4-Chlorotestosterone acetate metabolites in cattle after intramuscular and oral administrations

^a Address correspondence to this author at: LDH-LNR, Ecole Nationale Vétérinaire (Ministère de l'Agriculture), BP 50707, 44307 Nantes Cedex 03 France. Fax (33)-2-40-68-77-45; e-mail ldhlnr{at}vet-nantes.fr.

	Abstract

Top Abstract Introduction Methods and Materials Results and Discussion References

 
The use of 4-chlorotestosterone acetate by farmers for cattle fattening was recently demonstrated although the use of this anabolic steroid is strictly forbidden in the European Union. We investigated the metabolism of 4-chlorotestosterone acetate in the bovine species after intramuscular and oral administration. Nineteen metabolites were detected in urine after intramuscular injection, and eight metabolites were identified. For this purpose, preparative HPLC, mass spectrometry with different ionization modes (electronic impact and chemical ionization), and different acquisition techniques were used (high resolution, selected ion monitoring, and scan measurement). Metabolite stereoisomerism was determined on the basis of retention time and organic synthesis. 4-Chloroepitestosterone (M2), 4-chloroandrost-4-en-3-ol-17-one (M3), and 4-chloroandrost-4-ene-3,17-dione (M4) were identified as the main urinary markers of intramuscular administration. On the other hand, 4-chloroandrost-4-ene-3,17ß-diol (M7), 4-chloroandrostan-3ß-ol-17-one (M5), and M2 were the primary indicators of an oral administration. In addition, we have shown that 95% of the metabolites were sulfo-conjugated, except for M3, which was partially conjugated to glucuronic acid. Finally, the main metabolites (M2, M3, and M4) were easily identified for 1.5 months after intramuscular administration.

	Introduction

Top Abstract Introduction Methods and Materials Results and Discussion References

 
The use of anabolic steroids and thyreostats has been forbidden in the European Union since 1981 (directive 81/602/EEC). At the beginning of the 1990s, the large number of positive results for 4-chlorotestosterone acetate (CTA¹ ; Clostebol) after injection site analyses in France and elsewhere in Europe led us to focus on the metabolism of this androgenic steroid. Because bibliographic data relative to CTA biotransformations were limited in cattle at the beginning of our study, only 4-chlorotestosterone (CT) was screened in various biological matrices; therefore, no positive samples were found. The observation of a large discordance between the discovery of CTA in injection sites and the absence of CT in urine meant that the target analyte to detect in urine was not CT, as one could imagine; a metabolism study had to be initiated (1)(2)(3).

The first studies dealing with CTA metabolism were begun at the end of the 1950s. During this period, the synthesis of testosterone analogs substituted in the 4 position (4)(5)(6) showed an increase of the anabolic activity, especially with 4-halogenated compounds (7). CTA metabolism was first studied on guinea pig liver slices; it was shown that CT was oxidized in 4-chloroandrost-4-ene-3,17-dione (8). Later, CTA administration to ovario-suprarenalectomized women permitted the identification of another metabolite, 4-chloroandrost-4-en-3-ol-17-one (9)(10). Other investigations were carried out after oral administration to patients; toxicological (11) as well as metabolic (12) studies were performed. The two previously mentioned metabolites were identified. The incubation of CT with human liver and the use of mass spectrometry allowed the detection of 16 metabolites and the precise identification of 4 of them: 4-chloroandrost-4-ene-3,17ß-diol, 4-chloroandrost-4-ene-3ß,17ß-diol, and the corresponding hydroxylated derivatives (13). In the 1970s, the increasing use of CTA and of 1-dehydro-17-methyl-4-chlorotestosterone (Oral Turinabol^®) in sports led the antidoping control laboratories to study accurately the metabolism of these molecules. 4-chloroandrosterone and 4-chloroetiocholanolone, as well as 4-chlorotestosterone, were identified in human urine after the oral administration of CT (14). These metabolites were found to be conjugated. Schänzer and co-workers showed that after the oral administration of CTA in humans, the major metabolite found in urine was 4-chloroandrost-4-en-3-ol-17-one, this structure being confirmed by synthesis (15)(16). The metabolism of CTA in cattle was investigated only after 1993 [B. LeBizec, oral communication presented to the Institut National de la Recherche Agronomique, 1993 and (1)(2)(3)(17)(18)(19)]. Nevertheless, a limited number of phase I metabolites were identified after intramuscular (i.m.) administration. Phase II metabolites and the CTA residues in different excreta or organs were not studied. Leyssens and co-workers demonstrated the existence of 4-chloroandrost-4-ene-3,17-dione, 4-chloroandrost-4-ene-3,17ß-diol, 4-chloroandrost-4-en-3-ol-17-one, and 4-chloroepitestosterone (17)(18)(19). They also found two hydroxylated metabolites but lacked real certainty concerning the structure. No information was available on the conjugated forms of these metabolites.

The goal of the present study was to determine the structures of phase I and phase II urinary metabolites after the i.m. and oral administration of CTA to cattle. The urinary elimination kinetics of the main metabolites were studied to determine which withdrawal period still allows the detection of the metabolites.

	Methods and Materials

Top Abstract Introduction Methods and Materials Results and Discussion References

 
apparatus
The quadrupole mass spectrometer (MS) used was a Model 5989A coupled to a Model 5890 gas chromatograph (GC), both from Hewlett–Packard. The magnetic MS was a reverse-geometry, double-focusing SX102A instrument from Jeol coupled to a HP-5890 GC. The HPLC used was a Hewlett–Packard HP-1050, coupled to a fraction collector (Model 203; Gilson). The HPLC conditions were as follows: a 20-µL injection loop; flow rate, 2 mL/min; fractions collected each minute (2 mL). The solvent program is indicated in Table 1 .

gc-ms settings
GC.
The transfer line temperature was set at 280 °C and the split/splitless injector was maintained at 250 °C (time of splitless, 1 min). The column used was a HP-1 (30 m x 0.25 mm i.d.; film thickness, 0.25 µm; Hewlett–Packard). The GC settings were as follows: initial temperature, 120 °C (2 min), followed by a temperature change of 15 °C/min to 250 °C (0 min), then 5 °C/min to 300 °C (5 min). Helium (N55) was used as carrier gas at a flow rate of 1 mL/min.

MS.
MS determinations were performed under different modes. For electron impact (EI), the electronic beam energy was set at 70 eV in the EI mode. For positive chemical ionization (PCI), methane was used as the reagent gas. For high-resolution selected-ion monitoring, measurements were performed under EI conditions and in the selected-ion monitoring mode; the accelerating voltage was scanned while the magnetic field was kept constant. Perfluorokerosene ions were used as lock masses. The MS resolution was tuned on benzene and pyridine ions (m/z = 79) corresponding to C₅CH₆^· (79.05030) and C₅H₅N^· (79.04220) fragments, respectively. Slits were adjusted to reach a resolution of about 12 000.

reagents and reference compounds
Most of the reagents and solvents used were reagent grade products from Merck and from SDS. Helix pomatia juice was from Biosepra. The SPE C₁₈ columns (2 g of phase) were obtained from Varian, and the G60 silica gel was from Merck. N-Methyl-N-trimethylsilyltrifluoroacetamide (MSTFA) and trimethyliodosilane (TMIS) were purchased from Fluka; hydroxylamine hydrochloride and dithiothreitol (DTE) were from Aldrich. The reference steroids were obtained from Steraloids, Research Plus, and Sigma.

animals
CTA (500 mg) made soluble in sterilized peanut oil was injected i.m. into the cattle. The oil permitted slower diffusion of the steroid from the injection site. Urine was collected for 45 days. However, one calf received CTA by the oral route.

A plastic bag with one small magnet on each side was installed on each animal to collect urine. Each time the animal urinated, the bag walls and their magnets moved away one from each other, inducing a current shift. This current shift switched on a pump that aspirated urine directly into a refrigerator that was maintained below 4 °C. Twice a day, the total excreted volume was weighed, homogenized, and an aliquot was frozen (below -16 °C).

extraction procedure
Urine (20 mL) was hydrolyzed at least 15 h at 52 °C with 100 µL of Helix pomatia juice and 2 mL of acetate buffer (2 mol/L, pH 5.2). The analytes were adsorbed onto a SPE C₁₈ column previously conditioned with 5 mL of methanol and 5 mL of ultrapure water. The column was washed successively with 5 mL of ultrapure water and 5 mL of hexane; analytes were eluted with 5 mL of methanol/ethyl acetate (30:70, by volume). The eluate was washed with 2 mL of sodium hydroxide (1 mol/L) and evaporated. The dry residue was reconstituted in 1.5 mL of 1,1,1-trichloroethane/ethyl acetate (80:20, by volume). The residue was applied to a silica gel column (i.d., 1 cm; length, 8 cm) conditioned with 1,1,1-trichloroethane/ethyl acetate (80:20, by volume). Four milliliters of this mixture were used to wash the stationary phase. The metabolites were eluted with 15 mL of 1,1,1-trichloroethane/ethyl acetate (20:80, by volume). The eluate was evaporated to dryness under reduced pressure.

derivatization before the gc-ms analysis
The dry residue was derivatized according to two procedures: (reaction A) 60 min at 80 °C with DTE/TMIS/MSTFA (5 g of DTE and 5 ml of TMIS per liter of MSTFA) or (reaction B) 60 min at 60 °C with hydroxylamine/pyridine (30 g/L), followed by evaporation under a nitrogen stream, and then 60 min at 80 °C with DTE/TMIS/MSTFA (5 g of DTE and 5 mL of TMIS per liter of MSTFA).

Oxidation with pyridinium chlorochromate complex (PCC) was also used to simplify metabolite mass spectra. Purified dry residue from urine was dissolved in PCC/pyridine (1:1000, by volume). After 1 h at 60 °C, the mixture was evaporated under a gentle stream of nitrogen. The residue was derivatized according to reaction A.

synthesis of reference substances
Oxidation.
One milliliter of chromium trioxide/sulfuric acid/acetone (10 g of chromium trioxide and 20 mL of sulfuric acid per liter of acetone) was added to the dry substance residue (a standard of 4-chlorotestosterone to synthesize 4-chloroandrost-4-ene-3,17-dione and an HPLC fraction containing pure metabolite to determine the number of oxidizable alcohol functions). The mixture was allowed to react for 1 h at room temperature; it was then made alkaline with an equal volume of sodium hydroxide (1 mol/L) and extracted twice with 2 mL of diethyl ether. The organic layer was dried over sodium sulfate and evaporated under reduced pressure.

Reduction (15)
. 4-Chloroandrost-4-ene-3,17-dione (1 mg) was dissolved in 2 mL of anhydrous diethyl ether, and lithium aluminum hydride (LiAlH₄) was added (1 mol/L). After the mixture reacted for 15 min at room temperature, 2 mL of water was added; the mixture was extracted with 2 mL of diethyl ether. The organic layer was dried over sodium sulfate and evaporated under reduced pressure.

6-Hydroxylation.
(21)(22)(23) CTA (1 mg) was dissolved in 1 mL of perchloric acid (70%)/acetic anhydride (5:100, by volume). The reaction was maintained at 25 °C for 5 h; the organic layer was then washed with 1 mL of an aqueous solution of potassium hydroxide and dried under reduced pressure. The 4-chloroandrosta-3,5-diene-3,17-diol diacetate obtained from the reaction was dissolved in 1 mL of metachloroperbenzoic acid/dichloromethane (1 g/L) at 0 °C; the mixture was stirred for 12 h at room temperature. The organic layer was washed with 1 mL of an aqueous solution of potassium hydroxide and evaporated to dryness under reduced pressure. The 5,6-epoxy-4-chloroandrost-3-one 3-enol,17-diacetate obtained from the reaction was dissolved in ethanol and exposed to ultraviolet radiation or directly to sunlight for 12 h at 50 °C. After evaporation, the resulting 4-chloroandrost-4-en-6-ol-3-one 17-acetate was dissolved in 1 mL of hydrochloric acid (30%)/acetone (10:90, by volume) and heated for 4 h at 70 °C. The mixture was made alkaline with 1 mL of sodium hydroxide and extracted twice with 2 mL of diethyl ether; the solvent was evaporated to dryness under reduced pressure.

	Results and Discussion

Top Abstract Introduction Methods and Materials Results and Discussion References

 
detection of the urine metabolites
The urinary extract (derivatized according to reaction A) was first injected into the GC–MS (EI ionization, quadrupole mass spectrometer) in the scan measurement mode (masses from 300 to 650 u). Ion chromatograms corresponding to the control urine (J₀) and to a urine (J₅) taken 5 days after CTA administration were compared. All of the ionic signals that appeared in J₅ were considered as potential metabolites of CTA (Fig. 1 ); eight signals were detected (M1 to M8). The M8 metabolite is not shown because its molecular ion was not monitored. The mass spectra (shown in Fig. 2 ) of the fragments indicated that a chlorine atom was present in the compound. The interpretation of the different fragments will be made later.

View larger version (17K): [in this window] [in a new window]   Figure 1. Ion chromatograms showing the urinary metabolic profiles after CTA i.m. (left) and oral (right) administration to cattle (reaction A derivatization). Ions monitored: m/z 466, 468, 470, and 472.

View larger version (38K): [in this window] [in a new window]   Figure 2. Mass spectra of metabolites M1 to M8 derivatized according to reaction A and detected in the EI mode.

confirmation of urine metabolites
Chemical ionization.
We first tried to confirm the mass value of the molecular ion. The same derivatized extract analyzed above was analyzed another time in the GC–MS but in the PCI mode, using methane as the reagent gas. Quasi-molecular ions (MH) and their adducts (MC₂H₅) were observed for eight analytes. The measured m/z values are summarized in Table 2 . The M4 metabolite full mass spectrum is shown on Fig. 3 .

View larger version (16K): [in this window] [in a new window]   Figure 3. PCI (methane) mass spectrum of the M4 metabolite derivatized according to reaction A.

High resolution mode.
Exact masses of the molecular ions (Cl and Cl contributions) detected before were recorded in the EI mode. Because of the low metabolite concentrations, the following exact masses were not experimentally determined but were theoretically calculated: M₁, C₂₅H₄₃ClO₂Si₂ (466.2490 u and 468.2460 u);M₂, C₂₅H₄₃ClO₂Si₂ (466.2490 u and 468.2460 u); M₃, C₂₅H₄₃ClO₂Si₂ (466.2490 u and 468.2460 u); M₄, C₂₅H₄₁ClO₂Si₂ (464.2334 u and 466.2304 u); M₅, C₂₅H₄₅ClO₂Si₂ (468.2647 u and 470.2617 u); M₆, C₂₅H₄₇ClO₂Si₂ (470.2803 u and 472.2773 u); M₇, C₂₅H₄₅ClO₂Si₂ (468.2647 u and 470.2617 u); and M₈, C₂₈H₅₁ClO₃Si₃ (554.2835 u and 556.2805 u).

The previously mentioned analytes were detected again; this observation confirmed the elemental composition of the eight metabolites. The high resolution allowed the enhancement of detection specificity (Fig. 4 ). Almost no noise was recorded on the metabolite ion chromatograms. In addition, three additional metabolites were discovered: two monohydroxylated forms (m/z 554.2835) and one dihydroxylated (m/z 642.3179) form of chlorotestosterone.

View larger version (18K): [in this window] [in a new window]   Figure 4. High resolution ion chromatograms (r = 12 000) of a cow urinary extract (reaction A derivatization) 5 days after CTA i.m. administration. Metabolite peaks are colored black; 35Cl and 37Cl contribute to the M4 molecular ion (left, topand bottom) and M8 (right, top and bottom).

Preparative HPLC and full mass spectrum.
To know more than the elemental composition of the metabolites, full mass spectra with diagnostic ions were needed. Because of the low concentration of the metabolites and the concentration of the matrix background, a preparative HPLC was developed. We chose to use a normal phase, including silica as the stationary phase and strict organic solvents as the mobile phases to accelerate evaporation speed. The urinary extracts were split into 20 fractions; each fraction was then divided into two equivalent volumes, and each half was derivatized by reactions A or B. A low resolution, full mass spectrum was recorded on an electromagnetic MS. The cleanup was so impressive that the signal-to-noise ratio for each metabolite was considerably increased; 19 metabolites were detected. Seven compounds corresponded to reduced or oxidized CT, 11 were monohydroxylated CT forms, and 1 was identified as a dihydroxylated derivative. Molecular ion, base peak, and main fragment intensities of the 19 detected metabolites are summed up in Table 3 (reaction A derivatization and EI ionization).

molecular structure determination
Trimethylsilyl-enol (TMS-enol) and TMS-oxime derivatization mass spectrum studies.
Metabolite structures were determined by comparisons of the EI mass spectra after derivatization by reactions A or B (Fig. 5 ). The comparison of the M^· mass of the same metabolite after it had been derivatized by either reaction A or reaction B gives the number of keto groups directly. The results for metabolite M4 are shown in Fig. 6 . The molecular ions of the TMS (reaction A) and oxime derivatives (reaction B) were 464 and 494, respectively. The difference could be explained by a double reaction of hydroxylamine on the metabolite keto functions; this observation indicates a 4-chloro-androst-4-ene-3,17-dione structure.

View larger version (14K): [in this window] [in a new window]   Figure 5. Derivatization products obtained with reactions A and B. The presence of one keto group on a steroid structure leads to a 15-u increase of the molecular ion mass with reaction B compared with reaction A.

View larger version (20K): [in this window] [in a new window]   Figure 6. EI mass spectra of M4 derivatized with reaction A (top) and reaction B (bottom).

Fragmentation studies, using the EI-ionization, high-resolution acquisition mode, of testosterone analogs (6-hydroxytestosterone, 5-androstan-3-ol-17-one, 5-androstan-3ß,17ß-diol, and 17-epitestosterone) derivatized into the corresponding TMS-enol-TMS-ethers allowed us to correlate specific fragments with precise chemical structures:

(a) 17-keto steroids exhibit an intense 169 ion (C₉H₁₇OSi), which might correspond to a C- and D-ring fragmentation (15);

(b) androstane-3,17-diols release their trimethylsilanol and methyl groups very easily, leading to intense ions corresponding to (M-Me), (M-TMSOH)^· , (M-TMSOH-Me), and (M-2TMSOH-Me);

(c) androst-4-en-3-one compounds give rise to a 208 m/z fragment (C₁₂H₂₀OSi); according to an equivalent retro-Diels Alder mechanism, 4-chloroandrost-4-en-3-one compounds generate a 242 m/z ion (C₁₂H₁₉ClOSi); and

(d) contrary to androst-4-en-3-one (which is mainly converted to androst-3,5-diene when derivatized by reaction A), androst-4-en-3-ol yields a characteristic B-ring fragmentation by cleavage of the 9–10 and 5–6 bonds with charge retention on the A-ring. The resulting fragment for 4-chloroandrost-4-en-3-ol steroids is a 215 m/z ion whose elemental composition, C₁₀H₁₆ClOSi, was confirmed by high resolution measurement.

M1 and M2 increased their M^· mass by 15 u when derivatized according to reactions A and B, indicating that these metabolites had one keto function. The presence of the 242 m/z ion seemed to indicate 4-chloroandrost-4-en-3-one structures. Comparison with a standard of CT indicated that M1 was chlorotestosterone itself, and M2, whose retention time was shorter than the retention time of CT, could be 17-epichlorotestosterone.

After comparison of the reaction A and reaction B derivatives of M3 (15 u difference), it was possible to determine that one keto group was present. The high intensity of the 169 m/z ion could be interpreted as the existence of a 17-keto group trimethylsilylated to a TMS-enol ether. The chlorinated 215 m/z fragment observed in the derivatives of both reaction A and reaction B led us to theorize that M3 was a 4-chloroandrost-4-en-3-ol steroid. We therefore concluded that M3 was 4-chloroandrost-4-en-3-ol-17-one.

The increase of the M4 molecular ion by 30 u after reaction B indicated the presence of two keto groups. The intense 169 m/z ion detected in the mass spectrum of the reaction A derivative suggested a 17-keto group. The 229 m/z ion (C₁₀H₁₆ClNOSi) of the reaction B derivative (Fig. 5 ) was found to correspond to the rupture of the 5–6 and 9–10 bonds with charge retention on the A-ring (characteristic of a 4-chloroandrost-4-en-3-one TMS-oxime). The structure was concluded to be 4-chloroandrost-4-ene-3,17-dione.

The increase of the M5 molecular ion by 15 u after reaction B indicated the presence of one keto group. The abundant 169 m/z ion observed in the spectrum of the reaction A derivative of M5 could be interpreted as the fragmentation of a 17-keto group trimethylsilylated derivative to an TMS-enol derivative (15). The molecular weight of 324 for this compound thus indicated a 4-chloroandrostan-3-ol-17-one structure.

The mass spectrum of M6 remained unchanged no matter which derivatization was used (reaction A or reaction B), which meant that the metabolite had a diol structure. The M^· mass value (326 u) suggested a 4-chloroandrostane-3,17-diol.

The M7 mass spectrum was unchanged after reaction B, indicating a diol structure. The presence of the chlorinated 215 m/z fragment observed in both the reaction A and reaction B derivatives led us to hypothesize that M7 was a 4-chloroandrost-4-en-3-ol compound. Considering these two assumptions, M7 was concluded to be a 4-chloroandrost-4-ene-3,17-diol.

The molecular ion of the M8 TMS derivative was observed at m/z 554, indicating a hydroxylated metabolite. The mass spectra comparison of the reaction A and reaction B derivatives ( 15 u) allowed us to deduce the presence of one ketone and two alcohol functions.

Oxidation mass spectrum studies.
Some oxidation reactions performed on corticosteroids in our laboratory (J. Negriolli, university thesis, in preparation) and elsewhere (20) were used to give complementary data for the metabolite structure elucidation. The comparison of different oxidizing agents showed that PCC was one of the most powerful; it permits oxidation of primary or secondary alcohols even in a hindered position. The main metabolites (isolated from the other metabolites by HPLC) were oxidized with PCC. The mass spectrum study led us to the conclusion summarized in Fig. 7 . M1, M2, M3, and M7 were oxidized into 4-chloroandrost-4-ene-3,17-dione (m/z 320 u), confirming their 4 unsaturation. M5 and M6 were converted to 4-chloro-androstane-3,17-dione (m/z 322 u), proving the saturated character of the 4–5 bond. The M8 oxidized product was four mass units less than M8 itself (m/z 334 u), proving the presence of two hydroxyl groups, one keto group, and one unsaturated bond (Fig. 8 ). M4 was unchanged after oxidation.

View larger version (13K): [in this window] [in a new window]   Figure 7. Metabolite structures after oxidation with PCC.

View larger version (17K): [in this window] [in a new window]   Figure 8. Full mass spectrum of M8 after oxidation with PCC and derivatization according to reaction A (EI ionization).

metabolite synthesis
Synthesis started with CT. This steroid (M1) was synthesized from CTA that had undergone chemical hydrolysis. Alkaline conditions (potassium hydroxide in ethanol) were rejected because 4-hydroxytestosterone would be produced by nucleophilic substitution of the chlorine atom in the 4 position by a hydroxyl group. Acidic medium was preferred (hydrochloric acid in acetone). M4 was formed after the oxidation of CT with PCC. The reduction of this molecule (4-chloroandrost-4-ene-3,17-dione) with lithium aluminum hydride (LiAlH₄) produced a good yield of M5 and low quantities of M2, M3, and M6. The action of sodium borohydride (NaBH₄) on CT led to the formation of M7. We synthesized M8 by the 6-hydroxylation of CT (Fig. 9 ). The synthesized 6-hydroxylated steroid derivatized according to reaction A showed exactly the same mass spectrum and retention time as the M8 metabolite, 6-hydroxychlorotestosterone (Fig. 10 ).

View larger version (18K): [in this window] [in a new window]   Figure 9. Reaction steps leading to 6-hydroxychlorotestosterone.

View larger version (17K): [in this window] [in a new window]   Figure 10. EI mass spectrum of the synthesized 6-hydroxychlorotestosterone-3,6,17-tris-TMS (reaction A derivatization).

summary statement of cta urinary metabolites after i.m.AND ORAL ADMINISTRATIONS
A summary of the metabolite structures is presented in Table 4 . An indication of their relative importance after oral or i.m. administration is also given. The three main metabolites identified in urine after the i.m. administration of CTA to a cow were 4-chloroepitestosterone (M2), 4-chloroandrost-4-en-3-ol-17-one (M3), and 4-chloroandrost-4-ene-3,17-dione (M4). They constitute the main markers of illegal CTA i.m. administration. On the other hand, the presence of 4-chloroepitestosterone (M2), 4-chloroandrostan-3ß-ol-17-one (M5), and above all, 4-chloroandrost-4-ene-3,17ß-diol (M7) indicates CTA oral administration. In this case, the metabolites are mainly reduced, probably because of some preliminary biotransformations in the stomach.

study of the phase II METABOLITES
Our strategy to determine the conjugation forms of the metabolites was based on the affinity differences that exist between a conjugate and a free steroid for an organic and an aqueous phase.

To isolate free metabolites, urine (20 mL) was extracted with 2 x 20 mL of diethyl ether; after centrifugation, the solvent was evaporated to dryness. The dry residue was reconstituted in 1.5 mL of 1,1,1-trichloroethane/ethyl acetate (80:20, by volume) and purified on a silica gel column, as described previously.

To isolate glucuronide metabolites, the urine residue described above was submitted to a nitrogen stream to eliminate the possible remaining traces of diethyl ether and mixed with 2 mL of acetate buffer (2 mol/L, pH 5.2) and 200 µL of ß-glucuronidase (extracted from bovine liver, 5 x 10 U/L; Sigma). After the metabolites were extracted with 2 x 20 mL of diethyl ether, they were purified on a silica gel column.

To isolate sulfate metabolites, the remaining diethyl ether was eliminated under a nitrogen stream; 2 mL of acetate buffer (2 mol/L, pH 5.2) and Helix pomatia juice (200 µL; ß-glucuronidase, 10 FU/L; arylsulfatase, 10 RU/L; Biosepra) was then added. The sulfate and the mixed conjugated forms were extracted with diethyl ether and purified as described previously.

The results are summarized in Fig. 11 . We observed a very important excretion of the metabolites in their sulfate forms. 4-Chloroandrost-4-ene-3,17-dione was found in this fraction; the absence of hydroxyl groups on the structure led us to the following hypothesis. On one hand, it could mean that the sulfate is linked to an enol in the 3 or 17 position. This hypothesis was made by Gerhards et al. in 1965 in their study of the metabolism of methenolone acetate in man, in which they identified a 3-enol-glucuronide form for one metabolite (24). On the other hand, the presence of 4-chloroandrost-4-ene-3,17-dione (M4) could be explained by side activities of the Helix pomatia juice. It has been shown (25)(26) that some 3ß-hydroxy-5-ene-steroids could be converted into their 4-en-3-one forms by the isomerases and oxidases contained in Helix pomatia preparations. Nevertheless, we were not able to confirm these hypotheses; we plan to analyze the conjugated forms directly in LC-MS via an electrospray interface in a near future.

View larger version (25K): [in this window] [in a new window]   Figure 11. Conjugate forms of urinary CTA metabolites after i.m. administration. Free forms (top); glucuronide forms (middle); sulfate forms (bottom); y-axes, abundances in arbitrary units; and x-axes, metabolite names.

Less than 5% of the metabolites were excreted unconjugated (M2, M3, and M4). This observation made a preliminary deconjugation step necessary. M3 was the only metabolite excreted in its glucuronide form. Because of the fast action of ß-glucuronidase, it could be used as a screen for M3 in urine in the program to control illegal CTA use. Moreover, we did not observe a change in the percentage of conjugated forms during the days following i.m. administration. The differences between the conjugated forms recorded after oral and i.m. administration were minor.

excretion profile of cta urinary metabolites after i.m.ADMINISTRATION
No test was performed between J₁₅ and J₄₂. The urine collection system had the disadvantage of generating bedsores if it was maintained more than 15 days. We preferred to remove the collector before bedsores appeared.

The excretion of CTA urinary metabolites showed almost the same profile over the course of 43 days. The elimination curves of M1 and M2 are represented on Fig. 12 . The maximal concentration was reached between the days 12 and 14, with values close to 10 µg/L for M2. Chloroepitestosterone (M2) concentrations of 3 µg/L were still detected 1.5 months after i.m. administration. This concentration is well above the detection limit of the method, even when a quadrupole MS is used. The demonstration of CTA fraud is possible at least 1.5 months after i.m. administration, and most probably is possible for several months.

View larger version (43K): [in this window] [in a new window]   Figure 12. Urinary elimination profiles of M1 (top) and M2 (bottom) after CTA i.m. administration. Y-axes, abundances; x-axes, days after i.m. administration.

Nineteen metabolites were detected in the urine of a cow after CTA i.m. administration. A structure was proposed for the eight main metabolites detected. Six were unambiguously identified by the GC–MS in the low and the high resolution mode. EI and PCI, as well as selective ion monitoring and scan measurement mode acquisition, were used. The use of preparative HPLC increased the signal-to-noise ratio for each metabolite. The stereoisomerism of each metabolite was determined on the basis of its retention time. The remaining doubts about structure were erased by the organic synthesis of each metabolite. 4-Chloroepitestosterone (M2), 4-chloroandrost-4-en-3-ol-17-one (M3), and 4-chloroandrost-4-ene-3,17-dione (M4) were identified as the main urinary markers of i.m. administration. On the other hand, 4-chloroandrost-4-ene-3,17ß-diol (M7), 4-chloroandrostan-3ß-ol-17-one (M5), and M2 were found to be the main indicators of oral administration. Moreover, we have shown that 95% of the metabolites were excreted in sulfate form, except for M3, of which 25% was glucurono-conjugated. Finally, the main metabolites (M2, M3, and M4) were easily identified (good signal-to-noise ratio) 1.5 months after i.m. administration and, presumably, were detectable long after that time.

View larger version (9K): [in this window] [in a new window]   Figure .

	Acknowledgments

 
The work that has been described was based to a large extent on research carried out at LDH-LNR in which several scientists and technicians participated. Important contributions were made by Isabelle Gauthier, Cécile Roué, Claude Thernay, and Isabelle Gaudin.

	Footnotes

 
LDH-LNR, Ecole Nationale Vétérinaire (Ministère de l'Agriculture), BP 50707, 44307 Nantes Cedex 03 France.

¹ Nonstandard abbreviations: CTA, 4-chlorotestosterone acetate; CT, 4-chlorotestosterone; i.m., intramuscular; MS, mass spectrometer; GC, gas chromatograph; EI, electron impact; PCI, positive chemical ionization; MSTFA, N-methyl-N-trimethylsilyltrifluoroacetamide; TMIS, trimethyliodosilane; DTE, dithiothreitol; PCC, pyridinium chlorochromate complex; and TMS, trimethylsilyl.

	References

Top Abstract Introduction Methods and Materials Results and Discussion References

 

1. Le Bizec B, Montrade M-P, Monteau F, André F. Metabolites of some anabolic steroids in cattle after oral administration. Haagsma N Ruiter A eds. Proceedings of Euroresidue III, conference of residues of veterinary drugs in food, Veldhoven (May 6–8, 1996) 1996:248-252 Faculty of Veterinary Medicine Utrecht. .
2. Le Bizec B. Utilisation de la spectrométrie de masse pour le contrôle de l'usage des stéroïdes anabolisants en élevage. Application à l'étude du métabolisme de la 4-chlorotestostérone acétate chez le bovin. Nantes: Thèse d'Université, 1996:346..
3. André F, Le Bizec B, Montrade M-P, Maume D, Monteau F, Marchand P. Developments in residue assay and metabolism of growth promoters by mass spectrometric analysis. Analyst 1994;119:2529-2535. [Medline] [Order article via Infotrieve]
4. Camerino B, Patelli B, Vercellone A. Synthesis and anabolic activity of 4-substituted testosterone analogues. J Am Chem Soc 1956a;78:3540-3541.
5. Camerino B, Patelli B, Vercellone A, Meda F. Analoghi di ormoni steroidi. Nota II. 4-Idrossi e 4-alo-derivati. Il Farmaco-Ed Sc 1956;11:586-597.
6. Ringold HJ, Batres E, Mancera O, Rosenkranz G. Steroids. LXXXII. Synthesis of 4-halo hormone analogs. Steroids 1956;21:1432-1435.
7. Sala G, Baldratti G. Myotrophic (anabolic) activity of 4-substituted testosterone analogues. Proc Soc Exp Biol Med 1957;95:22-24.
8. Sala G, Castegnaro E. Metabolismo in vitro di un nuovo derivato del testoterone a prevalente azione anabolica: il 4-chlorotestosterone. Folia Endocrinol 1958;11:346-353.
9. Castegnaro E, Sala G. Isolamento del 4-chloro-4-androsten-3-ol-17-one dalle urine di pazienti trattati con 4-cloro-testosterone acetato. Folia Endocrinol 1961;14:581-589.
10. Molinatti GM, Camanni F, Pizzini A. Osservazioni sul metabolismo del 4-cloro-testosterone acetato nel soggetto normale ed ovarosurrenectomizzato. Folia Endocrinol 1961;14:528-533.
11. Krüskemper HL, Morgner KD. Untersuchungen zur klinischen Pharmakologie von Methyltestosteron, Mesterolon und 4-Chlortestosteron. Int J Clin Pharmacol 1968;1:455-460.
12. Castegnaro E, Sala G. Absorption and metabolism of 4-chlorotestosterone acetate by oral route. Steroids Lipids Res 1973;4:184-192. [ISI][Medline] [Order article via Infotrieve]
13. Stàrka L, Siekmann L, Hoppen HO, Breuer H. Stoffwechsel von 4-Chlortestosteron in der Leber des Menschen. Arzneim-Forsch 1969;19:2022-2025. [Medline] [Order article via Infotrieve]
14. Cartoni GP, Ciardi M, Giarrusso A, Rosati F. Capillary gas chromatographic mass spectrometric detection of anabolic steroids. J Chromatogr 1983;279:515-522. [ISI][Medline] [Order article via Infotrieve]
15. Schänzer W, Donike M. Metabolism of anabolic steroids in man: synthesis and use of reference substances for identification of anabolic steroid metabolites. Anal Chim Acta 1993;275:23-48.
16. Schänzer W. Metabolism of anabolic androgenic steroids. Clin Chem 1996;42:1001-1020. [Abstract/Free Full Text]
17. Leyssens L, Royackers E, Gielen B, Missoten M, Schoofs J, Czech J, et al. Metabolites of 4-chlorotestosterone acetate in cattle urine as diagnostic markers for its illegal use. J Chromatogr B 1994;654:43-54. [ISI][Medline] [Order article via Infotrieve]
18. Leyssens L, Royackers E. GC/MS for the detection of illegal use of steroids in cattle industry. GC-MS Varian application note. Varian 1994;33:1-4.
19. Hendriks L, Gielen B, Leyssens L, Raus J. Screening for the illegal use of clostebol acetate in cattle by identification of its urinary metabolites. Vet Rec 1994;134:192-193. [ISI][Medline] [Order article via Infotrieve]
20. Courtheyn D, Vercammen J, de Brabander H, Van denreyt I, Batjoens P, Van oosthuyze K, van Peteghem CH. Determination of dexamethasone in urine and faeces of treated cattle with negative chemical ionization-mass spectrometry. Analyst 1994;119:2557-2564. [Medline] [Order article via Infotrieve]
21. Edwards BE, Narasima RP. Perchloric acid catalysed acylations. Enol lactonization and enol acetylation of steroids. J Org Chem 1966;31:324.
22. Gardi R, Lusignani A. Autooxidation of steroids 3,5-diene-3-ol ethers. A simple route to 6ß-hydroxy 4–3-ketones. J Org Chem 1967;32:2647-2649.
23. Gauthier I. Synthèse de dérivés stéroïdiens. Rapport de DEA option "chimie-fine", LDH/LNR 1994:21 Université de Nantes Nantes. .
24. Gerhards E, Kolb KH, Schulze PE. Zum Stoffwechsel von 17ß-Acetoxy-1-methyl-1–5-androstenon-(3) (Methenolonacetat) beim Menschen. II. Isolierung und Identifikation von 3-Hydroxy-1-methylen-5-androstanon-(17) und anderen Metaboliten und zum Mechanismus der Ring-A-reduktion. Hoppe-Seyler's Z Physiol Chem 1965;342:40-62. [ISI][Medline] [Order article via Infotrieve]
25. Vanluchène E, Eechaute W, Vandekerckove D. Conversion of free 3ß-hydroxy-5-ene-steroids by incubation with Helix pomatia [Preliminary Note]. J Steroid Biochem 1982;16:701-703. [ISI][Medline] [Order article via Infotrieve]
26. Messeri G, Cugnetto G, Monetti G, Serio M. Helix pomatia induced conversion of some 3ß-hydroxy-steroids. J Steroid Biochem 1984;30:793-796.