Progesterone-Receptor Antagonists and Statins Decrease De Novo

Download PDF
1 downloads 0 Views 177KB Size Report
Comment
Sep 22, 2004 - DOI 10.1095/biolreprod.104.033878. Progesterone-Receptor Antagonists and Statins Decrease De Novo Cholesterol. Synthesis and Increase ...
BIOLOGY OF REPRODUCTION 72, 538–545 (2005) Published online before print 22 September 2004. DOI 10.1095/biolreprod.104.033878

Progesterone-Receptor Antagonists and Statins Decrease De Novo Cholesterol Synthesis and Increase Apoptosis in Rat and Human Periovulatory Granulosa Cells In Vitro1 Emilia Rung,3 P. Anders Friberg,3 Ruijin Shao,3 D.G. Joakim Larsson,3 Eva Ch. Nielsen,3 Per-Arne Svensson,4 Bjo¨rn Carlsson,4 Lena M.S. Carlsson,4 and Ha˚kan Billig2,3 Department of Physiology and Pharmacology3 and Research Centre for Endocrinology and Metabolism, Department of Internal Medicine,4 Go¨teborg University, SE-40530 Go¨teborg, Sweden ABSTRACT

stage-specific [3, 4], and atresia occurs with high incidence at all stages of follicular development [1]. The periovulatory interval is an exception to this, because the number of mature preovulatory follicles just before ovulation is approximately equal to the number of corpora lutea. Stimulation of the LH receptor and final recruitment of follicles have been shown to reduce the degree of granulosa cell apoptosis in rat [5], bovine [6], and chicken [7]. Thus, at the transition of preovulatory follicles into corpora lutea (the periovulatory interval), the incidence of atresia is low. Isolated LH receptor-stimulated granulosa/luteal cells are here defined as periovulatory granulosa cells. During in vitro incubation the granulosa cells undergo processes that mimic several of the aspects of luteinization. However, the in vivo process also includes, for instance, extensive vascularization and tissue remodeling. The apoptosis-resistant periovulatory interval in follicular development is characterized by LH-induced progesterone synthesis [8]. In addition to the effects of progesterone on traditional target organs, such as the uterus and mammary glands, autocrine/paracrine effects also occur in the ovary. Interestingly, induction of progesterone synthesis coincides with expression of the nuclear progesterone receptor (PR) in the follicular granulosa cells. In rat, the PR is transiently expressed during the periovulatory interval [9, 10]. However, in humans, it is prolonged, and PRs remain present in the corpus luteum [11, 12]. Our group has reported that the PR expression of both isoforms of PR (isoforms A and B) is transient in mouse granulosa cells [13], similar to the expression in rat. However, Gava et al. [14] have suggested that this is true for PR-A, whereas PR-B is expressed in mouse granulosa cells in follicles at earlier developmental stages as well as in corpora lutea. Evidence for direct effects of progesterone in the ovary includes the fact that mice lacking both PR isoforms are anovulatory [15, 16]. Mice lacking only the A isoform have decreased ovulation after stimulation [17], whereas mice lacking only the B isoform are fertile, with normal litter sizes [18]. The PR-mediated effects in periovulatory granulosa cells include induction of PACAP (pituitary adenylate cyclase-activating polypeptide) [19, 20] and its receptor PAC1 [21] and the proteases ADAMTS-1 and cathepsin L [22]. Studies by our group have shown that rat granulosa cells have decreased sensitivity to apoptosis after treatment with hCG, mimicking the LH surge. This decreased sensitivity is mediated, at least in part, by progesterone, because addition of the PR-antagonists Org 31710 or RU 486 dose-dependently diminish the effect [5]. Indeed, human granulosa cells, obtained from women undergoing treatment before in vitro fertilization (IVF) are affected by PR antagonists in

Progesterone-receptor (PR) stimulation promotes survival in rat and human periovulatory granulosa cells. To investigate the mechanisms involved, periovulatory rat granulosa cells were incubated in vitro with or without the PR-antagonist Org 31710. Org 31710 caused the expected increase in apoptosis, and expression profiling using cDNA microarray analysis revealed regulation of several groups of genes with functional and/or metabolic connections. This regulation included decreased expression of genes involved in follicular rupture, increased stress responses, decreased angiogenesis, and decreased cholesterol synthesis. A decreased cholesterol synthesis was verified in experiments with both rat and human periovulatory granulosa cells treated with the PR-antagonists Org 31710 or RU 486 by measuring incorporation of [14C]acetate into cholesterol, cholesterol ester, and progesterone. Correspondingly, specific inhibition of cholesterol synthesis in periovulatory rat granulosa cells using 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors (lovastatin, mevastatin, or simvastatin) increased apoptosis, measured as DNA fragmentation and caspase-3/7 activity. The increase in apoptosis caused by simvastatin was reversed by addition of the cholesterol synthesis-intermediary mevalonic acid. These results show that PR antagonists reduce cholesterol synthesis in periovulatory granulosa cells and that cholesterol synthesis is important for granulosa cell survival.

apoptosis, follicular development, mechanisms of hormone action, ovary, progesterone receptor

INTRODUCTION

Less than 0.1% of the follicles endowed in the human ovary at birth complete the path to ovulation. Instead, atresia, which at the cellular level is accomplished by apoptosis, is the most common fate of the developing follicles [1]. Granulosa cell apoptosis is an active cellular event that is dependent on transcription as well as translation [2]. Hormone and growth factor regulation of follicular atresia is Supported by grants 10380, 11295, 13141, and 13550 from the Swedish MRC and by grants from the Assar Gabrielsson Foundation, the Hjalmar Svensson Foundation, the Ollie and Elof Ericsson Foundation, the Lundberg Foundation, the Eva and Oscar Ahre´n Research Foundation, Stockholm and Swegene, Sweden. 2 Correspondence: Ha˚kan Billig, Department of Physiology, Go¨teborg University, P.O. Box 434, SE-40530 Go¨teborg, Sweden. FAX: 46 0 31 7733531; e-mail: [email protected] 1

Received: 30 June 2004. First decision: 7 July 2004. Accepted: 14 September 2004. 䊚 2005 by the Society for the Study of Reproduction, Inc. ISSN: 0006-3363. http://www.biolreprod.org

538

GRANULOSA CELL APOPTOSIS AND CHOLESTEROL SYNTHESIS

the same way [23, 24]. We also have shown that RU 486 increases apoptosis in periovulatory granulosa cells in vivo when administered to mice either 4 h before or together with hCG [13]. The progesterone-PR complex functions as a transcription factor, making it possible to study gene expression to learn more about PR-mediated effects in the periovulatory follicles. Because inhibition of gene transcription can inhibit granulosa cell apoptosis [2], this may include, in addition to other effects, transcriptional changes that regulate apoptosis. In the present study, we isolated rat granulosa cells from periovulatory follicles and incubated them in vitro for 24 h in the presence or absence of the specific PRantagonist Org 31710. Effects on gene transcription were then evaluated using the Affymetrix microarray technique. The expression of several genes with functional and/or metabolic connections was affected, and we focused on the down-regulation of cholesterol synthesis. MATERIALS AND METHODS

Granulosa Cell Isolation Follicle maturation was induced in 26-day-old female rats (SpragueDawley, B&K, Sweden) by s.c. injection of 10 IU of eCG (Sigma, St. Louis, MO). To mimic the endogenous LH surge, the animals were treated with an i.p. injection of 50 IU of hCG (N.V. Organon, Oss, Holland) 48 h after the eCG treatment. Ovaries were isolated after an additional 12 h, and granulosa cells were collected from mature preovulatory follicles by puncture. Human periovulatory granulosa cells were obtained in conjunction with ovum retrieval for IVF by follicle aspiration via transvaginal ultrasound-guided puncture. Isolated granulosa cells were purified further by isotonic Percoll centrifugation [23]. In each of the three experiments using human granulosa cells, the cells from three to four women were pooled before incubation. The local ethics committees approved the animal and human experiments, and all patients gave their consent after receiving both written and oral information.

Granulosa Cell Incubation Isolated granulosa cells were pooled and incubated (320 000 per 0.5 ml for rat granulosa cells and 64 000 per 0.5 ml for human granulosa cells) in Eagle minimum essential medium with glutamax-I, Earle salts, and Hepes (Life Technologies, Sweden) supplemented with penicillin (100 U/ml), streptomycin (100 ␮g/ml), and 0.1% BSA (fraction V; Sigma-Aldrich Chemie, GmbH, Germany). The incubations were carried out in tubes (12 ⫻ 75 mm; Falcon; Becton Dickinson, NJ) with or without addition of the PR-antagonists Org 31710 (10 ␮M; N.V. Organon) or RU 486 (10 ␮M; Exelgyn, Paris, France); the cholesterol synthesis-inhibitors mevastatin (10 ␮M; Sigma, Sweden), lovastatin (10 ␮M; Merck, NJ), or simvastatin (10 ␮M; Merck, NJ); mevalonic acid lactone (Aldrich, Germany) converted to mevalonic acid (1.6 mM) [25]; and the steroid synthesis-inhibitor cyanoketone (0.01–10 ␮M; Sterling Drug Inc., Rensselaer, NY) and/or [14C]acetate (10 ␮Ci; [1-14C]acetic acid, sodium salt, 50–62 mCi/mmol; Amersham-Pharmacia, Sweden). The RU 486, lovastatin, simvastatin, mevastatin, mevalonic acid, and cyanoketone were dissolved in ethanol, whereas Org 31710 was dissolved in dimethyl sulfoxide. The final concentration of solvent in the incubation medium was 0.2% or less in all experiments and did not affect any of the end points used. For experiments including cyanoketone, the incubation medium was replaced after 2 h of incubations to decrease further the presence of endogenously synthesized progesterone. All incubations were performed at 37⬚C in 5% CO2 and 95% humidified air for 24 h.

RNA Isolation and Hybridization to Microarray Chips Granulosa cells were isolated from four distinct experiments, each based on more than 20 rats, and incubated for 24 h with or without addition of 10 ␮M Org 31710. In previous experiments [5], 10 ␮M Org 31710 was the lowest concentration that significantly increased the degree of apoptosis. In each experiment, one batch of RNA was isolated from control cells and one batch from Org 31710-treated cells using RNeasy (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.

539

The average absorbance ratio (A260/A280) was 1.854 ⫾ 0.003. An aliquot of all batches of RNA was subjected to electrophoretic separation, showing no signs of degradation on visual inspection. Before hybridization to Affymetrix microarray chips, RNA from control and Org 31710-treated cells was pooled from the four experiments, and altogether, 8 ␮g of RNA from each group were used for array hybridization. Preparation of labeled target was performed according to the Affymetrix Gene Chip Expression Analysis manual [26]. Subsequently, two Rat Genome U34A arrays, representing approximately 8800 genes, were hybridized for each sample. Washing and staining was done using a fluidics station and a confocal scanner essentially according to the manufacturer’s instructions (Affymetrix). In brief, first staining was done (protocol EukGE-WS1), followed by signal amplification using biotinylated antistreptavidin antibody (final steps of protocol EukGE-WS2 starting with second stain) and scanning of the array.

Analysis of Gene Chip Data Gene expression in periovulatory rat granulosa cells after incubation for 24 h in the absence or presence of Org 31710 was analyzed on duplicate DNA microarray chips. Scanned output files were inspected visually for hybridization artifacts, and quantitative analysis of hybridization patterns and intensities was performed automatically by the Affymetrix GeneChip Analysis Suite 3.3 software (empirical algorithm) [26]. To allow comparison of gene expression, the DNA microarrays were scaled to an average intensity of 500. Comparative analysis was performed with the control group as baseline. Comparisons were made between the results from the duplicate control and the Org 31710 DNA microarray chips, generating a total of four comparisons. Genes that were considered by the Affymetrix software’s Diff Call parameter to be decreased (D) or marginally decreased (MD) in at least three of four comparisons were considered to be down-regulated. Likewise, genes with Diff Call increased (I) or marginally increased (MI) in at least three of four comparisons were considered to be up-regulated. Genes meeting these criteria were selected for further investigation. In addition, further analysis of the microarray results, with a focus on genes implicated in cholesterol synthesis, revealed other genes that were transcriptionally decreased after incubation in the presence of Org 31710 but that did not meet our primary, more stringent criteria for change in transcription. These genes were considered by the software to be either upor down-regulated in one or two of the four comparisons. They are reported here to be potentially regulated, because they are part of a bigger context (i.e., a metabolic pathway). An average fold-change of the four comparisons also was calculated for each gene. The number of comparisons in which the gene was either up- or down-regulated reflects the level of fidelity of the mRNA regulation, and the fold-change reflects the magnitude of mRNA regulation. The genes are referred to by their names and GenBank accessions numbers (http://www.ncbi.nlm.nih.gov). The expressed sequence tag (EST) sequences were subjected to standard nucleotide-nucleotide BLAST (blastn) for identification. To estimate the methodological error, comparisons between the duplicate chips (control vs. control and Org-treated vs. Org-treated) were made. Transcripts that in this comparison were determined by the software to be present in the sample and had a Diff Call of I, MI, MD, or D were considered to be false positives. The number of false positives in these single comparisons was 53 out of 8800 genes for the control chips and 28 out of 8800 genes for the Org 31710 chips.

Apoptosis Measurements Fragmentation of DNA was determined as described previously [27] with minor modifications [5]. In brief, the cells were pelleted and mildly lysed. The samples were centrifuged for 20 min at 15 700 ⫻ g to separate fragmented DNA (supernatant) from intact DNA (pellet), and the pellets were treated overnight with proteinase K. The DNA content in the supernatants and in the pellets was measured with a fluorescence spectrophotometer after addition of Hoechst dye (H33258). The apoptotic index reflects the ratio between low-molecular-weight DNA (isolated in the supernatant) and total DNA content (supernatant and pellet DNA). The ratio ranged from 6.1% to 10.3% in rat control cells and from 2.7% to 12.3% in human control cells. Caspase activity was measured using the Apo-ONE Homogenous Caspase-3/7 Assay or Caspase-Glo 3/7 Assay (Promega, Madison, WI) according to the manufacturer’s instructions. The Apo-ONE assay is based on cleavage of the profluorescent caspase substrate rhodamine 110, bis(N-CBZ-L-aspartyl-L-glutamyl-L-valyl-L-aspartic acid amide) (Z-DEVDR110). The amount of fluorescent product generated after caspase removal

540

RUNG ET AL.

of the DEVD-peptides is proportional to the amount of caspase-3/7 cleavage activity in the sample. Thirty-thousand granulosa cells per well were incubated in white 96-well plates (Nunc, Roskilde, Denmark) for 24 h using the medium and drugs described above. Fluorescence was detected 90 min after addition of the homogenous caspase-3/7 reagent at an excitation wavelength of 485 nm and an emission wavelength of 520 nm using a Fluostar Galaxy platereader (BMG Labtechnologies GmbH, Offenburg, Germany). Caspase-Glo is a very similar assay that detects cleavage of a proluminescent DEVD-substrate. Twenty-thousand granulosa cells per well were used, and luminescence was detected after 30 min. The value of the control group was set to 100% for all methods.

Analysis of De Novo Cholesterol Synthesis To verify the functional importance of the down-regulation of genes involved in cholesterol synthesis, the rate of de novo cholesterol synthesis in periovulatory granulosa cells was measured as the incorporation of 14Clabeled acetate into cholesterol, cholesterol ester, and progesterone. The [14C]acetate (10 ␮Ci) was added during the last 5 h of the 24-h incubation. After the incubation was completed, cells were pelleted by centrifugation (200 ⫻ g, 5 min), and extraction was performed as described previously [28] with modifications described elsewhere [29]. Cholesterol, cholesterol ester (cholesterol oleate), and progesterone were added as carriers. After drying under N2 gas, the samples were dissolved in chloroform. Cholesterol, cholesterol ester, and progesterone were separated by thin-layer chromatography on Silica Gel 60 plates (Merck, Darmstadt, Germany) with chloroform:methanol:water (65:24:4, v/v/v) followed by petroleum ether: diethylether:acetic acid (60:40:1, v/v/v). Spots were visualized by iodine staining and scraped off into scintillation vials. One milliliter of cyclohexane was added, and the radioactivity was determined in the presence of 10 ml of Ready-Safe scintillation mixture (Beckman, Sweden).

Progesterone Assay Progesterone in spent medium was analyzed by DELFIA time-resolved immunofluorometric assay [30] (Wallac Oy, Turku, Finland) according to the manufacturer’s instructions.

FIG. 1. Simplified overview of the cholesterol synthesis pathway. All genes directly involved in cholesterol synthesis that were present on the array are indicated together with the effect of Org 31710 (10 ␮M) on transcription levels. Each gene is identified by GenBank accession number and name, followed by the average fold-change. The numbers within parentheses show the range of fold-change in the four chip comparisons. Genes marked with an asterisk were down-regulated, and genes marked with a dagger did not meet our primary, more stringent criteria for regulation (but are included because they are part of the metabolic pathway). Genes in italics were present on the array but were not transcriptionally regulated. In addition to these genes, sterol carrier protein-2 was affected (M62763, sterol carrier protein 2: ⫺1.9 (⫺1.6 to ⫺2.2)*). N.C., No change.

Statistical Analysis Experiments were performed at least three times with at least three samples in each treatment group unless otherwise stated. Statistical analyses were performed by one-way ANOVA followed by the Dunnett posthoc test to compare multiple means to a single group or the Bonferroni post-hoc test (for the data shown in Fig. 4) to compare all groups to each other. Data are presented as the mean ⫾ SEM. A value of P ⬍ 0.05 was considered to be significant.

RESULTS

Expression Profile in Response to the PR-Antagonist Org 31710

Addition of Org 31710 (10 ␮M) increased the degree of DNA fragmentation to 133.4% ⫾ 3.6% compared to the control value of 100.0% ⫾ 2.1% (P ⬍ 0.01, n ⫽ 12 in each group), which is consistent with previously reported results [5]. Then, RNA isolated after 24 h of incubation was analyzed further with microarray. Of the 8800 genes on the array chips, approximately 3400 were considered by the algorithm used to be present in both control samples. Based on the criteria described in Material and Methods, 51 genes were down-regulated, and 49 genes were up-regulated. Groups of regulated genes were identified as belonging to metabolic or functional pathways. This includes down-regulation of cholesterol synthesis (see below), down-regulation of angiogenesis (e.g., ELK ligand LERK-2 [Eplg2]/ ephrin-B1 [U07560] by 2.9-fold), down-regulation of proteases possibly involved in follicular rupture (e.g., tissuetype plasminogen activator mRNA [M23697] by 2.1-fold and cathepsin H [M38135] by 2.9-fold), and increased stress responses (e.g., thioredoxin reductase mRNA [U63923] by 2.8-fold, growth arrest and DNA damage-in-

ducible protein GADD153 [U30186] by 2.3-fold, and liver glutathione S-transferase Ya subunit [K00136] by 3.7-fold). Further studies were focused on PR-mediated effects on cholesterol synthesis. Four of the down-regulated genes were identified as being involved in cholesterol synthesis. These genes were the mitochondrial and cytosolic 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) synthase, mevalonate kinase, and one EST sequence (GenBank accession no. H33491), which was identified as Rattus norvegicus sterol ⌬8-isomerase mRNA (GenBank accession no. AF071501) [31] after a BLAST homology search. The EST sequence is 98% identical (386/393 base pairs) to the rat sterol ⌬8-isomerase. In addition, one gene involved in cholesterol transport, sterol carrier protein-2, was among the down-regulated genes. Further analysis of the microarray results, with a focus on genes implicated in cholesterol synthesis, revealed other genes that were transcriptionally decreased after incubation in the presence of Org 31710 but that did not meet our primary, more stringent criteria for change in transcription. None of the genes involved in cholesterol synthesis showed a tendency to increased transcription. Figure 1 shows a simplified overview of the cholesterol synthesis pathway, indicating the function of affected genes. Effect of Org 31710 and RU 486 on De Novo Cholesterol Synthesis and Apoptosis in Rat Periovulatory Granulosa Cells

The PR-antagonist Org 31710 and the less specific PRantagonist RU 486 exhibited similar effects, and both decreased the rate of cholesterol synthesis (Fig. 2A), as im-

GRANULOSA CELL APOPTOSIS AND CHOLESTEROL SYNTHESIS

541

plied by the transcriptional down-regulation of cholesterol synthesis genes observed with the DNA microarray. Incorporation of [14C]acetate for 5 h into cholesterol was significantly suppressed to approximately 50% of control values in the granulosa cells incubated with Org 31710 or RU 486 (P ⬍ 0.01). The incorporation into cholesterol ester was reduced to between approximately 20% and 50% of the control value (P ⬍ 0.01). A significant decrease in incorporation after treatment with Org 31710 or RU 486 also was observed for progesterone (P ⬍ 0.05). When incorporation was performed in the presence of 25-fold excess of unlabeled acetate, the effects of Org 31710 remained the same (incorporation into cholesterol and cholesterol ester decreased to approximately 50% and 30%, respectively, of the control value), indicating that isotope dilution caused by different endogenous acetate concentrations could not explain our results. The degree of apoptosis was measured as caspase-3/7 activity using the Apo-ONE Homogenous Caspase-3/7 Assay. Org 31710 (10 ␮M) increased the substrate cleavage to 274% ⫾ 23%, and RU 486 (10 ␮M) increased the substrate cleavage to 323% ⫾ 10%, compared to the control value of 100% ⫾ 14% (P ⬍ 0.01, n ⫽ 4 in each group). The DNA fragmentation increased to 123.0% ⫾ 3.0% after treatment with Org 31710 and to 139.3% ⫾ 7.2% after treatment with RU 486, compared to the control value of 100% ⫾ 2.0% (P ⬍ 0.01, n ⫽ 11–24), which is in accordance with previous studies [5]. Effect of Org 31710 and RU 486 on De Novo Cholesterol Synthesis and Apoptosis in Human Periovulatory Granulosa Cells

Similar results were obtained using human granulosa cells (Fig. 2B) retrieved in conjunction with oocyte aspiration for IVF. Treatment with Org 31710 decreased the incorporation of [14C]acetate during 5 h to 39% for cholesterol, 35% for cholesterol ester, and 42% for progesterone (all P ⬍ 0.01) compared to control cells. The RU 486 exhibited very similar effects and decreased incorporation of [14C]acetate to 29% for cholesterol, 46% for cholesterol ester, and 31% for progesterone compared to control cells (all P ⬍ 0.01). In the same cell preparations, Org 31710 and RU 486 increased apoptosis, measured as DNA fragmentation, by 1.6- and 2.3-fold, respectively, compared to the control value, which is in agreement with the results of previous studies by our group [23]. Effects of HMG-CoA Reductase Inhibitors on Cholesterol Synthesis and Apoptosis in Rat Periovulatory Granulosa Cells

The HMG-CoA reductase-inhibitors lovastatin, simvastatin, or mevastatin were added during in vitro incubations of rat periovulatory granulosa cells to reveal if the decreased de novo cholesterol synthesis was connected to the induction of apoptosis. The presence of any of the three tested statins (10 ␮M) during incubations caused an expected decrease in [14C]acetate incorporation into cholesterol (P ⬍ 0.01) (Fig. 3A). Furthermore, all three statins increased the degree of apoptosis, measured as DNA fragmentation, compared to untreated control cells (P ⬍ 0.01) (Fig. 3B). The increased apoptosis was confirmed by an increase in caspase-3/7 activity as detected using the ApoONE Homogenous Caspase-3/7 Assay (Fig. 3C).

FIG. 2. Effects of Org 31710 (10 ␮M) and RU 486 (10 ␮M) on incorporation of [14C]acetate into cholesterol (Chol), cholesterol ester (CE), and progesterone (P) in (A) periovulatory rat granulosa and (B) periovulatory human granulosa cells in vitro. Primary isolated periovulatory granulosa cells were incubated in serum-free medium for 24 h, and [14C]acetate was added during the last 5 h (n ⫽ 9–14). * P ⬍ 0.05, ** P ⬍ 0.01.

Effect of Mevalonic Acid

Mevalonic acid (1.6 mM), the cholesterol synthesis intermediary immediately downstream of statin inhibition, did not affect the degree of apoptosis on its own, but it was able to reverse, at last in part, the increase in apoptosis observed after treatment with simvastatin (Fig. 4). As expected, mevalonic acid could not reverse the increase in apoptosis caused by RU 486 (Fig. 4). Inhibition of Progesterone Synthesis

Addition of the steroid synthesis-inhibitor cyanoketone (0.01–10 ␮M) to rat periovulatory granulosa cells resulted in a dose-dependent decrease of accumulated progesterone in spent medium (Fig. 5A). Cyanoketone also increased the degree of apoptosis measured as caspase-3/7 activity using the Caspase-Glo 3/7 Assay. This increase was reversed by addition of progesterone (Fig. 5B). Lovastatin, mevastatin, and simvastatin (10 ␮M) did not decrease accumulated progesterone to the extent needed to increase apoptosis by cyanoketone. All three statins decreased the amount of progesterone to approximately 10% of the control value (data not shown).

542

RUNG ET AL.

FIG. 4. Effect of mevalonic acid (Mev; 1.6 mM) on the degree of apoptosis in periovulatory rat granulosa cells incubated with or without addition of RU 486 (RU; 10 ␮M) or simvastatin (Sim; 10 ␮M). The degree of apoptosis was measured as DNA fragmentation (control set to 100%, n ⫽8–10). * P ⬍ 0.05, **P ⬍ 0.01. N.S., Nonsignificant.

FIG. 3. Effects of lovastatin, simvastatin, or mevastatin (all 10 ␮M) on periovulatory rat granulosa cells during 24-h incubation in serum-free medium on (A) incorporation of [14C]acetate into cholesterol (n ⫽ 11– 20), (B) apoptosis measured as DNA fragmentation (control set to 100%, n ⫽ 11–23), and (C) apoptosis measured as caspase-3/7 activity (control set to 100%; n ⫽ 4). **P ⬍ 0.01.

DISCUSSION

In the present study, we demonstrate that PR antagonists and statins inhibit the cholesterogenic pathway and activate apoptosis in rat and human periovulatory granulosa cells. We used expression profiling by DNA microarray analysis to monitor genome-wide changes in transcription after incubation of periovulatory rat granulosa cells in the presence or absence of the specific PR-antagonist Org 31710. Changes in expression levels were detected for more than 80 genes. A subset of these genes are involved in cholesterol homeostasis and form a functional cluster, suggesting a decreased de novo cholesterol synthesis in periovulatory gran-

ulosa cells treated with PR antagonists. The PR-antagonists Org 31710 and RU 486 both increase apoptosis [5, 23, 24] and decrease de novo synthesis of cholesterol, cholesterol ester, and progesterone. Inhibition of progesterone synthesis also increased apoptosis, which could be reversed by addition of progesterone. The pathway of cholesterol synthesis could also be inhibited by statins. In addition to inhibiting cholesterol synthesis, statins (including lovastatin, simvastatin, and mevastatin) also increased apoptosis in periovulatory granulosa cells. The effect of simvastatin was partially reversed by addition of the cholesterol-intermediate mevalonate. These findings show that PR is important for the regulation of cholesterol synthesis and that de novo cholesterol synthesis plays a role in the survival of periovulatory rat and human granulosa cells. In the present study, we focused on the PR antagonistregulated genes that are involved in de novo cholesterol synthesis (Fig. 1). The mitochondrial HMG-CoA synthase was among the down-regulated genes. The mitochondrial HMG-CoA synthase is reportedly involved in cholesterol synthesis in the testis and ovary (see ref [32]). Ruptured ovarian follicles and a subset of luteal cells express mitochondrial HMG-CoA synthase despite the lack of ketone body synthesis. In Leydig cells, sterol synthesis correlates closely with the levels of mitochondrial HMG-CoA synthase. To verify that the decreased mRNA levels for several genes involved in cholesterol synthesis also resulted in decreased cholesterol production, de novo cholesterol synthesis was measured as incorporation of [14C]acetate into cholesterol, cholesterol ester, and progesterone. The results showed a significant decrease of cholesterol synthesis in the PR antagonist-treated rat periovulatory granulosa cells, strengthening the results obtained by microarray analysis. In these experiments, we used two different PR antagonists to ensure that the effects are specific to PR. We previously showed that the increase in apoptosis observed after treatment with RU 486 or Org 31710 in human periovulatory granulosa cells is not caused by interactions with gluco-

GRANULOSA CELL APOPTOSIS AND CHOLESTEROL SYNTHESIS

corticoid, androgen, or gamma aminobutyric acid receptors [23]. That several genes involved in cholesterol synthesis were regulated in a similar way suggests a common regulator. To our knowledge, no functional progesterone response elements have been documented in the cholesterol synthesis genes. A possible candidate as a regulator is the sterol regulatory element-binding protein (SREBP) family, particularly SREBP-2 [33]. Both SREBP-1a and SREBP-2 were not represented on our microarray. In addition, no transcriptional regulation of ADD1, the rat homologue of SREBP-1c, was seen on the microarray. The activity of SREBP, however, is mainly regulated posttranslationally. Thus, lack of evidence on the transcriptional level does not exclude involvement of SREBP. A new sterol-insensitive form of SREBP-2 recently was reported in gonadal tissue [34]. Human granulosa cells were investigated to test if the effects of PR antagonists on cholesterol synthesis were unique to the rat. Human periovulatory granulosa cells share many, but not all, characteristics with rat periovulatory granulosa cells. For instance, in the rat, PR expression is transient during the periovulatory interval. In humans, however, PR expression is prolonged, and the PRs remain in the corpus luteum. Even so, PR antagonists increase apoptosis in human periovulatory granulosa cells [23, 24], as in rat periovulatory granulosa cells [5]. In addition, we show here that the effects of PR antagonists on de novo cholesterol synthesis were similar in rat and humans. Sources of cholesterol for steroid synthesis in ovarian granulosa cells include plasma lipoproteins, stored cholesterol esters, and de novo-synthesized cholesterol. Most of the cholesterol is usually provided by circulating lipoproteins, and the preference for different classes varies between species. Stimulation of the LH receptor has been reported to stimulate incorporation of acetate into sterols and steroids in granulosa cells [35, 36]. At the same time, tissue content of cholesterol and cholesterol ester decreased to 62% and 16%, respectively, of the control value, suggesting that LH-receptor stimulation both stimulates de novo synthesis of cholesterol and increases utilization of stored cholesterol and cholesterol ester for steroid production [35]. It also has been reported that progesterone inhibits esterification of cholesterol in fibroblasts [37] and hepatocytes [38]. Interestingly, the cholesterol synthesis inhibitors lovastatin, simvastatin, and mevastatin increased the degree of apoptosis in addition to efficiently decreasing the incorporation of [14C]acetate into cholesterol in periovulatory granulosa cells. The effect of simvastatin was reversed by addition of mevalonic acid, the cholesterol synthesis intermediary that has its formation blocked by statins. That statins increase the degree of apoptosis supports the idea that decreased cholesterol synthesis may, in turn, cause increased apoptosis. Mevalonic acid was unable to reverse the increase in apoptosis caused by RU 486. According to our microarray results, PR antagonists decrease the expression of several genes in the cholesterol synthesis pathway, including enzymes acting downstream from mevalonate. This explains why mevalonic acid was able to reverse the apoptotic effect of statins but not RU 486. Addition of cyanoketone dose-dependently decreased the progesterone production. The degree of apoptosis was increased at very low (undetectable) levels of progesterone in medium. This was reversed by addition of a high dose of progesterone. This high dose probably was required to

543

FIG. 5. A) Effect of cyanoketone (0.01–10 ␮M) on the level of accumulated progesterone in spent medium after incubations of periovulatory rat granulosa cells for 24 h (n ⫽ 12–24). Progesterone was undetectable in 10 of 12 samples treated with 1 ␮M cyanoketone and in all 12 samples treated with 10 ␮M cyanoketone. B) Effect of cyanoketone (0.01–10 ␮M) on the degree of apoptosis measured as caspase-3/7 activity. Black bars represent the addition of 3 ␮M progesterone (n ⫽ 24–36). * P ⬍ 0.05, **P ⬍ 0.01 compared to treatment with 10 ␮M cyanoketone alone.

achieve high enough intracellular progesterone concentrations compared to those in untreated cells. Statins, however, caused apoptosis at doses that did not lower progesterone levels to the same extent. This suggests that the increase in apoptosis caused by inhibition of cholesterol synthesis can be caused by depletion of factors other than steroids. Besides providing cholesterol for steroid synthesis, the de novo cholesterol synthesis pathway is also important for providing the cell with dolichols, ubiquinone, and substrates for isoprenylation of proteins [39]. The latter group has been shown to be important for cell survival [40, 41]. In conclusion, the specific PR-antagonist Org 31710 decreased the mRNA levels of several genes involved in de novo cholesterol biosynthesis. Org 31710 and RU 486 also decreased the incorporation of [14C]acetate into cholesterol, cholesterol ester, and progesterone in both rat and human granulosa cells and increased apoptosis in vitro. Pharmacological inhibition of cholesterol synthesis by the HMG-

544

RUNG ET AL.

CoA reductase-inhibitors lovastatin, simvastatin, and mevastatin not only decreased cholesterol synthesis but also increased apoptosis. Furthermore, mevalonic acid was able to reverse, at least in part, the effects of simvastatin but not RU 486 on apoptosis. This suggests that de novo synthesis of cholesterol in periovulatory granulosa cells can be regulated by the PR and that this pathway is important for regulating cell survival in the periovulatory interval. ACKNOWLEDGMENTS

18. 19.

20.

The authors thank Margareta Jerna˚ s for excellent technical assistance with the microarray hybridizations, Ann Wallin and Birgitta Weijdega˚ rd for performing the progesterone measurements, and Prof. Sven-Olof Olofsson for valuable help with setting up the thin-layer chromatographic analyses. The Org 31710 was generously provided by N.V. Organon and the RU 486 by Excelgyn. The lovastatin and simvastatin were a gift from Merck.

21.

REFERENCES

23.

1. Gougeon A. Dynamics of follicular growth in the human: a model from preliminary results. Hum Reprod 1986; 1:81–87. 2. Svanberg B, Billig H. Isolation of differentially expressed mRNA in ovaries after estrogen withdrawal in hypophysectomized diethylstilbestrol-treated rats: increased expression during apoptosis. J Endocrinol 1999; 163:309–316. 3. Hsueh AJ, Billig H, Tsafriri A. Ovarian follicle atresia: a hormonally controlled apoptotic process. Endocr Rev 1994; 15:707–724. 4. Markstro¨m E, Svensson EC, Shao R, Svanberg B, Billig H. Survival factors regulating ovarian apoptosis—dependence on follicle differentiation. Reproduction 2002; 123:23–30. 5. Svensson EC, Markstro¨m E, Andersson M, Billig H. Progesterone receptor-mediated inhibition of apoptosis in granulosa cells isolated from rats treated with human chorionic gonadotropin. Biol Reprod 2000; 63:1457–1464. 6. Porter DA, Harman RM, Cowan RG, Quirk SM. Susceptibility of ovarian granulosa cells to apoptosis differs in cells isolated before or after the preovulatory LH surge. Mol Cell Endocrinol 2001; 176:13– 20. 7. Johnson AL, Bridgham JT, Witty JP, Tilly JL. Susceptibility of avian ovarian granulosa cells to apoptosis is dependent upon stage of follicle development and is related to endogenous levels of bcl-xlong gene expression. Endocrinology 1996; 137:2059–2066. 8. Armstrong DT, O’Brien M, Greep RO. Effects of luteinizing hormone on progestin biosynthesis in the luteinized rat ovary. Endocrinology 1964; 75:488–501. 9. Park OK, Mayo KE. Transient expression of progesterone receptor messenger RNA in ovarian granulosa cells after the preovulatory luteinizing hormone surge. Mol Endocrinol 1991; 5:967–978. 10. Natraj U, Richards JS. Hormonal regulation, localization, and functional activity of the progesterone receptor in granulosa cells of rat preovulatory follicles. Endocrinology 1993; 133:761–769. 11. Iwai T, Nanbu Y, Iwai M, Taii S, Fujii S, Mori T. Immunohistochemical localization of oestrogen receptors and progesterone receptors in the human ovary throughout the menstrual cycle. Virchows Arch A Pathol Anat Histopathol 1990; 417:369–375. 12. Press MF, Greene GL. Localization of progesterone receptor with monoclonal antibodies to the human progestin receptor. Endocrinology 1988; 122:1165–1175. 13. Shao R, Markstro¨m E, Friberg PA, Johansson M, Billig H. Expression of progesterone receptor (PR) A and B isoforms in mouse granulosa cells: stage-dependent PR-mediated regulation of apoptosis and cell proliferation. Biol Reprod 2003; 68:914–921. 14. Gava N, Clarke CL, Byth K, Arnett-Mansfield RL, deFazio A. Expression of progesterone receptors A and B in the mouse ovary during the estrous cycle. Endocrinology 2004; 145:3487–3494. 15. Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA Jr, Shyamala G, Conneely OM, O’Malley BW. Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes Dev 1995; 9:2266–2278. 16. Lydon JP, DeMayo FJ, Conneely OM, O’Malley BW. Reproductive phenotypes of the progesterone receptor null mutant mouse. J Steroid Biochem Mol Biol 1996; 56:67–77. 17. Mulac-Jericevic B, Mullinax RA, DeMayo FJ, Lydon JP, Conneely

22.

24.

25.

26.

27. 28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

OM. Subgroup of reproductive functions of progesterone mediated by progesterone receptor-B isoform. Science 2000; 289:1751–1754. Mulac-Jericevic B, Lydon JP, DeMayo FJ, Conneely OM. Defective mammary gland morphogenesis in mice lacking the progesterone receptor B isoform. Proc Natl Acad Sci U S A 2003; 100:9744–9749. Park JI, Kim WJ, Wang L, Park HJ, Lee J, Park JH, Kwon HB, Tsafriri A, Chun SY. Involvement of progesterone in gonadotrophin-induced pituitary adenylate cyclase-activating polypeptide gene expression in preovulatory follicles of rat ovary. Mol Hum Reprod 2000; 6:238– 245. Ko C, In YH, Park-Sarge OK. Role of progesterone receptor activation in pituitary adenylate cyclase activating polypeptide gene expression in rat ovary. Endocrinology 1999; 140:5185–5194. Ko C, Park-Sarge OK. Progesterone-receptor activation mediates LHinduced type-I pituitary adenylate cyclase activating polypeptide receptor (PAC(1)) gene expression in rat granulosa cells. Biochem Biophys Res Commun 2000; 277:270–279. Robker RL, Russell DL, Espey LL, Lydon JP, O’Malley BW, Richards JS. Progesterone-regulated genes in the ovulation process: ADAMTS1 and cathepsin L proteases. Proc Natl Acad Sci U S A 2000; 97: 4689–4694. Svensson EC, Markstro¨m E, Shao R, Andersson M, Billig H. Progesterone receptor antagonists Org 31710 and RU 486 increase apoptosis in human periovulatory granulosa cells. Fertil Steril 2001; 76:1225– 1231. Makrigiannakis A, Coukos G, Christofidou-Solomidou M, Montas S, Coutifaris C. Progesterone is an autocrine/paracrine regulator of human granulosa cell survival in vitro. Ann N Y Acad Sci 2000; 900: 16–25. Magee AI. Protein labeling and immunoprecipitation. In: Bonifacino JS, Dasso M, Harford JB, Lippincott-Schwartz J, Yamada KM (eds.), Current Protocols in Cell Biology, vol. 1. Hoboken, NJ: Wiley; 2000: 7.5.1–7.5.8. Lockhart DJ, Dong H, Byrne MC, Follettie MT, Gallo MV, Chee MS, Mittmann M, Wang C, Kobayashi M, Horton H, Brown EL. Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol 1996; 14:1675–1680. Wyllie AH. Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 1980; 284:555–556. Olega˚ rd R, Svennerholm L. Fatty acid composition of plasma and red cell phosphoglycerides in full term infants and their mothers. Acta Paediatr Scand 1970; 59:637–647. Andersson M, Wettesten M, Bore´n J, Magnusson A, Sjo¨berg A, Rustaeus S, Olofsson SO. Purification of diacylglycerol:acyltransferase from rat liver to near homogeneity. J Lipid Res 1994; 35:535–545. Bengtsson G, Wallin A, Sjo¨blom P, Lindblom B. Effects of hyperosmolar glucose, prostaglandin-F2␣ and 15-methyl-prostaglandin-F2␣ on human placental cells in vitro. Hum Reprod 1995; 10:459–463. Bae S, Seong J, Paik Y. Cholesterol biosynthesis from lanosterol: molecular cloning, chromosomal localization, functional expression and liver-specific gene regulation of rat sterol ⌬8-isomerase, a cholesterogenic enzyme with multiple functions. Biochem J 2001; 353:689– 699. Hegardt FG. Mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase: a control enzyme in ketogenesis. Biochem J 1999; 338:569– 582. Brown MS, Goldstein JL. The SREBP pathway: regulation of cholesterol metabolism by proteolysis of a membrane-bound transcription factor. Cell 1997; 89:331–340. Wang H, Liu F, Millette CF, Kilpatrick DL. Expression of a novel, sterol-insensitive form of sterol regulatory element binding protein 2 (SREBP2) in male germ cells suggests important cell- and stage-specific functions for SREBP targets during spermatogenesis. Mol Cell Biol 2002; 22:8478–8490. Andersen JM, Dietschy JM. Relative importance of high- and lowdensity lipoproteins in the regulation of cholesterol synthesis in the adrenal gland, ovary, and testis of the rat. J Biol Chem 1978; 253: 9024–9032. Morris PW, Gorski J. Control of steroidogenesis in preovulatory cells. Luteinizing hormone stimulation of [14C]-acetate incorporation into sterols. J Biol Chem 1973; 248:6920–6927. Goldstein JL, Faust JR, Dygos JH, Chorvat RJ, Brown MS. Inhibition of cholesterol ester formation in human fibroblasts by an analogue of 7-ketocholesterol and by progesterone. Proc Natl Acad Sci U S A 1978; 75:1877–1881. Synouri-Vrettakou S, Mitropoulos KA. On the mechanism of the mod-

GRANULOSA CELL APOPTOSIS AND CHOLESTEROL SYNTHESIS ulation in vitro of acyl-CoA:cholesterol acyltransferase by progesterone. Biochem J 1983; 215:191–199. 39. Gru¨nler J, Ericsson J, Dallner G. Branch-point reactions in the biosynthesis of cholesterol, dolichol, ubiquinone, and prenylated proteins. Biochim Biophys Acta 1994; 1212:259–277. 40. Tanaka T, Tatsuno I, Uchida D, Moroo I, Morio H, Nakamura S, Noguchi Y, Yasuda T, Kitagawa M, Saito Y, Hirai A. Geranylgeranyl-

545

pyrophosphate, an isoprenoid of mevalonate cascade, is a critical compound for rat primary cultured cortical neurons to protect the cell death induced by 3-hydroxy-3-methylglutaryl-CoA reductase inhibition. J Neurosci 2000; 20:2852–2859. 41. Perez-Sala D, Mollinedo F. Inhibition of isoprenoid biosynthesis induces apoptosis in human promyelocytic HL-60 cells. Biochem Biophys Res Commun 1994; 199:1209–1215.