Studies on the determination of chlorotestosterone and its metabolites in bovine urine

Studies on the determination of chlorotestosterone and its metabolites in bovine urine Michaela Walshe, a Michael Keeffe a and Bruno Le Bizec b a Teagasc, The National Food Centre, Dunsinea, Castleknock, Dublin 15, Ireland b LDH/LNR, Ecole Nationale Vétérinaire, Ministère de l Agriculture et de la Pêche, BP 50707, 44307 Nantes Cedex 03, France Received 2nd July 1998, Accepted 16th September 1998 Chlorotestosterone and its metabolites were determined in urine samples from bovine animals treated with chlorotestosterone acetate by oral and intramuscular routes. Sample preparation, involving enzymatic deconjugation and solid-phase extraction, was optimised. The effect of different enzyme preparations, ph, and time of incubation were studied. An extraction/clean-up procedure based on solid-phase extraction (C 18 cartridge) and liquid liquid clean-up was developed. Determination of chlorotestosterone and its metabolites was by enzyme immunoassay and GC-MS. Metabolites were converted into their TMS-enol-TMS-ether and TMS-oxime-TMS-ether forms before GC-MS (EI) analysis. Introduction 4-Chlorotestosterone (clostebol) is an anabolic steroid which can be administered either orally or intramuscularly. This anabolic steroid is an important constituent of black-market preparations used for growth stimulation in animal husbandry. The European Union has prohibited the use of such hormones in animal breeding under directive 86/469. 1 Depending on the method of administration, different metabolites are found over a period of time. Previous work in this area was carried out by Le Bizec et al. 2 4 who examined the metabolism of 4-chlorotestosterone in cattle with determination by GC-MS. The animals had been treated intramuscularly and orally with chlorotestosterone acetate and 4-chloroepitestosterone, 4-chloro-4-androstene-3,17-dione and 4-chloro-4-androsten-3a-ol-17-one were the main urinary metabolites observed following intramuscular injection, while these metabolites together with 4-chloro-4-androstene-3a,17b-diol, 4-chloroandrostan-3b-ol-17-one and 4-chloroandrostane-3b,17a-diol were found following oral administration. Leyssens et al. 5,6 also examined the metabolites of chlorotestosterone acetate in the urine of cattle that had received an intramuscular injection. The metabolites were determined by high performance thin-layer chromatography (HPTLC) and capillary GC-MS following solid-phase extraction (SPE) clean-up and HPLC fractionation. A metabolite of clostebol has been determined in human urine following the consumption of meat contaminated with chlorotestosterone acetate. 7,8 The determination was carried out by GC-MS following a solid-phase clean-up on C 18 cartridges. The object of this work was to establish a routine method for the determination of 4-chlorotestosterone and its metabolites in bovine urine samples. A method was developed incorporating an efficient deconjugation step, SPE step and further clean-up by liquid liquid extraction. Determination was by enzyme immunoassay, and quantification of the 4-chloroandrostenedione, 4-chloroepitestosterone and 4-chlorotestosterone was carried out by GC-MS on a number of incurred samples. Presented at the Third International Symposium on Hormone and Veterinary Drug Residue Analysis, Bruges, Belgium, June 2 5, 1998. Experimental Animal treatments and samples All samples of bovine urine were stored frozen at 220 C until analysed. Control urine samples of known history, free of chlorotestosterone and/or its metabolites, were used in fortification studies. Urine samples were obtained from an animal treated with chlorotestosterone acetate (CTA) by two methods, oral and intramuscular routes. For samples following oral treatment, the animal was treated with a suspension of 500 mg of CTA in 30 ml of liquid paraffin, by oral administration in 2 3 15 ml doses. Urine samples were collected on a twice-daily basis for four days and each sample was filtered through Whatman filters (541 and GF/A) and then through a cellulose acetate membrane (0.45 mm), prior to storage at 220 C. For samples following intramuscular treatment the animal was injected with 500 mg of CTA. This had been dissolved in 0.1 ml of ethanol, 2 ml of benzyl alcohol and 8 ml of peanut oil and mixed prior to administration. The urine samples were collected on a daily basis for two weeks and filtered prior to freezing. The samples selected for analysis included a control (blank) urine taken prior to administration of CTA, denoted C1b. Three samples were selected following oral treatment, at 0, 1 and 3 days post administration and were denoted T2, T3 and T7, respectively. Four samples were selected following intramuscular injection, at 0, 3, 9 and 13 days post administration, and were denoted IM1, IM6, IM9 and IM14, respectively. Enzyme immunoassay Materials. Chlorotestosterone acetate was obtained from Sigma (St. Louis, M, USA), chlorotestosterone was obtained from Alltech (Applied Science Laboratories, State College, PA, USA) and 4-chloroandrostenedione was obtained from Dr. L. Leyssens (Dr. Willems-Instituut, Diepenbeek, Belgium) (Fig. 1). Water (doubly distilled), sodium acetate, methanol, hexane and ethyl acetate (BDH, Poole, England) were used for residue extractions. The C 18 Bond Elut SPE cartridges (1 g, 6cm 3 capacity) were obtained from Varian (Harbor City, CA, Analyst, 1998, 123, 2687 2691 2687

USA). Phosphate buffered saline (PBS) of ph 7.3 was supplied by xoid (Unipath Ltd., Basingstoke, England). Suc d Helix pomatia (SHP), containing 10 5 Fishman units of b-glucuronidase and 10 6 Roy units of sulfatase per millilitre, was supplied by Sepracor (Villeneuve La Garenne, France). Enzyme immunoassay kits were supplied by Laboratoire d Hormonologie (Marloie, Belgium), with antiserum that had been raised against chlorotestosterone hemisuccinate BSA. The principal cross-reactivities of the antibody are: 4-chloro-4-androstene- 3,17-dione (100%), 4-chlorotestosterone (28%), chlorotestosterone acetate (120%). The standard used in the immunoassay was 4-chloro-4-androstene-3,17-dione. Filter papers GFA and 541 were obtained from Whatman (Maidstone, UK) and the 0.45 mm cellulose acetate filters were obtained from Millipore (St. Louis, M, USA). Apparatus. The ethyl acetate methanol extracts were evaporated to dryness under nitrogen on a heating block (Stuart, UK) at 45 C. The water bath used for the incubation step was set at 37 C. During SPE, the C 18 columns were connected to a vacuum manifold (Waters, Milford, MA, USA). Absorbance measurements were carried out using a plate reader (Dynatech MRX Minireader, Dynatech Laboratories Ltd., Billingshurst, UK) with a 450 nm filter. Extraction procedure. An overall scheme for the clean-up of urine samples is presented in Fig. 2. Bovine urine (1 ml) was placed in a test-tube. 1.5 ml of PBS buffer containing 50 ml of Suc d Helix pomatia was added, and the tubes vortexed for 10 s. They were placed in a water bath set at 37 C overnight. Prior to SPE, 2.5 ml of PBS buffer was added to each tube and vortexed for 10 s. Solid-phase extraction. The C 18 SPE column was conditioned with 5 ml of methanol, followed by 5 ml of doubly distilled water and 5 ml of PBS buffer. The sample was added to the column and was allowed to pass through under gravity. The extraction column was washed with 10 ml of water and the column allowed to dry for 10 min by drawing air through the column. Hexane (3 ml) was passed through the column and the column was allowed to dry for 10 min. The residues were eluted from the column with 2 3 3 ml of ethyl acetate methanol (70 : 30). The organic layer was washed with 5 ml of 1 M potassium hydroxide, and mixing effected with a vortex mixer for 30 s. The samples were centrifuged at 2000g for 10 min. The organic layer was removed and the washing with potassium hydroxide was repeated. The organic layer was then evaporated at 45 C under nitrogen. The residue was reconstituted in 1 ml H of dilution buffer, and then further diluted 1 + 9 with dilution buffer for determination by enzyme immunoassay. Gas chromatography-mass spectrometry analysis Procedure. Methylchlorotestosterone and methyltestosterone-d 3 (internal standards, 100 ng each) were added to 10 ml samples of thawed bovine urine. A 200 ml portion of Suc d Helix pomatia and 2 ml of 2 M acetate buffer of ph 5.2 were added. The samples were incubated overnight at 52 C. The C 18 SPE column was conditioned with 10 ml of methanol, followed by 10 ml of doubly distilled water. The sample was added to the column and was allowed to pass through under gravity. The extraction column was washed with 10 ml each of water and hexane. The column was eluted with 5 ml of ethyl acetate methanol (70 : 30, v/v). The organic layer was washed with 2 3 5 ml of 1 M potassium hydroxide. The organic layer was then evaporated in a Speed-Vac system. The residue was reconstituted in 1.5 ml of trichloroethane ethyl acetate (80 : 20, v/v). The residue was added to a G60 silica gel column, which was prepared by mixing silica gel with trichloroethane ethyl acetate (80 : 20, v/v) and pouring the slurry into a glass column (8 3 1 cm). Trichloroethane ethyl acetate (80 : 20, v/v, 4 ml) was passed through the column and discarded. Trichloroethane ethyl acetate (80 : 20, v/v, 13 ml) and 12 ml of trichloroethane ethyl acetate (20 : 80, v/v) eluted the metabolites in two fractions (Fraction I, Fraction II). Derivatization. The organic fractions were reduced in a Speed-Vac system and transferred to a vial and 100 ng of norgestrel (external standard) were added. The extract was evaporated under a nitrogen stream and the dry residue was derivatised with 20 ml of MSTFA TMIS DTE, [N-Methyl-N- (trimethylsilyl) trifluoroacetamide trimethyliodo-silane dithioerythritol] at 60 C for at least 30 min. Gas Chromatography-mass spectrometry. Samples (2 ml) were injected into the GC-MS (splitless mode, SIM acquisition, electron impact ionisation) and the following masses recorded: 558/556/554/552 corresponding to mono-hydroxylated metabolites; 472/470/468/466/464 corresponding to oxidised or reduced non-hydroxylated metabolites; 449/301 d 3 -methyltestosterone; 456/316 norgestrel; 480/482 methylchlorotestosterone. Urine (1ml) -add 1.5 ml hydrolysis buffer (PBS buffer, 0.02 M, ph 7.0 containing 50 μl Suc d'helix pomatia) -incubate at 37 C, 16 h Hydrolysed urine (2.5 ml) 4-Chlorotestosterone 4-Chloroandrostenedione H rganic extract -add 2.5 ml of PBS -apply to C 18 cartridge (Bond Elut) (conditioned with 5 ml methanol, 5 ml water and 5 ml PBS) -wash with 10 ml water, 3 ml hexane -elute with 2 x 3 ml ethyl acetate methanol (70:30) 4-Chloroepitestosterone Fig. 1 Structures of 4-chlorotestosterone, 4-chloroandrostenedione and 4-chloroepitestosterone. ELISA (50 μl, in duplicate wells) -extract with 2 x 5 ml potassium hydroxide, 1 M -evaporate solvent at 45 C under nitrogen -dissolve residue in 1 ml buffer -dilute 1:10 with buffer Fig. 2 Procedure for extraction of 4-chlorotestosterone metabolites from bovine urine prior to determination by ELISA. 2688 Analyst, 1998, 123, 2687 2691

Results and discussion Enzyme immunoassay Conditions for extraction of 4-chlorotestosterone metabolites from urine samples prior to determination by ELISA were not available; the aim of this investigation was to develop an efficient procedure, with a minimum number of steps, which would provide a suitable sample extract for assay. ptimisation of the deconjugation step. ptimisation of the deconjugation step was carried out using a urine sample collected following intramuscular treatment. This step was optimised in terms of the incubation period, the ph of the buffer used for the incubation, and the amount of Suc d Helix pomatia added. Initial work involved examination of the buffer ph. Previous work by Le Bizec et al. 9 found that the optimum ph for deconjugation of steroid sulfates was ph 5.2 while for deconjugation of steroid glucuronides values in the ph range 4 7 were found to be suitable. Leyssens et al. 5 reported the use of ph 7.0 for deconjugation of 4-chlorotestosterone metabolites. Two values were investigated, ph 5.2 and ph 7.0, and the data in Table 1 indicate that the overall trend is for a higher yield from deconjugation at ph 7.0. The second parameter examined was incubation time, two values being selected (2 and 16 h). From the values obtained (Table 1) the overall trend indicates that an overnight incubation is necessary with higher results being obtained after the longer incubation time. Regarding the amount of enzyme, Suc d Helix pomatia, two concentrations were examined, equivalent to 500 and 5000 Fishman units of b-glucuronidase. The data in Table 1 indicate that the higher concentration yields higher responses in ELISA. The final conditions selected following optimisation were 0.02 M PBS, ph 7.0, with 5000 units of b- glucuronidase, and incubation for 16 h at 37 C. Table 1 Conditions for deconjugation of metabolites in incurred urine sample with variation of ph, time and concentration of deconjugation agent Incubation Suc d Helix pomatia/ ELISA response/ ph time/h Fishman units ng ml 21 urine 7.0 2 500 13.5 5000 14.5 16 500 11.8 5000 > 50 5.2 2 500 7.5 5000 4.9 16 500 7.7 5000 > 50 Development of the solid-phase extraction step. Following optimisation of the deconjugation step, development of a SPE step on C 18 was carried out. ther workers involved in the determination of 4-chlorotestosterone metabolites found that water 3 (100%) or water methanol 5 (30:70) were suitable as wash solutions. Both of these washes were examined, with the water wash (100%) being selected due to the production of cleaner extracts. A variety of eluents are reported as being suitable for eluting 4-chlorotestosterone metabolites from C 18 SPE cartridges, ethyl acetate, 7 methanol 3 and a mixture of ethyl acetate methanol 2 (70 : 30). Ethyl acetate was found to give incomplete recovery while methanol was found to increase the response obtained for all samples, including the control (blank) urine, and therefore was unsuitable for use with subsequent ELISA determination. The mixture of ethyl acetate methanol (70 : 30) was found to give good recovery and cleaner extracts. A further step was required, following SPE, as the clean-up procedure was insufficient for enzyme immunoassay determination. Previous work in this area 5 utilised a combination of silica and amino columns to further purify extracts containing steroids. ther workers 7 used an amino column in conjunction with the C 18 column for further purification. Le Bizec et al. 2 previously described the use of 1 M KH as an additional cleanup step to follow elution from C 18 cartridges. The wash step with 1M KH was added as a clean-up step as high responses for control (blank) urine samples indicated that the SPE step alone was insufficient clean-up for ELISA determination. The overall procedure is shown in Fig. 2. Enzyme immunoassay determination. As the metabolites do not undergo any separation step ELISA gives an overall response due to the presence of 4-chlorotestosterone and/or its metabolites, depending on their cross-reactivities with the antiserum. nly cross-reactivities for 4-chlorotestosterone, 4-chloroandrostenedione and chlorotestosterone acetate have been established. The response in the ELISA represents a combined response to all cross-reacting metabolites present. Results for urine samples following oral (T3) and intramuscular (IM9) treatments are shown in Table 2. Gas chromatography-mass spectrometry Ion chromatograms for urine samples collected before chlorotestosterone acetate administration were compared with those from samples collected post treatment. All signals present in the post treatment samples and absent from the control (pretreatment) samples, and having (M + )/(M + + 2 amu) ions in a ratio indicating the presence of a chloride atom, were considered to be potential metabolites from the chlorotestosterone acetate treatments. Table 2 Quantification of 4-chlorotestosterone (CLS), 4-chloroepitestosterone (EPICLS) and 4-chloroandrostenedione (CLAD) in incurred samples GC-MS results/ ELISA response/ ng ml 21 urine ng ml 21 urine (n =3) Sample identification CLS EPICLS CLAD Total metabolites (b) Post oral treatment T2 2.0 3.5 30.9 na a T3 na a na a na a 29 ± 3.5 T7 0.0 5.4 3.5 na a (a) Post intramuscular treatment IM6 0.8 13.5 42.6 na a IM9 1.2 14.4 51.0 199 ± 29 IM14 2.9 68.7 133.5 na a a na = Not analysed Table 3 Major metabolites of 4-chlorotestosterone identified in urine samples by GC-MS G 60 Fraction(s) Treatment Retention where Metabolite (ion) time/min predominant ral IM 4-Chloroandrostenedione (464) 21.0 I, II a a 4-Chloro-4-androsten-3a-ol- 17-one (466) 19.75 I, II b b 4-Chloroepitestosterone (466) 20.66 II a a 4-Chlorotestosterone (466) 21.33 II a b 4-Chloro-4-androstene-3a,17bdiol (468) 19.05 I, II a c 4-Chloroandrostan-3b-ol-17-one (468) 20.43 I b c a High abundance. b Intermediate abundance. c Low abundance. Analyst, 1998, 123, 2687 2691 2689

The major metabolites identified in urine samples after oral and intramuscular treatment are shown in Table 3, together with the fraction(s) from the G60 silica gel column in which they occurred predominantly, and their retention times on the chromatogram. Fig. 3 shows representative SIM chromatograms of ions 464, 466 and 468 for urine samples following oral and intramuscular treatments with chlorotestosterone acetate, by comparison with chromatograms for control urine samples. Table 2 shows the quantification of 4-chlorotestosterone, 4-chloroepitestosterone and 4-chloroandrostenedione in the urine samples by GC-MS. f these metabolites, 4-chloroandrostenedione is the major component in all samples from the treated animal. However, other metabolites, such as 4-chloro- 4-androsten-3a-ol-17-one and reduced metabolites were found to be the major components in some samples post intramuscular and oral treatment, respectively. These results confirm that 4-chloroandrostendione is a suitable target metabolite for analysis for the illegal use of chlorotestosterone acetate in animal production. Conclusions This paper describes the development of a screening method for 4-chlorotestosterone and/or its metabolites in bovine urine samples. An extraction procedure was developed for use in conjunction with enzyme immunoassay determination. Enzyme Fig. 3 Representative SIM chromatograms of ions a, 464; b, 466 and c, 468 for urine samples following i, oral and ii, intramuscular treatment. 2690 Analyst, 1998, 123, 2687 2691

immunoassay gave an overall value for 4-chlorotestosterone metabolites from bovine urine samples. GC-MS was used for the identification of the major metabolites in urine following oral and intramuscular administration of chlorotestosterone acetate to a bovine animal and for the specific quantification of 4-chlorotestosterone, 4-epichlorotestosterone and 4-chloroandrostenedione in incurred urine samples. Acknowledgements This work was carried out within the framework of the EC-SMT programme [Project no. SMT4-CT96-2092, Rapid control systems for screening veterinary drug residues in food (producing animals) ]; the authors (MW and MK) acknowledge financial support from this programme. Dr. L. Leyssens (Dr. L. Willems-Instituut, Diepenbeek, Belgium) is thanked for providing the standard material, 4-chloroandrostenedione. Mr. T. Darby, Grange Research Centre, Teagasc and Mr. P. Byrne, The National Food Centre, Teagasc are thanked for their assistance with animal treatment and sampling. References 1 EEC Directive 86/469, ff. J. Eur. Commun. 26/09/86 (1986), No. L275/36. 2 B. Le Bizec, M.-P. Montrade, F. Monteau and F. André, in Proc. EuroResidue III Conference, ed. N. Haagsma and A. Ruiter, University of Utrecht, Faculty of Veterinary Medicine, Utrecht, The Netherlands, 1996, 248. 3 F. André, B. Le Bizec, M.-P. Montrade, D. Maume, F. Monteau and P. Marchand, Analyst, 1994, 119, 2529. 4 B. Le Bizec, M.-P. Montrade, F. Monteau, I. Gaudin and F. André, in. Chem., 1998, 44(5), 973. 5 L. Leyssens, E. Royackers, B. Gielen, M. Missotten, J. Schoofs, J. Czech, J. P. Noben, L. Hendricks and J. Raus, J. Chromatogr. B, 1994, 654, 43. 6 L. Hendricks, B. Gielen, L. Leyssens and J. Raus, Vet. Record, 1994, 134, 192. 7 G. Debruyckere, R. De Sagher and C. Van Peteghem, in. Chem., 1992, 36, 1869. 8 G. Debruyckere, R. De Sagher, C. Van Peteghem, G. Van Vyncht, G. Maghuin-Rogister and E. De Pauw, Anal. Chim. Acta, 1994, 291, 155. 9 B. Le Bizec, M.-P. Montrade, F. Monteau and F. André, Anal. Chim. Acta, 1993, 275, 123. Paper 8/05125I Analyst, 1998, 123, 2687 2691 2691