Endocrinology Vol. 138, No. 7 2754-2762
Copyright © 1997 by The Endocrine Society
Characterization of Multiple Promoters Directing Tissue-Specific Expression of the Human Gonadotropin-Releasing Hormone Gene1
Ke-Wen Dong,
Kei-Li Yu,
Zheng-Guang Chen,
Ya-Di Chen and
James L. Roberts
Jones Institute for Reproductive Medicine (K.-W.D., Z.-G.C.),
Eastern Virginia Medial School, Norfolk, Virginia 23507; Department of
Zoology (K.-L.Y., Y.-D.C.), The University of Hong Kong, Pokfulam Road,
Hong Kong; The Dr. Arthur M. Fishberg Research Center for Neurobiology
(J.L.R.), Mount Sinai School of Medicine, New York, New York 10029
Address all correspondence and requests for reprints to: Ke-Wen Dong, Ph.D., Eastern Virginia Medical School, Jones Institute for Reproductive Medicine, 601 Colley Avenue, Norfolk, Virginia 23507.
 |
Abstract
|
|---|
Two promoters directing tissue-specific expression of GnRH gene in
neuronal and reproductive tissues were characterized by functional
analyses of GnRH promoter-luciferase (LUC) constructs in transfected
placental cells (JEG) and hypothalamic neuronal cells (GT1–7). Results
indicate that the downstream promoter directs the expression in a
neuronal cell-specific manner, whereas the upstream promoter is fully
active in the nonneural placental cell line. Transfection studies
carried out in several tumor cell lines derived from human reproductive
tissues verified that the upstream GnRH promoter construct was much
more active in directing luciferase expression in reproductive tissue.
The use of both upstream and downstream promoters in various human
tumor cell lines derived from reproductive tissues were demonstrated by
RT-PCR. Our studies also demonstrate that the reproductive
tissue-specific messenger RNA transcribed from upstream promoter is
capable of directing synthesis of preproGnRH protein.
Serial deletion studies localized a cell-specific upstream promoter
region that directs reproductive tissue expression. DNase I footprint
analysis using nuclear extract obtained from the JEG cells indicated
DNA/protein interactions in four specific sequence elements of the
upstream promoter region. The interaction between nuclear binding
proteins present in the JEG cells (but not the GT1–7 cells) and the
four specific sequences in the upstream promoter region was confirmed
by gel mobility shift analysis.
|
Introduction
|
|---|
GnRH PLAYS an integral role in the
regulation of reproductive processes. It is synthesized in hypothalamus
and secreted in a pulsatile manner into hypothalamic-pituitary portal
blood system, where it reaches the anterior pituitary and acts to
elicit the episodic release of gonadotropin. In mammals, GnRH is
derived from a 10-kDa precursor protein encoded by a single copy gene
comprised of four exons and three introns in rodents (1, 2, 3, 4) and humans
hypothalamus (5, 6). Initially, it was originally thought that GnRH was
produced exclusively by hypothalamic neurons, although GnRH and GnRH
receptor were observed in a variety of reproductive tissues (7). In
addition to its well-documented role on pituitary LH and FSH secretion
(1, 2, 3), GnRH has also been implicated in reproductive functions of
gonads and placenta (8). In the ovary, GnRH is involved in the
regulation of steroidogenesis (9). Because of the important functions
of GnRH in reproductive tissues and the hypothalamus, we isolated the
human GnRH (hGnRH) gene promoter and 5' flanking region so as to study
its regulation. In both JEG-3 and human mammary gland tumor cell lines
(MDA), we previously identified a discrete upstream transcriptional
start site 579 bp upstream from transcription start site used in the
hypothalamus (10). RT-PCR quantitation demonstrated that the upstream
start site is the major transcriptional start site, representing about
70% of the cytoplasmic transcripts in JEG-3 and MDA cells
In the present investigation, we further analyzed the human GnRH
multiple promoters for tissue or cell-specific expression. The upstream
promoter region was analyzed for its contribution to reproductive
tissue-specific expression in JEG-3 cells. The utilization of these two
promoters in a number of tumor cell lines derived from reproductive
tissues including mammary gland (MCF-7), ovary (PA-1), and prostate
gland (LNCaP) was also demonstrated. To study the utilization of these
promoters in neuronal cells, a mouse immortalized hypothalamic neuronal
cell line (GT1–7) that expresses mouse GnRH gene (11) was employed for
this study. To demonstrate that the reproductive tissue-specific
promoter could produce a functional GnRH molecule, we performed a
cell-free translation analysis and identified that reproductive
tissue-specific messenger RNA (mRNA) is capable of directing synthesis
of the preproGnRH protein.
|
Materials and Methods
|
|---|
Construction of hGnRH promoter-reporter deletion constructs
A 1.7-kb of Xha1/AflII fragment containing
the upstream transcriptional start site of human GnRH gene and a 0.65
kb of BamHI/Sph1 fragment containing the
downstream transcription start site of human GnRH gene were ligated to
a promoterless luciferase reporter plasmid, pxp2-Luc. Fragments of 5'
flanking region of the human GnRH gene were prepared using convenient
endonuclease restriction sites and were inserted into promoterless
pxp2-Luc vector. The positive clones were identified by restriction
mapping and southern blot analysis. The bacteria containing the
positive clone were grown in Luria-Bertani medium overnight and
harvested by centrifugation. Plasmid DNA was prepared by Wizard Maxi
Preps DNA purification system (Promega, Madison, WI).
Cell culture and transfection assay
The human JEG-3, MCF-7, PA1, and LNCaP cells were purchased from
American Type Culture Collection (Rockville, MD). All cell lines were
grown to 50–60% confluence, and the media of each plate replaced
12 h before transfection. The calcium-phosphate method of
transfections was carried out in triplicate 60 mm plates using 5 µg
of test plasmid per plate (12). After 12 to 14 h of incubation,
cells were washed three times with ice-cold 1x PBS followed by
addition of DMEM containing 5% FBS and 10% horse serum. Cells were
harvested and luciferase activity was measured as described by
Wondisford et al. (13). Protein content of the cellular
extract was determined by Bradford method (Bio-Rad, Hercules, CA) and
used to normalize the luciferase activities. To correct for the
different transfection efficiencies of the various luciferase
constructs, a pCMV ß-galactosidase construct (GUS) was cotransfected
into cells with each GnRH promoter-luciferase construct. Portion of the
harvested cell extract (10%) was used to detect ß-galactosidase
activity based on the conversion of
4-methyl-ambellifery-ß-D-galactoside to the highly
fluorescent molecule methylambelliferone. A promoterless pxp2-Luc
vector was used as a negative control for each transfection
analysis.
RNA isolation
Human placenta and various human tumor cell lines derived from
ovary (PA-1), prostate gland (LNCaP), mammary gland (MCF-7), and
placenta (JEG) were used as sources of RNA. All cell lines were grown
in DMEM supplemented with 5% FBS and 10% horse serum. Cells were
grown to 50–60% confluence before RNA extraction. For extraction of
RNA, plates were rinsed once in cold 1x PBS and cells removed by
scrapping. After centrifugation, the cells were resuspended in 500 µl
of ice-cold lysis buffer (10 mM Tris-HCl pH 7.5, 1.5
mM MgCl2, 0.3 M sucrose, 0.15%
Triton X-100) and triturated 10 times. Isolation of heteronuclear and
cytoplasmic RNA from cells was performed as previously described by
Jakubowski et al. (14).
RT-PCR
Approximate 10 µg of total RNA was hybridized with oligo
(dT)12–18 in a reverse transcription buffer (Life
Technologies, Gaithersburg, MD) and the reaction was carried out using
MMLV reverse transcriptase (Life Technologies) for 2 h at 37 C.
The reaction was terminated by heating for 15 min at 68 C, and the
reaction mixture was diluted to a final volume of 100 µl. Ten
microliters of the reverse transcription reaction was added into a
final volume of 50 µl in 10 mM Tris-HCl, pH 8.4, 2.5
mM MgCl2, 250 µM (dATP, dGTP,
dCTP, and dTTP), 0.5 µg each of sense and antisense strand primer and
2.5 U Taq polymerase (Perkin-Elmer Cetus, Foster City, CA).
The polymerase amplification was carried out for 30 cycles using a 94 C
denaturing cycle (1 min), a 55 C annealing cycle (30 sec) and a 72 C
extension cycle for 2 min followed by a final extension for 10 min. The
PCR products were then visualized by electrophoresis in an agarose gel
with ethidium bromide.
Cell-free translation of in vitro transcribed RNAs
The two human GnRH complementary DNA (cDNA) clones were used for
coupled in vitro translation according to instructions
provided by TNT coupled Reticulocyte Lysate Systems (Promega). The
in vitro translation products were labeled with
35S-methionine and separated on a 10% SDS polyacrylamide
gel. Radioactivity in the 10-kDa proGnRH band was quantitated using a
Molecular Dynamic Phosphorimager (Image Quant Software Program).
Nuclear extract and DNase footprinting
The JEG and GT1–7 cells were grown in 100-mm plates to 80%
confluence and then harvested by trypsinization. Cells were then
resuspended in five packed cell volumes of ice-cold buffer A [10
mM HEPES-NaOH, pH 7.9, 10 mM KCl, 1.5
mM MgCl2, 1 mM DTT, 50 µg/ml
phenyl-methyl-sulfonylfluoride (PMSF)] and incubated on ice for 10
min. The cells were spun down again and resuspended in ice-cold buffer
A. Cells were homogenized in a Dounce homogenizer and briefly pelleted
at 40 C for 10 min. The nuclei were separated from the cytoplasmic
supernatant by centrifugation in a SS-34 rotor, 12,500 rpm at 4 C
for 25 min, resuspended in 0.25 packed cell volume of ice-cold buffer C
(20 mM HEPES-NaOH, pH 7.9, 10 mM KCI, 1.5
mM MgCI2, 1 mM DTT, 50 µg/ml
PMSF, 0.25 mM EDTA, and 25% glycerol) and homogenized in
Dounce homogenizer. After rocking at 4 C for 30 min, the nuclear
extract was spun down and the supernatant was dialyzed against 150 vol
buffer D (20 mM HEPES-NaOH, pH 7.9, 100 mM KCI,
1 mM DTT, 50 µg/ml PMSF, 0.2 mM EDTA, 20%
glycerol). The BamHI/HindIII fragment containing
the 5'-flanking region of human GnRH gene was labeled by [gamma-
32P] ATP using T4 polynucleotide kinase. After digestion
with AflII, the labeled fragment of 325 bp, containing the
GnRH promoter sequence, was purified by 5% PAGE and eluted at 37 C
overnight with ammonium acetate (0.5 M). DNA-binding was
performed in 50 µl of mixture containing 1–2 fmol (1–3 x
104 cpm) DNA probe, 10–40 µg nuclear extract, 50
mM KCl, 1 mM MgCl2, 1
mM CaCl2, 10% glycerol, 20 mM
Tris, pH 8.0, 1 mM DDT, 1 mM MC, 1
mM EDTA and 1 µg of poly(dI-dC), for 30 min at room
temperature. DNase I (2 U) was used for partial digestion of the
radiolabeled template for 2 min. at room temperature, and the reaction
was terminated by adding 100 µl of DNase stop solution containing 32
mM EDTA, 0.14% SDS and 6 µg of yeast RNA. The reaction
mixture was extracted with phenol and chloroform, precipitated with
ethanol and resolved on a 6% denaturing polyacrylamide gel.
Gel-shift analysis
Double stranded oligonucleotides were designed according to the
nucleotide sequence of the upstream GnRH promoter region that exhibited
a clear DNase I foot print and labeled with [gamma-32P]
ATP. One microliter of nuclear extract (5 µg/µl) was preincubated
with and without competitor DNA in a 20 µl reaction containing 1x
gel-shift buffer (10 mM Tris-HCL, pH 8; 5% glycerol; 50
mM KCL; 1 mM EDTA and 1 mM DTT), 4
µg of poly (dI-dC) at room temperature for 20 min. One nanogram of
32P-end labeled probe was added and incubated for an
additional 10 min at room temperature, and then the reaction was
separated on a polyacrylamide gel at 4 C.
|
Results
|
|---|
Functional characterization of multiple human GnRH promoters by
transfection assay
Functional transfection studies were performed to determine
whether the two GnRH promoters could direct the expression of a
luciferase reporter gene in a cell-specific fashion. We tested three
reporter gene constructs (Fig. 1B
). The first construct
contains approximately 2.0 kb of full-length human GnRH 5'-flanking
region (XbaI/SphI fragment) containing both the
upstream as well as the downstream promoters. The second construct
contains a 1.7 kb XbaI/AflII fragment including
only the upstream transcriptional start site of human GnRH gene. The
third construct contains a 0.65 kb BamHI/Sph1
fragment that includes only the downstream transcription start site. As
shown in Fig. 1A, the upstream promoter-Luc construct was much more
active than the downstream promoter-Luc construct in JEG cells. In
contrast, GT1–7 cells exclusively utilize the downstream promoter,
suggesting that tissue-specific expression of the GnRH gene is mediated
by a different promoter. The full length GnRH promoter-Luc construct
(hU/D) displayed a luciferase activity similar to that of hD construct
in GT1–7 cells. Furthermore, this construct demonstrated the similar
luciferase activity as hU construct in JEG-3 cells. These results
suggest that both hU and hD constructs are fully functional in
controlling GnRH promoters activity.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. Basal promoter activity of the human GnRH gene in
placental cells and hypothalamic cells. A, Each vector in B was
transfected into JEG and GT1–7 cells by the calcium phosphate method,
and luciferase activity was measured in cell extracts after 16–18 h.
To correct for the different transfection efficiencies of various
luciferase constructs and different cell lines, a pCMV
ß-galactosidase (pCMV-ß-GUS) construct was cotransfected into the
cells with each luciferase construct and used as an internal control.
B, Structure of full-length human GnRH promoter region (hU/D) as well
as two human GnRH promoter regions (hU and hD) inserted into
promoterless pxp2-luc vectors. The hU construct was generated by
ligating 1687 bp XbaI/AflII fragment
containing only upstream GnRH promoter with pxp2-Luc vector. The hD
construct was established by ligation of a 663 bp
BamH1/SphI fragment containing the
downstream GnRH promoter to pxp2-Luc vector.
|
|
The differential abilities of the upstream and downstream promoters to
direct reproductive tissue specific gene expression were further tested
by transient transfection assay in kidney cells (293, ATCC) and a
number of reproductive tissue derived tumor cells including JEG, MCF-7,
LNCaP, and PA-1 cells (Fig. 2). The data revealed that
the upstream GnRH promoter-Luc construct (hU) was much more active,
compared with the downstream promoter-Luc construct (hD), in directing
luciferase expression in cell lines derived from reproductive tissues,
but not in kidney cells.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 2. Tissue-specific expression of the GnRH gene in
various human reproductive cell lines. The GnRH upstream promoter (hU)
and downstream promoter (hD) were fused to a promoterless luciferase
(Luc) reporter construct (Fig. 1B ) and transfected into various human
reproductive cell lines by the calcium phosphate method and luciferase
activity was measured in cell extracts after 16–18 h.
|
|
Analysis of the different GnRH mRNA species in various reproductive
tissue derived tumor cells
We next wanted to determine whether the upstream GnRH gene
transcriptional start site is actually used in producing mRNA in tumor
cells derived from human reproductive tissues. RT-PCR with two sets of
DNA oligonucleotide primers were used to examine the different GnRH
mRNA species present in various tumor cell lines. The first pair (A and
D) was designed to identify transcripts initiated from the upstream
start site (Fig. 3C). The second pair of primers (B and
D) recognizes transcripts initiated from both upstream and downstream
start sites (Fig. 3C). When primer B/D was used, RT-PCR analysis of the
first strand cDNA derived from the RNA isolated from human placenta and
tumor cell lines (PA-1, LNCaP, MCF-7, and JEG) yielded a 399 bp and
1269 bp fragments, whereas GT1–7 cells only produced an 399 bp
fragment (Fig. 3A). When primers A/D were used, the 979 and 1,849 bp
fragments were also observed in the placenta and the various cell lines
but not in GT1–7 cells. Southern blot analysis showed that those
RT-PCR products hybridized to the human GnRH cDNA probe (Fig. 3B). When
reverse transcriptase was omitted or when cDNA from human muscle tissue
was used, neither pair of primers produced any amplification products
by PCR (data not shown). Together, both RT-PCR analysis and Southern
blot assay indicated that PA-1, LNCaP, MCF-7, JEG-3, and GT1–7 cells
express the proGnRH gene. The upstream transcriptional start site,
however, is used only in the cell lines derived from reproductive
tissues.
In vitro translation of multiple human GnRH mRNAs
Our previous study revealed that the major transcript in the human
reproductive tissues is a 1949 bp mRNA (10). Computer translation
analysis of this reproductive tissue-specific GnRH mRNA reveals a
series of possible translation start sites in addition to the one used
in the smaller GnRH mRNA to synthesize preproGnRH. However, none of
these would be translated into a protein larger than 47 amino acids
except the GnRH precursor because of the close proximity of in-frame
stop codon. To determine if the 1949 bp GnRH mRNA could be translated
into the 10-kDa preproGnRH protein, two human GnRH cDNA clones were
created by RT-PCR cloning with two pairs of human GnRH primers. For the
first pair the sense primer A (AATTTTAAACACACAACTGTCTG) was located at
1 to 23 bases 3' to the reproductive tissue-specific upstream start
site. The second pair used sense primer B (CAGTACTCAACCTACTTCAAGG)
located at 2 bp 3' to the hypothalamic start site. Both pairs used the
same antisense primer I (ATTCACAACACASTTACTTTATTATG), located one base
5' to the poly A tail. The RT-PCR products of 1949 bp (reproductive
tissue-specific cDNA) and 499 bp (hypothalamic specific cDNA) were then
purified by Geneclean (bio-101, Vista, CA) and inserted into a
pBluescript vector. DNA sequence of these two human GnRH cDNA clones
were determined on both strands and found to be identical to those
previously reported (1, 5). Analysis of the products generated from
in vitro coupled reaction of these two human GnRH clones by
SDS-polyacrylamide gel showed that both GnRH mRNAs produced the 10-kDa
preproGnRH protein (Fig. 4) as the major translation
product, but the translation efficiency of the reproductive
tissue-specific GnRH mRNA appeared to be approximately 4- to 5-fold
lower than that of the hypothalamic mRNA (estimated by molecular
dynamic phosphorimager). Sequence analysis of the 5'-UTR of the
reproductive tissue-specific mRNA revealed three inverted repeat
sequences that can potentially generate stem-loop structures (data not
shown). These secondary structures may be responsible for the
difference in translational efficiency of the two mRNAs.
View larger version (30K):
[in this window]
[in a new window]
|
Figure 4. In vitro transcription and
translation of human GnRH cDNA. The hypothalamic GnRH cDNA (499 bp) and
reproductive specific cDNA (1949 bp) were transcribed and translated
in vitro by TNT coupled Reticulocyte Lysate Systems
(Promega) from both 5' and 3' direction. A 10.1-kDa protein (indicated
by arrow) was observed in both cDNAs only in 5'
direction. A cDNA encoding luciferase was transcribed and translated as
a positive control.
|
|
Location of the tissue-specific regulation region in the GnRH
upstream promoter
To locate the specific upstream and downstream promoter region
that mediates the basal promoter activity, a series of constructs
containing 5'- or 3' deletions of the human GnRH promoter were fused to
the luciferase reporter gene and tested for their ability to drive
expression of luciferase activity (Fig. 5). A pCMV
ß-galactosidase (GUS) construct was cotransfected into JEG cells with
each construct as an internal control. Deletion of the 5' flanking
sequence of the human GnRH promoter up to 548 bp (-1048,
HindIII site) from the upstream transcription start site
significantly decreased the basal promoter activity (Fig. 5). However,
further deletion of the 5' sequence to 169 bases (-723,
AflII site) from the upstream transcription start site
greatly diminished the basal promoter activity (Fig. 5) indicating that
the region between -1048 to -723 is important for the upstream
promoter activity. Deletion of 352 bp 3' flanking sequence from the
-1.7 kb human GnRH upstream promoter (BamHI site, Fig. 5)
did not result in significant change in transcription of luciferase
gene in JEG cells suggesting that the activities of various 5' deletion
promoter constructs tested were not dependent on the DNA sequences
between the upstream and downstream transcription start sites. As
expected, transfection of GT1–7 cells revealed that all luciferase
constructs containing only the upstream promoter were transcriptionally
much weaker than the constructs containing the downstream promoter.
Deletion of 352 bp (-202, AflII site) 5' flanking sequence
from the 0.65 kb downstream promoter retained nearly 80% of the
transcriptional activity in GT1–7 cells suggesting that the 202 bp
sequence 5' to the downstream transcription start site contains the
promoter element directing neuron-specific expression in GT1–7
cells.
View larger version (16K):
[in this window]
[in a new window]
|
Figure 5. Location of the tissue-specific regulation region
in the human GnRH upstream promoter. To localize the specific promoter
region that mediated the basal upstream promoter activity, a series of
constructs containing 5'- or 3'-deletion of human GnRH promoter were
fused to the luciferase reporter gene and tested for their luciferase
activity in JEG and GT1–7 cells. Deletion of the 5'-flanking sequence
of human GnRH promoter up to 548 bases (HindIII site)
away from the upstream transcriptional start site significantly
decreased the basal promoter. Further deletion of the 5' sequence to
169 bases (AflII site) away from the upstream
transcriptional start site greatly diminished the basal promoter
activity. Activity was expressed as relative light unit. Transfections
were performed in duplicate in three separate experiments.
|
|
Identification of a reproductive tissue-specific element in the
GnRH upstream promoter region
To determine whether any regulatory factors bind to the
-1048/-723 region important to upstream promoter activity,
footprinting analysis and gel shift assay were performed. Footprinting
analysis of the HindIII/AflII fragment of the
human GnRH 5' flanking region revealed four elements (•, Fig. 6) that bound to nuclear extract from the JEG cell (Fig. 7 lanes 3–5). To verify this observation, gel shift
analysis was performed. Gel-shift assay of the four elements with
nuclear extract of JEG and GT1–7 cells demonstrated that these four
elements could only bind to protein factors from JEG cell (Fig. 8, lane 2) but not from the GT1–7 cell (Fig. 8, lane
4). These mobility shifts could be competed with a 1000-fold excess of
cold oligonucleotide (Fig. 8, lane 3) indicating their specificity.
These data suggest that the four elements are involved in
tissue-specific expression of GnRH gene in placental cells.
View larger version (55K):
[in this window]
[in a new window]
|
Figure 6. DNA sequence of human GnRH 5'-flanking region. The
numbering is relative to the downstream transcriptional start site of
GnRH gene. The upstream transcriptional start site is shown as a
bent arrow. The footprinting gel shift regions
are underlined with a bold solid line.
|
|
View larger version (45K):
[in this window]
[in a new window]
|
Figure 7. Footprinting analysis of the human GnRH
5'-flanking region. The HindIII/AflII
fragment (Fig. 6) containing the human GnRH gene 5' flanking region was
end-labeled with 32P and incubated with nuclear extract
from JEG cell. Ten microgram BSA (lane 1), 20 µg BSA (lane 2), 10
µg JEG nuclear extract (lane 3), 20 µg JEG nuclear extract (lane
4), and 40 µg JEG nuclear extract (lane 5). The mixture were then
partially digested with DNase (2 units) and resolved on a 6% denatured
polyacrylamide gel.
|
|
View larger version (47K):
[in this window]
[in a new window]
|
Figure 8. Gel-shift assay of human GnRH 5'-flanking regions
that have footprint with JEG nuclear extract (Fig. 7). The double
strand oligonucleotides, located at -849/-876, -896/-919,
-940/-960 and -968/-987, were end-labeled with
[gamma-32P] ATP and mixed with 5 µg of nuclear extract
from JEG cell and GT1–7 cell. The specificity of the shift was
determined by shift competition with 1,000 fold higher concentration of
the cold test oligonucleotides. All experiments were performed in the
presence of 4 µg of poly(dI-dC).
|
|
|
Discussion
|
|---|
The presence of GnRH or GnRH-like peptides has not only been
detected in the hypothalamus, but also in human reproductive tissues,
such as placenta, ovary, testis, and mammary gland. This, together with
the presence of GnRH receptor in these reproductive tissues (7), has
resulted in the suggestion that GnRH may have a paracrine or autocrine
role in the local regulation of reproductive cell biological processes.
Using RT-PCR and primer extension, results of our previous study (10)
revealed that the human GnRH gene has two transcription start sites,
with a second one 5' to the previously describes hypothalamic
transcription start site (6). The upstream transcription start site,
however, appeared to be significantly used only in reproductive tissues
(10).
In the present study, the existence of two independent promoter regions
responsible for tissue-specific expression of human GnRH gene was
confirmed in the experimentally more tractable human reproductive tumor
cell lines. Using transfection analysis we have demonstrated that in
placental cells the GnRH upstream promoter construct expressed
luciferase activity as strongly as the full-length GnRH promoter
construct. However, the downstream promoter construct only demonstrated
minimal activity in directing luciferase expression in placental cells.
In contrast, in hypothalamic GT1–7 cells the human downstream promoter
construct produced luciferase activity similar to the full-length
construct, whereas the upstream promoter construct had almost no
activity (Fig. 1). This observation is in complete agreement with our
previous work (10) showing that approximate 70% of GnRH mRNA is
transcribed from the upstream promoter in the placental cells while
most all of the GnRH mRNA in hypothalamic cells was transcribed from
the downstream promoter (10). We also demonstrated, by RT-PCR, that the
upstream GnRH promoter is also used in various human tumor cell lines
derived from reproductive tissues such as MCF-7 cells(mammary gland),
PA-1 (ovary) and LNCaP (prostate). This observation was verified in
promoter/reporter assays where the upstream start site alone had
stronger promoter activity than the downstream start site in various
reproductive tissue derived tumor cells. The results of these transient
transfections demonstrated that the upstream transcriptional start site
is the major transcription start site in these reproductive cell lines.
These data also show differential usage of the two human GnRH gene
promoters: two nonoverlapping 5' control elements (Fig. 1B), each
containing only one of the two transcriptional start sites, that are
capable of directing reporter gene expression in tumor cells derived
from reproductive tissues or hypothalamic neurons.
Results in the present and previous studies revealed the presence of
four mRNA transcripts from the GnRH gene in human placental tissues and
reproductive tissue-derived tumor cell lines. Two of the mRNAs were
transcribed from downstream transcription start site and the other two
from the upstream start site. Both types of transcripts were found with
or without the first intron (1, 6, 10). The question arises as to
whether these mRNAs are functional, that is, could these mRNAs be
translated into proGnRH or a protein other than GnRH? A computer
analysis of the nucleic acid sequences of the reproductive tissue
specific GnRH mRNAs reveals a series of possible translation start
sites. However, only five regions could be translated into peptides
longer than 30 amino acids, including the GnRH translation region, and
the other four possible translation regions. Close examination of the
four possible translation regions discloses that all represent poor
Kozak translation consensus initiation sequences (15, 16), and each
region has a terminator codon within 47 amino acids; ergo, the
preproGnRH protein would be expected to be the major product of the
reproductive specific GnRH mRNAs. Results of the in vitro
translation study demonstrate for the first time that in
vitro synthesized RNA, corresponding to both the hypothalamic type
of GnRH mRNA (transcribed from the downstream start site without intron
1), and the larger reproductive tissue specific GnRH mRNA (transcribed
from upstream start site retaining intron 1), can act as a template for
protein synthesis in cell-free lysates. Because the upstream
transcription start site is used at a significant level only in
reproductive tissue (about 70% of the GnRH mRNA is transcribed from
this site) (10), the possible translation of these reproductive
specific GnRH mRNAs becomes an important issue for characterizing the
role of GnRH in the reproductive function of these tissues.
The use in placenta of a promoter other than the hypothalamic GnRH
downstream promoter could be reflected in differential control of GnRH
gene expression. Using a series of deletion constructs we have
determined the sequences mediating the reproductive tissue-specific
upstream promoter activity of GnRH in placental cells. Deletion of the
544 bp 3' flanking sequence (in reference to the upstream transcription
start site) from the 1.7 kb human GnRH promoter did not result in a
significant change in promoter activity). However, truncation of 5'
sequence of the upstream promoter from 1.7 kb to shorter fragments of
1158 bp (to BstXI site) and 846 bp (HindIII site)
produced gradual decreases in promoter activity. Further truncation to
169 bp (-723 bp AflII site) 5' to the upstream
transcription start site drastically decreased promoter activity
indicating that the region of this 325 bp segment (-1048 to -723,
Fig. 7) confers transcriptional activity required for tissue specific
GnRH gene expression in JEG cells.
To determine whether any regulatory factors bind to this upstream
region, footprinting and gel-shift assays were performed. We
demonstrated that there are four elements in this region that bind to
nuclear extract from placental cells (JEG) but not from hypothalamic
neuronal cells (GT1–7). Although the results in this study suggest
these four elements may be involved in reproductive tissue-specific
expression of the human GnRH gene in JEG-3 cells, the relative
importance of each protein-binding promoter-elements for the activity
of the reproductive tissue-specific GnRH promoter awaits further
promoter deletion and linker scanning analysis. Several reports have
suggested the involvement of POU family, especially Pit-1, in mediating
tissue-specific gene expression in the placenta (17, 18). Closer
examination of the four sequence elements indicates that the B element
(Fig. 6) GAGATTTAAATAG is a short AT-rich sequence. This sequence is
highly similar to the cis-active sequence of the POU family and
homeobox protein, such as Pit-1, OTF II, Mat a2 protein and eve
protein, suggesting tissue-specific function of these elements.
Moreover, a short AT-rich sequence (TAAAT), similar to those
cis-regulatory sequences of POU family homeobox proteins, is present in
all four elements, thus supporting our suggestion that these four
elements may be involved in POU family mediated reproductive
tissue-specific expression of the GnRH gene in the human placental
cells.
In summary, the human GnRH gene uses two transcription start sites, one
upstream and the other downstream, to produce different GnRH mRNAs. The
downstream transcription start site is fully active in the hypothalamic
neuronal cells (GT1–7). However, the upstream transcription start site
is primarily used in the placental cell and cell lines derived from
nonhypothalamic reproductive tissues. Computer analysis coupled with
in vitro translation revealed that the GnRH mRNAs
transcribed from both the upstream and downstream promoters are capable
of translating into preproGnRH protein. Serial deletion studies
demonstrated a specific region, localized on the upstream promoter
region, directs reproductive tissue-specific expression. DNase I
footprint study using nuclear extract from JEG-3 cells indicated four
specific sequence elements that are capable of binding to protein. The
interaction between DNA and protein (only nuclear extract of JEG-3
cells but not of GT1–7 cells) in four specific sequences was confirmed
by gel mobility shift analysis.
|
Acknowledgments
|
|---|
We would like to thank K. W. Cheng for technical assistance, I.
Sarfati for editorial/secretarial assistance, A. Gore, Evan Kelly, and
Greg Johnston for critical reading of the manuscript.
|
Footnotes
|
|---|
1 This work was supported by a NIH Grant (IR29HD/CA30244 to K.W.D.) and
Croucher Foundation, Research Grant Council (to K.L.Y.). 
Received January 13, 1997.
|
References
|
|---|
-
Adelman JP, Mason AJ, Hayflick JS, Seeburg PH 1986 Isolation of the gene and hypothalamic cDNA for the common
precursor of gonadotropin-releasing hormone and prolactin
release-inhibitory factor in human and rat. Proc Natl Acad Sci USA 83:179–183[Abstract/Free Full Text]
-
Bond CT, Hayflick JS, Seeburg PH, Adelman JP 1989 The rat gonadotropin-releasing hormone: SH locus: structure and
hypothalamic expression. Mol Endocrinol 3:1257–1262[Abstract/Free Full Text]
-
Mason AJ, Hayflick JS, Zoeller RT, Young III WS,
Phillips HS, Nikolics K, Seeburg PH 1986 A deletion truncating the
gonadotropin-releasing hormone gene is responsible for hypogonadism in
the hpg mouse. Science 234:1366–1371[Abstract/Free Full Text]
-
Kepa JK, Wang C, Neeley CI, Raynolds MV, Gordon DF,
Wood WM, Wierman ME 1992 Structure of the rat gonadotropin
releasing hormone (rGnRH) gene promoter and functional analysis in
hypothalamic cells. Nucleic Acids Res 20:1393–1399[Abstract/Free Full Text]
-
Seeburg PH, Adelman JP 1984 Characterization of
cDNA for precursor of human luteinizing hormone releasing hormone.
Nature 311:666–668[CrossRef][Medline]
-
Radovick S, Wondisford FE, Nakayama Y, Yamada M,
Cuttler Jr GB, Weintraub BD 1990 Isolation and
characterization of the human gonadotropin-releasing hormone gene in
the hypothalamus and placenta. Mol Endocrinol 4:476–480[Abstract/Free Full Text]
-
Stojilkovic SS, Reinhart J, Catt KJ 1994 Gonadotropin-releasing hormone receptors: structure and signal
transduction pathways. Endocr Rev 15:462–499[Abstract/Free Full Text]
-
Leung PCK, and Steele GL 1992 Intracellular
signaling in the gonads. Endocr Rev 13:476–498[Abstract/Free Full Text]
-
Siler-Khodr TM, Khodr GS, Valenzuela G, Rhode J 1986 Gonadotropin-releasing hormone effects on placental hormones
during gestation: II. Progesterone, Estrone, Estradiol and Estriol.
Biol Reprod 34:255–264[Abstract]
-
Dong K-W, Yu K-L, and Roberts JL 1993 Identification of a major up-stream transcription start site for the
human progonadotropin-releasing hormone gene used in reproductive
tissues and cell lines. Mol Endocrinol 7:1654–1666[Abstract/Free Full Text]
-
Mellon PL, Windle JJ, Goldsmith PC, Padula CA, Roberts
JL, Weiner RI 1990 Immortalization of hypothalamic GnRH neurons by
genetically targeted tumorogenesis. Neuron 5:1–10[CrossRef][Medline]
-
Sambrook J, Fritsch EF, and Maniatis T 1989 Standard protocol for calcium phosphate-mediated transfection of
adherent cells. In: Sambrook J, Fritsch EF, Maniatis F (ed) Molecular
Cloning - A Laboratory Manual (ed. 2), Cold Spring Harbor Laboratory
Press, pp 16.33–37
-
Wondisford FE, Farr E A, Radovick S, Steinfelder H,
Moates JM, McClaskey JH, Weintraub BD 1989 Thyroid hormone
inhibition of human thyrotropin ß-subunit gene expression is mediated
by a cis-acting element located in the first exon. J Biol Chem 264:14601–14604[Abstract/Free Full Text]
-
Jakubowski M, Roberts JL 1992 Multiplex solution
hybridization-RNase protection assay for quantitation of different RNA
transcripts from snap-frozen neuroendocrine tissue of individual
animals. J Neuroendocrinol 4:79–89
-
Kozak M 1981 Possible role of flanking nucleotides
in recognition of the AUG initiator codon by eukaryotic ribosomes.
Nucleic Acids Res 9:5233–5252[Abstract/Free Full Text]
-
Kozak M 1984 Compilation and analysis of sequence
upstream from the translational start site in eukaryotic mRNA. Nucleic
Acids Res 12:857–872[Abstract/Free Full Text]
-
Catanzaro DF, Sun J, Gilbert MT, Yan Y, Black T, Sigmund
C, Gross KW 1994 A pit binding site in human renin gene promoter
stimulates activity in pituitary, placental and juxtaglomerular cells.
Kidney Int 46:1513–1515[Medline]
-
Sun J, Oddoux C, Gilbert MT, Yan Y, Lazarus A, Campbell
WG, Catanzaro DF 1994 Pituitary specific transcription factor
(pit-1) binding site in the human renin gene 5'-flanking DNA stimulates
promoter activity in placental cell primary cultures and pituitary
lactosomtotropic cell lines. Circ Res 75:624–629[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
I. S. Parhar, T. Soga, S. Ogawa, S. Ogawa, D. W. Pfaff, and Y. Sakuma
Nonmammalian gonadotropin-releasing hormone molecules in the brain of promoter transgenic rats
PNAS,
April 19, 2005;
102(16):
5880 - 5885.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. L. Petersen, E. N. Ottem, and C. D. Carpenter
Direct and Indirect Regulation of Gonadotropin-Releasing Hormone Neurons by Estradiol
Biol Reprod,
December 1, 2003;
69(6):
1771 - 1778.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. L. Gray, B. A. Adams, C. M. Warby, K. R. von Schalburg, and N. M. Sherwood
Transcription and Translation of the Salmon Gonadotropin-Releasing Hormone Genes in Brain and Gonads of Sexually Maturing Rainbow Trout (Oncorhynchus mykiss)
Biol Reprod,
November 1, 2002;
67(5):
1621 - 1627.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. A. Lawson, L. A. Macconell, J. Kim, B. T. Powl, S. B. Nelson, and P. L. Mellon
Neuron-Specific Expression in Vivo by Defined Transcription Regulatory Elements of the GnRH Gene
Endocrinology,
April 1, 2002;
143(4):
1404 - 1412.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. A. Wuensch, R. Y. Huang, J. Ewing, X. Liang, and J. T. Y. Lau
Murine B cell differentiation is accompanied by programmed expression of multiple novel {beta}-galactoside {alpha}2,6-sialyltransferase mRNA forms
Glycobiology,
January 1, 2000;
10(1):
67 - 75.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
);
mRNA transcription from upstream start site with unspliced intron 1
(
).