Evidence for false-positive results for Boldenone testing of veal urine due to faecal cross-contamination during sampling

Transcription

1 Food Additives and Contaminants, Vol. 21, No. 8 (August 2004), pp Evidence for false-positive results for Boldenone testing of veal urine due to faecal cross-contamination during sampling C. A. Sgoifo Rossi{*, F. Arioli{, A. Bassini{, L. M. Chiesa{, V. Dell Orto{, M. Montana{ and G. Pompa{ { Dipartimento di Scienze e Tecnologie Veterinarie per la Sicurezza Alimentare, Universita` degli Studi di Milano, Facolta` di Medicina Veterinaria, Via Celoria 10, I Milan, Italy { UNIRELAB, Settimo Milanese, Milan, Italy concentrations lower than 2 ng ml 1. These results further support the supposition that faecal contamination of the urine during sample collection can lead to false-positive results during boldenone analysis. Keywords: promotor, boldenone, veal, urine (Received 12 November 2003; revised 5 May 2004; accepted 18 May 2004) Introduction European Directive 96/22/EC, which controls veterinary residues in animals, does not permit the presence of synthetic growth promoters in products of animal origin or in livestock. Boldenone is categorized in class A3 (growth promoters steroids) and is thus a banned substance. Testing of veal urine for banned substances is part of the European Union statutory programme for animals going into the food chain. In relation to this monitoring, three studies were conducted to investigate the apparent presence of the banned growth promoter boldenone in veal urine, which was suspected as being caused by interference from faecal contamination of the sample. In the first study, urine samples were collected at different times (time 0 and after 30 min) using (1) a conventional zoonotechnical apron and (2) a technique designed specifically to avoid faecal contamination ( kettle ). This resulted in samples that were, respectively, positive and negative for the presence of -boldenone (-BOL). In a second study, urine samples negative to -BOL were collected from eight veal calves, but became positive after deliberate faecal contamination. In a third study, data obtained from the Italian RNP (Residual National Program) indicated that 18.1% of 3295 urine samples collected using the zootechnical apron were positive for -BOL and 2.1% for -boldenone (-BOL), whilst of 902 samples collected using the kettle, -BOL was not detected in any samples and only 0.2% were positive to -BOL, in * To whom correspondence should be addressed. carlo.sgoifo@unimi.it 1,4-Androstadien-17b-ol-3-one, or simply boldenone or dehydrotestosterone, is a steroid, which was synthesized for the first time in The steroid is pharmacologically classed as a growth promoter with androgenic effects and its use has been studied in horses (O Connor et al. 1973), in dogs (Williams et al. 2000) and in pigeons (Hagedorn et al. 1996). Like other androgenic steroids, it has been classified by the International Agency for Research on Cancer (IARC) as a probable human carcinogen, with a carcinogenicity index higher than that of other androgens, such as nandrolone, stanozolol, testosterone and clostebol (De Brabander et al. 2004). A recent study has also demonstrated the role of boldenone in the development of human prostate carcinomas implanted in mice (Baisch et al. 1998). European Directive 96/22/EC controls residues of veterinary substances in animals and does not permit the presence of synthetic growth promoters in products of animal origin or in livestock. Boldenone has been categorized in class A3 (growth promoters and banned drugs steroids), as described in European Directive 96/23/EC. Testing of veal urine is part of the European Union (EU) statutory testing for banned substances of animals going into the food chain. The first experiments concerning the presence of boldenone in veal urine were performed by Arts et al. (1996), who concluded that 17 a-boldenone (a-bol), a metabolite of b-boldenone (b-bol), Food Additives and Contaminants ISSN X print/issn online # 2004 Taylor & Francis Ltd DOI: /

2 Boldenone testing of veal urine due to faecal cross-contamination 757 can be found at levels up to 3 ng ml 1 in the urine of veal calves known not to have been treated. Therefore, the presence of 17 a-bol in urine could not be considered as necessarily indicating use of a banned substance. b-bol was never found in concentrations above 0.1 ng ml 1 in this study, but its presence in animals that were definitely untreated could not be excluded. A study concerning bovine faeces was carried out by Van Puymbroeck et al. (1998), who analysed 50 samples of faeces positive to a-bol during a normal screening in a slaughterhouse, and observed the presence of 1,4-androstadiene 3,17-dione (ADD) and occasionally of b-bol and 4-androstene-3,17-dione (AED), precursor of testosterone. These findings are in accordance with results of Nielen et al. (2004), who detected a-bol and ADD in bovine skin swabs at higher levels and more frequently than b-bol. During 2001 and 2002, in the National Residual Program (NRP) in Italy hundreds of urine samples of veal calves fed with liquid feed were all found to be positive for a-boldenone and, in small number, also for b-bol. These circumstances, related to economic and sanitary aspects, shifted researchers attention to the origin of boldenone (a-bol and b-bol) in veal urine, and pointed out that in the EU, until 2002, different opinions still existed on the presence of boldenone in bovine urine. Whilst in the Netherlands and in other European countries, such as Belgium, the presence of a- BOL was considered as natural when below 2 ng ml 1. In Italy, any concentration of this steroid was considered as evidence of use of a banned substance. In order to evaluate the apparent origin of boldenone in veal calves, as indicated in urine found positive in the NRP sampling, in this study the role of faecal contamination during drawing was considered, as Van Puymbroeck et al. (1998) had reported high concentrations of boldenone and ADD in faeces. Following these studies, it is possible to make a hypothesis about faeces being transferred into urine collected by zoonotechnical apron, which does not avoid contamination. Materials and methods Materials Analytical-grade solvents and reagents were purchased from Merck (Darmstadt, Germany). b-glucuronidase from Helix pomatia was from Sigma-Aldrich (St Louis, MO, USA). 17 a-boldenone was from Tecnalab (Trieste, Italy). 17-b Boldenone was from AllTech (Milan, Italy). Immunoaffinity columns and washing buffer were purchased from Randox Laboratories Ltd (Crumlin, UK). Samples preparation and analysis Urine samples were centrifugated for 10 min at 3000 rpm. An aliquot (5.0 ml) was brought to ph 5.0 with 1 M HCl and 50 ml b-glucuronidase/arylsulphatase enzyme solution (H. pomatia) was added. After incubation for 18 h at 37 C, the ph of the mixture was adjusted to with 1 M NaOH. The sample was purified by immunoaffinity column according to the conditions suggested by the manufacturer. Briefly, after equilibration with 15.0 ml diluted washing buffer, the column was loaded with the sample, washed with 10.0 ml diluted washing buffer and with 5.0 ml water. Finally, the analytes were eluted with 4.0 ml ethanol water (70:30, v/v) mixture. The eluate was evaporated to dryness under nitrogen at 65 C and the residue dissolved in 100 ml methanol. A volume of 50 ml was analysed by LC-MS-MS. Analyses were carried out on a Waters Model 2695 HPLC instrument. Chromatographic separations were obtained under isocratic conditions using a Purospher RP-18 column (250 4 mm I.D., 5 mm) (Merck, Darmstadt, Germany ) at 35 C, with a mobile phase of acetonitrile: water 1:1 and at a flow rate of 1 ml min 1. Mass spectral analyses were carried out on a Thermo Finnigan LCQ Duo LC-MS-MS instrument, equipped with an atmospheric pressure chemical ionization source (APCI) and heated nebulizer interface operating in positive-ion mode at 450 C. Discharge current and capillary voltage were set at 5 ma and 15 V, respectively. A collision energy of 28 ev was chosen for the collision-induced dissociation (CID) in the MS-MS experiments. The protonated molecule at m/z 287 for a-bol and b-bol was the precursor ion for CID, and three product ions were identified to carry out selected reaction monitoring LC-MS-MS analyses. The precursor product ion combinations of m/z 287! 269, 287! 173 and 287! 135 were used for both a-bol and b-bol. Quantitative analyses were carried out by comparing the peak areas of analytes with those of standard

3 758 C. S. Sgoifo Rossi et al. Figure 1. Sampling methodology with a zoonotechnical apron. Figure 2. Cleaning and trimming of the sheath area. solutions. The calibration curve was linear in the range ng ml 1 (r 2 ¼ 0.988). The recoveries for samples spiked with 5 ng ml 1 of both analytes ranged between 87 and 109%. In this study, three investigations were conducted:. First investigation: five veal calves of the Holstein breed were studied with ages between 215 and 235 days. They were bred in a multiple barn and fed with milk replacers and corn silage. All animals were positive for the presence of boldenone in their urine during the NRP 2001 sampling. On first occasion, urine was collected by methodology normally used by National Veterinary Inspectors (zoonotechnical apron on the animal) (figure 1) and, on a second occasion, after 30 min by the alternative methodology, upon cleaning and trimming of sheath area (figure 2), using a kettle kept far away from body to avoid faecal contamination (figure 3).. Second investigation: eight veal calves of the Holstein breed were studied (aged between 180 and 220 days). They were bred in a multiple barn and fed with milk replacers and corn silage. Urine was collected using a kettle kept under the body of the animal, after cleaning and trimming of the animal s coat, thus avoiding any faecal contamination. Every sample (30 ml) was divided in two aliquots (15 ml each). The first was deliberately contaminated with faeces (0.5 g) taken from the floor near each animal. Before analysis, both aliquots were left at room temperature for 12 h.. Third investigation: data on the analysis of urine sampled in Lombardy and Piedmonte by the Figure 3. Sampling by kettle. National Veterinary Service during NRP were evaluated. Samples had been collected by the zoonotechnical apron until July 2002 (n ¼ 3295) and then by kettle, upon cleaning and trimming of the sheath area (n ¼ 902), and were analysed by Istituti Zooprofilattici. Results and discussion First investigation No samples showed the presence of b-bol, so the results in table 1 quote only urine analyses concerning

4 Boldenone testing of veal urine due to faecal cross-contamination 759 Table 1. -BOL levels (ng ml 1 ) in urine sampled by a zoonotechnical apron and kettle upon cleaning and trimming of the sheath area. Zoonotechnical apron Kettle and sheath area cleaning Veal ID H. pomatia deconjugation No deconjugation H. pomatia deconjugation Table 2. Concentration of -BOL and -BOL in urine samples collected by kettle upon cleaning and trimming of the sheath area and in the aliquots contaminated with faeces. a-bol (ng ml 1 ) b-bol (ng ml 1 ) Veal ID Non-contaminated urine Contaminated urine Non-contaminated urine Contaminated urine 1 0 1, , ,4 0 0, , , , , ,4 0 0 a-bol. Two of the five animals studied were positive when sampled by zoonotechnical apron, both in the presence or absence of deconjugation during the extraction procedure; the concentrations of a-bol in the deconjugated and non-deconjugated urine were comparable: this steroid was therefore present in its non-conjugated form. Three hypotheses about the presence of a-bol can be formulated:. First, it is excreted as a conjugate and then hydrolysed by enzymes from faecal microflora.. Second, the urinary elimination of a-bol occurs in the free state (Nielen et al. 2004). No evidence exists concerning this hypothesis. In fact both Arts et al. (1996) and Van Puymbroeck et al. (1998) carried out hydrolysis of the urine of veal calves treated with boldenone, so one cannot deduce if it was excreted as a conjugated or free molecule. On the other hand, samples of urine from the same animals, taken by kettle after 30 min and upon cleaning and trimming of sheath area and extracted after hydrolysis by H. pomatia, were negative.. Third, that the transfer of non-conjugated a-bol from faeces to urine is the most likely. These preliminary data suggest that faecal contamination of urine occurred during sampling by zoonotechnical apron and could interfere with boldenone analysis. Second investigation It was assessed whether the presence of boldenone could be reproduced in the laboratory by deliberate contamination of urine samples. Urine samples from eight veal calves, collected by kettle after cleaning and trimming of the sheath area, were negative for the presence of boldenone. On the other hand, aliquots (15 ml each) of urine contaminated with faeces collected from the floor (0.5 g) were always positive for a-bol and, only in one case, also for b-bol (table 2). Therefore, the results of this investigation indicate that only faeces could be responsible for the presence of boldenone in the veal urine samples. Regarding the presence of boldenone in faeces, one could make a hypothesis based on scientific studies.

5 760 C. S. Sgoifo Rossi et al. Figure 4. Urine samples collected with a zoonotechnical apron and those collected by kettle upon cleaning and trimming of animal sheath area. In rats, oral administration of a mixture of vegetable sterols (b-sitosterol, campesterol and stigmasterol) could lead to the presence of ADD in the intestinal tract through intervention of gut microflora (Song et al. 2000). Vegetable sterols are normally present in coconut and palm oil, and in part or totally used as replacers of lipids of animal origin in veal calf nutrition plans. Therefore, a natural transformation in ADD cannot be excluded (Poelmans et al. 2003). However, in the present circumstances, no evidence exists regarding ADD transformation in boldenone (a-bol and/or b-bol) in rat or bovine intestinal tract. Third investigation In figure 4, the percentage of urine samples found positive for the presence of boldenone during NRP are reported: those collected with a zoonotechnical apron and the those collected by kettle upon cleaning and trimming of animal sheath area. In a population of 902 veal calves of urine samples collected by kettle upon cleaning and trimming of the sheath area, only 0.2% of animals was positive for a- BOL with concentrations above 2 ng ml 1. Thus, the frequency of positive samples was drastically lower than those observed by Arts et al. (1996), who found nearly 90% for positive samples, and affirmed that a-bol can be present naturally in veal urine at concentrations lower than 3 ng ml 1. The authors do not talk about sampling methodology, nor about possible faecal contamination. As pointed out above, the European Community Reference Laboratory also believes the presence of a-bol in urine should be considered as natural below 2 ng ml 1. However, the results of this study clearly indicate that faecal contamination during sampling could be the factor responsible for the presence or absence of a-bol and/or b-bol in urine. Moreover, it is clear that a diligent control of urine contamination during sampling can drastically reduce the rate of samples detected as being positive for a-bol could reduce to zero the number of samples positive for b-bol. Conclusions Urine samples obtained using a zoonotechnical apron are usually contaminated by faecal material present on the coat, sheath or abdomen of the animal because of prolonged contact of urine with faeces. The extent of this contamination is greater if the time of sampling is also extended, and this procedure usually forces the animal with a zoonotechnical apron to stay in a recumbent position. These conditions can cause important differences in contamination between urine sampled by the zoonotechnical apron and urine sampled with a kettle upon cleaning and trimming of sheath area (figure 5). Therefore, this study suggests that urine sampled from the same animal by zoonotechnical apron or by kettle that does not permit any contamination can be, respectively, positive and negative, and it is very

6 Boldenone testing of veal urine due to faecal cross-contamination 761 Figure 5. Urine sampled by kettle upon cleaning and trimming of the sheath area (left) and urine sampled by a zoonotechnical apron (right). important if one considers that in Italy sampling in veal calves is usually carried out by zoonotechnical apron. Besides confirming the hypothesis regarding the faecal origin of a-bol, results from this study confirm that it could also have a non-endogenous origin. Regarding food safety, this statement could be important. In fact, the faecal origin of boldenone in urine excludes a real presence of this steroid as an indication of possible illegal treatment. Thus, there is a need to decide whether sampling urine not contaminated by faecal material is necessary to reveal the presence of boldenone in bovine urine, and to decide if it could be an indication of a banned treatment. In the same way, the definition of a legal natural value of boldenone concentration in bovine urine must be based on samples made by excluding any possibility of faecal contamination. Acknowledgements The authors are grateful to Dr G. Corcelli, Laboratorio Marini s.r.1., Rivoli (TO, Italy), for useful cooperation. References Arts, C. J. M., Schilt, R., Schreurs, M., and Van Ginkle, L. A., 1996, Boldenone is a naturally occurring (anabolic) steroid in cattle. In Proceedings of the Euroresidue III Conferences, Veldhoven 6 8 May 1996, N. Haagama and A. Ruiter (eds), University of Utrecht, Utrecht, The Nederlands, p Baisch, H., Otto, U., and Fack, H., 1998, Growth of human prostate carcinomas with and without hormone alpha dehydrotestosterone in nude mice. European Urology, 34, De Brabander, H. F., Poelmans, S., Schilt, R., Stefany, R.W., Le Bizec, B., Draisci, R., Sterk, S. S., Van Ginkel, L. A., Courtheyn, R., Van Hoof, N., Macri', A., and De Wasch, K., 2004, Presence and metabolism of the anabolic steroid boldenone In various animal species: a review. Food Additives and Contaminants (in press). Hagedorn, H. W., Zankl, H., Grund, C., and Schulz, R., 1996, Nachweis von Dopingsubstanzen in der Brieftaube. Berliner und Munchener Tierarztliche Wochenschrift, 109, Nielen, W. F. M., Rutgers, P., Van Bennekom, E. O., Lasaroms, J. J. P., and Van Rhijn, J. A. H., 2004, Confirmatory analysis of 17-beta-boldenone, 17-alfa-boldenone and androsta-1,4-diene- 3,17-dione in bovine urine, faeces, feed and skin swab samples by liquid chromatography-electrospray ionisation tanden mass spectrometry. Journal of Chromatography B, 801, O Connor, J. J., Stillions, M. C., Reynolds, W. A., Linkenheimer, W. H., and Maplesden, D. C., 1973, Evaluation of boldenone undecylenate as an anabolic agent in horses. Canadian Veterinary Journal, 14, Poelmans, S., De Wasch, K., Martele', Y., Schilt, R., Van Hoof, N., Noppe, H., Versylcke, T., Janssen, C., and Courtheyn, D., 2003, The possible transformation of phytosterols to

7 762 C. S. Sgoifo Rossi et al. boldenone. Proceedings Euro Food Chem. XII, Bruges, Belgium, pp Song, Y. S., Jin, C., and Park, E. H., 2000, Identification of metabolites of phytosterols in rat faeces using GC/MS. Archives of Pharmacological Research, 23, Van Puymbroeck, M., Kuilman, M. E. M., Maas, R. F. M., Witkamp, R. F., Leyssens, L., Vanderzande, D., Gelan, J., Raus, J., and Renger, F., 1998, Identification of some important metabolites of boldenone in urine and feces of cattle by GC/MS. Analyst, 123, Williams, T. M., Kind, A. J., Hyde, W. G., and Hill, D. W., 2000, Characterization of urinary metabolites of testosterone, methyltestosterone, mibolerone and boldenone in greyhound dogs. Journal of Veterinary Pharmacology and Therapeutics, 23,