| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Division of Reproductive Biology Research,3 Department of Obstetrics and Gynecology, Northwestern University, Feinberg School of Medicine, Chicago, Illinois 60611-3095
Department of Pathology,4 Tohoku University School of Medicine, Sendai 980-8575, Japan
ABSTRACT
The opposing actions of estrogen and progesterone during the menstrual cycle regulate the cyclical and predictable endometrial proliferation and differentiation that is required for implantation. Progesterone indirectly stimulates the expression of 17beta hydroxysteroid dehydrogenase type 2(HSD17B2), which catalyzes the conversion of biologically potent estradiol to weakly estrogenic estrone in the endometrial epithelium. We previously demonstrated upregulation of theHSD17B2gene in human endometrial epithelial cells by factors secreted from endometrial stromal cells in response to progesterone. We investigated the underlying mechanism by which these stroma-derived, progesterone-induced paracrine factors stimulateHSD17B2expression. Here, we show that transcription factors SP1and SP3 interact with specific motifs inHSD17B2promoter to upregulate enzyme expression in human endometrial epithelial cell lines. Conditioned medium (CM) from progestin-treated stromal cells increased levels of SP1 and SP3 in endometrial epithelial cells and inducedHSD17B2mRNA expression. Mithramycin A, an inhibitor of SP1-DNA interaction, reduced epithelialHSD17B2promoter activity in a dose-dependent manner. Serial deletion and site-directed mutants of theHSD17B2promoter demonstrated that two overlapping SP1 motifs (nt –82/–65) are essential for induction of promoter activity by CM or overexpression of SP1/SP3. CM markedly enhanced, whereas anti-SP1/SP3 antibodies inhibited, binding of nuclear proteins to this region of theHSD17B2promoter. In vivo, we demonstrated a significant spatiotemporal association between epithelial SP1/SP3 and HSD17B2levels in human endometrial biopsies. Taken together, these data suggest thatHSD17B2expression in endometrial epithelial cells, and, therefore, estrogen inactivation, is regulated by SP1 and SP3, which are downstream targets of progesterone-dependent paracrine signals originating from endometrial stromal cells.
endometrium, estradiol, estradiol receptor, estrogen metabolism, HSD17B2, menstrual cycle, paracrine regulation, progesterone receptor, SP1, SP3, uterus
INTRODUCTION
Estradiol (E2) and testosterone (T) regulate a large number of genes in a variety of tissues throughout the body [1–5]. In turn, the actions of these steroid hormones are modulated by the enzyme 17beta hydroxysteroid dehydrogenase type 2 (HSD17B2), which was cloned more than a decade ago [6]. HSD17B2 catalyzes the conversion of E2 to estrone (E1) and T to androstenedione in a variety of tissues, such as the placenta, liver, kidney, small intestine, breast, and uterus [6–11]. By decreasing the local levels of biologically active steroid hormones, HSD17B2 may play a role in limiting the action of these hormones on target tissues.
HSD17B2 is highly expressed in the epithelial cells of the endometrium, where its expression is enhanced by progesterone [12, 13]. In nonpregnant women, progesterone-induced HSD17B2 expression in endometrial epithelium reduces local E2 concentrations [6]. In this context, the function of HSD17B2 is critical for normal endometrial growth and differentiation, which is regulated by the opposing actions of fluctuating levels of E2 and progesterone during the menstrual cycle. The proliferative phase (the first half of the menstrual cycle) is associated with ovarian follicle growth and increased estrogen secretion, which stimulates endometrial growth [14]. During the secretory phase (the second half of the menstrual cycle), progesterone secreted by the corpus luteum becomes dominant, and stimulates the secretion of a mucinous substance by endometrial epithelial cells in preparation for implantation. The alternating dominance of E2 and progesterone during the menstrual cycle is essential for the cyclical development of endometrium [14], and HSD17B2 plays an important role in the shift from the E2-directed endometrial proliferative phase to the progesterone-directed secretory phase.
In previous studies, we demonstrated that progesterone indirectly stimulates the expression of HSD17B2 in endometrial epithelial cells [13]. Progesterone, acting through the progesterone receptor (PR) in stromal endometrial cells, induces the release of a number of water-soluble paracrine factors that then stimulate HSD17B2 expression in adjacent endometrial epithelial cells [13]. These stromal cell-derived, progesterone-induced paracrine factors act as intermediaries to the epithelial cells to minimize the local effects of active E2 in the presence of progesterone. Consistent with this hypothesis, recent studies have shown that retinoids stimulate the expression of HSD17B2 gene in an endometrial epithelial cell line [15, 16]. Thus, retinoids may constitute a component of the hormonal cocktail secreted by endometrial stromal (ES) cells.
Subsequent studies demonstrated that paracrine induction of HSD17B2 in human endometrial cancer cells by PR-dependent stromal factors is conferred by a proximal promoter region (nt –200/–1) [13]. This region contains several potential cis-acting elements, including two SP1 binding sites (Fig. 1) [13]. The transcription factor SP1 was previously reported to bind to a GC box (GGGCGG or CCCGCC) and activate transcription of a subset of genes that contain this motif [17, 18]. SP1 stimulates eukaryotic transcriptional initiation by aiding in the formation of a functional pre-initiation complex consisting of RNA polymerase II, activator-proteins, and the target DNA [19–21]. The presence of two or more SP1 binding sites in a promoter usually enhances its transcriptional activation [20, 22].
|
The SP family of transcription factors contains four members: SP1 to SP4 [23, 24]. SP1 and SP3 compete for similar binding sites, and the ratio of these two transcription factors has been shown to modulate the relative rates of transcription of several genes [25–27]. SP1 and SP3 are ubiquitously expressed, and their functional significance in transactivating the promoters of various genes, such as murine ornithine decarboxylase [28], tissue factor [29], interleukin-6 [30], and prolactin receptor [31], have previously been reported. The presence of multiple SP1 binding sites in the HSD17B2 promoter prompted us to investigate whether progesterone-induced stromal factors stimulate epithelial HSD17B2 gene expression via SP1 and/or SP3.
MATERIALS AND METHODS
Cell Cultures and Generation of Conditional Media
Term human placentas were obtained from women with normal pregnancies and vaginal deliveries. The isolation of cells from decidua parietalis tissue dissected from the fetal membranes was started within 1–2 h after delivery. Collagenase-dispersed cells were isolated as described previously [13]. Human endometrial tissue from normally menstruating women was obtained by endometrial suction biopsy through S.E.B. For both the placental and endometrial tissue studies, informed consent was obtained, and the studies were approved by the institutional review boards at Northwestern Memorial Hospital and Northwestern University. The cells were dispersed by collagenase digestion, and the stromal cells were separated from contaminating epithelial cells by selective trypsinization, as described previously [13].
Three types of endometrial cells were used in these experiments: 1) human decidual fibroblasts (HuDF); 2) human ES cells; and 3) two lines of human malignant endometrial epithelial cells, ECC2 (a kind gift from Prof. Asgerally Fazleabas, University of Illinois at Chicago) and Ishikawa cells (a kind gift from Dr. Masato Nishida, Kasumigaura National Hospital, Tsuchiura, Ibaraki, Japan). Both HuDF and ES cells were used as ES cells in these experiments. These cells were originally isolated from separate endometrial tissues, spontaneously immortalized by serial passages, and found to express prolactin in response to treatment with progesterone [13]. The ECC2 and Ishikawa cell lines were derived from endometrial epithelial cells and express estrogen receptor (ER) 
, PR, and HSD17B2 [32, 33]. HuDF, ES, and ECC2 cells were cultured to confluence in RPMI 1640 medium (Invitrogen Life Technologies Inc., New York, NY) containing 10% fetal bovine serum (FBS). Ishikawa cells were cultured in Eagle minimal essential medium (Invitrogen Life Technologies Inc.) containing 10% FBS.
For these experiments, conditioned medium (CM) from progestin-treated HuDF or ES cells (chromatin immunoprecipitation-PCR [ChIP] assays only) was collected and used to treat ECC2 or Ishikawa epithelial cell lines in order to mimic the stromal-epithelial paracrine interactions present in the endometrium. CM was collected according to a previously published protocol [13]. Briefly, media from HuDF or ES cells was changed at 48-h intervals until the cells grew to 95% confluence. Media were then aspirated, and the cells were washed twice with phosphate-buffered saline (PBS). Cells were incubated in serum-free RPMI 1640 overnight. The next day, serum-free RPMI 1640 containing PR agonist R5020 (10–7 M) was added to HuDF cells (30 ml/T75 flask). CM was collected at the end of a 48-h period. Collected CM was routinely charcoal-stripped to remove any residual R5020, and was stored at –80°C. Frozen CM was thawed at 37°C immediately before addition to endometrial epithelial cell cultures. The serum-free RPMI1640 was used as control medium.
Reverse Transcription-Polymerase Chain Reaction
Total RNA was isolated from epithelial cells (ECC2 or Ishikawa) treated with CM or control media using the RNase mini kit (Qiagen, Valencia, CA) following the protocol suggested by the manufacturer. For reverse transcription-polymerase chain reaction (RT-PCR) analysis of HSD17B2 and glyceraldehydes-3-phosphate dehydrogenase (GAPDH) mRNA, the SuperScript First-Strand Synthesis System (Invitrogen Life Technologies Inc.) was used to synthesize the first strand cDNA with 5 µg total RNA, as instructed by the supplier. Primer sequences used for detection of HSD17B2 were 5'-AGA ATG AGC ACT TTC TTC TCG GAC-3' and 5'-GGC TTG TCT TGG CCA AAA TGT CTT-3', which yielded a 1 117 bp fragment. GAPDH was used as a control to ensure equal use of total RNA under different conditions. Primer sequences used for detection of GAPDH were 5'-CCA CCC ATG GCA AAT TCC ATG GCA-3' and 5'-TCA GAC GGC AGG TCA GGT CCA CC-3', which produced a 594-bp fragment. PCR conditions were as follows: 39 cycles of denaturing at 94°C for 30 sec, annealing at 55°C for 50 sec, and extension at 72°C for 60 sec.
Real-Time Quantitative PCR
Real-time quantitative PCR was performed with a 7000 Sequence Detection System (Applied Biosystems, Foster City, CA) employing an SYBR green Master Mix (Applied Biosystems). The oligonucleotides for detection of SP1 were 5'-CTT GGT ATC ATC ACA AGC CAG TT-3' and 5'-TCC CTG ATG ATC CAC TGG TAG TA-3', and primers for detection of SP3 were 5'-TTG ACT ACA TCT AGT GGG CAG GT-3' and 5'-TAC AAC AGG CTG TGC TGT AGA AA-3'. In each instance, the amount of real-time quantitative PCR product for the gene of interest was normalized to the amount of GAPDH in the same sample.
Transient Transfections and Luciferase Assays
Transient transfection of ECC2 cells was carried out in 35-mm dishes with the Lipofectamine plus reagent (Invitrogen Life Technologies Inc.) with the following plasmids: 1) 1 µg pGL3-basic luciferase reporter plasmid, containing nt –1244/–1 fragment, progressively truncated HSD17B2 promoter fragments, or SP binding-site mutations in the nt –200/–1 construct [13]; 2) 1 ng pRL-TK plasmid as an internal control (Promega, Madison, WI); and 3) 1 µg pCMV4 expression vector that contained SP1 cDNA (a generous gift form Dr. Dimitris Kardassis, Crete, Greece) or pCMV4-Pfu expression vector that contained SP3 cDNA (a generous gift from Drs. Johnathan Horowitz, North Carolina State University, Raleigh, NC, and Graciela Krikun, Yale University, New Haven, CT). After transfection for 4 h in serum-free medium, medium was changed to RPMI 1640 with antibiotics, 10 mM HEPES, and 10% FBS. After an overnight incubation, cells were starved in serum-free medium for 12 h, then treated for 48 h with control media or CM collected from R5020-pretreated HuDF cells. For SP1 and SP3 dose-response experiments, cells were transfected with increasing amounts of SP1 and SP3 expression vectors (0, 1, 10, 100, and 500 ng). After treatment, transfected cells were washed twice in PBS and lysed in 250 µl of lysis buffer (0.1 M potassium phosphate [pH 7.8], 1% Triton X-100, 1 mM dithiothreitol, 2 mM EDTA). Luciferase assays were performed with 10 µl of cell lysate employing a Dual-Luciferase Reporter Assay System kit (Promega). Luminescence was measured with a LUMAT LB9507 luminometer (EG&G Berthold, Bad Wildbad, Germany), and luciferase activity was normalized to pRL-TK values.
Immunoblotting
Nuclear proteins used for immunobloting were prepared as previously described [34]. ECC2 nuclear proteins were electrophoresed on 7.5% polyacrylamide sodium dodecyl sulfate (SDS)-gels (5% stacking, 7.5% separating gels) at 25 mA for 15 min and then at 45 mA for 50 min. Proteins were then transferred to a nitrocellulose membrane in Transblot buffer (25 mM Tris, 192 mM glycine, and 20% methanol) at 4°C for 12 h at 50 V (Bio-Rad Laboratories, Hercules, CA). Membranes were then blocked with 5% milk in Tris-buffered saline (TBS) buffer (20 mM Tris-Borate [pH 7.2], 140 mM NaCl) overnight, and incubated with an anti-SP1 monoclonal antibody at 1:1 000 dilution or anti-SP3 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:1 000, for 1 h at room temperature. Membranes were then washed five times for 10 min with 0.1% Tween-20 in TBS and incubated with a 1:2 500 dilution of secondary antibody (anti-mouse IgG-horseradish peroxidase; Santa Cruz Biotechnology) for 1 h at room temperature. Next, membranes were washed five times (10 min each) with 0.1% Tween-20 in TBS. The signal was detected using chemiluminescence (SuperSignal Ultra Chemiluminescent Kit; Pierce, Rockford, IL) according to the manufacturer's protocol.
Site-Directed Mutagenesis
To generate plasmids bearing mutated consensus-binding sequences for transcription factors SP1 and SP3, site-directed mutagenesis was performed using the QuickChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA) following the manufacturer's instructions. A chimeric luciferase construct containing the nt –200/–1 region of HSD17B2 promoter [13] was used as a template for site-directed mutagenesis. Briefly, PCR using the DNA template (50 ng), mutagenesis primers (125 ng), deoxyribonucleotide triphosphate, and PfuTurbo DNA polymerase (2.5 U/µl) was carried out for 15 cycles at 95°C for 30 sec, 55°C for 60 sec and 68°C for 10 min. Dpn I restriction enzyme (10 U/µl) was added directly to the PCR products and incubated at 37°C for 60 min to eliminate the DNA template. The mutagenesis reaction was then used to transform DH-5-competent cells. Mutation of the binding consensus site was confirmed by DNA sequencing. The oligonucleotides used for mutagenesis are as follows, with the mutated nucleotides underlined (the complementary strand sequence is not shown): mutant SP1 (nt –76/–65), 5'-TAAAGGAGAGGCG-3'; mutant SP1 (nt –82/–72), 5'-AGAATGTAGGG-3'.
Electrophoretic Mobility Shift Assays (EMSA)
Nuclear extracts used for EMSA were prepared as previously described [34]. ECC2 cells were incubated in serum-free RPMI1640 medium for 24 h before they were processed for the extraction of nuclear proteins. The sequence of the double-stranded oligonucleotide used for EMSA was 5'-AGG GGG TGG GGA GGG GCG AGA GCG GT-3', which is identical to a 26-bp-long sequence (nt –82/–57) within the regulatory region of the HSD17B2 gene and contains two overlapping SP1 binding sites (nt –82/–72 and nt –76/–65), was synthesized by Invitrogen Life Technologies, Inc. After renaturing two single-stranded oligonucleotides to a duplex oligonucleotide, we end-labeled this with [
-32P]ATP using T4 kinase. The free [-32P]ATP was removed by Bio-Spin 6 chromatography columns (Bio-Rad). EMSAs were performed as previously described [13]. Briefly, 2-µg samples of nuclear extracts were incubated with the radiolabeled double-stranded oligonucleotide probe (20 000 cpm/reaction) for 15 min at room temperature in a reaction buffer containing 20 mM HEPES (pH 7.6), 75 mM KCl, 0.2mM EDTA, 20% glycerol and 2 µg of poly(dI-dC)-poly(dI-dC) as a nonspecific competitor. Protein-DNA complexes were resolved on 6% nondenaturing polyacrylamide gels with 0.5x Tris-borate-EDTA running buffer and visualized by autoradiography. For competition EMSA, a nonradiolabeled double-stranded oligonucleotide was added simultaneously with the labeled probe. Immunodepletion was performed after the addition of 0.5 µl of an antibody against SP1 (Santa Cruz Biotechnology) or SP3 (Sigma Inc., St. Louis, MO) to the binding reaction, followed by a 30 min incubation on ice before electrophoresis.
ChIP Assays
ChIP assays were performed using a ChIP kit (Upstate Biotechnology, Inc., Lake Placid, NY) following the manufacturer's instructions. Briefly, Ishikawa cells were incubated for 48 h with: 1) CM from R5020-treated ES cells; 2) mithramycin A (Sigma Inc.); or 3) CM from R5020-treated ES cells plus mithramycin A. Cells were then cross-linked with 1% formaldehyde at 37°C for 10 min. After being washed with cold PBS containing protease inhibitors, the cells were harvested and suspended in lysis buffer (50 mM Tris-HCl [pH 8.1], containing 1% SDS and 10 mM EDTA). The lysate was sonicated to generate DNA fragments with an average length of 200 to 1 000 bp. After removal of cell debris by centrifugation, the supernatant was incubated with 10 µg of specific antibodies to SP1, SP3, or nonspecific rabbit IgG (Santa Cruz Biotechnology). The immune complexes were collected using protein A-agarose/DNA beads and washed five times using low salt, high salt, lithium chloride, and Tris-EDTA buffers. The absorbed immune complexes were recovered by incubation with elution buffer (1% SDS and 0.1 M NaHCO3). After reversing cross-links at 65°C for 4 h, the genomic DNA was purified by phenol extraction. The HSD17B2 promoter region was detected by PCR amplification using primers flanking the SP1 binding sites (sense 5'-TCC AGT TAG TCA TCG CTC CAG-3' and antisense 5'-TTC CTA TTT GAC CGC TCT GCG-3') that yielded the predicted 128-bp product.
Human Tissues and Immunohistochemistry
Endometrial samples were collected at the time of hysterectomy from 25 premenopausal women (age range: 36–47 yr) and were immediately fixed in 10% formalin and embedded in paraffin. The cycle phase for each specimen was determined independently by two gynecologic pathologists (T.K. and H.S.) according to histological criteria. Written informed consent approved by the institutional review boards of Tohoku and Northwestern Universities was obtained before the collection of samples.
Immunohistochemistry to localize HSD17B2, SP1, and SP3 was performed as previously described on 2.5-µm-thick sections mounted on silane-coated slides by the biotin-streptavidin-amplified technique [13, 35, 36]. Monoclonal antibodies against SP1 and SP3 were purchased from Santa Cruz Biotechnology. The monoclonal antibody against HSD17B2 was produced as previously described [37].
Immunoreactivity was quantified by an H-scoring system described previously with modifications [38]. H-scores were generated by adding 2x percentage of strongly-stained nuclei (SP1 and SP3) or cytoplasms (HSD17B2) in 10 high-power fields and 1x percentage of weakly-stained nuclei or cytoplasms in 10 high-power fields, giving a range of scores from 0–200 [13, 35].
Statistical Analysis
Statistical differences between sample means were determined by analysis of variance followed by post-hoc multiple comparison testing using the Student-Neuman-Keuls test. Values are expressed as mean ± SEM, and P < 0.05 is considered statistically significant.
RESULTS
Mithramycin A Decreased both Basal and CM-Induced HSD17B2 mRNA Levels
Mithramycin A (also known as aureolic acid, mithracin, or plicamycin) is an aureolic acid-type polyketide produced by the soil bacteria Streptomyces argillaceus [39]. Mithramycin A binds to GC-rich regions of chromatin and interferes with the transcription of genes that bear GC-rich motifs [40, 41]. To determine whether the expression of HSD17B2 is regulated by SP1, which binds to the GC-rich motifs GGGCGG or CCCGCC, we treated ECC2 cells with increasing concentrations of mithramycin A (0, 1, 40, 100, and 400 nM) and then measured HSD17B2 mRNA levels. As shown in Fig. 2A, mithramycin A decreased basal HSD17B2 mRNA levels in a dose-dependent manner, suggesting that SP1 or other GC-targeted transcription factors may be involved in regulation of HSD17B2 gene expression in epithelial cells. Mithramycin A treatment did not alter GAPDH mRNA levels.
|
To determine whether progesterone-mediated stromal paracrine induction of HSD17B2 mRNA levels in ECC2 or Ishikawa epithelial cells is mediated via GC-targeted transcription factors, such as SP1, we treated epithelial cells with CM or with CM plus mithramycin A. As shown in Fig. 2B, direct treatment of epithelial cells with CM resulted in a robust increase in HSD17B2 mRNA level, whereas the addition of mithramycin A reduced CM-induced HSD17B2 mRNA level by 60% (P < 0.001). This suggests that the induction of epithelial HSD17B2 expression by stroma-derived paracrine factors may be, at least in part, mediated by GC-targeted transcription factors, such as those in the SP family. Comparable effects on HSD17B2 expression were obtained using Ishikawa cells (data not shown).
Mithramycin A Inhibited Basal and CM-Induced HSD17B2 Promoter Activity in Transfected ECC2 Cells
To investigate the role of GC-targeted transcription factors in mediating the induction of HSD17B2 promoter activity by CM, we transfected a chimeric construct of the HSD17B2 promoter coupled with a luciferase gene (pHSD[–200/–1]-Luc) into ECC2 or Ishikawa cells and then treated with CM or control media. As shown in Fig. 3A, CM induced HSD17B2 promoter activity by approximately 3-fold in ECC2 cells, whereas mithramycin A inhibited CM-induced promoter activity by over 60% (Fig. 3A). Mithramycin A also decreased basal HSD17B2 promoter activity in a dose-dependent fashion in ECC2 cells (Fig. 3B). Comparable effects of CM and mithramycin A were seen in HSD17B2 reporter-transfected Ishikawa cells (data not shown).
|
SP1 and SP3 Overexpression Stimulated HSD17B2 Promoter Activity in Endometrial Epithelial Cells
To determine whether SP1 or SP3 mediate HSD17B2 gene transcription, cotransfection experiments were performed in ECC2 or Ishikawa cells with pHSD(–200/–1)-Luc and increasing concentrations of expression vectors encoding wild-type SP1 or SP3. As shown in Fig. 4, SP1 and SP3 overexpression stimulated HSD17B2 promoter activity in ECC2 cells in a dose-dependent fashion, with 0.5 µg/well of SP1 and SP3 stimulating HSD17B2 promoter activity by 9- and 6-fold, respectively.
|
Deletion Analysis of the HSD17B2 Promoter
To identify the region(s) of the HSD17B2 promoter that are critical for transactivation by SP1 or SP3, a series of deletion constructs (nt –1244 to –1) were cotransfected into ECC2 cells with either the SP1 or SP3 expression vector. As shown in Fig. 5, SP1 induced pHSD(–1244/–1)-Luc, pHSD(–750/–1)-Luc, and pHSD(–200/–1)-Luc by 6-, 6-, and 9-fold, respectively. Truncation of the promoter to nt –65 abolished SP1 and SP3 activation of HSD17B2. These results indicate that sequences between the nt –200 to –65 are critical for SP1 or SP3 stimulation of HSD17B2 gene expression in ECC2 cells. Transfection of deletion constructs and SP expression vectors into Ishikawa cells produced similar results that were consistent with previous reports (data not shown) [13].
|
Site-Directed Mutagenesis of GC-Rich Motifs in HSD17B2 Promoter
To characterize the putative GC-rich motifs in the nt –200 to –65 region of the HSD17B2 promoter, we designed a series of reporter constructs in which the binding sites were individually mutated by site-directed mutagenesis. Mutation of the GC-rich motif at either nt –82/–72 or nt –76/–65 not only decreased basal HSD17B2 promoter activity by 50% (P < 0.05), but also blocked the stimulatory effect of SP1 overexpression by 90% (Fig. 6). Mutation of both GC-rich motifs abolished SP1 induction of HSD17B2 promoter activity (Fig. 6).
|
To determine whether the two GC-rich motifs on the HSD17B2 promoter are critical for transactivation by CM, we transfected ECC2 cells with the wild-type reporter construct or its GC-rich motif mutants and treated with CM or control media. As shown in Fig. 7, disruption of either GC-rich motif resulted in a marked decrease in CM-induced promoter activity (Fig. 7). Interestingly, mutation of both GC-rich motifs did not lead to a further reduction in CM-stimulated promoter activity (Fig. 7), suggesting that additional cis-acting elements may also be involved in mediating CM-stimulated gene expression.
|
Gel Shift and Supershift Studies of GC-Rich Motifs
To elucidate whether nuclear proteins bind specifically to the these GC-rich motifs, a gel shift assay was performed with a 26-bp probe, containing the two critical GC-rich motifs at nt –82/–72 and nt –76/–65, and nuclear extracts from ECC2 cells. Fig. 8 shows that a specific DNA-protein complex was formed when a 32P-labeled oligonucleotide probe was combined with a nuclear extract from the ECC2 cells (lane 2, arrow). The formation of the complex was competed by an unlabeled wild-type oligonucleotide (lane 4), but not by an oligonucleotide containing mutated SP binding sites (lane 3). Treatment with CM enhanced binding activity to the probe (lane 5), which was decreased when antibodies against SP1 (lanes 6 and 8) or SP3 (lanes 7 and 9) were added. These data suggest that both SP1 and SP3 significantly contribute to the formation of the protein complex in this region of the HSD17B2 promoter. Normal mouse IgG did not cause a depletion of this complex (lanes 10 and 11).
|
CM Increases SP1 and SP3 mRNA and Protein Levels in ECC2 Cells
Because both CM and overexpression of SP1 or SP3 increased HSD17B2 expression in ECC2 cells, we next determined whether CM had a direct effect on SP1 and SP3 mRNA and protein levels. As shown in Fig. 9A, CM from untreated stromal cells did not increase SP1 or SP3 protein levels (lane 2). CM from R5020-treated stromal cells, however, significantly increased SP1 and SP3 levels (lane 3) in ECC2 cells. Fig. 9B showed that CM from R5020-treated stromal cells significantly increased SP1 and SP3 mRNA levels in Ishikawa cells, whereas CM from untreated stromal cells did not have this effect. These results indicate that stromal-derived progesterone-induced paracrine factors may stimulate epithelial HSD17B2 gene transcription by upregulating SP1 and SP3 levels.
|
Endogenous SP1 and SP3 Interaction with the HSD17B2 Promoter Region Containing GC-Rich Motifs Is Stimulated by CM and Inhibited by mithramycin A
ChIP assays were performed to determine whether the HSD17B2 promoter region containing GC-rich motifs is occupied by endogenous SP1 and SP3 proteins. Sheared chromatin isolated from Ishikawa cells, treated with CM, mithramycin A, or CM plus mithramycin A, were immunoprecipitated with antibodies against SP1 or SP3. The presence of HSD17B2 promoter DNA in the immunoprecipitation was determined by PCR amplification. As shown in Fig. 10, HSD17B2 DNA was specifically immunoprecipitated with anti-SP1 (lanes 5–8) or anti-SP3 (lanes 9–12) antibodies, but not with a nonspecific rabbit IgG (lanes 1–4). Compared with controls (lanes 5–9), CM treatment increased SP1 (lane 7) and SP3 (lane 11) interaction with the HSD17B2 promoter by 7- and 2-fold, respectively. Furthermore, mithramycin A not only decreased SP1 (lane 6) and SP3 (lane 10) binding to the HSD17B2 gene, but also blocked binding stimulated by CM (lanes 8 and 12). These results demonstrate that stroma-derived progesterone-induced paracrine factors in CM increased endogenous SP1 and SP3 binding to the HSD17B2 promoter region, most likely at the GC-rich motifs.
|
In Vivo Temporal and Spatial Association between HSD17B2, SP1, and SP3 in Human Endometrial Epithelial Cells
In vivo, levels of circulating progesterone strikingly increase during the secretory phase of the human menstrual cycle compared with the proliferative phase. Thus, immunohistochemical analysis was performed to determine whether the expression of SP1 and SP3 proteins is increased in endometrial epithelial cells during the secretory phase of the menstrual cycle, when HSD17B2 expression is induced. Sections of human endometrium from the proliferative and secretory phases of the menstrual cycle were incubated with SP1, SP3, and HSD17B2 monoclonal antibodies. As shown in Fig. 11A, HSD17B2 was exclusively expressed in glandular epithelial cells of human endometrium (Fig. 11A, panels a and b), only during the secretory phase of the menstrual cycle (Fig. 11B). Conversely, SP1 was present in both stromal and epithelial cells, as previously reported (Fig. 11A, panels c and d) [18], yet only epithelial SP1 immunoreactivity increased significantly during the secretory phase (Fig. 11B). No statistically significant change in stromal SP1 immunoreactivity was observed with respect to cycle phase. We observed significantly increased epithelial SP3 immunoreactivity during the secretory phase (Fig. 11B). The parallel increases in SP1 and SP3 protein levels and HSD17B2 levels in endometrial epithelial cells during the secretory phase of the menstrual cycle suggest that SP1 and SP3 may be involved in mediating the progesterone-dependent increase in HSD17B2 in endometrial epithelium.
|
DISCUSSION
The human endometrium undergoes a series of morphological and biochemical changes involving proliferation (proliferative phase) and differentiation and secretion (secretory phase) during the human menstrual cycle. These processes are strictly controlled by estrogen and progesterone [42]. Estrogen generally acts as a potent mitogen on endometrial tissues, whereas progesterone reverses this estrogenic effect. The inactivation of E2 to E1 by HSD17B2 in the endometrial epithelium during the secretory phase has been viewed as an important protective mechanism against overproliferation of this E2-responsive tissue. Previous studies have shown that this protective mechanism is not present in the pathologic endometriotic tissues, thus giving rise to abnormally increased local levels of E2 and consequent growth in endometriosis [37].
HSD17B2 is highly expressed in the glandular epithelial cell fraction of the human endometrium during the secretory phase in response to progesterone [12, 13]. Previous studies revealed that this response was indirect in that the HSD17B2 gene is actually induced by stroma-derived, progesterone-induced paracrine factors [13]. As a natural follow-up to our previous work, we have demonstrated that, in response to these paracrine factors, there is an increase in SP1 and SP3 transcription factor levels and binding to a critical region in the HSD17B2 promoter. A putative model of the mechanism by which progesterone indirectly regulates HSD17B2 via stromal-epithelial paracrine communication and SP transcription factor activation is shown in Fig. 12.
|
Numerous studies have shown that gene expression is tightly controlled by the combined action of multiple transcription factors that interact with specific DNA sequences [43, 44]. Transcription factors may also bind co-activators and repressors that enhance or repress transcriptional activity. Because neither mithramycin A treatment nor site-directed mutagenesis of SP binding sites completely blocked CM-induced HSD17B2 expression in ECC2 cells, it is highly likely that other cis regulatory element(s) in the HSD17B2 promoter are involved in upregulation of this gene in response to stromal paracrine factors. The proximal region of HSD17B2 promoter harbors binding sites for several sequence-specific transcription factors, including EGR1, CEBPs, and JUN/FOS [13]. It will be important to elucidate whether these transcription factors are important for basal or stromal paracrine factor-induced transcription of the HSD17B2 promoter.
There are a growing number of examples of the overlapping roles of SP1 and SP3 in the regulation of target genes [20, 22, 26, 45–48]. Our finding that the proximal region of the HSD17B2 promoter binds both SP1 and SP3, and that mutations in each site dramatically impair its transcriptional activity, strongly suggest cooperation between the two SP binding sites. Significant cooperation between multiple adjacent SP1 sites has been demonstrated as necessary for maximal activation of several promoters (e.g., tissue factor gene) [26, 49]. Synergism and the possible physical interaction of SP1 and SP3 need to be analyzed. In addition, the binding and transactivating properties of SP1 can be modulated by posttranslational modifications, such as glycosylation and phosphorylation. In this study, protein levels of SP1 or SP3 were found to be increased in response to stromal paracrine factors, but the role of phosphorylation in SP1-stimulated HSD17B2 expression needs to be further investigated.
In summary, the proximal nt –200/–1 region upstream of the transcription initiation site of HSD17B2 contains two GC-rich motifs that are required for basal expression of human endometrial epithelium. Transcription factors SP1 and SP3 bind to these regulatory elements, and appear to be important for regulation of HSD17B2 by stroma-derived, progesterone-induced paracrine factors acting on endometrial epithelial cells. We have also shown that expression of SP1 and SP3 transcription factor is significantly increased in endometrial epithelial cells during the secretory phase of the menstrual cycle with a pattern similar to that of HSD17B2, suggesting that SP1 and SP3 may play critical roles in the regulation of morphological and biochemical changes of human endometrium in vivo. Studies are under way to identify partners of SP1 and SP3 in the transcriptional complex that enhance HSD17B2 promoter activity. We are also analyzing the CM for the isolation of progesterone-dependent stromal factors that induce epithelial HSD17B2.
ACKNOWLEDGMENTS
We thank Sijun Yang, Zongjuan Fang, Zhihong Lin, Joy Innes, and Scott Reierstad for their technical assistance, and Dr. Julie Kim for her suggestions. We thank Dr. Alf Stefan Andersson for his suggestions and review of the manuscript, Dr. Dimitris Kardassis for the human SP1 expression vector, and Drs. Johnathan Horowitz and Graciela Krikun for the human SP3 expression vectors.
FOOTNOTES
1 Supported by National Institutes of Health grant HD40093 to S.E.B. and grants from the Friends of Prentice and AVON Foundation. 
2 Correspondence: Serdar E. Bulun, Northwestern University, Feinberg School of Medicine, Robert H. Lurie Medical Research Center, 303 E. Superior Street, Suite 4–123, Chicago, IL 60611-3095. FAX: 312 503 0095; s-bulun{at}northwestern.edu
Received: 23 February 2006.
First decision: 14 March 2006.
Accepted: 13 June 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  L. Aghajanova, A. Hamilton, J. Kwintkiewicz, K.C. Vo, and L.C. Giudice Steroidogenic Enzyme and Key Decidualization Marker Dysregulation in Endometrial Stromal Cells from Women with Versus Without Endometriosis Biol Reprod, January 1, 2009; 80(1): 105 - 114. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-H. Cheng, P. Yin, Q. Xue, B. Yilmaz, M. I. Dawson, and S. E. Bulun Retinoic Acid (RA) Regulates 17{beta}-Hydroxysteroid Dehydrogenase Type 2 Expression in Endometrium: Interaction of RA Receptors with Specificity Protein (SP) 1/SP3 for Estradiol Metabolism J. Clin. Endocrinol. Metab., May 1, 2008; 93(5): 1915 - 1923. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J Su, Y.-H. Cheng, R. T Chatterton, Z.-H. Lin, P. Yin, S. Reierstad, J. Innes, and S. E Bulun Regulation of 17-Beta Hydroxysteroid Dehydrogenase Type 2 in Human Placental Endothelial Cells Biol Reprod, September 1, 2007; 77(3): 517 - 525. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
