Developmental Regulation of Thyrotropin Receptor Gene Expression in the Fetal and Neonatal Rat Thyroid: Relation to Thyroid Morphology and to Thyroid-Specific Gene Expression¹

Abstract

The TSH receptor plays a pivotal role in thyroid gland function, growth, and differentiation, but little is known about its role or regulation in the fetus and neonate. To explore these questions, we systematically evaluated TSH receptor gene expression at the level of messenger RNA (mRNA) in thyroid glands obtained from rat fetuses and neonates, from 14 days gestation to day 5 of postnatal life. Results were compared with histological evidence of differentiation and to thyroid-specific gene expression. Northern blot and RT-PCR analysis revealed that TSH mRNA was first detected at low levels on fetal day 15, but it increased 3- to 15-fold on fetal days 17–18. Up-regulation of TSH receptor mRNA on fetal day 17–18 was accompanied by the first appearance of colloid formation and of follicular development on morphological examination. It was also paralleled by increased expression of the thyroid-specific genes thyroglobulin (Tg) and thyroid peroxidase. Unexpectedly, TSH mRNA abundance was 2- to 3-fold higher in pregnant dams than in nonpregnant adult females or adult males.

In view of the 8-day lapse between the first appearance of the thyroid diverticulum and up-regulation of TSH receptor gene expression, we conclude that pituitary TSH, acting through its receptor, plays an important role in terminal thyroid maturation, but it is not involved earlier in gestation. Similarly, these data support previous evidence that the weak thyrotropic activity of human CG could not be of significance in early fetal thyroid gland development. The increased TSH receptor mRNA on fetal day 17–18 may be attributable to up-regulation by TSH, which is first secreted into the fetal circulation at this time. The significance of the increased TSH receptor expression during pregnancy remains to be explored.

THE TSH RECEPTOR plays a pivotal role in thyroid gland growth, function, and differentiation (1, 2). Through effects mediated primarily by the cAMP signal-transduction pathway, the TSH receptor exhibits transcriptional control of the genes for the major thyroid-specific proteins thyroglobulin (Tg) and thyroid peroxidase (TPO), and it stimulates an array of cellular events, including iodine uptake and organification, as well as thyroid hormone synthesis and secretion (1, 2). A member of the subgroup 2, G protein-coupled receptor superfamily, it is composed of a large, extracellular domain, 7 hydrophobic transmembrane-spanning regions, and a short intracytoplasmic tail. The N-terminal extracellular domain seems to be sufficient for binding of hormone, whereas the cytoplasmic loops and C-terminal tail are important in signal transduction (1, 2).

Despite major advances in knowledge of the structure, function, and molecular biology of the TSH receptor in the adult, very little is known about its expression and regulation in fetal life or its role in fetal thyroid gland development. In thyroid follicular cells obtained from 15-day-old rat fetuses, TSH can stimulate folliculogenesis and iodine organification in vitro (3), an effect that is mimicked by forskolin (4). Recently, evidence of TSH receptor message also has been reported, at 15.5 days gestation, by in situ hybridization (5). Despite the presence of TSH receptor messenger RNA (mRNA) and protein at fetal days 15–15.5, evidence of TSH-dependent function, such as a follicular structure (6), TPO activity (7), and thyroid hormonogenesis (6), does not appear until 2 days later (fetal day 17). The present studies were initiated to provide insight into whether this delay might be related to quantitative changes in TSH receptor expression.

Materials and Methods

Animals

Female Sprague Dawley rats (Taconic Farms, Inc., Germantown, PA) were housed in an environmentally controlled room (22 C; lights on from 0700 to 1700 h) and mated with males overnight. The day a vaginal plug was seen was designated day 0 of pregnancy; the day of birth was designated as day 0 of life. After ketamine/rompun anesthesia was administered, the thyroid gland was removed, immediately frozen in liquid N_2, and stored at −70 C until analyzed. Control tissues were removed, frozen, and stored in a similar manner. Thyroidectomies in fetal rats were performed under a dissecting microscope. On fetal days 14 and 15, the entire cervical portion of the fetus was removed; and, on day 16, the thyroid was removed (attached to the trachea). Thereafter, the thyroid glands were obtained free of surrounding tissue. Pools of up to 8–12 fetal glands were used to obtain sufficient tissue for study. The study was approved by the Animal Care Committee of the University of Massachusetts Medical Center.

RNA preparation

Thyroid tissue was homogenized on ice using a Dounce glass-Teflon homogenizer; occasionally, a power (polytron) homogenizer was used to homogenize thyroid tissue from older animals. Total cellular RNA was extracted by a monophasic solution of phenol and guanidine isothiocyanate (TRIzol Reagent, Life Technologies, Inc., Grand Island, NY), as described by Chomczynski et al. (8). In brief, addition of chloroform-isoamyl alcohol (24:1) was followed by centrifugation. RNA in the supernatant was precipitated with isopropranol, washed with 75% ethanol, air dried, and dissolved in water treated with diethyl pyrocarbonate. RNA was quantitated by absorbance at 260nm and stored at− 70 C. Intactness of the RNA and equality of loading were monitored by electrophoretic fractionation on a 6.6% formaldehyde, 1.2% agarose gel, and ethidium bromide staining.

Northern hybridization analysis

Northern hybridization was performed as described previously, with minor modifications (9). In brief, 5 μg of total cellular RNA was electrophoresed in a 1.2% denaturing agarose gel, transferred to a Zeta-probe membrane (Bio-Rad Laboratories, Inc., Hercules, CA), cross-linked to filter by exposure to UV light for 1 min, and stored in a plastic bag at 4 C until use. The rat TSH receptor gene probe employed (kindly provided by Dr. Leonard Kohn, NIH) was the purified insert from clone T8AFB and represents residues −54 to 2780 of the nucleotide sequence reported for the rat FRTL-5 TSH receptor (10). It was labeled withα -³²P-deoxycytidine 5′-triphosphate by the random primer method to a specific activity of at least 1 × 10⁹ dpm/μg DNA. Prehybridizations and hybridizations were performed in 50% formamide, 5 × SSC (20 × SSC is 0.3 m sodium chloride, 0.3 m sodium citrate), 10 × Denhardt’s solution (100 × Denhardt’s solution is 2% Ficoll, 2% polycinylpyrolidone), 50 mm sodium phosphate (pH 6.5), 1% SDS, 250 μg/ml Escherichia coli DNA at 42 C for 4 and 18 h, respectively. Blots were washed with 2 × SSC, 0.1% SDS at room temperature, and twice with 0.1 × SSC, 0.1% SDS at 65 C and then exposed to either preflashed XAR-5 x-ray film (Eastman Kodak Co., Rochester, NY) using a Cronex Lightning Plus screen at −70 C or, in later experiments, a Storm 840 phosphor screen. The amount of radioactivity corresponding to each area was quantitated within the linear range of signals using, respectively, scanning laser densitometry (LKB 2400 GelScan XL, Amersham Pharmacia Biotech, Inc., Piscataway, NJ) or by the Image Quant statistical program (Molecular Dynamics, Inc., Sunnyvale, CA). To control for variability in loading, the relative abundance of TSH receptor message was normalized by comparison with expression of the housekeeping gene, rat glyceraldehyde phosphate dehydrogenase (GAPDH). For this purpose, the oligonucleotide probe for GAPDH (Oncogene Science, Inc., Uniondale, NY) was used.

Thyroid histology

After thyroidectomy, thyroid glands obtained from the same litters as those used for Northern analysis were immediately placed in Bouin’s solution for overnight fixation, washed in running water, and stored in 10% formalin until processing. Hematoxylin-eosin and periodic acid Schiff (PAS) stains were performed using standard methods.

Slot blot RNA analysis

Slot blot analysis was performed as described previously (9). In brief, 1 and 2 μg of each RNA sample were immobilized on a Zetaprobe membrane (Bio-Rad Laboratories, Inc.) using a Minifold ll slot blot system (Schleicher & Schuell, Inc., Keene, NH), cross-linked to filters by exposure to UV light for 1 min, and stored in plastic bags at 4 C. The labeling of probes, prehybridzation and hybridizations, and washing steps were performed exactly as described for Northern analysis. Homologous rat probes included the TSH receptor (described above) and the thyroid-specific proteins Tg and TPO. The Tg gene probe is a 0.64-kb complementary DNA (cDNA) kindly provided by G. Vassart (11). The TPO gene probe, a 2.8-kb fragment of FRTL-5 TPO cDNA inserted in pUC9, was kindly provided by S. Kimura, NIH (12). Results were quantitated using a Storm 840 phosphoimager.

Results

Ontogeny of TSH receptor gene expression in the fetal and neonatal rat thyroid

To address the question as to when the TSH receptor is first expressed during gestation, TSH receptor mRNA was initially evaluated by Northern hybridization analysis and quantitated by densitometry after monitoring the mRNA with respect to representation of ribosomal RNA (28S and 18 S) as an internal standard. Results were also normalized by comparison with the housekeeping gene, rat GAPDH (Fig. 1). The representation of TSH receptor mRNA transcript levels in total cellular RNA was very low on fetal days 15 and 16 but increased 3- to 15-fold by fetal day 17. There was a subsequent slight increase in relative mRNA abundance on neonatal day 5, but the extent of this increase was variable in repeated experiments. Two major transcripts of 5.6-kb and 3.3-kb were observed at all ages, similar to those observed previously in rat FRTL-5 cells. An unexpected finding was the 2- to 3-fold increase in TSH receptor gene expression in mothers, as compared with 5-day neonates.

Figure 1

Cellular representation of rat TSH (rTSH) receptor mRNA levels in thyroid glands obtained from fetal and neonatal rats, pregnant dams, and fetal liver (negative control). A (top), Total cellular RNA (5 μg) was prepared from thyroid glands and subjected to Northern blot hybridization analysis, as described in Materials and Methods. The rTSH receptor gene probe employed represents residues −54 to 2780 of the nucleotide sequence reported for the rat FRTL-5 TSH receptor. The exposure time was 5 days at 70 C, with two intensifying screens. Middle, The above blot was stripped and probed with the housekeeping gene GAPDH. Bottom, Ethidium bromide (EtBr) staining of total RNA resolved on a 1.2% agarose gel. B, Relative TSH receptor abundance in arbitrary densitometric units, normalized by comparison with GAPDH and expressed as mean +sem of three separate experiments. Note the unexpected 2- to 3-fold increase in TSH mRNA levels in mothers used as a positive control. f, Day of fetal life; d, day of postnatal life.

The first appearance of TSH receptor mRNA on fetal day 15 was verified by RT-PCR analysis using primers that amplified a 302-bp fragment of the C-terminal portion of the extracellular domain specific for the rat TSH receptor. The aforementioned amplified fragment also hybridized with the TSH receptor cDNA probe on Southern hybridization analysis (data not shown).

Induction of TSH receptor mRNA correlates with formation of colloid and follicles

To compare changes in TSH receptor gene expression with morphological evidence of thyroid maturation thyroid glands obtained from the same litters as those used for Northern analysis and RT-PCR were studied under light microscopy (Fig. 2). Before fetal day 17, the thyroid gland was difficult to distinguish from surrounding tissue. On fetal day 17, a well-defined thyroid gland, containing clusters of mitotic epithelial cells separated by a scanty stromal network (1A) and with minute amounts of PAS-positive material (1B), indicative of early colloid formation, was first observed. At 18 days gestation, the follicular structure was much better developed (2A), and increased amounts of PAS reactive colloid material were seen (2B). Further increase in thyroid size and follicular development was observed at 10 days postnatally (3A and 3B).

Figure 2

Morphology of the fetal and neonatal thyroid gland (350× magnification). Thyroid glands obtained from the same litters as those used for Northern analysis were processed for hematoxylin-eosin[ H & E (A)] and PAS (B) stains. Before fetal day 17, the thyroid gland was difficult to recognize on inspection. 1A (fetal day 17), Clusters of epithelial cells containing numerous mitoses and separated by a scanty stromal network are seen; 1B (fetal day 17), minute amounts of PAS-positive material, indicated by the arrows, first appear; 2A (fetal day 18), clear evidence of follicular development is observed; 2B (fetal day 18), increased colloid (indicated by the arrows) is observed; 3A and B (postnatal day 10), further follicular development is seen. Follicles are filled with colloid.

Induction of TSH receptor mRNA correlates with increased expression of Tg and TPO

Further studies were performed to determine whether TSH receptor gene expression was related to other aspects of thyroid maturation, such as expression of the thyroid-specific genes Tg and TPO. As seen in Fig. 3, up-regulation of TSH receptor mRNA was accompanied by a concomitant increase in expression of both Tg and TPO. In contrast, no signal was observed in the lane containing the negative control, liver.

Figure 3

Temporal association between gene expression of the TSH receptor (TSHr) and gene expression of the major thyroid-specific proteins Tg and TPO. Total cellular RNA was isolated from fetal thyroid glands of various ages, slotted in 1 and 2 μg samples, and hybridized with specific probe, as described under Materials and Methods.

Up-regulation of TSH receptor mRNA in pregnancy

Because of the observation that TSH receptor gene expression in mothers was 2- to 3-fold that found in 5 day neonates (Fig. 1), experiments were performed to determine whether this was simply attributable to a further increase in TSH receptor gene expression with maturation into adulthood or whether the TSH receptor was up-regulated in pregnancy. As noted in Fig. 4, TSH receptor mRNA abundance was again 2-fold higher in pregnant females, as compared with adult nonpregnant females and adult males.

Figure 4

Relative abundance of TSH receptor mRNA in thyroid glands obtained from pregnant (preg) and nonpregnant female rats, adult male rats, and liver (negative control). A (top), Northern hybridization analysis of rTSH receptor mRNA was performed as described in Materials and Methods; middle, the above blot was stripped and probed with the housekeeping gene GAPDH; bottom, ethidium bromide staining of total RNA resolved on a 1% gel. B, relative TSH receptor abundance, normalized by comparison with rat GAPDH and expressed in arbitrary densitometric units as mean +sem of two separate experiments.

Discussion

In the present studies, we have begun to explore the role of the pivotal thyroid regulator, the TSH receptor, in thyroid gland development by systematically examining the ontogeny of gene expression in the thyroid gland of the rat fetus and newborn, a frequently used model of human thyroid development. Using a cDNA probe specific for the full-length rat TSH receptor and a short portion of the 5¹- and 3¹-untranslated regions, we detected two major transcripts, of 5.6 kb and 3.3 kb, similar to those described previously in rat FRTL-5 cells (10). No apparent difference in the relative proportions of these transcripts, at the fetal and neonatal ages examined (as compared with adult animals) was observed.

Although TSH receptor message was detected as early as fetal day 15, morphological evidence of thyroid differentiation was not present until fetal day 17. This is consistent with previous studies that have shown that follicular development, iodine organification, and thyroid hormone synthesis do not appear until fetal day 17 (6, 7). Immunoreactive Tg, although demonstrable at 15 days of fetal life, undergoes a similar increase in abundance on fetal day 17 (6). Because the fetal rat thyroid can respond to both TSH (3) and forskolin (4) by fetal day 15, the reason for this 2-day delay presumably is not merely a lag in translation into functional protein. One important reason would seem to be that the physiological ligand, pituitary TSH, is not found in the fetal circulation until fetal day 17 (13). Our findings would suggest that coincident up-regulation of TSH receptor gene (and, presumably protein) expression on fetal day 17 may be an additional reason for this 2-day lag period.

The regulation of steady-state TSH receptor mRNA levels is complex and involves not only the homologous hormone TSH acting through its cAMP signal (14, 15) but the coordinate action of multiple hormones[ insulin (15, 16), thyroid hormone (17, 18)], growth factors[ insulin-like growth factor, IGF-1 (15, 16)], and DNA binding proteins [thyroid-transcription factor, TTF-1 (19), an as-yet-unidentified single-strand DNA-binding protein (20), a Y box protein (21), and a cAMP-response element modulator (CREM) (22)]. Multiple studies have shown that stimulation with TSH leads to desensitization of the TSH receptor (homologous desensitization), a phenomenon that involves not only a reduction in available receptors but a decreased coupling of the TSH receptor and the G_s subunit of the signal transduction machinery (23–26). In the best-studied model, the rat FRTL-5 cell line, TSH inhibits TSH receptor expression at a transcriptional level (14, 15); although, at lower concentrations of TSH, a stimulatory effect has been described (15, 25). In human thyroid monolayer cell cultures, TSH up-regulates TSH receptor mRNA at concentrations of TSH in the physiological range (27). Thus, the dosage of TSH, study conditions, and species may all play a role in determining the results obtained. In vivo, the TSH receptor is down-regulated by TSH in the adult rat (17). Our finding that TSH receptor mRNA increases on fetal day 17–18, a time when pituitary TSH is first secreted into the circulation, suggests that (contrary to the adult situation) the TSH receptor may be up-regulated by TSH in the fetus, though the mechanism remains unclear. This pattern is reminiscent of the LH receptor, which is up-regulated by LH in fetal life but down-regulated in the adult (28).

The association between TSH receptor mRNA expression and both morphological and molecular evidence of thyroid differentiation observed in this study provide strong evidence that TSH, acting through its receptor, plays an important role in terminal thyroid maturation but is not involved in early thyroid maturation. It is possible that other factors play a role in the thyroid maturation observed on fetal day 17. For example, both insulin and IGF-1 have important effects on both thyroid growth and function (14–16, 29), but their developmental expression within the thyroid gland is unknown. Other candidate genes are TTF-1, TTF-2, and Pax 8, thyroid-specific transcription factors that are important not only for commitment of progenitor cells to a thyroid-specific phenotype but for the expression of thyroid-specific gene expression (30). Evidence to date would suggest that TTF-1 and Pax 8, which are expressed 5 days before the expression of the genes for Tg, TPO, and the TSH receptor, are important in early thyroid maturation and are necessary, but not sufficient, for the development of the fully differentiated thyroid phenotype (30). Clearly, the pivotal importance of the TSH receptor in thyroid maturation is underscored by the severe hypothyroidism and hypoplastic thyroid glands found in hyt/hyt mice (31), an inborn strain of mice with a loss of function mutation of the TSH receptor (32). Similar findings have been reported in infants and children with inactivating mutations of the TSH receptor (33) and in babies born to mothers with potent TSH receptor blocking antibodies (34, 35).

In summary, these data suggest strongly that pituitary TSH, acting through its receptor, plays an important role in terminal thyroid maturation but is not involved earlier in gestation. The increased TSH receptor mRNA on fetal day 17–18 may be attributable to up-regulation by TSH, which is first secreted into the fetal circulation at this time. The demonstration that the TSH receptor gene is not expressed until relatively late in gestation provides further evidence that placental factors with thyrotropic activity, such as human CG, do not play a significant role in early fetal thyroid gland development. The significance of the increased TSH receptor gene expression in pregnancy remains to be explored.

Acknowledgments

We thank Jack Green for assistance with the figures.

¹

Presented, in part, at the 69th Annual Meeting of The American Thyroid Association, November 13–17, 1996, San Diego, CAlifornia. This work was supported, in part, by NIH Grant DK-32520 (to the University of Massachusetts Medical School).

²

Current address: Amgen, Inc., Newbury Park, California 91320-1789