| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Zoology, University of Hong Kong (Y.-C.I., W.M.L.), Hong Kong, China; and Departments of Obstetrics and Gynecology, and Pharmacology and Toxicology (Y.-C.I., G.L.H.), University of Western Ontario, London Regional Cancer Center, London, Ontario, Canada N6A 4L6
Address all correspondence and requests for reprints to: Dr. Will M. Lee, Department of Zoology, University of Hong Kong, Pokfulam Road, Hong Kong, China. E-mail: hrszlwm{at}hku.hk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the testis, expression of the gene encoding SHBG (shbg) gives rise to an identical protein, known as the androgen-binding protein, which is thought to participate in androgen-dependent sperm maturation within the male reproductive tract (10). However, plasma SHBG and testicular androgen-binding protein are separated by the blood-testis barrier, and the protein probably functions quite differently in these two tissue compartments. It has also been shown that shbg is expressed in rat brain (11), human placenta (12), and human uterine endometrium (13), but the biological significance of these observations is not known.
A SHBG-like protein has been identified in the blood of a wide variety of mammals and other vertebrate species, including fish, amphibians, and reptiles (14, 15). However, the age dependence and hormonal control of SHBG gene (shbg) expression in different tissues varies considerably among species. In rodents, such as rats (16), mice (17, 18), and hamsters (19, 20), the liver does not produce SHBG postnatally, and it is only produced transiently in the liver during late fetal life. Consequently, the trace amounts of SHBG in the blood of male and female rats (21) is probably derived from sources other than the liver. By contrast, shbg is expressed postnatally in the both the liver and testis of humans (1, 22) and rabbits (23). To gain insight into these differences in shbg expression between species, we have compared the rabbit shbg (rbshbg) sequence with the human (1) and rat (24) shbg sequences and have studied the molecular characteristics and functional activity of its 5'-flanking region in several different cell lines.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cloning and sequence analysis of rbshbg
Primers corresponding to various positions of the
rbshbg complementary DNA (cDNA) sequence (23) were used to
amplify introns using rabbit genomic DNA as template. Multiple PCR
reactions were conducted using different combinations of paired primers
to ensure that the same sequence was amplified at least twice. The
amplified DNA was cloned into pBluescript II
(KS+), and overlapping genomic fragments were
sequenced by the dideoxy-chain termination method using an ALF DNA
sequencer (Amersham Pharmacia Biotech, Uppsala, Sweden)
with universal vector primers.
RNA extraction
Total RNA was prepared from tissues using Trizol reagent, as
suggested by Life Technologies, Inc. (Burlington, Canada).
The concentration of RNA was determined by spectrophotometry at 260 nm,
and its integrity was assessed by agarose gel electrophoresis.
Polyadenylated RNA was prepared by oligo(deoxythymidine) affinity
chromatography using the PolyATract System IV (Promega Corp., Madison, WI).
Analysis of the transcription start site
Primer extension reactions were conducted using a standard
method (25), as described previously (26). A synthetic oligonucleotide
(5'-ATA GAT CTC AGG CCT CTG CTC TCC AC-3') complementary to 18
nucleotides (nt) upstream of the translation start codon of rbshbg
messenger RNA was end labeled with
[
-32P]deoxy-ATP using T4 polynucleotide
kinase (Life Technologies, Inc.). Total RNA (100 µg)
from adult rabbit liver, testis, kidney, or spleen was hybridized
overnight at 42 C with the labeled oligonucleotide in Tris buffer (10
mM; pH 8.3) containing 0.15 M KCl and 1
mM EDTA. RT was initiated by adding 20 U Superscript
reverse transcriptase (Life Technologies, Inc.) and was
performed at 42 C. After termination of the reaction, the products were
subjected to ribonuclease A digestion at 37 C for 15 min and ethanol
precipitation before analysis by urea-acrylamide gel electrophoresis.
The sizes of the primer extended products were determined by direct
comparison with a sequencing reaction generated from the same
primer.
Rapid amplification of cDNA ends (5'RACE)
5'RACE was performed using a Marathon cDNA Amplification Kit
(CLONTECH Laboratories, Inc., Palo Alto, CA).
Polyadenylated RNA (1 µg) from liver and testis of the same adult
rabbit was reverse transcribed into cDNAs. A specially designed adapter
sequence provided in the Marathon kit was ligated to their ends, and
the adapter primer served as the forward primer. An antisense exon
5-specific primer (5'-CT GAA TTC TGC GTG AGT CCC TGG AGG-3') derived
from nt 688–713 in the rbshbg cDNA, and an antisense exon
3-specific primer (5'-CG GAG CTC AAA GGA GGA GGA GGC TTT G-3') derived
from nt 207–233, served as the outer and nested primers, respectively.
Major PCR products were cloned into pBluescript II
(KS+) and sequenced.
Identification of the rbshbg 5'-flanking sequence
The 5'-flanking sequence of rbshbg was obtained using
the protocol described in the Universal GenomeWalker Kit
(CLONTECH Laboratories, Inc.). Briefly, five separate
walker libraries were constructed by ligating a specially designed
adapter sequence to rabbit genomic DNA (CLONTECH Laboratories, Inc.), and each was digested by a different restriction enzyme.
The antisense gene-specific primers were designed from the
rbshbg exon 1 sequence. The outer primer (5'-TAT TCC TGC ACC
TGG GTG GCC AGA AC-3') was complementary to nt 137–160 in the
rbshbg (GenBank accession no. AF144712). The nested primer
(5'-AT TCT AGA GCT TCT CAG GGC CAG CCG ATG GTG-3') was complementary to
nt 113–136. Major PCR products amplified with the adapter primers were
cloned into pBluescript II (KS+) and sequenced.
Gene-specific primers designed for subsequent walkings were as follows:
5'-AT TCT AGA ATC CGT TTG TAC AGG ACC CAA CA-3' (complementary to
-561/-539 in Fig. 4
), 5'-ATC CGC CCC CCA CTC GCA ACT ATC TTT CCT T-3'
(-538/-510), 5'-AT TCT AGA AAG CAC GAG GCG AGG CTG TGA CTT-3'
(-762/-739), and 5'-ATG CTT AAG TTG AAA ACC ATG CAC ACT TCC-3'
(-731/-704). The 5'-flanking sequence was finally confirmed by direct
PCR from rabbit genomic DNA (CLONTECH Laboratories, Inc.)
using gene-specific primers (5'-G GGG AGA TCT AAT ATG TGG GGG CAG GCA
G-3' and 5'-A TGG GAG CTC AGG CTG GAG CGC CCG GAC-3') and a
proof-reading Taq polymerase (ID-PROOF DNA Polymerase, ID
Labs Biotechnology, London, Canada). Various portions of the
5'-flanking region were directly amplified by PCR from rabbit genomic
DNA and were cloned into pGL2 Basic (Promega Corp.) for
analysis of their transcriptional activities.
|
-32P]deoxy-ATP and T4 polynucleotide kinase
(Life Technologies, Inc.). Unincorporated radiolabeled
nucleotides were removed with a MicroSpin G-25 column (Amersham Pharmacia Biotech, Inc., Baie d’Urfé, Canada), and the
radiolabeled DNA fragments were further gel purified. DNase I (fast
protein liquid chromatography pure; Amersham Pharmacia Biotech) was used at a 1:40 dilution (0.56 U/µl) to digest
approximately 20,000 cpm end-labeled DNA in an in vitro
footprinting assay (27). The DNase I digestions were terminated, and
the products were treated with proteinase K before analysis on an 8%
urea-acrylamide DNA sequencing gel along side a Maxam-Gilbert
sequencing reaction of the footprinted fragment (29). The resulting gel
was examined by autoradiography.
Electrophoretic mobility shift assay
The double stranded oligonucleotide used in the assay was
5'-gttcac TGG CTT TCT TGG CAA TGA GGG Atgg-3'/3'-aagtg ACC GAA AGA ACC
GTT ACT CCC Tac cg-5', which corresponds (in uppercase
letters) to the rbshbg footprinted region FPa. This
fragment was end labeled by [
-32P]deoxy-CTP
and Klenow fragment of DNA polymerase I (Life Technologies, Inc.). Rabbit liver nuclear protein (5 µg) extract was
incubated in the presence and absence of an excess of unlabeled
competitor oligonucleotide (2.5 pmol) in a final volume of 15 µl
containing 10 mM HEPES (pH 7.6), 50
mM KCl, 25 mM
MgCl2, 10% glycerol, 1 mM
dithiothreitol, and 3 µg poly(dI:dC). After a 10-min incubation on
ice, 25 fmol 32P-labeled, double stranded
oligonucleotide were added and incubated at room temperature for 15
min. The reaction mixture was then loaded onto a 6% polyacrylamide
gel. The gel was electrophoresed at 180 V for 30 min, dried, and
autoradiographed.
Site-directed mutagenesis
Mutagenesis was performed with the Altered Sites In
Vitro Mutagenesis System (Promega Corp.). Single
stranded DNA comprising the region to be mutated was prepared by the
pSELECT-1 vector. It was then isolated and annealed to an ampicillin
repair primer and a mutagenic primer (5'-CTC ACC CCG TTT GCC TGG GCA
GGG GTC AAG GGT CAG TGG CCC T-3'). The mutated strand was synthesized,
ligated by addition of 10 U T4 DNA polymerase and 2 U T4 DNA ligase,
and incubated at 37 C for 3 h. The mutant vector was then
transformed into Escherichia coli BMH 71–18 mut S cells by
heat shock. The plasmid DNA was extracted and transformed into E.
coli JM107 cells, and the resulting mutant DNA was sequenced to
confirm that only the targeted mutation had occurred.
Cell culture and transfections
The culture media and reagents used for tissue culture
experiments were obtained from Life Technologies, Inc.
Human hepatoblastoma (HepG2) cells, mouse Sertoli (TM4) cells, and
Chinese hamster ovary (CHO) cells were cultured in DMEM (high glucose),
supplemented with 10% FBS, 2 mM L-glutamine,
100 U/ml penicillin, and 100 µg/ml streptomycin. Reporter constructs
were transfected using the Lipofectamine reagent, according to the
protocol suggested by Life Technologies, Inc. Briefly,
cells were plated in six-well plates at approximately 2.5 x
105 cells/well 24 h before transfection.
Before addition of DNA-liposome complexes, cells were rinsed with
serum-free DMEM. For each transfection, 1.2 µg reporter construct
were cotransfected with 0.2 µg pCMVlacZ control plasmid in 1 ml
serum-free DMEM by incubation at 37 C for 5 h. An equal volume of
DMEM containing 20% FBS was then added, and the cells were incubated
overnight at 37 C. The culture medium with the DNA-liposome mixture was
replaced by DMEM containing 10% FBS on the following day. Forty-eight
hours after the start of transfection, cells were rinsed twice with PBS
(10 mM sodium phosphate and 0.15 M NaCl, pH 7.5) and
harvested by scraping in 40 mM Tris-Cl (pH 7.5), 1
mM EDTA, and 0.15 M NaCl. Cells were pelleted
by centrifugation for 1 min and resuspended in 100 µl 0.25
M Tris, pH 8.0. Cell extracts were prepared by lysing with
three freeze-thaw cycles. For luciferase assays, cell extracts (20
µl) were mixed with 100 µl luciferase assay reagent (Promega Corp.) for detection in a luminometer. For ß-galactosidase
assay, cell extracts (5 µl) were incubated with 100 mM
sodium phosphate (pH 7.2), 1 mM
MgCl2, 50 mM ß-mercaptoethanol, and
665 µg/ml
o-nitrophenyl-D-galactopyranoside in
a total volume of 100 µl at 37 C until a yellow color was present.
The reaction was stopped by the addition of 150 µl 1
M sodium carbonate. Absorbance at 405 nm was
measured and used to correct for transfection efficiency. Relative
luciferase activities were calculated by dividing luciferase light
units by OD reading from the ß-galactosidase assay. Fold increases in
the relative luciferase activities of various constructs were
determined in relation to the background luciferase activity of the
promoterless pGL2 Basic. All transfection experiments were performed in
duplicate and were repeated at least three times.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Intron-exon junctions (Fig. 1B)
were determined by comparing the genomic sequence with a
rbshbg cDNA sequence (23). Altogether, rbshbg
spans 2.5 kb and contains eight exons. The sequence of intron-exon
boundaries reveals no deviation from the consensus sequence for the
5'-donor and 3'-acceptor splice sites (30) and follows the GT/AG rule
(Fig. 1B). Exon 1 contains a 52 bp 5'-untranslated region and the
initiation codon for the SHBG precursor polypeptide (see also below).
The coding region of the gene is distributed over eight exons with
sizes ranging from 92–208 bp, whereas introns are 98–316 bp in length
(Fig. 1B).
|
).
The size heterogeneity of these products obtained from liver and testis
RNA was very similar and ranged between 57–71 nt in length, with the
most abundant product being 60 nt in length in both cases. Although
this suggests multiple transcription start sites, a major transcription
start site is clearly used in both liver and testis, and this is
located 52 bp upstream from translation initiation codon in exon 1.
|
Potential regulatory elements in the rabbit shbg 5'-flanking
region
To identify potential sequence elements involved in the
transcriptional regulation of rbshbg, we isolated a region
(893 bp) 5' to the translation initiation codon in exon 1 (Fig. 3). This sequence does not contain a
canonical TATA box, a CCAAT box or an initiator sequence, but there is
a GC box motif located around nt -518/-504 relative to the major
transcription start site (marked by an arrow and defined as
+1), as shown in Fig. 3. A computer-assisted search using MatInspector
(31) revealed several putative response elements with homology to known
binding sites for transcription factors. These include two binding
sites for hepatocyte nuclear factor-3ß (HNF-3ß; consensus
TATTKRYTY) (32) at -365/-353 and -619/-607; three C/EBPß sites
(consensus TKNNGYAAK, ATTGCGCAAT) (32) at -43/-30, -340/-329, and
-542/-528; two upstream stimulator factor (USF) sites (consensus
GGCCACGTGACC) (32) at -687/-675 and -749/-741; two overlapping
binding sites for nuclear factor I (NF-I; consensus
TTGGCN5–7GCCAA) (33) at -170/-156 and
-162/-149.
|
). In addition, an electrophoretic
mobility shift assay of rabbit liver nuclear proteins using a double
stranded FPa oligonucleotide revealed that two protein-DNA complexes
formed on this template, and addition of excess unlabeled
oligonucleotides demonstrated the specificity of these interactions
(Fig. 4B). A region within the hshbg promoter that overlaps
FPa in the rbshbg proximal promoter is also protected with
liver nuclear extracts in a DNase I footprinting assay (FP6) (26), and
these regions contain two overlapping binding sites for NF-1 (Fig. 3).
Transcriptional activity of the rbshbg 5'-flanking region
To localize putative cis-acting elements within
the rbshbg proximal promoter and to evaluate its
transcriptional activity, various lengths of rbshbg
5'-flanking region were inserted into the pGL2 Basic luciferase
reporter gene vector (Fig. 5A). The
resulting plasmids were transiently transfected into HepG2, TM4, and
CHO cells by lipofection to examine the cell type specificity of their
transcriptional activities. In general, all constructs showed an
increase in relative luciferase activity in all cell lines tested as
the 5'-end extended from -234 to -841 (Fig. 5B). However, the
activities in HepG2 and TM4 cells were always much higher than those in
CHO cells. As shown in Fig. 5B, 5'-deletions from -841 to -806
decreased transcriptional activity, whereas deletions from -806 to
-497 increased it. This suggests that repressive elements exist within
the -806/-498 region, and these can be modulated by elements within
the -841/-806 region.
|
) provided additional evidence that the
two most proximal DNase I footprinted regions (FP1 and FP2) are
conserved between species. However, we failed to detect any footprints
within this region of the rbshbg sequence by the DNase I
footprinting assay despite the fact that positive footprints were
obtained on the human sequence when used as a positive control together
with either mouse or rabbit nuclear proteins (data not shown). Just
upstream from these conserved regions, a cis-element that
interacts with USF in the hshbg proximal promoter (26) is
lacking in the rabbit and rat shbg 5'-flanking regions (see
FP4 in Fig. 6A). This lack of sequence conservation in the
shbg promoters between species extends into the 5'-region of
FP3 in the hshbg promoter sequence (26), which binds the
chicken ovalbumin upstream promoter-transcription factor (COUP-TF) and
HNF-4. We therefore assessed the functional importance of the latter
binding site for COUP-TF/HNF-4 by changing the corresponding segment in
rbshbg to a perfect hshbg FP3 sequence by
oligonucleotide-directed mutagenesis (Fig. 6A). As shown in Fig. 6B, this mutation increased the rbshbg proximal promoter
activity by 7.3- and 7.1-fold in HepG2 and TM4 cells, respectively, but
only by 3.2-fold in CHO cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Like shbg promoter sequences from humans and rats (1, 24), the 5'-flanking region of rbshbg showed no obvious TATA box, CAAT box, or initiator sequences. However, together with other common features, such as multiple transcription start sites and the presence of a GC box consensus sequence, the rbshbg 5'-flanking region displays typical features of a TATA-less type promoter. In particular, clusters of GA- and CT-rich regions forming short and imperfect repeats typically found in promoter sequences were observed around the -200 region of rbshbg, and computer analysis of the rbshbg 5'-flanking region showed consensus sequences for several liver-specific transcription factors. In addition, a protein-binding site at -170/-149 was demonstrated by DNase I footprinting and gel shift assays. This partially overlaps a footprinted region (FP6) in the hshbg proximal promoter (26), and contains a sequence that resembles the consensus binding site for the transcription factor NF-1. Members of NF-1 family have been shown to be highly expressed in liver (34) and are able to modulate liver-specific transcription by interacting with HNF-3 (35). The proteins binding to the FPa were not identified, but their possible involvement in modulating shbg expression in the liver should be explored further.
The rbshbg 5'-flanking region preferentially supports gene expression in human hepatoblastoma cells and mouse Sertoli cells, and this is in line with our previous Northern hybridization and RT-PCR studies on the tissue specificity of rbshbg expression, where expression was confined to liver and testis (23). Several important regulatory sequences have been identified within the -298 region of the hshbg promoter (26), which shares partial sequence homology with the corresponding region of the rbshbg promoter. However, this region of the rbshbg promoter was relatively inactive in hepatoblastoma cells compared with the hshbg proximal promoter. Sequence comparisons of this region of the shbg promoter showed that at least one footprint (FP1) is highly conserved between species, and this cis-element is a binding site for HNF-4, which actively recruits transcriptional machinery to the hshbg promoter in the absence of a binding site for the TATA-binding protein (26). This loading of HNF-4 onto FP1 might require the cooperative binding of transcriptional factors to FP3, as evidenced by the fact that the conserved FP1 sequence on rbshbg does not bind nuclear proteins, but transcriptional activity is greatly enhanced when its adjacent sequence is mutated to provide a perfect COUP-TF/HNF-4-binding site.
The first 497 bp of the rbshbg promoter was equally active in hepatoblastoma and Sertoli cells, and 5'-deletions of the promoter modified its activity similarly in both cell types. Interestingly, the repression in transcriptional activity caused by the presence of sequences between -806 and -497 was restored in the construct containing an additional 35 bp of 5'-flanking sequence. This short sequence does not contain any obvious transcription factor-binding sites, and its effects on rbshbg promoter activity remain to be defined. However, if this sequence is functionally important its activity is unlikely to be conserved between species, because the similarity between shbg promoter sequences between species only extends to approximately 600 bp 5' from their transcription start sites. Our studies of the rbshbg also revealed large gaps in shbg promoter sequence alignments between species within this region, and it is very likely that these contribute to differences in their activities.
The most obvious difference in the shbg proximal promoter sequences between species is the lack of sequences in rabbit and rat promoters that correspond to an USF-binding site at FP4 in the hshbg proximal promoter (24, 26). In addition, a COUP-TF/HNF-4-binding site at FP3 in the hshbg promoter is poorly conserved in the corresponding sequences of other species. In previous studies, this latter site did not contribute significantly to its transcriptional activity of the hshbg promoter (26), but its inclusion added significantly to the transcriptional activity of the rbshbg promoter, which lacks the adjacent USF-binding site. This further supports the concept that a functional interplay occurs between USF and COUP-TF/HNF-4 at their closely positioned recognition sequences in the hshbg promoter. Unlike the shbg in rodents, the shbg in humans and rabbits is expressed in the liver during postnatal life, and the absence of these cis-elements in the rabbit and rat shbg promoters cannot therefore account for species differences in the expression of shbg postnatally. However, it may contribute to species-specific differences in the way the gene responds to hormonal or environmental stimuli, such as nutritional status, in different tissues, and this needs to be studied further. In support of this, the addition of a human shbg FP3 sequence to the rabbit shbg promoter specifically enhanced its activity in hepatoblastoma cells and mouse TM4 Sertoli cells compared with the CHO cells. It is therefore likely that additional studies of the rbshbg promoter characterized in this report will yield new information about the tissue specificity and regulation of shbg expression in various mammalian species including humans.
Stady II—International Symposium on Signal Transduction in Health and Disease 12–15 September 2000, Tel Aviv, Israel
For further information from Prof. Zvi Naor, Department of Biochemistry, Tel Aviv University, Tel Aviv Israel. Tel: +972 3 640 9032/6417057; Fax: +972 3 640 6834; e-mail: stady2000@unitours.co.il or naorzvi@post.tau.ac.il.
|
Footnotes |
|---|
Received October 12, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K.-M. Ng, M.-T. So, and W. M. Lee Expression of Rabbit Sex Hormone-Binding Globulin during Pregnancy and Prenatal Development and Identification of a Novel Isoform Endocrinology, April 1, 2005; 146(4): 1965 - 1972. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Cheng, C. H. Cheung, and W. M. Lee Mouse Testin: Complementary DNA Cloning, Genomic Organization, and Characterization of Its Proximal Promoter Region Biol Reprod, April 1, 2003; 68(4): 1376 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Munell, C. A. Suarez-Quian, D. M. Selva, O. M. Tirado, and J. Reventos Androgen-Binding Protein and Reproduction: Where Do We Stand? J Androl, September 1, 2002; 23(5): 598 - 609. [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Cheng, C. M. Yeung, B. K. C. Chow, and P. C. K. Leung Characterization of a New Upstream GnRH Receptor Promoter in Human Ovarian Granulosa-Luteal Cells Mol. Endocrinol., July 1, 2002; 16(7): 1552 - 1564. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
