| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Zoology (E.S.W.N., B.K.C.C.), University of Hong Kong, Hong Kong; and the Department of Obstetrics and Gynecology (P.K.W.C., P.C.K.L.), University of British Columbia, Vancouver, British Columbia, Canada V6T 3V5
Address all correspondence and requests for reprints to: Dr. Billy K. C. Chow, Department of Zoology, University of Hong Kong, Pokfulam Road, Hong Kong. E-mail: bkcc{at}hkusua.hku.hk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T3–1 gonadotropes. Subsequent deletion
analysis indicated that the most proximal 173 bp within the first exon
of the gene, although not a promoter itself, contains a critical
regulatory element(s) essential for the basal expression of the hGnRHR
gene. The functional roles of the putative gonadotrope-specific
elements (GSE; consensus 5'-CTGA/TCCTTG-3')
residing at positions -5, -134, and -396 were studied by
site-directed mutagenesis, and it was found that only the mutation at
position -134 significantly reduced the promoter activity (80%
reduction; P < 0.05). The attenuation effect of
this GSE mutant was cell specific, as it was restricted to T3–1
cells, but not to COS-7 and human ovarian adenocarcinoma (SKOV-3)
cells. Competitive mobility shift assays using either T3–1 nuclear
extract or recombinant SF-1 protein clearly indicated that SF-1 is able
to interact specifically with this GSE element positioned at -134.
Using a SF-1 antibody that completely abrogated complex formation in
the gel shift assays, the involvement of endogenous nuclear SF-1 was
further evidenced. By competitive gel shift assays using oligoprimers
with 2-bp scanning mutations, the sequences essential for the
interaction with SF-1 were identified
(5'-TTGA/TCCCTG-3',
underlined sequences were important). To study the
in vivo function of SF-1, vector directing expression of
sense or antisense SF-1 messenger RNA (mRNA) was cotransfected with the
hGnRHR promoter-luciferase construct into T3–1, SKOV-3, and COS-7
cells. Overexpression of the SF-1 mRNA was able to enhance promoter
activities in all of the cells tested. On the contrary, expression of
the antisense SF-1 mRNA reduced the hGnRHR promoter activity only in
T3–1 cells, not in COS-7 or SKOV-3 cells. In summary, the data
reported here provide conclusive evidence that SF-1 interacts with the
GSE motif at position -134 within the first exon of the hGnRHR gene to
mediate its cell-specific expression. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The anterior pituitary provides an excellent model to study the
ontogeny and development of various pituitary cell types. For the
gonadotropes, the role of steriodogenic factor-1 (SF-1) and its
interaction with gonadotrope-specific element (GSE) in conferring
phenotypic expression in gonadotropes has received much attention
recently. This element was found to be essential for the expression
of the glycoprotein hormone -subunit (7), LHß (8), and mouse GnRHR
(9) in the pituitary, and the P450 (10, 11, 12, 13), aromatase and Mullerian
inhibiting substance (14) in steriodogenic tissues. In fact, the
presence of SF-1 at multiple levels of the
hypothalamic-pituitary-gonadal axis (15, 16, 17) further suggests the
regulatory role of SF-1 as a master protein directing its effect at
multiple levels of the reproductive axis. In the mouse GnRHR gene, it
has been shown that the proximal 500 bp relative to the ATG site with
two putative GSE motifs is crucial to the gonadotrope-specific promoter
activity (18, 5). More recently, the GSE site residing at -250 to
-232 was identified to be one of the cis-elements within a
tripartite tissue-specific enhancer to confer gonadotrope-specific
expression of the murine GnRHR gene (9, 19). Mutation of this site led
to a 58% reduction in promoter activity, whereas a novel
GnRHR-activating sequence (GRAS) and an activator protein-1 (AP-1)
motif within the enhancer also contributed, if not more importantly, to
the cell-specific expression. In addition, the impaired expression of
GnRHR in Ftz-F1 (SF-1 homolog)-disrupted mice (17) further supports the
hypothesis that SF-1 is the trans-acting
gonadotrope-specific mediator for the murine GnRHR gene.
In this report, we propose that SF-1 mediates the cell-specific expression of the hGnRHR gene at the pituitary level by interacting with putative GSEs within the first exon of the hGnRHR gene. We found that SF-1 interacts with only one of three putative GSE homologs at position -134 relative to the ATG site, and this interaction alone is largely responsible for conferring cell specificity. The functional importance of this GSE motif and the involvement of nuclear SF-1 were tested vigorously using various assays. It also seems that the mechanisms to regulate gonadotrope-specific expression in human and mouse GnRHR genes are different, possibly due to the structural differences between the two promoters.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T3–1 and COS-7 were grown in DMEM supplemented with 10% FBS
(Life Technologies, Gaithersburg, MD). SKOV-3 cells were
grown in medium 199 containing 10% FBS. All cells were incubated at 37
C with 5% CO2 in medium supplemented with 100 U/ml
penicillin G and 100 µg/ml streptomycin (Life Technologies).
Plasmids and DNA manipulations
Escherichia coli strains JM109 and DH5 were used
as the host strains for subcloning and sequencing. All DNA
manipulations were performed as previously described (20). A 2.3-kb
HindIII DNA fragment containing the 5'-flanking region of
the gene was obtained by PCR using the plasmid with the first exon and
the 5'-end of the hGnRHR gene as a template. In the PCR, a
HindIII site was introduced into the 5'-untranslated region
of the hGnRHR gene immediately 5' to the ATG start site using primers
MP-1 and MP-2 (Table 1
). This DNA
fragment was subcloned into the HindIII site of pGL2-basic
(Promega Corp., Madison, WI) in both forward and reverse
orientations (p2300-Luc F and p2300-Luc R). Deletion mutants,
p2200-Luc, p227-Luc, and p167-Luc, were constructed by digestion with
BglII or exonuclease III (Pharmacia LKB
Biotechnology, Piscataway, NJ). The construct p2200/-173-Luc was
generated by the deletion of a HpaI-HindIII
fragment from p2200-Luc. All mutants were identified by restriction
mapping and confirmed by DNA sequence analysis. For the SF-1 expression
vector, a 1.4-kb mouse SF-1 complementary DNA (cDNA) was obtained by
RT-PCR from messenger RNA (mRNA) prepared from T3–1 cells using
primers SF1–5 and SF1–3 (Table 1). After DNA sequence analysis, the
PCR fragment was subcloned into pRC-cytomegalovirus (pRC-CMV;
Invitrogen, San Diego, CA) in both forward and reverse
orientations (pRC-CMV/SF-1 and pRC-CMV/anti SF-1). Plasmid DNA for
transfection was prepared using the QIAGEN Midi Preps kit
(Qiagen, Valencia, CA). All enzymes and oligoprimers were
purchased from Life Technologies.
|
) and universal primers MP, T7, and MP-5 (Table 1). Mutation
was confirmed by DNA sequence analysis after the mutagenesis
reaction.
Transient transfection assay
One day before transfection, cells were plated at a density of
2 x 105 for COS-7, 5 x 105 for
T3–1, and 1.5 x 105 for SKOV-3 cells/35-mm well
(six-well plate, Costar, San Diego, CA). A mixture
containing 5 µg promoter-luciferase construct, 2.5 µg pRSV-ß-gal
(ß-gal, ß-galactosidase; RSV, Rous sarcoma virus), 20 µg
lipofectamine (Life Technologies), and 200 µl of
serum-free medium was prepared, and transfection was performed
following the manufacturer’s protocol. For the sense and antisense
coexpression experiments, 1, 2, or 5 µg pRC-CMV/SF-1,
pRC-CMV/anti-SF-1, or the control pRC-CMV were cotransfected with 5
µg p2200-Luc and 2.5 µg pRSV-ß-gal. After overnight incubation, 1
ml of the medium supplemented with 20% FBS was added. Cell lysate was
prepared 48 h later by washing the cells twice with ice-cold PBS
followed by the addition of 200 µl reporter lysis buffer
(Promega Corp.).
Luciferase and ß-galactosidase measurements
To assay for the luciferase activity, 100 µl luciferase
substrate solution (Promega Corp.) were automatically
injected into 20 µl cell lysate, and luciferase activity was measured
as light emission using a luminometer (Lumat LB9507, EG&G
Berthold, Bad Wildbad, Germany). ß-Galactosidase activity was
determined by incubating the cell lysate (100 µl) in 100
mM sodium phosphate buffer (pH 7.3), 1 mM
MgCl2, 50 mM ß-mercaptoethanol, and 0.7 mg/ml
o-nitrophenyl-ß-D-galactopyranoside for
15 min at 37 C, and the absorbance at 420 nm was measured using a
spectrophotometer (UV160A, Shimadzu, Columbia, MD). For each
transfection study, luciferase activity was determined and normalized
based on the ß-galactosidase activity.
Northern blot analysis
Total RNA was prepared from the cells 2 days after transfection.
Five or 15 µg of total RNA were size-fractionated by electrophoresis
in a denaturing formaldehyde gel, followed by transblotting and UV
cross-linking onto a Hybond N+ membrane
(Amersham, Arlington Heights, IL). A full-length SF-1 cDNA
was labeled by the RadPrime DNA labeling kit (Life Technologies) and [-32P]deoxy-ATP (3000
Ci/mmol). After overnight hybridization in the Rapid-Hyb buffer
(Amersham) at 65 C, the membrane was washed three times in
0.1 x SSC-0.1% SDS at 65 C and then exposed to BioMax film
(Eastman Kodak Co., Rochester, NY) for 16 h at -70
C. To serve as an internal control, the blot was stripped and reprobed
with 32P-labeled ß-actin cDNA.
Expression of the SF-1 as a glutathione-S-transferase (GST) fusion
protein
The SF-1 cDNA (1.4 kb) was directionally subcloned into the
NotI/EcoRI sites of the prokaryotic expression
vector pGEX 4T-3 (Pharmacia) in-frame with the N-terminal GST fusion
partner to produce pGST/SF-1. The plasmid was transformed into E.
coli strain BL21 (Pharmacia) for the production of
the recombinant SF-1 fusion protein. In summary, bacterial culture was
grown at 37 C with vigorous agitation to an OD600 reading
of 0.6 and was induced with 0.1 mM
isopropyl-ß-D-thiogalactopyranoside for 90 min.
Afterward, the cells were harvested by centrifugation at 2500 x
g for 5 min at 4 C using a Beckman Coulter, Inc. JA14 rotor (Palo Alto, CA). The pelleted cells were
resuspended in PBS and treated with 1 mg/ml lysozyme for 2 min at 4 C,
then lysed by vortexing in the presence of 1% Tween-20 and 1% Triton
X-100. Cell debris was removed by centrifugation at 15,000 x
g for 20 min at 4 C. Cell lysate was recovered and stored at
-70 C in the presence of 20% glycerol and 0.5
mM phenylmethylsuflonylfluoride. The protein
concentration of the lysate was determined by a Bradford protein assay
kit (Bio-Rad Laboratories, Inc., Richmond, CA).
Gel mobility shift assays
Double stranded oligonucleotide DNA was prepared by heating
complementary oligonucleotides at 95 C for 3 min in 10 mM
Tris (pH 8.0), 100 mM NaCl, and 1 mM EDTA,
followed by slow cooling to room temperature. The annealed probe was
end labeled by T4 polynucleotide kinase and [
-32P]ATP
(Amersham) with the Ready-To-Go T4 polynucleotide kinase
labeling kit (Pharmacia). Nuclear extract was prepared
from T3–1 (22), and a mobility shift assay was performed (23) with
1 µg poly(dI-C) and 0–10 µg nuclear extract or GST fusion
protein. The 16-bp oligo, GSE1 (corresponding to -1 to -16 of the
hGnRHR gene), and the mutants, GSE mut 1, 1a, and 1b (Table 1), were
used as the probes for the gel shift assay. In addition, the 24-bp GSE2
(Table 1) corresponding to position -127 to -150 of the hGnRHR gene
was used as both a probe and a competitor. For the competition assays,
a 100-fold molar excess of the unlabeled double stranded mutants, GSE
mut 2 and GSE mut 2a to 2h (Table 1), was used. In antibody abrogation
gel shifts, 6 µg nuclear extract were incubated with the rabbit
polyclonal antibodies directed against the DNA-binding domain of SF-1
(Upstate Biotechnology, Inc., Lake Placid, NY) for 30 min
at room temperature before addition of the radiolabeled probe. The
binding reaction was performed at room temperature for 20 min using 0.2
pmol probe (200,000 cpm). Free and bounded probes were separated by
electrophoresis for 2 h at 160 V in a 5% polyacrylamide gel in
0.5 x Tris-borate. After electrophoresis, the gel was
dried and exposed (Biomax MR film, Eastman Kodak Co.) for
16 h at -70 C with intensifiers (Amersham).
Statistical analysis
The promoter-luciferase construct was tested by three
independent transfection experiments within each study, and the study
was repeated two (n = 6) or three times (n = 9). When
appropriated and unless otherwise stated, the transfection data were
analyzed by either one- or two-way ANOVA followed by Tukey’s test
(Figs. 1 and 2) or Dunnett’s test (Fig. 7), with the
negative control (p2300-Luc R in Figs. 1 and 2, pRC-CMV in Fig. 7) as
the independent variable (24).
|
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T3–1 cellsT3–1 cell (25). This 2.3-kb fragment was found
to be transcriptionally active, as it could induce a 3.6 ±
0.2-fold increase in luciferase activity compared with the control
plasmids (p2300-Luc R and pGL2-basic; P < 0.05; Fig. 1). As the hGnRHR gene contains over 650 bp of 5'-UTR and is much
longer than those in mouse and rat GnRHR genes (5, 6), we next tested
whether this 5'-UTR is necessary for promoter activity. A series of 5'-
and 3'-deletion mutants was constructed (Fig. 1). Deletion of the most
proximal 173 bp (p2200/-173-Luc) resulted in an almost complete lost
of promoter function (5.7 ± 0.9% of promoter activity remained;
P < 0.05). To test whether this 173-bp region can act
as a stand-alone promoter, two 5'-deletion mutants (p227-Luc and
p167-Luc) were prepared, but it was found that these DNA fragments were
insufficient to support full promoter activity (15.4 ± 1.4% and
14.0 ± 0.8%, respectively; P < 0.05). Probably
this 173-bp DNA fragment, although not a promoter itself, contains a
critical regulatory element(s) essential for the basal expression of
the hGnRHR gene in gonadotropes.
A putative GSE homolog residing at -134 of the hGnRHR 5'-UTR is
essential for gene expression
Knowing that the proximal 173-bp fragment contains a
cis-acting element(s) to regulate hGnRHR expression, we have
identified three putative GSE homologs with core sequence
TG(A/T)CC (16) within the 5'-UTR of the
hGnRHR gene at positions -5 to -13, -134 to -142, and -396 to
-404 relative to the ATG start codon. Three site-directed mutants
corresponding to these putative GSE homologs were constructed. Mutation
of the putative GSE homolog at -134 resulted in a 79.9 ± 2.2%
(P < 0.05) reduction in promoter activity (Fig. 1),
suggesting that the drop in promoter activity in the p2200/-173-Luc
construct may be caused by the deletion of this motif. On the other
hand, no significant change in promoter activity was observed in the
other two mutants, indicating that these putative sites are not
functional in the hGnRHR promoter. To test whether this functional
-134 GSE motif is cell specific, the promoter activities of the
wild-type and mutant constructs were tested in T3–1, SKOV-3, and
COS-7 cells. Consistently, luciferase activity detected in T3–1
cells (3.3 ± 0.4-fold; P < 0.05) was higher than
that in the other two cell lines (1.7 ± 0.4-fold in SKOV-3 and
2.1 ± 0.3-fold in COS-7; P < 0.05). More
importantly, it was found that mutation at -134 attenuates the
expression of the reporter gene only in T3–1 cells, but not in the
other two cell types (Fig. 2; 80% reduction; P <
0.05). This finding strongly indicated that this functional GSE motif
is largely responsible for the gonadotrope-specific expression of the
gene. In addition, the p2200-luc construct was able to drive a robust
reporter gene expression in COS-7 and SKOV-3 cells, and this expression
was not affected by the mutation at -134. It appears that there are
more stringent requirements for expression in gonadotropes, or at least
in T3–1 cells. In summary, based on the functional assays, the GSE
homolog within the first exon of the hGnRHR gene at position -134 was
essential for promoter function, especially in determining
cell-specific expression.
A protein factor interacts specifically with the GSE homolog
residing at -134 of the hGnRHR gene
After we demonstrated the function of the -134 GSE homolog,
synthetic overlapping oligonucleotides (-127 to -150) containing this
GSE homolog were used in gel shift assays using the T3–1 nuclear
extract to show specific protein-DNA interaction. With an increasing
concentration of nuclear extract (0–10 µg protein), a dose-dependent
increase in the intensities of a single retarded band was observed
(Fig. 3, left). This
protein-DNA interaction was specific, as there was again a
dose-dependent reduction in the intensities of the shifted band when an
increasing amount of the unlabeled competitor probe (0- to 100-fold)
was added to the binding reaction (Fig. 3, middle). A 3-fold
molar excess of the unlabeled probe was already effective, whereas a
10-fold molar excess of the unlabeled probe was able to completely
abolish complex formation (Fig. 3, middle). In contrast,
even at a 100-fold molar excess, the mutant GSE homolog (Table 1, GSE
mut 2) failed to abrogate GSE-protein complex formation (Fig. 3, right). The sequence of the mutant GSE used in
this competition assay was
5'-TCAAATTTgaattCTGAGATACTT-3', with the putative GSE core
motif TGTCC replaced. Our data indicated that the binding
between the nuclear protein factor and the GSE motif is specific at or
around the putative GSE motif (the underlined region). We
next sought to define the precise sequence motif that is required for
binding to occur. In this study, competitor oligoprimers containing
successive two-base mutation of the functional GSE sequence were
prepared (mutants 2a to 2h; see Table 1). As shown in Fig. 4, mutants 2, 2a, 2c, 2d, 2e, 2g, and 2h
had markedly impaired abilities to compete for the GSE-protein complex.
On the other hand, the other mutants (mutants 2b and 2f) were effective
in abrogating the interaction between the wild-type GSE sequence and
the protein factor. This experiment clearly identified sequences within
the motif that are essential for the interaction with a T3–1
nuclear protein (5'
TTGA/TCCCTG-3';
underlined sequences were important).
|
|
) and other GSE homologs in
the promoter regions of the - and ß-subunits of gonadotropins were
different, and there was a two-bases deviation from the consensus
sequence (CTGA/TCCTTG;
Fig. 8). This observation prompted us to investigate the identity of
the protein factor that interacted with the -134 GSE homolog by
recombinant SF-1 protein and antibody abrogation gel shift assays. The
mouse SF-1 cDNA (1.4 kb) was initially cloned into the bacterial
expression vector pGEX 4T-3, and the recombinant GST/SF-1 fusion
protein was prepared. The ability of this fusion protein to interact
with the -134 GSE homolog was examined by gel shift assays. A single
retarded band similar in size to that seen in the previous assays was
observed using the bacterial extract (0–2 µg) containing the
GST/SF-1 fusion protein. On the other hand, no shifted band was found
when the control GST recombinant extract (0–2 µg) was used (Fig. 5). The presence of GST/SF-1 and GST
proteins in these bacterial extracts was confirmed by Western blot
analysis using an anti-GST antibody (data not shown). To further
identify the protein factor, antibody abrogation gel shift was
conducted using a rabbit polyclonal antibody raised against the
DNA-binding domain of the murine SF-1 protein. The addition of 1.2,
2.4, and 4.8 µg of the anti-SF-1 antibody consistently abrogated the
complex formation in a dose-dependent manner (Fig. 6). This anti-SF-1 antibody has been
shown previously to be able to block mouse SF-1 binding to the DNA
sequence (13); hence, the result presented here strongly indicated that
the protein factor within the DNA-protein complex in previous gel shift
assays is SF-1.
|
|
|
T3–1 cells. Overexpression
of SF-1 mRNA (2 µg pRC-CMV/SF-1/plate; Fig. 7A) in SKOV-3, COS-7, and T3–1 cells
significantly (P < 0.01) up-regulated the activity of
the wild-type hGnRHR promoter (p2200-Luc; Fig. 7) by 77 ± 13%,
160 ± 31%, and 150 ± 16%, respectively. In contrast,
repression of the endogenous SF-1 by overexpressing the antisense SF-1
mRNA (pRC-CMV/anti SF-1; Fig. 7A) resulted in a significant
(P < 0.01) down-regulation of promoter function only
in T3–1 cells (29.2 ± 3.4% or 60.9 ± 4.8% using 2 or
5 µg, respectively), not in SKOV-3 and COS-7 cells. The lack of
effect of the antisense SF-1 transcripts in SKOV-3 and COS-7 is due to
the fact that SF-1 is not expressed in these cells (Fig. 7B). Elevated
expression of both sense and antisense SF-1 transcripts after
transfection in all of the cells tested was confirmed by Northern blot
analysis using the full-length SF-1 cDNA as a probe. Interestingly, for
COS-7 cells, the hybridization signal of the antisense transcripts was
consistently (in three independent experiments) much lower than that of
the sense transcripts. It is possible that the antisense transcripts
were less stable, particularly in COS-7 cells. In summary, the
coincident changes in hGnRHR promoter activities with the expression
levels of SF-1 in gonadotropes and other cell types further supported
the hypothesis that SF-1 is a key in vivo factor to regulate
hGnRHR gene expression.
An anonymous protein of a higher mol wt can interact with DNA
sequence adjacent to the -5 GSE motif
Although the proximal GSE homolog at position -5 is not
functional, it shares a high degree of sequence identity with the GSE
consensus (Fig. 8). A gel shift assay was
performed to investigate whether this GSE motif can interact with SF-1.
A retarded band was observed, but the protein was not SF-1 because the
complex was of a higher mol wt, and the recombinant GST/SF-1 fusion
protein was unable to interact with the GSE1 probe (Table 1). As
confirmed by scanning mutation coupled to gel shift assay, the
sequences within this probe that are required for interacting with the
anonymous protein were at the 5'-region of the GSE consensus core
(-16-GCTCTGTCCTGGGAAA-1, the binding motif overlaps with
the underlined region; Fig. 9, right panel). Mutation of the probe from
-16-GCTC-13 to -16-atct--13
resulted in the loss of the retarded band as shown in Fig. 9
(right panel).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit, and LHß are the phenotypic
markers of pituitary gonadotropes. It seems that the
gonadotrope-specific expressions of these genes are differentially
regulated by a combination of various cis-acting elements
and, hence, transcription factors. For example, in the -subunit
glycoprotein, several cis-acting motifs are involved, such
as the pituitary glycoprotein basal element (26, 27), the
-ACT element (28), several E boxes (29), and the GSE (30, 16). In the mouse GnRHR gene, a tripartite gonadotrope-specific
enhancer, including an AP-1 site, a GRAS, and a GSE present in the
promoter region, was found to be crucial to expression (19). Similarly,
the LHß promoter contains multiple tissue-specific elements.
Nevertheless, GSE remains as the only element that is found to be
responsible for the regulation of all of these phenotypic markers of
gonadotropes, and it is likely that the GSE-binding factor, SF-1, is
the key mediator for the gonadotrope differentiation and
development. In light of the central position of GnRHR in human reproductive functions and the presence of SF-1 in the human pituitary gonadotropes (15, 31), we sought to investigate the mechanisms underlining the role of SF-1 in hGnRHR go-nadotrope-specific expression. To study hGnRHR gene regulation, 2.3 kb of the 5'-flanking region were previously characterized in our laboratory (3). Compared with the mouse GnRHR gene, the hGnRHR gene contains a much longer 5'-UTR (>650 bp in human and <200 bp in mouse and rat). The difference in the structure of the human and mouse 5'-flanking regions suggests that they can be regulated by different mechanisms. The 5'-UTR of the mouse gene does not seem to play a functional role in gene expression (5). On the other hand, the human 5'-UTR, i.e. the first exon, is essential in determining the basal and gonadotrope-specific expression of the hGnRHR gene. Within the 5'-UTR, deletion of the most proximal 173 bp almost completely abolished the transcriptional activity of the promoter. However, this 173-bp DNA fragment, when tested alone, retained only a minimal basal activity. Analysis of this 173-bp DNA fragment revealed that there are two putative sites homologous to GSE, and by site-directed mutation, only the GSE homolog at -134 was found to be functional. Mutation of this site resulted in a dramatic drop (20% activity remains) in promoter strength, whereas mutation of other putative GSE sites (GSE mut 1 and mut 3) had no effect on promoter activity. However, the mutation of the GSE homolog at position -134 can account for most, but not all, of the loss in promoter activity of the 3'-deletion mutant (p2200/-173-luc; 6% activity remains). Our result suggests that in addition to this functional -134 GSE homolog, there could be another cis-element(s) present in this 173-bp DNA fragment that also contributes to regulate hGnRHR gene expression.
The function of the -134 GSE homolog as a cell-specific regulator was
illustrated by the restricted effect of the GSE mut 2 in T3–1
cells, but not in COS-7 or SKOV-3 cells. It appears that there are more
stringent requirements for expression in gonadotropes, and the
regulation of GnRHR expression should require the interplay between a
cell-specific regulatory factor(s) and a general transcription
factor(s) (32). A ubiquitous transcription factor, AP-1, and possibly
two other tissue-specific regulatory proteins, SF-1 and GRAS-binding
protein, were found to collectively regulate the gonadotrope-specific
expression of the mouse GnRHR gene (19). Nevertheless, it is unlikely
that human and mouse GnRHR genes share the same set of modules in
regulating gonadotrope-specific expression, because neither the
GRAS-like element nor the AP-1 site is found in the vicinity of this
functional GSE motif. It is possible that the SF-1 that binds to the
hGnRHR gene interacts with a totally different set of protein factors
to confer tissue-specific expression. The ability of SF-1 to interact
with a variety of protein factors was demonstrated in the human
steroidogenic acute regulatory protein gene. The two GSEs function in
different manners by interacting with two sets of transcription factors
(33). Although we still do not know the target protein for SF-1 to
trans-activate gene expression, the in vivo
effect of SF-1 on regulating hGnRHR gene expression was studied. There
was a 150% increase in hGnRHR promoter activity produced by
overexpressing SF-1 mRNA and presumably SF-1 protein in T3–1 cells.
This enhancement was also observed in COS-7 and SKOV-3 cells. In
contrast, overexpression of the antisense SF-1 mRNA leading to the
reduction of endogenous SF-1 protein expression alleviated hGnRHR
promoter function by 61%. This study clearly indicates the in
vivo effect of SF-1 to mediate gonadotrope-specific
expression.
It is of particular interest that the GSE motif residing in the 5'-UTR within the first exon has a significant contribution to the gonadotrope-specific expression of hGnRHR. However, this finding is not unique to the hGnRHR gene. Many housekeeping and growth control genes also require both upstream and downstream elements for promoter activity (34). In addition to transcription initiation, downstream elements may influence processes such as RNA elongation, processing, and translation. Moreover, intragenic enhancers or activators have been described for numerous genes, including the rat insulin-like growth factor I gene (35), the human antithrombin gene (36), the human tissue inhibitor of metalloproteinases 1 gene (37), and human and porcine choline acetyl-transferase genes (38).
Comparison of the GSE sites in the 5'-UTR of the hGnRHR gene with those
found in other gonadotrope-specific genes reveals that the
human GSE homologs have a one- to three-base deviation from the
consensus sequence (CTGA/TCCTTG; Fig. 8). The
putative GSE site locating at -396 (CTGTCCaac) has
three-base substitutions, which is probably the reason why this motif
is not functional (GSE mut 3) and is not capable of binding to SF-1
(data not shown). The proximal GSE homolog at -5
(CTGTCCTgG) has only a single base substitution from T to
g. However, this -5 motif is still nonfunctional and is unable to
interact with recombinant or nuclear SF-1. On the other hand, the -134
GSE homolog (tTGTCCcTG) contains two-base
substitutions. Although both residues are conserved in other functional
GSE homologs, this -134 motif is still transcriptionally active and is
able to interact specifically with the nuclear SF-1 as well as the
GST/SF-1 recombinant protein, as indicated in the competitive and
antibody abrogation gel shifts. The selective binding of SF-1 with the
-134, but not with the -5, GSE homolog raised the issue of what are
the sequences within the consensus that contribute most to SF-1
interaction. The competitive gel shift assays indicated that the most
important sequences within this 9-bp binding motif are the CC
dinucleotide in the center and the flanking sequences T and G
(5'-TTGA/TCCCTG-3').
As both the -5 and -134 motifs contain all of the sequences crucial
for SF-1 interaction, the present data suggest that in addition to the
sequence, the spatial arrangement of GSE in relation to other
cis-acting motifs may also contribute to confer its
regulatory function. A similar observation was made in the mouse GnRHR
gene, in which the proximal GSE homolog immediately 5' to the ATG start
codon is also nonfunctional (5). By competitive gel shift studies using
T3–1 nuclear extract and the proximal GSE motif as a probe, an
unidentified protein of a higher mol wt was found to interact
specifically with the sequence adjacent to the -5 proximal GSE motif.
As an alternative explanation for the lack of function for this GSE
homolog, this anonymous protein may block the entry of SF-1 at this
position, which may serve as a regulatory mechanism to control the
promoter activity of the hGnRHR.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received May 18, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
subunit gene to pituitary and placenta. J Biol Chem 270:26497–26504-subunit gene in pituitary gonadotropes. Mol Endocrinol 8:878–885[Abstract]
-subunit gene. Mol Endocrinol 6:893–903[Abstract]
-subunit promoter by a
LIM-homeodomain transcription factor. Mol Cell Biol 14:2985–2993-subunit gene in the
placenta and pituitary gland. Mol Cell Biol 14:5592–5602-glycoprotein hormone subunit gene in
pituitary cells. Mol Endocrinol 9:278–291[Abstract]
-subunit gene is regulated by a gonadotrope-specific
protein. Mol Cell Biol 12:2143–2153This article has been cited by other articles:
 |
|
 |
|
  Y. Shima, M. Zubair, T. Komatsu, S. Oka, C. Yokoyama, T. Tachibana, T. A. Hjalt, J. Drouin, and K.-i. Morohashi Pituitary Homeobox 2 Regulates Adrenal4 Binding Protein/Steroidogenic Factor-1 Gene Transcription in the Pituitary Gonadotrope through Interaction with the Intronic Enhancer Mol. Endocrinol., July 1, 2008; 22(7): 1633 - 1646. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Faurholm, S. Cochrane, R. R Millar, and A. A Katz Gene structure and promoter functional analysis of the marmoset type II GnRH receptor J. Mol. Endocrinol., August 1, 2007; 39(2): 91 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-Y. Kim, K.-C. Choi, N. Auersperg, and P. C K Leung Mechanism of gonadotropin-releasing hormone (GnRH)-I and -II-induced cell growth inhibition in ovarian cancer cells: role of the GnRH-I receptor and protein kinase C pathway. Endocr. Relat. Cancer, March 1, 2006; 13(1): 211 - 220. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-M. Yeung, B.-S. An, C. K. Cheng, B. K.C. Chow, and P. C.K. Leung Expression and transcriptional regulation of the GnRH receptor gene in human neuronal cells Mol. Hum. Reprod., November 1, 2005; 11(11): 837 - 842. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Karpova, J. Presley, R. R. Manimaran, S. P. Scherrer, L. Tejada, K. R. Peterson, and L. L. Heckert A Ftz-F1-Containing Yeast Artificial Chromosome Recapitulates Expression of Steroidogenic Factor 1 in Vivo Mol. Endocrinol., October 1, 2005; 19(10): 2549 - 2563. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Cheng and P. C. K. Leung Molecular Biology of Gonadotropin-Releasing Hormone (GnRH)-I, GnRH-II, and Their Receptors in Humans Endocr. Rev., April 1, 2005; 26(2): 283 - 306. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B.-S. An, J.-H. Choi, K.-C. Choi, and P. C. K. Leung Differential Role of Progesterone Receptor Isoforms in the Transcriptional Regulation of Human Gonadotropin-Releasing Hormone I (GnRH I) Receptor, GnRH I, and GnRH II J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 1106 - 1113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. T.-O. Lee, K.-C. Tan-Un, R. T.-K. Pang, D. T.-W. Lam, and B. K.-C. Chow Regulation of the Human Secretin Gene Is Controlled by the Combined Effects of CpG Methylation, Sp1/Sp3 Ratio, and the E-Box Element Mol. Endocrinol., July 1, 2004; 18(7): 1740 - 1755. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Granger, V. Ngo-Muller, C. Bleux, C. Guigon, H. Pincas, S. Magre, D. Daegelen, A. Tixier-Vidal, R. Counis, and J.-N. Laverriere The Promoter of the Rat Gonadotropin-Releasing Hormone Receptor Gene Directs the Expression of the Human Placental Alkaline Phosphatase Reporter Gene in Gonadotrope Cells in the Anterior Pituitary Gland as well as in Multiple Extrapituitary Tissues Endocrinology, February 1, 2004; 145(2): 983 - 993. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Cheng, B. K. C. Chow, and P. C. K. Leung An Activator Protein 1-Like Motif Mediates 17{beta}-Estradiol Repression of Gonadotropin-Releasing Hormone Receptor Promoter via an Estrogen Receptor {alpha}-Dependent Mechanism in Ovarian and Breast Cancer Cells Mol. Endocrinol., December 1, 2003; 17(12): 2613 - 2629. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. B. R. Jacobs, D. Coss, S. M. McGillivray, and P. L. Mellon Nuclear Factor Y and Steroidogenic Factor 1 Physically and Functionally Interact to Contribute to Cell-Specific Expression of the Mouse Follicle-Stimulating Hormone-{beta} Gene Mol. Endocrinol., August 1, 2003; 17(8): 1470 - 1483. [Abstract] [Full Text] [PDF]
|
|
