| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Unité de Recherches sur la Glande Thyroïde et la Régulation Hormonale, XR 96 INSERM (E.J.H., C.J., B.S.), Unité de Recherches Hormones et Reproduction, U-135, INSERM (G.M.), and Laboratoire Virus, Neurones et Immunité, Faculté de Médecine (C.H.), 94276 Le Kremlin-Bicêtre; Unité de Recherches sur le Développement, Vieillissement et la Pathologie de la Rétine, U-450 INSERM (P.C., B.O.), 75016 Paris; and Service Commun de Biologie Moléculaire de l’Hôpital Saint-Antoine (E.C.), 75012 Paris, France
Address all correspondence and requests for reprints to: Dr. Bertrand Saunier, Bât. INSERM Gregory Pincus, Hôpital de Bicêtre, 80 rue du Général Leclerc, 94276 Le Kremlin-Bicêtre Cedex, France.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Besides its expression in thyroid tissue, the presence of a TSH receptor in nonthyroid tissues has also been described. Birnbaumer and Rodbell reported the stimulation of cAMP production by TSH in adipocytes (13), in which TSH binding was characterized soon afterward (14, 15). Similarly, Pekonen and Weintraub reported the existence of TSH-binding sites at the surface of lymphocytes (16). More recently, expression of TSH receptor messenger RNA (mRNA) has been reported in both cell types (17, 18). These results strongly suggest that TSH may play another role besides regulating the thyroid gland (19). In fact, the presence of TSH in cerebrospinal fluid (20) and brain (21) has been described, raising the question of a hypothetical local function of TSH in the central nervous system.
A few years ago we described the existence of specific [particularly compared with LH/CG (22)] and saturable binding of radiolabeled TSH at the surface of astroglial cells (23) in primary cultures from newborn rat cerebrum. However, TSH failed, surprisingly, to enhance cAMP production in these cells, contrary to what was reported in thyroid. Instead, TSH stimulated the release of arachidonic acid (23). Moreover, we found that activity of the type II 5'-iodothyronine-deiodinase, known to be regulated by other factors in these cells (24), was potently stimulated by TSH in a sustained manner (23). In astroglial cells TSH also induced a stimulation of mitogen-activated protein kinase isozymes, Erk1 and Erk2 (25). Together, these results suggested that a functional TSH receptor was effectively expressed in these cells. Thus, although TSH exerts several effects in astroglial cells, uncoupling of TSH receptor from adenylyl cyclase may suggest the existence of a TSH receptor variant.
In this report we first determined the sequence of the corresponding cDNA in astrocytes. By in situ hybridization, using a specific riboprobe, we then showed that, as in thyroid, the TSH receptor gene was expressed in cultured astroglial cells and in brain cryoslices. The protein was detected with anti-TSH receptor antibodies in primary cultured astrocytes in humans and in brain cryoslices in both rats and humans.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cells
(Escherichia coli) were purchased from
Stratagene (La Jolla, CA). The laser confocal microscope
was purchased from Carl Zeiss {[LSM410, (New York,
NY)]; lens, x40; 1.2 C-apochromat; excitation wavelengths:
FITC = 488 nm and rhodamine = 543 nm}
and was driven by Philippe Leclerc (INSERM, Le Kremlin-Bicêtre,
France).
Sequencing of the TSH receptor cDNA
Total RNA was isolated, as previously described
(26), from rat primary cultured astrocytes, and cDNA was
then synthesized as previously described (27). The entire
length of the astrocytic TSH receptor cDNA was sequenced by PCR with
eight pairs of oligonucleotides (Fig. 1
)
designed from the sequence of the TSH receptor cDNA in rat thyroid
cells (6): 1) 5'-ACAGCGCGCAACGATGAAGT-3' and
5'-GCATCAGGGTCTATGTAAGT-3'; 2) 5'-GCGACTGGAGCCACATTCTT-3' and
5'-GGAAGGAAGAGCAGTAACGC-3'; 3) 5'-CAAAGATGCATTTGGAGGAG-3' and
5'-CCGGATACTACTCTCATTAC-3'; 4) 5'-TACCCAAGTCACTGCTGTGC-3' and
5'-AAAGACATTGCCCAGGAG-AG-3'; 5) 5'-CAGGAAGAGACTCTACAAGC-3'
and 5'-AGAAGGAAG-CAGCAAACCCA-3'; 6)
5'-TGCAACACGGCTGGTTTCTT-3' and 5'-TTCACATAGCAGGAGCAGAC-3'; 7)
5'-GTACATCGCCCTTGTTCTCC-3' and
5'-TGTTGGGACAGACTCTCTGG-3'; and 8)
5'-TTCATCCTACTCAGCAAGTT-3' and
5'-AATACTCCCTAGAGATGAAC-3'. Both strands of the eight amplimers were
then submitted to sequencing (PE Applied Biosystems,
Foster City, CA) (28).
|
Plasmid construction and cloning for rat TSH receptor antisense RNA
synthesis
Specific PCR primers were designed from the sequence of the rat
TSH receptor cDNA (6), and one restriction site, for
EcoRI or HindIII (underlined),
respectively, was added to the 5'-end of each oligonucleotide:
5'-GTCAGAATTCGCTCTCCTGGGCAACGTCTT-3' and
5'-GACTAAGCTTGGCGAAGGTGATGGCATA-3'. The PCR product (see
Fig. 1
), of about 350 bp, was gel extracted, digested with both
EcoRI and HindIII restriction enzymes, and
inserted by ligation into a Bluescript plasmid, which was then
transfected into competent DH5 cells. Positive bacterial cell clones
were selected and grown for plasmid extraction and purification
(32). Then both strands were sequenced using primers
corresponding to T3 and T7 promoters.
In situ hybridization
Rat brains were fixed overnight in 4% paraformaldehyde in PBS
and then incubated in 30% sucrose. Brain sections of 15 µm were
freeze-cut. TSH receptor riboprobes, sense and antisense, were
synthesized with T3 or T7 RNA polymerase (see above). The
riboprobes were DIG-labeled (DIG-deoxy-UTP), as described in the
DIG-riboprobe kit protocol (Roche). The hybridization of
TSH receptor mRNA in brain sections was performed overnight at 50 C
with 10 µl of probes (10–20 ng/µl) in a solution containing 50%
formamide, 5 x SSC (saline sodium citrate solution), 0.02% SDS,
5% blocking reagent, and 0.1% N-lauryl sarcosine. Washing
was performed for 30 min at 50 C in 2 x SSC, for 20 min at 60 C
in 2 x SSC plus 50% formamide, and for 30 min in 2 x SSC,
then for 30 min in 1 x SSC at room temperature. Sections were
rapidly rinsed in 100 mM Tris-HCl, pH 7.5, plus
150 mM NaCl and then incubated in blocking
reagent. A conjugated antibody (anti-DIG alkaline phosphatase
conjugate) was used to detect the probes, and a subsequent alkaline
phosphatase-catalyzed color reaction was performed with X-phosphate and
nitro blue tetrazolium salt producing a precipitate. The signal
obtained with colloidal gold-coupled antibody (Roche) was
enhanced by incubating the preparation for a few minutes with BL
silver-enhancing blotting kit (British BioCell International, UK) and
observed with polarized light. Five in situ hybridization
experiments were performed independently, with four to five rats in
each.
Glial fibrillary acidic protein (GFAP), neuron-specific enolase
(NSE), and neurofilament detection by immunofluorescence
Double staining was performed by in situ
hybridization with TSH receptor RNA probes, and immunocytochemistry was
performed with anti-GFAP, anti-NSE, or antineurofilament antibody.
Immunodetection of the RNA probes was realized with an FITC-conjugated
antibody (anti-DIG fluorescein isothiocyanate conjugate). After RNA
detection, sections were incubated overnight with anti-GFAP (DAKO Corp.), anti-NSE, or antineurofilament antiserum diluted 1:50 in
a solution of 0.1% Triton X-100 in PBS and then with a 1:30 solution
of F(ab')2 goat antirabbit horseradish
peroxidase-conjugate serum (Caltag Laboratories, Inc.) for
2 h. After washing twice with PBS, sections were placed on slides
and mounted in Mowiol (mix of glycerol and polyvinyl alcohol) for
observation. When possible, a colocalization pattern was determined
that was the result of a computer-assisted pixel by pixel analysis. For
each pixel, the intensity of the input signal in each wavelength
channel was considered significant only when it was above a given
threshold. For the significant pixels, these values were reported in a
two-axis graph, each axis corresponding to one wavelength channel. A
"colocalization cloud" was then obtained, that was submitted to
further analysis to determine which pixels had the most significant
input signal in both wavelength channels, simultaneously.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) to
establish the sequence of the detected cDNA. These PCR products had the
same apparent sizes as those observed in thyroid (not shown) and were
subsequently sequenced. Comparison of our data to those in GenBank
showed that the sequence of the TSH receptor cDNA in astrocytes was
identical to that described in the thyroid (6).
TSH receptor gene expression in primary cultured astrocytes
We then performed in situ hybridization experiments
(Fig. 2) using RNA probes (see
Materials and Methods and Fig. 1) corresponding to part of
the last exon in the human TSH receptor gene (33) (as yet,
the structure of the rat TSH receptor gene has not been determined). In
primary cultured astroglial cells, the results of experiments performed
with sense riboprobe were negative (not shown). With antisense
riboprobe, this mRNA was detectable in only some cultured astrocytes,
but then at a very high level (Fig. 2, A–C). Many positive cells were
associated together in islets (Fig. 2B), but isolated-labeled cells
were also found (Fig. 2, A and C). Whereas all of the TSH receptor
mRNA-positive cells were also GFAP labeled (Fig. 2, D and E),
GFAP-positive cells did not all express the TSH receptor gene (Fig. 2, F and G). In addition, none of the GFAP-negative cells was detected as
TSH receptor positive.
|
, B and D). In sharp contrast,
staining was observed predominantly in discrete areas of young rats
brains in experiments performed with antisense probe (Fig. 3, A and
C).
|
In the hippocampus, both the dentate gyrus and Ammon’s horn were
labeled (Fig. 3, A and E). However, staining was clearly most prominent
in the granule cell layer of dentate gyrus. In Ammon’s horn, staining
was located at the level of the pyramidal cell layer and spatially
decreased from the subfield CA4 to the subfield CA1. In piriform and
postcingulate cortex, cortical layers 2–3 were the most stained in a
pattern of adjoining spots, whereas in the fiber layer above no
staining was detected (Fig. 3, F and G). In the deeper layers of the
cortex (piriform and postcingulate cortex) and in the white matter
below staining was observed; overall it was much less than in layers
2–3, but contained a few intense spots (Fig. 3C) not corresponding to
vessels. In the hypothalamus, only some nuclei were labeled (Fig. 3H).
In the brains of adult rats, these regions were not stained.
Combined TSH receptor mRNA and GFAP labeling in brain
As some astrocytes in primary culture expressed the TSH receptor
gene, we studied whether this was also the case in brain. Using laser
confocal microscopy, TSH receptor mRNA was detected in the surrounding
region of the hippocampus (Figs. 3A and 4A). Whereas many GFAP-positive cells
were detected in this area (Fig. 4B), only a few of them expressed the
TSH receptor gene (Fig. 4, C and D). Hence, most of the GFAP-positive
cells were negative for the expression of TSH receptor mRNA (Fig. 4, B–D). The same result was observed in the areas beneath the piriform
and postcingulate cortex (not shown).
|
), staining was
detected in NSE-expressing cells in both piriform cortex (Fig. 5, A–D)
and hippocampus (Fig. 5, E–L), suggesting that TSH receptor mRNA was
indeed located within neuronal cells.
|
, E–H). This contrasted with
the Ammon’s horn’s area where only a few cells were doubly labeled,
but most were singly labeled (Fig. 5, I–L). Moreover, TSH receptor
mRNA was not colocalized with axon-specific neurofilament (Fig. 5, M–O), suggesting that it was confined to the neuronal cell bodies
(Fig. 5, P–R). We also noticed labeling consistent with the expression
of the TSH receptor gene in surrounding blood vessels (Fig. 5, A–D),
but it was not colocalized with other labeling.
Combined labeling in the ependymal cell and subventricular area
Interestingly, TSH receptor mRNA was detected in the ependymal
cells layer, where coexpression with GFAP was sometimes detected,
albeit at very low level (Fig. 6, A–D).
It is, however, beneath the ependymal cell layer, in the subventricular
zone, that dual labeling was mainly detected (Fig. 6, A–D). Moreover,
in the cells of this area that expressed NSE (Fig. 6, E–G) or
axon-specific neurofilament (not shown), TSH receptor mRNA was not
detected. In the deeper structures of the brain, except for isolated
cells, GFAP-positive cells generally did not express the TSH receptor
gene (Fig. 6, A–D).
|
We further tested whether the TSH receptor itself could be detected in
these cells. Figure 7A shows that several
astrocytes were labeled using antihuman TSH receptor antibody (provided
by Edwin Milgrom), whereas no labeling was detected with nonspecific
antibody (Fig. 7B). This labeling was observed with three different
antibodies, each directed against a distinct part of the extramembrane
subunit of the human TSH receptor (34). This staining was
abolished when the anti-TSH receptor antibody was first incubated with
the purified extramembrane subunit of the human TSH receptor. Together
these data suggested that the observed labeling was specific, and
therefore that TSH receptor was indeed present in primary cultured
human astrocytes.
|
, A–D and G). Then, rat and human
brain cryoslices were incubated with anti-TSH receptor antibody. Albeit
stronger in humans, which may be explained by a higher affinity of this
antibody for the human TSH receptor compared with that in the rat, a
positive signal was detected in both cases (Fig. 8, E and F, and H and
I). This demonstrates that the TSH receptor transcript detected in the
brain is translated in vivo as well.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The possibility that TSH receptor was expressed in cultured astrocytes has been previously considered by others (35). Although adenylyl cyclase was functional in these cells and effectively coupled to other G protein-coupled receptors, the lack of stimulation of cAMP production by TSH led some authorities to believe that no functional TSH receptor was expressed in astrocytes (35). Nevertheless, our previous findings in astroglial cells were discordant with the interpretation that an absence of stimulation of cAMP production necessarily meant an absence of TSH receptor. Indeed, in these cells, we identified several TSH effects that appeared to be independent of stimulation of the cAMP pathway and yet still specific for this hormone (23, 25). Thus, all of these TSH effects were specifically blocked by monoclonal anti-TSH antibody, monoclonal anti-idiotypic anti-TSH antibody (36, 37), polyclonal anti-TSH receptor peptide antibody (38), and human Igs containing a high level of anti-TSH receptor-blocking activity (39, 40).
It is unlikely that the product of another TSH receptor gene could explain the uncoupling of adenylyl cyclase from the TSH receptor observed in astroglial cells (23). Indeed, according to the similar dimeric structure of the pituitary glycoprotein hormones (TSH, LH, and FSH) (41), the high sequence homology of the cDNAs (42), and the very close spatial structure (43) of their receptors, for evolutionary reasons (41) these striking similarities make unlikely the existence of another gene coding for a high affinity TSH receptor that would be completely different from that already identified. Lower affinity binding sites were also found in cultured astroglial cells (23), as also described in thyroid (44, 45). The concentration of TSH required to produce specific effects in astroglial cells (23, 25) tends, however, to indicate that high affinity binding sites mediate the effects of TSH. The existence of alternative spliced variants of the TSH receptor transcript has been observed in thyroid (46). In fact, we found that the TSH receptor cDNA in primary cultured rat astroglial cells was identical to that of the main type in rat thyroid cells. A priori, this result excluded that an alternate splicing of the TSH receptor pre-mRNA explains the change in functional coupling in astrocytes. Processing of the TSH receptor depends on the type of cell in which it is expressed (47). Such posttranslational differences might tentatively explain the uncoupling of this receptor from adenylyl cyclase in astrocytes.
Using a riboprobe corresponding to the TSH receptor gene, the expression of which was previously identified in thyroid and now in cultured astrocytes, we observed two main types of localization. One type of localization was transient, in neuronal cells, and mostly found in the dentate gyrus of the hippocampus. This was established by detection of TSH receptor mRNA in NSE-positive cells, with laser confocal microscopy analysis. The other localization was in astrocytes, in accordance with our previous reports. The latter expression was mostly detected in the ependymal and subventricular areas. This restriction of expression is, perhaps, comparable to our present observation of primary cultured rat and human cells that suggests that TSH receptor expression is confined to particular astrocytes.
Albeit as yet unknown, the pattern of this expression may suggest various physiological roles. In a previous report we formulated the hypothesis that TSH receptor expression in the brain may be involved in local thyroid homeostasis when we reported the stimulation of type II iodothyronine deiodinase activity with TSH in astrocytes (23). This is consistent with the pattern of expression of type II iodothyronine deiodinase in the brain, predominating in astrocytes and in tanycytes of the ependymal area (48). In addition, the description by others of TSH receptor gene expression in sheep hypothalamus (49) may be consistent with a neuroendocrine feedback of TSH upon TRH release or expression, for example.
The pattern of expression suggests additional physiological roles. In the dentate gyrus, where TSH receptor gene expression in neuronal cells culminated, neuronal precursors have been observed in rodents (50, 51). Progenitor cells, which can ultimately differentiate into both neuronal and astroglial cells (52), are found in the central system during development. Most precursors migrate toward their final location, and they fully differentiate (53). Whereas astroglial cells are well known to keep a potential for proliferation in adult brain, neurons were thought, until recently (50, 51), to have lost this potential. Moreover, enhancement of the TSH receptor transcript level in the brain coincided with a dramatic rise in thyroid hormone ß-receptor expression (54, 55). The thyroid system is implicated in neural development within the neonatal period (56). It was therefore tempting to link the enhancement of expression of TSH receptor to that of thyroid hormone ß-receptor, perhaps in relationship with the commitment of neuronal precursors in the brain at birth.
Finally, further reports have appeared recently in the literature, consistent with the persistence of stem cells in the central nervous system of adult mammals (57, 58, 59). It is generally accepted that embryonic stem cells are initially located within the inner part of the central nervous system. Later, this part will give rise to the ependymal and subventricular areas. As one might expect, stem cells in the adult were found mostly within the ependymal cells layer (57) and/or astrocytes of the subventricular zone (58). Whether astroglial and neuronal intermediary precursors are obligatory steps in their ultimate differentiation process is still under debate (59). Coincidentally, in our study TSH receptor gene expression in NSE-negative cells culminated in those latter areas. Although not yet proved, our observations are consistent with an expression in brain precursor and/or stem cells. This hypothesis would satisfactorily explain the prominent labeling of neuronal cells in the dentate gyrus, the staining detected in ependymal cells and astrocytes of the subventricular zone, and the existence of isolated GFAP-positive cells that might correspond to migrating cells.
The factors allowing TSH receptor expression in the brain remain unknown. Interestingly, TSH receptor gene expression in the hippocampus and cortex was regulated throughout the neonatal period (our data) and correlated positively with the level of circulating TSH in serum (60). In humans, a peak of TSH is also detectable in serum after birth, although it lasts only a few hours (61). In thyroid, TSH-induced regulation of TSH receptor expression has been reported (6), suggesting that TSH may also regulate TSH receptor expression in the brain. In the thyroid, this regulation was, however, cAMP dependent; such a mechanism was excluded in astrocytes (23). Nevertheless, other processes have been implicated in TSH receptor expression in the thyroid, such as TTF-1 transcription factor (1). Albeit its expression in the thyroid gland is well known, TTF-1 is, in fact, also important during brain development (62). Moreover, its activity was reported to be regulated in a cAMP-independent manner in nonthyroid tissue (63). Therefore, this left open the possibility for TSH to regulate this expression through TTF-1 in astrocytes also. The presence of TSH or TSH-like peptide in the brain and cerebrospinal fluid that has been reported in rats and primates (21) and in humans (20) might, therefore, be of physiological relevance.
Taken together, our results demonstrate the expression of the TSH receptor gene in both neurons and astrocytes in the brain. Whether TSH receptor expression in the brain is related to the commitment of precursor or stem cells after birth and with the increased level of TSH observed in the serum or plays a role later, as suggested by the more frequent occurrence of neurodegenerative disorders during hypothyroidism with elevated level of circulating TSH (64) or by the description of enhanced TSH receptor expression in Alzheimer’s disease (65), are questions remaining to be addressed.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Recipient of a postdoctoral fellowship from INSERM (Paris,
France).
3 Present address: Building 10, Room 8C112, National Institutes of
Health, 9000 Rockville Pike, Bethesda, Maryland 20892.
Received June 20, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-adrenergic receptors in cultured rat thyroid cells. J Biol
Chem 260:9230–9236 and ß
c-erbA genes in developing rat brain. Mol Endocrinol 5:1339–1350[Abstract]
- and ß-thyroid hormone receptor mRNAs,
including the ß2-subtype, in the developing mammalian nervous system.
J Neurosci 12:2288–2302[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  K. M. Webber, G. Perry, M. A. Smith, and G. Casadesus The Contribution of Luteinizing Hormone to Alzheimer Disease Pathogenesis Clin. Med. Res., October 1, 2007; 5(3): 177 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kurne, O. F. Aydin, and R. Karabudak White Matter Alteration in a Patient With Graves' Disease J Child Neurol, September 1, 2007; 22(9): 1128 - 1131. [Abstract] [PDF]
|
|
|
|
 |
|
 B. Saunier How Low Can Maternal Thyroxin Go? Endocrinology, June 1, 2007; 148(6): 2591 - 2592. [Full Text] [PDF]
|
|
|
|
 |
|
 R. P Peeters, W. M van der Deure, and T. J Visser Genetic variation in thyroid hormone pathway genes; polymorphisms in the TSH receptor and the iodothyronine deiodinases. Eur. J. Endocrinol., November 1, 2006; 155(5): 655 - 662. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. V. H. Grommen, S. Taniuchi, T. Janssen, L. Schoofs, S. Takahashi, S. Takeuchi, V. M. Darras, and B. De Groef Molecular Cloning, Tissue Distribution, and Ontogenic Thyroidal Expression of the Chicken Thyrotropin Receptor Endocrinology, August 1, 2006; 147(8): 3943 - 3951. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. E. Kempers, A. S. P. van Trotsenburg, D. A. van Tijn, E. Bakker, B. M. Wiedijk, E. Endert, J. J. M. de Vijlder, and T. Vulsma Disturbance of the Fetal Thyroid Hormone State Has Long-Term Consequences for Treatment of Thyroidal and Central Congenital Hypothyroidism J. Clin. Endocrinol. Metab., July 1, 2005; 90(7): 4094 - 4100. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Calvo, E. Jauniaux, B. Gulbis, M. Asuncion, C. Gervy, B. Contempre, and G. Morreale de Escobar Fetal Tissues Are Exposed to Biologically Relevant Free Thyroxine Concentrations during Early Phases of Development J. Clin. Endocrinol. Metab., April 1, 2002; 87(4): 1768 - 1777. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. W. Szkudlinski, V. Fremont, C. Ronin, and B. D. Weintraub Thyroid-Stimulating Hormone and Thyroid-Stimulating Hormone Receptor Structure-Function Relationships Physiol Rev, April 1, 2002; 82(2): 473 - 502. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||
, 5'- and part of 3'-untranslated regions;
, translated region; 
, PCR primers; 
, region corresponding to
the RNA probe used for in situ hybridization experiments
(starting nucleotide 1336). The size of each PCR product is
framed.
