194
Bioconjugate Chem. 2002, 13, 194−199
Enzyme-Assisted Synthesis and Structure Characterization of Glucuronide Conjugates of Methyltestosterone (17r-methylandrost-4-en-17β-ol-3-one) and Nandrolone (estr-4-en-17β-ol-3-one) Metabolites Tiia Kuuranne,†,‡ Olli Aitio,‡ Mikko Vahermo,† Eivor Elovaara,§ and Risto Kostiainen*,†,‡ Viikki Drug Discovery Technology Center, FIN-00014 University of Helsinki, Finland, Division of Pharmaceutical Chemistry, Department of Pharmacy, P.O. Box 56, Viikinkaari 5E, FIN-00014 University of Helsinki, Finland, and Laboratory of Toxicokinetics and Metabolism, Department of Industrial Hygiene and Toxicology, Finnish Institute of Occupational Health, FIN-00250 Helsinki, Finland . Received March 15, 2001; Revised Manuscript Received October 5, 2001
A new and useful method based on enzyme-assisted synthesis was developed for producing 3R-O-βD-glucuronide conjugates from synthetic phase I metabolites of methyltestosterone and nandrolone. The formed glucuronide conjugates of 17R-methyl-5R-androstane-3R,17β-diol (I), 17R-methyl-5βandrostane-3R,17β-diol (II), 5R-estran-3R-ol-17-one (III), and 5β-estran-3R-ol-17-one (IV) are urinary metabolites, indicating the human misuse of the above-mentioned anabolic androgenic steroids (AAS). The common lack of reference material precludes the use and validation of these biomarkers in human doping control. Liver microsomes from Aroclor 1254-induced rats were used as a highly active source of mammalian UDP-glucuronosyltransferases (UGT, EC 2.4.1.17). After purification by protein precipitation, liquid-liquid extraction (dichloromethane), C-18 solid-phase extraction, and lyophilization, the steroid glucuronide structures were characterized by 1H and 13C NMR spectroscopy and tandem mass spectrometry. The enzymatic method was highly stereoselective, producing a single major conjugate from the parent steroids I-IV. The stereochemically pure steroid glucuronide conjugates were recovered in milligram amounts (1.0-2.8 mg, yield 12-29%), which is sufficient for veterinary and human doping control analyses; for pharmaco-, toxico-, and enzyme kinetic studies in the pharmaceutical industry; for clinical laboratories; and for forensic medicine. A new sensitive LC-MS method was developed for controlling the product purity in syntheses, as well as for enzyme kinetic characterization of AAS-metabolizing UGT activities in rat liver toward the aglycones I-IV. In this study, the UGT enzymes responsible for the formation of 3R-O-linked glucuronides from the substrates I, II, III, and IV exhibited the specific enzyme activity values: 25, 124, 48, and 212 nmol/mg microsomal protein in a 2-h incubation, respectively.
INTRODUCTION
Methyltestosterone (17R-methylandrost-4-en-17β-ol-3one) and nandrolone (estr-4-en-17β-ol-3-one) are exogenous anabolic androgenic steroids (AAS) and widely misused in human sports. Because of their nonpolar character, AAS are extensively metabolized by routes involving functionalization (phase I) and conjugation (phase II) reactions prior to excretion in urine. AAS molecules with a 3-keto-4-ene structure undergo enzymatic reduction of the C4-C5 double bond in the A-ring as a rate-limiting step, which is catalyzed by 5R- and 5βreductases (Figure 1). As soon as the double bond is reduced, the 3-keto-group is transformed mainly to a 3Rhydroxy structure (1, 2). Phase I reactions thus provide attachment sites for subsequent conjugation reactions mediated by phase II metabolism. For AAS, the main conjugation pathway is glucuronidation (3), an SN2 reaction between the donating coenzyme uridine-5′* Corresponding author, Tel.: +358-(0)9-191 59 134, Fax: +358-(0)9-59 556, e-mail:
[email protected]. † Viikki Drug Discovery Technology Center, University of Helsinki. ‡ Division of Pharmaceutical Chemistry, University of Helsinki. § Finnish Institute of Occupational Health.
Figure 1. Structures and nomenclature of AAS glucuronides. (a) 17R-methyl-5R-androstane-17β-ol-3R-O-(β)-glucuronide, (b) 17R-methyl-5β-androstane-17β-ol-3R-O-(β)-glucuronide, (c) 5Restran-17-one-3R-O-(β)-glucuronide, and (d) 5β-estran-17-one3R-O-(β)-glucuronide.
10.1021/bc010038g CCC: $22.00 © 2002 American Chemical Society Published on Web 02/06/2002
Enzyme-Assisted Synthesis of Steroid Glucuronides
Bioconjugate Chem., Vol. 13, No. 2, 2002 195
Table 1. Yields of the Syntheses and Kinetic Data of Steroid Glucuronides
bulk substratea
product (glucuronide)
I
5R-MTG 17R-methyl-5R-androstan17β-ol-3R-O-glucuronide 5β-MTG 17R-methyl-5β-androstan17β-ol-3R-O-glucuronide 5R-NG 5R-estran-17-one-3RO-glucuronide 5β-NG 5β-estran-17-one-3RO-glucuronide
II III IV
synthesisb
yield (mg) yield (%)
reaction velocity:c enzymic glucuronidation activity (nmol/mg protein per 2 h) at different substrate concentrations (µM) 1
10
50
100
250
1000
exogenous source [ref]
1.0
12
0.6
4.9
18
21
20
25
mestanolone, methyltestosterone, oxymetholone [1, 2, 39]
2.8
29
0.9
8.3
40
79
124
124
methandriol, methyltestosterone, metandienone [1, 2, 39]
1.3
16
0.7
3.7
18
30
45
48
nandrolone [1, 2, 39]
1.1
13
1.0
10.1
44
83
178
212
nandrolone [1, 2, 39]
a (I) 17R-methyl-5R-androstane-3R,17β-diol; (II) 17R-methyl-5β-androstane-3R,17β-diol; (III) 5R-estran-3R-ol-17-one; (IV)5β-estran-3Rol-17-one. b 1 mM substrate and 5 mM UDPGA concentrations used for the biosyntheses (duration 24 h). c Assayed with liver microsomes from Aroclor 1254 treated rats under conditions essentially the same as in bulk syntheses.
diphospho-R-D-glucuronic acid (UDPGA) and a nucleophilic steroid aglycone, yielding β-D-glucuronides through a configurational inversion of the R-glycosidic bond (4). Steroids possessing either a 3R- or 17β-hydroxyl group are predominantly metabolized by the UGT-mediated pathway (5), whereas steroids with the 3β-hydroxyl structure are mostly excreted as sulfates (6, 7). Studies with isoenzyme preparations of animal and human UGTs, obtained by protein purification or by cDNAexpression techniques, have revealed that members of both the UGT1 and UGT2 families carry out C18 and C19 steroid glucuronidation (3, 8-16). The development of a synthesis method for the production of AAS glucuronides is of great interest due to the current lack of commercial reference compounds, especially in the case of exogenous AAS. In addition to veterinary and human doping control analyses, synthetic steroid glucuronides are applicable as reference material for analyses in the pharmaceutical industry, clinical laboratories, and forensic medicine. Regarding recent advances in method development, the adoption of liquid chromatography-mass spectrometry (LC-MS) techniques provides an alternative pathway for direct and straightforward analysis of nonvolatile steroid conjugates, offering a rapid analysis system under the growing pressure of sample input. Reports of chemical syntheses of steroid glucuronide conjugates have been widely published (17-22). The problem with the classical synthesis is the possible formation of a racemic mixture of R/β-anomer and additional byproducts (18). Enzyme-assisted synthesis may offer an elegant way to produce stereochemically pure steroid glucuronides by virtue of the stereoselectivity encountered in nature for biochemical catalysts, i.e., UGT enzymes (3). Methods using rat liver preparations as a source of conjugating enzymes have been described for a structurally diverse group of substrates, e.g., p-nitrophenol (23); a series of nitrogen-containing antidepressants (24); nitrocatechols (25, 26); androsterone, androstanediol, dihydrotestosterone (27); epitestosterone (28); and testosterone (29-31). The purpose of the present study was to develop and optimize a synthesis method for enzyme-assisted production of O-glucuronide conjugates derived from four AAS metabolites relevant to human metabolism and the abuse of exogenous steroids. The steroid aglycones 17R-methyl5R-androstane-3R,17β-diol (I), 17R-methyl-5β-androstane3R,17β-diol (II), 5R-estran-3R-ol-17-one (III), and 5βestran-3R-ol-17-one (IV) each yielded a major product corresponding to the 3R-O-glucuronides referred to as 5R-
MTG, 5β-MTG, 5R-NG, and 5β-NG (Figure 1). The synthesized products were characterized by NMR and tandem mass spectrometry and used as reference material for the development of a sensitive LC-MS method suitable for our present and future UGT enzyme kinetic studies. EXPERIMENTAL SECTION
Chemicals. Steroid aglycones I, III, and IV (see Table 1) were purchased from Steraloids (Wilton, NH). Compound II was synthesized in our laboratory by the synthesis route of Shinohara (32), and the structure of the product was confirmed with the GC-EI/MS method against a standard reference of 17R-methyl-5β-androstane-3R,17β-diol. UDPGA was from Boehringer-Mannheim, and D-saccharic acid 1,4-lactone and androsterone glucuronide (internal standard) from Sigma Chemicals (St. Louis, MO). Aroclor 1254 (RCS-088/Analabs, Lot No K040) was from the Foxboro Company (North Haven, CT 06473). Solvents were of HPLC-grade. All other reagents were of analytical grade. Preparation of Liver Microsomes. Adult female Wistar rats from Helsinki University Breeding Centre (n ) 8, weight 224 ( 18 g) were killed on day 8 after administration of Aroclor 1254 (500 mg/4 mL olive oil/ kg body weight on day 1, and 125 mg/1 mL/kg on day 6, intraperitoneally). The livers were perfused in situ with saline before being removed and stored at -70 °C. Liver microsomes were prepared by differential centrifugation (33), and the protein concentration (37.6 mg/mL) was determined according to the method of Lowry et al. (34). The treatment of the animals was approved by the local Ethical Committee for Animal Studies. Bulk Synthesis. The reaction mixtures for enzymeassisted synthesis of steroid glucuronides consisted of 1 mM of the lipophilic steroid substrates I, II, III, or IV (dissolved in 2 mL of methanol); 5 mM MgCl2; 5 mM D-saccharic acid 1,4-lactone (a β-glucuronidase inhibitor); and 5 mM UDPGA in 50 mM K-Na-phosphate buffer (pH 7.4) and contained 20 mg of rat liver microsomal protein in a final volume of 20 mL. The concentration of methanol was limited to 10% to avoid protein denaturation. The enzyme reactions (started by UDPGA addition) were carried out at 37 °C (in a water bath) for 24 h in a glass vial with magnet stirring. The synthesis reaction was terminated with 2 mL of water-saturated dichloromethane (29), which was added for protein precipitation (15 min in an ice bath) and extraction of the residual steroid aglycone and nonpolar impurities from the synthesis mixture. A clear aqueous phase was separated by
196 Bioconjugate Chem., Vol. 13, No. 2, 2002
centrifugation (14000 rpm, 10 min). The supernatant obtained was acidified with 4 N perchloric acid to ensure the neutral form of steroid glucuronides for their retention and cleanup by solid-phase extraction (SPE) with a C-18 cartridge (100 mg, International Sorbet Technology). The column was conditioned with 2.5 mL of methanol and then acidified with 2.5 mL of 1 N HCl. After loading of the sample (5 mL per a cartridge), polar impurities were removed with 2.5 mL of water and the final elution was performed with 2.5 mL of methanol. Methanol was evaporated to dryness (N2, 60 °C), and the dry residue, which was reconstituted to 3 mL with water, was frozen and lyophilized. This step enabled the formation of white crystals of steroid glucuronides. Mass Spectrometric Characterization. The mass spectra were recorded by Perkin-Elmer Sciex API 300 (Sciex, Concord, Canada) triple quadrupole instrument equipped with electrospray ion source (ESI). Samples were introduced by loop injection, and both positive and negative ion MS and MS/MS spectra were measured. Operation parameters, as well as the MS and MS/MS behavior of the glucuronides studied, have been presented in detail earlier (35). LC-MS Procedure. The LC-MS method was developed for the detection of synthesis impurities and for the enzyme kinetic studies. Liquid chromatography consisted of Perkin-Elmer LC 200 Series micro pump and Series 200 autosampler, and Phenomenex Luna C-18 (150 × 1.0 mm, particle size 5 µm) micro column. The total eluent flow of 100 µL/min was split after the column by a ratio of 1:10 (ICP-04-20, LC-Packings). HPLC eluent A was 15 mM ammonium acetate/formic acid buffer (pH 4.2) in water:acetonitrile (9:1, v/v), and eluent B included the same buffer and acetonitrile but in a reversed ratio (1:9, v/v). The gradient was set as a linear increase of the eluent B content from 10% to 50% in 15 min, corresponding to an actual acetonitrile gradient from 18% to 50%. The column was equilibrated for 11 min after each analysis. The injection volume was 5 µL. NMR Spectroscopy. Steroid glucuronides were dissolved in 40 µL of acetone-d6. The NMR experiments were carried out on a Varian Unity 500 spectrometer at 23 °C using a gHX nano-NMR probe. For the DQFCOSY (36) experiment (16 scans per t1 value), a matrix of 2k × 256 points was collected and zero-filled to 2k × 512. A cosinebell weighting function was used in both dimensions. For the HMQC (37) and HMBC (38) spectra (32 and 64 scans per t1 value, respectively), matrixes of 2k × 256 and 2k × 128 points were recorded and zero-filled to 2k × 512 and 2k × 256 points, respectively. A cosine-bell weighing function was used. The average 1H-13C coupling constant was estimated to be 150 Hz and ∆2 was 60 ms. The 1H and 13C chemical shifts were referenced to residual acetone, 2.05 and 29.92 ppm, respectively. Enzyme Activity Studies. The kinetic data obtained for microsomal UGT activity toward the parent steroid aglycones I, II, III, and IV represent mean values of duplicate incubations performed and analyzed all on the same day. In enzyme activity assays the final incubation volume was 100 µL, the microsomal protein content 100 µg, and the test concentration of each substrate 1, 2.5, 5, 10, 25, 50, 100, 250, 500, and 1000 µM. Otherwise, the composition of the reaction mixture was the same as in bulk synthesis. The glucuronidation reaction (preincubated for 11min) was started with the addition of UDPGA, incubated for 2 h at 37 °C before being stopped with 10 µL of 4 N perchloric acid and by chilling the sample tubes for 15 min in an ice bath (protein precipitation). An internal standard, androsterone glucuronide (10
Kuuranne et al.
µM), was added prior to centrifugation (14000 rpm, 10 min). The supernatant was cleaned up by the same C-18 SPE procedure as in bulk synthesis, with the exception that the final elution was in 500 µL acetonitrile. The sample was evaporated to dryness (N2, 60 °C) and diluted in 100 µL of eluent A for LC-MS analysis. The four most concentrated samples of each series (substrate concentration 100-1000 µM) were additionally diluted 10-fold with eluent A to ensure that the measurement was within the dynamic range of ESI-MS. The steroid glucuronide formation was calculated by using the ratio of the peak area of ammonium adduct [M + NH4]+ of analyte (AA) and that of the internal standard (AISTD), i.e., v ) AA/ AISTD. RESULTS AND DISCUSSION
Bulk Synthesis. Figure 1 shows the steroidal 3R-Oβ-D-glucuronides (5R- and 5β-MTG, 5R- and 5β-NG) formed from 17R-methyl-5R-androstane-3R,17β-diol (I), 17R-methyl-5β-androstane-3R,17β-diol (II), 5R-estran-3Rol-17-one (III), and 5β-estran-3R-ol-17-one (IV) by means of enzyme-driven syntheses. The compounds I and 5RMTG can be detected in human urine after dosing with mestanolone, methyltestosterone, or oxymetholone, and the compounds II and 5β-MTG after methandriol, methyltestosterone, or metandione intake (39). The compounds III, IV, 5R-NG, and 5β-NG originate from nandrolone metabolism (39). Notably, we used liver microsomes from female rats pretreated with a mixture of polychlorinated biphenyls (Aroclor 1254) as the enzyme source. According to human excretion studies, the metabolites synthesized here (Figure 1 and Table 1) are, evidently, also formed in man (1, 2, 39). The method described in this work is suitable for producing milligram quantities of stereochemically pure steroid glucuronides at reasonable costs. This amount is enough for metabolic in vitro studies as well as for doping control analysis. The concentrations in these analyses are typically below 1 µg/ mL, and therefore standard material of 1 mg of steroid glucuronide is enough for more than 1000 analyses. The data in Table 1 summarize the results for bulk syntheses of the four steroid glucuronides illustrated in Figure 1. We also show the results from the short-term kinetic studies describing the reaction velocities of the present enzyme-driven syntheses. The UGT activities in rat liver microsomes were tested toward the parent steroids I - IV at 10 different substrate concentrations (1-1000 µM) using a reaction time of 2 h and a 5 mM UDPGA concentration. These findings suggested that at a 250-1000 µM substrate concentration the reaction velocity is maintained nearly at maximum for each substrate (I-IV). The efficiency of the rat enzyme source to glucuronidate different steroids exhibited an 8-fold difference. This was revealed by the specific enzyme activities varying from 25 to 212 nmol/mg protein per 2 h (Table 1). Moreover, the 5R-MTG and 5R-NG products were clearly formed at lower specific rates than the 5βMTG and 5β-NG anomers. Despite careful optimization of the initial reaction conditions (substrate, UDPGA, and microsomal protein), the yields of single syntheses remained lower than expected by the reaction velocities measured for the biosyntheses of 2 h (Table 1). A plausible explanation may lie in the uncertainty of maintaining the optimal reaction conditions throughout an incubation time extended over up to 24 h. Another obvious reason for low yields is the poor solubility of the substrates in aqueous incubation media. This problem could be only partly resolved by addition of methanol,
Enzyme-Assisted Synthesis of Steroid Glucuronides
Bioconjugate Chem., Vol. 13, No. 2, 2002 197
Table 2. Mass Spectrometric Data of Steroid Glucuronides; m/z (rel abundance)a ion
5R-MTG
5β-MTG
5R-NG
5β-NG
Positive Ion MS 505 (36) 505 (38) n.d.b 475 (43) 500 (100) 500 (100) 470 (100) 470 (100) Negative Ion MS [M - H]481 (100) 481 (100) 451 (100) 451 (100) Positive Ion MS/MS, Precursor [M + NH4]+, Collision Offset, 25 V [M + NH4]+ 500 (3) 500 (1) 470 (7) 470 (4) [M + H]+ n.d. n.d. 453 (7) 453 (2) + [M + H - 2H2O] 447 (1) 447 (1) 417 (5) 417 (20) [M + H - Glu]+ n.d. n.d. 277 (19) 277 (72) + [M + H - Glu - H2O] 289 (16) 289 (22) 259 (100) 259 (100) [M + H - Glu- 2H2O]+ 271 (100) 271 (100) 241 (54) 241 (68) + [Glu + H] 177 (6) 177 (2) 177 (25) 177 (7) [Glu + H - H2O]+ 159 (10) 159 (3) 159 (37) 159 (15) + [Glu + H - 2H2O] 141 (12) 141 (6) 141 (56) 141 (25) Negative Ion MS/MS, Precursor [M - H]-, Collision Offset, 40 V [M - H]481 (89) 481 (100) 451 (52) 451 (57) [M - H - H2O]463 (6) 463 (6) 433 (5) 433 (3) [M - H - CH3COOH] 421 (2) 421 (3) 391 (5) 391 (5) [M - H - 118] 363 (5) 363 (4) n.d. 333 (4) [M - H - 134] 347 (1) 347 (2) 317 (4) 317 (2) [M - H - Glu - 2H] 303 (1) 303 (4) 273 (12) 273 (3) [Glu - H] 175 (28) 175 (15) 175 (2) n.d. [Glu - H - H2O] 157 (36) 157 (18) 157 (11) 157 (9) [Glu - H - H2O - CO] 129 (16) 129 (9) n.d. 129 (6) m/z ) 113 (90) (59) (75) (55) m/z ) 95 (10) (6) (12) (7) m/z ) 85 (100) (41) (100) (100) m/z ) 75 (89) (49) (82) (82) m/z ) 59 (12) (3) (6) (5) [M + Na]+ [M + NH4]+
a
See nomenclature in Figure 1. b n.d. ) not detected.
because the amount of organic solvent must be less than 10% to avoid denaturation of enzyme catalysts. The purity of the synthesis products was characterized with LC-MS. The total ion chromatograms showed only one intense peak corresponding to the steroid glucuronide, and no impurities were detected. Despite a complex incubation matrix containing biological material, a simple procedure based on protein precipitation and Table 3.
1H
and
13C
liquid-liquid extraction with dichloromethane followed by C-18 SPE was adequate for efficient and rapid cleanup of the syntheses. The chemical stability of aryl and alkyl glucuronides has earlier been reported to be good (18). However, during our LC-MS studies it was evident that acid-catalyzed methylation of the glucuronide moiety occurred in standard samples, which were stored in a freezer for more than two weeks in the presence of methanol and at pH 4.2. Stock solutions were therefore prepared in acetonitrile and stored in a freezer for several months without any detectable changes in the steroid glucuronide structure. Mass Spectrometry. The abundant ammonium adduct ions [M + NH4]+ and deprotonated molecules [M H]- recorded in positive and negative ESI-MS, respectively, indicate the correct molecular weights of the synthesized steroid glucuronides. The product ion spectra of [M + NH4]+ and [M - H]- show diagnostic product ions [M + H - Glu]+, [M + H - Glu - H2O]+, [M + H Glu - 2H2O]+ and [M - H - Glu - 2H]-, indicating clearly the presence of the glucuronide moiety (Table 2). These ions also provide strong evidence for the expected O-glucuronidation. The product ions [Glu + H]+, [Glu + H - H2O]+, [Glu - H]-, and [Glu - H - H2O]- provide further evidence for the presence of the glucuronide moiety. The isomers 5R-NG and 5β-NG can be distinguished according to the ratio of relative abundances of product ions [M + H - Glu]+/[M + H - Glu - H2O]+, which is significantly higher for 5β-NG than for 5R-NG. However, the product ion spectra of 5R-MTG and 5β-MTG pair were strongly analogous and thus without a distinction of isomers. A more detailed discussion about MS and MS/MS behavior of the compounds is presented in our earlier paper (35). NMR Spectroscopy. For structural analysis, the 1H and 13C spectra of each steroid glucuronide were assigned. The 1H signals were assigned from DQFCOSY,
NMR Chemical Shifts [ppm]a
1H
5R-MTG
5β-MTG
5R-NG
5β-NG
13C
5R-MTG
5β-MTG
5R-NG
5β-NG
GH1 GH2 GH3 GH4 GH5 H1 H1′ H2 H2′ H3 H4 H4′ H5 H6 H6′ H7 H7′ H8 H9 H10 H11 H11′ H12 H12′ H14 H15 H15′ H16 H16′ H17(CH3) H18(CH3) H19(CH3)
4.443 3.254 3.463 3.621 3.838 1.395 1.395 1.528 1.823 3.957 1.454 1.454 1.644 0.117 1.173 0.890 1.668 1.435 0.720 n.d.d 1.277 1.576 1.240 1.487 1.213 1.222 1.55c 1.590 1.825 1.169 0.842 0.831
4.511 3.204 3.445 3.609 3.842 0.961 1.826 1.307 1.791 3.680 1.547 1.820 1.396 1.247 1.877 1.110 1.437 1.492 1.43c n.d. 1.29c 1.48c 1.314 1.507 1.326 1.227 1.548 1.617 1.833 1.192 0.839 0.958
4.451 3.249 3.455 3.619 3.842 1.264 1.643 1.321 2.014 4.006 1.183 1.795 1.507 1.019b 1.572b 1.019b 1.776b 1.284 0.717 0.808 1.138 1.839 1.185 1.693 1.326 1.532 1.921 1.990 2.378 0.849 -
4.512 3.191 3.436 3.604 3.838 1.538 1.691 1.197 1.787 3.686 1.615 1.678 1.800 1.256 1.941 1.192 1.553 1.282 1.320 1.416 1.106 1.763 1.264 1.712 1.415 1.526 1.930 1.996 2.376 0.848 -
GC1 GC2 GC3 GC4 GC5 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C14 C15 C16 C17(CH3) C18(CH3) C19(CH3)
102.35 74.32 77.14 72.56 75.78 33.08 25.83 74.07 34.99 39.91 29.03 32.05 36.99 55.00 n.d. 21.03 32.38 51.56 23.81 39.28 26.24 14.35 11.59
102.16 74.48 77.14 72.59 75.86 35.86 27.33 78.94 34.93 42.91 27.77 26.91 37.55 41.30 n.d. 21.11 32.60 51.45 23.91 39.44 26.31 14.36 23.65
102.37 74.44 77.15 72.56 75.82 24.50 29.86 73.68 39.90 36.87 34.00b 30.46b 41.41 48.98 47.45 25.46 32.47 51.32 21.99 35.86 13.85 -
102.26 74.42 77.16 72.62 75.87 31.99 26.54 78.95 34.97 36.52 26.43 25.68 42.06 39.39 40.86 25.55 32.51 51.08 22.13 35.93 13.89 -
a
See nomenclature in Figure 1. b C6-H6,H6′ and C7-H7,H7′ may have to be interchanged. c From HMQC.
d
n.d., not determined.
198 Bioconjugate Chem., Vol. 13, No. 2, 2002
Figure 2. Expansion of 1D, HMQC and HMBC spectra of 5βNG. In the 1D spectrum, a typical aglycone H3 resonance (AH3) of a H3β-H5β configuration is seen. The solid lines in HMQC and HMBC spectra with JGH1,GH2 ) 7.7 Hz establish the glycosidic linkage to be β(1f3).
and the corresponding 13C resonances were located from the HMQC spectrum (Table 3). The anomericity of the glucuronide moieties was determined from the vicinal coupling constant 3JGH1,GH2. As the measured coupling constants of 7.7-7.8 Hz are typical values between two axial protons (40), the β-configuration was confirmed. The aglycone H3 resonance in the 1H spectra of 5R-MTG and 5R-NG appears as a quintet with a coupling constant of ∼2.5 Hz as expected for the H3β-H5R steroids (41). On the other hand, the corresponding H3 resonance in 1H spectra of 5β-MTG and 5β-NG is a multiplet of seven peaks (Figure 2), which is the typical pattern for the H3βH5β-configuration (41). These results confirm the expected anomericity at H5 for each steroid. Finally, the positions of the glycosidic linkages of each steroid glucuronide were identified from the interglycosidic correlation in the HMBC spectra. In each HMBC spectra the
Kuuranne et al.
resonance correlating H1 of the glucuronide to C3 of the aglycone was observed (Figure 2). Liquid Chromatography-Mass Spectrometry. A sensitive LC-MS method was developed for the determination of steroid-glucuronidation activities. Steroid glucuronide structures possess neither chromophores nor fluorophores, and therefore conventional LC methods based on UV or fluorescence detection are not applicable. The method showed a good linearity as indicated by the calibration curves, which were linear over a wide concentration range of 10 to 5000 ng/mL and exhibited correlation coefficients (r2) between 0.9989 and 0.9993. The quantitation limit of the present method was sufficient for the determination of steroid metabolizing UGT activity in small-volume enzyme assays (100 µL) with a very low aglycone (I, II, III, IV) concentration, starting from 1 µM and upward (Table 1). It is of the utmost importance to work with diluted samples due to the limited dynamic range of ESI-MS. The signal saturation at higher concentrations than 10-50 µM is due to droplet surface saturation in the ESI process (42). Enzyme Kinetics. The velocity of the pathway forming glucuronides linked to the 3R-OH moiety of the parent AAS substrates I-IV was measured in hepatic microsomes from Aroclor-induced rats and with the LC-MS method described above. This UGT activity was readily quantitated even at a substrate concentration as low as 1 µM. The actual reaction rates and their dependence on the aglycone concentrations were measured at a 2 h time point and under the same conditions as used in the bulk syntheses. Because the glucuronidation activities obtained (Table 1) were not fully consistent with the assay conditions required for the determination of enzyme kinetic parameters, no Km or Vmax values are given. Our results on enzyme kinetics are in line with reports suggesting that the glucuronidation of hydroxyandrogens shows a high degree of stereo- and regioselectivity (5). The 5R-structured steroids, both C18 and C19 steroids, were glucuronidated in all test concentrations (1-1000 µM) at a clearly lower velocity than the corresponding 5β-anomers. It is also noteworthy, that our NMR data supports the assumption that 17R-CH3 substitution hinders 17β-OH glucuronidation, and the glucuronides detected in the present biosyntheses were linked only to the 3R-hydroxy positions. Yields of the steroid glucuronides 5R-MTG, 5R-NG, and 5β-MTG were reflected by the specific activities assayed at 0.1 to 1 mM substrate concentrations. In this context, 5β-NG was a clear exception. ACKNOWLEDGMENT
Ms. Leena Luukkanen is gratefully acknowledged for her excellent guidance with enzyme-assisted syntheses, Mr. Antti Leinonen and Dr. Kimmo Kuoppasalmi for scientific support, and the National Technology Agency (TEKES) together with the University Pharmacy (YA) for financial support. LITERATURE CITED (1) Scha¨nzer, W. (1996) Metabolism of anabolic androgenic steroids: 5R- and 5β-reduction of 3-keto-4-ene steroids, Recent advances in doping analysis (4), in Proceedings of the Manfred Donike Workshop on Dope Analysis, March 17-22, 1996, pp 185-201. (2) Scha¨nzer, W. (1996) Metabolism of anabolic androgenic steroids. Clin. Chem. 42, 1001-1020. (3) Mackenzie, P. I., Rodbourne, L., and Stranks, S. (1992) Steroid UDP glucuronosyltransferases. J. Steroid Biochem. Mol. Biol. 43, 1099-1105.
Enzyme-Assisted Synthesis of Steroid Glucuronides (4) Mulder, G. J., Coughtrie, M. W. H., and Burchell, B. (1990) Glucuronidation. Conjugation reactions in drug metabolism (G. J. Mulder, Ed.) p 51, Taylor & Francis, Great Britain. (5) Jin, C.-J., Mackenzie, P. I., and Miners, J. O. (1997) The regio- and stereoselectivity of C19 and C21 hydroxysteroid glucuronidation by UGT2B7 and UGT2B11. Arch. Biochem. Biophys. 341, 2, 207-211. (6) Makin, H. L. J., and Trafford, D. J. H. (1972) The chemistry of the steroids. Clinics Endocrin. Metab. 1, 333-360. (7) Shackleton, C. H. L. (1986) Profiling steroid hormones and urinary steroids. J. Chromatogr. Biomed. Appl. 379, 91-156. (8) Mackenzie, P. I., and Owens, I. S. (1983) Purification of a form of mouse liver UDP glucuronosyltransferase which glucuronidates androgens. J. Steroid Biochem. 19, 10971102. (9) Beaulieu, M., Le´vesque, E., Barbier, O., Turgeon, D., Be´langer, G., Hum, D. W., and Be´langer, A. (1998) Isolation and characterization of a simian UDP-glucuronosyltransferase UGT2B18 active on 3-hydroxyandrogens. J. Mol. Biol. 275, 785-794. (10) Be´langer, G., Barbier, O., Hum, D. W., and Be´langer, A. (1999) Molecular cloning, expression and characterization of a monkey steroid UDP-glucuronosyltransferase, UGT2B19, that conjugates testosterone. Eur. J. Biochem. 260, 701-708. (11) Guillemette, C., Hum, D. W., and Be´langer, A. (1995) Specificity of glucuronosyltransferase activity in the human cancer cell line LNCaP, evidence for the presence of at least two glucuronosyltransferase enzymes. J. Steroid Biochem. Mol. Biol. 55, 355-362. (12) Coffman, B. L., King, C. D., Rios, G. R., and Tephly, T. R. (1998) The glucuronidation of opioids, other xenobiotics, and androgens by human UGT2B7Y(268) and UGT2B7H(268). Drug Metab. Dispos. 26, 73-77. (13) Falany, C. N., Green, M. D., and Tephly, T. R. (1987) The enzymatic mechanism of glucuronidation catalyzed by two purified rat liver steroid UDP-glucuronosyltransferases. J. Biol. Chem. 262, 1218-1222. (14) Falany, C. N., Green, M. D., Swain, E., and Tephly, T. R. (1986) Substrate specificity and characterization of rat liver p-nitrophenol, 3R-hydroxysteroid and 17β-hydroxysteroid UDP-glucuronosyltransferases. Biochem. J. 238, 65-73. (15) Oguri, K., Kurogi, A., Yamabe, K., Tanaka, M., Yoshisue, K., Ishii, Y., and Yoshimura, H. (1996) Purification of a phenobarbital-inducible UDP-glucuronosyltransferase isoform from dog liver which catalyzes morphine and testosterone glucuronidation. Arch. Biochem. Biophys. 325, 159-166. (16) Hum, D. W., Be´langer, A., Le´vesque, E Ä ., Barbier, O., Beaulieu, M., Albert, C., Valle´e, M., Guillemette, C., Tchernof, A., Turgeon, D., and Dubois, S. (1999) Characterization of UDP-glucuronosyltransferases active on steroid hormones. J. Steroid Biochem. Mol. Biol. 69, 413-423. (17) Kaspersen, F. M., and Van Boeckel, C. A. (1987) A review of the methods of chemical synthesis of sulphate and glucuronide conjugates. Xenobiotica 17, 1451-1471. (18) Stachulski, A. V., and Jenkins, G. N. (1998) The synthesis of O-glucuronides. Nat. Prod. Rep. 15, 173-186. (19) Conrow, R. B., and Bernstein, S. (1971) Steroid conjugates. VI. An improved Koenings-Knorr synthesis of aryl glucuronides using cadmium carbonate, a new effective catalyst. J. Org. Chem. 36, 863-870. (20) Sanaullah, and Bowers, L. D. (1996) Facile synthesis of [16, 16, 17-2H3]-testosterone, -epitestosterone and their glucuronides and sulfates. J. Steroid Biochem. Mol. Biol. 58, 225-234. (21) Chung, B. C., Mallamo, J. P., Juniewicz, P. E., and Shackleton, C. H. L. (1992) Synthesis of deuterium-labeled 5R-androstane-3R, 17β-diol and its 17β-glucuronide. Steroids 57, 530-536. (22) Rao, P. N., Rodriguez, A. M., and Miller, D. W. (1986) Synthesis of 5R-androstane-3R, 17β-diol 3- and 17-glucuronides. J. Steroid Biochem. 25, 417-421. (23) Lehman, J. P., Ferrin, L., Fenselau, C., and Yost, G. (1981) Simultaneous immobilization of cytochrome P-450 and glucuronyltransferase for synthesis of drug metabolites. Drug Metab. Dispos. 9, 15-18.
Bioconjugate Chem., Vol. 13, No. 2, 2002 199 (24) Lehman, J. P., Fenselau, C., and Depaulo, J. R. (1983) Quaternary ammonium-linked glucuronides of amitriptyline, imipramine, and chlorpromazine. Drug Metab. Dispos. 11, 221-225. (25) Luukkanen, L., Kilpela¨inen, I., Kangas, H., Ottoila, P., Elovaara, E., and Taskinen, J. (1999) Enzyme-assisted synthesis and structural characterization of nitrocatechol glucuronides. Bioconjugate Chem. 10, 150-154. (26) Lautala, P., Kivimaa, M., Salomies, H., Elovaara, E., and Taskinen, J. (1997) Glucuronidation of entacapone, nitecapone, tolcapone, and some other nitrocatechols by rat liver microsomes. Pharm. Res. 14, 1444-1448. (27) Rittmaster, R. S., Leopold, C. A., and Thompson, D. L. (1989) Androgen glucuronyl transferase activity in rat liver, evidence for the importance of hepatic tissue in 5R-reduced androgen metabolism. J. Steroid Biochem. 33, 1207-1212. (28) Falany, C. N., and Tephly, T. R. (1983) Separation, purification and characterization of three isoenzymes of UDPglucuronosyltransferase from rat liver microsomes. Arch. Biochem. Biophys. 227, 248-258. (29) Rao, G. S., Haueter, G., Rao, M. L., and Breuer, H. (1976) An improved assay for steroid glucuronosyl transferase in rat liver microsomes. Anal. Biochem. 74, 1, 35-40. (30) Numazawa, M., Suzuki, K., and Nambara, T. (1977) Glucuronidation of testosterone by rat liver microsomes. Chem. Pharm. Bull. 25, 420-426. (31) Pacifici, G. M., Gucci, A., and Giuliani, L. (1997) Testosterone sulphatation and glucuronidation in the human liver: interindividual variability. Eur. J. Drug Metab. Pharmacokinet. 22, 253-258. (32) Shinohara, Y., Baba, S., and Kasuya, Y. (1984) Synthesis of deuterium labeled 17-methyl-testosterone. Steroids 44, 253-260. (33) Luukkanen, L., Elovaara, E., Lautala, P., Taskinen, J., and Vainio, H. (1997) Characterization of 1-hydroxypyrene as a novel marker substrate of 3-methylcholantrene-inducible phenol UDP-glucuronosyltransferase(s). Pharmacol. Toxicol. 80, 152-158. (34) Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. (35) Kuuranne, T., Vahermo, M., Leinonen, A., and Kostiainen, R. (2000) Electrospray and atmospheric pressure chemical ionization tandem mass spectrometric behavior of eight anabolic steroid glucuronides. J. Am. Soc. Mass Spectrom. 11, 722-730. (36) Marion, D., and Wu¨thrich, K. (1985) Application of phase sensitive two-dimensional correlated spectroscopy (COSY) for measurements of 1H-1H spin-spin coupling constants in proteins, Biochem. Biophys. Res. Commun. 117, 967-974. (37) Summers, M. F., Marzili, L. G., and Bax, A. (1986) Complete proton and carbon-13 assignments of coenzyme B12 through the use of new two-dimensional NMR experiments. J. Am. Chem. Soc. 108, 4285-4294. (38) Bax, A., and Summers, M. F. (1986) 1H and 13C assignments from sensitivity-enhanced detection of heteronuclear multiple-bond connectivity by 2D multiple quantum NMR. J. Am. Chem. Soc. 108, 2093-2094. (39) Scha¨nzer, W. and Donike, M. (1993) Metabolism of anabolic steroids in man: synthesis and use of reference substances for identification of anabolic steroid metabolites. Anal. Chim. Acta 275, 23-48. (40) Van Halbeek, H. NMR spectroscopy of carbohydrates. A tutorial, American Chemical Society 205th National Meeting, Denver, CO, 1993. (41) Kirk, D. N., Gaskell, S. J., and Marples, B. A. (1995) Spectroscopic methods of steroid analysis, in Steroid analysis (H. L. J. Makin, D. B. Gower, and D. N. Kirk, Eds.) p 72, Chapman & Hall, Great Britain. (42) Kostiainen, R., and Bruins, A. (1994) Effect of multiple sprayers on dynamic range and flow rate limitations in electrospray and ionspray mass spectrometry. Rapid Commun. Mass Spectrom. 8, 549-558.
BC010038G