Induction of synthesis of cholesterol side chain cleavage cytochrome P-450 and adrenodoxin by follicle-stimulating hormone, 8-bromo-cyclic AMP, and low density lipoprotein in cultured bovine granulosa cells

Synergism between FSH and activin in the regulation of proliferating cell nuclear antigen (PCNA) and cyclin D2 expression in rat granulosa cells

Follicular development is associated with both proliferation and differentiation of granulosa cells under the control of FSH. We show that regulation of genes involved in cellular proliferation by FSH can be functionally separated from the regulation of genes involved in granulosa cell differentiation by synergistic actions of activin and T. Incubation of undifferentiated rat granulosa cells with FSH, forskolin, activin-A, or T alone did not influence either the expression of the proliferation-associated genes cyclin D2 and proliferating cell nuclear antigen or the differentiation-associated genes P450 aromatase, LH receptor, P450 cholesterol side-chain cleavage enzyme, and 3 beta-hydroxysteroid dehydrogenase. However, when granulosa cells were stimulated with either FSH or forskolin in the presence of activin-A, significant increases (P < 0.05) were observed for cyclin D2 and proliferating cell nuclear antigen at both the mRNA and protein levels as well as mRNAs for P450 aromatase, LH receptor, P450 cholesterol side-chain cleavage enzyme and 3 beta-hydroxysteroid dehydrogenase. Although T synergized with FSH to increase the expression of mRNAs for P450 aromatase, LH receptor, P450 cholesterol side-chain cleavage enzyme, and 3 beta-hydroxysteroid dehydrogenase, it did not interact with FSH to increase the expression of mRNAs for cyclin D2 and proliferating cell nuclear antigen. The differences in the actions of activin and T could provide a cellular mechanism by which FSH-regulated granulosa cell proliferation could be functionally separated from FSH-regulated granulosa cell differentiation.

Prostaglandin F(2alpha) induces a rapid decline in progesterone production and steroidogenic acute regulatory protein expression in isolated rat corpus luteum without altering messenger ribonucleic acid expression

With interest in steroidogenic acute regulatory protein (StAR) involvement in the luteolytic process, we studied changes in serum progesterone levels and the concomitant expression of StAR mRNA and protein (37-, 32-, and 30-kDa forms) in postovulatory Day 7 corpora lutea (CL) isolated from rats 1 h after injection with prostaglandin F(2alpha) (PGF(2alpha), n = 6) or saline (n = 6). Serum progesterone levels were determined by RIA, StAR and beta-actin mRNA expression by Northern analysis, and StAR and beta-actin protein expression by Western analysis. Adrenal, brain, and spleen from control animals were used as positive and negative controls for StAR expression. Scanning optical densitometry measurements were standardized by dividing the signal strength from each StAR autoradiogram lane by that from the corresponding beta-actin autoradiogram lane. ANOVA was used for significance testing, with alpha set at 0.05. The 37-, 32-, and 30-kDa forms of StAR protein were expressed in all adrenal samples, whereas only the 37- and 30-kDa forms were found in CL. Serum progesterone levels and expression of the 30-kDa and 37-kDa forms of the StAR protein in CL were all found to be significantly lower in the PGF(2alpha)-treated than the saline-treated group. StAR mRNA expression was not significantly different in the saline- and PGF(2alpha)-treated rats. The rapid decline in StAR protein expression that accompanies PGF(2alpha) induced luteolysis, therefore, does not result from significant decline in mRNA expression.

Regulation of cytochrome P450scc synthesis and activity in the ovine corpus luteum

The rate-limiting step in luteal biosynthesis of progesterone consists of cleavage of the side chain of cholesterol by mitochondrial cytochrome P450 side-chain cleavage enzyme (P450scc) to form pregnenolone. Luteal mRNA encoding P450scc, quantitated on selected days of the 16-day ovine estrous cycle, was similar on days 3 and 6, increased by 2-fold on day 9 (P < 0.05) and remained elevated on day 15. Levels of P450scc mRNA on day 15 of pregnancy were not different from those found on any day of the cycle (P < 0.05). To determine whether levels of mRNA encoding P450scc are hormonally regulated, ewes on day 10 of the estrous cycle were injected with hCG or prostaglandin F2 alpha (PGF2 alpha). P450scc mRNA was not increased for up to 36 h after injection of hCG, nor decreased within 8 h after injection of PGF2 alpha (P < 0.05). An assay for P450scc activity was developed which utilized ovine small and large luteal cells in the presence of 22R-hydroxycholesterol and ovine high density lipoprotein. Enzyme activity was quantitated by measurement of progesterone production. In small luteal cells activation of the protein kinase A (PKA) second-messenger system by treatment with LH resulted in 910% increase in progesterone production without altering activity of P450scc. Activation of the protein kinase C (PKC) second-messenger system with phorbol 12-myristate 13-acetate caused a 51% reduction in progesterone secretion from large luteal cells but did not alter activity of P450scc. These findings suggest that in mature luteal tissue steady state levels of mRNA encoding P450scc, and enzyme activity are independent of acute regulation by activation of PKA or PKC second-messenger systems.

Plasticity of cell organization during differentiation of normal and oncogene transformed granulosa cells

Granulosa cells, which nurse the oocyte and serve as a major source for estradiol and progesterone production, undergo major morphological changes which correlate very well with modulation of their steroidogenic capacity. These include changes in intercellular contacts and communication, in cell membrane receptors, and in the development and organization of organelles associated with steroidogenesis (i.e., mitochondria, smooth endoplasmic reticulum, lipid droplets, and lysosomes). These biochemical and morphological changes can also be obtained in primary cultures as well as in oncogene transformed granulosa cell lines established recently in our laboratory. A growing body of evidence suggests that plasticity of the cytoskeleton plays a major role in the biochemical and morphological differentiation of granulosa cells as well as in other steroidogenic cells.

Characterization of progesterone biosynthesis in a transformed granulosa cell line

The study of regulation of steroidogenesis in primary cultures of rat granulosa cells is difficult because the cells do not undergo more than one cell doubling in culture. Furthermore, there is size and steroidogenic heterogeneity in granulosa cells and it is difficult to obtain pure, functionally defined populations. Hence, it is advantageous to develop a homogeneous population of granulosa cells. In this report we describe the characterization of one such cell line (Rao-gcl-29) developed from diethylstilbestrol treated immature rat granulosa cells by transformation with SV40 T antigen. In this cell line cyclic AMP analogs induce high levels of progesterone biosynthesis, though there was no effect on estradiol biosynthesis. Also, FSH and hCG have no effect on progesterone biosynthesis. In the presence of FBS medium (20% fetal bovine serum in DMEM/F-12) and enriched medium (10% fetal bovine serum, 10% horse serum and 2% UltraSer G in DMEM/F-12 medium), 1 mM cAMP analogs induce high levels of progesterone biosynthesis up to 96 h. Ultrastructural features of the cell line resemble those of primary granulosa cells, in addition to forming gap junctions. Cyclic AMP analogs also induced cytochrome P450scc mRNA in this cell line by 48 h, and this effect is apparent by 24 h. Thus, this cell line could be useful in understanding the molecular mechanisms of regulation of cytochrome P450scc gene regulation.

Steroidogenic enzymes: structure, function, and role in regulation of steroid hormone biosynthesis

In the pathways of steroid hormone biosynthesis there are two major types of enzymes: cytochromes P450 and other steroid oxidoreductases. This review presents an overview of the function and expression of both types of enzymes with emphasis on steroidogenic P450s. The final part of the review on regulation of steroidogenesis includes a description of the normal physiological fluctuations in the steroid output of adrenal cortex and gonads, and provides an analysis of the relative role of enzyme levels in the determination of these fluctuations. The repertoire of enzymes expressed in a steroidogenic cell matches the cell's capacity for the biosynthesis of specific steroids. Thus, steroidogenic capacity is regulated mainly by tissue and cell specific expression of enzymes, and not by selective activation or inhibition of enzymes from a larger repertoire. The quantitative capacity of steroidogenic cells for the biosynthesis of specific steroids is determined by the levels of steroidogenic enzymes. The major physiological variations in enzyme levels, are generally associated with parallel changes in gene expression. The level of expression of each steroidogenic enzyme varies in three characteristics: (a) tissue- and cell-specific expression, determined during tissue and cell differentiation; (b) basal expression, in the absence of trophic hormonal stimulation; and (c) hormonal signal regulated expression. Each of these three types of expression probably represent the functioning of distinct gene regulatory elements. In adult steroidogenic tissues, the levels of most of the cell- and tissue-specific steroidogenic enzymes depend mainly on trophic hormonal stimulation mediated by a complex network of signal transduction systems.

Regulation of steroid hydroxylase gene expression is multifactorial in nature

In summary, regulation of steroid hydroxylase gene expression is complex and multifactorial, involving cAMP-dependent and -independent mechanisms required for maintenance of optimal steroidogenic capacity, tissue-specific mechanisms which lead to different steroidogenic pathways in different tissues, and developmental mechanisms which lead to fetal imprinting of steroid hydroxylase expression and which probably overlap with both maintenance and tissue-specific mechanisms. Future studies will involve identification of the trans-acting factors associated with each of these aspects of the multifactorial regulation and characterization of the cis-regulatory elements to which they bind. Such studies will inevitably lead to the identification of genes encoding these trans-acting factors and investigation of their regulation. In this way, it will be possible to work outward from the steroid hydroxylase genes toward the cell surface receptors in order to elucidate the series of events which lead to cAMP-dependent and -independent regulation of steroid hydroxylase gene expression.

Production of C21 steroids in rat fetal ovaries

Female gonads of fetal and neonatal rats were cultured for 24 h in medium 199 supplemented with dibutyryl cyclic AMP (dcAMP), cycloheximide, spironolactone (an inhibitor of 17 alpha-hydroxylase) or hydroxylated cholesterol derivatives. Radioimmunoassays of pregnenolone (P5) and progesterone (P4) were performed in the media. In control medium, progesterone production developed during the second week of life whereas ovaries treated with dcAMP (1 mM) produced progesterone as early as day 14.5 of gestation. The effect of the nucleotide was reversible and the lag period of responsiveness was 12 h. Cycloheximide (1 microgram/ml) added with dcAMP completely inhibited the production of progesterone. Addition of spironolactone to the basal medium was without effect but significantly increased P5 and P4 productions in the presence of dcAMP. The inhibitory effects of spironolactone and trilostane on steroidogenesis were tested in fetal testis. Lastly, the dcAMP effect could not be mimicked with hydroxylated cholesterol derivatives (20 microM). However, an additive effect of 22-hydroxycholesterol was observed in the presence of the nucleotide. These results indicate that the in vitro production of P5 and P4 by fetal ovaries in the presence of dcAMP suggests that dcAMP induces enzymes involved in C21 steroid synthesis. Levels of action of the nucleotide are discussed.

An improved method for primary culture of ovarian androgen-producing cells in serum-free medium: effect of lipoproteins, insulin, and insulinlike growth factor-I

Although luteinizing hormone (LH) alone stimulates ovarian interstitial cells cultured in serum-free medium to synthesize large amounts of androgens, there seem to be additional factors in vivo that modulate the time course and magnitude of the cellular responses to LH. In an attempt to develop a more nearly physiologic cell culture model, lipoproteins, insulin, and insulinlike growth factor-I (IGF-I) were added to the serum-free medium. The effects of these modifications on androgen biosynthesis by dispersed cells from ovaries of hypophysectomized immature rats cultured in 96-well tissue culture plates were examined. A saturating dose of LH stimulated a 25-fold increase in androsterone synthesis at 2 d, which decreased at 4 and 6 d. Addition of human high density (hHDL) or human low density lipoprotein (hLDL) caused a 2.5-fold increase in LH-stimulated androsterone synthesis. Cells were approximately twice as sensitive to hHDL (ED50 = 5.5 +/- 0.5 micrograms cholesterol/ml) compared to hLDL (ED50 = 9.1 +/- 1.1 micrograms cholesterol/ml). Surprisingly, rat HDL caused only a 40% increase in LH-stimulated androsterone synthesis. When insulin alone was added to cells cultured with a saturating dose of LH, there was a 2.8-fold increase in androsterone synthesis. Addition of hHDL and insulin together caused a synergistic increase in LH-stimulated androsterone synthesis. In contrast to hHDL, which did not change the time course of LH-stimulated androsterone production, insulin prolonged maximal LH-stimulated androsterone synthesis at 4 and 6 d. Inasmuch as the ED50 for insulin action (1.3 +/- 0.1 micrograms/ml) was supraphysiologic, the effects of IGF-I on LH-stimulated androgen synthesis were examined. IGF-I mimicked all of the effects of insulin, but at a physiologic concentration (ED50 = 2.5 +/- 0.3 ng/ml). Ovarian cells cultured in serum-free medium supplemented with hHDL and insulin or IGF-I exhibit responses that closely approximate the physiologic responses observed in vivo. These results suggest that lipoproteins and IGF-I are important physiologic stimulators of ovarian theca-interstitial cell androgen biosynthesis which, when added to the serum-free medium, make the cellular responses in this in vitro model more nearly approximate the responses in vivo.

Ferredoxin reductase levels in the ovaries of pigs and superovulated rats during follicular cell growth and luteinization

The concentration of ferredoxin reductase, a component of the mitochondrial steroidogenic electron transport chain, was measured in the ovaries of pigs and superovulated rats by a protein blotting procedure using polyclonal antibodies to the purified protein. The concentration of ferredoxin reductase in porcine granulosa cells doubled during growth of follicles from small (1-2 mm diameter) to large (6-12 mm diameter). The concentration doubled again during the period of luteinization. This is in contrast to the rate of cholesterol side-chain cleavage, which showed little change during follicular growth but increased by more than tenfold during luteinization. A similar large increase in cholesterol side-chain cleavage occurs during the period of luteinization in the ovaries of superovulated rats, but as for the pig, only a small increase in ferredoxin reductase was observed. A threefold increase in the yield of mitochondrial protein from tissue homogenates was found between the granulosa cells of small-medium follicles and the cells of the corpora luteum. The increase in ferredoxin reductase during follicular development and luteinization, therefore, correlates well with the increase in mitochondria in the cells, but does not correlate with the dramatic increase in cholesterol side-chain cleavage activity which occurs during luteinization. Based on these results, it is unlikely that the level of ferredoxin reductase limits the expression of the full steroidogenic activity of the granulosa cells of the ovary.

Stoichiometry of mitochondrial cytochromes P-450, adrenodoxin and adrenodoxin reductase in adrenal cortex and corpus luteum. Implications for membrane organization and gene regulation

We have estimated the concentrations of cytochromes P-450scc and P-45011 beta and the electron-transfer proteins adrenodoxin reductase and adrenodoxin in the adrenal cortex and corpus luteum using specific antibodies against these enzymes. While in the adrenal cortex the concentrations of these enzymes are relatively constant in different animals and show no significant sex differences, in corpora lutea they vary considerably and can increase at least up to fifty-fold over the levels found in the ovary. The average relative concentrations of adrenodoxin reductase, adrenodoxin and P-450 are 1:3:8 in the adrenal cortex (which has two cytochromes P-450, P-450scc and P-450(11) beta, in equal concentrations) and 1:2.5:3 in the corpus luteum (which has only P-450scc). We further present evidence that the levels of cytochrome c oxidase also show a degree of correlation with the levels of the mitochondrial steroidogenic enzymes.

Regulation of cholesterol side-chain cleavage cytochrome P-450 gene expression in adrenal cells in monolayer culture

Utilizing a cDNA probe specific for bovine cytochrome P-450scc and primary monolayer cultures of bovine adrenocortical cells and human fetal adrenal cells, it has been shown that the chronic action of ACTH on the adrenal cortex includes regulation of P-450scc gene expression at the transcriptional level. The bovine P-450scc cDNA hybridizes strongly to human, pig and rat RNA. Advantage was taken of the cross-reactivity of the bovine P-450scc cDNA with human P-450scc RNA to examine the regulation of P-450scc gene expression by ACTH in human fetal adrenal cells. This process is mediated by cyclic AMP and is inhibited by cycloheximide, in a fashion similar to bovine adrenocortical cells, suggestive that a protein factor(s) activates the response in both species. Hence, the actions of ACTH to regulate P-450scc gene expression in bovine adult adrenocortical cells and human fetal adrenal cells appear to proceed by similar mechanisms.

Regulation of the biosynthesis of cytochromes P-450 involved in steroid hormone synthesis

The actions of ACTH to regulate the synthesis of the various enzymes involved in steroid hormone biosynthesis have been studied using bovine adrenocortical cells in monolayer culture. ACTH causes an increase in the synthesis of both the mitochondrial and the microsomal forms of cytochrome P-450 involved in steroid hormone biosynthesis, as well as of the iron-sulfur protein involved in transferring electrons to the mitochondrial forms of cytochrome P-450, namely, adrenodoxin. This increased synthesis is reflective of an increase in translatability of mRNA species specific for these various proteins, and appears in each case to be mediated by cyclic AMP. Whereas the mitochondrial proteins are synthesized as precursors of higher molecular weight which are processed upon insertion into the mitochondria, the microsomal proteins are synthesized as species identical in molecular weight to the mature forms. In order to determine whether the action of ACTH to increase the rate of synthesis of these proteins is the result of an increase in the levels of specific mRNA species, cDNA clones complementary to these mRNA species are being isolated. These probes will also make it possible to characterize the genes encoding the steroidogenic enzymes, as well as to identify regulatory elements which control their transcription.