| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Physiology, Institute of Biomedicine (P.M., J.K., P.P., I.T.H.), University of Turku, FIN-20520 Turku, Finland; Department of Physiology (M.T.-S.), University of Córdoba, Adva Menéndez Pidal s/n, 14004 Córdoba, Spain; Department of Cell Biology and Biochemistry (D.M.S.), Texas Tech University Health Sciences Center, Lubbock, Texas 79430
Address all correspondence and requests for reprints to: Ilpo T. Huhtaniemi, Department of Physiology, Institute of Biomedicine, University of Turku, Kiinamyllynkatu 10, FIN-20520 Turku, Finland. E-mail: ilpo.huhtaniemi{at}utu.fi
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylase and 3ß-hydroxysteroid dehydrogenase decreased.
The diminished response in steroidogenesis following chronic
T3 exposure was not a result of alterations in StAR mRNA
stability, but rather was due to inhibition of transcription of the
StAR gene. Similar acute stimulatory and chronic inhibitory responses
to T3 were found when LHR mRNA expression and LHR ligand
binding were examined. Transfections with an LHR or StAR
promoter/luciferase reporter construct demonstrated that a 173-bp
fragment of the LHR promoter containing an SF-1 binding motif was
involved in T3 response, as was the SF-1 recognition site
at -135 bp in the StAR promoter. Furthermore, the importance of SF-1
in T3 function was also verified employing mutation in the
bases of SF-1 sequences using electrophoretic mobility shift assays.
The potential physiological relevance of these findings was
demonstrated when similar responses were obtained in mice rendered hypo
and hyperthyroid. Collectively, these observations further characterize
the thyroid-gonadal connection and provide insights into the mechanisms
for a dual regulatory role of thyroid hormone in Leydig cell functions. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thyroid hormone, through its nuclear receptor, plays a crucial role in regulating differentiation, growth, and metabolism in higher organisms (8, 9). Thyroid hormone receptors (TRs) and steroid hormone receptors share many properties, including ligand-dependent activation, nuclear site of action, sequence-specific DNA recognition sites and the ability to regulate gene transcription. A novel C-erb-A gene, encoding the TR, has been detected in a human testicular complementary DNA (cDNA) library (10). To date, multiple TRs, binding T3 with high affinity and specificity, have been detected in different tissue (11). The level of TR expression in rat testis is highest during fetal and perinatal life, and decreases thereafter, whereas the morphological and functional development of the testis has been shown to be dependent upon thyroid hormone action (12, 13, 14). Recent studies show that mesenchymal and immature Leydig cells contain higher levels of TR messenger RNA (mRNA) than adult Leydig cells (15). It has also been reported that Leydig cells increase in number in adult rat after neonatal hypothyroidism (16).
A recent report from this laboratory (17) demonstrated the presence of a single class of high-affinity thyroid hormone receptors and demonstrated T3-induced StAR expression and progesterone (P) production in both primary and tumorous (mLTC-1, Ref. 18) murine Leydig cells. Furthermore, mLTC-1 cells, which express the nuclear receptor SF-1, coordinately increased T3-mediated StAR expression and steroidogenesis. Conversely, overexpression of DAX-1 (dosage-sensitive sex reversal-adrenal hypoplasia congenita critical region on the X-chromosome) markedly diminished SF-1 expression and concomitantly abolished the T3 responses.
The importance of thyroid hormone has long been implicated in mammalian reproduction because disturbed thyroid function is frequently associated with abnormalities in sexual function including impaired fertility (12, 19). Several lines of evidence suggest that thyroid malfunction affects hormones involved in reproduction, especially LH, FSH, and testosterone (T). Thyroid hormone deficiency also results in alterations in testicular morphology, and these responses can vary greatly from species to species (12, 20, 21). A preliminary report demonstrates that hypothyroidism in rats accounts for augmentation of LH/hCG binding to ovarian membranes, whereas chronic administration of T3 markedly diminished its level (22). Despite the apparent importance of thyroid hormone in mammalian reproduction, little information is available on its role in the regulation of gonadal steroidogenesis and LHR function.
Specific T3 binding has also been demonstrated in mechanically dispersed Leydig cells of goat testes, where the effects of T3 on androgen production are mediated through a thyroid hormone-induced protein (TIP) that has not yet been characterized (23). Thyroid hormone has been shown to potentiate cellular differentiation and LH/hCG receptor formation in pig granulosa cells in vitro in response to FSH stimulation (24). In MA-10 mouse Leydig tumor cells, it has been reported that longer stimulation with cAMP analog resulted in suppression of StAR mRNA, StAR protein, and P production (25). However, the mechanisms responsible for the long term attenuating responses of Leydig cells to T3 remain to be established.
The present findings document, for the first time, that chronic exposure to T3 significantly reversed its observed increased acute response (8 h) on StAR expression and steroidogenesis, and that this reversal is in large part due to decreased cholesterol supply to the mitochondrial inner membrane. In addition, T3 acutely increased LHR gene expression and ligand binding, whereas its longer effects progressively suppressed these levels. The findings in cultured cells correlated with in vivo data obtained in mice with experimentally induced hypo and hyperthyroidism. Furthermore, mLTC-1 cells transfected with LHR promoter-driven luciferase constructs demonstrated a positive thyroid hormone response element within 173-bp upstream of the transcription start site of the LHR gene. These findings provide novel evidence that thyroid hormone plays a dual role in regulating LHR function, and control of steroidogenesis.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Isolation, purification, and culture of primary mouse Leydig
cells
Mouse testicular Leydig cells were prepared as described
previously (17). Briefly, after removing the testicular
capsule, interstitial cells were dissociated from the testicular tissue
by treating with 0.2% collagenase (20 min, 34 C, 95%
O2/5% CO2), separated by
filtration, washed, and purified by continuous Percoll (Amersham Pharmacia Biotech, Uppsala, Sweden) density gradient
centrifugation (1.01–1.126 kg/liter). Cell types were separated and
collected based on their buoyant densities with Leydig cells being
found in the zone equivalent to 1.07 kg/liter of Percoll. The purity of
Leydig cells in this fraction, as assessed by 3ß-hydroxysteroid
dehydrogenase (3ß-HSD) staining, was found to be 75–80% (17, 26). Cells were subcultured in Waymouth’s medium supplemented
with 9% heat-inactivated horse serum (Life Technologies, Inc.) and 4.5% FCS (Bioclear UK Ltd., Mile Elm Calne,
Wiltshire, UK) containing penicillin and streptomycin.
RNA extraction and RT-PCR
Total RNA was extracted from isolated Leydig cells of different
experimental groups employing the single step acid guanidinium
thiocyanate-phenol-chloroform extraction method (27). The
isolation and amplification of mouse (mLTC-1) StAR (4),
LHR (28), and L19 (29) cDNAs were carried out
as previously described (17). Briefly, the following
primer pairs were used:
StAR (forward) 5'-GACCTTGAAAGGCTCAGGAAGAAC-3' (bases -51 to -27)
StAR (reverse) 5'-TAGCTGAAGATGGACAGACTTGC-3' (bases 931 to 908)
LHR (forward) 5'-CTCTCACCTATCTCCCTGTC-3' (bases 179 to 195)
LHR (reverse) 5'-TCTTTCTTCGGCAAATTCCTG-3' (bases 878 to 858)
L19 (forward) 5'-GAAATCGCCAATGCCAACTC-3' (bases 154 to 173)
L19 (reverse) 5'-TCTTAGACCTGCGAGCCTCA-3' (bases 559 to 540)
To control for variation in RT-PCR efficiency, a 405-bp fragment of the
L19 ribosomal protein gene was coamplified in each sample. RT and PCR
of the target genes were sequentially run in the same assay tube, as
described previously (17, 30). Two micrograms of total RNA
were reverse transcribed and the cDNAs generated were further amplified
by PCR in a 50 µl reaction mixture containing 1 nmol/liter of
oligoprimers, 200 mmol/liter of a deoxy-NTP mixture, including
[32P]-CTP, 20 U RNasin, 12.5 U avian
myeloblastosis virus reverse transcriptase (AMV-RT) and 2.5 U
Dynazyme-DNA polymerase in 1x PCR buffer (10 mM Tris-HCl,
50 mM KCl, 1.5 mM MgCl2
and 0.1% Triton X-100, pH 8.8) (Finnzymes, Espoo, Finland). The
reaction was initiated (see Ref. 17 for StAR) for mouse
LHR at 50 C for 10 min (RT), followed by denaturation at 97 C for 3
min. The PCR was then run for 30 cycles, defining denaturation at 96 C
for 1.5 min, annealing at 53 C for 1.5 min and extension at 72 C for 2
min (PTC-200, Peltier Thermal Cycler, MJ Research, Inc.).
A final cycle of extension at 72 C for 10 min was included. The
molecular sizes of the PCR products (StAR, LHR, and L19) were
determined in 1.2% agarose gel. The gels were then vacuum dried, and
exposed to Fuji Photo Film Co., Ltd. x-ray film (Tokyo,
Japan) at 4 C for 1–3 h. The relative mRNA levels of StAR, LHR, and
L19 were quantified by phosphorimaging and densitometry (Tina 2.0
Program, Raytest Isotopenmeßgeräte GmbH, Straubenhardt,
Germany).
Preparation of mitochondria, and immunodetection of the 30-kDa StAR
protein
Isolation of mitochondria was carried out from control and
treated mLTC-1 cells as described previously (25, 31). In
brief, cells were washed with 0.01 mol/liter PBS, collected and
homogenized at 4 C (30 strokes at 1200 rpm) with a
Potter-Elvehjem homogenizer, fitted with a serrated pestle. The
homogenate was centrifuged at 600 x g for 20 min to
remove broken cell debris and nuclei, and the resulting supernatant was
further centrifuged at 10,000 x g for 25 min. The
pellet containing mitochondria was washed twice in the same buffer and
pelleted each time at 9,000 x g for 15 min.
Mitochondrial protein (20 µg) from each group was solubilized in sample buffer (25 mmol/liter Tris-Cl, pH 6.8, 1% SDS, 5% ß-mercaptoethanol, 10% glycerol, and 0.01% bromophenol blue) and loaded onto a 12% SDS-polyacrylamide gel (Mini Protean II System, Bio-Rad Laboratories, Inc., Hercules, CA), as described by Laemmli (32). Electrophoresis was performed at 200 V for 1 h, and proteins were electrophoretically transferred onto a nitrocellulose membrane (Hybond, Amersham Pharmacia Biotech, Buckinghamshire, UK), and the membranes were processed as described previously (31). The immunodetection of StAR protein was performed using an ECL Western blotting detection kit (Amersham Pharmacia Biotech), and the membranes were exposed for 1–5 min to Fuji Photo Film Co., Ltd. x-ray film and quantitated as above.
Northern blot analysis
Total RNA (20 µg) from different treatment groups was assessed
using Northern hybridization analysis. Briefly, antisense complementary
RNA probes, a NotI fragment (960 bp) of the mouse StAR cDNA,
a BglII fragment (410 bp) of the extracellular domain of the
rat LHR, and an EcoRI fragment (298 bp) of P450scc were
produced by in vitro transcription (Promega Corp., Madison, WI) with T7 RNA
polymerase, dNTPs and [32P]-UTP
(Amersham Pharmacia Biotech). For preparing cDNA probes,
an EcoRI-PstI fragment (780 bp) of SF-1, a
SacI-Bgl ll fragment (906 bp) of 3ß-HSD, and a
NotI fragment (1713 bp) of P450c17 cDNA were labeled with
[32P]-CTP using the Prime-a-Gene labeling
method (Promega Corp.). The labeled probes were purified
using Sephadex G-50 nick columns (Amersham Pharmacia Biotech). Prehybridization (12 h) and hybridization (16
h) of the ribo- and cDNA probes were carried out at 66 C and 42 C,
respectively, under stringent conditions as previously described
(17, 31, 33). Following hybridization, the membranes were
washed twice at room temperature for 20 min with 2x SSC containing
0.1% SDS, followed by 1–3 h at 66 C or 42 C with 0.1x SSC and 0.1%
SDS until removal of the background counts. To assess the variation of
mRNA levels, the membranes were subjected to rehybridization with a
cDNA probe of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and
exposed to Fuji Photo Film Co., Ltd. x-ray film for 36–48
h at -80 C.
[125I]iodo-hCG binding
Highly purified hCG (CR-127, NIDDK, NIH, Bethesda, MD) was
radioiodinated with Na[125I]-iodide (IMS 300,
Amersham Pharmacia Biotech), using a solid phase
lactoperoxidase method (34). The specific activity of the
labeled hormone was found to be between 28–32 µCi/µg. The hCG
binding studies were carried out under optimized conditions, as
described previously (35). In brief, after washing, 3
x 105 cells were incubated with
[125I]iodo-hCG (
105
cpm/incubation) either in the absence (total) or presence (nonspecific)
of 50 IU of unlabeled hCG (Pregnyl, Organon Oss, The
Netherlands). The binding affinity was determined by Scatchard
analysis, by incubating similar aliquots of cell suspension with
increasing amounts of labeled hCG (1.5 to 60 x
104 cpm/tube) in the presence or absence of a
fixed concentration (50 IU/tube) of Pregnyl. The reaction was
terminated after overnight incubation by addition of 3 ml of ice-cold
Dulbecco’s-PBS containing 0.1% BSA. After centrifugation, the
supernatant was discarded by aspiration and the pellet was counted in a
-spectrometer (1260 Multigamma II, AGG Wallac, Inc.).
Plasmids, transfections, and determination of
luciferase activity
The murine LHR promoter/luciferase fusion genes were constructed
as described previously (36). Briefly, the 2040-bp
fragment of the 5'-flanking region (-1/-2040 bp), 173-bp deleted
fragment from 2040-bp (-174/-2040 bp) and 173-bp (-1/-173 bp)
fragment (in relation to translation initiation codon) of the mouse LHR
gene, were linked to the coding sequence of the luciferase reporter
construct (36, 37). The role of SF-1 mediated
T3 function was also assessed in mLTC-1 cells
using 2 µg of pBKCMV-hDAX-1 (obtained from Dr. R. Yu, Northwestern
University Medical School, Chicago, IL) expression plasmid. The mLTC-1
cells were transfected at 60–70% confluency by using the FuGENE 6
transfection reagent (Roche Molecular Biochemicals GmbH,
Mannheim, Germany). Two micrograms each of an LHR promoter construct,
and a pSV-ß-galactosidase expression vector (Promega Corp.), were used for transfection under optimized conditions
(17), following instructions of the manufacturer.
Thirty-six hours after transfection, cells were stimulated without or
with T3 for 8 h and assessed for either SF-1
binding (EMSA) or reporter assay (luciferase measurement). Luciferase
activity was determined from cell lysates by chemiluminescence
following addition of luciferin using the 1251 luminometer (BioOrbit,
Turku, Finland) (32). The corresponding ß-galactosidase
activity was measured from the same sample to determine the variation
in transfection efficiency.
The 5'-flanking regions of the mouse StAR gene were generated or obtained from Dr. B. J. Clark (Department of Biochemistry, University of Louisville School of Medicine, Louisville, KY). Full-length and deleted fragments of the 5'-StAR gene were placed upstream of the luciferase reporter gene into the pGL2 basic vector (Promega Corp.). The plasmids carrying mutations in the SF-1 sites (-135 and -42 bp) were generated by site-directed mutagenesis. Briefly, SF-1 sites at positions -135 and -42 bp were mutated from (CCAAGGTGG to TACGTAGTT) and (AGGCTG to TACGTA), respectively. For promoter analysis, mLTC-1 cells were cotransfected with 1 µg of plasmids and 1 µg of pRL-SV40 vector (a plasmid that constitutively expresses Renilla luciferase) to normalize the transfection efficiency. Following 36 h, cells were stimulated and luciferase activity was determined, as above.
Preparation of nuclear extracts and electrophoretic
mobility shift assays (EMSA)
The nuclear extracts (NE) from different experimental groups,
were prepared as described previously (36), with slight
modifications to improve their purity. Briefly, after washing the cells
twice with PBS, they were collected (107
cells/300 µl), and kept for 5 min in ice-cold lysis buffer A
containing protease inhibitors (20 mmol/liter HEPES, 20 mM
NaCl, 0.5% Nonidet P-40, 1 mM dithiothreitol, 1
mM phenylmethylsulfonyl fluoride, 2 mg/liter
leupeptin, 2 mg/liter aprotinin, pH 8.0). Following
centrifugation, the crude nuclear pellet was resuspended and allowed to
swell for 15 min at 4 C in 75 µl buffer C (buffer A containing 1.5
mM MgCl2, 0.2 mM EDTA,
20% glycerol, pH 7.9), followed by intermittent mixing of the
suspension for another 15 min. After removing the debris by
centrifugation at 12,000 x g for 5 min, the NE was
assayed directly or stored at -80 C.
The doubled-stranded DNA probes were engineered and synthesized from the mouse LHR promoter (37) sequences [consensus (con) and mutated (mut) bases in the SF-1 binding sites are underlined], by heating sense (s) and antisense (as) primers to 72 C for 5 min in annealing buffer (10 mM Tris-Cl, 100 mM NaCl, 1 mM EDTA, pH 7.5), followed by cooling at room temperature.
SF-1-cons 5'- GGGTGGCCACAGTTCAAGGTCAAGGAGAA-3'
SF-1-conas 5'- GGGTTCTCCTTGACCTTGAACTGTGGCCA-3'
SF-1-muts 5'- GGGTGGCCACAGTGTAATATCAAGGAGAA-3'
SF-1-mutas 5'- GGGTTCTCCTTGATATTACACTGTGGCCA-3'
The 5'-GGG overhangs present in the double-stranded oligonucleotides
(100 ng) were labeled with [32P]-dCTP (3000
Ci/mmol, Amersham Pharmacia Biotech) using Klenow
(Promega Corp.) fill-in reaction at 37 C for 1 h, and
probes were purified by Nick column (Amersham Pharmacia Biotech). Ten micrograms of NE were incubated in 20 µl
reaction buffer (12 mM HEPES, 60 mM KCl, 12%
glycerol, 4 mM Tris-Cl, 1 mM EDTA, 1
mM dithiothreitol, pH 7.9) containing 2 µg of poly
(dI-dC):poly (dI-dC). The NE protein was incubated at 4 C for 30–40
min in the presence of molar excess (0- to 10-fold) cold competitors
(unlabeled consensus SF-1 and mutated SF-1 nucleotides), followed by
additional incubation for 45–50 min with approximately 0.5 ng of
labeled probe. The reaction mixture (protein-DNA complexes) was then
subjected to electrophoresis in low ionic strength 0.25x TBE buffers
(90 mM Tris, 90 mM boric acid and 2
mM EDTA) on 5% polyacrylamide gels. The gels were then
vacuum dried, exposed to Fuji Photo Film Co., Ltd. x-ray
film and quantified.
Determination of pregnenolone, progesterone (P), and
testosterone (T)
The concentrations of pregnenolone, P and T in the media were
determined following extraction with diethyl ether using specific RIAs
as previously described (38, 39, 40).
Data analysis
The data presented are the mean ± SEM.
Statistical analysis was performed by one-way ANOVA, using the Statview
program (Abacus Concepts Inc., Berkeley, CA), followed by Fisher’s
protected least significant differences test (Fisher’s PLSD).
P < 0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
show
that acute (8 h) stimulation with T3 (37.5
pM) resulted in a 260 ± 15% increase in StAR mRNA
expression (determined by quantitative RT-PCR) over basal level,
whereas the elevated levels clearly declined following longer exposure
to T3 (Fig. 1A). The concentration of
T3 used in these experiments was based on our
previous observations (17). Similar results were obtained
when mitochondrial protein samples were measured for content of the
30-kDa StAR protein. In these samples, a 170 ± 10% increase in
immunoreactive StAR protein was seen following 8 h of
T3 treatment, and a subsequent decline occurred
thereafter up to 30 h (Fig. 1B). P concentration in the media
observed at 8 h was increased to 300 ± 18% by
T3, and thereafter declined, following StAR
expression. Compilation of the StAR mRNA, StAR protein and P data
demonstrated an intimate correlation on
T3-induced StAR expression and steroidogenesis
(Fig. 1C).
|
). A full-length mouse
StAR probe hybridized with two major transcripts at 3.4 and 1.6 kb, and
two minor ones at 2.7 and 1.4 kb. All transcripts were of sufficient
size to encode functional StAR protein, and showed coordinate
regulation (200–500%) by T3. P production in
those cells exhibited similar response, suggesting an intimate
association between StAR expression and steroid production (Fig. 2C).
|
). Subsequently, cells stimulated for
30 h with T3 displayed a >50% reduction in
P synthesis, when compared with 8 h stimulation. Incubation with
either 22R hydroxycholesterol (30 µM, a membrane-permeant
analog of cholesterol), or pregnenolone (20 µM,
Sigma), in combination with T3 for
8 h, additively (P < 0.01) increased P
production. When these incubations were extended to 30 h, a
partial restoration of the T3-mediated inhibition
of P synthesis was found, reaching 61 ± 4.6% and 72 ±
3.7% of the respective 8 h levels (Fig. 3A). These findings
suggest that the inhibitory effect of T3 occurs,
at least in part, before pregnenolone production. Another explanation
is increased metabolism of the P accumulated during the first 8 h
of culture between 8–30 h. The acute and chronic effects of
T3 on P synthesis were further assessed by
determining testosterone (T) production. In a similar experimental
paradigm, acute and chronic effects of T3 on T
production of mLTC-1 cells, though quantitatively much lower, were
qualitatively similar to those of P (data not shown).
|
4 and 5 steroidogenic
pathways, i.e. the 3ß-HSD inhibitor, trilostane (10
µM, Sanofi Pharmaceuticals, Inc.,
Malvern, PA) and the P450c17 inhibitor, SU 10603 (25
µM). As demonstrated in Fig. 3B
, mLTC-1 cells
stimulated with T3 for 8 h significantly
(P < 0.001) increased pregnenolone production, whereas
a marked decrease occurred in these levels at 30 h. Incubation of
cells pretreated for 30 h with T3 in the
presence of 22R hydroxycholesterol for 8 h significantly restored
(80 ± 5.3% of 8 h T3 stimulated)
pregnenolone synthesis. These data indicate that longer exposure of
T3 decreases cholesterol delivery to the
mitochondrial inner membrane and probably does so through repression of
the StAR protein.
Long-term effect of T3 on StAR mRNA stability and on
steroidogenic enzyme mRNA levels
Considering the effects of long-term T3
incubation, StAR mRNA stability was evaluated. Cells prestimulated in
the absence or presence of T3 (37.5
pM) for 10 h were further incubated with actinomycin D
(10 mg/liter) for up to 24 h. As shown in Fig. 4, in both transcriptionally arrested
control and T3-treated cells, the half-lives of
StAR mRNA were found to be approximately 11 h, indicating that
long-term exposure to T3 does not affect StAR
mRNA stability, and thus, the decreased StAR mRNA levels are a result
of decreased transcription.
|
) revealed that
stimulation with T3 (37.5
pM) for longer time periods resulted in an
83 ± 5.4% increase in P450scc mRNA. Conversely, chronic
T3 incubation (30 h) caused significant
(P < 0.01) decreases in P450c17 and 3ß-HSD mRNA
levels by 56 ± 3.5% and 34 ± 3.1%, respectively.
|
. [125I]iodo-hCG
binding was maximally increased at 16 h (228 ± 10.4% of
control) following incubation with T3. Longer
exposure to T3 markedly reduced the hCG binding
(46 ± 5.5% at 72 h) in a time-dependent manner when
compared with controls (Fig. 6A). Both the acute and chronic effects of
T3 on LHR mRNA expression were analyzed by
Northern hybridization. An LHR-specific probe (nucleotides 441–849)
revealed multiple transcripts of 6.9, 4.2, 2.6, and 1.8 kb in these
cells, and all of them were coordinately up-regulated by acute
T3 stimulation, whereas down-regulation of the
elevated levels occurred following longer treatment (Fig. 6B). Under
these incubation conditions, no significant changes were observed in
the hCG binding affinity (Kd) (data not shown).
The effects of T3 on LHR levels closely
correlated with the binding data, and strongly implicate the
involvement of thyroid hormone in the regulation of LHR gene
expression.
|
. Serum
T3 and T4 levels were
decreased with THU and increased with T4
treatments, respectively, providing evidence for the altered thyroid
status of the treated mice. Serum LH increased moderately but
consistently in hyperthyroid animals and decreased in the hypothyroid
group. There were no noticeable alterations in serum T levels; however,
serum PRL levels increased significantly following THU treatment when
compared with controls. No changes were observed in testis weights
(Table 1).
|
illustrate [125I]iodo-hCG binding to Leydig
cells isolated from testes of mice of the different treatment groups.
Interestingly, hypothyroidism caused a 175% elevation in binding (Fig. 7A). In contrast, T4 treatment significantly
decreased hCG binding to 56% of that seen in controls. Scatchard
analysis revealed that the affinity of hCG binding showed no major
changes with the mean Kd values being 4.47
± 0.11, 4.29 ± 0.26, and 4.52 ± 0.13 x
10-10 M in
hypothyroid, hyperthyroid and control groups, respectively (Fig. 7B).
|
). Northern blot analysis
employing an LHR specific probe confirmed the RT-PCR results, and
revealed that all of the LHR mRNA splice variants were coordinately
regulated (data not shown).
|
). Conversely, cells transfected with
173-bp deleted segment (-174/-2040 bp), which includes a nuclear SF-1
binding site, displayed reduced luciferase activity (66 ± 5.6%
of full-length promoter) and no response to T3
was observed. In contrast, the T3-mediated
luciferase activity increased to 136% of the basal level when mLTC-1
cells were transfected with the promoter fragment consisting of the
first 173-bp (-1/-173 bp) upstream of the translation start site,
indicating the presence of putative thyroid hormone binding motif(s) in
this basal promoter region.
|
, T3 treatment
significantly augmented an apparent dose-dependent increase in
protein-DNA complexes, using nuclear extracts prepared from mLTC-1
cells. Competition binding studies demonstrated that the binding to
protein-DNA complexes was effectively inhibited by increasing doses of
the homologous sequence (cold competitor). Importantly, mutation of
bases in the SF-1 consensus sequences abolished the binding competition
obtained with consensus SF-1 (Fig. 10). Interestingly, cells expressing
DAX-1 markedly suppressed the basal and
T3-stimulated formation of protein-SF-1 DNA
complexes, suggesting a relationship between T3
action and SF-1.
|
). Transient transfection of
mLTC-1 cells with a p-966StAR/Luc plasmid carrying a mutation of the
SF-1 site at position -135 bp, accounted for an approximately 40%
decrease of basal and T3-stimulated luciferase
activity compared with p-966StAR/Luc alone. Mutation at both SF-1 sites
(positions -135 and -42 bp) did not show further impairment of basal
and T3-induced luciferase activity. Although PGL2
control showed low stimulation of luciferase activity by
T3, it does not confound the conclusion
concerning functional significance of the -135 bp SF-1 site in StAR
promoter function. These results indicated the importance of the SF-1
site at position -135 bp, and further documented the specific
involvement of SF-1 in T3 function.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The key second messenger in trophic hormone stimulated steroidogenesis is cAMP which, in turn, triggers a regulatory cascade resulting in mobilization and delivery of cholesterol to the inner mitochondrial membrane where P450scc catalyzes its conversion to pregnenolone. The transport of cholesterol is the first, truly regulated and rate-limiting step in steroidogenesis, and it is mediated by the StAR protein (1, 2, 4, 5). Recently, mechanisms of StAR action have been evaluated by N- and C-terminal truncated forms, which demonstrate the importance of the C-terminus, and provide evidence that StAR acts on the outside of the mitochondria to transfer cholesterol (41, 42).
Our results show that acute induction of P production and StAR expression in mLTC-1 cells by T3 requires on-going RNA and protein synthesis, which are in general agreement with previous observations (2, 17, 43). Chronic T3 treatment, which can clearly be seen at 30 h, diminished the stimulatory responses on P synthesis and StAR expression obtained at 8 h. Inhibition of P production did not appear to be as pronounced as inhibition of StAR expression suggesting the possible involvement of StAR independent processes in P production. Consistent with our observations, transcriptional inhibition in MA-10 cells markedly affected synthesis of the StAR protein, whereas 10–20% of the steroidogenic potency of the cells occurred through a StAR-independent mechanism (44). In mLTC-1 cells, four StAR transcripts were observed, two major (3.4 and 1.6 kb) and two minor ones at 2.7 and 1.4 kb, all of which showed coordinate regulation following T3 stimulation (17). Our data also document that the decrease of StAR expression following T3 treatment was due to StAR gene transcription, rather than degradation of the message. The half-life of the StAR mRNA in transcriptionally arrested control and T3-treated cells was found to be approximately 11 h in both cases. Thus, the present findings clearly show that T3 regulates steroidogenesis through induction of the StAR protein that, in turn, is preceded by SF-1 expression. In an earlier report, we demonstrated that mLTC-1 cells expressing SF-1 coordinately augmented StAR expression and steroid production, whereas DAX-1 overexpression drastically decreased these responses (17). Consideration of our previous data led to the hypothesis that longer T3 stimulation may induce DAX-1 expression. DAX-1 has been reported to repress SF-1 function, but no clear-cut regulation of DAX-1 protein was observed in relation to acute or chronic T3 treatments (data not shown).
The present results confirm our earlier observations and extend them by
illustrating the mechanisms of chronic
T3-mediated attenuation of the acute stimulated
levels of StAR expression and steroidogenesis. This was clearly shown
in experiments in which addition of 22R hydroxycholesterol or
pregnenolone to the T3-inhibited cells
significantly reversed the inhibition of P synthesis. These results
were also confirmed by measuring pregnenolone production from 22R
hydroxycholesterol in the presence of the 3ß-HSD inhibitor trilostane
and the P450c17 inhibitor SU 10603. These inhibitors have recently been
demonstrated to be potent in inhibiting pregnenolone metabolism in
bovine adrenocortical cells (45). In addition, P
production was greatly reduced (90%) by these inhibitors in
luteinized primate granulosa cells (46). It is also
possible that longer exposure to T3 decreases the
low-density lipoprotein (LDL) receptors, which have been demonstrated
to be instrumental in cholesterol delivery to adrenocortical cells
(47).
Contrary to the inhibitory effect that chronic treatment with T3 had on steroidogenesis and StAR, P450scc mRNA levels increased significantly during T3 stimulation. These results clearly reinforce and support the data obtained in cultured rat or MA-10 cells, which indicated that treatment with either hCG, (Bu)2cAMP or forskolin increased the expression of P450scc mRNA and protein (48, 49). However, P450c17 and 3ß-HSD mRNAs displayed moderate but consistent decreases following longer exposure to T3. Previous studies also demonstrated that chronic stimulation with cAMP represses the levels of T-mediated stimulation of P450c17 and 3ß-HSD mRNA levels in cultured mouse Leydig cells (50). This possibility cannot be excluded in mLTC-1 cells, as they possess basal StAR expression and T production in contrast to MA-10 mouse Leydig tumor cells. On the other hand, increased levels of 3ß-HSD activity, protein synthesis, and mRNA levels were also reported in cultured rat Leydig cells in relation to LH, cAMP or forskolin stimulation (51).
The role of thyroid hormone in gonadal function has been well documented and is known to affect the development, growth, and function of essentially all organs and tissues and has been implicated in mammalian reproduction leading to gonadal abnormalities with impaired fertility (8, 9, 10, 11). The present findings provide interesting evidence that acute T3 treatment markedly elevates LHR mRNA expression and receptor binding, whereas chronic treatment dramatically decreased these levels. The biphasic response of thyroid hormone on LHR regulation demonstrated the antagonistic interrelationship between the thyroid and testis in gonadotropin action.
To address the physiological relevance of thyroid hormone action on LHR expression, we carried out experiments in hypo and hyperthyroid mice, where serum T3 and T4 levels decreased in the former and increased in the latter. Our present results with mouse serum hormone levels corroborate previous findings, which demonstrate that prepubertal hypothyroidism is related to decreased serum LH and FSH, normal or low T levels, and hypersecretion of PRL in most cases, together with increasing number of Leydig cells (15, 16, 21, 24, 52). The in vivo results provided evidence indicating the effect of hypothyroidism in the augmentation of LHR expression, whereas conversely, hyperthyroidism diminished these responses. These results may be interpreted as indicating that increased testicular LHR levels in hypothyroid mice sensitize this organ to LH. In contrast, decreased LHR levels in hyperthyroidism have an opposite effect, and the maintenance of normal T secretion requires elevated LH secretion. In agreement with this observation, previous studies have demonstrated an enlargement of the interstitium with increased numbers of Leydig cells in hypothyroid mice, whereas fibrous thickening of the interstitium with reduced Leydig cell numbers were observed with T4 treatment (not illustrated) (12, 14, 15). A discrepancy seems to exist with regards to in vitro and in vivo receptor function, possibly due to the repressor-like activity of thyroid hormone in receptor protein interaction. However, it should be taken into account that pituitary TSH levels may interact with LHR and inhibit in vivo receptor function, although other factors might also be involved.
The effects of thyroid hormone on adult testis have not been thoroughly investigated, possibly because of the findings that T4 action in vivo did not affect their oxygen consumption (53). Although thyroid hormone has nonnuclear actions, the major actions of T3 are thought to be mediated by nuclear T3 receptors (TRs) (54). TR is a member of the steroid/thyroid hormone receptor superfamily that binds to the hexameric motif AGGTCA, termed the nuclear receptor half-site, to activate gene transcription (55). The cloning and characterization of the 5'-flanking region of the mouse and rat LHR gene demonstrated that the basal promoter region (173 bp) is highly GC rich containing AP1 and several SP1 sites, which are involved in transcriptional regulation (36, 37, 56). The importance of this region was further confirmed, by expressing either the 173-bp fragment alone or its deletion from the full-length fragment (2040 bp) in mLTC-1 cells. Comparison with the 2040-bp fragment, in relation to T3-mediated luciferase activity indicated the presence of putative thyroid hormone response element(s) in this region. These results support our previous findings that regulation of T3-induced StAR expression and P production is dependent on SF-1, as inhibition of the latter by DAX-1 abolished the T3-mediated responses (17).
To evaluate the mechanisms involved in thyroid hormone action, and to further understand the role of SF-1, EMSA studies were carried out with T3-stimulated nuclear extracts using oligonucleotides corresponding to an SF-1 binding site. Our data suggest potential involvement of SF-1 in T3 function, as mutations in the SF-1 consensus sequences rendered it incapable of competing with protein-DNA complexes. The importance of SF-1 was also assessed in cells expressing DAX-1 in relation to T3. Further insight into these mechanisms was observed in studying the 5'-flanking analyses of the mouse StAR gene, which demonstrate that the proximal 966-bp fragment is sufficient for full promoter activity in transiently transfected mLTC-1 cells. Concerning the p-966-bp fragment, it is noteworthy that mutations in the bases of the SF-1 site at position -135 bp significantly decreased both basal and T3 stimulated StAR promoter activity. These data suggest that the -135-bp SF-1 element is important for basal StAR expression and not for the T3 response, which further supported previous findings with regard to the cAMP-dependent StAR promoter activity in MA-10 cells (43, 57). These results further strengthen the crucial involvement of SF-1 in thyroid hormone-regulated gonadal function.
Importantly, the 5'-flanking region of the SF-1 promoter contains an E box, a CAAT box, and SP1 elements that are required for steroidogenic-specific expression of the SF-1 gene in the adrenal gland and gonads (58). There is a conspicuous lack of thyroid hormone response elements (TREs) in the promoter sequences of SF-1 (58), in the human and mouse StAR (43, 59), and in murine and rat LHR (37, 56), suggesting the involvement of additional factor(s). Indeed, recent studies demonstrate interactive cooperation of SF-1 and C/EBPß (CCAAT/enhancer binding protein-ß) in the regulation of StAR gene transcription and steroidogenesis (57). In addition, the binding motif of the GATA-4 transcription factor, located at position -61/-66 in the mouse and human StAR promoters, is also involved in mediating the acute response to hormones (60).
Taken together, our data clearly demonstrate that thyroid hormone acutely increased StAR expression, steroidogenesis, and LHR levels, whereas these responses were suppressed with chronic treatment. The latter effects of T3 occurred partly due to diminished cholesterol delivery to the mitochondrial inner membrane. The precise nature of the TR isoform(s) in Leydig cells, and other factors involved in the thyroid hormone action on LHR regulation, will need to be addressed in further detail. Importantly, the present clear-cut in vitro and in vivo findings provide evidence for the role of thyroid hormones in LHR regulation in mouse Leydig cells, explaining further the pathophysiology of gonadal function in association with thyroid malfunction.
|
Acknowledgments |
|---|
-hydroxylase/C17–20 lyase cDNA (P450c17); Dr. A. J. W.
Hsueh, Stanford University School of Medicine (Stanford, CA) for the
rat LHR cDNA template; and Dr. R. Voutilainen, University of Helsinki
(Helsinki, Finland), for the P450scc probe. Thanks are due for the
generous gift obtained from Dr. R. Yu, Center for Endocrinology,
Metabolism and Molecular Medicine (Chicago, IL), for the pBKCMV-hDAX-1
expression plasmid, and to Dr. B. J. Clark, University of
Louisville School of Medicine (Louisville, KY), for the mouse StAR
promoter constructs. We gratefully acknowledge the gift of the 3ß-HSD
enzyme inhibitor (Trilostane) obtained from Sanofi Pharmaceuticals, Inc., Malvern, PA. The superb technical
assistance of Ms. Tarja Laiho, Nina Lehtimäki, and Riikka
Kytömaa is gratefully acknowledged. |
Footnotes |
|---|
Received April 18, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
