| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Faculty of Medicine René Descartes (G.S., A.C., M.P.), Paris V, Site Necker, Institut National de la Santé et de la Recherche Médicale Equipe Mixte 0363, and Pediatric Endocrine Unit, Assistance Publique-Hôpitaux de Paris (AP-HP), Hôpital Necker Enfants-Malades, 75743 Paris, France; Department of Clinical Biology (L.L., J.-M.B.), Institut Gustave-Roussy, 94805 Villejuif, France; Faculty of Medicine Denis Diderot (F.G., A.-L.D.), Paris VII, and Department of Developmental Biology, AP-HP, Hôpital Robert Debré, 75019 Paris, France; Department of Pathology (M.T., B.C.), Institut Gustave-Roussy, 94805 Villejuif, France; Department of Cytogenetics (J.M., M.V.), Hôpital Necker Enfants-Malades, 75743 Paris, France; Biostatistics and Epidemiology Unit (S.M.), Institut Gustave-Roussy, 94805 Villejuif, France; and Department of Nuclear Medicine (M.S.), Institut Gustave-Roussy, 94805 Villejuif, France
Address all correspondence and requests for reprints to: Professor Michel Polak, M.D., Ph.D., Service d’Endocrinologie Pédiatrique, Assistance Publique-Hôpitaux de Paris, Hôpital Necker Enfants-Malades and Faculté de Médecine René Descartes, Paris V, Institut National de la Santé et de la Recherche Médicale Equipe Mixte 0363, 149, rue de Sèvres, F-75743 Paris, Cedex 15, France. E-mail: michel.polak{at}nck.aphp.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objective: This study aimed to investigate the ontogeny of thyroglobulin (Tg), thyroid peroxidase (TPO), sodium/iodide symporter (NIS), pendrin (PDS), dual oxidase 2 (DUOX2), thyroid-stimulating hormone receptor (TSHR), and thyroid transcription factor 1 (TITF1), forkhead box E1 (FOXE1), and paired box gene 8 (PAX8) in the developing human thyroid.
Design: Thyroid tissues from human embryos and fetuses (7–33 GW; n = 45) were analyzed by quantitative PCR to monitor mRNA expression for each gene and by immunohistochemistry to determine the cellular distribution of TITF1, TSHR, Tg, TPO, NIS, and the onset of T4 production. A broken line regression model was fitted for each gene to compare the loglinear increase in expression before and after the onset of T4 synthesis.
Results: TITF1, FOXE1, PAX8, TSHR, and DUOX2 were stably expressed from 7 to 33 GW. Tg, TPO, and PDS expression was detectable as early as 7 GW and was correlated with gestational age (all, P < 0.01), and the slope of the regression line was significantly different before and after the onset of T4 synthesis at 11 GW (all, P < 0.01). NIS expression appeared last and showed the highest fit by the broken line regression model of all genes (correlation age P < 0.0001, broken line regression P < 0.0001). Immunohistochemical studies detected TITF1, TSHR, and Tg in unpolarized thyrocytes before follicle formation. T4 and NIS labeling were only found in developing follicles from 11 GW on.
Conclusion: These results imply a key role of NIS for the onset of human thyroid function.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Structural differentiation of the human thyroid gland can been described by a precolloid, a beginning colloid, and a follicular growth stage (1). The precolloid stage is found in fetuses from 7 to 10 GW, when the thyroid consists of strands of compact unpolarized thyrocyte precursors. The beginning colloid stage is characterized by the appearance of small follicles formed by polarizing thyrocytes from 10 to 11 GW. From 12 GW on, progressive follicular growth occurs. The fetal thyroid acquires the capacity of iodide accumulation and thyroid hormone synthesis during the beginning colloid stage (2).
Since these landmark studies, progress of molecular biology allowed the identification of genes involved in thyroid development (Titf1, Foxe1, Pax8, Tshr) and thyroid hormone synthesis (Tg, Tpo, NIS, Pds, Duox2), and the ontogeny of these genes has been described in part in murine models (3, 4, 5, 6, 7, 8, 9, 10).
However, the molecular mechanisms underlying the onset of human thyroid function remain largely unknown. Recently, we showed that human thyrocyte precursors express the thyroid-enriched transcription factors thyroid transcription factor 1 (TITF1), forkhead box E1 (FOXE1), and paired box gene 8 (PAX8) from developmental d 32 and 33 on (11).
The present study aimed at determining the ontogeny of mRNA and protein expression of the genes that define the terminally differentiated thyrocytes and are involved in thyroid hormone synthesis. For this purpose, we analyzed the expression of the sodium/iodide symporter (NIS), thyroperoxidase (TPO), thyroglobulin (Tg), pendrin (PDS), dual oxidase 2 (DUOX2), and the thyroid-stimulating hormone receptor (TSHR) in the human embryonic and fetal thyroid. Our results imply a key role of NIS for the onset of human fetal thyroid function.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
After approval by the Institutional Ethical Committee of the experimental design and protocols, embryonic and fetal thyroid tissues (n = 45) were obtained from either elective termination of pregnancy or autopsies. No gross anatomical abnormalities were documented in any of these cases. Gestational age was determined from several developmental criteria: postovulatory age, as estimated from crown-rump length measured during an ultrasound scan (12, 13); hand and foot morphology, as defined by descriptive embryology (14, 15); and foot length (16).
The number of sampled thyroid tissues per GW were as follows: six at 7 GW; seven at 8 GW; six at 9 GW; four at 10 GW; four at 11 GW; three at 12 GW; two at 13 GW; one each at 14, 15, and 16 GW; two each at 17, 18, and 22 GW; one each at 23 and 24 GW; and two at 33 GW (n = 23 from 7–10 GW before onset of thyroid hormone synthesis; n = 22 from 11–33 GW after onset of thyroid hormone synthesis). Furthermore, a pool of six normal adult thyroid tissues obtained from surgical specimens was included. Tissue samples were snap-frozen and stored at –80 C before RNA analysis.
For immunohistochemical studies, three tissues per GW were fixed by immersion in 3.7% buffered formalin and then embedded in paraffin. Subsequently, 5-µm-thick sections were mounted on poly-L-lysine-coated slides and processed for immunohistochemistry.
RNA extraction and cDNA synthesis
Total RNA was purified from frozen thyroid tissues using RNeasy extraction, according to the manufacturer’s instruction (QIAGEN, Hilden, Germany). For all thyroid samples older than 9 GW, total RNA from each lobe was extracted separately. Amount and purity of total RNA extracted was measured with Nanodrop spectrophotometer (Nyxor Biotech, Paris, France). The quality of RNA preparation, based on the 28S/18S ribosomal RNAs ratio, was assessed using the RNA 6000 Nano Labchip (Agilent Technologies, Palo Alto, CA). Only tissues with a ratio greater than 1.5 were selected for further analysis.
Reverse-transcription was performed, using 500 ng of total RNA for each sample, in 20-µl reaction volume containing 50 U Moloney murine leukemia virus reverse transcriptase, 20 U ribonuclease inhibitor, 5 mmol/liter MgCl2, 10 mmol/liter Tris-HCl (pH 8.3), 10 mmol/liter KCl, 50 pmol/liter random hexamers (Applied Biosystems, Foster City, CA), and 1 mmol/liter deoxynucleoside triphosphates (Amersham Pharmacia Biotech, Uppsala, Sweden). The complementary DNAs were then diluted in nuclease-free water for quantitative real-time PCR.
Real-time quantitative PCR
Oligonucleotide primers and TaqMan probes specific for ribosomal protein, large, P0 (RPLP0), peptidyl propyl isomerase A (PPIA), Tg, TSHR, PAX8, NIS, TPO, PDS, and DUOX2 were designed to be intron spanning and were published earlier (17, 18, 19, 20). Primers and probes for 18S, TITF1, and FOXE1 were purchased from Assays-on-Demand (Applied Biosystems). The real-time PCRs were carried out using TaqMan Universal PCR Master Mix and were run on ABI-PRISM 7700 Sequence Detector System (Applied Biosystems). Each reaction was performed with an amount of cDNA equivalent to 2.5 ng of total RNA in a final volume of 25 µl, and expression of each target gene was determined simultaneously for all samples in a 96-well plate. Expression of each target gene was measured in duplicate and normalized relative to a set of three reference genes (18S, RPLP0, and PPIA) (21). The mean of the reference normalized expression measurements (
Ct) for duplicates was used for statistical analysis. When appropriate, the average expression of two independent samples of the same tissue was used as the final result for each tissue. Gene expression values were calculated according to the 2–Ct method. A pool of six normal adult thyroid tissues was used as calibrator. Results from individual expression levels of seven tissues from 14 to 18 GW and four tissues from 22 to 24 GW were pooled.
Statistical analysis
For each gene, we wanted to evaluate the relationship between the logarithm of mRNA expression and gestational age through a linear regression line and also to test whether the slope of this line was significantly different between the time period before the onset of thyroid hormone synthesis at 11 GW (7–10 GW) and after (11–33 GW). For this purpose, a broken line regression model (22) was constructed for each gene using reference normalized measurements (Ct). The onset of thyroid hormone synthesis at 11 GW was defined a priori. In this regression model, the P age tests a linear relationship between expression and gestational age, whereas the P broken line tests whether there is a significant difference in the slope of the regression lines before and after 11 GW. The adjusted R2 values represent the fit of the model or the predictive ability of the model to explain the observed variation in the mRNA expression. All P values were two-sided and were considered significant only if P < 0.01 to account for multiple testing.
Immunohistochemistry
After deparaffinization and rehydration, sections were subjected to a 12-min microwave treatment in 0.01 mol/liter citrate buffer (pH 6.0) or Tris/EDTA buffer (pH 8.0) for antigen retrieval. After inhibition of endogenous peroxidases with PBS solution containing 3% H2O2 for 10 min, the sections were incubated with the blocking solution for 15 min (Biogenex, Menarini Diagnostics, Chevilly-Larue, France). The sections were incubated for 60 min with primary antibody (see Table 2
) in a humidified chamber at room temperature. The streptavidin-biotin-peroxidase complex was used with diaminobenzidine as chromogen (Biogenex, Menarini Diagnostics), producing brown staining, and the streptavidin-biotin-phosphatase complex was used with Fast Red as chromogen (Dako France, Trappes, France) producing red staining. Finally, slides were counterstained with hematoxylin (Hematoxylin Mayer’s, Dako France) and mounted. Negative controls were performed by omission of the primary antibodies. Positive controls were made on sections of normal adult thyroid tissues.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Three distinct gene expression patterns were observed during the precolloid stage and before 10 GW (Fig. 1A). Figure 1B shows the expression (mean ± SD) of each studied gene as a function of the gestational age and histological stage. The expressions of the TITF1, FOXE1, PAX8, TSHR, and DUOX2 mRNA were detected in all samples, and their expression levels remained stable before and after 11 GW. Table 1 summarizes the results of the broken line regression analysis. The expression of TPO, Tg, and PDS mRNA was detected as early as 7 GW, however at low levels. Their expression levels increased significantly with age (all P age <0.01), and this increase was significantly different before and after 11 GW (all, P broken <0.01). The resulting fold change between 9 and 11 GW was of five times (PDS), of six times (Tg), and of eight times (TPO). NIS expression was detected only in one of seven tissues at 8 GW and in two of six tissues at 9 GW, and at very low levels. A tremendous increase of NIS expression (up to 277-fold) was observed between 9 and 11 GW. As early as 13 GW, NIS expression reached a level equivalent to levels observed in adult thyroid tissues. Its increase in expression was significantly different before and after 11 GW (P broken < 0.0001) and was associated with the highest adjusted R2 value (0.73) of all investigated genes.
|
|
Tissues of 7, 11, and 15 GW (n = 3 for each GW), representing the precolloid, the beginning colloid, and the follicular growth stage, were used for immunohistochemical studies of TITF1, TSHR, Tg, TPO, NIS, and T4 (Table 2
). TITF1 staining was detected in the nuclei of thyrocytes at 7, 11, and 15 GW (Fig. 2, A–C). TSHR staining was detected in the cytoplasm of thyrocytes at 7 GW, whereas at 11 and 15 GW immunoreactivity was observed at the basolateral membrane (Fig. 2, D–F). Tg staining was observed from 7 GW and was cytoplasmic with small intracellular accumulations. At 11 and 15 GW, Tg staining was detected in the follicular lumen (Fig. 2, G–I). Iodinated Tg or T4 was not present at 7 GW, was first detected at 11 GW in the lumen of the developing follicles, and showed a homogenous staining pattern of the colloid within the follicles at 15 GW (Fig. 2, P–R). TPO staining was not detected at 7 GW and was located to the apical membrane at 11 and 15 GW (Fig. 2, J–L). NIS staining was absent at 7 GW and was predominantly perinuclear with only limited basolateral staining at 11 GW, whereas at 15 GW, all follicles displayed strong basolateral staining (Fig. 2, M–O).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The present mRNA expression data show that terminal differentiation of the human thyroid follows a precisely timed gene expression program with three different gene expression patterns of the functional genes: DUOX2 expression is stable from as early as 7 GW, is not correlated with time, and is not modulated at the onset of thyroid hormone synthesis at 11 GW. Tg, TPO, and PDS represent a second group of genes, which are expressed at low levels at 7 GW but increase significantly before 11 GW. Their increased expression is significantly correlated with the onset of thyroid function. NIS is the last gene to appear, and it shows the strongest up-regulation at 10 GW. Its expression is the most strongly correlated with the sudden onset of thyroid hormone synthesis at 11 GW. This expression pattern suggests that NIS expression is the main limiting step for the onset of thyroid function in man.
Our immunohistochemical studies support these gene expression data. The detection of immunoreactivity for cytoplasmic Tg at 7 GW and T4 from 11 GW on are in accordance with previous biochemical, electron microscopic, and autoradiographic studies (2, 23). Furthermore, our results show that T4 synthesis occurs only in developing follicles, confirming the close interaction of structural and functional maturation during terminal differentiation. NIS expression is detected by PCR at the transcript level and by immunohistochemistry at the protein level coincidently with T4 synthesis. However, the NIS staining at this stage is predominantly perinuclear and may represent NIS protein synthesis within the endoplasmatic reticulum. The NIS staining of the basolateral membrane observed at this stage in very few thyrocytes may represent the beginning of NIS protein accumulation at the basolateral membrane. On one hand, the immunohistological technique might not be sensitive enough to detect the beginning of NIS integration in the basolateral membrane at this stage. On the other hand, NIS has a very high affinity to iodide (Km 20–40 µM) (24); thus, low NIS concentrations at the plasma membrane might result in sufficient iodide trapping for the onset of T4 synthesis. Strong basolateral staining is observed at 15 GW and in the adult thyroid gland. In contrast, TPO immunoreactivity at 11 GW is well established at the apical membrane. This result further supports the concept that basolateral NIS accumulation is the final step of terminal differentiation of thyrocytes.
The thyroid-specific expression of Tg, TPO, NIS, and PDS requires a subset of transcription factors, which are TITF1, FOXE1, and PAX8. Binding sites of these transcription factors are present in the promoters of Tg and TPO (TITF1, FOXE1, PAX8), in the NIS upstream enhancer (TITF1 and PAX8), and the PDS promoter (TITF1) (3, 24, 25). Furthermore, PAX8 plays a key role for terminal differentiation in vitro and acts synergistically with TITF1 in vivo (26, 27). Our quantitative PCR data show, however, that the expression levels of these genes are not modulated during thyroid differentiation. Thus, besides TITF1 and PAX8, either prolonged exposure to these factors or additional factors may be necessary to complete the transcriptional activation of the target genes.
Because activation of the TSH/Tshr pathway in the developing thyroid of the mouse occurs concomitantly with thyroid hormone biosynthesis, and Tshr-null mice show abolished NIS and TPO expression, the TSH signaling cascade represents a further key regulator of terminal differentiation (9). In our study, TSHR was expressed as early as 7 GW and showed no significant increase of mRNA levels during terminal differentiation. At the protein level, TSHR staining was predominantly cytoplasmic at 7 GW and became basolateral during the beginning colloidal stage at 10 to 11 GW, concomitantly with up-regulation of Tg, TPO, and PDS mRNA levels and with the onset of NIS gene expression. These results suggest that maturation of the basolateral membrane with appropriate targeting of TSHR might be a prerequisite for the onset of TSHR signaling and consecutive NIS expression. Along the same line is the recent demonstration that folliculogenesis in primary porcine thyrocyte cultures has a direct effect on NIS expression and response to TSH stimulation, suggesting a close interaction between structure and function of thyrocytes (28). However, further studies are needed to investigate the regulatory mechanisms of terminal differentiation.
In a more general context, our results are supported by the following observations. 1) Accidental radioablation of the fetal thyroid was observed in pregnant women exposed to radioactive iodide at the end of the first trimester (29). Thus, NIS expression provides a molecular explanation for the onset of radiosensitivity of the fetal thyroid. 2) T4 was detected in the fetal blood in vivo from 10–12 GW on, an observation that is in accordance with our results in vitro (30, 31, 32). 3) The same temporal gene expression pattern of Tg, TPO, and NIS is found in the developing thyroid of the mouse and rat, although no quantitative data from qPCR experiments are available in the murine model (7, 8, 9). 4) NIS expression is lost very early during the dedifferentiation process in thyroid tumors, whereas TSHR expression is retained (17). Furthermore, the fold increases in NIS (>100x) and PDS (5x) mRNA expression during thyroid differentiation were comparable to the fold decrease of NIS and PDS levels in thyroid carcinomas (17, 18). Thus, dedifferentiation during tumorigenesis might follow the inverse gene expression pattern observed during normal differentiation. 5) NIS mediates the active iodide trapping from the blood stream. Patients with inactivating NIS mutations suffer from congenital hypothyroidism due to iodide transport defect, indicating that no compensatory mechanism for NIS-mediated iodide transport exists in the thyrocytes. However, the phenotype is dependent on iodide intake (24, 33). Thus, iodide supply by active iodide trapping is the first limiting step of thyroid hormone biosynthesis, both for the onset of thyroid function in the developing thyroid and for normal thyroid function later in life.
In conclusion, our results show that terminal differentiation follows a precisely timed gene expression program with three different gene expression patterns. They confirm data on the onset of iodide accumulation and thyroid hormone synthesis in the human thyroid. The temporal and structural correlation of thyroid hormone synthesis with folliculogenesis supports the concept that structural and functional maturation are closely related. Finally, the data extend the crucial role of NIS for the onset of thyroid function in the human fetus and provide the molecular explanation for the onset of radiosensitivity of the fetal thyroid. Further studies will be necessary to define the molecular mechanisms of terminal differentiation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Conflict of Interest: There is no conflict of interest for any of the authors.
First Published Online October 31, 2006
Abbreviations: DUOX2, Dual oxidase 2; FOXE1, forkhead box E1; GW, gestational week(s); NIS, sodium/iodide symporter; PAX8, paired box gene 8; PDS, pendrin; PPIA, peptidyl propyl isomerase A; RPLP0, ribosomal protein, large, P0; Tg, thyroglobulin; TITF1, thyroid transcription factor 1; TPO, thyroperoxidase; TSHR, thyroid-stimulating hormone receptor.
Received July 7, 2006.
Accepted October 25, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Milenkovic, X. De Deken, L. Jin, M. De Felice, R. Di Lauro, J. E Dumont, B. Corvilain, and F. Miot Duox expression and related H2O2 measurement in mouse thyroid: onset in embryonic development and regulation by TSH in adult J. Endocrinol., March 1, 2007; 192(3): 615 - 626. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
