V. Luu-The, Oncology and Molecular Endocrinology Research Center, Laval University Medical Center (CHUL), 2705 Laurier Boulevard, Québec (QC) G1V 4G2, Canada. Fax: + 1 418 654 2761, Tel.: + 1 418 654 2296, E-mail: email@example.com
Recently, we have shown that the biosynthesis of androstenol, a potential endogenous ligand for the orphan receptors constitutive androstane receptor and pregnane-X-receptor, requires the presence of enzymes of the steroidogenic pathway, such as 3β-hydroxysteroid dehydrogenase, 5α-reductase and 3α-hydroxysteroid dehydrogenase. In this report, we examine at the molecular level whether the enzyme 17α-hydroxylase/17,20-lyase (P450c17), which possesses dual 17α-hydroxylase and 17,20-lyase activities and catalyzes the production of precursors for glucocorticoids and sex steroids, is also able to catalyze the formation of a third class of active steroids, 16-ene steroids (including androstenol). The role of components of the P450 complex is also assessed. We transfected human embryonic kidney (HEK-293) cells with various amounts of vectors expressing P450c17, NADPH-cytochrome P450 reductase, and cytochrome b5. Our results showed that P450c17 possesses a 16-ene-synthase activity able to transform pregnenolone into 5,16-androstadien-3β-ol, without the formation of the precursor 17-hydroxypregnenolone. Cytochrome b5 has a much stronger effect on the 16-ene-synthase activity than on the 17α-hydroxylase/17,20-lyase activities. On the other hand, P450reductase has a drastic effect on the latter, but a negligible one on 5,16-androstadien-3β-ol synthesis. Our results therefore demonstrate that human P450c17, as other enzymes of the classical steroidogenic pathway, is involved in the biosynthetic pathway leading to the formation of androstenol.
It has been established that human cytochrome P450c17 (product of the CYP17 gene) has two distinct activities responsible for the synthesis of glucocorticoid and sex steroid precursors from pregnenolone (preg). A 17α-hydroxylase activity, which converts preg into 17α-OHpreg, is necessary for cortisol synthesis, and a 17,20-lyase activity further transforms 17α-OHpreg into dehydroepiandrosterone (DHEA), the precursor of sex steroids (Fig. 1). Several different studies have revealed that these two activities are differentially modulated by many factors, two of the most important being the abundance of the redox partner cytochrome P450reductase (P450red) [1,2] and the interaction with cytochrome b5 (cyt b5), an allosteric effector [3–6].
As for 16-androstenes, the precise mechanism by which they are biosynthesized has been until now subject for debate. Early studies reported testosterone to be a precursor for these steroids [7–12]. Later, however, testosterone and a large number of other compounds, including epitestosterone, DHEA, 16α-hydroxypregnenolone and 16α-hydroxyprogesterone, were excluded as precursors for 16-androstenes, whereas preg and progesterone were found to be putative precursors. Multiple pathways have been suggested for the transformation of C21-steroids into 16-unsaturated C19-steroids in porcine testicular homogenates. These included 20β-reduction (preg → pregnenediol → 5,16-androstadien-3β-ol) , 21-hydroxylation (preg → 21-OHpreg → 5,16-androstadien-3β-ol) [10,13] and 16–17-dehydrogenation (preg → 17α-OHpreg → 16-dehydro-preg → 5,16-androstadien-3β-ol) [7,14]. A concerted process (preg → 5,16-androstadien-3β-ol) has also been suggested. Finally, results published by Weusten et al. provided evidence that androstadienol was synthesized from preg in a single step by a 16-ene-synthase enzyme system in human testicular homogenates . However, the molecular mechanism responsible for this biosynthesis remains to be elucidated.
It is well recognized that 16-androstenes are produced by Leydig cells of porcine testis  and that these steroids have pheromonal activity in pigs. In humans, the physiological role of 16-androstene steroids is still ill-defined. It has been proposed that these compounds may have significant effects on behavior, namely reducing nervousness, tension and other negative emotional states in women . Another study demonstrated a positive relationship between menstrual synchrony and the ability to smell certain 16-androstene steroids . Recent reports in the literature show that androstenol (5α-16-androsten-3α-ol) could modulate the activity of two orphan receptors, the recently renamed CAR (constitutive androstane receptor) , previously known as constitutively active receptor , and PXR (pregnane-X-receptor) . It has been suggested that androstenol is an endogenous ligand for these receptors , which share a common hetero-dimerization partner, RXR (retinoid-X-receptor), and are subject to cross talk interactions with other nuclear receptors and with a broad range of other intracellular signaling pathways [21,22].
The purpose of this research is to examine, at the molecular level, whether the human P450c17 overexpressed in HEK-293 cells possesses 16-ene-synthase activity and how it differs from the 17α-hydroxylase and 17,20-lyase activities. We also compare the human P450c17 with its porcine counterpart.
Construction of P450red, cyt b5 and P450c17 expression vectors
The cDNA fragments containing the entire coding regions of human NADPH-cytochrome P450reductase (P450red, EC 126.96.36.199) [23,24] and cyt b5 were isolated as previously described . The cDNAs were then subcloned into a pCMV expression vector. Porcine P450c17 cDNA was amplified by PCR using Taq DNA polymerase (Perkin-Elmer Cetus, Emerville, CA, USA)  and an oligo-primer pair (5′-GGGGTCGACATGTGGGTGCTCTTGGTTTTCTTCTTG-3′ and 5′-GGGGTCGACTCAGGAGGTACTCCCCTCAGTGTGGGC-3′) and poly(A)+ RNA isolated from pig testis. The cDNA was then subcloned into a pCMV expression vector. The cDNA coding for human P450c17 (EC 188.8.131.52) was kindly provided by Y. Tremblay (CHUL Research Center, Quebec, Canada).
Transient expression in transformed human embryonic kidney (HEK-293) cells
Vectors expressing P450c17 (pCMV-P450c17), P450red (pCMV-P450red) and cyt b5 (pCMV-cyt b5) were transfected into HEK-293 cells using the Ex-gene kit according to the manufacturer's instructions (MBI Fermentas, Amherst, NY, USA). Cells were initially plated at 5 × 105 cells per well in six-well falcon flasks and grown in Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY, USA) supplemented with 10% (v/v) fetal bovine serum (Hyclone, Logan, UT, USA) at 37 °C under a 95% air, 5% CO2 humidified atmosphere.
Assay of enzymatic activity
Determination of the activities was performed in intact cells transiently transfected with P450c17 and/or P450red and/ or cyt b5 as previously described . Briefly, [3H]preg, [3H]17α-OHpreg or [3H]DHEA was added to freshly changed culture medium in six-well culture plates. For enzymatic assays performed with intact cells in culture, we have previously established 16 h to be an appropriate incubation time period, as the activity vs. time graph still shows linearity. After 16 h of incubation, the steroids were extracted twice with 2 mL of ether. The organic phases were pooled and evaporated to dryness. The steroids were solubilized in 50 µL of dichloromethane, applied to a Silica Gel 60 TLC plate (Merck, Darmstadt, Germany) before separation by migration in the toluene/acetone (4 : 1, v/v) solvent system. Substrates and metabolites were identified by comparison with reference steroids, revealed and quantified using phosphoimaging, Storm 860 (Molecular Dynamics Inc., Sunnyvale, CA, USA). Nonlabeled reference steroids were revealed with a solution of molybdate/sulfuric acid (10 : 10, v/v).
Analysis by HPLC
3H-Labelled steroids were analyzed using Waters Nova-Pak reverse-phase C18 HPLC column (3.9 × 150 mm, 4 µm). The mobile phase was MeOH/H2O (80 : 20, v/v) with 2 mm ammonium acetate and 0.1% acetic acid, with a flow rate of 1 mL·min−1. Radioactivity was monitored in the eluent using Beckman 171 HPLC Radioactivity Monitoring System. Nonlabeled androstadienol and preg standards were monitored using UV at 216 nm.
Identification of metabolites by HPLC analysis and corecrystallization
To verify the nature of the metabolites obtained from the transformation of preg by human and porcine P450c17, we identified by HPLC analysis, extracts of HEK-293 cells transfected with P450c17 and cyt b5. 3H-Labeled preg, 17α-OHpreg, and DHEA, used as standards, showed elution peaks at 4.70, 2.30 and 2.50 min, respectively (Fig. 2). In both porcine and human assays using preg as a substrate, an additional peak of elution appeared at 15 min (panels C and D, respectively). This additional peak coincides with the elution time of nonlabeled commercial androstadienol monitored using UV at 216 nm. This data shows that one of the metabolites obtained in assays using human and porcine P450c17 is androstadienol. In addition to comigratory behavior on both HPLC and TLC analyses, the identity of the radiolabeled androstadienol product was confirmed by cocrystallization with commercial steroid (data not shown).
Assessment of the 16-ene-synthase, 17α-hydroxylase and 17,20-lyase activities of human and porcine P450c17
In order to produce DHEA from preg, P450c17 first transforms preg into 17α-OHpreg through its 17α-hydroxylase activity and then transforms this intermediate into DHEA through its 17,20-lyase activity. In order to determine whether the transformation of preg into androstadienol requires prior synthesis of 17α-OHpreg or DHEA, we performed enzymatic assays using human and porcine P450c17 in the presence of various substrates – preg, 17α-OHpreg and DHEA – and analyzed androstadienol formation from each substrate. As observed in Fig. 3, the biosynthesis of androstadienol in humans (A) and pigs (B) does not require prior formation of 17α-OH-preg and DHEA. The lack of androstadienol synthesis in the presence of 1 µm of ketoconazole (C), an inhibitor of cytochrome P450, further demonstrates the specific implication of P450c17 in the formation of this metabolite.
Formation of androstadienol by human and porcine P450c17, with and without cyt b5
Using porcine and human P450c17 expressed in HEK-293 cells in culture, we compared the formation of androstadienol from preg in the pig and the human. As illustrated in Fig. 4, both human and porcine enzymes have the ability to produce androstadienol in presence of cyt b5. When exogenous cyt b5 is omitted from the transfection assays, both human and porcine P450c17 poorly catalyze the formation of androstadienol from preg (less than 2% of preg transformation). Porcine P450c17 shows a slightly stronger stimulation by cyt b5, its activity increasing to 15% of preg transformation while the activity of human P450c17 increases to 12%. These results show that human and porcine P450c17 have similar catalytic activities and that, in both species, P450c17 is involved in the biosynthesis of androstadienol.
Effect of cyt b5 on DHEA and androstadienol biosynthesis
In order to determine the effect of cyt b5 in P450c17 16-ene-synthase and 17α-hydroxylase/17,20-lyase activities, we performed transfection assays with increasing amounts of DNA fragments encoding cyt b5 and monitored the formation of androstadienol and DHEA. As shown in Fig. 5 the stimulation of DHEA and androstadienol production from preg increases with increasing amounts of cyt b5 in presence of endogenous levels of P450red. In presence of these low levels of P450red, cyt b5 shows a slight stimulatory effect on DHEA formation. However, we observe a more profound effect on the synthesis of androstadienol. More precisely, an increase of androstadienol formation was observed at a cyt b5/P450c17 ratio of 5 : 1 (Fig. 5). The activity reached a maximum at a ratio of 12 : 1. Thus, the influence of human cyt b5 changes dramatically as the cyt b5/P450c17 ratio varies.
Effect of P450red on DHEA and androstadienol synthesis
To further investigate the modulation of human P450c17 16-ene-synthase activity, we proceeded to analyze the relative effect of P450red on 17α-hydroxylase/17,20-lyase and 16-ene-synthase activities. To do so, we cotransfected P450c17 and cyt b5 in amounts determined to be optimal for androstadienol formation along with increasing amounts of P450red. It can clearly be seen in Fig. 6 that the addition of P450red, even in small amounts, has a profound effect on 17α-hydroxylase/17,20-lyase activities. In presence of only endogenous P450red levels, DHEA formation from preg is below 10%. Increasing P450red up to 0.25 µg causes a drastic increase of DHEA formation, reaching levels up to 50% of preg transformation. On the other hand, increasing amounts of P450red do not significantly stimulate 16-ene-synthase activity. For this activity, endogenous levels of P450red seem to be sufficient for optimal cyt b5 stimulation as increasing amounts of P450red do not further stimulate androstadienol production. These results clearly show a differential modulation of 17α-hydroxylase/17,20-lyase and 16-ene-synthase activities by P450red and cyt b5.
It is already known that human cytochrome P450c17 possesses two distinct activities, a 17α-hydroxylase and a 17,20-lyase activity, responsible for the biosynthesis of glucocorticoid and sex steroid precursors. In this report we show that human and porcine P450c17 also possess a 16-ene-synthase activity that catalyzes the transformation of preg into androstadienol, a precursor in the biosynthesis of androstenol (Fig. 1). This reaction differs from the production of DHEA through the 17α-hydroxylase/17,20-lyase activity, in that it does not require the 17α-OH-preg formation step and it is strongly stimulated by optimal amounts of cyt b5. Indeed, results obtained from assays using 17α-hydroxypregnenolone as a substrate demonstrate that the synthesis of androstadienol diverges from the biosynthetic pathway of sex steroids at the level of preg transformation and that it does not involve the 17α-hydroxylase activity. Furthermore, time course experiments (data not shown) using preg as a substrate do not show the production of any intermediates in the formation of androstadienol which suggests that it is synthesized from pregnenolone in a single step. Therefore, although the synthesis of glucocorticoid/sex steroid precursors and 16-ene-steroids result from the same enzyme, the activities responsible for their formation and the regulation of these activities are distinct. Gower et al. had previously demonstrated, using porcine testis microsomes, the formation of 16-androstenes from 17α-hydroxypregnenolone . However, this situation is not observed in the intact transfected cell system using porcine or human P450c17. In combination, these results suggest the presence, in pig testis, of another system that can use 17α-hydroxypregnenolone to form a different product.
P450c17 is therefore a crucial enzyme, not only in the formation of sex steroid precursors, but also in the production of androstadienol which is considered to be an intermediate in the 16-androstene pathway leading to the biosynthesis of androstenol and of 5α(16)androsten-3-one, a pheromonally active steroid in the pig [8,14]. A recent report by our group  shows that further transformation of androstadienol into androstenol involves the classic enzymes of the steroidogenic pathway, namely 3β-HSD, 5α-reductase and 3α-HSD. Interestingly, because of the lack of either a 17-keto or a 17-hydroxy group, this pathway does not require the enzyme 17β-HSD which is specific to sex steroid biosynthesis.
Although it is well known that androstenol is a pheromone in the pig, its role in the human is still ill-defined. Recent findings show that androstenol is able to modulate the expression of certain cytochrome P450s and alcohol dehydrogenases through interactions with the orphan receptors CAR and PXR [18,20]. These receptors belong to the P450-regulatory nuclear receptors, in the subfamily NR1 (nuclear binding site 1) . Other members of this NR1 orphan nuclear receptor gene subfamily are PPAR (peroxisome proliferator-activated receptor), LXR (liver X receptor) and FXR (farnesol-X-receptor). They share a common hetero-dimerization partner, the RXR, and are subject to cross-talk interactions with other nuclear receptors and with a broad range of other intracellular signaling pathways, including those activated by certain cytokines and growth factors [21,22]. It has been shown that the steroids androstenol [16(5α)-androsten-3α-ol] and 5β-pregnanedione (5β-pregnane-3,20-dione) modulate the action of these receptors and thus are putative endogenous ligands for these receptors . Upon binding to the ligands, PXR and CAR bind DNA as a heterodimer with the RXR and modulate the expression of cytochromes P450, especially CYP2B and CYP3A families. Because P450s play an essential role in the detoxification of drugs and of a large series of exogenous compounds from the environment, androstenol and 5β-pregnanedione that modulate cytochrome P450 levels could have a profound effect on the detoxification process.
Because of the low affinity of these orphan nuclear receptors (in the range of 1–10 × 10−6m) and their relatively broad spectrum of ligand specificity, many researchers that are familiar with classic steroid receptors, namely androgen, estrogen, progesterone, glucocorticoid and mineralocorticoid receptors, that bind to their corresponding specific ligand with a very high affinity (10−10−10−9m), are skeptical about the idea of androstenol and 5β-pregnanedione being ligands for these orphan receptors. However, it is noteworthy that, as classic active steroids are diluted in the blood, their concentration is very low and thus they require high affinity receptors to pick them up. On the other hand, the orphan receptor ligands are most probably produced locally in the various tissues. Because of the small volume of the cell, the production of a little amount of ligand will give a relatively high concentration (up to 1–10 × 10−6m). We hypothesize that this low affinity combined with the local biosynthesis of ligands represents a mechanism allowing the selective regulation of the action of the receptor: ligands that enter the cell or tissue by chance will not have a high enough concentration to turn on the receptor. Only for ligands that are produced locally or accumulated in the tissue (probably through active transport or hydrophobicity) is the concentration high enough to modulate the receptor activity.
As suggested above, local biosynthesis in various tissues such as the liver, could constitute a way to selectively regulate the activity of nuclear orphan receptors such as CAR and PXR. Although there is no evidence of P450c17 expression in the human liver, many other enzymes such as 3β-HSD, 5α-reductase and 3α-HSD, whose activities lead to androstenol synthesis from androstadienol, are present in the liver and many other peripheral tissues. Furthermore, androstadienol is found in circulation suggesting that it is synthesized at the sites of expression of P450c17 and eventually converted to androstenol by different enzymes in peripheral tissues such as the liver and adipose tissue. Our study aimed at elucidating the nature and the mechanism of the reactions involved in the local formation of androstenol is thus of major importance.
This work has been supported by a grant from the Canadian Institutes of Health Research. The authors would like to thank Guy Reimnitz, Nathalie Paquet and Mei Wang for their technical assistance and Sylvie Méthot for careful reading of the manuscript.