| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Cell Biology and Neuroscience, University of South Carolina School of Medicine, Columbia, South Carolina 29208
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5) was cloned from pig ovary. RNase protection assays revealed that ERß mRNA was expressed in the preovulatory follicles and early, midluteal, and regressing corpora lutea (CL) of eCG ± hCG-primed gilts. ERß and ERß5 transcripts were shown by semiquantitative reverse transcription polymerase chain reaction to be expressed at a ratio of approximately 2:1 in granulosa cells, small, medium, and large antral follicles, and midluteal phase corpora lutea of unprimed animals. Immunoreactive ERß proteins corresponding to the size of in vitro translated ERß and ERß5 were detected by immunoblot. Full-length ERß was detected in granulosa, small, medium, and large antral follicles, and midluteal phase CL of unprimed animals. Putative ERß5 immunoreactive bands were abundant only in granulosa cell extracts. In COS-1 cells, transfected ERß5 had no effect on basal transcription of an estrogen-responsive reporter construct but did repress wild-type ERß transactivation when cotransfected at 10-fold excess plasmid. No repression of ER
transactivation was observed. In primary granulosa cell cultures, transfected ERß5 plasmid did not inhibit basal reporter activation. ERß5 was shown by immunofluorescence to localize to the nucleus in transfected COS-1 cells. In vitro translated ERß5 proteins bound estrogen response elements in DNA in electrophoretic mobility shift assays, as indicated by supershift analysis. ERß is abundant in porcine ovary, and a naturally occurring splice variant missing exon 5 may have biological function.
corpus luteum, estradiol receptor, granulosa cells, ovary, steroid hormone receptors
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(ER) and the more recently identified estrogen receptor ß (ERß) [5]. The cDNAs of both ER and ERß have been identified in the ovaries of mammals, including human and nonhuman primates, rodents, cows, and sheep [6–10]. A consistent observation has been the expression of ERß in granulosa cells, but its expression in different stages of corpus luteum (CL) is more variable among these species. The ability of ovarian estrogen to exert various autocrine and paracrine actions is most likely dependent on its temporal expression and abundance of ER subtypes in specific ovarian cells. Therefore, it is necessary to determine the cycle-dependent expression and distribution of ER isoforms and their variants to understand ovarian estrogen action in different species.
Estrogen is luteotrophic in the pig [11–14] and exerts time-dependent inhibitory and stimulatory effects on steroidogenesis in cultured granulosa cells [15]. Porcine follicles (3–5 mm) exhibit high-affinity nuclear and cytosolic estrogen binding sites [16]. To better understand the role of ERß in porcine ovary, we cloned its cDNA. In the process, we isolated a splice variant, ERß5, that lacks 139 base pairs (bp) resulting from a deletion of exon 5. This variant transcript has recently been detected in bovine and sheep ovaries and in human testis and mammary gland, but whether a corresponding protein exists has not been evaluated [10, 17–19]. Little is known about the ER isoforms in the porcine ovary. Slomczynska et al. [20, 21], using histological sections of porcine ovaries and heterologous probes, localized ERß message and protein to granulosa and theca of all sizes of antral follicles and all stages of CL. The ERß antibody used in those studies recognized the carboxyl-terminus of ERß and thus would not detect the truncated translation product of the exon 5 variant. Also, the in situ hybridization studies did not differentially detect ERß variants. Our goal was to quantify the total amount of ERß mRNA in hormonally primed gilts to evaluate ERß mRNA temporal expression in dated ovarian structures. We also wanted to determine the relative abundance of the wild-type and ERß5 variant at both the mRNA and protein level in unprimed porcine ovaries.
ER transcripts lacking exon 5 (ER5) similar to the ERß5 variant have been identified in other species [22]. Depending on the cell type studied, transfected ER5 demonstrated no significant transactivation ability, dominant negative activity, or constitutive activity at estrogen response elements (EREs) [23–25]. The potential activity of ERß5 may therefore be influenced by its cellular context. In this study, our goal was to characterize the transactivation abilities of ERß5 in COS cells and granulosa cells and to determine its nuclear localization and ability to bind DNA.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
General chemicals including protease inhibitors were purchased from Sigma (St. Louis, MO). Plasmids pCRII and pcDNA3.1 were purchased from InVitrogen (Carlsbad, CA). MMLV-H+ and H- reverse transcriptases, random hexamers, restriction enzymes, RNase-free RQ DNase, T4 polynucleotide kinase, and plasmid pRLtkLuc were obtained from Promega Corp. (Madison, WI). Radioisotopes were purchased from Amersham Pharmacia Biotech (Piscataway, NJ). Cell culture reagents, Trizol, LipofectAMINE, and oligonucleotides were obtained from Life Technologies (Gaithersburg, MD). Plasmid purification reagents were obtained from Qiagen (Valencia, CA). Porcine ovaries were obtained from a local abattoir with the exception of those used for RNase protection analysis.
Cloning of Porcine ERß Isoforms
RNA was extracted from adult porcine ovary tissue using Trizol reagent and homogenization. One microgram of total RNA was reverse transcribed using MMLV-RNase H- reverse transcriptase and the SMARTRACE cDNA Amplification kit (Clontech Laboratories, Palo Alto, CA) according to the manufacturer's suggestions. Only a partial cDNA for porcine ERß was available in GenBank (accession no. AF164957), so a combination of bovine (accession no. AF110402) and porcine oligonucleotide primers (shown in bold) were used to isolate the complete porcine coding sequence by polymerase chain reaction (PCR). A 477-bp amino terminal cDNA was isolated with primers A: 5'-GCTGTTACCTACTCAAGACATGGAT-3' and B: 5'-GCATAATCACTGCAGACAGCACAGAAGTGG-3' using the Advantage II polymerase and PCR kit (Clontech). The 3' remainder of the full-length ERß was isolated by PCR using the primers C: 5'-CATCCATTGCCAGCCGTCACTTCTGTATGC-3' and D: 5'-TCACTGAGCCTGGGGTTTCTGGGAGCC-3'. The exon 5-deleted cDNA was initially detected with primers C and D. Subsequently, cDNAs for the full-length (1581 bp) and the exon 5 deletion variant (1442 bp) were isolated using primers A and D. Gel-purified PCR products were cloned into the pCRII plasmid for sequence verification and subsequently cloned into the BstXI site of pcDNA3.1. The full-length porcine ERß sequence has been submitted to GenBank (accession no. AF267736). Sequences were confirmed in 2 animals.
The porcine ER complete coding sequence was cloned by PCR using oligonucleotide primers 5'-GGCTGTGCTCTTCTTCCAGGTGG-3' and 5'-GCTGGATGCATGCCAGAGTGTG-3' derived from GenBank (accession nos. AF034972 and AF034974) and ovary cDNA made with the SMARTRACE kit. PCR was performed using the Advantage-GC 2 PCR kit (Clontech). Gel-purified PCR products were cloned into the pCRII plasmid for sequence verification and subsequently cloned into the EcoRI site of pcDNA3.1.
Hemagglutinin (HA)-tagged estrogen receptor sequences were synthesized by PCR using Pfu polymerase (Stratagene, La Jolla, CA) and the above plasmids containing the estrogen receptor cDNAs. Primers were designed to include the last 15 nucleotides of the ER carboxyl-terminal coding sequence plus an additional 27 nucleotides coding in frame for the influenza hemagglutinin peptide YPYDVPDYA [26]. Sequences were inserted into pCRII, excised, and subsequently ligated into either the EcoRI site (ER) or the BstXI sites (both ERß cDNAs) of pcDNA3.1 in frame with a stop codon. All plasmid inserts were sequenced to confirm products and orientation of DNA (DNA sequencing facility, University of Maine, Orono, ME).
Ribonuclease Protection Assays
RNA was isolated from porcine tissues of eCG ± hCG-stimulated gilts collected as part of a previously published study [27]. Dated tissues were collected relative to the day of the hCG injection (Day 0). RNA included unstimulated ovary (Day -3), preovulatory follicles from eCG-treated gilts (Day -1 and Day 0), post-hCG-treated preovulatory follicles (Day 1), and CL (Days 2, 8, 12, and 20). The pCRII plasmid containing the N-terminal region ERß 477-bp cDNA was linearized at an internal SalI site and transcribed using SP6 polymerase to yield a 338-nucleotide riboprobe that contains nucleotides 200–458 of the ERß coding sequence. Riboprobes were synthesized using the Maxiscript (ERß) or Megascript kit (18S) (Ambion, Austin, TX) and -32P-UTP. Total RNA (10 µg) was cohybridized with gel-purified saturating amounts of high-specific-activity ERß riboprobe and low-specific-activity 18S riboprobe. RNase protection assays were performed using the RPA II as previously described [27]. Hybridized products were electrophoresed on 6% acrylamide gels containing 8% urea. Gels were exposed to BioMAX MR film (Kodak, Rochester, NY) with intensifying screens for up to 6 days. Autoradiographic bands were quantified using a Molecular Imager FX system and Quantity One v. 4.2.1 software (Bio-Rad, Hercules, CA).
Reverse Transcription PCR Analysis
A riboprobe designed to span ERß exons 4–6 was unable to detect ERß in ovarian tissues; therefore, semiquantitative reverse transcription (RT)-PCR was utilized to detect both the intact ERß transcript and the exon 5 deletion. Oligonucleotide primers corresponding to porcine ERß sequences in exons 4 and 6, 5'-ACACAACCCGAGTGAAGGAG-3' and 5'-TCATGGCCTTGACACAGAGA-3', were used for PCR. Porcine ribosomal protein S16 was used as an internal control with primers 5'-GCCTTCCAAGGGTCCTCTAC-3' and 5'-CACACCAGCAAATCGTTCC-3' derived from GenBank (accession no. AW787100). RNA was isolated using Trizol reagent from granulosa cells (from 1- to 5-mm antral follicles) of prepubertal gilts, and small follicles (1–2 mm), medium follicles (2–4 mm), a mix of large follicles (5–10 mm), healthy preovulatory large follicles (8–10 mm), and midluteal phase CL dissected from the ovarian stroma of cycling pigs. Ten to 40 µg of RNA for each sample was treated with RNase-free RQ DNase for 30 min at 37°C, followed by reextraction with Trizol reagent. DNase-treated RNA (3 µg) in sterile water was mixed with 1.5 µg random hexamers and reverse transcribed using MMLV-H+ polymerase (200 U) in 25-µl reactions for 1 h at 37°C according to the manufacturer's suggestions. RNA concentrations were measured spectrophotometrically prior to and after DNase treatment.
PCR was carried out using 5 µl of each cDNA reaction, 100 pmoles of each primer, 2 mM MgCl2, 1 µCi -32P-dCTP, 100 µM dNTPs, and 1.25 U Taq polymerase in 1x PCR buffer B (Fisher Scientific, Fair Lawn, NJ). Amplification conditions consisted of denaturation at 95°C for 2 min, followed by 27 cycles of 95°C for 60 sec, 47°C for 45 sec, and 72°C for 1 min, followed by a final extension at 72°C for 7 min. This cycle number was determined to be in the linear range for each primer set in preliminary experiments. A water control was carried through all reactions. Five microliters of each PCR was loaded onto a native 6% polyacrylamide gel and electrophoresed for 2 h at 240 V. Gels were dried and exposed to BioMAX MR film, and autoradiographic bands were analyzed as for the RNase protection assays. Expected products were 459 bp for intact ERß, 320 bp for exon 5-deleted ERß cDNA, and 196 bp for porcine S16 cDNA.
Protein Extraction and Immunoblot Analysis
Whole cell lysates were prepared by homogenizing tissues in 7–10 volumes of ice-cold buffer containing 50 mM Tris, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 mM sodium vanadate, and 1 mM sodium fluoride using 10 strokes of a Dounce homogenizer. Cells were incubated on ice for 30 min and then centrifuged at 13 000 x g for 15 min at 4°C to pellet cellular debris. Supernatants were retained. Nuclear extracts were isolated by resuspending granulosa cells in approximately 10 volumes of buffer A (50 mM Hepes, pH 7.5, 10 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 0.5 mM PMSF, 20 µg/ml aprotinin, 10 µg/ml leupeptin, 1 µg/ml pepstatin, and 1 µg/ml soy bean trypsin inhibitor). Triton X-100 was added (0.5%), and nuclei were pelleted for 30 sec (5000 rpm). Nuclei were resuspended in 2 volumes of buffer C (same as buffer A except with 400 mM KCl) and allowed to incubate for 15 min on ice prior to centrifugation to pellet debris. Whole cellular extracts and nuclear extracts were snap frozen and stored at -70°C. Protein concentrations were determined using Bio-Rad dye reagent.
Equivalent amounts of protein (50 or 100 µg) were mixed with Laemmli buffer, boiled for 5–10 min, and electrophoretically separated on 10% or 15% SDS polyacrylamide gels. Proteins were electrotransferred to nitrocellulose Hybond membranes (Amersham). Membranes were blocked for 1 h in 5% nonfat milk in Tris-buffered saline containing 0.05% Tween (TTBS) and incubated overnight at 4°C with either 2 µg/ml anti-ERß (H-150; Santa Cruz) or normal rabbit IgG (Santa Cruz) in TTBS plus 1% milk. The ERß H-150 antibody recognizes the first 150 amino acids of recombinant human ERß, which shares 81% identity with the porcine sequence. No blocking peptide is available for this antibody. Following washing in TTBS, membranes were incubated with either 1:1500 (for 50-µg blots) or 1:3000 (for 100-µg blots) double-staining grade horseradish peroxidase goat anti-rabbit IgG secondary antibody (65-6120; Zymed, South San Francisco, CA) in TTBS plus 5% milk for 1 h. Following extensive washing in TTBS, immunoreactive bands were detected by enhanced chemiluminescence (Amersham).
In vitro translated 35S-methionine-labeled proteins were generated using the TNT Rabbit Reticulocyte Lysate kit (Promega) and 1 µg of each plasmid, T7 polymerase, and a methionine- amino acid mixture as recommended by the manufacturer's protocol. Two or 5 µl of each reaction was separated by SDS-PAGE (10% or 15% gels) to mimic the gel conditions used for immunoblots. Gels were fixed in methanol/acetic acid/glycerol solution and then dried prior to autoradiography.
Transfections
ER-negative COS-1 cells were maintained in phenol red-free Dulbecco modified Eagle medium (DMEM), 1% nonessential amino acids (NEAA), 10% newborn calf serum (NCS), and 1% antibiotic/antimycotic solution. For transfection, COS-1 cells were plated in DMEM, 1% NEAA, 10% NCS at approximately 105 cells/35-mm well. Cells were transfected at 80–90% confluency. The same lot of serum was used for all experiments. Prior to transfection, cells were rinsed with DMEM only and subsequently transfected using 12 µl LipofectAMINE reagent and 2 µg plasmid DNA per well in DMEM plus 1% NEAA for 5 h. Transfected plasmids included a vitA2ERE-tk-Luc, which contains 1 copy of the vitellogenin A2 ERE (obtained from Dr. Daniel Noonan, University of Kentucky, Lexington, KY) and a minimal thymidine kinase (tk) promoter upstream of the firefly luciferase gene [28], pRLtkLuc, a renilla luciferase vector included as a transfection control, ER cDNAs in pcDNA3.1 expression vector, and pcDNA3.1 itself. Duplicate wells were transfected. Following transfection, medium was replaced with DMEM, 1% NEAA, and 10% NCS, and cells were incubated for 16 h, followed by an additional 24 h with 10 nM 17ß-estradiol (E2). Cells were lysed using 1x Passive Lysis buffer, and 20 µl of lysate was assayed using the Dual Luciferase Assay kit (Promega) and a Turner 20E luminometer according to the manufacturer's recommendations.
Ovarian granulosa cells were isolated from 1- to 5-mm antral follicles by needle aspiration and cultured as previously described [29]. Cells were plated at 8 x 106 cells/35-mm well in phenol red-free modified Eagle medium (MEM) containing antibiotics and antimycotics and 3% FCS. After 39–43 h, cells were rinsed with MEM only and transfected using 12 µl LipofectAMINE reagent and 2 µg plasmid DNA per well in MEM for 5 h. Following transfection, medium was replaced with serum-free MEM, and cells were treated with 0.05% ethanol vehicle or 10 nM E2 for 24 h. Cells were lysed and assayed as described for COS-1 cells.
Immunofluorescence
COS-1 cells were plated onto ultraviolet light-treated coverslips in 35-mm wells essentially as described above. Cells were transfected with 2 µg of each HA-tagged ER plasmid or pcDNA3.1 vector only by lipofection for 5 h as described above. Following transfection, medium was replaced with DMEM, 1% NEAA, and 10% NCS containing 10 nM E2. After 24 h, cells were fixed in 4% fresh paraformaldehyde for 10 min at 4°C. Cells were washed 3 times, for 10 min each time, with cold 1x PBS and stored at 4°C.
Cells were permeabilized with 0.1% Triton X-100 in PBS for 3 min and then blocked in PBS containing 1% BSA (fraction V) and 0.1% Tween-20 for 30 min. Cells were incubated with fluorescein-conjugated mouse monoclonal anti-HA (1666878; Roche, Indianapolis, IN) in blocking solution (5 µg/ml; Roche) for 1 h. Cells were washed 3 times in PBS, for 5 min each time. DNA was stained with 4',6'-diamidino-2-phenylindole (DAPI) for 10 min to identify nuclei. Coverslips were mounted on slides using ProLong Antifade solution (Molecular Probes, Eugene, OR).
Cells were visualized by epifluorescence at 40x magnification using a Zeiss microscope (Thornwood, NY). Images were captured and merged using a SPOT camera and software (Diagnostic Systems Inc., Sterling Heights, MI). Final images were converted to grayscale using Photoshop 5.5 (Adobe Systems Inc., San Jose, CA).
Electrophoretic Mobility Shift Assays
Two double-stranded oligonucleotides containing EREs were used for electrophoretic mobility shift assays (EMSA). A consensus sequence (cs-ERE; 5'-GGATCTAGGTCACTGTGACCCCGGATC-3') was purchased from Santa Cruz Biotech. A vitellogenin A2 ERE derived from the Xenopus gene was constructed by annealing the oligonucleotide 5'-CAAAGTCAGGTCACAGTGACCTGATC-3' and its reverse complement under high-salt conditions followed by purification on a nucleotide removal column (Qiagen). Both the cs-ERE and vitellogenin A2 ERE were labeled with 
-32P-ATP (3000 Ci/mmol) and T4 polynucleotide kinase. In vitro translated proteins were generated using the TNT Rabbit Reticulocyte Lysate kit (Promega) and 500 ng of each plasmid, T7 polymerase, complete amino acid mixture, and 4 µM E2. Five microliters of each in vitro translation reaction was incubated with or without (4 µg) polyclonal antibody or monoclonal antibody (0.8 µg) and labeled ERE (
100 000 cpm) in a buffer containing 5 mM Tris, pH 8, 80 mM KCl, 6% glycerol, 1 mM dithiothreitol, 0.05% NP-40, 50 ng denatured salmon testes DNA, and 2 µg poly dI-dC at room temperature for 20 min. Competition assays with cold oligos were incubated for 10 or 30 min on ice prior to addition of labeled oligo and then incubated at room temperature for 20 min. Samples were mixed with loading buffer (25 mM Tris, pH 7.5, 4% glycerol) and electrophoresed on native 5% polyacrylamide gels at 150 V for 3 h in 0.5x Tris-borate-EDTA (with 1 buffer exchange at 90 min). Gels were dried and exposed to BioMAX MR film at -70°C with intensifying screens. The antibodies used for supershift analysis were anti-ER (MC-20; Santa Cruz Biotech), anti-ERß (H-150; Santa Cruz), and mouse monoclonal anti-HA (1-583-816; Roche), which recognizes the sequence YPYDVPDYA.
Statistical Analyses
Raw data from RT-PCR and transfection studies were subjected to ANOVAs. When differences were significant, post hoc comparisons of means were made using a Fisher protected least significant difference (PLSD) test. RNase protection assay relative data (optical density units of ERß mRNA autoradiographic bands normalized for the corresponding 18S rRNA band) were analyzed by nonparametric Kruskal-Wallis test. A P value of <0.05 was considered significant. Comparisons were made with StatView software (SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5 transcript appears to result from the splicing of nucleotide 943 with nucleotide 1082 using BLAST alignment [30]. The ERß5 transcript has also been detected in bovine, sheep, and human tissues [10, 17–19]. The predicted translations of these transcripts are shown in Figure 1. The full-length porcine ERß has an open reading frame of 526 amino acids and notably lacks 1 amino acid between positions 413 and 414 when compared with several other mammalian species (rodents, primates, cows, and sheep). The exon 5 deletion variant introduces a novel stop codon shortly downstream of the splice site, resulting in a carboxyl terminus of 5 unique amino acids. The exon 5-deleted variant includes the DNA-binding domain (DBD) and the major nuclear localization signals but lacks the majority of the ligand-binding domain (LBD). MaxHom alignment revealed amino acid identities of 92% with cow and sheep, 87% with human, rat, and mice ERß coding sequences, and 48.8% overall identity with porcine ER [31].
|
RNase protection assays employing a riboprobe for an amino-terminal segment of ERß (exons 1 and 2) were used to evaluate the abundance of ERß in several ovarian structures of known developmental age from hormonally primed gilts (Fig. 2A). Two predominant protected bands were observed corresponding to the full-length cDNA protected fragment (258 bp) and a second band approximately 50 bp smaller (Fig. 2B). These bands were consistently proportionate and were quantified together. The smaller band may represent a splice variant in exon 1 or 2 or a polymorphism that we did not detect during cloning. Ovarian preovulatory follicles and CL isolated from eCG ± hCG-treated gilts revealed the presence of ERß mRNA in all structures. ERß mRNA did not vary significantly among tissue types; however, regressing CL (Day 20) tended to have the lowest expression, as determined by Kruskal-Wallis ranking.
|
A second riboprobe designed to span exon 4–6 and detect both the intact sequence and the exon 5 deletion was unable to detect any ERß mRNA, suggesting low abundance of each transcript or secondary structure complications (data not shown). As an alternative, semiquantitative RT-PCR was performed on select porcine ovarian tissues from unprimed animals to detect the presence of both transcripts. Figure 3A shows representative RT-PCR autoradiographic bands for ERß, ERß5, and internal control ribosomal S16 protein cDNAs. Figure 3B shows the analysis of these transcripts for 3 sets of tissues. Neither ERß nor ERß5 transcripts differed significantly among granulosa cells (from 1- to 5-mm antral follicles) from unprimed prepubertal pigs, pools of follicles of different sizes (including theca, basement membrane, and granulosa cells), or midluteal phase CL from naturally cycling pigs. The ratio of ERß5 to ERß was similar among the tissue types and ranged from a low mean of 50% in granulosa cells to a high mean of 57% in small follicles.
|
To determine the approximate sizes of ERß and ERß5 proteins, ER plasmids were in vitro transcribed and translated with 35S-methionine. Figure 4A shows a representative autoradiogram of these products separated with 15% SDS-PAGE. Proteins of the approximate predicted molecular mass of 58.8 and 35.5 kDa for porcine ERß and ERß5, respectively, and 66.4 kDa for ER were detected [32]. Both in vitro translated ERß and ERß5 products appeared as doublets. The lower molecular mass band in each lane may be due to the use of a second in-frame AUG located at amino acid 46. Figure 4B shows an immunoblot of cell lysates separated with 15% SDS-PAGE under conditions identical to those illustrated in Figure 4A. Immunoreactive full-length ERß protein was detected in the lysates of all unprimed ovarian tissues examined. Smaller proteins, some which corresponded to the sizes shown with in vitro translated products, were present predominantly in the nuclear extracts and whole-cell extracts from fresh granulosa cells of 1- to 5-mm follicles from unprimed animals. The smaller proteins appeared to be less abundant or not detected in follicles of increasing size or CL. Figure 4C shows a control blot incubated with normal rabbit IgG in place of the primary antibody.
|
The human equivalent of ERß5 cDNA isolated from testis has been reported to be a negative regulator of wild-type ER and ERß when transiently transfected into ER-negative COS-7 cells [18]. We thus tested the ability of our porcine constructs to transactivate the vitA2ERE-tk-luciferase reporter construct in the well-studied ER-negative cell line, COS-1. We determined the maximally effective amounts of plasmid for transactivation to be 50 ng for ER (3.47 ± 0.74-fold, versus vector control) and 500 ng plasmid for ERß (2.60 ± 0.49-fold). Thus, ER was approximately 10-fold more potent at activating the ERE reporter construct in COS-1 cells. ER did self-squelch at 1 µg plasmid, but ERß did not (data not shown). We do not know if these differences in transactivating abilities are due to possible differences in transcription or translation efficiencies for each plasmid cDNA. ERß5 had no significant activity in COS-1 cells at any concentration tested when compared with vector alone. We thus tested the effect of ERß5 to modify ERß and ER transactivation. At 10-fold (1 µg) plasmid excess, ERß5 significantly inhibited ERß transactivation (100 ng; Fig. 5A). Up to 100-fold excess ERß5 plasmid did not significantly attenuate ER (10 ng) plasmid-mediated transactivation (Fig. 5B).
|
Because COS-1 cells are ER negative, we examined expression of ERß and ERß5 in granulosa cells that naturally express the transcripts and immunoreactive proteins. This population of granulosa cells has little full-length ER (unpublished data) [21]. Granulosa cells transfected with vector alone were used as a reference for endogenous ER activity, and this activity was normalized to 1.0 for each experiment. Transfected wild-type ERß was a potent activator of the ERE reporter construct in serum-free primary cultures of granulosa cells (P < 0.05). The amplitude of the E2 response did not change with overexpression of full-length ERß. ERß5 did not inhibit endogenous activation of the ERE reporter construct but appeared to possess modest transactivation ability in these ovarian cells (3- to 4-fold versus vector). The amplitude of the E2 response was not significantly affected by overexpression of the exon 5 variant.
ERß5 lacks most of the LBD but has the nuclear localization signals. To confirm that ERß5 can localize to the nucleus, direct immunofluorescence was used with COS-1 cells transfected with HA-tagged ER and ERß plasmids (Fig. 6, A–C). All ERs localized predominantly to the nucleus (DAPI positive) in transfected COS-1 cells. Vector only showed no specific fluorescein staining (Fig. 6D).
|
ERß5 possesses the DBD, so we used EMSA to test the capacity of ERß and ERß5 to bind 2 DNA oligonucleotides containing EREs (Fig. 7, A and B). ER plasmids or vector only were in vitro transcribed and translated and used in EMSA. Unprogrammed lysate (no plasmid) was also used as a control. ER was used as a positive control and exhibited strong binding. ERß bound both the cs-ERE and the vitA2 ERE oligonucleotides and migrated as multiple bands, most likely because of the doublet generated during in vitro translation (Fig. 4A). ERß5 protein bound in the DNA:protein complex generated by the lysate but could be shifted with both the ERß-specific polyclonal antibody or the anti-HA antibody in the case of the HA-tagged protein. The monoclonal anti-HA antibody tended to produce a weaker supershift, probably because it only recognizes the single epitope. All supershifted bands were eliminated by preincubation with a 100-fold excess of cold ERE oligonucleotide, implying specific binding (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5 were detected predominantly in granulosa cells from 1- to 5-mm follicles from unprimed animals. In addition to ERß5, several other ERß splice variants have been described. Two variants of exon 3 deletions, ERß3, lack part of the DBD but retain the LBD [33]. The ERß3 variants display altered subnuclear localization and fail to transactivate at a consensus ERE but can activate at AP-1 sites [33]. Rodent ERß2 contains an in-frame insertion between exons 5 and 6 that adds 18 amino acids to the LBD, has reduced ligand-stimulated activity, and can repress wild-type ER transactivation [34, 35]. ERßcx found in humans has a different exon at the C-terminus and acts to repress transactivation by both ER and ERß [36]. The ERß5 transcript has recently been detected in bovine and sheep ovaries and human testis and mammary gland [10, 17–19]. However, it was not determined whether a protein of predicted size was present. In 2 of these studies, function of ERß5 in transfected clonal cells was evaluated. Inoue et al. [18] isolated an ERß5 variant from human testis cDNA and showed that human ERß5 cDNA transfected into COS-7 cells exerted a repressor activity at 10-fold excess plasmid when cotransfected with either wild-type ER or ERß (with more potent repression of ERß). Similarly, our porcine studies showed repression of ERß transactivation with 10-fold excess ERß5 plasmid (1 µg). The physiological relevance of such inhibition at 10-fold plasmid excess in COS cells is questionable given that our mRNA studies revealed that ERß5 expression is only half that of the full-length ERß5 transcript in ovarian tissue.
In contrast, COS-1 transfections did not show repression of ER transactivation, even with up to 100-fold excess ERß5 plasmid. This inconsistency may be due to species differences, specific experimental conditions, or differences in vector construction. It may also be due to the fact that human ER in COS-7 cells was less active than ERß at the same concentration, whereas our porcine ER was about 10 times more potent than ERß in COS-1 cells. Walter et al. [17] identified a bovine exon 5-deleted variant in ovary that had no activity alone or when cotransfected with ERß in HEK293 cells. This latter difference may be due to the host cell used for transfection or to the fact that the bovine cDNAs were artifactually truncated, resulting in a 54-amino acid deletion in the N-terminus of both the full-length and exon 5-deleted isoforms. Studies with the human full-length and N-terminal truncated ERß isoforms have shown that the full-length transcript has a higher transactivating capacity, implying a functional component to the missing 54 amino acids [37].
In the current study, ERß5 transfected into primary granulosa cells did not inhibit basal ERE reporter activity by endogenous ERs. Rather, ERß5 exhibited a 3- to 4-fold increase above basal levels, although this increase was not significant when compared with the robust stimulation by transfected wild-type ERß, which yielded 8- to 10-fold (-E2) and 14- to 15-fold (+E2) increases. Taken together, these studies imply that the cellular context for ERß isoform expression is important. One explanation for the differences observed between COS-1 cells and primary granulosa cells may be the abundance or presence of coactivator molecules. A recent analysis of coactivator transcripts in the porcine ovary has shown abundant expression of SRC-1 and the presence of RAC3 [38]. Alternatively, the transactivation ability observed for ERß5 may be an artifact of overexpressing the variant at levels not seen endogenously in granulosa cells. Our endogenous ERß mRNA and protein data revealed stronger expression of the wild-type ERß than of the variant.
The predicted ERß5 protein lacks most of the LBD, including -helixes 4–12 associated with dimer formation, coactivator recruitment, AF2 transactivation, and hormone binding [39]. The hormone binding capacity of this variant has not been tested. The variant does possess the AF1-containing A/B domain, the DNA binding domain, and the nuclear localization signals. We have provided evidence that ERß5 localizes to the nucleus and binds DNA, as predicted from its protein structure. The human ERß5 also localizes to the nucleus [18]. Although we cannot rule out the possibility that our transfection procedure may influence the subcellular distribution of plasmid products, in another study proteins expressed from the pcDNA3 plasmid distributed randomly throughout the cell in the absence of a nuclear localization signal [40].
Whether ERß5 can dimerize with wild-type estrogen receptors is unknown. In studies with ER truncated in the LBD, the receptor was able to bind DNA as dimers, although with lower affinity than the ligand-bound intact ER [41]. Because the bulk of the LBD needed for dimerization of ERs is deleted, one would predict that dimerization with wild-type ERs would occur via the DBD. In addition, some coactivators such as p300 can bind the A/B region of human ER and ERß [42]. Potentially ERß5 could recruit or squelch cofactors with its intact A/B amino terminus. These speculations warrant further study.
The mRNAs for ERß and ERß5 were detected by semiquantitative RT-PCR in all the ovarian structures isolated from unprimed animals, with ERß being predominant. The full-length ERß protein was detected in the same tissues and was most abundant in the granulosa cells and midluteal phase CL. The putative immunoreactive protein bands for ERß5 observed predominantly in granulosa cells were less abundant than full-length ERß. These smaller bands were also less abundant or absent in large follicles or CL even with up to 100 µg of protein for each sample (data not shown). This finding suggests a developmental uncoupling of transcription and translation for the ERß5 variant or perhaps a high turnover rate for the variant protein in some cells.
Unlike studies in the rat, we did not observe a downregulation of ERß mRNA after hCG treatment in either preovulatory follicles (Day 1) or new CL (Day 2) [43, 44]. We may have either missed a window of mRNA decrease in the pig given our sampling interval or we may have needed to evaluate granulosa cells from staged follicles rather than whole follicles. Perhaps a decrement does not occur after the gonadotropin surge in this species. ERß mRNA and protein expression do recover in rat CL of pregnancy, suggesting a luteal function [45]. In several nonrodent species, nonpregnant animals exhibit abundant ERß in the CL. ERß mRNA and protein are strongly expressed throughout the luteal phase, including late CL, in rhesus monkeys [7]. RT-PCR revealed human ERß mRNA in all stages of CL, with a decrease in levels during the late secretory phase [46]. Sheep also expressed ERß and ERß5 transcripts in all stages of CL examined by RT-PCR, with a decrease in ERß5 in CL during pregnancy [10].
The pig was one of the first models for the study of ovarian cell function [4]. Although estrogen-binding capacity in follicles has been demonstrated and estrogen has a vital function in luteal life span in pigs [13, 16, 47], little is known about the ER expression in porcine ovary. Slomczynska et al. [20, 21], using heterologous probes, localized ERß mRNA and protein in porcine small, medium, and large follicles and CL, consistent with our findings. Our studies show that wild-type ERß is strongly expressed in multiple follicular structures and that ERß5 protein found in granulosa cells may have biological activity. Further studies are needed to clarify the role of ERs in the follicular development and luteal function of pigs.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported in part by NIH grant HD-38945 (to H.A.L.). 
2 Correspondence: Holly A. LaVoie, Department of Cell Biology and Neuroscience, Bldg. 1, School of Medicine, University of South Carolina, Columbia, SC 29209. FAX: 803 733 3212; hlavoie{at}med.sc.edu
Accepted: October 10, 2001.
Received: August 23, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Malekinejad, E. J Schoevers, I. J.J.M Daemen, C. Zijlstra, B. Colenbrander, J. Fink-Gremmels, and B. A.J Roelen Exposure of Oocytes to the Fusarium Toxins Zearalenone and Deoxynivalenol Causes Aneuploidy and Abnormal Embryo Development in Pigs Biol Reprod, November 1, 2007; 77(5): 840 - 847. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Benoit, G. L. McCoy, and C. A. Blake Localization of Fertility Factor SP22 to Specific Cell Types Within the Anterior Pituitary Gland Experimental Biology and Medicine, November 1, 2005; 230(10): 721 - 730. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Gillio-Meina, Y. Y. Hui, and H. A. LaVoie Expression of CCAAT/Enhancer Binding Proteins Alpha and Beta in the Porcine Ovary and Regulation in Primary Cultures of Granulosa Cells Biol Reprod, May 1, 2005; 72(5): 1194 - 1204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Rusovici, Y. Y. Hui, and H. A. LaVoie Epidermal Growth Factor-Mediated Inhibition of Follicle-Stimulating Hormone-Stimulated StAR Gene Expression in Porcine Granulosa Cells Is Associated with Reduced Histone H3 Acetylation Biol Reprod, April 1, 2005; 72(4): 862 - 871. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. A. LaVoie, D. Singh, and Y. Y. Hui Concerted Regulation of the Porcine Steroidogenic Acute Regulatory Protein Gene Promoter Activity by Follicle-Stimulating Hormone and Insulin-Like Growth Factor I in Granulosa Cells Involves GATA-4 and CCAAT/Enhancer Binding Protein {beta} Endocrinology, July 1, 2004; 145(7): 3122 - 3134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Gillio-Meina, Y. Y. Hui, and H. A. LaVoie GATA-4 and GATA-6 Transcription Factors: Expression, Immunohistochemical Localization, and Possible Function in the Porcine Ovary Biol Reprod, February 1, 2003; 68(2): 412 - 422. [Abstract] [Full Text] [PDF]
|
|
| ||
) and immunoreactive proteins corresponding in size to the in vitro translated ERß
