(Received for publication, November 27, 1996)
From the Departments of Obstetrics & Gynecology and Cell Biology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-2515, the § Department of Molecular Biology, Graduate School of Medical Science, Kyushu University, Higashi-ku Fukuoka 812, Japan, and the ¶ Department of Clinical Biochemistry, University of Edinburgh, Edinburgh EH3 9YW, Scotland
Steroidogenic factor-1/adrenal
4-binding protein (SF-1/Ad4BP) is an orphan nuclear
receptor/transcription factor known to regulate the P450 steroid
hydroxylases; however, mechanisms that regulate the activity of
SF-1/Ad4BP are not well defined. In addition, little is known about the
mechanisms that regulate the human steroidogenic enzyme, type II
3-hydroxysteroid dehydrogenase (3
-HSD II), the major gonadal and
adrenal isoform. Regulation of the 3
-HSD II promoter was examined
using human adrenal cortical (H295R; steroidogenic) and cervical (HeLa;
non-steroidogenic) carcinoma cells. H295R cells were transfected with a
series of 5
deletions of 1251 base pairs (bp) of the 3
-HSD II
5
-flanking region fused to a chloramphenicol acetyltransferase (CAT)
reporter gene followed by treatment with or without phorbol ester
(phorbol 12-myristate 13-acetate; PMA). CAT assay data indicated that
the region from
101 to
52 bp of the promoter was required for
PMA-induced expression. A putative SF-1/Ad4BP regulatory element,
TCAAGGTAA, was identified by sequence homology at
64 to
56 bp of
the promoter. Cotransfection of HeLa cells with the
101 3
-HSD-CAT
construct and an expression vector for SF-1/Ad4BP increased CAT
activity 49-fold. Subsequent treatment with PMA induced an unexpected
synergistic increase in transcriptional activity 540-fold over basal.
Mutation of the putative response element (TCAA
TAA to
TCAA
TAA) abolished SF-1-induced CAT activity and the
synergistic response to PMA. Gel mobility shift assays confirmed that
SF-1/Ad4BP interacts with the putative element and transcripts for
SF-1/Ad4BP were detected in H295R cells by Northern analysis. These
data are the first to demonstrate 1) regulation of a non-cytochrome
P450 steroidogenic enzyme promoter by SF-1/Ad4BP, 2) a powerful
synergistic effect of PMA on SF-1/Ad4BP-induced transcription, and 3)
the importance of the SF-1/Ad4BP regulatory element in the regulation
of the 3
-HSD II promoter.
The steroidogenic enzyme 3-hydroxysteroid
dehydrogenase/
5-
4-ene-isomerase
(3
-HSD)1 is essential for the
biosynthesis of all classes of steroid hormones and catalyzes the
dehydrogenation and isomerization of
5-3
-hydroxysteroids including pregnenolone,
17
-hydroxypregnenolone, dehydroepiandrosterone, and
5-androstene-3
,17
-diol to the
4-3-ketosteroids
progesterone, 17
-hydroxyprogesterone, androstenedione, and
testosterone, respectively. Subsequent tissue-specific metabolism of
4-3-ketosteroids by various cytochrome P450 enzymes
results in the production of glucocorticoids, mineralocorticoids,
estrogens and androgens. This crucial enzyme is present in classical
steroidogenic tissues such as the adrenal cortex, ovary, testis, and
placenta, and was more recently localized to peripheral tissues such as prostate, mammary gland, and skin (1). In the human, 3
-HSD exists as
two isoforms (type I and II) derived from the tissue-specific expression of two highly related but distinct genes (2, 3). Human type
I 3
-HSD is predominantly expressed in placenta and skin and is the
major form found in breast tissue (2). In contrast, type II 3
-HSD
expression is almost exclusively localized to the adrenal, ovary, and
testis (2).
The essentiality of this enzyme in adrenal and gonadal steroidogenesis
is underscored by the severe physiological consequences that arise in
cases of 3-HSD deficiency. Congenital adrenal hyperplasia, which can
be fatal if not detected and treated early, occurs in response to
deficiencies in any one of the steroidogenic enzymes involved in the
biosynthesis of cortisol, including 3
-HSD (4). However, because
3
-HSD is also involved in gonadal steroidogenesis, insufficient
levels of 3
-HSD may impair sexual differentiation, resulting in
pseudohermaphroditism with incomplete masculinization of the external
genitalia in males and mild virilization in females (5, 6).
Furthermore, ovarian synthesis of progesterone is required for the
establishment and maintenance of early pregnancy (7) and a reduction in
the duration or concentration of systemic progesterone, luteal phase
insufficiency, is associated with impaired fertility and repeated first
trimester abortion (8, 9).
Numerous studies have demonstrated the role of the transcription factor
steroidogenic factor-1 (SF-1; also called adrenal 4-binding protein;
Ad4BP) in the cAMP-mediated transactivation of cytochrome P450 steroid
hydroxylase genes (10-15) in adrenal and gonadal tissues. This
transcription factor is a member of the steroid hormone receptor
superfamily (11, 16) and is classified as an orphan nuclear receptor
because an endogenous ligand has not been identified. SF-1/Ad4BP and a
closely related isoform, embryonal long terminal repeat-binding protein
(ELP), are transcribed from the same gene (17, 18) and are homologs of
fushi tarazu factor 1 (Ftz-F1), an orphan nuclear
receptor that controls the fushi tarazu homeobox gene in
Drosophila (19). In addition to its regulatory actions on
steroid hydroxylase gene expression, targeted disruption of the
Ftz-F1 gene in mice proved this transcription factor to be
essential for sexual differentiation and development of the adrenal
gland and gonads (20) and critical for normal development of the
ventromedial hypothalamus and pituitary gonadotrophs (21).
Interestingly, this nuclear receptor mediates transcriptional control
over a number of genes that are involved in various aspects of
reproductive function including the -subunit of pituitary glycoprotein hormones (22), Müllerian inhibiting substance (23),
and oxytocin (24) genes. Additionally, pituitaries of Ftz-F1-disrupted mice lack transcripts for
gonadotropin-releasing hormone receptor, as well as
-subunits of
luteinizing hormone and follicle-stimulating hormone (25), all of which
are requisite for reproductive competence.
Although it is well established that 3-HSD is vital for the
synthesis of essential adrenal glucocorticoid and mineralocorticoid hormones, as well as gonadal production of progesterone, estrogens, and
androgens (26), research aimed at elucidating the factors that regulate
expression of the type II 3
-HSD gene has lagged behind that of the
cytochrome P-450 steroid hydroxylase enzymes. Human adrenocortical
carcinoma (H295R) cells have provided a good physiological model in
which to study 3
-HSD-II gene regulation, in that they express the
type II 3
-HSD isozyme and treatment of the cells with either
angiotensin II or phorbol ester results in enhanced expression of
3
-HSD mRNA and synthesis of aldosterone, presumably through
activation of protein kinase C (PKC; Refs. 27 and 28). The objective of
the present study was to identify the specific regions of the type II
3
-HSD promoter that confer basal and phorbol ester-induced
regulation of transcription in adrenal cortical cells as a first step
toward elucidating the response elements and transcription factors
involved.
H295R (human adrenocortical tumor; Ref. 27) cells were cultured in Dulbecco's modified Eagle's medium/Ham's F-12 medium (1:1, 15 mM HEPES; Life Technologies, Inc.) supplemented with 2% fetal calf serum (HyClone Laboratories, Logan, UT) and insulin (6.25 mg/ml), transferrin (6.25 mg/ml), selenium (6.25 ng/ml), and linoleic acid (5.35 mg/ml; 1% ITS+, Collaborative Research, Bedford, MA). HeLa (human cervical carcinoma) cells were cultured in Dulbecco's modified Eagle's medium/F-12 with 10% fetal calf serum. All media contained 50 µg/ml gentamicin (Sigma). Phorbol ester (phorbol 12-myristate 13-acetate (PMA), Sigma) treatment and control media supplied to cells during transient transfection, or cells destined for nuclear extract preparation or RNA isolation contained 200 nM PMA (added as 10,000-fold stock in dimethyl sulfoxide; Me2SO) or an equivalent amount of Me2SO as carrier, respectively.
Plasmids and Reporter Plasmid ConstructionPlasmid
(BlueScript KS II+ vector; Stratagene, San Diego, CA)
containing 1251 bp of 5-flanking and 820 bp of downstream sequence (
1251 to +820 bp; relative to the transcriptional start site) of the
h3
HSD-II gene (3) subcloned into the HindIII site was a
kind gift of Dr. Van Luu-The, Laval University, Quebec, Canada. Initially, sequence from
1251 to +45 bp was amplified by the polymerase chain reaction (PCR) using Taq polymerase
(Promega, Madison, WI), a 5
oligonucleotide primer identical to the
published T7 promoter sequence in the BlueScript vector and a 3
oligonucleotide primer that spanned from +27 to +45 bp of untranslated
exon I and was designed to contain a penultimate HindIII
site. Following agarose gel purification, the
1251 to +45 bp fragment
was subcloned into the PCR II vector (InVitrogen, San Diego, CA),
subjected to HindIII digestion, and ultimately fused to the
chloramphenicol acetyltransferase (CAT) gene in the pCAT-Basic vector
(
1251 CAT; Promega).
Progressive 5 deletion mutants of the
1251 to +45 bp fragment of the
h3
HSD-II gene were generated using PCR. Five 5
oligonucleotide primers were used in separate reactions with the previously described 3
oligonucleotide primer to generate five constructs with sequence progressively deleted from the 5
end of the promoter. The five primer
sequences were identical to
1051 to
1028 (
1051 CAT),
701 to
678 (
701 CAT),
301 to
278 (
301 CAT),
101 to
78 (
101
CAT), and
52 to
32 (
52 CAT) bp of the promoter. All
oligonucleotides contained a penultimate HindIII restriction
endonuclease site to facilitate subcloning. Following PCR, expected
sizes of the promoter fragments were confirmed by agarose gel
electrophoresis, extracted from the gel, and subcloned into the PCR II
vector. Clones containing the five promoter fragments were digested
with HindIII and the resulting fragments subjected to
agarose gel purification prior to ligation into the pCAT-basic vector.
All constructs were confirmed by sequencing through the ligation sites
using the dideoxy chain-termination method (29) and Sequenase version
2.0 DNA sequencing kit (United States Biochemical Corp.).
The 101 mutant CAT promoter construct (
101M CAT) was generated by
synthesizing the top and complementary DNA strands (
101 to +1 bp) in
three sections that, when annealed, yielded 3 double-stranded oligonucleotides containing 6-bp overhangs to facilitate ligation. The
putative SF-1 response element was mutated by substituting two
thymidine residues for two guanine residues (underlined) in the
sequence so that TCAA
TAA became TCAA
TAA.
Both 5
and 3
ends of the construct were designed to contain
HindIII restriction sites to facilitate subcloning into the
pCAT-basic vector. The entire
101 mutant promoter CAT construct was
sequenced in both directions to confirm mutation of the putative
SF-1/Ad4BP response element.
The pRC/RSV expression vector (InVitrogen) containing the full-length SF-1/Ad4BP cDNA (RSV-SF-1/Ad4BP) was described previously (30). The pSV2-luciferase (pSV2-luc) plasmid was used to normalize for differences in transfection efficiency among samples. All reporter constructs and expression vectors were prepared for transfection using the Qiagen Plasmid Mega kit (Qiagen Inc., Chatsworth, CA).
Transient TransfectionsH295R and HeLa cells were
transiently transfected using a modification of the calcium phosphate
co-precipitation method (31). Briefly, adherent H295R and HeLa cells
were cultured to 55-65% confluency in 100-mm tissue culture dishes
(Corning Scientific Products, Corning, NY) in 10 ml of the appropriate
medium. Calcium phosphate-DNA co-precipitates were formed by dropwise
addition of equal volumes (0.5 ml) of solution A (0.24 M
CaCl2 containing 10 µg of promoter-CAT construct, 10 µg
of control pGEM-3Z plasmid (Promega), plus 2 µg of pSV2-luc plasmid
DNA for H295R cells and 10 µg of promoter-CAT construct, 10 µg of
RSV-SF-1/Ad4BP, and 2 µg of pSV2-luc plasmid DNA for HeLa cells) to
Solution B (2 × Hepes-buffered saline; 50 mM Hepes, 1.4 mM Na2HPO4, 0.28 M NaCl (pH 7.1)). Calcium phosphate:DNA precipitates were incubated at 23 °C for at least 20 min and added to single 100-mm dishes of cells
containing 9 ml of fresh medium. H295R and HeLa cells were incubated
with precipitate for 4 h at 37 °C (5% CO2 and 95%
air), shocked for 3 min or 30 s, respectively, with 15% (v/v)
glycerol in Dulbecco's phosphate-buffered saline (D-PBS; 0.137 M NaCl, 0.137 M NaCl, 0.5 mM
MgCl2, 6.45 mM Na2HPO4,
1.5 mM KH2PO4), washed with D-PBS,
and incubated at 37 °C for 24 or 36 h, respectively. During the
final 24 h of incubation, cells were cultured in the presence or
absence of PMA or carrier as described previously. Cells were harvested
using trypsin/EDTA (Life Technologies, Inc.), pelleted, resuspended in
0.25 M Tris-HCl (pH 7.4), and stored at 70 °C until
assayed for CAT activity. Transfections were performed in triplicate
with mock (no plasmid) serving as negative controls. Experiments were
repeated identically at least twice and at least three times with
modifications.
Frozen cell pellets were thawed
on ice and lysed by sonication. Soluble extract was separated from cell
debris by centrifugation, divided into aliquots for CAT and luciferase
assays, and stored at 70 °C prior to use. Prior to CAT assay,
extracts were heated to 60 °C for 5 min to denature any endogenous
acetylase/deacetylase enzymes. Fluorescent CAT assays were performed as
described (32) with some modification using the FLASH CAT assay kit
(Stratagene). Acetyl coenzyme A (CoA) was synthesized by reaction of
CoA (Pharmacia Biotech Inc.) with acetic anhydride
(Sigma) as described elsewhere (33) and stored at
70 °C until use. H295R cell extracts (45 µl) and HeLa cell
extracts (1 µl of a 1:10 dilution) were incubated in 0.25 M Tris-HCl (pH 7.4) in a total reaction volume of 125 µl
with acetyl-CoA (8.2 mM) and fluorescent
borondipyrromethene difluoride (BODIPY) chloramphenicol (CAM) substrate
(1:12.5 dilution) at 37 °C for 8 and 1 h, respectively.
Reactions were terminated by addition of cold ethyl acetate (850 µl),
followed by vigorous vortexing. An aliquot (800 µl) of extracted
substrate and acetylated products was removed (organic phase), dried
under vacuum, and resuspended in ethyl acetate (20 µl) prior to
separation on thin-layer chromatography plates (LK6, Whatman, Clifton,
NJ) with chloroform:methanol (9:1) for 30 min. Substrate and products
were visualized under long-wave UV light (366 nm) and photographed
(Type 55 positive/negative film, Polaroid, Cambridge, MA). Substrate
and combined product bands were scraped from the plates, extracted and
diluted 1:10 in methanol prior to quantification by fluorescence
spectrophotometry at excitation and emission wavelengths of 490 nm and
512 nm, respectively, using a fluorometer. Percent conversion of BODIPY
CAM substrate to 1-, 3-, and 1,3-acetylated BODIPY CAM products was
computed after correcting samples for background activity.
Differences in transfection efficiency between samples were monitored using a portion of the cell extracts for luciferase assays (34) with minor modification. Briefly, 10 µl of cell extract was added to 350 µl of reaction buffer (25 mM glycylglycine, 5 mM ATP, and 15 mM MgSO4 (pH 7.6)). Luciferin (1 mM; Boehringer Mannheim) was injected into the reaction and relative light output determined using a Monolight 2010 luminometer (Analytical Luminescence Labs, San Diego, CA). Transfection efficiency did not vary significantly between triplicate culture dishes within any single treatment group or between treatment groups.
Preparation of Nuclear ExtractsSeveral 225-cm2 flasks of H295R cells were maintained as described previously. At 55-65% confluency, cells were cultured in the presence or absence of 200 nM PMA for 24 h at which time cells were harvested using trypsin/EDTA, pelleted, washed with D-PBS, pelleted and resuspended in D-PBS (1.0 ml), and maintained on ice. Crude nuclear extracts were prepared according to the method of Dignam and co-workers (35) as modified by Andrews and Faller (36). Briefly, after pelleting and removal of D-PBS, cells were allowed to swell for 10 min on ice in a hypotonic buffer (10 mM HEPES-KOH (pH 7.9) at 4 °C, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride) followed by vortexing and centrifugation. The supernatant was discarded and the pellet resuspended in a high salt buffer (20 mM HEPES-KOH (pH 7.9), 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride) for 20 min to extract nuclear proteins. The suspension was centrifuged, and the supernatant containing nuclear proteins was aliquoted and stored in liquid nitrogen. Protein concentrations were determined using the BCA method (Pierce) as modified for the presence of sulfhydryl reagents (37).
Electrophoretic Mobility Shift AssaysElectrophoretic
mobility shift assays (EMSA) were performed as described (38) with some
modification. Single-stranded oligonucleotides containing the putative
SF-1/Ad4BP element (underlined) and 10-11 bases of 5 and 3
flanking
3
HSD-II promoter sequence
(5
-GTGGCAGGAGT
TAAGGGCTGA-3
and opposite strand
5
-TCAGCCCTTA
ACTCCTGCCAC-3
) were synthesized
using an automatic DNA synthesizer (Oligo 1000, Beckman Instruments
Inc., Palo Alto, CA). Double-stranded SF-1/Ad4BP probe was prepared by
annealing 50 ng of each oligonucletide strand for 2 min at 65 °C,
followed by slow cooling to room temperature. The probe was end-labeled
using [
-32P]ATP (3000 Ci/mmol; DuPont NEN) and T4
polynucleotide kinase (New England Biolabs, Beverly, MA) and purified
using G-50 spun columns as described (39) or ProbeQuant G-50 Micro
Columns (Pharmacia Biotech Inc.) following the supplied protocol.
Nuclear extracts (20 µg) from control and PMA-treated H295R cells
were preincubated in the presence (1 µl) or absence of SF-1/Ad4BP
antiserum (30) or as a nonspecific control, antiserum against a
secreted epididymal sperm-binding protein (40) for 30 min on ice prior
to the addition of poly(dI·dC)·poly(dI·dC) (2 µg, Pharmacia
Biotech Inc.) in 15.0 mM HEPES (pH 7.9), 50 mM
KCl, 42 mM NaCl, 0.15 mM MgCl2, 1 mM EDTA, 1 mM dithiothreitol, 2.5% glycerol,
4% Ficoll, and 32P-labeled SF-1/Ad4BP oligonucleotide
(~4 × 104 cpm) to a final reaction volume of 20 µl and incubated for an additional 30 min on ice. In additional
competition experiments, reactions contained unlabeled, double-stranded
SF-1/Ad4BP oligonucleotide (50 or 500 molar excess) or double-stranded
oligonucleotide (50 or 500 molar excess) containing a mutated form of
the putative SF-1/Ad4BP element (used to generate the
101
mutant CAT construct, top strand:
5
-ATGTGGCAGGAGT
TAAGGGCTGAGACACAA-3
),
prolactin-inducible element (PIE; Ref. 41) or
-activated site
(GAS; Ref. 42) to assess the specificity of DNA binding. DNA-protein
complexes were resolved using native PAGE (5%;
acrylamide:bisacrylamide, 37.5:1) with 0.5 × Tris borate-EDTA
(44.5 mM Tris, 44.5 mM boric acid, 1 mM EDTA) for 2 h at 150 V. Gels were dried under
vacuum at 70 °C for 1 h and exposed to Kodak BioMax MR film
(Eastman Kodak, Rochester, NY) for 29 h at 23 °C.
Prior to
isolation of mRNA, four 225-cm2 flasks of H295R were
cultured to 65% confluency, at which time cells were treated with
media containing 200 nM PMA or an equivalent amount of
Me2SO for 24 h. In addition, one 75-cm2
flask of untreated HeLa and two 75-cm2 flasks of mouse F9
embryonic teratocarcinoma cells were cultured until cells were
confluent. Cells were harvested using trypsin/EDTA, and the resulting
pellets were stored at 70 °C.
Isolation of enriched polyadenylated (poly(A)+) mRNA
was performed using a messenger RNA isolation kit (Stratagene)
following the protocol, including the poly(A)+ enrichment
procedure, provided by the manufacturer. After elution from
oligo(dT)-cellulose push columns, the poly(A)+-enriched
mRNA was ethanol/sodium acetate-precipitated overnight at
20 °C using glycogen as a carrier. Following centrifugation, mRNA samples were resuspended in a small volume of elution buffer, aliquoted, and stored at
20 °C in an excess of absolute ethanol.
Northern analysis was conducted using Me2SO-glyoxal RNA
sample preparation and agarose gel electrophoresis as described (39) with some modification. Samples (0.5 or 1.0 µg) or RNA standard (Life
Technologies, Inc.) were denatured in 1.08 M deionized
glyoxal, 54% (v/v) Me2SO, and 10 mM sodium
phosphate (pH 7.0) in a reaction volume of 25 µl for 1 h at
55 °C and resolved using agarose (1.5% w/v in 10 mM
sodium phosphate (pH 7.0) and 10 mM iodoacetic acid) gel
electrophoresis with 10 mM sodium phosphate (pH 7.0). RNA was transferred to a nylon membrane (Duralon-UV; Stratagene) by capillary action with 20 × SSC (1 × SSC = 0.15 M NaCl, 15 mM Na3 citrate) and
fixed by UV-cross-linking and baking for 2 h at 80 °C. The blot
was de-glyoxalated in 0.1 M sodium phosphate (pH 8.0) for
1 h at 65 °C, followed by prehybridization in 100 ml of
hybridization buffer (50% v/v deionized formamide, 5 × SSC,
0.025 M sodium phosphate buffer (pH 7.0), 5% w/v SDS,
0.1% w/v BSA, 0.1% w/v Ficoll 400, 0.1% w/v polyvinylpyrrolidone,
and 0.05% w/v fish sperm DNA) for 1 h at 42 °C prior to
addition of 50 µCi of [-32P]dATP-labeled cDNA
probe. SF-1/Ad4BP probe was generated using random primers,
exo
Klenow fragment (Prime-it II kit, Stratagene) and a
~1.7-kilobase pair fragment obtained by SacI digest of the
RSV-SF-1/Ad4BP plasmid. After hybridization (19 h), the blot was washed
once with 2 × SSC, 1% (w/v) SDS at 45 °C for 15 min; twice
with 2 × SSC, 0.1% (w/v) SDS at 45 °C for 15 min; and once
with 0.1 × SSC, 0.1% (w/v) SDS for 15 min at 60 °C.
Autoradiography was performed at
70 °C with intensifying screens
using Kodak X-Omat AR or BioMax MR film. Following autoradiography,
blots were stripped of probe by washing for 1 h at 80 °C in 0.2 mM EDTA, 0.05% (w/v) pyrophosphate, 0.1% (w/v) SDS, 0.2%
(w/v) BSA, 0.2% (w/v) Ficoll 400, and 0.2% (w/v) polyvinylpyrrolidone
and rehybridized (19 h) with a [
-32P]dATP-labeled
constitutive probe (625-bp fragment of CHO-B; ribosomal protein S2;
Refs. 43 and 44) generously provided by Dr. Kelly Mayo, Northwestern
University, Evanston, IL. After hybridization, the blot was washed as
described above, except that the duration of the final wash was 10 min.
Autoradiographs and Polaroid negatives were scanned using an Agfa Arcus II flatbed scanner and Adobe Photoshop software. Digitized images were saved as TIFF files, and all cropping and text enhancements were carried out using Aldus Pagemaker or Freehand programs.
To determine the region(s) of the promoter that confer phorbol
ester-mediated transcriptional regulation of the type II 3-HSD gene,
a series of 5
deletions of the promoter fused to a CAT reporter gene
were used to transiently transfect human adrenocortical carcinoma
(H295R) cells followed by treatment for 24 h in the absence
(basal) or presence of PMA. As shown in Fig. 1, deletion of the promoter sequence from
101 to
52 bp resulted in a nearly 4-fold and 500-fold reduction in basal and PMA-stimulated CAT activity,
respectively, suggesting that this region is important for both basal
and PMA-induced regulation of type II 3
-HSD expression. Additionally, the region from
1251 to
301 bp may contain a negative regulatory element that inhibits PMA-induced transcription since an
approximately 5-fold increase in CAT activity was observed following
the deletion of
1251 to
301 bp of the promoter.
A subsequent search of the 101 to
52 bp promoter sequence revealed
a putative SF-1/Ad4BP regulatory element, TCAAGGT
A, at
64 to
56 bp that differed from the reported SF-1 (PyCAAGGPyCPu; Ref. 13) and Ad4BP (C/TCAAGGT/CC/T; Ref. 10) elements by a single
nucleotide (underlined). To determine if the putative SF-1/Ad4BP element was functional, non-steroidogenic HeLa cells, which do not
express SF-1/Ad4BP, were transiently cotransfected with the
101 CAT,
52 CAT, and
101M CAT constructs in the presence and absence of an
expression vector for SF-1/Ad4BP (RSV-SF-1/Ad4BP). In addition,
transfected cells were cultured for 24 h in the presence or
absence of PMA (200 nM) to investigate the ability of
phorbol ester to modulate SF-1/Ad4BP-mediated activity of the 3
-HSD
promoter (Fig. 2).
Cotransfection of HeLa cells with 101 CAT and RSV-SF-1/Ad4BP
increased CAT activity 49-fold over basal, whereas only small increases
were observed (6- and 2-fold, respectively) in those cells transfected
with
52 CAT, which lacked the putative SF-1/Ad4BP element, or
101M
CAT, in which two nucleotides in the putative response element had been
mutated. PMA treatment increased CAT activity 20-fold over basal in
those cells transfected with
101 CAT alone. PMA had a similar effect
on cells transfected with
52 CAT (17-fold increase); however, CAT
activity in response to PMA was considerably less in those cells
transfected with
101M CAT (5-fold). Interestingly, an unexpectedly
powerful synergistic effect of SF-1/Ad4BP and PMA on CAT activity was
observed in cells cotransfected with
101 CAT and RSV-SF-1/Ad4BP and
treated with PMA. Overexpression of SF-1/Ad4BP in HeLa cells
transfected with
101 CAT and subsequently treated with PMA increased
promoter activity 540-fold as compared with basal. An increase was
observed (16-fold) in similarly treated cells transfected with the
52 CAT construct plus RSV-SF-1/Ad4BP and was most likely due to the action
of PMA alone. No similar increase was observed with the
101M CAT
construct. Collectively, these data suggest that SF-1/Ad4BP can
specifically regulate the promoter of the human type II 3
-HSD gene
and that SF-1/Ad4BP-induced expression of the gene in HeLa cells is
greatly enhanced by phorbol ester.
Although phorbol ester activation of the human type II 3-HSD
promoter in non-steroidogenic HeLa cells appeared to be primarily mediated by SF-1, it was also necessary to evaluate the response in a
physiologically relevant human steroidogenic cell line. To address this
issue, adrenocortical carcinoma (H295R) cells were transfected with the
101,
52, and
101M CAT constructs followed by treatment with or
without phorbol ester. As shown in Fig. 3, PMA treatment
of H295R cells transfected with the
101 CAT construct containing the
putative SF-1 regulatory element increased promoter activity 18-fold
over basal. In contrast, deletion (
52 CAT) or mutation (
101M CAT)
of the putative SF-1 regulatory element dramatically reduced PMA
responsiveness to 3-fold and 2-fold over that for basal, respectively.
Collectively, these data and those from the preceding HeLa cell
experiments clearly demonstrate the importance of the SF-1 regulatory
element in mediating phorbol ester regulation of the human type II
3
-HSD promoter in both steroidogenic and non-steroidogenic
cells.
Interaction of SF-l with the putative regulatory element was examined
by EMSA using nuclear extracts prepared from H295R cells cultured in
the presence or absence of PMA and a 32P-labeled
double-stranded oligonucleotide containing the putative SF-1/Ad4BP site
and 10-11 bases of 5 and 3
flanking type II 3
-HSD promoter
sequence. As shown in Fig. 4, multiple protein-DNA complexes were formed when nuclear extracts from untreated and PMA-treated H295R cells were incubated with the oligonucleotide probe.
However, when extracts were incubated in the presence of increasing
concentrations of unlabeled oligonucleotide, the appearance of one
complex was markedly diminished as compared with the others. Preincubation of the extracts with SF-1/Ad4BP antiserum abolished the
formation of that particular complex, confirming the participation of
SF-1/Ad4BP. Additionally, the intensity of the band representing the
SF-1/Ad4BP specific protein-DNA complex appeared to be slightly darker
for control versus PMA treatment. These data may be
interpreted to suggest reduced DNA binding in those extracts derived
from cells treated with PMA as compared with control. However, given that the cells were harvested at only one time point and that the
kinetics of PMA-induced expression of SF-1/Ad4BP mRNA are currently
unknown, the results of this comparison appear to be of limited
value.
To further test the specificity of DNA binding, nuclear extracts from
untreated H295R cells were incubated in the presence of a nonspecific
control consisting of antiserum against a secreted epididymal protein
in rats or increasing concentrations of unlabeled heterologous
oligonucleotides (containing GAS or PIE response elements) or unlabeled
oligonucleotide (101M) containing a mutated form of the putative
SF-1/Ad4BP response element flanked by type II 3
-HSD promoter
sequence (Fig. 5). Incubation of nuclear extracts with
nonspecific antibody or increasing concentrations of unlabeled
101M,
GAS, or PIE oligonucleotides failed to diminish the appearance of the
complex that was abolished by SF-1/Ad4BP antiserum, suggesting that
formation of this protein-DNA complex is specific to SF-1/Ad4BP.
Finally, Northern analysis was used to determine if SF-1/Ad4BP was
expressed in H295R, HeLa, and mouse F9 teratocarcinoma cells. As shown
in Fig. 6, SF-1/Ad4BP transcripts were detected in H295R
cells and expression appeared to increase slightly upon treatment
with PMA. In addition, a slightly smaller transcript was detected in F9
cells after longer autoradiography times. As anticipated, no
transcripts for SF-1/Ad4BP were detected in HeLa cells. Collectively,
results of the EMSA and Northern analysis demonstrate that SF-1/Ad4BP
is expressed in adrenal cortical carcinoma cells and interacts with the
putative SF-1/Ad4BP regulatory element present in the promoter of the
human type II 3-HSD gene.
In the present study we sought to determine regions of the human
type II 3-HSD promoter important for regulation using a physiologically relevant cell line, H295R adrenal cortical carcinoma cells (27, 28). Using deletion mutagenesis we determined that the
region from
101 to
52 bp of the promoter was essential for PMA-mediated transcription of the reporter gene and contained a
putative SF-1/Ad4BP regulatory element TCAAGGTAA from
64 to
52 bp.
Interestingly, the human type I 3
-HSD gene is also regulated by PMA
(45); however, it does not contain a functional SF-1/Ad4BP element
(TCAAAGTGA; Ref. 46) because it differs from the consensus element by
one critical core nucleotide (10). As a result, the mechanism
conferring PMA responsiveness to the type I 3
-HSD gene, which is
predominantly expressed in the placenta and skin, most likely differs
from that of the type II isoform. The putative SF-1/Ad4BP element found
in the type II 3
-HSD promoter is another variant of the shared motif
AGGTCA, the core binding sequence of several zinc-finger DNA-binding
proteins including the estrogen, thyroid hormone, retinoic acid, and
vitamin D3 receptors (47). This sequence has been reported
to confer cAMP responsiveness to a number of the cytochrome P450
steroid hydroxylase genes in gonadal (13, 14) as well as adrenal (15,
48, 49) tissues. In these tissues, steroid biosynthesis is regulated by
the interaction of luteinizing and/or follicle-stimulating hormone or
adrenocorticotrophic hormone, respectively, with their cognate cell
membrane receptors, resulting in elevated production of cAMP and
enhanced tissue-specific expression of the appropriate steroidogenic
enzyme.
Expression of 3-HSD and synthesis of aldosterone in H295R cells are
regulated through the action of angiotensin II via the type I
angiotensin II receptor coupled to polyphosphoinositidase-C and
subsequent increases in intracellular calcium (28). Because this effect
can be mimicked by phorbol ester (27), it is presumed to be mediated
via protein kinase C. Our results are consistent with PKC-mediated
regulation of 3
-HSD expression in H295R cells. PMA treatment of
H295R cells transfected with a series of 5
deletion mutants of the
3
-HSD promoter increased reporter gene activity, for all constructs
except
52 CAT, greater than that observed for untreated transfected
cells. The novel aspect of this finding is that the PMA-induced
increase in transcriptional activity appeared to be mediated by
SF-1/Ad4BP. Deletion of promoter sequence from
101 to
52 bp, later
found to contain a putative SF-1/Ad4BP regulatory element, abolished
PMA-stimulated CAT activity. Additionaly, mutation of the putative
SF-1/Ad4BP regulatory element also inhibited PMA-induced promoter
activation. To date, no similar effect of phorbol ester has been
reported for any of the genes known to be regulated by SF-1/Ad4BP.
Cotransfection of HeLa cells with the 101 CAT construct and
RSV-SF-1/Ad4BP expression vector yielded a significant 49-fold increase
in transcriptional activity due to SF-1/Ad4BP alone, as well as a
distinct synergistic effect of PMA on SF-1/Ad4BP-mediated transcription. The 49-fold increase in CAT activity in the absence of
any other treatment is greater than that reported for isolated SF-1/Ad4BP elements from other genes similarly cotransfected into non-steroidogenic cell lines (13, 15, 23, 24, 49). SF-1/Ad4BP alone was
unable to activate MIS gene expression in HeLa cells, whereas
coexpression of a mutant form of SF-1/Ad4BP lacking the putative ligand
binding domain slightly increased transcriptional activity (23),
suggesting that ligand or a cofactor is necessary for activation of the
MIS promoter by SF-1/Ad4BP. In the case of some steroid hydroxylase
genes, the addition of protein kinase A activators such as cAMP and
forskolin or coexpression of the catalytic subunit of protein kinase A
is necessary for appreciable activation by SF-1/Ad4BP in
nonsteroidogenic cells. In our case, high levels of transcriptional
activity due to SF-1/Ad4BP alone may result from promoter-specific
sequence that confers a greater sensitivity of the type II 3
-HSD
gene to stimulation by SF-1/Ad4BP. A number of reports have suggested
that nucleotides 5
of the response elements for SF-1/Ad4BP and another
orphan nuclear receptor, nerve growth factor-inducible factor B
(NGFI-B), may influence binding affinity (10, 15, 23, 50).
Additionally, a brief survey of reported functional SF-1/Ad4BP response
elements in comparison to that for type II 3
-HSD indicates that the
putative 3
-HSD SF-1/Ad4BP element is a unique variant of the AGGTCA
consensus motif in that the penultimate cytidine at the 3
end of the
sequence has been replaced with an adenine resulting in AGGTAA, which
has not been reported for other genes. This difference could be in part
responsible for the enhanced SF-1/Ad4BP activation of 3
-HSD gene
expression in HeLa cells.
The mechanism underlying the stimulatory effect of PMA on 3-HSD
promoter activity in HeLa cells in the absence of SF-1/Ad4BP is less
clear and may be independent of a functional SF-1/Ad4BP response
element because removal of the element,
52 CAT, failed to reduce CAT
activity in response to PMA. It is possible that a yet unidentified
regulatory element is present in the sequence from
52 to +45 bp of
the promoter and perhaps PMA induces the synthesis of a HeLa
cell-specific regulatory protein(s) that is responsive to PMA in the
absence of SF-1/Ad4BP. Loss of PMA-induced CAT activity after mutation,
but not removal, of the putative SF-1/Ad4BP response element (
101M
CAT) is seemingly more complex. However, it is possible that the
PMA-responsive region of the promoter may partially overlap that for
SF-1/Ad4BP binding and that the loss of several nucleotides from the
5
-end of the PMA responsive sequence may be well tolerated whereas
mutation of several of those nucleotides severely compromises PMA
responsiveness. Interestingly, HeLa cells express an orphan nuclear
receptor, chicken ovalbumin upstream promoter transcription factor
(COUP-TF; Ref. 51), that has been reported to bind to recognition
elements that partially and totally encompass the SF-1/Ad4BP elements
in the 17
-hydroxylase (15) and oxytocin (24) promoters,
respectively.
The highly synergistic effect of PMA on SF-1/Ad4BP activation of the
3-HSD promoter in HeLa cells has not been reported previously, and
the mechanism responsible for this cell specific response is presently
unknown. Because PMA is a potent activator of PKC and SF-1/Ad4BP has 10 consensus PKC phosphorylation sites by sequence homology (16) and is a
phosphoprotein,2 it is tempting to
speculate that the profound response to PMA is due, at least in part,
to direct phosphorylation of the transcription factor. Alternatively,
it is possible that treatment with PMA induces the production and(or)
phosphorylation of a HeLa cell factor, that significantly augments
SF-1/Ad4BP activation of the type II 3
-HSD promoter. Synergistic
effects of phorbol ester and cholera toxin on estradiol-stimulated
transcription of a synthetic estrogen-responsive reporter gene were
also found to be cell-specific and appeared to result from the
stabilization or facilitation of the receptor with components of the
transcriptional apparatus, possibly as a result of phosphorylation of
the receptor or other necessary proteins (52). It is also possible that
PMA provokes synthesis of an undiscovered ligand for SF-1/Ad4BP in
nonsteroidogenic HeLa cells. Recent data indicate that SF-1/Ad4BP
expression and action are not limited to steroidogenic tissues and the
regulation of steroidogenic enzymes (22-25). Thus, it is possible that
the undiscovered ligand for SF-1/Ad4BP is not a steroid, as has been hypothesized (15, 17, 53); instead, it may be a molecule that is
present in steroidogenic as well as nonsteroidogenic tissues, with
specificity of activation relying solely on the tissue-specific expression of SF-1/Ad4BP.
There is evidence to indicate that other members of the steroid/thyroid
hormone receptor superfamily, normally activated by ligand binding, may
elicit their actions at steroid-responsive regulatory elements through
ligand-independent processes that involve cross-talk between
membrane-bound receptor signaling pathways and the specific nuclear
steroid receptor. Dopamine can activate progesterone, estrogen (ER),
vitamin D, and thyroid hormone receptor-, but not glucocorticoid,
receptor-mediated activation of target response elements in transfected
cells in the absence of steroid ligand and in the case of progesterone
receptor, activation required the presence of a specific serine
phosphorylation site on the receptor (54). More recently, epidermal
growth factor (EGF) was reported to activate estrogen-independent
transcription of a consensus estrogen response element cotransfected
with an expression vector for the mouse ER into human endometrial
adenocarcinoma cells (55). In the presence of estrogen, EGF had a
synergistic effect on transcription. Although not determined, ER
phosphorylation and(or) activation of other regulatory proteins were
hypothesized as plausible mechanisms mediating the effects of EGF.
In HeLa cells, co-transfection of SF-1/Ad4BP followed by treatment with PMA results in promoter activity levels that greatly exceed the additive effect of each treatment alone and fulfills the criteria for synergism in transcriptional activation as discussed by Herschlag and Johnson (56). Therefore, the synergism implies that the two activating mechanisms function in the same pathway. While these high levels of activation might occur without physical interaction between the two activating agents, it is also highly likely that the activation of protein kinase C isoforms by phorbol ester treatment may alter the phosphorylation state of SF-1 with a corresponding change in activity.
Our investigation of basal and phorbol ester-mediated regulation of
type II 3-HSD gene expression in human adrenocortical carcinoma
cells has resulted in several novel findings. Both basal and
PMA-induced transcription of the gene in H295 cells required the
presence of promoter sequence containing an SF-1/Ad4BP recognition element. This is the first demonstration of SF-1/Ad4BP-mediated regulation of a non-cytochrome P450 steroidogenic enzyme promoter. Cotransfection of the reporter construct containing the putative SF-1/Ad4BP element or a mutated version of the element and an expression vector for SF-1/Ad4BP into HeLa cells confirmed the essentiality and functionality of this response element in a
non-steroidogenic cell line. We also discovered a previously unreported
synergistic effect of PMA on the regulation of transcription by
SF-1/Ad4BP. Additionally, results of EMSA with SF-1/Ad4BP antiserum
indicated that H295R cell nuclear extracts contained SF-1/Ad4BP, which
specifically interacted with the putative 3
-HSD SF-1/Ad4BP response
element. Northern analysis also confirmed the presence of SF-1/Ad4BP
transcripts in H295R but not HeLa cells. Collectively, these data
provide considerable evidence to support a role for SF-1/Ad4BP in the regulation of the type II 3
-HSD gene in adrenal cortical cells. However, further research is needed to more clearly define the mechanisms underlying phorbol ester-mediated regulation of this gene in
both H295R and HeLa cells.