo) to 20% at 3 weeks both in vitro and in vivo, suggesting effective auto-regulation of excessive growth in both conditions. However, these FLC were again much higher than the 3% FLC in simultaneously assessed host thyroids. The difference remained throughout the 45-day period studied. We conclude that FRTL-5 cells growing as monolayers and as three-dimensional spheroids in vitro or after xenotransplantation in vivo invariably show much higher proliferation rates under comparable environmental conditions than the normal follicular epithelium in the thyroids of host mice. The one exception is the confluent monolayer with near-zero growth, while densely packed three-dimensional transplants still grow intensively. Although growth-retarding cell to cell interactions are also clearly operative in growing FRTL-5 cells, they are less effective than those dampening the replication rate of the thyrocytes within the monolayer hull of normal follicles. A local failure of these mechanisms, allowing growth rates comparable to those of grafted FRTL-5 cells results in adenoma formation in normal thyroids. These observations call for caution in the transfer of in vitro growth studies with FRTL-5 cells to in vivo conditions prevailing in the normal thyroid.

Transient neonatal hypothyroidism: characterization of maternal antibodies to the thyrotropin receptor.

Transient neonatal thyroid disease is known to occur as a result of transplacental passage of maternal immunoglobulin G (IgG) that contains antibodies to the TSH receptor (TRAb). Thyroid-stimulating antibody (TSAb) produces hyperthyroidism, and antibody that blocks TSH binding (TBIAb) results in hypothyroidism. We have analyzed in detail the IgG from four women who gave birth to children with transient neonatal hypothyroidism and have shown stimulating and inhibiting antibodies to coexist in three. Human and/or rat thyroid (FRTL5) cells were used to show stimulatory effects in vitro. Inhibition was assessed as prevention of stimulation of these cells (by TSH or TSAb) or by the blocking of binding of [125I] TSH to its receptor. The IgG from two mothers was tested to identify whether the inhibitory and stimulating bioactivities resided in molecules characterized by either or both kappa- or lambda-light chains. Evidence for restricted heterogeneity (implying oligoclonality) was obtained, in that with one, purely inhibitory IgG all activity was with IgG kappa. With the other, stimulating and inhibitory activities were predominantly in IgG kappa and IgG lambda, respectively. In addition, the latter IgG contained a second stimulator that was not suppressed by either its own or other inhibitory IgG. Despite the presence of stimulatory antibodies in these IgG, the clinical effect was neonatal hypothyroidism, reflecting the greater potency of the inhibitory IgG in all instances. Based on the histories of these four women and their offspring it is apparent that TRAb, and in particular TBIAb, can develop at any point in the course of autoimmune thyroid disease, i.e. either at the onset or long after the autoimmune process has been established.

Clinical review 3: The clinical use of thyrotropin receptor antibody measurements.

There are relatively few circumstances wherein the assay of TRAb, as TSAb or TBIAb, is merited in the routine clinical management of patients with autoimmune thyroid disease. These include: 1. the prediction of neonatal hyperthyroidism; 2. the confirmation, by assay of maternal serum, and perhaps newborn blood, that neonatal hypothyroidism, identified on routine screening, may be transient and due to transplacental passage of TBIAb; 3. the prediction of relapse of hyperthyroidism at the end of a course of antithyroid drugs; this information should lead to more careful observation to identify relapse as early as possible; 4. the confirmation that adult agoitrous hypothyroidism reflects the effect of an antibody (TBIAb) blocking the action of TSH although, as discussed, there is little, if any, practical gain from this information. Additional applications of the routine use of TRAb assays will require the development of a procedure that combines the sensitivity, specificity, availability, and ease found variably in current TSAb and TBI procedures, and at much less cost than what is provided today.

Influence of cytokines on growth and differentiated function of FRTL5 cells.

A functioning rat thyroid cell line (FRTL5) was used to study interactions of TSH with interferon-gamma (IFN gamma) and tumor necrosis factor-alpha (TNF alpha). We examined effects on growth and differentiated function. Growth was assessed by DNA, incorporation of [3H]thymidine ([3H]Tdr) into DNA, and cell number; uptake of 125I (I- uptake) and the concentration of cAMP were measured simultaneously with growth assessment. IFN gamma stimulated the 30-min I- uptake and enhanced the effect of TSH. TNF alpha had minimal effects on growth indices (slight increase in [3H]Tdr incorporation) and had no influence on I- uptake; it inhibited TSH stimulation of both growth and I- uptake. When combined, IFN gamma and TNF alpha synergized in inhibiting TSH-stimulated growth. By itself TNF alpha inhibited stimulation of I- uptake by TSH, but augmented the enhancement seen with IFN gamma. The influence of calf serum (CS) was to increase the rate of incorporation of [3H]Tdr, but a similar qualitative pattern for the actions of the cytokines remained. A reverse profile (stimulation by IFN gamma, inhibition by TNF alpha, and stimulation by the combination) was seen for I- uptake, with CS increasingly diminishing all values. TSH stimulation of growth was progressively effective with increments of CS in the medium, but consistently there was inhibition that was greater with IFN gamma than with TNF alpha and was almost total with the combined cytokines. Stimulation of I- uptake by TSH was enhanced by IFN gamma, reduced by TNF alpha, and, when serum was present, increased to a degree that was greater than additive by the combined cytokines. Growth stimulation by insulin or insulin-like growth factor-I was inhibited partially by the individual cytokines and completely by the combination. Both insulin and insulin-like growth factor-I inhibited TSH stimulation of I- uptake, but similar stimulation by the cytokines was not affected. Simultaneous with inhibition of TSH-stimulated growth, both IFN gamma and TNF alpha enhanced cAMP accumulation. The mechanism of these multiple effects of IFN gamma and TNF alpha, especially on the actions of TSH, may not currently be fully explained, but they almost certainly reflect differing modes of action. The relevance to thyroid function in man is conjectural. Especially in Graves' disease, where thyroid infiltration with cells that secrete these cytokines is common, it seems probable that both IFN gamma and TNF alpha will have significant influences on both growth and differentiated cell function.

Variations in the culture medium for FRTL5 cells: effects on growth and iodide uptake.

We evaluated the composition of the medium in which FRTL5 cells were incubated for effects on basal and stimulated growth and differentiated function. Variables included the concentration of calf serum (CS) from 0% (0.2% BSA) to 5% and the addition of insulin or insulin-like growth factor-I (IGF-I); stimulation was by TSH, with some experiments using thyroid-stimulating antibody-immunoglobulin G. The basal rate of incorporation of [3H]thymidine and the content of DNA per well were enhanced progressively by the increasing concentration of CS, whether in a pretreatment period of 3 days or over a subsequent 2 days of incubation. Concomitantly, I- uptake was progressively inhibited. Stimulation of growth by TSH required serum and was greatest with 5% CS; TSH effect on I- uptake was progressively decreased by increments of CS. Insulin and IGF-I had effects comparable to each other, but much smaller concentrations of IGF-I were required; they both augmented growth and enhanced growth stimulation by TSH. These actions were progressively annulled by increasing concentrations of CS. I- uptake (basal and stimulated) was inhibited by insulin and IGF-I, but this inhibitory effect was abolished by CS, particularly at 5% concentrations. Overall, the influences of varying the concentrations of CS and the effects of insulin or IGF-I were consistent; if there was enhancement of growth stimulation, there was inhibition of the TSH effect on I- uptake. Underlying mechanisms for these observations have yet to be elucidated.

Graves disease presenting as painful thyroiditis.

Hyperthyroidism associated with subacute (painful, viral) thyroiditis is well-recognized as a clinical entity; the thyroid gland in Graves disease is minimally, if ever, tender and painful. We describe a 10-year-old girl with hyperthyroidism whose initial clinical presentation was predominantly a painful, tender goiter. Graves disease was established by high uptake of 131I with a diffuse pattern of distribution of radioactivity on scan and the presence of thyroid-stimulating antibody. Thyrotropin-binding inhibiting IgG and antibody to thyroid microsomal antigen were both positive. She responded well to treatment with propylthiouracil and had spontaneous regression of her thyroid pain. The cause of the severe pain and tenderness remains speculative.

Evidence of hyperthyroidism in apparently euthyroid patients treated with levothyroxine.

A concentration of serum thyroxine (T4) above the accepted normal range has recently been recognized to result from commonly prescribed replacement dosages of levothyroxine sodium. To determine if the high levels of serum T4 have undesirable metabolic effects, despite the fact that the subjects are accepted as euthyroid, we studied 28 patients receiving long-term levothyroxine therapy; 19 patients were considered to be receiving replacement dosages and nine suppressive dosages of levothyroxine. To assess the effect of levothyroxine on target tissue, we measured the thyrotropin response to protirelin and systolic time intervals obtained by simultaneous electrocardiography and echocardiography. Thyrotropin response to protirelin was suppressed in patients with elevated serum T4 levels and normal serum triiodothyronine levels. These patients also had shortened systolic time intervals typical of hyperthyroidism. Our data indicate that commonly given replacement dosages of levothyroxine may induce undesirable metabolic consequences and that these patients perhaps ought to be seen as having "subclinical hyperthyroidism." The prescribed dosage of levothyroxine as therapy for hypothyroidism is still frequently excessive.

A study of hepatic metabolism of thyroxine in BB/W rats treated with L-thyroxine.

The effect of thyroxine (T4) on T4 conversion to triiodothyronine (T3) and reverse T3 (rT3) was studied in BB/W rats. A colony of 38 BB/W rats was obtained and half were treated with thyroxine (T4), 1 mg per liter of drinking water. At 106 days of age the following groups were identified: nondiabetic, no T4 treatment, 8 rats; nondiabetic, T4 treated, 8 rats; diabetic, no T4 treatment, 10 rats; diabetic, T4 treated, 7 rats. All animals with diabetes were treated with insulin. T4 conversion to T3 and rT3 was assessed in liver homogenates in 0.1 M Tris-HCl buffer, pH 7.4, with or without 5 mM dithiothreitol (DDT). Serum T4 and rT3 were significantly elevated in both T4-treated groups (P less than 0.001), while serum T3 was not affected in either. Basal T4 deiodination to T3 by the liver homogenate did not change on treatment with T4; the addition of DTT increased T3 production in the homogenate from T4 treated nondiabetic animals (P less than 0.05). In both nondiabetic and insulin-treated diabetic rats there was no effect of T4 on the rate of rT3 production. Since, in the rat, 30-40% of circulating T3 is a direct contribution of thyroid gland secretion, and that would be absent in our T4-suppressed animals, the normal serum T3 may reflect increased absolute peripheral T3 production from the greater concentration of circulating T4.

Effects of gamma-interferon and tumor necrosis factor alpha on thyroid cells: induction of class II antigen and inhibition of growth stimulation.

A functioning rat thyroid cell line (FRTL5) was used to study interactions of thyrotropin (TSH) and various cytokines on expression of class I and II major histocompatibility complex (MHC) antigens and on growth stimulation. Only gamma-interferon (gamma-IFN) affected MHC antigen expression, i.e., to enhance class I, that was constitutive, and to induce class II. A concomitant, but probably not directly related, effect of gamma-IFN was to diminish growth stimulation, as effected by TSH and other activators of adenylate cyclase and measured by DNA increase and enhanced incorporation of [3H]thymidine into DNA. Stimulation of growth by tetradecanoylphorbol ester was also decreased by gamma-IFN. These effects of gamma-IFN were mimicked to some degree by tumor necrosis factor but there was major synergism between the two cytokines. Enhanced accumulation of cAMP by TSH and other agents was not diminished in these experiments. Flow cytometry analysis showed that inhibition of growth stimulation involved blocking of the passage of cells from the G0/1 phase to the S phase. The data may have relevance to goiter size in autoimmune thyroid disease.

Evidence supporting the identity in Graves' disease of thyroid-stimulating antibody and thyroid growth-promoting immunoglobulin G as assayed in FRTL5 cells.

This paper addresses the question: in Graves' disease is there a thyroid-growth stimulating IgG (TGI) separate from thyroid-stimulating antibody (TSAb)? Using the functioning rat thyroid line (FRTL5) cells for TGI (incorporation of [3H]-thymidine into DNA) and TSAb (increase in cAMP concentration) assays, we tested IgG from 30 Graves' patients. Positive TGI assay occurred only if cAMP increased in the cells and responses correlated, i.e., r = 0.95, P less than 0.001. With one very potent TSAb-IgG we showed that Fab was active as TGI and TSAb, IgG with pI of 8.5-9.0 was the most potent fraction in both systems and an inhibitory IgG prevented the action of both TSAb-IgG and TSH in both the TSAb and TGI assays. In the last example, the action was on the cell membrane and not on the TSH or IgG. These data are entirely compatible with the view that in Graves' disease, at least as tested in FRTL5 cells, the same IgG is active in stimulating both growth and adenylate cyclase.