Use of molecular probes to study regulation of aromatase cytochrome P-450.

Aromatase, an enzyme complex localized in the endoplasmic reticulum of estrogen-producing cells, is composed of NADPH-cytochrome P-450 reductase, and aromatase cytochrome P-450 (cytochrome P-450AROM). To define the molecular mechanisms involved in the multifactorial regulation of cytochrome P-450AROM in estrogen-producing cells, we have isolated a cDNA specific for human cytochrome P-450AROM and have used this cDNA to isolate the human cytochrome P-450AROM gene. The cDNA sequence encodes a polypeptide of 503 amino acids and contains--near the carboxy-terminus, a region of high homology with the putative heme-binding regions of other P-450 cytochromes. COS1 cells transfected with an expression plasmid containing the cytochrome P-450AROM cDNA had the capacity to aromatize testosterone, androstenedione and 16 alpha-hydroxyandrostenedione, suggesting that a single polypeptide catalyzes all steps of the aromatization reaction using either of the three major C19-substrates. The human cytochrome P-450AROM gene is greater than 52 kb in size and consists of 10 exons and 9 introns. Hormonally induced changes in aromatase activity of human ovarian granulosa and adipose stromal cells are associated with comparable changes in cytochrome P-450AROM gene expression and synthesis, whereas the reductase component is only modestly affected. Studies are in progress to define the molecular mechanisms involved in the regulation of cytochrome P-450AROM gene expression in estrogen-producing cells.     

Immunolocalization of aromatase, 17 alpha-hydroxylase and side-chain-cleavage cytochromes P-450 in the human ovary

Immunohistochemical localization of cholesterol side-chain-cleavage, 17 alpha-hydroxylase and aromtase cytochromes P-450 was performed in 35 morphologically normal human premenopausal ovaries by using specific antibodies against the enzymes. In well-developed ovarian follicles in the late stages of follicular growth, immunoreactivity of P-450AROM was only seen in granulosa cells while P-450(17 alpha) and P-450SCC activity was confined to theca interna cells, confirming that follicular oestrogen is produced in granulosa cells by the aromatization of androgens derived from the theca interna cells. In the corpus luteum, this functional differentiation is maintained, since immunoreactivity of P-450AROM was exclusively present in luteinized granulosa cells while that of P-450(17 alpha) was present in luteinized theca calls. Immunoreactivity of P-450SCC was present in both types of cells in the corpus luteum.     

Expression of 3beta-hydroxysteroid dehydrogenase, cytochrome p450 17alpha-hydroxylase/17,20-lyase and cytochrome p450 aromatase enzymes in corpora lutea of diestrous and early pregnant mares

In the pregnant mare, luteal estrogen production increases at the onset of equine chorionic gonadotropin (eCG) secretion by endometrial cups. In previous studies, we have demonstrated that eCG stimulates luteal androgen and estrogen production in pregnant mares. To further elucidate the regulation of steroidogenesis within the equine corpus luteum (CL) of pregnancy, we examined the expression of 3beta-hydroxysteroid dehydrogenase (3beta-HSD), cytochrome P450 17alpha-hydroxylase/17,20 lyase (P450(17alpha)) and cytochrome P450 aromatase (P450(arom)) in luteal tissue samples collected during diestrus (Days 7 to 10) and pregnancy before (Days 29 to 35) and after (Days 42 to 45) the onset of eCG secretion. Immunoblot analyses revealed a single protein per enzyme with molecular weights of 48 kDa (3beta-HSD), 58 kDa (P450(17alpha)) and 56 kDa (P450(arom)). Steady-state levels of 3beta-HSD were lower in luteal tissue of diestrus than pregnancy, but expression did not change during pregnancy. Steady-state expression of P450(17alpha) in CL of diestrus was not significantly different from that of pregnancy. During pregnancy, P450(17alpha) expression was significantly higher after the onset of eCG secretion. Steady-state expression of P450(arom) in CL of diestrus was not significantly different from that of pregnancy. During pregnancy, luteal expression of P450(arom) was significantly lower after the onset of eCG secretion. These data support the hypotheses that eCG has a differential effect on the expression of luteal steroidogenic enzymes, that the eCG-induced increase in luteal estrogen production is the result of an increase in available aromatizable androgen due to an increase in P450(17alpha) expression and activity, and that increased luteal estrogen production is not due to an increase in aromatase expression.     

Immunoreactive cytochrome P-450(17 alpha) in rat and guinea-pig gonads, adrenal glands and brain.

The cytochrome P-450(17 alpha)-hydroxylase, 17----20 lyase (P-450(17 alpha)) is the key enzyme responsible for the biosynthesis of androgens in steroidogenic organs. Its cellular localization has been examined with an immunohistochemical technique. In immature rat ovary, P-450(17 alpha) was first detected in sparse interstitial cells on postnatal Day 8. The number of immunoreactive interstitial cells increased thereafter and the intensity of P-450(17 alpha) staining in these cells was highest at 3 weeks of age. The intensity of staining then started to decline and was very faint at Day 35. From 6 weeks on, the distribution of immunoreactive P-450(17 alpha) was of the adult type: it was detected exclusively in the thecal cells of the large antral, preovulatory, follicles. P-450(17 alpha) was not detectable during pregnancy except on the day of parturition, when thecal cells were transiently immunoreactive. The staining had vanished 24 h after delivery. Human chorionic gonadotrophin (hCG), injected into immature females on Days 24 to 26, induced P-450(17 alpha) prematurely in thecal cells. When injected on Days 12 to 14 of pregnancy, hCG also induced P-450(17 alpha) in the thecal cells surrounding the largest follicles, whereas the interstitial and luteal cells were not immunostained. The antiprogestin RU486, injected on Day 16 of pregnancy, reinstated P-450(17 alpha) (and P-450scc) immunoreactivity in the thecal cells. Oestradiol selectively suppressed P-450(17 alpha) expression in the thecal cells of RU486-treated females. In immature guinea-pig ovary, P-450(17 alpha) was immunostained in thecal cells, not in interstitial cells, although the interstitial cells expressed the delta 5-3 beta-hydroxysteroid dehydrogenase. P-450(17 alpha) was also immunolocalized in the Leydig cells of rat and guinea-pig testes, and in the guinea-pig adrenal cortex (zonae fasciculata and reticularis), but not in the rat adrenal cortex. P-450(17 alpha) was not detectable in the brain of either rat or guinea-pig.     

Changes in content of cytochrome P450(17)alpha, cytochrome P450scc, and 3-hydroxy-3-methylglutaryl CoA reductase in developing rat ovarian follicles and corpora lutea: correlation with theca cell steroidogenesis

The following study was undertaken to determine which hormones (luteinizing hormone, LH, and prolactin, PRL) and enzymes (cytochrome P450(17)alpha, nicotinamide adenine dinucleotide phosphate [NADPH]-cytochrome P450 reductase, 3-hydroxy-3-methylglutaryl [HMG] CoA reductase, cholesterol side-chain cleavage cytochrome P450 [P450scc], and adrenodoxin) were associated with the regulation of androgen biosynthesis by developing rat follicles and corpora lutea in vivo as well as by thecal explants maintained in culture. Immunoblots of soluble cell extracts of small antral (SA), preovulatory (PO), and luteinizing (PO + human chorionic gonadotropin [hCG], 7 h) follicles, newly formed corpora lutea (PO + hCG, 24 h), and corpora luteal isolated on Day 15 of pregnancy, demonstrated that cytochrome P450(17)alpha was low in SA follicles, selectively increased 4-fold in PO follicles, and decreased to less than 10% within 7 h after hCG. Filter hybridization assays using a 32P-labeled cytochrome P450(17)alpha cDNA probe demonstrated that changes in the content of P450(17)alpha mRNA exhibited a pattern similar to that of the enzyme. Conversely, immunoblots for other microsomal enzymes either exhibited no change (NADPH cytochrome P450 reductase) or a transient increase after the hCG surge (HMG CoA reductase), whereas the mitochondrial enzymes either increased markedly in association with luteinization (cytochrome P450scc) or were increased in a more transient manner (adrenodoxin). The LH-induced loss of cytochrome P450(17)alpha in vivo was not associated with loss of androgen biosynthesis when luteinizing theca were placed in culture in medium containing either LH or LH and PRL, suggesting that other hormones, or the presence of other cell types, are required to maintain the decrease in cytochrome P450(17)alpha in vivo. Conversely, the LH-induced increase in cytochrome P450scc in vivo was associated with the maintenance of elevated progesterone production by theca in culture, suggesting that cytochrome P450scc may be constitutively expressed in luteinized theca. Thus, thecal cell cytochrome P450(17)alpha and the regulation of its content and mRNA by LH are pivotal to the biosynthesis of androgens, the obligatory precursors for estradiol biosynthesis and the consequent development of preovulatory follicles. The molecular basis for the different effects of low versus elevated concentrations of LH on cytochrome P450(17)alpha, as well as cytochrome P450scc, remain to be determined.     

Regulation of the biosynthesis of steroidogenic enzymes

Recombinant DNA technology can permit study of the regulation of steroid hydroxylase gene expression at three levels. The first of these is cAMP-regulated gene expression. In the adrenal, ACTH, via cAMP, increases the expression of the genes for all of the cytochrome P-450 species involved in the steroid biosynthetic pathway, as well as the iron-sulfur protein, adrenodoxin. This action of cAMP is inhibited by cycloheximide, suggestive of the involvement of a regulatory protein factor in mediating this action of cAMP. The second level is tissue-specific regulation of steroid hydroxylase gene expression. An example of this which we have studied is the expression of cholesterol side-chain cleavage cytochrome P-450 (P-450sec) and 17 alpha-hydroxylase cytochrome P-450 (P-450(17) alpha) in the bovine ovary. P-450sec is expressed at high levels in the corpus luteum but at low levels in follicles, whereas P-450(17)alpha is expressed in follicles, but is undetectable in the corpus luteum. The third level is fetal imprinting. A number of the cytochrome P-450 species involving in the steroidogenic pathway are expressed in the fetal adrenal at a time when exposure of the gland to ACTH is very low, suggestive that factor(s) other than pituitary ACTH mediate this expression in fetal life.     

Import and processing of P-450SCC and P-450(11) beta precursors by corpus luteal mitochondria: a processing pathway recognizing homologous and heterologous precursors

Maturation of the precursor forms of bovine cholesterol side-chain cleavage cytochrome P-450 (P-450SCC) and 11 beta-hydroxylase cytochrome P-450 (P-450(11)beta) was investigated using mitochondria from bovine corpus luteum. The results show that both precursors, whose synthesis was directed by bovine adrenocortical RNA, can be imported and proteolytically processed to their corresponding mature forms by bovine corpus luteal mitochondria, even though P-450(11)beta is not expressed in this tissue. Furthermore, the efficiency of processing of pre-P-450(11)beta by corpus luteal mitochondria is similar to that of pre-P-450SCC, an endogenous enzyme of these mitochondria. However, the P-450(11)beta precursor is not processed by mitochondria from a nonsteroidogenic tissue (heart), a result observed previously for the P-450SCC precursor (M. F. Matocha and M. R. Waterman (1984) J. Biol. Chem. 259, 8672-8678). This discriminatory processing of pre-P-450(11)beta by heterologous mitochondria suggests that the precursor forms of P-450SCC and P-450(11)beta are processed via a common pathway in steroidogenic mitochondria and that this pathway is absent in nonsteroidogenic mitochondria.     

Expression of retinol-binding protein and cellular retinol-binding protein in the bovine ovary

Retinol (vitamin A) is essential for reproduction, and retinoids have been suggested to play a role in ovarian steroidogenesis, oocyte maturation, and early embryonic development. Retinol is transported systemically and intercellularly by retinol-binding protein (RBP). Within the cell, cellular retinol-binding protein (CRBP) functions in retinol accumulation and metabolism. Since the actions of retinoids are mediated, in part, by retinoid-binding proteins, the objective of this study was to investigate cell-specific expression of RBP and CRBP in the bovine ovary. Immunocytochemical analysis (ICC) localized RBP to the thecal and granulosa cell layers of antral and preantral follicles with the most intense staining in the cells of large, healthy follicles. The tunica adventitia of arterial blood vessels also exhibited RBP staining. Immunostaining of CRBP was most intense in the granulosa cells of preantral follicles and present, but diminished, in thecal and granulosa cells of antral follicles. Within the corpus luteum, both proteins were observed in large luteal cells, but only RBP was observed in small luteal cells. Northern blot analysis demonstrated that thecal and granulosa cells from antral follicles and luteal tissue expressed RBP and CRBP mRNA. Synthesis and secretion of RBP by thecal cells, granulosa cells, and luteal cells were demonstrated by immune-complex precipitation of radiolabeled RBP from the medium of cultured cells or explants, followed by SDS-PAGE and fluorography. Follicular fluid was collected from small (<5 mm) and large (8-14 mm) follicles, pooled according to follicular size, and analyzed for retinol, RBP, estradiol-17beta, and progesterone. Concentrations of retinol, RBP, and estradiol were greater in the fluid of large follicles. Results demonstrate retinoid-binding protein expression by bovine ovaries and provide physical evidence that supports the concept that retinoids play a role in ovarian function.     

In vitro evaluation of aromatase enzyme in granulosa cells using a [11C]vorozole binding assay.

An in vitro method for measuring aromatase cytochrome P450 enzyme (P450AROM) in human granulosa cells (GC) has been developed, based on binding of the 11C-labeled aromatase inhibitor vorozole. GC were obtained following superstimulation during in vitro fertilisation. The method revealed a binding affinity (Kd) of 0.4 nM and a maximum binding (Bmax) at 11 fmol/4000 cells which is equal to 1.6 million binding sites per cell. Linear Scatchard plots indicated a single type of binding site. P450AROM concentrations measured by [11C]vorozole binding correlated positively with aromatisation of [1beta-3H]androst-4-ene-3,17-dione measured as [3H]water release, and a positive association was also found with the ovarian in vivo response to follicle-stimulating hormone (FSH) stimulation expressed as 1000 times the ratio of the number of oocytes recovered from a patient and the total dose of recombinant FSH administered. Frozen cells could be used for P450AROM quantitation, provided the correct freezing procedure was used. Quantitation of P450AROM, based on binding of [11C]vorozole is an accurate and sensitive in vitro method, which might be extended to the measurement of aromatase expression by a noninvasive technique in the intact ovary in vivo using positron emission tomography.     

Inverse relationship between the expression of messenger ribonucleic acid for peroxisome proliferator-activated receptor gamma and P450 side chain cleavage in the rat ovary

Messenger RNA for peroxisome proliferator-activated receptor gamma (PPARgamma) has been found in granulosa cells, and its expression decreases after the LH surge. We determined which developmental stage of ovarian follicle expresses mRNA for PPARgamma and evaluated the impact of PPARgamma agonists on steroidogenesis. Ovaries were collected from immature eCG/hCG-treated rats at 0 (no eCG), 24, and 48 h post-eCG and 4 and 24 h post-hCG. Ovarian tissue was serially sectioned and processed for in situ hybridization to localize mRNA corresponding to PPARgamma, aromatase, and the LH receptor, and P450 side chain cleavage (P450SCC) and to determine whether apoptotic cells were present. During follicular development, there was no correlation between the expression of mRNAs for PPARgamma and aromatase or the presence of apoptotic cells, but a general inverse correlation was observed between the expression of PPARgamma mRNA and LH receptor mRNA. At 4 h post-hCG, follicles expressing P450SCC mRNA had lost expression of PPARgamma mRNA. This inverse pattern of expression between PPARgamma and P450SCC mRNAs was also observed 24 h post-hCG, with developing luteal tissue expressing high levels of P450SCC mRNA but little or no PPARgamma mRNA. To determine the impact of PPARgamma on steroidogenesis, granulosa cells were collected from ovaries 24 h post-eCG and cultured alone, with FSH alone, or with FSH in combination with the PPARgamma agonists ciglitazone or 15-deoxy-delta 12,14-prostaglandin J2 (PGJ2). Treatment of granulosa cells with PGJ2 stimulated basal progesterone secretion, whereas ciglitazone or PGJ2 had no significant effect on FSH-stimulated steroid production. These findings suggest that 1) PPARgamma may regulate genes involved with follicular differentiation and 2) the decline in PPARgamma in response to LH is important for ovulation and/or luteinization.     

Cell-specific variations and hormonal regulation of immunoreactive cytochrome P-450scc in the rat ovary

Specific rabbit antibodies to the bovine cholesterol side-chain cleavage cytochrome P-450 (P-450scc) were used to cross-react with the enzyme in the rat ovary. The luteal cells of cyclic, pregnant, and pseudopregnant rats were immunostained. P-450scc was also expressed in the interstitial cells of prepubertal and cyclic adult rats, and in the thecal cells lining the preovulatory follicles. In cyclic females, RU 486 and oestradiol increased the intensity of P-450scc immunostaining. The granulosa cells of ovarian follicles whatever their stage of development, including preovulatory follicles, were not labelled, except after ovulation. The intensity of immunostaining of thecal and interstitial cells decreased during early pregnancy or pseudopregnancy, and disappeared after Day 9, whereas these cells were intensely labelled 24 h after parturition. The immunostaining of thecal and interstitial cells was again detected in 18-day pregnant rats, treated with the antiprogesterone RU 486. It is therefore concluded that both oestradiol and progesterone are involved in P-450scc regulation.     

Regulation by follicle-stimulating hormone of the synthesis of aromatase cytochrome P-450 in human granulosa cells

The effects of FSH to increase the activity of aromatase, as well as the synthesis of the components of the aromatase enzyme complex, have been studied in human ovarian granulosa cells obtained from women undergoing oocyte retrieval. FSH increased aromatase activity, as well as the synthesis of aromatase cytochrome P-450 (P-450AROM) in a time-dependent fashion, whereas in the absence of FSH, both activity and synthesis declined with duration of culture. The effect of FSH was mimicked by forskolin, an activator of adenylate cyclase. FSH also increased the synthesis of NADPH-cytochrome P-450 reductase, but to a relatively modest extent. The levels of hybridizable mRNA species encoding cytochrome P-450AROM of lengths 3.0, 2.4, and 1.6 kilobases were also increased with FSH treatment. It is concluded that the regulation of aromatase activity by FSH in human granulosa cells is mediated primarily by changes in the synthesis of cytochrome P-450AROM, that this action of FSH is mediated by cAMP, and that the changes in cytochrome P-450AROM synthesis are the consequences of changes in the levels of mRNA encoding this enzyme.     

A novel basal lamina matrix of the stratified epithelium of the ovarian follicle.

Basal laminas are important sheets of specialized extracellular matrix that underlie and surround groups of cells, such as epithelia or endothelia, enabling the cells to orientate their basal/apical polarity and creating a microenvironment for them. Basal laminas can also individually encapsulate whole cells, such as muscle cells, thereby forming a microenvironment but not polarizing the enclosed cells. Other mesenchymal or stromal cells exist with no basal lamina. In the course of studying the bovine follicular basal lamina which underlies the multilayered epithelium of the ovarian follicle, we identified a developmentally regulated novel extracellular matrix (which we call focimatrix for focal intra-epithelial matrix). Focimatrix is composed of basal lamina-like material deposited as plaques or aggregates between the multilayers of the epithelial granulosa cells. The focimatrix does not encapsulate individual or groups of cells and therefore does not form a microenvironment for them. Focimatrix contains collagen type IV subunits alpha1 and alpha2 (but not alpha3-alpha6), and laminin chains alpha1, beta2 and gamma1 (but not alpha2 or beta1), and nidogen-1 and perlecan (but not versican). The amount of focimatrix increases with increasing follicular size, and its appearance precedes the expression by granulosa cells of the enzymes for steroid hormone synthesis, cholesterol side-chain cleavage cytochrome P450 (SCC) and 3beta-hydroxysteroid dehydrogenase (3beta-HSD), in the days preceding ovulation. The expression in granulosa cells of two components examined, nidogen-1 and perlecan, also increases substantially when follicles enlarge to a sufficient size capable of ovulating. Following ovulation the follicular basal lamina is degraded, and presumably focimatrix is too since it is not detected in corpora lutea that develop from the ovulating follicles. During this development the granulosa cells undergo an epithelial-mesenchymal transition (EMT) into luteal cells following ovulation, and substantially increase their expression of steroidogenic enzymes in the process. During EMT epithelial cells lose polarity. Since focimatrix exists on more than one side of the granulosa cells, we propose that it disrupts the polarity induced by the follicular basal lamina in the lead up to ovulation. Hence focimatrix maybe a key part of the follicular/luteal EMT.