This article was downloaded by: [University of Victoria] On: 26 February 2015, At: 03:32 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Food Additives & Contaminants: Part A Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tfac20

Investigation of the presence of prednisolone in bovine urine a

a

b

c

a

Eva de Rijke , Paul W. Zoontjes , Danny Samson , Sabrina Oostra , Saskia S. Sterk & Leendert A. van Ginkel

a

a

RIKILT part of Wageningen UR, European Union Reference Laboratory, Wageningen, the Netherlands b

Mead Johnson Nutrition, Nijmegen, the Netherlands

c

National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands Accepted author version posted online: 07 Jan 2014.Published online: 03 Mar 2014.

Click for updates To cite this article: Eva de Rijke, Paul W. Zoontjes, Danny Samson, Sabrina Oostra, Saskia S. Sterk & Leendert A. van Ginkel (2014) Investigation of the presence of prednisolone in bovine urine, Food Additives & Contaminants: Part A, 31:4, 605-613, DOI: 10.1080/19440049.2013.878479 To link to this article: http://dx.doi.org/10.1080/19440049.2013.878479

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Food Additives & Contaminants: Part A, 2014 Vol. 31, No. 4, 605–613, http://dx.doi.org/10.1080/19440049.2013.878479

Investigation of the presence of prednisolone in bovine urine Eva de Rijkea*, Paul W. Zoontjesa, Danny Samsonb, Sabrina Oostrac, Saskia S. Sterka and Leendert A. van Ginkela a RIKILT part of Wageningen UR, European Union Reference Laboratory, Wageningen, the Netherlands; bMead Johnson Nutrition, Nijmegen, the Netherlands; cNational Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands

Downloaded by [University of Victoria] at 03:32 26 February 2015

(Received 19 August 2013; accepted 18 December 2013) Over the past two years low levels of prednisolone have been reported in bovine urine by a number of laboratories in European Union member states. Concentrations vary, but are reported to be below approximately 3 µg l–1. Forty per cent of bovine urine samples from the Dutch national control plan had concentrations of prednisolone between 0.11 and 2.04 µg l–1. In this study the mechanism of formation of prednisolone was investigated. In vitro conversion of cortisol by bacteria from faeces and soil, bovine liver enzymes and stability at elevated temperatures were studied. In vitro bovine liver S9 incubation experiments showed a significant 20% decrease of cortisol within 6 h, and formation of prednisolone was observed from 0.2 g l–1 at t = 0 to 0.5 g l–1 at t = 6. Under the influence of faeces, the stability of cortisol in urine is reduced and cortisol breaks down within 50 h. Prednisolone is formed up to 4 µg l–1 at 70°C after 15 h. However, this decreases again to zero after 50 h. With soil bacteria, a slower decrease of cortisol was observed, but slightly higher overall formation of prednisolone, up to 7 µg l–1 at 20°C. As opposed to incurred urine, in fortified urine incubated with faeces or soil bacteria no prednisolone was detected. This difference may be explained by the presence of natural corticosteroids in the incurred sample. With UPLC-QToF-MS experiments, in urine and water samples incubated with faeces, metabolites known from the literature could be (tentatively) identified as 20β-hydroxy-prednisolone, cortisol-21-sulfate, oxydianiline, tetrahydrocortisone-3-glucuronide and cortexolone, but for all compounds except 20β-hydroxy-prednisolone no standards were available for confirmation. Based on the results of this study and literature data, for regulatory purposes a threshold of 5 µg l–1 for prednisolone in bovine urine is proposed. Findings of prednisolone in concentrations up to 5 µg l–1 in bovine urine can, most likely, originate from other sources than illegal treatment with growth promoters. Keywords: metabolism; mass spectrometry; corticosteroids; ultra-performance liquid-chromatography; prednisolone; cortisol; conversion; bacteria; faeces; urine

Introduction Prednisolone, a corticosteroid with glucocorticosteroid activity, is used for the treatment of a wide range of inflammatory and autoimmune conditions. The use of corticosteroids in livestock is regulated in the European Union for therapeutic purposes. Prednisolone is one of the registered veterinary drugs for intra-mammary administration, which also include, for example, amoxycilin and clavulinic acid. The maximum residue limit after therapeutic use of prednisolone in bovine animals is 4 µg kg–1 in muscle and fat, 10 µg kg–1 in liver and kidney, and 6 µg kg–1 in milk (EU Commission Regulation 37/2010). In the Netherlands corticosteroids are classified as group A3 substances (anabolic/unauthorised substances) because of their steroidal structure, whereas other European Union member states classify them as B2f (other pharmacologically active substances) (EU Council Directive 96\23\EC). No recommended concentration for control was established in the Community Reference Laboratories Guidance Paper (2007). The illegal use of corticosteroids as growth promoters can, however, not be excluded. Corticosteroids can *Corresponding author. Email: [email protected] © 2014 RIKILT

prolong the effect of growth-promoting substances, such as anabolic steroids and β-agonists, in the last weeks before slaughter. Next to that, low doses of glucocorticoids can result in improved feed intake, increased live weight gain, a reduced feed conversion ratio, reduced nitrogen retention, and increased water retention and fat content (Courtheyn et al. 2002; Baiocchi et al. 2003; Dusi et al. 2011). In the last few years, low levels of prednisolone in bovine (European Commission Staff Working Document 2010) and equine (Fidani et al. 2012) urine have been reported by several European Union member states. Low levels of prednisolone have also been found in porcine urine (Annual Report of the Belgian Hormone Commission 2011). Because prednisolone differs in structure from cortisol by only one double bond, the formation of prednisolone could resemble the process described for the formation of boldenone from testosterone (Blokland et al. 2007; Arioli et al. 2008) in the presence of faeces in the urine sample. The possibility of in vivo formation of prednisolone from endogenously present cortisol (Figure 1) has indeed been reported

606

Downloaded by [University of Victoria] at 03:32 26 February 2015

Figure 1.

E. de Rijke et al.

Molecular structures of (A) prednisolone and (B) cortisol.

(Arioli et al. 2010; Bredehöft et al. 2010; Cannizzo et al. 2011; Ferranti et al. 2011; Pompa et al. 2011; Fidani et al. 2012; Vicenti et al. 2012). The hypothesis that faecal contamination causes formation of prednisolone in urine is supported by the in vitro experiment of Arioli et al. (2010) in which an aqueous faecal solution was incubated with cortisol and cortisol-glucuronide, leading to deconjugation of the glucuronide and 1-delta dehydrogenation resulting in formation of unquantifiable traces of prednisolone. In the animal experiment with dexamethasone of Ferranti et al. (2011), prednisolone was found both in bovine urine samples from control animals taken at a slaughterhouse (56% positive at levels of 0.4–1.4 µg l–1) and among animals treated with dexamethasone (90% positive at levels of 0.4–1.5 µg l–1). The relation of stress to the formation of prednisolone from cortisol was investigated by Pompa et al. (2011) in an experiment with dairy cows (n = 3, 44– 106 months). Prednisolone was detected occasionally in unstressed situations. When stress was simulated by injection with tetracosactide hexaacetate, a synthetic analogue of adrenocorticotropic hormone, prednisolone was constantly found in concentrations between 1.01 and 4.08 µg l–1. This can be considered as proof that endogenous formation of prednisolone can occur. Another study into the formation of prednisolone from cortisol (Bredehoft et al. 2010) describes in vitro experiments with different bacterial strains that resulted in formation of prednisolone, e.g. with the soil bacteria Rhodococcus erythropolis. The first findings of prednisolone in the horse were described by Fidani et al. (2012). In 78% of the horse urine samples investigated prednisolone was detected in low concentrations, with an average of 1 µg l–1. Its origin may be endogenous but no proof for this was provided by the authors. Further, the 2011 annual report of the Belgian Hormone Commission (Hormonencel) (Annual Report of the Belgian Hormone Commission, 2011) stated that prednisolone was found in low concentrations in porcine urine, but no further scientific information was provided. In an Italian field survey by Vincenti et al. (2012) when the urine of 131 guaranteed untreated cows aged between 2.5 and 8 years from beef and dairy farms were investigated for prednisolone, no prednisolone (>0.7 µg l–1) was

found in 124 samples, and the remaining seven samples showed only traces of prednisolone between 0.1 and 0.3 µg l–1. Interestingly, in these seven samples the cortisol levels were also higher. Because the results in the literature discussed above are not always consistent and conclusive, in the current study the natural occurrence of prednisolone and the possibility of in vivo formation of prednisolone from endogenous cortisol was systematically investigated under different experimental conditions. This was done to study the contamination of urine samples with faeces or soil bacteria that could occur during sampling, due to faeces or soil on the skin of the animal which could then come in contact with the urine during excretion, or by ingestion of soil. For this purpose an experimental protocol was designed in which the factors most likely to cause prednisolone formation were taken into account, i.e. faecal and soil bacterial contamination, metabolic conversion and temperature. The current study provides information on: (1) the levels of prednisolone in randomly selected samples of bovine urine, (2) in vitro metabolism of cortisol under the influence of liver S9 fractions, (3) the degradation of cortisol and formation of prednisolone in bovine urine under the influence of soil bacteria or faeces, and (4) the formation of the other metabolites under the influence of faeces.

Materials and methods Analysis of 100 bovine urine samples From different cows in Dutch farms and slaughterhouses across the country in the annual control plan 100 bovine urine samples were collected: 33 urines were sampled directly from the bladder in different slaughterhouses and 67 urines were sampled at the farm during urination. The samples were analysed for the following corticosteroids: prednisolone, cortisol, cortisone, prednisone, isoflupredone, methylprednisolone, betamethasone, dexamethasone, flumethasone, beclomethasone, triamcinoloneacetonide and clobetasol. These were samples from the Dutch regulatory control programme and not from guaranteed non-treated animals.

Food Additives & Contaminants: Part A

Downloaded by [University of Victoria] at 03:32 26 February 2015

In vitro cortisol metabolism using incubations with bovine liver S9 fractions To investigate the formation of metabolites of cortisol, including the possible formation of prednisolone, first an in vitro experiment was performed. According to the method of Rijk et al. (2008), 250 μg cortisol were incubated with bovine liver S9 fraction. Incubations were carried out in a glass tube containing 0.1 M sodium phosphate buffer (pH 7.4), 33 mM potassium chloride, 8 mM magnesium chloride, 2 mg ml–1 bovine liver S9 fraction and 4 mM NADPH. The final volume was 1 ml; the mixture was incubated at 39°C in a water bath for 6 h and metabolic reactions were terminated at t = 0 and 6 h by adding 1 ml acetonitrile. Control incubations without the S9 enzyme fraction were included to monitor potential non-enzymatic reactions within the 6–h incubation period. The reaction mixture was centrifuged (3600g, 10 min, 4°C) and acetonitrile was evaporated from the supernatant (40°C). To the residue (approximately 0.9 ml) 18 µl acetic acid were added and the sample was mixed. The mixture was extracted twice with 500 µl ethyl acetate. After each extraction the mixture was centrifuged (2000g, 5 min). The organic layers were combined and evaporated (40°C) and re-dissolved in 100 µl water/acetonitrile, 90:10 containing 5 mM formic acid.

Cortisol degradation studies Three types of samples were used: an incurred urine sample (±50 µg cortisol l–1 and ±25 µg cortisone l–1), a cortisolfree blank urine sample (fortified with 125 µg l–1 cortisol) and a MilliQ water sample (fortified with 125 µg l–1 cortisol). The cortisol-free urine was selected from the set of 100 samples analysed, and contained no detectable amounts of cortisol. Fortified water samples were included to exclude other degradation mechanisms due to the urine matrix. The samples were incubated under three conditions: (1) by the addition of faecal suspension, (2) by the addition of a soil bacterial solution (R. erythropolis) or (3) as a control (no addition). For each treatment, three temperature regimes (20°C, 40°C and 70°C) were applied. This means that in total nine different experimental conditions were tested on each of the three sample types. The different incubation temperatures were selected to include RT (20°C), a physiologically relevant temperature for bacterial and enzymatic conversion (40°C), and the effect of (extreme) high temperature (70°C) that could occur during shipment or storage of the samples, or during sample treatment in the laboratory. The soil bacterial solution was prepared as follows: a Rhodococcus erythropolis strain (43066 GYM/MH 1 28/48 Micr 066; DSMZ, Braunschweig, Germany) was inoculated onto a blood plate and incubated for 24 h at 30°C. A single colony was transferred to 100 ml brain heart broth (BHI) in

607

duplicate and the plate count was determined after incubation for 3 days at 30°C. The incubation sample with the highest plate count (1.2 × 108 counts ml–1) was kept at 4°C until further use. An aliquot of 10 µl contained approximately 1 million soil bacteria. The faecal suspension was prepared using a method modified from Arioli et al. (2008, 2010). Briefly, rectal faeces from a veal calf was collected at the farm (Wageningen, the Netherlands), diluted 1:100 in 0.9% saline (NaCl) solution and shaken overnight at RT. The reason for diluting the faecal sample was to simulate conditions that could occur during faecal contamination while sampling urine from a cow, as the small amounts of faeces on the skin of the animal are diluted by the urine. To the blank urine, incurred urine and water samples fortified with 125 µg l–1 cortisol either 50 µl faecal suspension or 10 µl soil bacterial solution (R. erythropolis) was added per 2 ml of urine. Samples (2 ml) were taken from the incubates collected after 0, 1, 2, 4, 8, 16, 24, 30, 40, 50 and 64 h and from the control (blank urine fortified with 125 µg l–1 cortisol) after 0, 5, 10, 15 and 20 days. The incubated and control samples were heated at 80°C for 15 min to deactivate the soil bacteria and kept at −80°C until all the samples were collected and analysed. To check the cortisol stability during heating, the concentration was measured before and after heating at 80°C proving the stability of cortisol under these conditions. The samples were cleaned and concentrated according to the method described in urine sample preparation below. Urine sample preparation for UPLC-MS analysis An aliquot of 2 ml of urine was fortified with internal standards prednisolone-D8, cortisone-D8 and cortisol-D4, adjusted to pH 5.2 and hydrolysed with Suc d’Helix Pomatia juice (Merck, Darmstadt, Germany), for 2 h at 55°C. After hydrolysis, the sample was extracted twice with tert-butyl methyl ether (TBME) and dried. Acetonitrile and acetate buffer were added and the sample was transferred to a polymeric reversed-phase SPE cartridge. The cartridge was washed with methanol/NH4OH and water and dried under vacuum. After elution with MeOH, the extract was dried, dissolved in water/acetonitrile and centrifuged, 20 µl were analysed with UPLC-MS as described below. UPLC-QqQ-MS analysis The UPLC-QqQ-MS system consisted of an Acquity Ultra Performance LC (Waters, Milford, MA, USA) equipped with an Acquity UPLC BEH C18 column (1.7 µm, 100 × 2.1 mm) coupled to a Micromass, Quattro Ultima Platinum MS (Waters) with an ESI interface. Of each sample, 20 µl were injected onto the column and gradient elution was performed using a binary mobile phase of (A) 0.1% acetic acid in water and (B) 0.1% acetic acid in acetonitrile. The gradient was

608

E. de Rijke et al.

Downloaded by [University of Victoria] at 03:32 26 February 2015

Table 1. MS/MS corticosteroids.

fragmentation

conditions

for

selected

Compound

Precursor ion mass (m/z)

Product ion mass (m/z)

Collision energy (eV)

Prednisone

359.3

Prednisolone

361.3

Cortisol

363.3

Cortisone

361.3

Prednisolone-D8 Cortisone-D8 Cortisol-D4

369.3 369.3 367.3

237.3 323.3 343.3 325.2 121.2 327.2 163.1 121.1 351.2 168.2 121.2

16 8 8 8 20 13 23 28 8 23 20

0–0.5 min, 25% B; 0.5–0.7 min, linear increase to 30% B, with a hold of 2.3 min; 3.0–5.5 min, linear increase to 100% B and then a hold of 2.5 min. The LC flow was 0.4 ml min–1, the column temperature was set at 40°C and the vial tray temperature at 12°C. Chromatograms were recorded in positive ESI mode. The capillary voltage was 3.2 kV and the source and desolvation temperatures were 120 and 350°C, respectively. The desolvation and cone gas flows were set at 780 and 120 l h–1, respectively, and the CID gas was argon (p = 3*10−3 mbar; purity > 99.998%). SRM MS/MS chromatograms were recorded in positive ionisation mode for the different corticosteroids. The MS/MS transitions are presented in Table 1. With this method cortisol, cortisone, prednisone and prednisolone can be confirmed in bovine urine at a level of 0.25 µg l–1.

were set at respectively 150 and 450 l h–1, and the CID gas was argon (p = 2.2*10–3 mbar; purity > 99.998%).

Results Presence of prednisolone in bovine urine samples randomly collected at Dutch farms The analytical method was validated according to European Union criteria (Commission Decision 2002/ 657/EC), and resulted in decision limits (CCβ) of 0.13, 0.25 and 0.15 µg l–1 for prednisolone, cortisol and cortisone and detection capabilities (CCα) of 0.20, 0.40 and 0.24 µg l–1, respectively. Adequate precisions and intralaboratory reproducibility (RSD < 20%) were obtained for all compounds and linearity was > 0.99 for all compounds. The presence of prednisolone was confirmed in 40 samples, in concentrations between 0.11 and 2.04 µg l–1. Cortisol was confirmed in 58 samples (at concentrations between 0.27 and 122.0 µg l–1) and cortisone was confirmed in 40 samples (at concentrations between 0.06 and 70.1 µg l–1). Prednisone and the other synthetic corticosteroids were not found. The results are plotted in Figure 2. Cortisol and cortisone concentrations showed a correlation (Figure 2A), while no clear correlation was observed between prednisolone and cortisol. However, a clear difference can be seen between samples taken from slaughtered animals or living animals at the farm. Much higher levels of cortisol and cortisone were found in samples from slaughtered animals (a high stress level) than in samples taken at the farm (a low stress level). Prednisolone levels in the urine samples, however, are not influenced by the origin (slaughtered or alive) of the sample (Figure 2B).

UPLC-QToF-MS Twenty-four samples from the faeces experiments (samples at the beginning, middle and end of the decay curves of fortified and incurred urine with and without faeces at 20 and 70°C) were analysed on a UPLC-QToF-MS system to screen for metabolites based on their accurate mass. Identification was performed by a combination of a database search and comparison with published metabolite data. In the method, next to the compounds in Table 1, six other commercially available metabolites were included: 6α/βhydroxy-prednisolone, 20α/β-hydroxy-prednisolone, and 20α/β-hydroxy-prednisone (all from Steraloids, Newport, RI, USA). Of each sample extract 5 µl were injected onto an Acuity Ultra Performance LC system (Waters, San José, CA, USA) equipped with a Acquity UPLC BEH C18, (1.7 µm, 150 × 2.1 mm), coupled to a MicroTof-Q (Bruker Daltonics, Bremen, Germany) system. The LC conditions were the same as above. Chromatograms were recorded full scan in positive ESI mode. The capillary voltage was 2.5 kV and the source and desolvation temperatures were 120 and 450°C, respectively. The desolvation and cone gas flows

In vitro formation of prednisolone The bovine S9 incubation experiments with cortisol showed a clear decrease of cortisol and an increase of prednisolone after 6 h. The estimated cortisol concentration in the sample at t = 0 was 1.8 g l–1 and at t = 6 approximately 1.5 g l–1. This means that approximately 20% of the cortisol was converted. The estimated prednisolone concentrations were 0.2 g l–1 at t = 0 and 0.5 g l–1 at t = 6, corresponding to a formation of approximately 170%. In the control sample without S9 enzymes no prednisolone was observed, which indicates that the formation of prednisolone only occurs after addition of S9. Formation of prednisolone in urine samples under the influence of soil bacteria or faeces Control experiments The results of the cortisol degradation control experiments are summarised qualitatively in Table 2. Control experiments

Food Additives & Contaminants: Part A

609

140

Cortisol concentration (ppb)

120 y = 1.4341x R² = 0.6577

100 80 60 40 20 0

0

10

20

30

40

50

60

70

80

Cortison concentration (ppb)

Prednisolon concentration (ppb)

Downloaded by [University of Victoria] at 03:32 26 February 2015

2.5 2.0 1.5 1.0 0.5 00 0.0

0

20

40

60

80

100

120

140

Cortisol concentration (ppb)

Figure 2. Concentrations of (A) cortisol versus cortisone (y = 1.4341x and R2 = 0.6577); and (B) prednisolone versus cortisol in bovine urines randomly collected in the Netherlands. Black markers indicate urine samples collected at the farm; and grey markers indicate urines collected at the slaughterhouse.

were performed in order to assess the thermal stability of cortisol in water and the effect of the addition of soil bacteria or faeces. From the results included in Table 2 it is concluded that cortisol is stable in water at temperatures up to 70°C for at least 40 h. The presence of soil bacteria showed a significant influence on the stability of cortisol and cortisone, especially at 20°C. This influence is most clearly visible when cortisol is added to water. Without the addition of soil bacteria cortisol proved to be stable, after the addition of soil bacteria the estimated half-life of cortisol is approximately 50 h. At higher temperatures the half-life of cortisol shortens, but the differences between the control samples become less. Quantitatively, no clear correlation was observed between the rate of decay and the temperature. The addition of faeces had no influence on the levels of cortisol and cortisone, irrespective the temperature. Soil bacteria The results for the fortified and incurred samples are shown in Figure 3. The fortified and incurred urine samples at 20 and 40°C showed a gradual decrease of cortisol (approximately 10–20%) up to 50 h (Figure 3B, C, E, F).

However, the same trend was observed for the fortified urine experiments at 40°C without additives, while in the same experiment at 20°C cortisol surprisingly showed a linear decrease of 50% over 20 h (Figure 3A). In the experiments at 70°C an exponential decrease to zero was observed over 50 h (Figure 3G, H, I). Formation of prednisolone was observed in incurred urine in the presence of soil bacteria at all investigated temperatures (Figure 4C). At 20°C, prednisolone was formed up to 7 µg l–1 after 64 h, at 40°C a constant level of approximately 2.5 µg l–1 prednisolone was formed after 10 h, and at 70°C formation of prednisolone up to 4 µg l–1 was observed, followed by a decrease to zero after 40 h. However, a similar trend for prednisolone was observed in incurred urine without additives (Figure 4B), which may indicate that the formation of prednisolone is related to temperature rather than bacteria. Unfortunately, the experiment with incurred urine without additives at 20°C was stopped after 20 h, therefore it is not certain whether without soil bacteria also relatively high levels of prednisolone will be observed after 40–60 h. Formation of prednisolone was not observed in the spiked urine samples incubated with soil bacteria. This is

610 Table 2.

E. de Rijke et al. Degradation profiles of cortisol in water under different experimental conditions.

Temperature (°C) 20 40

Downloaded by [University of Victoria] at 03:32 26 February 2015

70

Treatment

Trend in cortisol concentration

No addition Faeces Bacteria No addition Faeces Bacteria No addition Faeces Bacteria

No significant change over 500 h No significant change over 100 h Linear decrease, level two-fold lower after 50 h No significant change over 50 h No significant change over 50 h Linear decrease to zero over 50 h No significant change over 50 h No significant change over 50 h Linear decrease of 25% after 50 h

Figure 3. Concentrations of cortisol and cortisone (µg l–1) monitored over time at 20°C, 40°C or 70°C in urine fortified with 125 µg l–1 cortisol (A, D, G), urine with soil bacteria solution fortified with 125 µg l–1 cortisol (B, E, H), or incurred urine with soil bacteria solution (C, F, I).

in contrast to cortisol degradation profiles, which are the same in fortified urine and incurred urine at all temperatures (Figure 3). Formation of prednisone was not observed in any of the experiments.

Faeces The influence of the presence of faeces is demonstrated in Figure 5. At 20°C there seems to be an initial period with very limited decay of cortisol, but after 20 h the speed of decay increased rapidly (Figure 5B, C). At 40°C the results seem similar, but the decrease after 20 h seems less in the spiked samples (Figure 5E). At 70°C an exponential decay is observed, also in the control samples (Figure 5G, H, I), indicating that the exponential

degradation of cortisol at 70°C is most likely caused by temperature rather than by the influence soil bacteria or faeces. Results for incurred cortisol and cortisone are similar (Figure 5C, F, I). At 70°C low levels of prednisolone were formed in the incurred urines incubated with faeces (Figure 4D). The highest level of prednisolone was found after 15 h (4.1 µg l–1) and after approximately 50 h the level decreased again to zero. In the fortified urine samples incubated with faeces no prednisolone was detected; this was also observed for the soil bacteria experiments and may be explained by the presence of natural corticosteroids in the incurred sample that are not present in the fortified sample. Formation of prednisone was not observed in any of the experiments.

Downloaded by [University of Victoria] at 03:32 26 February 2015

Food Additives & Contaminants: Part A

611

Figure 4. Concentrations of prednisolone (µg l–1) monitored over time at 20°C, 40°C or 70°C in spiked or incurred urine with (A, B) no added bacteria or faeces, (C) added soil bacteria, and (D) added faeces.

Figure 5. Concentrations of cortisol and cortisone (µg l–1) monitored over time at 20°C, 40°C or 70°C in urine fortified with 125 µg l–1 cortisol (A, D, G), urine with faeces suspension fortified with 125 µg l–1 cortisol (B, E, H) or incurred urine with faeces suspension (C, F, I).

Identification of other metabolites using targeted metabolomics The incubated urine samples were investigated on UPLC-QToF-MS to identify possible metabolites. Samples were collected at the beginning, middle and end (approximately t = 2, 20 and 50 h) of the incubation

period for incurred urine without additives, and with soil bacteria or faeces at 20 and 70°C (Figures 4C, I and 3B, E). In the literature several metabolite pathways for corticosteroids are described, such as (de)conjugation, oxidation at C-11, reduction at C-3 and/or C-20 and/or C-21 and formation of ring A saturated steroids (Palme et al.

Downloaded by [University of Victoria] at 03:32 26 February 2015

612

E. de Rijke et al.

1996). For cortisol, these reactions generate metabolites such as cortisol-21-glucuronide, cortisol-21-sulfate, cortisone-21-sulfate, tetrahydrocortisone-3-glucuronide, deoxycorticosterone-21-glucoside (Antignac et al. 2002), and 11,17-dioxoandrostanes such as 11-oxoetiocholanolone (Möstl et al. 1999, 2002). Major hydroxylated metabolites of prednisone and prednisolone found in human urine were 20β-hydroxy-prednisone, 6β-, 20α- and 20βhydroxy-prednisolone (Garg & Jusko 1991). The incubated samples were screened for all these metabolites and compared with the samples without additives. In the fortified water samples incubated with faeces an increase of m/z 363.3 was observed, which was confirmed as 20β-hydroxy-prednisolone with a commercially available standard. Concentrations ranged from 0 μg l–1 at t = 0 h up to approximately 0.8 μg l–1 at t = 50 h. In fortified urine with faeces none of these hydroxylated prednisone and prednisolone metabolites was detected. In the incurred urine incubated with faeces 20α-hydroxy-prednisolone, 20β-hydroxy-prednisolone and 20β-hydroxy-prednisone were confirmed, but their concentrations (i.e. approximately 0.1, 0.1 and 0.5 μg l–1, respectively) remained constant during the time of incubation. Furthermore, an increase of two peaks with m/z 441.2 and 201.1 was observed; they most likely correspond to cortisol-21-sulfate and oxydianiline, respectively, but no standards were available for confirmation. Also, an increase of m/z 305.2 was observed, but this peak could not be (tentatively) identified, and two peaks, m/z 539.2 and m/z 347.2, showed a clear decrease and may correspond to tetrahydrocortisone-3-glucuronide and cortexolone. Furthermore, an increase of m/z 289.2 and 361.2, corresponding respectively to testosterone and dromostanolone, was observed after incubation, but they were not considered to be cortisol metabolites. The abovementioned compound trends were observed at all temperatures investigated, but were most pronounced at 40°C.

Conclusions Prednisolone was detected in 40% of the bovine urine samples randomly collected within Dutch routine monitoring programmes, showing concentrations between 0.11 and 2.04 µg l–1. It was shown that this high incidence of low levels of prednisolone may be caused by conversion from cortisol. In vitro bovine liver S9 incubation experiments showed a 20% decrease of cortisol within 6 h, and formation of prednisolone was observed from 0.2 g l–1 at t = 0 h to 0.5 g l–1 at t = 6 h. Under the influence of faeces, cortisol in incurred urine breaks down within 50 h. The addition of faeces clearly increased the speed of degradation. However, the presence of soil bacteria did not seem to have any influence. At all investigated temperatures, formation of prednisolone was observed: up to 4.1 µg l–1 at 70°C after 15 h, but decreasing after 40 h to zero.

In incurred urine incubated with soil bacteria, a slower decrease of cortisol was observed than with faeces incubation, but slightly higher overall formation of prednisolone was observed at all temperatures: up to 7 µg l–1 at 20°C. In fortified urine incubated with faeces or soil bacteria no prednisolone was detected. This difference may be explained by the fact that the incurred urine contains other natural corticosteroids that may contribute to the prednisolone catabolism. With UPLC-QToF-MS experiments, metabolites known from literature could be (tentatively) identified as 20β-hydroxy-prednisolone, cortisol-21-sulfate, oxydianiline, tetrahydrocortisone-3-glucuronide and cortexolone, but for all compounds except 20β-hydroxy-prednisolone no standards were available for confirmation. Although the experiments described do not give a complete insight into the mechanism for degradation of cortisol and formation of prednisolone, it was demonstrated here that prednisolone can be formed in vitro in samples containing endogenous levels of cortisol. The formation of prednisolone is enhanced by the presence of faeces in the sample. It is proposed to establish a threshold level for findings of prednisolone for regulatory purposes based on the following calculation: 1.45 (mean of 100 urine samples) ± 3*1.08 (3*SD) = 4.69 µg l–1. Based on the results of this study and previously published data, for regulatory control purposes a threshold level for findings of prednisolone in bovine urine of 5 µg l–1 is proposed. Findings of prednisolone in concentrations up to 5 µg l–1 in bovine urine most likely originate from other sources than illegal treatment with growth promoters. Additional research to understand further the mechanism of formation will be carried out in the near future. Currently, quantitative proficiency tests are being conducted and isotope ratio MS studies are being performed to differentiate unambiguously between endogenous and exogenously administered natural hormones, including corticosteroids.

Funding The financial support of the EC DG SANCO and the Dutch Ministry of Economic Affairs, Agriculture and Innovation [project 72708.01] for the EU Reference Laboratory tasks of RIKILT, within which framework this study was undertaken is greatly acknowledged.

References Annual Report of the Belgian Hormone Commission. 2011. Multidisciplinary Belgian Hormone Commission. [Internet]. [cited 2013 Aug 15]. Available from: http://issuu.com/fedpolbelgium/docs/jaarverslag_mdhc_2011 Antignac JP, Le Bizec B, Monteau F, André F. 2002. Study of natural and artificial corticosteroid phase II metabolites in bovine urine using HPLC-MS/MS. Steroids. 67:873–882. Arioli F, Fidani M, Casati A, Fracchiolla ML, Pompa G. 2010. Investigation on possible transformations of cortisol,

Downloaded by [University of Victoria] at 03:32 26 February 2015

Food Additives & Contaminants: Part A cortisone and cortisol glucuronide in bovine faecal matter using liquid chromatography–mass spectrometry. Steroids. 75:350–354. Arioli F, Gavinelli MP, Fracchiolla ML, Casati A, Fidani M, Ferrer E, Pompa G. 2008. Evaluation of boldenone formation and related steroids transformations in veal faeces by liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom. 22:217–223. Baiocchi C, Brussino M, Pazzi M, Medana C, Marini C, Genta E. 2003. Separation and determination of synthetic corticosteroids in bovine liver by LC-ion-trap-MS-MS on porous graphite. Chromatographia. 58:11–14. Blokland MH, van Rossum HJ, Sterk SS, van Ginkel LA, Stephany RW. 2007. Development of a method which discriminates between endogenous and exogenous β-boldenone. Anal Chim Acta. 586:147–153. Bredehöft M, Baginski R, Parr MK, Thevis M, Schänzer W. 2010. Investigations of the microbial transformation of cortisol to prednisolone in urine samples. J Steroid Biochem Mol Biol. 129:54–60. Cannizzo FT, Capra P, Divari S, Ciccotelli V, Biolatti B, Vincenti M. 2011. Effects of low-dose dexamethasone and prednisolone long term administration in beef calf: chemical and morphological investigation. Anal Chim Acta. 700:95–104. Commission Decision 2002/657/EC. Commission Decision of 12 August 2002 implementing Council Directive 96/23/EC concerning the performance of analytical methods and the interpretation of results [Internet]. [cited 2013 Aug 15]. Available from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do? uri=OJ:L:2002:221:0008:0036:EN:PDF European Union. 2009. Commission regulation (EU) No 37/ 2010 of 22 December 2009 on pharmacologically active substances and their classification regarding maximum residue limits in foodstuffs of animal origin. Off J Eur Union. L15:1–72. European Union. 1996. Council Directive 96/23/EC of 29 April 1996 on measures to monitor certain substances and residues thereof in live animals and animal products and repealing Directives 85/358/EEC and 86/469/EEC and Decisions 89/ 187/EEC and 91/664/EEC. Available from: http://eur-lex. europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:1996L 0023:20090807:EN:PDF Courtheyn D, LeBizec B, Brambilla G, De Brabander HF, Cobbaert E, Van de Wiele M, Vercammen J, DeWasch K. 2002. Recent developments in the use and abuse of growth promoters. Anal Chim Acta. 473:71–82. CRL 2007. CRLs view on state of the art analytical methods for national residue control plans. [Internet; cited 2013 Aug 15]. Available from: http://www.bvl.bund.de/SharedDocs/

613

Downloads/09_Untersuchungen/EURL_Empfehlungen_Konz entrationsauswahl_Methodenvalierungen.pdf?__blob=publica tionFile Dusi G, Gasparini M, Curatolo M, Assini W, Bozzoni E, Tognoli N, Ferretti E. 2011. Development and validation of a liquid chromatography–tandem mass spectrometry method for the simultaneous determination of nine corticosteroid residues in bovine liver samples. Anal Chim Acta. 700:49–57. European Commission Staff Working Document. 2010. Commission staff working document on the implementation of national residue monitoring plans in the member states in 2010 (Council Directive 96/23/EC) [Internet]. [cited 2013 Aug 15]. Available from: http://ec.europa.eu/food/food/ chemicalsafety/residues/docs/workdoc_2010_en.pdf Ferranti C, delli Quadria F, Palleschi L, Marchiafava C, Pezzolato M, Bozzetta E, Caramelli M, Draiscia R. 2011. Studies on the presence of natural and synthetic corticosteroids in bovine urine. Steroids. 76:616–625. Fidani M, Pompa G, Mungiguerra F, Casati A, Fracchiolla ML, Arioli F. 2012. Investigation of the presence of endogenous prednisolone in equine urine by high-performance liquid chromatography mass spectrometry and high-resolution mass spectrometry. Rapid Commun Mass Spectrom. 26:879–886. Garg V, Jusko WJ. 1991. Simultaneous analysis of prednisone, prednisolone and their major hydroxylated metabolites in urine by high-performance liquid chromatography. J Chromatogr B. 567:39–47. Möstl E, Maggs JL, Schrötter G, Besenfelder U, Palme R. 2002. Measurement of cortisol metabolites in faeces of ruminants. Vet Res Commun. 26:127–139. Möstl E, Messmann S, Bagu E, Robia C, Palme R. 1999. Measurement of glucocorticoid metabolite concentrations in faeces of domestic livestock. Zentralblatt fur Veterinarmedizin Reihe A. 46:621–631. Palme R, Fischer P, Schildorfer H, Ismail MN. 1996. Excretion of infused C-14-steroid hormones via faeces and urine in domestic livestock. Anim Reprod Sci. 43:43–63. Pompa G, Arioli F, Casati A, Fidani M, Bertocchi L, Dusi G. 2011. Investigation of the origin of prednisolone in cow urine. Steroids. 76:104–110. Rijk JCW, Bovee TFH, Groot MJ, Peijnenburg AAC, Nielen MWF. 2008. Evidence of the indirect hormonal activity of prohormones using liver S9 metabolic bioactivation and an androgen bioassay. Anal Bioanal Chem. 392:417–425. Vincenti M, Leporati M, Capra P, Gatto AA, Barbarino G, Nebbia C. 2012. A field survey on the presence of prednisolone and prednisone in urine samples from untreated cows. Food Addit Contam Part A. 29:1–8.