Characterization of Testosterone 11 -hydroxylation Catalyzed by Human Liver Microsomal Cytochromes P450

A combination of accelerator mass spectrometry (AMS) and liquid chromatography-tandem mass spectrometry has been used to clarify some new aspects of testosterone metabolism. The main pathway of testosterone oxidative metabolism by human liver microsomes is the formation of 1 -, 2 -/ -, 6 -, 15 -, and 16 hydroxytestosterones, mainly catalyzed by cytochromes P450 2C9, 2C19, and 3A4. We now report the first determination that 11 hydroxytestosterone (11 -OHT) can also be formed by human liver microsomal fractions. The structures of five hydroxylated metabolites of testosterone (2 -, 6 -, 11 -, 15 -, and 16 -OHT) and the C-17 oxidative metabolite androstenedione were determined by liquid chromatography with UV detection at 240 nm and liquid chromatography-tandem mass spectrometry. Corresponding results were obtained by high-performance liquid chromatographyAMS analysis of incubations of [4-C]testosterone with human liver microsomes. 6 -Hydroxylation was always the dominant metabolic pathway, but 2 -, 15 -, and 16 -OHT, and androstenedione were also formed. The previously undetected hydroxytestosterone, 11 -OHT, was found to be a minor metabolite formed by human liver microsomal enzymes. It was formed more readily by CYP3A4 than by either CYP2C9 or CYP2C19. 11 -Hydroxylation was inhibited by ketoconazole (IC50 30 nM) at concentrations similar to the IC50 (36 nM) for 6 -hydroxylation Therefore, CYP3A4 could be mainly responsible for testosterone 11 -hydroxylation in the human liver. These findings identify human hepatic biotransformation of testosterone to 11 -OHT as a previously unrecognized extra-adrenal metabolic pathway. Cytochrome P450 (P450) enzymes, a superfamily of more than 160 known members, play a major role in the metabolism of numerous physiological substrates, and liver microsomal P450 enzymes are responsible for the biosynthesis or catabolism of steroid hormones. Hydroxylation of active androgens is usually associated with a decrease in biological certainty (Hobkirk, 1979). Endogenous and exogenous testosterone undergoes oxidative metabolism by the human liver cytochrome P450 enzymes (Wood et al., 1988; Yamazaki and Shimada, 1997; Rendic et al., 1999). -Hydroxylation at either the C6 or C16 position is the major route of testosterone oxidative metabolism, whereas 1 -, 2 / -, and 15 -hydroxytestosterone are produced as minor metabolites. Human liver enzymes are also found to oxidize testosterone at the C17 position to form androstenedione (Fig. 1). Various P450 enzymes form hydroxylated steroids highly stereospecifically and regioselectively (Wood et al., 1983). An alteration of the steroid hydroxylation metabolism in liver microsomes therefore indicates altered expression of various P450 enzymes. Liver tumor microsomes, whether from chemically induced or spontaneous tumors, showed a radically modified steroid hydroxylation pattern (Feuer et al., 1986; Lange et al., 1989). The 11 -hydroxylase is the most thoroughly studied mammalian enzyme involved in steroid hormone biosynthesis. The observation that 11-deoxycorticosterone and 11-deoxycortisol are hydroxylated at the 11 -carbon position by the enzyme in the mitochondrial fraction of the adrenal suggests that a single reactive site on the enzyme is involved in both cases (Sharma et al., 1962). Biochemical abnormalities are completely linked to the congenital adrenal hyperplasia that causes altered excretion of the tetrahydro-metabolites of these compounds (White et al., 1994; Choi et al., 2002). 11 -Hydroxylation was confirmed in human adrenal homogenate as the major metabolic pathway of testosterone (Chang et al., 1963) as well as of certain neutral and phenolic steroids (Knuppen and Breuer, 1962; Hudson and Killinger, 1972). On incubating testosterone with a testicular tumor, 11 -hydroxytestosterone (11 OHT) also was observed (Savard et al., 1960). Although the liver is responsible for most of the metabolism of cholesterol and the steroid hormones, to our knowledge, no metabolic study showing the roles of 11 -hydroxylase on the endogenous steroid metabolism in the human liver has been published. Whether testosterone may undergo hydroxylation at the C11 position by human liver microsomal P450s has apparently been overlooked, perhaps due in part to techniques that were less specific and less sensitive than those now available. In many studies (van der Hoeven, 1984; Smith et al., 1992; Tachibana and Tanaka, 2001), testosterone metabolites were assayed by HPLC with UV detection This study was supported by the National Cancer Institute, Pfizer, and the Massachusetts Institute of Technology Center for Environmental Health Sciences (National Institute of Environmental Health Sciences Grant P30-ES02109). These data were presented in part at the American Chemical Society National Meeting, September 7–11, 2003, New York, NY. Article, publication date, and citation information can be found at http://dmd.aspetjournals.org. doi:10.1124/dmd.104.003327. ABBREVIATIONS: OHT, hydroxytestosterone; LC, liquid chromatography; LC-UV, high-performance liquid chromatography with UV detection; HPLC, high-performance liquid chromatography; AMS, accelerator mass spectrometry; CID, collision-induced dissociation; KTZ, ketoconazole. 0090-9556/05/3306-714–718$20.00 DRUG METABOLISM AND DISPOSITION Vol. 33, No. 6 Copyright © 2005 by The American Society for Pharmacology and Experimental Therapeutics 3327/3032631 DMD 33:714–718, 2005 Printed in U.S.A. 714 at A PE T Jornals on A uust 4, 2017 dm d.aspurnals.org D ow nladed from (LC-UV) and/or gas chromatography-mass spectrometry or LC-mass spectrometry in which 11 -OHT, which may have been present only at trace levels, might not have been detected. In the present study, we have re-examined hepatic microsomal metabolism of testosterone, using C labeling to ensure detection of all metabolites with equal sensitivity and LC-tandem mass spectrometry for structural identification. Individuals with congenital adrenal hyperplasia suffer from hypertension due to a complete deficiency of adrenal 11 -hydroxylase (Eberlein and Bongiovanni, 1956). In most classic cases of this deficiency, there is overproduction of steroid precursors proximal to the blocked 11 -hydroxylase, leading to pathological effects (Zachmann et al., 1983). In this paper, we demonstrate that the human liver is capable of generating 11 -oxygenated steroids from precursors such as testosterone. This could potentially have implications for either diagnosis or therapeutic treatment of adrenal insufficiency diseases. Materials and Methods Chemicals and Reagents. Testosterone and its hydroxylated metabolites, as well as androstenedione and 17 -methyltestosterone, were purchased from Steraloids (Newport, RI). [4-C]Testosterone (specific activity, 50 mCi/ mmol) was obtained from American Radiolabeled Chemicals (St. Louis, MO). Pooled human liver microsomes and selectively expressed human cytochrome P450 enzymes (CYP2C9, CYP2C19, and CYP3A4) were purchased from BD Gentest (Woburn, MA). Liver microsome protein content was 20 mg/ml in 250 mM sucrose. Recombinant P450 isoforms were expressed in insect cells selectively transfected with a baculovirus expression system containing the cDNA for human CYP2C9 (total protein concentration, 4 mg/ml; P450 content, 500 pmol/mg), CYP2C19 (3 mg/ml, 360 pmol/mg), and CYP3A4 (5 mg/ml, 400 pmol/mg). The catalytic activities for CYP2C9 (according to diclofenac 4 -hydroxylase), CYP2C19 [(S)-mephenytoin 4 hydroxylase), and CYP3A4 (testosterone 6 -hydroxylation) were 8000, 900, and 2200 pmol/mg P450/min, respectively. Single human liver samples were obtained from organ donors or patients undergoing liver resection (Tennessee Donor Service, Nashville, TN). Each liver was tested for pathogenicity using a polymerase chain reaction protocol. Liver samples (0.3 g) were homogenized in 10 ml of ice-cold 10 mM potassium phosphate buffer (pH 7.4) using a Polytron homogenizer. The cell debris, nuclei, and mitochondria were removed by centrifugation at 10,000g for 20 min at 4°C, and then the supernatant was ultracentrifuged at 120,000g for 2 h at 4°C. The pellet was resuspended in ice-cold phosphate buffer (pH 7.4) containing 1 mM EDTA and 20% glycerol (v/v; Guengerich, 1994). Aliquots of microsomal fractions were quickly stored at 80°C until used. Protein concentrations of the microsomal fractions were measured in triplicate, with bovine serum albumin as a calibration standard, using commercial protein assay kits (Bio-Rad, Hercules, CA). Cytochrome P450 was estimated by absorbance differences between 450 and 490 nm based on the method of Omura and Sato (1964). The NADPH-regenerating system used for all NADPH-requiring oxidase assays was obtained from BD Gentest. The system consists of two solutions (solution A, 26.1 mM NADP , 66 mM glucose 6-phosphate, and 66 mM MgCl2 in H2O; solution B, 40 U/ml glucose-6-phosphate dehydrogenase in 5 mM sodium acetate). A stock solution of steroids including 17 methyltestosterone was prepared at concentrations of 1 mol/ml in methanol. The stock solution was used to prepare a working solution of varying concentrations (1–100 mol/ml) in methanol. Microsomal Assays. A 0.3-ml reaction mixture containing 0.33 mg/ml human liver microsomal protein or 120 pmol/ml human CYP2C9, CYP2C19, or CYP3A4, 50 l of NADPH-regenerating solution A, 20 l of solution B, and 0.2 mol of testosterone in 100 mM potassium phosphate (pH 7.4) was incubated at 37°C for 20 min. After incubation, the reaction was terminated by the addition of 0.5 ml of acetonitrile, and 50 l of methyltestosterone (1 mM) was added. The mixture was then centrifuged (12,000 rpm) for 10 min to precipitate protein. The supernatant, filtered with a VariSep Nylon Syringe Filter (0.2 m 4 mm; Varian, Inc., Palo Alto, CA), was dried under a stream of nitrogen and then reconstituted with 50 l of 50% methanol. Recovery Test and Calculations. Solutions containing testosterone and its six metabolites at varied concentrations were prepared for method validation. All quantitative calculations were based on the peak area ratios relative to that of the internal standard. To prepare the calibration curve, enzyme-free incubation solutions with increasing concentrations of added analyte (0.1–2 mM testosterone and 0.2–10 M for all metabolites) and a fixed concentration of added 17 -methyltestosterone (50 l; 1 mM) were analyzed according to the procedure described above. The recovery of the procedure was measured by comparing the responses obtained from the extracted samples to those obtained from the corresponding unextracted reference standards prepared at the same concentrations. Instrumentation and Chromatography. LC analyses were carried out with an Agilent 1100 binary pumping system (Agilent Technologies, Palo Alto, CA), using a 9725 Rheodyne injector (Rheodyne, Rohnert Park, CA) and a G1314A variable wavelength UV detector set at 240 nm. An aliquot of the sample (2 l) from each incubation was injected onto a Capcell Pak C18 UG120 column (2.0 150 mm, 5m particle size; Shiseido Fine Chemicals, Tokyo, Japan) and eluted at a flow rate of 0.2 ml/min by gradients of solvent A (20 mM ammonium acetate in 10% methanol) and solvent B (90% methaFIG. 1. Oxidative metabolism of testosterone by human liver microsomal cytochrome P450 enzymes. 715 TESTOSTERONE 11 -HYDROXYLATION BY HUMAN LIVER at A PE T Jornals on A uust 4, 2017 dm d.aspurnals.org D ow nladed from nol). The typical gradient (solvent B) was as follows: 0 min, 10%; 0 to 10 min, 10%; 10 to 20 min, 30%; 20 to 30 min, 55%; 30 to 38 min, 55%; 38 to 45 min, 100%; 45 to 50 min, 100%. The gradient was then returned to the initial condition (10% B) and held for 8 min before analysis of the next sample. On-line mass spectrometric analyses were performed using an ion trap (Agilent MSD-Trap XTC) equipped with an electrospray ionization interface. When there was insufficient material for on-line mass spectral analysis, multiple fractions were collected, pooled, and concentrated prior to off-line analysis using flow injection. Collision-induced dissociation (CID) spectra were collected on positive ions using the auto-CID function of the instrument. Nitrogen gas from a liquid-nitrogen dewar boil-off was used as nebulizer (12 psi) and curtain gas (12 l/min). Helium was the collision gas. The heated nebulizer probe temperature and voltage were maintained at 350°C and 5000 V, respectively. AMS analyses were conducted by the Biological Engineering Accelerator Mass Spectrometry Lab at Massachusetts Institute of Technology. The AMS instrument has previously been described in detail (Liberman et al., 2004a). The interface used to generate CO2 from liquid samples has also been described in detail (Liberman et al., 2004b), and an example of its application to the analysis of HPLC effluent in continuous fashion has been presented (Skipper et al., 2004).