Search

Liraglutide is a glucagon-like peptide-1 (GLP-1) analog developed for type 2 diabetes. Long-term liraglutide exposure in rodents was associated with thyroid C-cell hyperplasia and tumors. Here, we report data supporting a GLP-1 receptor-mediated mechanism for these changes in rodents. The GLP-1 receptor was localized to rodent C-cells. GLP-1 receptor agonists stimulated calcitonin release, up-regulation of calcitonin gene expression, and subsequently C-cell hyperplasia in rats and, to a lesser extent, in mice. In contrast, humans and/or cynomolgus monkeys had low GLP-1 receptor expression in thyroid C-cells, and GLP-1 receptor agonists did not activate adenylate cyclase or generate calcitonin release in primates. Moreover, 20 months of liraglutide treatment (at >60 times human exposure levels) did not lead to C-cell hyperplasia in monkeys. Mean calcitonin levels in patients exposed to liraglutide for 2 yr remained at the lower end of the normal range, and there was no difference in the proportion of patients with calcitonin levels increasing above the clinically relevant cutoff level of 20 pg/ml. Our findings delineate important species-specific differences in GLP-1 receptor expression and action in the thyroid. Nevertheless, the long-term consequences of sustained GLP-1 receptor activation in the human thyroid remain unknown and merit further investigation.

Specific thyroid hormone (T3) receptors are present in thyroid follicular cells, including the rat FRTL5 clonal line, but little is known about the effects of T3 on the growth and differentiated function of the thyroid. Unlike primary cultures of animal or human thyroid cells, FRTL5 do not secrete appreciable amounts of thyroid hormones. We now have studied the effects of T3 by itself and in combination with TSH and insulin-like growth factor-I (IGF-I) on [3H]thymidine incorporation into DNA, iodide uptake, and cAMP production in FRTL5. We also have investigated the expression of different c-erbA mRNAs in these cells. Specific binding of T3 to FRTL5 cell nuclei in intact cells occurred with a binding capacity of 0.1-0.15 ng T3/mg DNA and an apparent Kd of 0.4 nM. Using an RNase protection assay on total cellular FRTL5 RNA and specific cRNA probes, we demonstrated the presence of c-erbA alpha and -beta mRNAs, both encoding T3 receptors. Biological effects were assessed in serum-free medium or buffer containing 0.1% BSA after maintaining quiescent culture of cells for at least 5 days in hormone-free medium containing 5% calf serum. T3 alone stimulated a dose-dependent increase in [3H]thymidine incorporation that reached a plateau at 188% of the control value at 10 nM T3. At 10(-11) M TSH, T3 potentiated TSH-stimulated [3H]thymidine incorporation (2.2-fold), but at TSH concentrations greater than 5 x 10(-11) M, T3 had no effect or reduced the response to TSH. T3 potentiated the [3H]thymidine response to 2 and 10 ng/ml IGF-I by 1.5- to 1.7-fold. T3 alone had no effect on iodide uptake, but attenuated iodide uptake stimulated by TSH. T3 was more potent in inhibiting TSH-stimulated iodide uptake than in enhancing TSH-stimulated DNA synthesis. T3 did not affect either basal or TSH-stimulated cAMP accumulation. Thus, in FRTL5 thyroid follicular cells 1) T3 receptors are expressed, as measured by direct binding assays and by the expression of c-erbA mRNAs; and 2) T3 acts as a growth factor and weak antidifferentiation factor. We suggest that T3 may modulate the actions of TSH and growth factors in thyroid epithelium.

We have characterized the binding of 125I-IGF-I and 125I-IGF-II to plasma membranes purified from human thyroid tissue. IGF binding was time- and temperature-dependent. At 4 degrees C, maximal specific binding of 125I-IGF-I was 17.3 +/- 2.5% and of 125I-IGF-II was 8.8 +/- 2.0% (mean +/- SD/60 micrograms membrane protein). 125I-IGF-I binding was inhibited completely by unlabeled IGF-I, IGF-II, insulin, and the type-I IGF receptor monoclonal antibody, alpha IR-3. 125I-IGF-II was inhibited completely by unlabeled IGF-II and nearly completely by IGF-I. 125I-IGF-II binding also was inhibited significantly by insulin, suggesting that much or all of the IGF-II was bound to the type-I IGF receptor. Scatchard analysis revealed a single class of binding sites with a Kd of 6.0 +/- 4.2 x 10(-10) M for IGF-I binding and 5.7 +/- 1.3 x 10(-10) M for IGF-II binding. IGF-I binding was inhibited by a variety of salts in a dose-dependent manner, calcium and magnesium salts being more effective than sodium or potassium salts. Affinity crosslinking of 125I-IGF-I and -II showed clear evidence only for type-I IGF receptors. Thus, a crude plasma membrane fraction of human thyroid tissue expresses predominantly type-I IGF receptors.

Multiple signal transduction pathways interact in FRTL5 cells to promote thyroid follicular cell differentiated function and cell proliferation. In these cells, TSH is a tissue-specific mitogen that promotes DNA synthesis primarily through activation of adenylate cyclase. To further test the role of adenylate cyclase in regulating cell growth and differentiated function we have introduced into FRTL5 the human beta 2-adrenergic receptor (BAR) complementary DNA and have studied the ability of isoproterenol, alone and in combination with insulin-like growth factor I (IGF-I), to stimulate cAMP accumulation, iodide transport, [3H]thymidine incorporation into DNA, and cell growth. Wild-type FRTL5 were infected with a PLJ retroviral construct containing the BAR in either a sense (FRTL BAR) or antisense (FRTL RBAR) orientation, and cell populations were selected on the basis of resistance to the antibiotic geneticin. FRTL BAR expressed approximately 1.3 x 10(5) high affinity binding sites per cell for the beta 2-specific ligand, CGP-12177, while neither FRTL5 wild-type nor RBAR cells demonstrated any specific binding. FRTL BAR had significantly higher levels of intracellular cAMP, [3H]thymidine incorporation, and iodide uptake in the absence of added isoproterenol than FRTL RBAR or wild-type cells. In FRTL BAR, but not RBAR cells, isoproterenol stimulated a dose-dependent accumulation of cAMP, iodide uptake, [3H]thymidine incorporation, and cell growth. FRTL BAR and RBAR cells were equally responsive to TSH and to IGF-I. Isoproterenol enhanced the ability of IGF-I to stimulate [3H]thymidine incorporation in BAR but not RBAR cells. Isoproterenol partially inhibited the ability of TSH to stimulate cAMP generation and DNA synthesis. These studies demonstrate that activation of adenylate cyclase through the BAR introduced into FRTL5 cells by retroviral infection reproduces the range of biological effects in these cells stimulated by TSH and suggest that activation of adenylate cyclase is sufficient to stimulate thyroid differentiated function and cell growth. FRTL BAR cells will provide an interesting model system with which to study the heterologous regulation of both TSH and BARs through activation of a common signal transduction pathway, adenylate cyclase.

Somatostatin, a cyclic tetradecapeptide, is both a hypothalamic hormone and a paracrine peptide, with effects on many tissues. Despite the fact that somatostatin can inhibit various cellular events in a number of cell lines, somatostatin is a constituent of medium defined for optimal growth of FRTL5, a line of differentiated and nontransformed rat thyroid follicular cells. In the present study we have evaluated the role of somatostatin in the control of DNA synthesis in FRTL5 cells and have investigated the mechanisms of somatostatin interaction with pathways stimulated by TSH and insulin-like growth factor-I (IGF-I). Somatostatin inhibits TSH-stimulated DNA synthesis and cell proliferation in FRTL5 cells. Maximal effects are observed at somatostatin concentrations of 0.1-10 ng/ml, and the effects are diminished at somatostatin concentrations above 10 ng/ml. Somatostatin also inhibits (Bu)2cAMP-stimulated DNA synthesis, suggesting that the loci of somatostatin action are both proximal and distal to activation of adenylate cyclase. Somatostatin also inhibits DNA synthesis stimulated by insulin-like growth factor-I (IGF-I), a pleiotropic growth factor that works through non-cAMP-dependent pathways. The somatostatin analog octreotide is more potent than native somatostatin in inhibiting DNA synthesis stimulated by either TSH or IGF-I. Somatostatin does not alter TSH or IGF-I binding to FRTL5, demonstrating that somatostatin affects the postreceptor signal transduction pathways stimulated by these factors. We conclude that 1) the use of somatostatin in hormone-supplemented medium for FRTL5 is unnecessary and may inhibit cell growth; 2) somatostatin can inhibit the direct effects of IGF-I on peripheral tissues in addition to its ability to interfere with IGF-I synthesis by inhibiting the synthesis and release of pituitary GH; and 3) somatostatin is a useful tool for dissecting the pathways involved in mediating differentiated function and growth of FRTL5 cells.

We have studied the effects of thyrotropin (TSH) on the growth and on the levels of the mRNAs of the cellular proto-oncogenes, c-myc, and c-fos, in the specific target of TSH action, the thyroid follicular cell. FRTL5 cells, a cloned line from normal rat thyroid gland that depends upon TSH for its replication, were maintained in a quiescent state for 5 days by keeping them in a medium devoid of serum or TSH. The addition of bovine TSH (bTSH, 1 nM) increased DNA synthesis and stimulated cell proliferation after a lag period of 24 h. This growth response was anteceded by prompt, but transient, increases in the levels of c-myc and c-fos mRNAs, with peak responses at 60 and 30 min, respectively. The minimally and maximally effective concentrations of bTSH were 0.01 mM and 1.0 nM, respectively. Dibutyryl cAMP (Bt2cAMP) stimulated cell growth and increased the level of c-myc mRNA in a concentration-dependent manner, with maximum effects at a Bt2cAMP concentration of 1 mM. At the single concentration tested (1 mM), Bt2cAMP also increased the level of c-fos mRNA. Hence, bTSH-stimulated mitogenesis in quiescent FRTL5 cells is associated with rapid, but short-lived, increases in the levels of the mRNAs of the proto-oncogenes, c-myc and c-fos. Since bTSH is known to stimulate adenylate cyclase in these cells, and since the effect of TSH on c-myc and c-fos mRNAs is mimicked by Bt2cAMP, it is possible that these responses to bTSH are mediated, at least in part, by cAMP.

The present studies were undertaken to examine the factors that influence the growth of cells of endocrine gland origin, particularly the possible interactions between "nonspecific" growth factors and the trophic hormone for a target endocrine cell. As a model system, we explored the individual and conjoint effects of insulin-like growth factor-I (IGF-I) and TSH on the growth of FRTL5 cells, a nontransformed line of cloned rat thyroid follicular epithelium. In these cells, IGF-I and TSH each produced a dose-dependent enhancement of DNA synthesis and cell proliferation. When added together, IGF-I and TSH were markedly synergistic in stimulating DNA synthesis, producing increases in 3H-thymidine incorporation that were far greater than the sum of the effects of each alone. A similar effect of IGF-I was evident in the case of the stimulation of DNA synthesis produced by immunoglobulin G (IgG) preparations from the blood of patients with Graves' disease. Such IgG bind to the TSH receptor and mimic the actions of TSH therein. It is suggested, therefore, that there exist in the FRTL5 cell line at least two mechanisms for the regulation of growth, one activated at the level of the IGF-I receptor and the other at the level of the TSH receptor. When the two pathways are activated concurrently, a synergistic enhancement of DNA synthesis takes place. The findings indicate that the FRTL5 cell line is an excellent model in which to study these complex interactions and that IGF-I may be a determinant of thyroid cell growth, both normally and in certain thyroid diseases.

Adenosine inhibits TSH-stimulated [3H]thymidine incorporation into DNA in FRTL5 thyroid follicular cells by both inhibiting cAMP generation and acting at a locus beyond adenylate cyclase. On the other hand, adenosine markedly potentiates DNA synthesis in FRTL5 stimulated by insulin-like growth factor-I (IGF-I). The mechanisms of this latter effect are unknown, but require the coincubation of adenosine and IGF-I and not mediated by an increase in intracellular cAMP concentration. Adenosine increases the maximal response of FRTL5 to [3H]thymidine incorporation stimulated by IGF-I and increases the sensitivity of FRTL5 to IGF-I. These effects of adenosine are reflected by an increase in nuclear labeling as well as by an increase in [3H]thymidine incorporation into DNA. Adenosine also plays a role as an autocrine growth factor in FRTL5, since adenosine deaminase increases the response of these cells to TSH. The effects of adenosine on both TSH- and IGF-I-stimulated DNA synthesis are shared by guanosine and inosine, although with different potencies for the various guanine nucleosides. Inosine potentiates IGF-I-stimulated DNA synthesis, but inhibits TSH-stimulated DNA synthesis only weakly. Adenosine interacts with multiple receptors and with multiple postreceptor pathways in FRTL5 to produce divergent effects on the control of cell replication by two growth factors (TSH and IGF-I) that act through different postreceptor pathways.

Insulin-like growth factor I (IGF-I) is one of a number of mitogenic factors in the serum of animals and humans. We demonstrated previously that IGF-I is a potent mitogen for FRTL5 cells, a line of rat thyroid follicular cells. In this study, we assessed the relevance of this finding with respect to the levels of IGF-I found in human serum by comparing the effects of normal serum and serum from patients with untreated acromegaly or hypopituitarism on DNA synthesis in quiescent FRTL5 cells. As expected, when added to cells maintained in Coon's modified Ham's F-12 medium containing 0.1% BSA, but devoid of insulin, transferrin, TSH, or calf serum, normal serum produced a dose-dependent stimulation of [3H]thymidine incorporation into DNA. A similar, but more marked, effect was produced by the addition of serum from patients with untreated acromegaly. In multiple experiments, a standard concentration (0.5%) of acromegaly serum was more stimulatory to DNA synthesis than was normal serum. In a single experiment designed to eliminate interassay variation and define the relationship between the ability of serum to stimulate DNA synthesis and its IGF-I concentration, studies were performed with 0.5% concentrations of serum from 9 normal subjects, 15 patients with untreated acromegaly, and 3 patients with panhypopituitarism. On the average, [3H]thymidine incorporation in the presence of serum from patients with acromegaly was approximately 3 times, and that in the presence of serum from patients with hypopituitarism only one fourth, that in the presence of serum from normal subjects; there was no overlap of individual values in the three groups. For the entire study group, we found a highly significant correlation (r = 0.86) between the serum IGF-I concentration and the ability of that serum to stimulate thymidine incorporation into the DNA of FRTL5 cells. The stimulatory effects of serum from both normal and acromegalic subjects were decreased or abolished by the addition of a monoclonal antibody against IGF-I. In hypophysectomized rats, GH increases the thyroid to body weight ratio and enhances the effect of TSH to promote thyroid growth. Further, an inordinate frequency of nontoxic goiter in patients with acromegaly has been reported. Taken together with these observations, our findings suggest that the effect of IGF-I to promote thyroid cell growth in vitro has a counterpart in the living animal or patient.

Insulin-like growth factors (IGFs) are potent mitogens for FRTL5 rat thyroid follicular cells. IGFs also synergize the independent mitogenic effects of thyrotropin-stimulating hormone (TSH) and other agents that increase intracellular AMP concentration. We examined whether FRTL5 cells and M12 cells, a TSH-independent mutant cell line derived therefrom, secrete IGF that regulates the growth of rat thyroid follicular cells. Immunoreactive IGF-II, but not IGF-I, was found in media conditioned by FRTL5 cells; media from M12 cells contained four- to fivefold higher concentrations. Medium conditioned by FRTL5 and M12 both stimulated [3H]thymidine incorporation in FRTL5 and amplified the mitogenic effects of TSH. M12-conditioned medium was more potent than FRTL5-conditioned medium. Sm-1.2, a monoclonal antibody that recognizes IGF-I and IGF-II but not insulin, inhibited basal DNA synthesis in FRTL5 and M12 cells and the mitogenic effects in FRTL5 of agents that are synergized by IGF, such as TSH, forskolin, Bt2cAMP, and Graves'-IgG. Sm-1.2 did not inhibit the mitogenic response to insulin. Thus, rat insulin-like growth factor II (rIGF-II) is an autocrine growth factor that regulates FRTL5 growth, in part by amplifying the mitogenic response to TSH. Results with M12 cells raise the possibility that endogenous rIGF-II may partially mediate the TSH-independent growth of these cells.

In previous studies we have shown that insulin-like growth factor I (IGF-I) has a mitogenic effect in a line of rat thyroid follicular cells, the FRTL-5. In view of this effect, we undertook studies to identify and characterize some physicochemical and binding properties of the receptor for IGF-I in these cells and to determine what role it plays in the mitogenic activity of insulin and insulin-like growth factors in the FRTL-5 cell. Binding of 125I-labeled IGF-I (biosynthetic Thr59-IGF-I) to FRTL-5 was a function of time, temperature, and pH and was completely inhibited by high concentrations of unlabeled IGF-I. Scatchard plots of four saturation studies revealed a single apparent binding site with an average Ka of 4.2 +/- 0.6 X 10(9) M-1 (mean +/- SD) and an average maximum binding capacity of 20 +/- 2 pm/100 micrograms cellular protein. Rat IGF-II (rIGF-II) and insulin were far less potent that IGF-I in inhibiting the binding of [125I] IGF-I, and bovine TSH was without effect. 125I-Labeled IGF-II also bound to FRTL-5 cells. Binding was completely inhibited by unlabeled rIGF-II and, with lesser potency, by IGF-I. Even at high concentrations, insulin failed to inhibit the binding of [125I]IGF-II. Disuccinimidyl suberate cross-linked [125I]IGF-I to a moiety in FRTL-5 that had an apparent mol wt of approximately 135,000, as judged from sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions. Cross-linking of [125I]IGF-I was inhibited in a dose-dependent manner by unlabeled IGF-I and, with far lower potency, by rIGF-II and insulin. All three peptides stimulated the incorporation of [3H]thymidine into the DNA of FRTL-5 cells, IGF-I being the most potent, followed in decreasing order of potency of rIGF-II and insulin. The mitogenic activities of these polypeptides correlated well with their abilities to inhibit the binding of [125I]IGF-I. These data indicate that the FRTL-5 cell possesses a receptor for IGF-I that resembles in its binding and physicochemical properties the receptor for IGF-I in other tissues (type I IGF receptor) and that mediates the mitogenic response to IGF-I and insulin in these cells. FRTL-5 cells also contain a receptor for IGF-II (type II IGF receptor), but its role vis-à-vis that of the type I IGF receptor in relation to the mitogenic effect of IGF-II in these cells is uncertain.

We have evaluated whether the type I adenosine receptor mediates adenosine's ability to inhibit thyrotropin-stimulated cyclic AMP generation and DNA synthesis in FRTL5 cells. The xanthine derivative 1,3-dipropyl-8-cyclopentylxanthine, a selective antagonist for the type 1 adenosine receptor, binds to FRTL5 with high affinity and specificity. 1,3-Dipropyl-8-cyclopentylxanthine does not alter basal cyclic AMP levels but does reverse adenosine's ability to inhibit thyrotropin-stimulated cyclic AMP generation. 1,3-Dipropyl-8-cyclopentylxanthine also potently inhibits thyrotropin-stimulated and dibutyryl cyclic AMP-stimulated [3H]-thymidine incorporation into DNA in FRTL5 cells. Thus, in FRTL5 cells, 1,3-dipropyl-8-cyclopentylxanthine displays both adenosine antagonist and adenosine agonist properties, the latter occurring at a site distal to cyclic AMP generation.

We have studied the binding of recombinant human insulin-like growth factor I (hIGF-I), and hIGF-II, and rat IGF-II [rIGF-II (multiplication-stimulating activity)] to the human amniotic fluid IGF-binding protein placental protein-12 (PP12). PP12 displayed a 5- to 10-fold higher affinity for IGF-I compared to hIGF-II or rIGF-II. These differences in binding affinity were confirmed by both saturation binding analysis and competitive binding analysis using 125I-labeled IGF-I, hIGF-II, and rIGF-II and each of the unlabeled ligands. PP12 produced dose-dependent inhibition of IGF-I-stimulated [3H]thymidine incorporation in the rat thyroid follicular cell line FRTL5. Inhibition of IGF-I-stimulated thymidine incorporation paralleled the ability of PP12 to inhibit IGF-I binding to the surface of FRTL5. At a high concentration, PP12 also inhibited TSH-stimulated DNA synthesis but did not inhibit the binding of 125I-labeled TSH to FRTL5. Insulin did not inhibit the binding of 125I-labeled IGFs to PP12, and PP12 did not inhibit the ability of insulin to stimulate DNA synthesis. These data suggest that the ability of PP12 to inhibit TSH-stimulated DNA synthesis is through the inactivation of IGF produced endogenously by FRTL5. Low concentrations of PP12 produced a statistically significant enhancement of TSH-stimulated DNA synthesis; the mechanism by which this occurs remains unclear.