| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
IHF Institute for Hormone and Fertility Research, University of Hamburg, 22529 Hamburg, Germany; and the Laboratory of Biochemistry, National Cancer Institute, National Institutes of Health (C.V.), Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Birgit Gellersen, Ph.D., IHF Institute for Hormone and Fertility Research, Grandweg 64, 22529 Hamburg, Germany. E-mail: gellersen{at}ihf.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and PGE2 and therefore serves as an important determinant
in regulating their local concentrations. To gain insights into the
transcriptional regulation of this enzyme, we have isolated 3.5 kb of
the 5'-flanking sequence of the human PGDH promoter and characterized
its control in hemopoietic cells and cells of myometrial and placental
origin. Several potential binding sites for cAMP-responsive
element-binding protein (CREB), Ets, and activating protein-1 (AP-1)
transcription factors are present within 2368 bp of the 5'-flanking
region. This region and deletions thereof were fused to the luciferase
reporter gene and used for transient transfection experiments. In
Jurkat leukemic T cells, which express PGDH endogenously, the
transfected PGDH promoter was strongly induced by phorbol ester.
Induction was reversed by coexpression of A-Fos, a dominant negative to
AP-1. In primary cultures of myometrial smooth muscle cells (SMC), the
Ets family members Ets-1, Ets-2, and PEA3 potently stimulated
transcriptional activity of the PGDH promoter. PEA3-mediated activation
was partially repressed by A-Fos, suggesting an involvement of AP-1
proteins, which might be conferred by a distal and a proximal Ets/AP-1
composite element. The distal Ets/AP-1 element is flanked by two
CRE-like sequences. Cotransfection of A-CREB, a dominant negative to
CREB, inhibited stimulation of PGDH-2368/luc3 by PEA3 in myometrial
SMC, whereas treatment with 8-bromo-cAMP moderately enhanced promoter
activity. Progesterone is believed to be an important stimulus for PGDH
expression in the utero-placental unit, thus contributing to the
maintenance of a quiescent uterus during pregnancy. In myometrial SMC,
both isoforms of the progesterone receptor, PR-B and PR-A, caused a
ligand-dependent activation of PGDH-2368/luc3. Transcriptional activity
of PR-B, but not PR-A, was further enhanced by the addition of
8-bromo-cAMP. We could not confirm a recently proposed transcriptional
control of PGDH by mineralocorticoid receptor. No effect of
mineralocorticoid receptor, in the absence or presence of aldosterone,
with or without 8-bromo-cAMP, was observed on PGDH-2368/luc3. Taken
together, these findings demonstrate control of the PGDH promoter by
multiple pathways and provide evidence for cross-talk among Ets, AP-1,
cAMP, and PR-mediated signaling, suggesting complex regulatory
mechanisms for the expression of PGDH. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
is a vasoconstrictor
and strong stimulant of smooth muscle contraction, whereas
PGE2 causes vasodilation and can relax or excite
smooth muscle depending on which of its receptor subtypes are present
(7, 8). PGF2 and PGE2
are rapidly catabolized in vivo into their biologically
inactive 13,14-dihydro-15-keto metabolites, PGFM and PGEM,
respectively. This conversion is a two-step process carried out
sequentially by the enzymes NAD+-dependent
15-hydroxyprostaglandin dehydrogenase (PGDH; also known as type-I
15-PGDH; EC 1.1.1.141), catalyzing the reversible oxidation of the
15-hydroxyl group of PGs, and PG 
13-reductase
(9, 10, 11). PGDH belongs to the family of short chain dehydrogenases, as
it contains a region of 20 amino acids that are highly conserved within
this family of enzymes (12). PGDH as a key catabolic enzyme in the
inactivation of PGs serves as an important determinant in regulating
concentrations of biologically active PGs. It is expressed in many
mammalian tissues and has been purified to homogeneity from placenta,
lung, and kidney (12, 13, 14, 15). In the lungs it is responsible for
inactivating PGs in the venous blood before reaching the arterial
circulation and keeps the plasma concentrations of primary PGs
extremely low (<10 pg/ml) (16). In the feto-placental unit, PGDH has
been implicated in the maintenance of pregnancy, as high PGDH activity
in the chorion protects the myometrium from high levels of
PGF2 produced by the amnion (17, 18, 19).
Consistent with the important role of PGs in the inflammatory process,
PGDH is expressed at relatively high levels in cells of hemopoietic
origin (20, 21, 22). The PGDH complementary DNA (cDNA) has been cloned from
human, rat, mouse, and guinea pig (23, 24, 25, 26). In addition to the
full-length form consisting of a 798-bp coding sequence, two
C-terminally truncated forms have been reported with a reduced
translated region (27, 28). The human gene is located on chromosome 4,
but has not been cloned except for a 1.1-kb 3'-portion containing three
exons, which has been sequenced and suggests generation of truncated
messenger RNAs (mRNAs) by alternative splicing (29). Despite the physiological importance of PGDH, little is understood about the cellular or molecular mechanisms involved in regulating its expression. Indirect evidence suggests that uterine PGDH is under the control of progesterone. During the menstrual cycle, PGDH enzymatic activity is higher in endometrium of the secretory compared with the proliferative phase (30). When the antiprogestin onapristone is administered 2 days after the LH surge, there is a pronounced reduction in immunostaining for PGDH within the endometrial glands (31). Women receiving the antigestagen mifepristone (RU486) in early pregnancy have significantly reduced concentrations of PGDH enzyme in decidual tissue (32). Small blood vessels in the decidua of control individuals are negative for PGE2 but positive for the metabolite PGEM, whereas in mifepristone-treated women PGE2 is present, but PGEM is undetectable. This rise of PGE2 in blood vessels is suggested to result from a direct inhibitory effect of the antigestagen on PGDH expression and may be a major reason for the abortive effect of RU486, by synergizing with leukocyte chemotactic cytokines such as IL-8 to stimulate neutrophil infiltration and tissue destruction (33). Antiprogestagenic reduction of PG catabolism sensitizes the uterus by lowering the threshold above which PGs cause myometrial contractions (34). Not only local uterine, but also systemic, inactivation of PGs in the lung has been proposed to be controlled by steroids, as PG inactivation in rabbits is enhanced during pregnancy or progesterone treatment (35).
In early pregnancy, PGDH is abundant in cytotrophoblastic cells but absent from syncytiotrophoblast (32). At term, PGDH immunoreactivity is localized to the syncytiotrophoblast of placenta and the trophoblast layer of the chorion (19). PGDH in the chorionic trophoblast is considered to serve as a protective barrier to prevent PGs synthesized by amnion or chorion from reaching decidua and myometrium. In a subset of patients with preterm labor, in particular in the presence of infection, PGDH immunoreactivity, activity, and mRNA are reduced in chorionic trophoblast (19, 36). At term labor, a regional loss of PGDH is observed in the chorion of the lower uterine segment, potentially facilitating local generation of bioactive PGs to promote cervical ripening (37).
In cultured human trophoblast cells, PGDH activity and mRNA levels are increased by treatment with the synthetic progestins R5020 and medroxyprogesterone acetate (MPA), inhibited by the progesterone receptor (PR) antagonists RU486 and onapristone, and by cortisol (38). Because these cells produce progesterone, they were also treated with trilostane, an inhibitor of 3ß-hydroxysteroid dehydrogenase (3ßHSD) which converts pregnenolone to progesterone. The observation that trilostane also reduces PGDH expression supports autocrine/paracrine regulation.
Compared with the proposed steroid-mediated control of PGDH in the reproductive system, far less is known about regulation of the enzyme in other tissues. As opposed to the glucocorticoid-mediated inhibition of PGDH in trophoblast cells, the synthetic glucocorticoid dexamethasone (DEX) has been reported to stimulate PGDH synthesis and activity in human erythroleukemia cells, as did progesterone in this system (20). The only other known stimulant of PGDH synthesis in erythroleukemia cells and human promyelocytic leukemia cells (HL-60) is the phorbol ester, O-tetradecanoylphorbol 13-acetate (TPA) (20, 21).
It is not known whether control of human PGDH expression occurs at the transcriptional level. Recently, the genomic structure of the mouse PGDH gene has been reported (39). Characterization of 1.6 kb of the promoter region in the monocytic cell line U937 revealed a moderate transcriptional induction by phorbol ester.
The aim of the present study was to elucidate the molecular mechanisms governing human PGDH expression by isolating the promoter region of the gene, and to characterize PGDH transcriptional regulation in a number of different cell types of uterine, placental, and hemopoietic origin. Furthermore, we were particularly interested in testing the recently proposed hypothesis that human PGDH might be the first gene identified that is directly regulated by the mineralocorticoid receptor (MR) (40).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
FIX II
(Stratagene, La Jolla, CA) was screened with a
digoxygenin-labeled PGDH cDNA probe prepared by PCR on template
pGEM-humPGDH using primer PGDH-1 (position 6–30) and PGDH-3 (antisense
to position 765–738) and digoxygenin-11-deoxy-UTP (DIG-11-dUTP;
Roche Molecular Biochemicals, Mannheim, Germany). Four
positive clones were identified, from which one, named clone 2D, was
subjected to further analysis.
The genomic insert was released from phage DNA by digestion with
NotI, XbaI, or SstI. Resulting
fragments were subcloned into the NotI restriction site of
the pCR-Script SK+ plasmid
(Stratagene) or the XbaI or SstI
sites of pGEM-3Zf+ plasmid (Promega Corp.), respectively, by means of a shot-gun cloning method,
using 200 ng of the total restriction digests for ligation with the
respective linearized recipient plasmid. Resultant clones were digested
with the above enzymes and fragments resolved in a 0.7% agarose gel
before Southern transfer onto positively charged nylon membranes
(Roche Molecular Biochemicals). The following subclones
were used for further study (numbers give the approximate insert sizes
in kilobases): pCR/Not-6, pCR/Not-4.3, pGEM/Xba-2, pGEM/Xba-0.5,
pGEM/Xba-6, pGEM/Sst-6.7, and pGEM/Sst-4. The positions of the genomic
subfragments within clone 2D were determined by restriction digest
mapping and are illustrated in Fig. 1
. To
identify clones containing the immediate 5'-flanking region of the PGDH
gene, oligonucleotide PGDH-8 was labeled with DIG-11-dUTP and terminal
deoxynucleotidyl transferase (Roche Molecular Biochemicals) and used as a probe. Blots were hybridized in a
simplified high SDS buffer (41), and detection was performed with
anti-DIG-antibody conjugated to alkaline phosphatase and the
chemiluminescent substrate CDP-Star (Roche Molecular Biochemicals).
|
-35S]ATP and the Sequenase 2.0 DNA
sequencing kit (Amersham Pharmacia Biotech Buchler,
Braunschweig, Germany). Sequencing was performed along both strands of
DNA. Putative transcription factor-binding sites were identified with
the TFSEARCH program version 1.3 by Yutaka Akiyama
(http://www.rwcp.or.ja/papia/) or with MatInspector v2.2
(http://transfac.gbf.de/) (42) using the TRANSFAC databases (43).
Determination of the transcription start site
The transcription start site of the PGDH gene was determined by
means of ribonuclease protection assay (RPA) using total RNA from human
term placenta. Genomic template for generation of the riboprobe was
constructed via restriction digest of pGEM/Xba-6 with BamHI
(position -388 relative to ATG start codon) and NarI
(position +85). This fragment was then subcloned into the
BamHI and Nar I sites of pCR-Script
SK+ (Stratagene). The recombinant
plasmid was subsequently linearized with BamHI and used as
template for riboprobe synthesis. An antisense complementary RNA (cRNA)
probe was transcribed in the presence of
[-32P]CTP (Amersham Pharmacia Biotech Buchler) using T7 RNA polymerase (Roche Molecular Biochemicals). The labeled riboprobe was purified by
electrophoresis through a 6% denaturing polyacrylamide gel. The
full-length band was excised from the gel and eluted in probe elution
buffer from the HybSpeed RPA kit (Ambion, Inc., Austin,
TX) at 37 C overnight. For RPA analysis, cRNA probe was hybridized
against 30 µg human placental total RNA or control yeast RNA followed
by digestion with ribonucleases A/T1, using the HybSpeed RPA kit,
following the specified protocol. Protected fragments were resolved in
a 6% denaturing polyacrylamide gel against a
32P-labeled RNA marker, generated with T7 RNA
polymerase in the presence of [-32P]CTP on
RNA Century marker templates (Ambion, Inc.) and a dideoxy
chain termination DNA sequencing reaction as size markers.
Autoradiography was performed at -80 C against x-ray film (Fuji Photo Film Co., Ltd., Tokyo, Japan).
Cell culture
Cell lines used in the current study included the human
choriocarcinoma cell lines JEG-3 and BeWo (ATCC HTB 36 and ATCC CCL 98,
American Type Culture Collection, Manassas, VA), the human
leiomyosarcoma cell lines SKUT-1 (ATCC HTB 114) and SKN (RCB0513, Riken
Gene Bank, Tsukuba, Japan), Jurkat E6.1 human leukemic T lymphoblasts
(European Collection of Animal Cell Cultures, Salisbury, UK), and the
human promyelocytic leukemia cell line HL-60 (ATCC CCL 240). In
addition, primary cultures of human myometrial smooth muscle cells
(SMC) and human endometrial stromal cells were prepared as described
previously (44, 45). BeWo and SKN cells were cultured in Ham’s F-12
with 10% FCS, JEG-3 cells in DMEM with 10% FCS, and SKUT-1 and
primary myometrial cells in a 1:1 mixture of DMEM and Ham’s F-12 with
10% FCS. Jurkat and HL-60 cells were maintained in Iscove’s MEM with
10% FCS. All media were supplemented with 100 U/ml penicillin and 100
µg/ml streptomycin.
RT-PCR and Southern blot analysis
RT-PCR was used to examine expression of PGDH mRNA in different
cell types. Total RNA was extracted from cultured cells at
approximately 80% confluence using RNAzol B (Biotecx Laboratories,
Houston, TX), according to the manufacturer’s protocol. Complementary
DNA was synthesized from 5 µg total RNA using Superscript
ribonuclease H- reverse transcriptase
(Life Technologies, Inc., Eggenstein, Germany) and
oligo(deoxythymidine) primers. PCR was performed using 200
nM sense and antisense primers, 0.2 U Biotherm DNA
polymerase (Genecraft, Muenster, Germany), and 200 nM dNTPs
in a final volume of 50 µl. A touchdown program was used to amplify
PGDH cDNA (two cycles each at annealing temperatures of 68, 64, 60, and
56 C, followed by 40 cycles at 52 C for 45 sec, extension in each cycle
for 1 min at 72 C, and denaturation for 45 sec at 95 C), in a Touchdown
thermocycler (Hybaid, Teddington, UK). RT-PCR with glyceraldehyde
3-phosphate dehydrogenase (GAPDH)-specific primers was used to assess
the integrity of cDNA template in each sample (30 cycles; annealing at
50 C for 30 sec, extension at 72 C for 1 min). Oligonucleotide primers
were PGDH-1 (sense; position 6–30 relative to ATG start codon in the
human PGDH mRNA), PGDH-2 (antisense; complementary to position
396–419), GAPDH-sense (corresponding to position 278–298 in GenBank
accession no. M33197; HUMGAPDH), and GAPDH-antisense (complementary to
position 508–528) (46).
The PCR products were fractionated in a 1% agarose gel before Southern blot analysis, as described above, with DIG-labeled oligonucleotide probes designed internally to the amplifying primers against the published cDNA sequence: PGDH-5 (antisense to position 265–289 relative to ATG start codon) and GAPDH-internal (corresponding to position 452–469).
RT-PCR and Southern hybridization for members of the Ets family of transcription factors were performed as follows. For PEA3, primers were PEA-5 (position 584–604 in human PEA3 cDNA) (47) and PEA-3 (antisense to position 992–969), probe PEA-int corresponds to position 817–839. For amplification of Ets-1 and Ets-2 cDNAs, a common downstream primer was used, ets-AS, against the highly homologous DNA-binding domain (antisense to human Ets-1A cDNA, position 1344–1363, and Ets-2A cDNA, position 1600–1619, respectively; ACRAAGCGGTACACACGTAGCG) (48). Sense primers were Ets1-S (position 728–750 in human Ets-1A cDNA) or Ets2-S (position 510–530 in human Ets-2A cDNA), and internal probes for Southern hybridization were Ets1-AS (antisense to position 1024–1047) or Ets2-AS (antisense to position 920–941), respectively. The same touchdown program as that outlined above was employed.
Reporter gene constructs and expression vectors
A 2.4-kb fragment of human PGDH 5'-flanking DNA including a
small portion of exon 1 was prepared as follows. On template
pCR/Not-4.3 (see Fig. 1
) PCR was performed with the following primers:
M13 universal primer (anchored in the polylinker of pCR-Script 5' to
the genomic insert) and PGDH-18
(GTGCATGGTCGACCCACTGCTGG),
antisense to position -17/+6 relative to the ATG start codon
(boldface), introducing a SalI site
(underlined) through three point mutations
(italics). The PCR product was restricted with
NotI at the natural site at position -54 and with
SalI at the artificial site at -7 in the genomic insert.
Clone pGEM/Xba-6 was cut at the same internal NotI site and
with SalI in the polylinker of
pGEM-3Zf+ 3' to the genomic insert. The released
genomic fragment was removed and replaced by the
NotI-SalI PCR fragment (-54/-7) described
above. The resultant plasmid contained PGDH 5'-flanking DNA from the
XbaI site at -2368 to the artificial SalI site
at -7 in pGEM-3Zf+ and was termed
pGEM/PGDH-2368.
For construction of the PGDH-2368/luc3 luciferase reporter gene construct, the complete 2.4-kb insert of pGEM/PGDH-2368 vector was isolated via restriction digest with KpnI and SalI and inserted into the KpnI and XhoI sites of pGL3-Basic (Promega Corp.). 5'-Deletion constructs were generated from PGDH-2368/luc3 as follows. The PGDH-1024/luc3 vector was constructed by releasing the fragment between the BglII restriction sites at position -1024 and in the 3'-polylinker region of the PGDH-2368/luc3 construct. This was then subcloned into the BglII site of the pGL3-Basic vector. PGDH-388/luc3 was generated by removing the fragment between restriction sites BamHI at position -388 and BglII in the polylinker region of PGDH-2368/luc3. This fragment was blunt ended by incubation with Pfu polymerase (Stratagene, Heidelberg, Germany) and dNTPs for 30 min at 72 C and subcloned into the SmaI site of the pGL3-Basic vector. PGDH-203/luc3 was generated by HindIII restriction digest of the PGDH-2368/luc3 vector to remove the fragment from -203 bp to the HindIII site in the 3'-polylinker. This fragment was inserted into the HindIII site of pGL3-Basic.
The construct PGDH-2368(-636)/luc3 is equivalent to the full-length 2.4-kb construct with 636 bp from -1024 to -388 deleted. This construct was generated by isolating the fragment between the BamHI site at -388 and the BglII site in the 3'-polylinker region of PGDH-2368/luc3. This fragment was then subcloned into the PGDH-2368/luc3 vector, which had been restricted with BglII to remove the fragment from -1024 to the BglII site in the 3'-polylinker region. The identity and orientation of all inserts were confirmed by sequence analysis.
An expression vector for human Ets-1 (pSG5-hEts1) and the human Ets-2
cDNA in pBluescript SK were provided by Dr. B. Wasylyk (INSERM,
Illkirch, France) (48). The full-length Ets-2 cDNA (position 152-2271
in GenBank accession no. HUMETS2A) was excised with EcoRI
and inserted into the EcoRI site of pSG5
(Stratagene) to give the expression vector pSG5-hEts2; a
clone in which the cDNA was inserted in the reverse orientation,
pSG5-hEts2rev., was used as a negative control. The expression vector
pCMVE1A-F, carrying the human PEA3 cDNA, was provided by Dr. K. Yoshida
(Sapporo Medical University, Sapporo, Japan) and is referred to as
pCMV-hPEA3 in this report (47). The hPR-B and hPR-A expression vectors
N-hPR0 and N-hPR2 in pRc/CMV (Invitrogen, Groningen,
Netherlands) have been described previously (49). N-hPR0rev. carries
the hPR-B insert in the reverse orientation and was used for controls.
The expression vector for human glucocorticoid receptor- (hGR)
was provided by Dr. M. G. Rosenfeld (University of California-San
Diego, La Jolla, CA). The human MR expression vector, pRS-hMR (50) was
provided by Dr. R. M. Evans (Howard Hughes Medical Institute, San
Diego, CA).
Expression vectors for the dominant negative inhibitors of activating protein-1 (AP-1) and cAMP-responsive element (CRE)-binding protein (CREB), CMV500/A-Fos and CMV500/A-CREB, consist of an acidic amphipathic extension fused onto the N-terminus of the Fos or CREB leucine zipper domains, respectively, inserted into pCMV500 (51, 52).
The steroid-inducible control reporter plasmid mouse mammary tumor virus (pMMTV-Luc) carries the mouse mammary tumor virus long terminal repeat in front of the luciferase gene and has been described previously (49). For construction of a progesterone-responsive reporter plasmid, oligonucleotides GRE-a (CTGTACAGGATGTTCTAGCTACG) and GRE-b (GATCCGTAGCTAGAACATCCTGTACAGAGCT) were annealed. The resultant double stranded oligonucleotide contains the glucocorticoid-responsive element (GRE) from the tyrosine aminotransferase gene (bold face) (53) and BamHI and SstI overhangs (underlined). It was dimerized via the BamHI ends and ligated into the SstI site of pGL2-Basic (Promega Corp.). The insert was excised with Acc65I and XhoI and ligated into the corresponding sites in the polylinker of pGL3-Basic 5' to the minimal promoter element -32/+65 of the human decidual PRL promoter (54) to yield PRE/-32wt/luc3. The cAMP-responsive reporter construct CRE/-36rPRL/luc3 has been described previously (45).
The phorbol ester-inducible reporter plasmid 3xAP1/luc3 contains three
AP-1 consensus elements in front of the minimal Ig
promoter in
pGL3-Basic and was provided by Dr. R. Schmid (University of Ulm, Ulm,
Germany).
Transient transfections
Equimolar amounts of each PGDH promoter/luciferase reporter gene
vector were used for transient transfection analysis, keeping the total
amount of DNA constant by the inclusion of the promoterless plasmid
p0GH (Nichols Institute Diagnostics, San Juan Capistrano,
CA).
Jurkat T lymphoma cells were transfected via electroporation at 250 V and 960 microfarads in a Gene Pulser (Bio-Rad Laboratories, Inc., Munich, Germany). Approximately 8 x 106 cells were transfected with 10 µg DNA in 800 µl growth medium. In cotransfection studies a maximum of 6.5 µg reporter and 2–4 µg expression vector were used. After electroporation, cells were added to 12.5 ml fresh growth medium and plated at 5 x 105 cells/well in 12-well cell culture plates.
All other cell lines were transfected by means of calcium phosphate coprecipitation using the ProFection Mammalian Transfection System (Promega Corp.). On the day before transfection, cells were plated in 12-well culture plates at a density of 1.5 x 105 cells/well. Medium was changed 3 h before transfection. Cells were transfected with a maximum of 2 µg reporter vector/well. Primary cultures of myometrial SMC or endometrial stromal cells were seeded in 12- or 24-well plates at a density of 0.8 or 0.4 x 105 cells/well, respectively, and transfected by calcium phosphate coprecipitation as described above. For cotransfection experiments in 24-well plates, a maximum of 0.9 µg reporter and 0.1–0.4 µg expression vector were used. For steroidal stimulations, cells were plated in medium supplemented with 10% FCS that had been depleted of steroids by treatment with dextran-coated charcoal. After overnight exposure to the DNA, the solution was aspirated, and fresh medium was added to the cells. Treatments included 10-7 or 10-6 M progesterone, DEX, or aldosterone, 5 x 10-8 M TPA (all from Sigma, Deisenhofen, Germany), or 0.5 mM 8-bromo-cAMP (8-Br-cAMP; Biolog, Bremen, Germany). Cells were harvested 24 h later in reporter lysis buffer (Promega Corp.) for luciferase assay. Twenty microliters of extract were mixed with 50 µl luciferase reagent (Promega Corp.), and relative light units (RLU) were measured in a Berthold luminometer (Isernhagen, Germany). All experiments were performed in triplicate in at least three independent experiments and expressed as the mean ± SD unless indicated otherwise.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Partial sequencing of subclones pGEM/Xba-2 and
pCR/Not-6 and of the 5'-end of clone 2D revealed complete homology with
positions 106246–115391 of the locus HUMFLNG6PD, a region
on the human chromosome X from the filamin to the glucose-6-phosphate
dehydrogenase gene (GenBank accession no. L44140). The portion of this
locus that is contained in phage clone 2D comprises the entire gene (11
exons) for Rab GDP-dissociation inhibitor- (Rab GDI) (55). Beyond
position 106246, representing a Sau3AI site, locus
HUMFLNG6PD had no homology with clone 2D. The PGDH gene has
been assigned to chromosome 4 (29). To clarify whether this fusion of
regions from chromosomes X and 4 was an artifact generated during
library construction or whether chromosome 4 carries a duplicated
portion of the X chromosome, we performed Southern blot analysis.
Genomic DNA from human pituitary, the human breast tumor cell line
T47D, and the human bladder carcinoma line T24 were digested with
NotI, SstI, XbaI, EcoRI,
BamHI, or PstI and hybridized with an
XhoI-NcoI fragment isolated from subclone
pGEM/Xba-2 of clone 2D, which represents HUMFLNG6PD position
106263–107402 (Fig. 1). The resultant bands corresponded exclusively
to those predicted from HUMFLNG6PD and not to those deduced
from the restriction map of clone 2D as outlined in Fig. 1 (data not
shown). We therefore had to conclude that phage clone 2D is an
artifactual hybrid and performed all subsequent experiments with the
genomic sequence downstream of the Sau3AI at the point of
divergence from HUMFLNG6PD sequence.
Sequencing of the PGDH-specific portion of 2D yielded novel sequence of
3505 bp 5' to the start ATG codon of the published human PGDH cDNA
(23). About 1070 bp downstream of the start codon were also sequenced
and aligned to the published mouse PGDH gene (39) (Fig. 2). Conserved splice donor and acceptor
sequences suggest high similarity of the genomic structures. Exons 1
and 2 are separated by an intron 1 of about 0.2 kb in both species. The
second exon is followed by an intron of 3.5 kb in the mouse; the human
clone 2D contains intronic sequence of about 3 kb and terminates
upstream of exon 3.
|
). The protected
fragments were 120 and 122 bases in size, locating the major
transcription start sites at -34 and -36 relative to the ATG start
codon. Sequence of 5'-flanking region to position -2369
(XbaI site) of the hPGDH gene is shown in Fig. 4A. Alignment of the human
with the published mouse PGDH promoter region extending to -1.6 kb
revealed a high degree of homology within 240 bp of the proximal
promoter region, including an AP-1 site, a Sp1/AP-2 site, the
TATA box, and the location of the transcriptional start site, which had
been mapped to -35 in the mouse (39) (Fig. 4B). The remaining
5'-flanking sequence of the human PGDH gene diverges from that of the
mouse. It contains several potential binding sites for members of the
C/EBP and c-Ets families of transcription factors and for AP-1. The
entire fragment was inserted into the luciferase reporter plasmid
pGL3-Basic for transient transfection experiments. In addition to this
construct (PGDH-2368/luc3), a series of 5'-deletion constructs was
generated as outlined in Fig. 5:
PGDH-1024/luc3, PGDH-388/luc3, and PGDH-203/luc3. Finally, the portion
-1024/-388 was removed from PGDH-2368/luc3 to yield
PGDH-2368(-636)/luc3 (Fig. 5).
|
|
|
). As previously reported, term
placenta and the promyelocytic leukemia cell line HL-60 were positive
for PGDH (21, 23). In addition, strong signals were obtained with cDNA
from the leukemic T cell line Jurkat and from myometrial tissue. The
choriocarcinoma cell lines BeWo and JEG-3, however, were negative, as
were primary cultures of myometrial SMC and the leiomyosarcoma cell
lines SKUT-1 and SKN. The activity of the endogenous PGDH promoter seen
in freshly isolated placenta and myometrium is apparently not
maintained in tumor cell lines derived from these tissues, whereas
leukemic cell lines display significant mRNA expression. We next tested
the ability of these cell lines to activate the transfected PGDH
promoter and observed cell type-specific profiles for induction of the
various deletion constructs relative to the promoterless pGL3-Basic
vector (Fig. 6B). While Jurkat cells showed maximal induction with
PGDH-388/luc3 (30-fold over pGL3-Basic), the choriocarcinoma cell lines
JEG-3 and BeWo activated the longest construct, PGDH-2368/luc3, most
effectively (50- and 80-fold inductions, respectively). Primary
myometrial SMC only displayed a small induction of this construct, but
it was preferentially used by the leiomyosarcoma cell lines SKUT-1 and
SKN. PGDH-1024/luc3 displayed relatively low basal activities in all
cell types. The drop in activation between PGDH-388/luc3 and
PGDH-1024/luc3 prompted us to delete this 636-bp region from the
full-length promoter construct to generate PGDH-2368(-636)/luc3. When
expression profiles were compared in transfected Jurkat and myometrial
SMC, a striking difference became apparent (Fig. 6C). Although deletion
of the 636-bp fragment from PGDH-2368/luc3 had no effect in Jurkat
cells, it brought the low activity of PGDH-2368/luc3 up to the level of
the most active construct, PGDH-388/luc3, in myometrial SMC. This
indicates that a cell-specific repressor element may be located between
-388 and -1024.
|
and treated with phorbol ester (TPA) to activate
AP-1-dependent transcription. As a positive control, a reporter
construct carrying a trimerized AP-1 response element was included
(3xAP1/luc3; Fig. 7). All reporter gene
fusions were clearly induced by TPA, including the shortest construct,
PGDH-203/luc3. This indicates that the promoter-proximal AP-1 site
significantly contributes to the induction. On the other hand, deletion
of the 636-bp region from the full-length construct did not affect TPA
responsiveness, indicating that the two AP-1 elements between -388 and
-1024 are not involved (see Fig. 5). Cotransfection of a dominant
negative to AP-1 proteins, pCMV/A-Fos (52), completely abrogated
TPA-induced stimulation. A-Fos inhibits AP-1 function by
heterodimerizing with Jun family members. It consists of the leucine
zipper region of the human Fos protein to which an acidic amphipathic
extension has been appended at the N-terminus, replacing the basic
region of the bZIP (basic region/leucine zipper) protein. In contrast
to the pronounced induction of the PGDH promoter by TPA in Jurkat
cells, numerous individual primary cultures of myometrial SMC showed
either no or a less than 2-fold activation of the PGDH promoter upon
addition of the phorbol ester.
|
). Expression vectors for Ets-1, Ets-2,
and PEA-3 affected a strong induction of the PGDH-388 promoter
construct in transfected myometrial SMC (7.7-, 6.3-, and 18-fold,
respectively; Fig. 8B). Addition of TPA did not modulate this response
(data not shown). However, cotransfection of the dominant negative
A-Fos significantly reduced the PEA-3-stimulated induction that was
obtained with all PGDH promoter constructs containing 5'-flanking
sequence of at least 203 bp (Fig. 8C). Among the 5'-deletion
constructs, PGDH-388/luc3 was the most highly responsive to PEA-3. As a
negative control to the A-Fos expression vector, we had included the
dominant negative A-CREB (51). Analogous to the A-Fos, it is composed
of an acidic extension fused to the CREB leucine zipper. This fusion
did not interfere with PEA-3-mediated induction of PGDH-203/luc3,
PGDH-388/luc3, or PGDH-1024/luc3. Interestingly, however, it repressed
induction of PGDH-2368/luc3 as strongly as did A-Fos. These results
provide evidence for two conclusions. Even though TPA, alone or in
combination with overexpressed PEA-3, does not cause activation of the
PGDH promoter in myometrial SMC, endogenous Jun/Fos proteins appear to
be involved in PEA-3-stimulated activation, even on the shortest
construct, PGDH-203/luc3. In addition, CREB-dependent signaling seems
to be involved in the PEA-3 induction conferred by the region
-2368/-1024.
|
Regulation of the PGDH promoter by steroid hormones
Studies on the level of PGDH protein expression and enzymatic
activity in numerous systems have established a role for
glucocorticoids and progesterone in its control (20, 31, 32, 34, 38). A
recent report on cortisol-dependent down-regulation of PGDH activity
and mRNA levels in placental and chorion trophoblast cells (56) gave
rise to the hypothesis that this effect might be mediated not by the
GR, but by the MR, and occur at the transcriptional level (40). We
therefore initiated studies on the roles of PR, GR, and MR in the
control of the PGDH promoter. Myometrial SMC were cotransfected with
PGDH-2368/luc3 and expression vectors for hPR-B, hGR, and hMR and
stimulated with progesterone, DEX, or aldosterone, respectively. As a
positive control reporter we employed pMMTV/luc containing the
steroid-responsive mouse mammary tumor virus long terminal repeat (Fig. 9). The PGDH promoter was unaffected by
MR or GR in the presence or absence of ligand, but was clearly
induced by PR-B in a ligand-dependent manner. The pMMTV promoter
was strongly inducible by ligand-activated MR and PR-B and massively
activated by ligand-activated GR, confirming functional expression
of the transfected receptors.
|
). Although addition of 8-Br-cAMP in
the absence of steroid was without effect, a pronounced synergism was
obtained in the presence of ligand. Liganded PR-A, on the other hand,
did not activate the reporter, and addition of 8-Br-cAMP to
progesterone-treated cells only effected a miniscule induction. A
strikingly different pattern was observed on the PGDH-2368 promoter
construct. It was induced by PR-B plus progesterone, whereas MR plus
aldosterone was completely ineffective as shown before in Fig. 9, but
interestingly PR-A plus progesterone was as effective as liganded PR-B
in PGDH promoter activation. 8-Br-cAMP alone caused a small, but
significant, 2-fold induction and enhanced progesterone-mediated
activation in the presence of PR-B, but not of PR-A. The lack of
transcriptional activity of liganded MR on the PGDH promoter could not
be overcome by the addition of 8-Br-cAMP (Fig. 10).
|
, left panel), all
PGDH promoter constructs were activated 2- to 3-fold by 8-Br-cAMP, an
effect also seen on the control reporter constructs carrying a PRE or a
CRE in front of a minimal promoter. However, only the PRE reporter
construct was induced by progesterone and exhibited pronounced
synergistic activation upon addition of 8-Br-cAMP plus progesterone.
The level of endogenous PR in endometrial stromal cells is, therefore,
sufficient to trans-activate through a PRE consensus binding
site, but does not act on the PGDH 5'-flanking DNA. In contrast, when
PR-B was cotransfected (Fig. 11, right panel), activation by
progesterone alone and pronounced synergy with 8-Br-cAMP were obtained
on all PGDH promoter constructs containing at least 388 bp of
5'-flanking DNA. The response of the PRE reporter construct was merely
elevated by overexpression of PR-B.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) (55) within the HUMFLNG6PD locus. The
PGDH-specific portion contained 3.5 kb of novel 5'-flanking
DNA, exons 1 and 2, intron 1, and about 3 kb of intron 2. The total
number of exons in human PGDH is not known; cloning of a
1.1-kb fragment comprising three exons (29), identification of splice
variants that are C-terminally truncated (27, 28), and alignment with
the genomic structure of mouse PGDH (39) suggest that the
1.1-kb genomic clone contains the exons corresponding to mouse exons 5
and 6 and part of exon 7, which carries the termination codon.
Alignment of clone 2D with the mouse gene shows a high degree of
conservation of splice junctions at exons 1 and 2, a similarly sized
intron 1 of 0.2 kb, and an intron 2 of at least 3 kb, with the mouse
intron 2 being 3.5 kb in length. Striking homology is seen in the
proximal promoter region of the mouse and human genes. The
transcriptional start sites mapped to -34/-36 by us correspond well
to the start site in the mouse promoter at -35, and conservation of
potential binding sites for Myo-D/AP-1 and Sp1/AP2 suggest
functional relevance. Even though the 5'-flanking regions diverge
beyond -230, there is an interesting conservation of a potential
upstream TATA box at -1.2 kb in the mouse and -1.26 kb in the human
gene. The significance of this TATA sequence has not been addressed as
yet. Analysis of 2369 bp of the 5'-flanking region revealed five AP-1
sites, two of which appeared particularly indicative of a
cis-acting role due to their context. At -224/-163, the
most proximal AP-1 site is neighbored by an Ets site, and at a
clustered region with coordinates -2055/-1964, an AP-1 site is
closely adjoined by two adjacent Ets sites, all three sites, in turn,
being flanked by two CREB-binding sites. By transient transfection analysis, we could demonstrate an involvement of both AP-1 and Ets family members in PGDH transcriptional regulation. TPA treatment of Jurkat cells resulted in a pronounced activation of all PGDH promoter constructs, which was completely reversible by cotransfection of the dominant negative A-Fos vector. Taken together with the fact that even the shortest promoter construct, PGDH-203/luc3, which only carries an AP-1 site, was TPA responsive, these results suggest that the TPA effect is predominantly transduced via activation of AP-1 in Jurkat cells. In myometrial SMC, Ets-1, Ets-2, and particularly PEA3 proved to be potent activators of the PGDH promoter. The most highly responsive deletion construct was PGDH-388/luc3, which contains both an AP-1- and an Ets-binding sequence. In fact, coexpression of dominant negative A-Fos caused a repression of PEA3-induced stimulation by at least 50%, pointing to a contribution of AP-1 to PEA3 trans-activation. Surprisingly, in most individual myometrial preparations, we could not elicit PGDH promoter activation by TPA, alone or in combination with Ets factors. This may, however, be due to the transfection technique employed. We observed in primary cultures of endometrial stromal cells that electroporation or liposome-mediated transfection allowed activation of responsive promoters by TPA, whereas the calcium phosphate coprecipitation method resulted in an elevated basal activity of the target promoter that was, however, refractory to further stimulation by TPA (unpublished observation). The latter phenomenon might represent a Ca2+-mediated triggering of signal transduction pathways that would otherwise be activated by TPA. As we use calcium phosphate for transfection of myometrial SMC (alternative methods being too inefficient), but electroporation for Jurkat cells, the lack of effect of TPA in myometrial cells as opposed to the effectiveness of TPA in Jurkat cells might be explained by such a mechanism. The suppressive effect of A-Fos on PEA3-mediated promoter activation suggests that endogenous AP-1 proteins are interacting with the transfected PEA3 on the PGDH promoter. A cooperation between Ets and AP-1 proteins has been well established (60). Ets/AP-1 composite sites are found in many promoters, such as those of the matrix metalloproteinases (reviewed in Ref. 61), granulocyte-macrophage colony-stimulating factor (62), or urokinase-type plasminogen activator genes (63), and are designated prototypical Ras-responsive elements (64). One mechanism for cooperative action of Ets and AP-1 proteins is a direct physical association between the Ets domain of Ets proteins and the basic domain of Jun, which then results in formation of a trimeric complex with Fos via leucine zipper interaction (65). In addition, both the Ets-1 and Ets-2 genes are transcriptionally up-regulated by Ets/AP-1 complexes (66, 67).
PEA3-mediated activation of PGDH-2368/luc3, but not of shorter promoter constructs, was repressed not only by A-Fos but also by the dominant negative A-CREB. A composite response element of two adjacent Ets-binding sites and an AP-1 site flanked by two CRE-like elements is located at -2055/-1964. The upstream sequence (TGACGTCC) corresponds to the CRE of the human T cell leukemia virus type 2 long terminal repeat, but was found to bind at least CRE modulator only very poorly (68). The downstream sequence (TGACCTCA), however, has been shown to bind CREB or activating transcription factor-2 (ATF-2) homodimers or ATF-2/c-Jun heterodimers with very high affinity comparable to the consensus CRE (TGACGTCA) (69). The dominant negative A-CREB most likely interacts not only with CREB, but also with the highly related leucine zippers of CREM or ATF-1 (51). PEA3 stimulation of the PGDH-2368 promoter construct, therefore, involves not only AP-1, as demonstrated by the repressive effect of A-Fos, but apparently also members of the CREB/CREM/ATF family of bZIP proteins, as evidenced by the interference of A-CREB. We have previously shown that the intracellular cAMP content of myometrial SMC increases with time in culture (44), suggesting an activated PKA pathway. It remains to be determined whether activated CREB/CREM/ATF proteins bind to the downstream CRE and if such binding involves heterodimerization with c-Jun. Among the growing number of transcription factors that use the CREB-binding protein CBP/p300 to mediate trans-activation of RNA polymerase II are not only CREB, c-Fos, and c-Jun (70), but also Ets-1 (71). The clustered Ets/AP-1/CREB element in the distal promoter region of the PGDH gene would therefore allow multifactorial input to be transduced by a single integrator molecule, CBP.
When comparing basal activities of PGDH promoter deletion constructs in various cell types, we noted a marked difference between Jurkat and myometrial SMC, on the one hand, and choriocarcinoma cells BeWo and JEG-3, on the other hand. Although the former two cell types displayed the highest activation with PGDH-388/luc3, the choriocarcinoma cell lines preferentially activated the longest construct, PGDH-2368/luc3. Assuming that CREB or related factors are involved in activation via the distal element, -2055/-1964, choriocarcinoma cells might stimulate PGDH transcription in a paracrine fashion. These cells not only produce high amounts of hCG, but are likely to express receptors for LH/hCG as well. Invasiveness of JEG-3 cells has been shown to be stimulated by hCG, providing indirect evidence for the presence of gonadotropin receptors (72). Using in situ hybridization and immunohistochemistry, choriocarcinomas have been demonstrated to contain LH/hCG receptors, expression levels exceeding those of normal placenta (73). Ligand binding to LH/hCG receptors leads to activation of the PKA pathway, providing a means of up-regulating PGDH expression during pregnancy. Only one report has dealt with the effect of cAMP on PGDH expression and seems to contradict our hypothesis. During in vitro differentiation of human term placental trophoblast cells into syncytiotrophoblasts, expression of PGDH was found to increase. The addition of 8-Br-cAMP, however, diminished the PGDH protein level and enzymatic activity (74). This is somewhat surprising in view of the fact that during differentiation of cultured trophoblast cells to form syncytial structures, the cellular levels of cAMP are known to rise, and differentiated endocrine functions, such as production of hCG and progesterone, are stimulated by cAMP analogs (75).
In myometrial SMC we observed a small, but significant, increase in PGDH-2368/luc3 activity after treatment with 8-Br-cAMP. More importantly, the cAMP analog enhanced progesterone-stimulated promoter activity in the presence of hPR-B. This is the first demonstration of a progesterone effect on PGDH expression at the transcriptional level. Not only hPR-B, but also the truncated receptor hPR-A, conferred progesterone responsiveness to the PGDH promoter, which was in marked contrast to the lack of trans-activation function of the PR-A on a reporter construct that carried an isolated PRE. In the context of the PGDH-2368 promoter construct, however, PR-A differed from PR-B in that its function could not be enhanced by 8-Br-cAMP. The secretory phase of the menstrual cycle and pregnancy are physiological conditions characterized by elevated levels of gonadotropins (LH and hCG, respectively) and progesterone. One might speculate that the enhanced expression of PGDH seen in endometrium, myometrium, and trophoblast layers under these conditions can at least in part be attributed to a stimulation of the PKA pathway by gonadotropins in combination with activated PR. A modulation of the response to gonadotropins might be brought about by a change in PR isoform ratio, a shift in favor of PR-A no longer being permissive to synergistic action with gonadotropins (or other agents that activate the PKA pathway). Distinct roles of PR isoforms in relation to PGDH expression are suggested by a longitudinal study in women using a levonorgestrel-releasing intrauterine system (76). After insertion of the system, there was a significant down-regulation of total PR protein, the PR-B isoform, and PGDH in the endometrial glands in response to the progestogen. After long term levonorgestrel delivery (6–12 months), PGDH activity increased in the endometrial glands along with an apparent rise in PR-A expression.
A link between progesterone- and PKA-mediated control mechanisms may also have the following implication in PGDH expression. One of the major PGs in many systems, PGE2, causes an increase in intracellular cAMP when binding to its receptor subtypes EP2 or EP4 (8). After up-regulation of PGDH by cAMP in synergy with progesterone, PG inactivation would be enhanced, thus leading to a reduction of cAMP formation and PGDH expression and establishing a negative feedback loop.
The augmentation of agonist-dependent PR activity by 8-Br-cAMP is well documented, but the mechanism is not clear. PR is a phosphoprotein, but the net phosphorylation status is not altered in response to 8-Br-cAMP, nor are hormone- or DNA-binding activities (77). The role of 8-Br-cAMP might rather lie in causing a disruption of interaction between PR and nuclear corepressors NCoR or SMRT (silencing mediator for retinoid and thyroid hormone receptor) (57). Yet no PRE consensus sequence is found in PGDH-2368, so the question remains to be answered whether PR exerts its effect by directly binding to the PGDH 5'-flanking sequence, by protein-protein interaction, or by inducing an intermediate factor, and how 8-Br-cAMP selectively enhances activity of PR-B but not PR-A. Deletion analysis in primary cultures of human endometrial stromal cells with overexpressed PR-B revealed that a region responsive to progesterone and to synergistic enhancement by 8-Br-cAMP resides within the first 388 bp of 5'-flanking DNA to the PGDH transcriptional start site. This region, however, did not respond to progesterone in the absence of cotransfected PR-B, whereas the endogenous PR present in endometrial stromal cells efficiently trans-activated a PRE control construct, particularly upon addition of 8-Br-cAMP. This discrepancy in response of a PRE-driven promoter compared with the PGDH promoter argues in favor of differential underlying mechanisms in PR action.
Effects of glucocorticoids on PGDH have been reported, but appear to be cell type-specific. Although DEX stimulated PGDH synthesis and enzymatic activity in human erythroleukemia cells (20), cortisol was found to decrease PGDH activity and mRNA in cultured human trophoblast cells from term placenta (38). In an effort to clarify whether glucocorticoids exerted their effects at the transcriptional level, we performed cotransfection analyses of PGDH promoter constructs with GR expression vectors. Although we observed a small ligand-dependent decrease in activity of the PGDH-2368 promoter construct in Jurkat and JEG-3 cells (data not shown), no effect of DEX on basal PGDH promoter activity was detectable in myometrial SMC. The control reporter construct PRE/-32wt/luc3 was strongly activated by GR upon addition of DEX in all cell types. At least under the conditions tested to date, GR does not appear to control the PGDH promoter to any significant extent.
In a recent study the inhibitory effects of cortisol and DEX on PGDH activity in cultured human trophoblast cells were suggested to be modified by local tissue-specific expression of 11ßHSD isoforms. Although the placental trophoblast mainly expresses 11ßHSD2, which converts cortisol to the inactive cortisone, 11ßHSD1 predominates in chorionic trophoblast to catalyze the reverse reaction (56). Thus, by modifying glucocorticoid levels, the 11ßHSDs could indirectly contribute to the regulation of PGDH activity. Dose-response curves showing inhibition of PGDH activity in placental cells in response to cortisol or DEX, performed in the absence or presence of the 11ßHSD inhibitor carbenoxolone, revealed a 100-fold higher EC50 for DEX compared with cortisol in the presence of carbenoxolone. This finding prompted J. Funder to postulate that the glucocorticoid effect must be mediated by the MR, which has a much lower affinity for DEX than for cortisol, rather than the GR, which binds both compounds with equal affinities (40). If this held true, it would be the serendipitous and intriguing discovery of the first directly MR-regulated gene. Having the PGDH promoter at hand, we set out to test this hypothesis. A prerequisite for a physiological role of MR in controlling PGDH expression would be local expression of the receptor. However, numerous attempts to detect aldosterone binding in tissues expressing 11ßHSD2, such as human term placental trophoblast (78), nonpregnant human myometrium and endometrium (79), or rat placenta (80), failed to detect MR expression. On the other hand, aldosterone may not be solely derived from the adrenal glands; SMC also elaborate aldosterone and may thus provide a local source of mineralocorticoids (81). We performed RT-PCR analyses for MR expression on various human tissues, including endometrium and myometrium from nonpregnant uterus, cultured myometrial SMC, term placenta, and Jurkat cells, and detected MR mRNA in all cases (data not shown). Although detection of MR by PCR does not allow conclusions as to the level of expression of functional protein, the potential presence of MR in these tissues of interest encouraged us to study the effect of MR on PGDH promoter control. Although MR upon addition of aldosterone strongly stimulated the PRE/-32wt/luc3 control reporter construct in Jurkat, JEG-3 (data not shown) and myometrial cells, no effect on the transfected PGDH promoter was observed. Transcriptional activity of MR, like that of PR, can be enhanced by 8-Br-cAMP. In contrast to PR, this augmentation of MR action can even occur in the absence of ligand, depending on promoter context (58). We therefore cotransfected myometrial SMC with MR expression vector and either PGDH-2368/luc3 or control construct PRE/-32wt/luc3 and treated the cells with 8-Br-cAMP or aldosterone alone or in combination. Although we did not obtain ligand-independent activation of the PRE construct with 8-Br-cAMP, the synergistic action of 8-Br-cAMP and aldosterone was very pronounced. However, no modulation of PGDH-2368/luc3 transcription was elicited by aldosterone alone, nor was the 2-fold induction of the PGDH promoter by 8-Br-cAMP affected by aldosterone. It remains to be elucidated whether the PGDH gene is transcriptionally regulated by MR through a control region outside 2368 bp of 5'-flanking DNA.
In conclusion, using 2.4 kb of 5'-flanking DNA to the human PGDH promoter, we have demonstrated the involvement of Ets, AP-1, cAMP, and PR in its transcriptional control. We provide evidence for cross-talk between multiple signaling pathways, which would allow for complex cell-specific regulatory mechanisms and await unraveling.
|
Acknowledgments |
|---|
Received August 19, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylase. Crit Rev Biochem Mol Biol 31:101–126[Medline]
