| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Reproductive Biology (A.H.P.), Department of Obstetrics and Gynecology, Stanford University School of Medicine, Stanford, California 94305; and Department of Physiology and Biophysics (D.B.H.), University of Illinois at Chicago, Chicago, Illinois 60612
Correspondence: Address all correspondence and requests for reprints to: Anita H. Payne, Division of Reproductive Biology, Department of Obstetrics and Gynecology, Stanford University School of Medicine, 300 Pasteur Drive, Stanford, California 94305-5317. E-mail: anita.payne{at}stanford.edu
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Cytochrome P450sIII. Hydroxysteroid...IV. Epilogue: Steroidogenic...References |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Cytochrome P450sIII. Hydroxysteroid...IV. Epilogue: Steroidogenic...References |
|---|
illustrates all of the enzymes involved in the biosynthesis of the adrenal steroid hormones, corticosterone, cortisol, and aldosterone; and the gonadal steroid hormones, progesterone, estradiol, and testosterone. These enzymes fall into two major classes of proteins: the cytochrome P450 heme-containing proteins (Table 1) and the hydroxysteroid dehydrogenases (Table 2). Tables 1 and 2 list the human and mouse genes that encode the enzymes described in this review, the gene symbol for each, the chromosomal location of each gene, the nomenclature and synonyms found in the literature for the proteins, the molecular mass and major tissue sites of expression of each enzyme, as well as the nucleotide and protein accession numbers.
|
|
|
An important discovery in the regulation of tissue-specific expression of gonadal and adrenal steroidogenic enzymes was reported in 1992 by two laboratories. These laboratories identified a transcription factor that is required for the tissue-specific expression of P450 steroidogenic enzymes as well as the gonadal and adrenal-specific isoform of 3ß-HSD in gonads and adrenal glands, but not in placenta (18). This transcription factor, a member of the steroid hormone receptor superfamily, was named steroidogenic factor-1 (SF-1) by Lala et al. (19) and adrenal 4-binding protein (Ad4BP) by Morohashi et al. (20). The role of SF-1 and the role of other factors involved in the regulation of expression of the steroidogenic enzymes in gonads, adrenal glands, and the placenta are discussed in the sections titled "Regulation of Expression" under each specific class of enzymes.
|
II. Cytochrome P450s |
|---|
|
TopAbstractI. IntroductionII. Cytochrome P450sIII. Hydroxysteroid...IV. Epilogue: Steroidogenic...References |
|---|
.
The P450 enzymes involved in steroid hormone biosynthesis are membrane-bound proteins associated with either the mitochondrial membranes CYP11A, CYP11B1, and CYP11B2, or the endoplasmic reticulum (microsomal) CYP17, CYP19, and CYP21. These P450 enzymes are members of a superfamily of heme-containing proteins found in bacteria, fungi, plants, and animals (21). They derive their name from the characteristic that, when complexed in vitro with exogenous CO, they absorb light maximally at 450 nm. In the biosynthesis of steroid hormones from cholesterol, cytochrome P450 enzymes catalyze the hydroxylation and cleavage of the steroid substrate. They function as monooxygenases utilizing reduced nicotinamide adenine dinucleotide phosphate (NADPH) as the electron donor for the reduction of molecular oxygen. The general reaction is
 |
In this reaction, the oxygen is activated by P450, and one oxygen atom is introduced into the substrate RH as a hydroxyl group, whereas the other atom of oxygen is reduced to H2O. The electrons from NADPH are transferred to the substrate by two distinct electron transfer systems. The mitochondrial transfer involves transfer of the high potential electron to a flavoprotein, adrenodoxin reductase (ferredoxin reductase); and then sequentially to adrenodoxin (ferredoxin), a nonheme iron-sulfur protein; to P450; and finally to the substrate (Fig. 2A). The microsomal electron transfer system involves only one protein, cytochrome P450 oxidoreductase, a protein that contains two flavins. The electrons are transferred from NADPH to a flavinadenine dinucleotide, followed sequentially by transfer to flavinmononucleotide, P450, and the substrate (Fig. 2B).
|
). The reaction requires three molecules of oxygen, three molecules of NADPH, and the mitochondrial electron transfer system described above. CYP11A catalyzes three sequential oxidation reactions of cholesterol with each reaction requiring one molecule of O2 and one molecule of NADPH (Fig. 3). The first reaction is hydroxylation at C22, followed by hydroxylation at C20 to yield 20,22R-hydroxycholesterol that is cleaved between C22 and C20 to yield the C21 steroid pregnenolone and isocaproaldehyde (22, 23). Isocaproaldehyde is then oxidized to isocaproic acid (24). Studies on purified proteins as well as studies with recombinant proteins from CYP11A cDNAs have provided conclusive evidence that a single protein catalyzes all three reactions at a single active site (25, 26). The electrons required for the reaction are transferred from NADPH to ferredoxin reductase, to ferredoxin, and finally to CYP11A (27). A model of the interactions between CYP11A and the electron transport proteins has been proposed based on the expression of mutants. The results of these studies indicate that the acidic residues, Asp 74 and Asp 76, of ferredoxin interact with the basic residues of ferredoxin reductase and CYP11A (28).
|
The structure of the cholesterol side-chain cleavage gene, CYP11A (Table 1), has been determined in human (33) and rat (36). The gene is at least 20 kb and consists of nine exons containing an unusual exon/intron junctional sequence that begins with GC in the sixth intron of both human (33) and rat genes (36). The human gene is located on chromosome 15q23-q24 (30), and the mouse gene is found on chromosome 9 at 31 cM (37).
3. Tissue- and developmental-specific expression.
The major sites of expression of CYP11A are in the adrenal cortex, ovary, testis, and placenta. In addition, CYP11A has been detected in the central and peripheral nervous system (38) and in human and rodent heart (15, 16). In the adrenal cortex, CYP11A is expressed in all three zones, the zona fasciculata, the zona reticularis, and the zona glomerulosa (39, 40). In the ovary, CYP11A is expressed in the theca interna and the granulosa cells of ovulatory follicles, but not in small antral follicles (36). The only site of expression in the testis is the Leydig cell (41). In the primate placenta, the synciotrophoblast cell is the site of expression of CYP11A (42), whereas in rodent placenta, CYP11A is expressed in giant trophoblast cells during midpregnancy (32, 43, 44). Before the expression of CYP11A in giant trophoblast cells during midpregnancy, expression of the CYP11A protein was demonstrated in decidual mitochondria at embryonic day (E) 6.5 (32, 43, 44).
During embryogenesis, the earliest expression of CYP11A mRNA in mouse was observed at E10.5 in the urogenital ridge, with significant expression observed in fetal testis by E12.5 (45). CYP11A mRNA expression in fetal testis continues throughout pregnancy with little if any expression observed in the fetal ovary (45, 46). Studies on the expression of steroidogenic enzyme mRNA in human fetal adrenal glands and gonads between 12 and 26 wk found that CYP11A was most abundant in the adrenal gland between 20 and 21 wk, followed by testis, with only minor expression in fetal ovaries (47). Testicular expression showed a decrease from a high at about 15 wk to significantly lower levels by 26 wk (47). Postnatally, in both rodents (48) and humans (49), testicular expression of CYP11A decreases due to the disappearance of the fetal Leydig cell population. As development of the adult population of Leydig cells occurs [in rodents after postnatal day 10 (48, 50) and in humans at the beginning of pubertal development (49)], there is a sharp increase in the expression of CYP11A reaching adult levels in mouse by postnatal day 25 (50). In mouse adrenal primordium, CYP11A expression was observed as early as E11 (45) and in rat fetal adrenal glands at E12 (51). Studies in primate fetal adrenal glands showed that human adrenal glands expressed CYP11A between 14 and 22 wk of gestation only in the fetal zone (FZ) and transitional zone (TZ) of the adrenal cortex, not in definitive zone (DZ) cells, and became detectable in the DZ after 23 wk (52, 53). In late gestation, monkey adrenal expression of CYP11A was observed in all three zones (53). Postnatally, adrenal CYP11A expression is essential for life (54, 55).
B. CYP17
1. Reaction catalyzed.
CYP17 (P450c17) catalyzes two mixed-function oxidase reactions utilizing cytochrome P450 oxidoreductase and the microsomal electron transfer system (Fig. 2B). The two reactions catalyzed by P450c17 are the 17
-hydroxylation of the C21 steroids, pregnenolone (
5 steroid) or progesterone (4 steroid), followed by the cleavage of the C17–20 bond to produce the C19 steroids, dehydroepiandrosterone (DHEA) or androstenedione, respectively (Fig. 4). Each reaction requires one molecule of NADPH and one molecule of molecular O2. In this two-step reaction, 17-hydroxypregnenolone or 17-hydroxyprogesterone is formed as an intermediate. Initially, it was believed that each reaction was carried out by distinct enzymes. However, studies by Hall and colleagues (56, 57) showed that a single purified protein catalyzed both 17-hydroxylation and C17–20 cleavage (lyase activity). Subsequent cloning of bovine P450c17 cDNA and expression of this cDNA confirmed that both reactions are catalyzed by a single protein (58). Although the CYP17 enzyme from various species catalyzes both the hydroxylation and the lyase reaction, there are marked species-dependent differences in the utilization of either 17-hydroxypregnenolone (5) or 17-hydroxyprogesterone (4) as substrate for the lyase activity. The major species-dependent differences have been observed in C17–C20 lyase activity. The human and bovine enzymes prefer 17-hydroxypregnenolone as the substrate yielding DHEA as the product compared with the rodent enzyme that utilizes 17-hydroxyprogesterone as the substrate yielding androstenedione as the product (57, 58, 59, 60, 61). These species-dependent differences in substrate preference are not related to differences in the aa sequence of the human, bovine, or rat enzyme, but are a property of the human and bovine enzyme for the requirement of the accessory protein cytochrome b5 in promoting the lyase activity of 17-hydroxypregnenolone but not of 17-hydroxyprogesterone (61, 62). Cytochrome b5 is an accessory protein that acts as an allosteric effector of the CYP17-oxidoreductase complex, thus increasing the Vmax of the lyase activity (62). Rat CYP17 lyase activity with 17-hydroxyprogesterone as substrate is also stimulated by b5, but the fold increase is small relative to the increase observed with the human enzyme (61).
|
). Comparison of the mouse aa sequence to rat and human aa sequences shows that they are 83 and 66% identical, respectively. The CYP17 protein of different species contains regions of high homology common to members of the P450 gene family (63). These are the putative binding regions for mouse aa 434–454 (63), human aa 435–455 (73), and the ozols tridecapeptide sequence (343–372 aa) (74) that may play a role in substrate specificity (75). Furthermore, there is a region specifically conserved among different species of CYP17 (296–319 aa) that may function in catalysis (76). Arginine346 in the rat enzyme (77) and arginine347 in the human enzyme (78) were found to be critical for catalyzing the lyase activity.
3. Tissue- and developmental-specific expression.
CYP17 is expressed in all the classic steroidogenic tissues; however, there are species-related differences in the expression of this enzyme in the adrenal gland and placenta. In testis of all species, CYP17 is expressed solely in the Leydig cell (40, 79, 80). In ovary, CYP17 expression is restricted to thecal cells that are the site of androgen production (81, 82, 83, 84). The general consensus is that granulosa cells and luteal cells do not express CYP17 (81, 82, 83, 84, 85). A recent report, however, suggests that human luteinized granulosa cells in culture do express CYP17 (86). In the adrenal glands of human, bovine, macaque, and guinea pig, CYP17 is expressed in the zona reticularis and the zona fasciculata (87, 88, 89) but not in the zona glomerulosa, which is the site of aldosterone synthesis (90, 91). CYP17 is not expressed in mouse (92) or rat adrenal glands (40, 61). The adrenals of these animals produce corticosterone rather than cortisol that occurs in human and other species. CYP17 is not expressed in human placenta. The C19 substrate for placental androgens and estrogens is derived from fetal adrenal glands as DHEA sulfate that is converted to 16-hydroxydehydroepiandrosterone sulfate in the fetal liver and transported to the placenta, where it is acted on by placental steroid sulfatase (93). In contrast, CYP17 is expressed in mouse and rat placenta starting at midpregnancy and declining just before parturition (32, 94).
Expression of CYP17 has been studied in human (52, 53) and monkey fetal adrenal glands (53). CYP17 mRNA and protein have been detected in the TZ and FZ throughout gestation, but not in the DZ. This differs somewhat from observations of the expression of CYP11A, which showed expression in the DZ after 23 wk gestation (52). Fetal monkey adrenal glands from late gestation monkey fetuses exhibited a similar pattern of CYP17 in the FZ and TZ as observed in human fetal adrenal glands. CYP17 was not detected in the DZ of the monkey adrenals (53). Keeney et al. (95) reported detection of CYP17 mRNA in mouse fetal adrenal cells between E12.5 and E14.5, after which expression disappears and, as reported earlier, CYP17 is not expressed postnatally in the adrenal glands of mice (92) or rats (94, 95). The expression of CYP17 mRNA in fetal mouse gonads follows the same expression pattern as observed for CYP11A (46). CYP17 is expressed in mouse fetal testes from E13, the earliest time examined, and throughout pregnancy. No expression has been observed in mouse fetal ovary until the day of birth (46). Expression of CYP17 in mouse testes postnatally is low between birth and 20 d, rising between 20 and 25 d, and reaching maximum expression after postnatal d 40 (50).
C. CYP19
1. Reaction catalyzed.
CYP19 (P450arom) catalyzes the conversion of the C19 androgens, androstenedione and testosterone, to the C18 estrogens, estrone and estradiol, respectively.
The reaction involves the microsomal electron transfer system, cytochrome P450 reductase, and three molecules each of oxygen and NADPH. The first two oxygen molecules are involved in the oxidation of the C19 methyl group by standard hydroxylation reactions, whereas the third oxygen molecule is used in a reaction postulated to be a peroxidative attack on the C19 methyl group combined with elimination of the 1ß hydrogen to yield a phenolic A ring and formic acid (Fig. 5) (96, 97).
|
The deduced aa sequence of human CYP19 compared with rat and mouse shows 81% homology (96). Both the human and the mouse protein consist of 503 aa having a molecular mass of 58 kDa (96). The CYP19 proteins of different species contain the same structural features described for the other cytochrome P450 enzymes: the heme-binding region containing a conserved cysteine residue that serves as the fifth coordinating ligand of the heme iron, and the substrate binding site in the amino-terminal I-helix region.
3. Tissue- and developmental-specific expression.
The expression of CYP19 is widely distributed. The major sites in humans (12) and rats (102) are in the preovulatory follicle, the corpus luteum of ovulatory humans, and the rat corpus luteum during the second half of pregnancy (102). In human pregnancy, the placenta is the major site of CYP19 expression (103). Rat and mouse placenta do not express CYP19 (103). Other nonprimates such as bovine, porcine, and equine do express CYP19 in the placenta (103). Testicular expression of CYP19 has been known for many years. In an early report, the testicular expression of aromatase was demonstrated by measuring the secretion of estradiol by human, simian, canine, and rat testes (104, 105). The testicular site of CYP19 expression in the testis has been studied by several investigators. Studies on separated seminiferous tubules and intact testicular tissue from human testes indicated that the major site of aromatization in human testes is in the interstitial tissue (106). Immunohistochemical studies in human testes showed that aromatase was detected in Leydig cells and absent in Sertoli cells in normal adult testes (107). Studies investigating the testicular site of aromatization of testosterone to estradiol using rat testes or isolated testicular cells demonstrated that CYP19 was expressed in Leydig cells from 25-d-old and 60- to 70-d-old rats and, furthermore, that human chorionic gonadotropin (hCG) can acutely stimulate Leydig cell aromatization of testosterone to estradiol (108, 109, 110). Aromatization was demonstrated in Leydig cells from 15-d-old immature rats (110). Studies by Dorrington and Khan (111) examining Sertoli cells from immature rat testes maintained in culture provided evidence for aromatase activity, with the highest expression in Sertoli cells from 5-d-old rats with decreasing expression with age becoming essentially undetectable in rats aged 30 d and older. Taken together, the above-described studies indicate that there is a shift in the site of testicular somatic cell expression of CYP19 from Sertoli cells in neonatal rats to the Leydig cells of pubertal and mature rats. More recently, it has been established that aromatase is expressed in germ cells from the pachytene spermatocyte stage of rats and mice (reviewed in Ref.112). Indirect evidence for aromatase expression in human spermatozoa has been obtained from the measurements of estradiol levels in ejaculates (113).
Expression of CYP19 mRNA during fetal development has been studied in humans and mice. Toda et al. (114), studying the expression of CYP19 mRNA in various human fetal tissues, detected high expression in fetal liver with lower expression in fetal intestines and minute or no expression in skin, lung, kidney, adrenal, or brain. CYP19 mRNA is undetectable in adult human liver (114). Harada and Yamada (115), studying the expression of CYP19 mRNA in brain tissues of male and female mice during fetal and postnatal development, detected expression in total brain as early as 12 d gestational age, with rapidly increasing expression reaching maximal levels 4 d postnatally followed by a continuous decrease to adult levels. Expression of CYP19 mRNA in fetal gonads of mice was first detected in fetal testes at d 17 postcoital and in both male and female gonads 1 d postnatally (46). For more detailed recent reviews on tissue-specific expression and regulation of CYP19, see Refs.12 and 103 .
D. CYP21
1. Reaction catalyzed.
CYP21 (P450c21) catalyzes the hydroxylation of C21 of progesterone and 17-hydroxyprogesterone yielding 11-deoxycorticosterone and 11-deoxycortisol, respectively. This reaction occurs in the endoplasmic reticulum and requires one molecule of NADPH and one molecule of oxygen and the microsomal electron transfer system (Fig. 6). Progesterone is the substrate for CYP21 in the adrenal zona glomerulosa that does not express CYP17. C21 hydroxylation is the catalytic step that determines the biosynthesis of adrenal steroid hormones (Fig. 1). 11-Deoxycorticosterone cannot serve as a substrate for hydroxylation at C17 by CYP17, thus 11-deoxycorticosterone cannot be an intermediate in the biosynthesis of cortisol.
|
) (123).
3. Tissue- and developmental-specific expression.
CYP21 is exclusively expressed in the adrenal cortex and is essential for the synthesis of adrenal-specific steroids, the glucocorticoids, cortisol and corticosterone, and the mineralocorticoid, aldosterone (124, 125). It is expressed in all three zones of the adrenal cortex—the zona reticularis, zona fasciculata, and zona glomerulosa. Expression has not been detected in kidney, liver, testis, or ovary (118). The fetal primate adrenal gland differs in structure from the structure of the postnatal gland. It consists of three functional zones: 1) the outer DZ that does not appear to be involved in adrenal steroidogenesis; 2) the TZ, located between the inner FZ and the DZ; and 3) the FZ that expresses the enzymes required for DHEA synthesis, and thus, is responsible for supplying the substrate for placental estrogen synthesis. Coulter and Jaffe (126) examined the expression of the CYP21 protein by immunohistochemistry in human fetal adrenal glands between 13 and 24 wk gestation and in pregnant rhesus monkeys between 109 and 125, 134 and 156, and 159 and 172 d of pregnancy. Very few immunoreactive cells were detected in human adrenal DZ cells between 13 and 24 wk; however, expression was observed in all cells of the TZ and FZ. In the fetal adrenal glands of rhesus monkeys between 109 d and term, all three zones expressed the CYP21 protein, but expression was markedly reduced in the FZ relative to the DZ and TZ. Expression of CYP21 during human fetal development differs from the expression observed in the adult adrenal gland. In the adult adrenal gland, intense expression was observed in the zona glomerulosa and zona fasciculata, with little or no expression in the zona reticularis (126).
In a more recent study, Narasaka et al. (52) examined the expression of steroidogenic enzymes by histochemical analysis of human fetal adrenal glands between 14 wk gestation and term. They report expression of CYP21 in the FZ and TZ between 14 wk and term and in the DZ starting at 24 wk and continuing to term.
E. CYP11B1 and B2
1. Reaction catalyzed.
CYP11B1 and CYP11B2 (P450c11b1 and P450c11b2) are located in the inner mitochondrial membrane. CYP11B1 catalyzes the 11ß-hydroxylation of 11-deoxycorticosterone or 11-deoxycortisol yielding corticosterone or cortisol, respectively (Fig. 7). CYP11B1 also has the capacity to hydroxylate C18 of 11-deoxycorticosterone or corticosterone to form 18-hydroxycorticosterone; however, it cannot catalyze the oxidation of the 18-hydroxy group to form aldosterone. Aldosterone synthesis from 11-deoxycorticosterone is catalyzed by CYP11B2, more commonly referred to as aldosterone synthase, which catalyzes three sequential reactions, each utilizing one molecule of NADPH and one molecule of oxygen and the mitochondrial electron transfer system. The three sequential reactions are: the 11ß-hydroxylation of 11-deoxycorticosterone, the hydroxylation of carbon 18, followed by oxidation of the carbon 18 hydroxyl group to yield the carbon 18 aldehyde group resulting in the formation of aldosterone (Fig. 8). It is believed that the conversion of 11-deoxycorticosterone to aldosterone by CYP11B2 occurs without the release of the intermediates in a manner similar to the cleavage of cholesterol in the biosynthesis of pregnenolone (127). It also was reported that CYP11B2 does not utilize corticosterone efficiently as a substrate for aldosterone synthesis, thus providing additional evidence that the biosynthesis of aldosterone occurs with 11-deoxycorticosterone as substrate and the catalysis is solely mediated by CYP11B2 and does not involve the sequential action of CYP11B1 followed by CYP11B2 (128). In rat, three CYP11B enzymes have been characterized, CYP11B1, B2, and B3 (129). The reactions catalyzed by B1 and B2 are the same as described above for the human and mouse enzymes. B3 can convert deoxycorticosterone to 18-hydroxycorticosterone, but lacks 18 oxidase activity and thus cannot synthesize aldosterone (129).
|
|
The aa sequences of the human CYP11B1 and B2 are 93% identical (127). The proteins are located in the inner mitochondrial membrane and thus are synthesized including a leader sequence of 24 aa that is cleaved in the mitochondria to yield a protein of 479 aa in human and 476 aa in mouse. Although CYP11B1 and B2 consist of the same number of aa when separated by SDS-PAGE, the apparent molecular mass of the human enzymes was reported as 51 and 49 kDa, respectively (127), and the rat enzymes as 51.5 and 49.5 kDa, respectively (134).
3. Tissue- and developmental-specific expression.
Expression of CYP11B1 and B2 is limited to the adrenal cortex. B2 appears to be expressed exclusively in adrenal zona glomerulosa cells, whereas the major site of CYP11B1 is in the adrenal zona fasciculata/reticularis, with some expression of B1 also observed in mitochondria of the adrenal zona glomerulosa (10, 134). Expression of CYP11B1 appears to be severalfold greater than expression of B2 in both human and rat adrenal glands (10, 135). In contrast, Domalik et al. (11), examining mouse adrenal cortex sections by in situ hybridization, found similar expression of the two CYP11B isoforms, with CYP11B1 expressed exclusively in the zona fasciculata/reticularis and CYP11B2 expressed exclusively in the zona glomerulosa. Rat CYP11B3 mRNA is expressed in the zona fasciculata/reticularis and not in the zona glomerulosa (129). However, developmental expression of rat CYP11B3 differs from that of CYP11B1 as described below (129).
Developmental expression of immunoreactive CYP11B1 and B2 in primate fetal adrenal glands has been studied by Coulter and Jaffe (126). The study did not differentiate between the expression of the B1 and B2 proteins due to the unavailability of an antibody that can distinguish between these two enzymes. Coulter and Jaffe examined human fetal adrenal glands between 13 and 24 wk gestation and fetal adrenal glands from rhesus monkeys from 109 d gestation to term. CYP11B1/B2 in the human fetal adrenal glands was absent in the DZ, but present in the TZ and FZ with higher expression in the FZ than the TZ. In the fetal rhesus monkey between 109 d and term, the B1/B2 protein was detected in all cells of the TZ and FZ, but absent in the DZ until near term. Metapyrone-induced ACTH stimulation in the rhesus monkey increased expression of the enzymes in the TZ and FZ and induced expression in the DZ (126). Developmental expression of CYP11B1, B2, and B3 mRNAs in rat adrenal glands between neonatal d 2 and 18 and from adult rats exhibit distinct expression patterns (129). Adrenal glands from neonatal d 2 exhibited a relatively large amount of B1 mRNA, with barely detectable B2 and no detectable B3 mRNA. At 10 d, B1 mRNA was markedly reduced, and B3 became the major CYP11B mRNA. This pattern of expression was found in adrenal glands until postnatal d 18 with no differences observed in expression in glands from male and female pups. In adrenal glands from mature rats, B3 expression is considerably less than B1.
F. Regulation of expression of the cytochrome P450 steroidogenic enzymes
A major nuclear factor that determines cell-specific expression of P450 steroidogenic enzymes in gonads and adrenal glands was identified in two laboratories in 1992. This nuclear DNA-binding protein, referred to as SF-1 by Lala et al. (19) or Ad4BP by Morohashi et al. (20), belongs to the orphan nuclear receptor family and binds to variants of an AGGTCA sequence motif found in the proximal promoter of all P450 steroidogenic enzymes (136). Although SF-1 is essential for cell-specific gonadal and adrenal expression, other factors are necessary for determining both maximal and cell-specific expression of these enzymes. Adrenal-specific expression of the CYP21 genes in both human (125) and mouse (137) is mediated by cell-specific elements located within the corresponding C4B and C4A genes. Three SF-1 sites have been identified within this distant promoter of the human CYP21 gene.
Chronic, but not acute, stimulation by pituitary peptide hormones (ACTH in the adrenal zona reticularis and zona fasciculata; LH in ovarian theca, corpus lutea, and testicular Leydig cells; and FSH in ovarian granulosa cells), acting via G protein-coupled receptors, activates adenylate cyclase thereby increasing cAMP, which in turn, leads to increased synthesis of the steroidogenic P450 enzymes specific for these cells [reviewed for bovine P450 enzymes by Waterman (138) and Waterman and Keeney (139)]. The regulation of hormonal stimulation via cAMP leading to increased synthesis of these enzymes, with some exceptions described below, is not mediated by the cAMP response element (CRE)/CRE binding protein (CREB) system. cAMP responsive sequences found in the promoters of steroidogenic CYP genes expressing these enzymes differ among genes and among the same genes of different species (69, 136, 139). The bovine CYP11B gene does contain a canonical CRE element in its promoter, which was shown to be required for cAMP-stimulated transcription (139, 140). In addition, it has been reported that cAMP acts via CRE/CREB in the rat CYP19 promoter (141) and in the PII human CYP19 promoter expressed in granulosa cells (142). Although, hormone-stimulated increases in cAMP enhance the expression of all of the steroidogenic P450 enzymes, additional factors are involved in maintaining maximal expression, with the exception of CYP17 whose expression appears to be entirely dependent on cAMP stimulation (138, 143).
CYP11B2 transcription, which is essential for aldosterone synthesis, is stimulated by angiotensin II, and by potassium ions that act by increasing the intracellular concentration of calcium ions (Ca2+) (144). Furthermore, it was reported that cAMP can independently increase CYP11B2 transcription and that intracellular Ca2+ and cAMP act via the same elements in the proximal promoter of the human CYP11B2 gene (144). Two elements located at –71/–64 and –129/–114 are required for both basal and maximal induction by either cAMP or Ca2+. The –71/–64 sequence resembles a consensus CRE/CREB element; the other element, –129/–114, binds SF-1 and the chicken ovalbumin upstream promoter transcription factor (144). Thus, increases in either cAMP or intracellular Ca2+ can independently increase transcription of the human CYP11B2 gene. These studies would indicate that the protein kinase C (PKC) pathway is not involved in mediating aldosterone-stimulated increases in CYP11B2 transcription. In a recent study, Bassett et al. (145) reported that although the CYP11B2 promoter contains a binding site for SF-1, SF-1 does not appear to be involved in the transcriptional regulation of CYP11B2.
Human placenta expresses two cytochrome P450 enzymes, CYP11A and CYP19. The placental-specific expression of these enzymes is dependent on factors distinct from those regulating expression in gonadal and adrenal cells. SF-1, which is essential for expression of these enzymes in adrenal and gonadal cells, is not present in human placenta (146). Identification of factors regulating the placental-specific expression of human CYP11A has long been sought. Hum et al. (147) identified a 55-kDa protein that appeared to act as a transacting factor for CYP11A expression in placental cells, but not in adrenal cells. However, this nuclear protein also was found in other nonsteroidogenic cells that do not express CYP11A (147). In a subsequent investigation, Huang and Miller (148) identified two transcription factors related to HIV-inducible LBP proteins, LBP-1B and LBP-9, that could modulate expression of human CYP11A in placenta but were not involved in determining its placental-specific expression. A recent study reported the isolation of a placental-specific nuclear protein that specifically binds to a conserved proximal region, SCC1, in the promoter of the human, rat, mouse, ovine, bovine, and porcine CYP11A gene (44). This placental nuclear protein turned out to be activating protein 2 (AP-2). The major AP-2 in human was AP-2 and in mouse giant trophoblast cells, AP-2
. Interestingly, the AP-2 binding sequence overlaps with the previously identified SF-1 sequence that determines the gonadal- and adrenal-specific expression of CYP11A. The authors demonstrated that coexpression of the rat promoter containing the SCC1 sequence with an AP-2 or AP-2 expression vector markedly increased transcriptional activity (44). It remains to be shown whether AP-2 factors are involved in the expression of CYP11A in human placenta and whether additional factors to AP-2 are required for determining the placental-specific expression of CYP11A.
Placental expression of CYP19 in human and nonhuman primates, ungulates, and rabbits is under the control of a placental-specific exon 1 (103). The placental-specific exon for human has been described above. CYP19 is not expressed in the placenta of either mouse or rat (12). Relatively little is known regarding the transcription factors that determine placental-specific expression of CYP19. Kamat et al. (149, 150) showed that elements within 501 bp of the proximal promoter flanking the placental-specific exon 1.1 determine placental-specific expression. From their studies, they conclude that this region contains both placental-specific positive elements and specific sequences that bind inhibitory transcription factors in nontrophoblast cells. Yamada et al. (151) identified a glia cell missing placental transcription factor that binds specifically to an element within 205 bp of the placental-specific promoter and may be involved in determining the placental-specific expression of CYP19.
|
III. Hydroxysteroid Dehydrogenases |
|---|
|
TopAbstractI. IntroductionII. Cytochrome P450sIII. Hydroxysteroid...IV. Epilogue: Steroidogenic...References |
|---|
A. 3ß-Hydroxysteroid dehydrogenase/isomerase
3ßHSD/isomerases are membrane-bound enzymes and are distributed to both mitochondrial and microsomal membranes depending on the type of cell in which they are expressed (see Section III.A.3). During the past decade, multiple isoforms of 3ßHSDs have been isolated and characterized in human (153), mouse (7, 154), and rat (8, 155, 156). The different isoforms are numbered in the order in which they were identified; therefore, the same numeral in different species does not indicate that they are orthologous (see Table 2 for human and mouse isoforms). The largest number of isoforms (six) was identified in mouse. These isoforms are highly homologous in their aa sequence, but fall into two distinct functional groups. Mouse 3ßHSD I, III, and VI function as classic dehydrogenase/isomerases (see Section III.A.2), whereas mouse 3ßHSD IV and V function exclusively as 3-ketosteroid reductases, are not involved in active steroid hormone biosynthesis (5, 6), and thus will not be discussed further in this review.
The enzymatic action of the 3ßHSD/isomerases are essential for the production of all active steroid hormones. Active steroid hormones are 4-3-ketosteroids, progesterone, testosterone, cortisol, or aldosterone, or are derived from a 4-3-ketosteroid, estradiol (Fig. 5). Two distinct isoforms have been identified in humans, human 3ßHSD I and II. Both of these isoforms function as steroid dehydrogenase/isomerases. In this review, human 3ßHSD I and II and mouse 3ßHSD I and VI will be described, because these are the sole isoforms involved in the biosynthesis of all active steroid hormones.
1. Reaction catalyzed.
Human 3ßHSD I and II and mouse 3ßHSD I and VI catalyze the conversion of the 5-3ß-hydroxysteroids, pregnenolone, 17-hydroxypregnenolone, and DHEA to the 4-3-ketosteroids, progesterone, 17-hydroxyprogesterone, and androstenedione, respectively. Two sequential reactions are involved in the conversion of the 5-3ß-hydroxysteroid to a 4-3-ketosteroid. The first reaction is the dehydrogenation of the 3ß-equatorial hydroxysteroid requiring the coenzyme NAD+ yielding a 5-3-keto intermediate and reduced NADH. The reduced coenzyme, NADH, then activates the isomerization of the 5-3-ketosteroid to yield the 4-3-ketosteroid (Fig. 9) (157, 158, 159). This reaction is catalyzed by a single dimeric protein without the release of the intermediate or coenzyme (159). Human and mouse enzymes also have the capacity to convert 5-androstane-3ß,17ß-diol to dihydrotestosterone in the presence of NAD+ (2, 3, 160). NADP+ was reported to be relatively ineffective as a cofactor for these reactions (160). Comparison of Km values for the expressed human and mouse isoforms demonstrated that Km values for mouse 3ßHSD VI were lower than for mouse 3ßHSD I with pregnenolone as a substrate (7). Similar results were reported for expressed human 3ßHSD I compared with expressed human 3ßHSD II (2, 161).
|
3. Tissue- and developmental-specific expression.
The isoforms of 3ßHSD are expressed in a cell- and tissue-specific manner. The expression of human 3ßHSD isoforms examined by RNase protection analysis demonstrated that human 3ßHSD I is expressed in placenta, skin, and breast tissue, whereas 3ßHSD II is expressed in adrenal gland, ovary, and testis (2, 153). A similar distribution of expression of the two orthologous mouse isoforms was demonstrated with mouse 3ßHSD I expressed in testis, ovary, and adrenal gland, and 3ßHSD VI observed in placenta and skin, as well as in testis (7). In situ hybridization in mouse testicular sections demonstrated the exclusive expression of 3ßHSD in the Leydig cells of the testis (166, 167). The expression of 3ßHSD VI mRNA and protein was demonstrated during early pregnancy in mouse (32). 3ßHSD mRNA and protein were observed in decidual cells after implantation until E7.5 and are observed in giant trophoblast cells at E8.5, reaching a maximum between E9.5 and E10.5 and thereafter decreasing to undetectable levels (32, 165, 167). Immunochemical localization of 3ßHSD was examined in sections of rat adrenal glands and gonads (40). 3ßHSD was seen in secretory cells of the zona reticularis and the zona glomerulosa. Immunolabeling was detected mostly in the endoplasmic reticulum, but also over the vesicular crista membranes of the mitochondria (40). As demonstrated in mouse testis, 3ßHSD was exclusively detected in the Leydig cell. According to studies by Pelletier et al. (40), immunostaining for 3ßHSD was found exclusively in the mitochondria of the Leydig cell. The intramitochondrial localization of 3ßHSD has been confirmed in bovine adrenal glomerulosa cells and rat adrenal fasciculata cells (168, 169, 170). In rat ovary, 3ßHSD was detected in thecal and granulosa cells of growing and preovulatory follicles with no detection observed in primary follicles. High expression was seen in interstitial and luteal cells. Immunoreactivity was detected mostly in the endoplasmic reticulum, with some labeling in the crista membranes of the mitochondria (40).
Expression of 3ßHSD during fetal and neonatal development has been examined most extensively in mouse testis. Expression of 3ßHSD in fetal gonads was first examined between E13 and postnatal d 1 (46). 3ßHSD I mRNA was detected in all fetal testicular samples, but not until postnatal d 1 in fetal ovaries. In subsequent studies on developmental changes in the testicular expression of 3ßHSD I and VI mRNA, Baker et al. (166) confirmed the testicular expression of 3ßHSD I from E13 to adulthood. A small amount of 3ßHSD VI was detected at E13, with no expression of 3ßHSD VI observed between E13 and postnatal d 10. Expression of 3ßHSD VI increased after postnatal d 10, becoming the predominant isoform by postnatal d 60 (166). The appearance of 3ßHSD VI in neonatal testes correlated with the appearance of the adult type Leydig cell (166). In studies reported by Payne and colleagues (7, 167), 3ßHSD I mRNA or protein always was found to be the predominant isoform in adult testes. In their investigations examining fetal human adrenal glands during midgestation by immunohistochemistry and in situ hybridization, Mesiano et al. (53) could not detect expression of 3ßHSD in any of the samples examined. However, expression was detected during late gestation in the monkey fetal adrenal gland. 3ßHSD was detected in the DZ with some expression in the TZ. Narasaka et al. (52) could not detect immunoreactive 3ßHSD in any cortical zones of human fetal adrenal glands before 22 wk gestation; however, after 23 wk gestation, expression was discernible in the TZ and DZ, but not in the FZ. The lack of expression of adrenal 3ßHSD in the human fetal adrenal zone explains the lack of synthesis of cortisol and aldosterone by human fetal adrenals until late during gestation. Immunohistochemical studies on the expression of 3ßHSD in human adrenal glands obtained between the ages of 7 months and 62 yr detected expression in all three adrenal zones between 7 months and 8 yr followed by a decrease in the zona reticularis with increasing age (171).
B. Regulation of expression of 3ßHSDs
The gonadal- and adrenal-specific expression of human 3ßHSD II and mouse 3ßHSD I also appears to be dependent on the SF-1 as described for the gonadal- and adrenal-specific expression of the P450 steroidogenic enzymes. Leers-Sucheta et al. (172) identified an SF-1 response element in the proximal promoter of the HSD3B2 gene that conferred adrenal-specific expression of the gene. In addition, these authors reported that this SF-1 response element also was found to be essential for mediating phorbal ester-induced transcriptional activity. An earlier study on the analysis of the mouse Hsd3b1 promoter identified three potential SF-1 consensus binding sites in the proximal promoter (162). In a subsequent study, it was shown that SF-1 also was required for the expression of mouse Hsd3b1 (165). In contrast, expression of human 3ßHSD I and mouse 3ßHSD VI is not dependent on SF-1 (165). These two isoforms are essential for the biosynthesis of progesterone in human placenta and in mouse giant trophoblast cells, respectively. As described above, SF-1 is not expressed in either human or mouse placenta (146, 173). However, two transcription factors have been identified that determine the specific expression of mouse 3ßHSD VI in giant trophoblast cells in midpregnancy. These two factors are AP-2, the same factor identified as that which determines trophoblast-specific expression of CYP11A, and another mouse trophoblast-specific transcription factor, Dlx 3 (165). These transcription factors are also expressed in human placenta. However, recent studies found that placental-specific expression of human 3ßHSD I is not mediated by either AP-2 or Dlx-3, but by transcription enhancer factor-5 and a GATA-like protein (174).
Tremblay and Beaudoin (175) reported that cAMP and phorbol myristate acetate increased the expression of 3ßHSD I mRNA in human JEG-3 cells and that the increase mediated via these two signaling pathways was additive. Subsequent studies with human 3ßHSD II promoter constructs in human adrenal cortical (H295R) cells demonstrated marked increases in transcriptional activity after treatment with phorbol ester and, as discussed above, maximal effect of phorbol ester was dependent on the presence of SF-1 (172). Additional studies on the regulation of 3ßHSD in the H295R cells demonstrated that treatment with the glucocorticoid receptor agonist, dexamethasone, increased the expression of 3ßHSD mRNA and that this effect was additive to the increase observed with phorbol ester treatment (176). This effect of dexamethasone on 3ßHSD transcriptional activity involved Stat5A and the Stat5A response element. Additional studies by the same investigators reported that epidermal growth factor and prolactin also increase the transcriptional activity of the human 3ßHSD II promoter and that the effect of each of these hormones involves the Stat5A response element in the type II promoter (177, 178). Investigations on the regulation of 3ßHSD expression in gonads have been limited mostly to studies in rodents. In vitro studies in cultured granulosa cells from immature rats showed that treatment with FSH increased rat 3ßHSD I mRNA, protein, and enzymatic activity (179). Furthermore, treatment with IGF-I enhanced the FSH-induced increase (179, 180). Eimerl and Orly (180) reported a high constitutive expression of 3ßHSD mRNA in granulosa cells that could be increased by treatment with FSH. In addition, they observed that IGF-I in the absence of FSH had the capacity to increase 3ßHSD mRNA that was distinct from their observations with the effect of IGF on increasing CYP11A mRNA that could only be observed in the presence of FSH. The role of gonadotropin and prolactin on ovarian 3ßHSD expression was investigated in hypophysectomized rats (181). These investigators found that in vivo treatment with hCG for 2, 3, and 9 d induced increases in 3ßHSD mRNA expression by 63, 145, and 146%, respectively, above the control levels in ovarian interstitial cells only. In contrast, treatment with prolactin resulted in a marked decrease in 3ßHSD mRNA in the corpus luteum.
In studies on the regulation of 3ßHSD mRNA in mouse Leydig cells in culture, it was shown that these cells exhibit high constitutive expression of 3ßHSD (182). Treatment with cAMP results in a marked decrease in the expression of 3ßHSD mRNA that could be reversed in the presence of aminoglutethimide, an inhibitor of cholesterol metabolism. It was demonstrated that testosterone produced during treatment of Leydig cells with cAMP repressed the cAMP induction of 3ßHSD mRNA (182). Furthermore, the glucocorticoid agonist, dexamethasone, negatively regulated 3ßHSD expression. Studies in rat Leydig cells in culture demonstrated that treatment with LH, forskolin, or cAMP increased 3ßHSD mRNA, protein, and activity after 24–72 h of treatment (183). Investigations using gonadotropin-deficient mice to study the role of LH/hCG in regulating the expression of 3ßHSD I and VI mRNA in the adult Leydig cell lineage showed that the expression of 3ßHSD I was independent of LH stimulation (184). In contrast, the expression of 3ßHSD VI mRNA was highly dependent on LH/hCG stimulation.
C. 17ß-Hydroxysteroid dehydrogenases
Like 3ßHSDs, the 17HSDs play essential roles in steroidogenesis. These enzymes catalyze the final step in the biosynthesis of active gonadal steroid hormones, estradiol, and testosterone, and unlike other steroidogenic enzymes described in this review, 17HSDs are not involved in the biosynthesis of adrenal steroids. The 17ßHSDs convert inactive 17-ketosteroids into their active 17ß-hydroxy forms. Although in in vitro studies these enzymes can function as either reductases or oxidases of the 17-keto group, in intact cells they function in a unidirectional manner (185). The 17HSDs are found as either membrane-bound or soluble enzymes (Table 2). To date, 11 different 17HSDs have been identified. So far, in human, nine different 17HSDs have been cloned, 1–5, 7, 8, 10, and 11 (9). Unlike the 3ßHSDs described above, there is very little homology among the different 17HSD isozymes. The 17HSDs differ in tissue distribution, catalytic preferences, substrate specificity, subcellular localization, and mechanisms of regulation. Among the many different forms of 17HSDs, three forms participate in the final step of biosynthesis of active steroid hormones in gonads, types 1, 3, and 7, and therefore will be the only types discussed in this review. For discussion of the other types of 17HSDs, see reviews in Refs.9 and 185, 186, 187 .
The recently introduced nomenclature for 17HSD is based on the genetic identity of the enzymes and their functionality. Furthermore, the 17HSDs were numbered chronologically as they were identified. Among species, orthologs are assigned the same number; for example, type 4 17HSDs in human, rodent, guinea pig, pig, or chicken are all orthologs of the same enzyme. The canonical gene name for 17HSDs is HSD17Bn (in human) and Hsd17bn (in other species) where "n" represents the enzyme type number (e.g., enzyme 17HSD type 1 corresponds to the HSD17B1 gene). However, several recently discovered 17HSDs initially were assigned other gene names because a different functionality was studied (9).
1. 17HSD1.
Type 1 17HSD was the first of the 17HSD/ketosteroid reductases to be cloned and characterized. 17HSD1 was first purified from human placenta by Jarabak et al. (188, 189). It was cloned originally from a human placental library and subsequently found to be expressed in ovary and the mammary gland (185).
a. Reaction catalyzed.
Human 17HSD1 has substrate specificity for estrogens, whereas the rodent enzyme can utilize both estrogens and androgens; NADPH is the preferred cofactor for conversion of estrone to estradiol (Fig. 10) (152, 186). Human 17HSD1 has a 100-fold higher affinity for C18 than for C19 substrates. Whereas human 17HSD1 predominantly catalyzes the conversion of estrone to estradiol, mouse and rat 17HSD1 also efficiently convert androstenedione to testosterone (190).
|
17HSD1 is a member of the short-chain alcohol dehydrogenase reductase (SDR) superfamily. The three-dimensional structure was solved in 1995 (196). The 17HSD1 protein exists as a homodimer, consisting of noncovalently bound but tightly associated subunits; stable dimerization of the enzyme is required for its enzymatic activity (194). The x-ray structure of the dimer indicates that there are two dimerization helices in each monomer forming a four-helix bundle (196). Disruption of the helices by site-directed mutagenesis of specific aa resul
