Reproductive Biology and Endocrinology Regulation of P450 Oxidoreductase by Gonadotropins in Rat Ovary and Its Effect on Estrogen Production

Background: P450 oxidoreductase (POR) catalyzes electron transfer to microsomal P450 enzymes. Its deficiency causes Antley-Bixler syndrome (ABS), and about half the patients with ABS have ambiguous genitalia and/or impaired steroidogenesis. POR mRNA expression is up-regulated when mesenchymal stem cells (MSCs) differentiate into steroidogenic cells, suggesting that the regulation of POR gene expression is important for steroidogenesis. In this context we examined the regulation of POR expression in ovarian granulosa cells by gonadotropins, and its possible role in steroidogenesis.


Background
The ovary is one of the major steroidogenic organs, and ovarian follicular development is initiated by follicle stimulating hormone (FSH) secreted by the pituitary gland. FSH-induced follicular development involves granulosa cell proliferation and differentiation [1,2]. FSH induces steroidogenic enzymes in ovarian granulosa cells leading to the production of steroid hormones, including estrogens. P450 aromatase (CYP19) is one of the steroidogenic enzymes regulated by FSH, and catalyzes the biosynthesis of estrogens from C19 steroids in the granulosa cells [3].
P450 oxidoreductase (POR) is an 82-kDa membranebound flavoprotein [4,5] responsible for the transfer of electrons from nicotine adenine dinucleotide phosphate to flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN), and finally to microsomal P450 enzymes, including aromatase, as well as to several non-P450 enzymes [6,7]. Several clinical studies have reported that mutations in POR are sufficient to cause Antley-Bixler syndrome (ABS), which is characterized by skeletal malformations. In some cases, ABS patients have genital ambiguity, and in some of the cases female ABS patients present with polycystic ovary syndrome [6,[8][9][10][11]. The clinical features are primarily explained by the impaired activities of CYP21A2, CYP17A1, and CYP19 involved in steroidogenesis [11,12].
Bone marrow-derived mesenchymal stem cells (MSCs) are defined as pluripotent cells, and have been shown to differentiate into adipocytes, chondrocytes, osteoblasts, and hematopoietic-supporting stroma, both in vivo and ex vivo [13][14][15]. We recently reported that stable transfection of MSCs with the transcription factor, steroidogenic factor-1 (SF-1), followed by treatment with cAMP, led to their conversion into steroidogenic cells [16]. These cells offer a possible clinical source of stem cells for the treatment of diseases of steroidogenic organs, and represent a powerful tool for studying the differentiation of the steroidogenic lineage.
In this study, we investigated the changes in gene expression in MSCs during differentiation into steroidogenic cells. POR gene expression, as well as that of various steroidogenic enzymes, was induced by SF-1 and cAMP. It was also induced by gonadotropins in rat ovary and granulosa cells. We also examined the effects of wild-type and mutant POR expression on aromatase activity, and the effects of knockdown of endogenous POR in mammalian cell lines. Our results suggest that up-regulation of POR expression by gonadotropins, as well as that of aromatase, could enhance ovarian estrogen synthesis.

Rat granulosa cell culture
Granulosa cells were obtained from immature, SLC: SD female rats (21 days old) that had received injections of 2 mg DES in 0.2 ml sesame oil once daily for 4 d. The ova-ries were then excised and granulosa cells were released by puncturing the follicles with a 26-gauge needle. Animals were treated according to NIH guidelines and the experiments were approved by the animal experiment committee of the University of Fukui. Granulosa cells were washed and collected by brief centrifugation, and cell viability was determined by trypan blue exclusion. The granulosa cells were then cultured in Ham's F-12/DMEM supplemented with 50 mg/l gentamicin sulfate and 0.1% bovine serum albumin on collagen-coated plates in a humidified atmosphere containing 5% CO 2 in 95% air at 37°C [19].

Western blot analysis
In order to determine protein expression levels, granulosa, KGN, and COS-7 cells (purchased from American type culture collection) were washed twice in ice-cold phosphate-buffered saline and lysed in RIPA buffer (50 mM Tris-HCl (pH 8.0) containing 1% Nonidet P-40, 0.1% sodium dodecyl sulfate (SDS), 0.5% deoxycholate, and 1 mM phenylmethylsulfonyl fluoride). The cell debris was removed by centrifugation at 15,000 × g at 4°C for 15 min, and supernatants were used as cell lysates. Proteins were quantified using a Bio-Rad Protein Assay kit. Equal amounts of protein (30 μg) were resolved by SDS-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes. Western blot analysis of POR, CYP19, and GAPDH was performed with antisera directed against POR (raised against amino acids 1-300 of human POR), CYP19 (synthetic peptide corresponding to amino acids 376-390 of human CYP19), and GAPDH, respectively. All antibodies were capable of reacting with both human and rat antigens, respectively. Western blot analysis of POR, aromatase, and GAPDH was performed with antisera directed against POR, CYP19, and GAPDH, respectively. Enhanced chemiluminescence western blot reagents (Amersham Pharmacia Biotech, Arlington Heights, IL) were used for detection.
Transfection COS-7 cells were cultured in DMEM supplemented with antibiotics and 10% FBS in a humidified atmosphere containing 5% CO 2 and 95% air at 37°C. DNA transfections were carried out using Lipofectamine Plus reagent. All plasmids used for transfections were prepared using a Genopure Plasmid Maxi Kit. Cells (4 × 10 5 ) were cultured in 60-mm dishes in 5 ml medium on the day before transfection. Three hours after co-transfection of human aromatase (1 μg/well) and rat POR (20, 50, or 100 ng/well) or pcDNA3, medium was changed to growth medium. After 48 h, the growth medium was changed to medium containing androstenedione (final concentration 10 -5 M). Culture medium was collected at various time points in order to observe the conversion of androstenedione to estrone.

EIA
The levels of estrone in the media were measured using an EIA assay kit (Cayman) according to the manufacturer's instructions. Briefly, 50 μl of diluted medium (1:1000), estrone tracer, and antiserum were added to each well. They were incubated at room temperature for 2 h and developed with Ellman's reagent solution. The developed plates were measured using a plate reader at 405 nm.

SiRNA transfection
KGN cells were cultured in DMEM/Ham's F12 supplemented with antibiotics and 10% FBS in a humidified atmosphere containing 5% CO 2 and 95% air at 37°C. Knockdown of endogenous POR was carried out using siRNAs and HiPerFect reagent. Twenty-four hours after transfection, the culture medium was changed to DMEM/ Ham's F12 supplemented with antibiotics and 10% FBS containing 1 mM 8-Br-cAMP and 10 μM androstenedione. Cells and medium were collected 48 h after transfection. SiRNAs for POR were POR #1 (sense: 5' GCACAUCU-GUGCGGUGGUUTT, anti sense: 5' AACCACCGCACA-GAUGUGCTT) and POR #2 (sense: 5' CCAAC UGG CUGCGGGCCAATT, anti sense: UUGGCCCGCAGCC AGUUGGTT). They were targeted to the region approximately 1496 and 1568 nucleotides downstream from the beginning of the human POR (Ref seq NM_000941), respectively.

Statistics
Data from each experiment were analyzed using the Student's t test. P < 0.01 was considered statistically significant.

Changes in gene expression in human MSCs during differentiation into steroidogenic cells
In previous studies, we converted MSCs into steroidogenic cell-lineages such as Leydig cells and adrenocortical cells by expression of SF-1 and treatment with cAMP [16]. DNA microarray analysis was performed using control and SF-1-transduced human MSCs in order to examine alterations of gene expression. Using more than 22,000 probe sets (18,400 transcripts and variants, including 14,500 well-characterized human genes), a number of differences in gene expression levels between control cells and UE7T-13 cells stably expressing SF-1 were detected: expression levels of about 1,700 genes, including POR, were found to be increased more than 5-fold in SF-1-transduced cells (representative data from two independent experiments are shown in Table 1). Gene expression of StAR [22], CYP11A1 [23], HSD3B2 [24], CYP19 [3], and CYP17A1 [25], whose transcriptions are known to be regulated by SF-1, were induced. Consistent with the results of DNA microarray, RT-PCR and real-time PCR analysis showed that POR mRNA was markedly increased more than 20fold in UE7T-13 cells transduced with SF-1. It was further induced by the addition of 8-Br-cAMP, and reached to about 31-fold ( Fig. 1A and 1B). These results indicate that, concomitant with microsomal P450 genes, the POR gene was induced during the differentiation of MSCs into steroidogenic cells, and suggest that its up-regulation may contribute to the development of the steroidogenic phenotype.

Induction of POR expression by gonadotropins in the ovary
In order to evaluate the relevance of above hypothesis, we examined the expression of POR in the ovary and in granulosa cells, an organ and cells that are well known to be involved in steroidogenesis or in the differentiation into steroidogenic by gonadotropins. Ovarian POR mRNA was detected at relatively low levels before gonadotropin treatment. It was induced after 48 h of eCG treatment, and was further increased within 2 h of the following hCG treatment ( Fig. 2A and 2B). Its expression then fell substantially at 8 h, and continued to decline thereafter. The pattern of POR expression induced by eCG and hCG was similar to that of Cyp19 expression.
POR mRNA and proteins were strongly induced by FSH in granulosa cells after 3-6 h. They fell temporarily at 12 h, and then increased again from 24-48 h (Fig. 3A, B and 3C). Cyp19 mRNA and protein were barely detectable before FSH treatment. They were strongly induced by FSH, and followed a similar pattern to that of POR, although Cyp19 levels responded more quickly than POR levels. Consistent with the results from MSCs, POR was induced in the ovary and granulosa cells when they were stimulated to differentiate into steroidogenic cells by gonadotropins. Induction of Cyp19 protein by FSH causes Regulation of POR and aromatase mRNA and protein levels by FSH. Granulosa cells isolated from immature DES-primed rats were cultured and treated with FSH (30 ng/ ml) for the indicated times, and the total RNA or protein was extracted. A: RT-PCR was performed with the specific primer sets for POR, aromatase, and GAPDH. B: Results of real-time PCR are also shown. Data are the mean ± SEM values of at least three independent experiments. *, P < 0.01. C: Western blot analysis was performed with antibodies to POR, aromatase, and GAPDH, using the same lysates.  [26]. These observations suggest that FSH-induced up-regulation of POR expression, as well as that of Cyp19, could enhance estrogen synthesis in granulosa cells.

Effects of POR expression on estrogen synthesis
In order to determine if estrogen synthesis was affected by the level of POR expression when aromatase expression remained constant, POR and aromatase expression vectors were transiently co-transfected into COS-7 cells that expressed an undetectable level of aromatase protein (identified by Western blotting) and low levels of POR expression (Fig. 4A and 4B). In order to investigate aro-matase enzyme activity, we measured the conversion of androstenedione to estrone in the culture medium. Estrone levels were low and were unaffected by co-transfection with both control vectors, or with the POR expression vector alone. However, time-dependent accumulation of estrone occurred when aromatase was expressed, and estrone levels were further increased by cotransfection with the POR expression vector. Co-transfection with the aromatase expression vector and 100 ng of the POR expression vector resulted in a 6-fold increase in the amount of estrone converted in 6 h, compared with the aromatase vector alone (Fig. 4A and 4B). These results suggest that estrogen synthesis is up-regulated by increas-Production of estrone by co-transfection of wild-type or mutant POR and aromatase expression vectors in COS-7 cells  ing levels of POR expression, even when aromatase expression is constant.
We then examined the effects of the expression of mutant POR genes on estrone production. It has been reported that aromatase activity was completely lost when the R457H or V492E POR mutant proteins were expressed instead of the wild-type protein [27]. In COS-7 cells cotransfected with R457H or V492E mutant POR and aromatase expression vectors, the amount of estrone was reduced to 25% compared to co-transfection with wildtype POR and aromatase. These estrone levels were similar to those seen when aromatase was expressed alone (Fig.  4C and 4D). These results suggest that estrogen synthesis is dependent not only on the expression levels of aromatase protein, but also on the expression levels of POR protein.

Effects of knockdown of endogenous POR protein on estrogen production
We investigated the effects of knockdown of endogenous POR by siRNA in granulosa cell-derived KGN cells. Androstenedione and 8-Br-cAMP were added to the culture medium 24 h after transfection with 10 nM siRNA and the amount of estrone in the medium was measured after 48 h. In contrast to the rat granulosa cells, POR expression was relatively high in KGN cells and was not greatly changed by 8-Br-cAMP. Aromatase expression, however, was induced by 8-Br-cAMP, as in rat granulosa cells treated with FSH (Fig. 5A). POR siRNA caused a reduction in the amount of estrone to about 70% of that in cells transfected with a scrambled control siRNA (Fig.  5C). No estrone was detected in the absence of 8-Br-cAMP.

Discussion
In our previous study, we reported that stable transfection of MSCs with SF-1, followed by treatment with cAMP, led to their conversion into steroidogenic cells [16]. These cells offer a possible clinical source of stem cells for the treatment of diseases of steroidogenic organs, and represent a powerful tool for studying the differentiation of the steroidogenic lineage [16,[28][29][30]. In this study, using DNA microarray, RT-PCR, and real-time PCR analysis, we observed that POR mRNA was induced during the differentiation of hMSCs into steroidogenic cells. We also demonstrated that the expression of the POR mRNA and protein were up-regulated by FSH during the differentiation of rat granulosa cells into steroidogenic cells. Similar phenomena have been observed in adrenocortical cell lines stimulated with adrenocorticotropic hormone (ACTH) or cAMP [31,32]. These results suggest that POR expression in steroidogenic cells is regulated by the pituitary hormone/cAMP pathway. In contrast, it has been reported that POR gene expression in hepatic cells and other organs was not changed by any stimuli, including cAMP, and in support of these observations, the promoter region of the POR gene supported constitutively high luciferase activity that was unaffected by any stimuli [33][34][35][36]. These results suggest that regulation of POR gene expression differs between steroidogenic and non-steroidogenic cells. Although the reasons for this difference remain unclear, the expression of SF-1 in steroidogenic cells may explain this phenomenon. SF-1 is expressed in steroidogenic cells such as adrenocortical cells, testicular Leydig cells, ovarian theca and granulosa cells, and regulates steroidogenic genes that encode cytochrome P450scc, StAR and CYP19, by binding to cAMP-response sequences [3,22,23]. Our results from MSCs indicate that the POR gene could represent a new target for SF-1 in steroidogenic cells. It is therefore conceivable that the expression of POR is regulated by the pituitary hormone/cAMP pathway. Such regulations may also cause similar fluctuations in POR and CYP19 expression in the ovary and in granulosa cells. However, CYP19 mRNA was induced more rapidly than POR mRNA, because the CYP19 gene is regulated by CREB, as well as by SF-1 [3]. On the other hand, our preliminary study indicated that CREB was not involved in the induction of POR mRNA in granulosa cells by FSH (data not shown). The POR gene spans more than 70 kb, including an untranslated first exon residing about 38.8 Effect of POR knockdown on estrone production and cAMP response of POR and aromatase genes in KGN cells   kb upstream of the coding region [4]. Further studies are needed to understand the mechanisms regulating POR gene expression.
Based on our observations in rat ovary and in granulosa cells stimulated by gonadotropins, it is likely that increasing POR expression induced by gonadotropins further upregulates aromatase activity, and that estrogen production is regulated not only by aromatase expression, but also by POR. This proposition is supported by the augmentation of aromatase activity coincident with increasing POR expression in COS-7 cells, despite a constant expression of aromatase. Hall and his colleagues, using purified enzymes, also reported that the activity of CYP17 was increased by the addition of POR protein in a dosedependent manner, resulting in increased androgen production [37]. Because almost all steroidogenic enzymes are up-regulated by the pituitary hormone/cAMP pathway, our model should be applicable to other microsomal steroidogenic P450s, such as CYP17 and CYP21. In support of this hypothesis, CYP17 expression was markedly increased, together with POR, in SF-1-transduced MSCs. Similarly, the expression of ferredoxin, which is a member of the electron transport chain in mitochondria and regulates pregnenolone production by P450scc, is known to be increased by pituitary hormones and cAMP [38][39][40]. Therefore, it is possible that, in addition to steroidogenic enzyme expression, up-regulation of electron transporter expression plays a vital role in steroid hormone production by gonadotropins and ACTH. This mechanism should be conserved between mitochondrial and microsomal steroidogenesis.
In COS-7 cells, a low level of estrone was detected in the culture medium in the absence of transfection of CYP19, even though aromatase proteins were not detected by western blotting. COS cells (COS-1 and COS-7) have often been used for measuring aromatase activity by ectopic expression of the CYP19 gene, and low aromatase activities have been reported under non-transfected conditions [41][42][43][44]. It is thus conceivable that aromatase proteins are expressed at very low levels in this cell line. It has been reported that the R457H and V492E POR mutations are associated with loss of P450 enzyme activities, because these mutations in the FAD domain hamper the ability of POR to transfer electrons to its FMN group, which ultimately provides electrons to P450 enzymes [27]. When these mutants were expressed concomitantly with aromatase in COS-7 cells, estrone production was similar to that seen with aromatase alone. This result indicates that the augmentation of aromatase activity by POR is not a side effect of overexpression of POR, but is due to the increased electron supply resulting from increased POR expression in COS-7 cells.
FSH, forskolin, and cAMP have been reported to cause increases in aromatase activity in KGN cells (a human ovarian granulosa-like tumor cell line), as well as in rat granulosa cells [45,46]. In contrast, POR expression in KGN cells was little changed by 8-Br-cAMP treatment, and was much higher than in primary cultures of rat granulosa cells. In addition, it has been reported that the lower amount of POR protein, compared with the amount of P450 protein, was still sufficient to carry out all metabolic reactions [47]. Hence, the amount of POR expression after inhibition by siRNA, still seemed to be sufficient to activate aromatase induced by 8-Br-cAMP. Therefore, estrone production should not be greatly reduced by POR siRNA. POR is involved not only in steroidogenesis, but also in drug and xenobiotic metabolism, and in tumorigenesis [48,49]. It is likely that the hormone-responsiveness of the POR gene is lost during immortalization of KGN cells. However, induction of POR in granulosa cells may play a more significant role in estrogen production, because the amount of POR protein in these cells is limited. In granulosa cells, POR and aromatase expression are regulated by gonadotropins, which in turn may be responsible for coordinated regulation of steroid hormone production at each stage of follicular growth.
In this study, we have demonstrated that estrogen production was caused not only by aromatase expression, but also through the promotion of aromatase activity by FSHinduced POR expression in ovarian granulosa cells. POR knockout mice are embryonically lethal, but POR heterozygous knockout mice are fertile, and do not have ambiguous genitalia [50]. Our study suggests that a comparison of ovarian function and morphology between POR heterozygous knockout mice and aromatase knockout mice [51] could provide further useful information. In addition, POR might be involved in the differentiation of these steroidogenic organs through activation of microsomal P450 enzymes.