Confirmatory analysis of in muscle using liquid chromatography - tandem mass spectrometry Mikael Pedersen, Jens Hinge Andersen

To cite this version:

Mikael Pedersen, Jens Hinge Andersen. Confirmatory analysis of steroids in muscle using liq- uid chromatography - tandem mass spectrometry. Food Additives and Contaminants, 2011, pp.1. ￿10.1080/19440049.2010.549153￿. ￿hal-00671233￿

HAL Id: hal-00671233 https://hal.archives-ouvertes.fr/hal-00671233 Submitted on 17 Feb 2012

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Food Additives and Contaminants

For Peer Review Only

Confirmatory analysis of steroids in muscle using liquid chromatography – tandem mass spectrometry

Journal: Food Additives and Contaminants

Manuscript ID: TFAC-2010-345.R1

Manuscript Type: Original Research Paper

Date Submitted by the 09-Dec-2010 Author:

Complete List of Authors: Pedersen, Mikael; DTU Food, Food Chemistry Andersen, Jens; DTU Food, Food Chemistry

Methods/Techniques: Chromatography - LC/MS, Method validation

Additives/Contaminants: Veterinary drug residues - anabolic steroids

Food Types: Animal products – meat

http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 1 of 23 Food Additives and Contaminants

1 2 3 4 5 1 Confirmatory analysis of steroids in muscle using liquid chromatography – 6 7 2 8 tandem mass spectrometry 9 10 3 11 12 4 Mikael Pedersen and Jens Hinge Andersen 13 14 15 5 16 For Peer Review Only 17 6 National Food Institute (DTU-FOOD), Mørkhøj Bygade 19, DK-2860 Søborg, Denmark 18 19 20 7 21 22 8 Abstract: 23 24 9 A method is described for screening and confirmation of synthetic and endogenous steroids in 25 26 27 10 muscle tissue. The method is sensitive, selective, rapid and the consumption of organic solvents is 28 29 11 low, compared to previously published methods. The procedure involves hydrolysis, defattening 30 31 12 with heptane and final clean up with SPE using C18 cartridge. After filtration the analytes are 32 33 34 13 analysed by LC/MS/MS, and quantification is performed using deuterated internal standards. 35 36 14 Decision limits (CC α) varied from 0.02 to 0.33 g kg -1 and the detection capabilities (CC β) were < 37 38 -1 39 15 0.50 g kg . The mean within-laboratory reproducibility ranged from 5 – 22% (%RSD IR ). 40 41 16 Endogenous steroids (ex. , and ), have been included in 42 43 17 the method, to provide insights into the levels of these steroids as findings of these endogenous 44 45 46 18 steroids have been made several times during the period where analysis has been made of imported 47 48 19 meat. 49 50 51 20 52 21 Keywords: Steroids, LC-MS/MS, Muscle, Method validation, LC-TOFMS, Import control 53 22 54 55 23 Introduction 56 57 58 24 The use of steroids is forbidden for fattening purposes in the European Community [1]. 59 60 25 The most suitable matrices for control of abuse is often , faeces, plasma and kidney fat, but for

1 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 2 of 23

1 2 3 4 5 26 control of abuse of these compounds in import from third countries to the EU, these matrices are not 6 7 27 available. Therefore a sensitive and specific method for screening and confirmation of these 8 9 28 substances in muscle, has been developed. In 2007 Denmark imported 6300 tons of beef directly 10 11 12 29 from third countries (primarily South America and ) corresponding to app. 5% of beef 13 14 30 consumed in Denmark. 15 16 31 For Peer Review Only 17 18 -1 19 32 Since the European Commission has set 1 g kg as the recommended concentrations for most 20 21 33 steroids in muscle tissue [2] a very sensitive method has to be used and it must be in compliance 22 23 24 34 and validated in accordance with the criteria of the Commission Decision 2002/657/EC [3]. 25 26 35 A confirmatory method has been developed for the determination of acetate, 27 28 36 acetate, acetate, acetate, , α/β- 29 30 31 37 , methylboldenone, α/β -trenbolone, α/β-nortestosterone, , testosterone, 32 33 38 epitestosterone, androstenedione and clostebolacetate (screening only) in muscle from bovine. The 34 35 36 39 steroids represent both gestagens and and both synthetic and endogenous steroids have 37 38 40 been included. The method has the advantages that the amount of organic solvents consumed is 39 40 41 very low, no evaporation steps are needed and because of appropriate cleanup and reduced amount 41 42 43 42 of sample material, the matrix effect is low. Since 2007, analysis for steroids in meat from third 44 45 43 countries has been included in the Danish residue plan and the developed method is part of the 46 47 48 44 flexible accreditation our laboratory has been assigned by the Danish Accreditation and Metrology 49 50 45 Fund - DANAK [4]. As a consequence of the survey on the use of veterinary medicinal products in 51 52 46 third countries prepared in 2009 [5], the substances altrenogest and clostebolacetat were included in 53 54 55 47 the method. Clostebolacetate, though, was not included in the accreditation because some criteria 56 57 48 regarding confirmation was not fulfilled. 58 59 49 60

2 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 3 of 23 Food Additives and Contaminants

1 2 3 4 5 50 The procedure involves hydrolysis, defattening with heptane, clean up with SPE using C18 6 7 51 cartridges and final filtration. Analytes were analysed using liquid chromatography (LC) coupled 8 9 52 with tandem mass spectrometry (MS 2), and quantification was performed using spiked standards in 10 11 12 53 combination with deuterated internal standards. Two transitions were monitored for each 13 14 54 analyte. The decision limits (CC α) and detection capabilities (CC β) were well below the 15 16 For Peer Review-1 Only 17 55 recommended concentration which is set at 1 g kg . The mean within-laboratory reproducibility 18 19 56 ranged from 5 to 22% (%RSD IR ). The method is sensitive and reliable and has been used for import 20 21 57 control during the last two years. In addition to the detection and quantification by LC-MS/MS, 22 23 24 58 detection by exact mass measurement using time-of-flight mass spectrometry (LC-TOFMS) has 25 26 59 been demonstrated. The three endogenous steroids testosterone, epitestosterone and 27 28 29 60 androstenedione, have been included in the method to get knowledge about the levels of these 30 31 61 steroids and findings of these steroids have been made several times. A fractional factorial design 32 33 62 was used to evaluate the performance of the method when introducing changes in possible critical 34 35 36 63 factors. 37 38 39 64 Materials and methods 40 41 42 65 Materials and reagents 43 66 All reagents were of analytical or HPLC grade and supplied by Merck (Darmstadt, Germany) 44 45 67 and Rathburn Chemicals (Walkerburn, Scotland). Water was ultra-purified using a Maxima 46 47 48 68 purification system from USF Elga (Bucks, UK). Protease Type VIII from B. Licheniformis was 49 50 69 from Sigma-Aldrich (St. Louis, MO, USA). Medroxyprogesteronacetate (MPA), Megestrolacetate 51 52 70 (MGA), Chlormadinonacetate (CMA), Methyltestosterone (MT), β-Nortestosterone (BNO), 53 54 55 71 Testosterone (TES), epitestosterone (ETE), Androstenedione (ASD), Clostebolacetate (CLAc) and 56 57 72 Altrenogest (ALTR) were purchased from Sigma-Aldrich, Melengestrolacetate (MLA) was 58 59 60 73 purchased from Steraloids Inc. (Newport, USA), α-Trenbolon (ATR) was bought from LGC

3 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 4 of 23

1 2 3 4 5 74 PromoChem (National Analytical Reference Laboratory, Australia), Methylboldenon (MBO), α- 6 7 75 Boldenon (ABO), β-Boldenon (BBO), β-Trenbolon (BTR) and α-Nortestosteron (ANO), 8 9 10 76 Medroxyprogesteronacetate-d3 (MPA-d3), Megestrolacetate-d3 (MGA-d3), Melengestrolacetate-d3 11 12 77 (MLA-d3), Testosteron-d2 (TES-d2), Methyltestosteron-d3 (MT-d3), β-Boldenone-d3 (BBO-d3), 13 14 15 78 Methylboldenone-d3 (MBO-d3), β-Nortestosteron-d3 (BNO-d3) and β-Trenbolon-d2 (BTR-d2) were 16 For Peer Review Only 17 79 purchased from RIVM (Bilthoven, The Netherlands). Stock solutions of each compound were 18 19 80 prepared separately in ethanol at concentrations of 1 mg mL -1. Stock solutions of internal standards, 20 21 22 81 however, were prepared by reconstitution of ampoules containing a fixed amount of material with 23 24 82 ethanol at a concentration of 0.1 mg mL -1. Stock solutions were stored at –18 °C until use for up to 25 26 -1 27 83 3 year. A final working solution containing all the compounds at 0.1 g mL , except internal 28 29 84 standards, was prepared by diluting stock solutions with ethanol. A final working solution of 30 31 -1 32 85 internal standards containing all nine internal standards at 0.1 g mL was also prepared by diluting 33 34 86 stock solutions with ethanol. Final working solutions were freshly made. Samples were cleaned up 35 36 87 on C18 SPE-columns (500 mg, 3 mL, Waters Corp. Milford, MA, USA) and filtrated through NH2 37 38 39 88 SPE columns (500 mg, 5 mL, Waters Corp.). 40 41 89 Samples 42 43 90 Bovine muscle was either collected at slaughterhouses, as part of the national plan for 44 45 46 91 monitoring drug residues in animals and animal products, or samples were obtained from local 47 48 92 supermarkets. As soon as the samples were received, fat and tendons were removed, they were 49 50 51 93 homogenized (appr. 100 g) in a blender and stored at –18 °C until analysis. Poultry muscle, used for 52 53 94 the factorial design, was bought from local supermarkets and treated as bovine muscle. 54 55 95 Sample preparation 56 57 58 96 Muscle tissue was homogenised, and 1 g was transferred to a centrifuge tube. 20 µL internal 59 60 97 standard working solution (0.1 µg mL -1) and 2 mL of freshly prepared Protease solution (1 mg mL -1

4 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 5 of 23 Food Additives and Contaminants

1 2 3 4 5 98 in Tris buffer, pH 9.5) was added and the sample was vortexed. The sample was then stored at 50 6 o 7 99 C for two hours (every 30 min. the sample was vortexed for one min.) or 3 hours during factorial 8 9 100 experiment. The sample was extracted twice with 2 mL of methanol. The sample was washed with 10 11 12 101 3 mL of heptane and 15 mL of water was added. The sample was applied to the C18 SPE-column 13 14 102 (500 mg) pre-activated with 3 mL of methanol and 3 mL of water. After washing with 2×2.5 mL of 15 16 103 water, 2×2.5 mL ofFor acetone/water Peer (20/80 v/v)Review and 2×2.5 mL heptane Only the column was vacuum-dried 17 18 19 104 for 5 min. 20 21 105 The analytes were eluted from the column with 5 mL of acetonitrile and filtered through a NH2 22 23 106 SPE-column (500 mg) from Waters, pre-activated with 5 mL of acetonitrile. Further 2 mL of 24 25 26 107 acetonitrile was filtered through the NH2 SPE-column. The extract (app. 7 mL) was dried using a 27 28 108 nitrogen evaporator at 40 °C. The residue was reconstituted with 200 µL of acetonitrile/water/acetic 29 30 31 109 acid (35/65/0.1). 32 33 110 LC-MS/MS analysis 34 35 111 HPLC separation was performed on an Agilent 1100 series LC system from Agilent 36 37 38 112 Technologies (Palo Alto, CA, USA) equipped with a high-pressure binary pump, degasser, auto 39 40 113 sampler and a column heater. The steroids were separated on an Atlantis C18 column 41 42 43 114 (2.1mm*150mm, 3µm) from Waters Corp. at 40 °C using a gradient method. Mobile phase A 44 45 115 contained a mixture of 0.1% acetic acid in water and acetonitrile (65:35, v/v) and mobile phase B 46 47 116 contained acetonitrile , and the flow rate was set at 0.2 mL min -1. The sample volume injected was 48 49 50 117 50 µL. 51 52 118 The mass spectrometer used was a Quattro Ultima Pt triple quadropole instrument (Waters 53 54 119 Corp.) with Masslynx v. 4.0 software (Waters Corp.). Ionisation of the analytes was achieved using 55 56 + 57 120 an electrospray interface in the positive ion mode (ESI ), and ionisation source parameters were as 58 59 121 follows: capillary voltage (V cap ), 3.5 kV; cone voltage (V cone ), 40 V; RF Lens 1 voltage (V Ref1 ), 15 60

5 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 6 of 23

1 2 3 4 5 122 V; desolvation temperature, 400°C; source temperature, 120°C. Nitrogen was used as the nebulising 6 -1 7 123 gas (maximum flow), desolvation gas (flow-rate of 600 L h ), and the cone gas (flow-rate of 120 L 8 9 124 h-1). Argon was used as the collision gas at a pressure of ~ 2.2 × 10 -3 mbar. Data acquisition was 10 11 12 125 performed in the multiple reaction monitoring (MRM) mode. MRM transitions and collision 13 14 126 energies for the analytes are listed in Table 1, and the dwell time was set to 0.2-0.3 s for trenbolon 15 16 For Peer Review Only 17 127 and 0.1 s for other transitions. 18 19 128 Spiked standards were used to quantify the analytes. The fortification levels of the validation 20 21 129 curves were: 0.0 (blank sample), 0.125, 0.25, 0.5, 1.0, 2.0 4.0, 8.0 g kg -1, and a fixed amount of 22 23 -1 24 130 internal standard (0.5, 2.0 or 4.0 g kg depending on the ) was added to all the samples. 25 26 131 Calibration curves were obtained by plotting response of the analyte versus nominal concentrations 27 28 -1 29 132 (ng mL ) added of the analyte and response is defined as the chromatographic peak area of the 30 31 133 analyte divided by the area of the related internal standard, multiplied with concentration of internal 32 33 134 standard. Deuterated labelled standards are suitable for quantification purposes and if a deuterated 34 35 36 135 standard was not available for a steroid a deuterated epimer standard was used or a deuterated 37 38 136 standard with a matching retention time was used. Therefore MLA-d3 was used as an internal 39 40 137 standard for CMA and CLAc, BBO-d was used as an internal standard for ABO (epimer), BTR-d 41 3 2 42 43 138 was used as an internal standard for ATR (epimer), BNO-d3 was used as an internal standard for 44 45 139 ANO (epimer) and MT-d3 was used as an internal standard for ASD, ALTR and ETE. 46 47 48 140 UHPLC-TOFMS analysis 49 50 141 UHPLC was performed on a Waters Acquity UHPLC system (Waters Corp.) equipped with a 51 52 142 high-pressure binary pump, degasser, autosampler and a column heater. 53 54 55 143 The steroids were separated on an Acquity UPLC BEH C18 column (2.1mm*50mm, 1.7µm) from 56 57 144 Waters at 40 °C using isocratic conditions. Mobile phase contained a mixture of 0.1% acetic acid 58 59 60

6 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 7 of 23 Food Additives and Contaminants

1 2 3 4 -1 5 145 in water and methanol (35:65, v/v), and the flow rate was set at 0.25 mL min . The sample volume 6 7 146 injected was 5 µL. 8 9 147 The mass spectrometer used was an LCT Premier TM /XE time-of-flight mass spectrometer 10 11 12 148 (Waters Corp.) with Masslynx v. 4.1 software (Waters Corp.). Ionisation of the analytes was 13 14 149 achieved using an electrospray interface in the positive ion mode (ESI +) and ionisation source 15 16 150 parameters were asFor follows: capillaryPeer voltage, Review 3.0 kV; desolvation Only temperature, 350°C; source 17 18 -1 19 151 temperature, 120°C; cone voltage, 30 V; desolvation gas (flow-rate of 700 L h ), and the cone gas 20 21 152 (flow-rate of 10 L h -1). Data were collected in W-optics centroid mode over a mass range of 70- 22 23 153 1000 Da with a scan time of 0.1 s. For the dynamic range enhancement (DRE) lockmass, a solution 24 25 -1 -1 26 154 of leucine-enkephalin at 50 g L was infused through an infusion pump at 10 L min . 27 28 29 155 Results and discussion 30 31 32 33 156 Method development 34 35 157 The method developed by Blasco et al [6] was selected as the starting point for method 36 37 38 158 development. It was decided not to include a hydrolysis step as less than 20% of testosterone and 39 40 159 less than 5% of β-oestradiol was in the conjugated form in muscle tissue [7-9]. The tissue was 41 42 43 160 initially enzymatically degraded by Protease Type VIII, leading to a homogenious suspension. The 44 45 161 MS/MS fragmentation observed for the steroids was similar to that reported in earlier studies [6, 10- 46 47 162 14]. Precursor and daughter selected for LC-MS/MS quantification and confirmation are 48 49 50 163 shown in Table 1. Due to the natural isotopic abundances of carbon and hydrogen, certain MRM 51 52 164 transitions produce a signal for both MPA and MGA. Use of 65% methanol as mobile phase and a 53 54 165 Symmetry LC-column, under isocratic conditions, was presented by S. Mortensen et al [15] for 55 56 57 166 separating these transitions. We have used acetonitrile under gradient conditions to achieve full 58 59 167 separation of MGA and MPA. Likewise the close eluting isomers of trenbolon ( α- and β-trenbolon) 60

7 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 8 of 23

1 2 3 4 5 168 were separated using the same gradient. Ion suppression due to matrix effect can be a serious 6 7 169 problem when using ESI ionization [16] and is primarily caused by the presence of endogenous 8 9 170 substances [17- 18]. Ion suppression was investigated during method development and it was found 10 11 12 171 to be less than 20 % for most steroids and less than 35% for the gestagens. It was shown that 13 14 172 increasing the amount of sample (and solvents used for cleanup), resulted in more ion suppression. 15 16 173 It was therefore decidedFor to usePeer only 1 g of tissueReview sample to reduce Only ion suppression. When using 1 g 17 18 19 174 of sample it was likewise possible to decrease the amount of organic solvents. Often dilution of the 20 21 175 sample or injecting less volume is used for reducing matrix effect, but this would in this study also 22 23 176 increase the detection limit which is not desirable. 24 25 26 177 Method validation 27 28 178 The method was initially validated for the determination of 15 androgens in bovine muscle tissue 29 30 31 179 according to Commission Decision 2002/657/EC [3] to be used for control of muscle tissue 32 33 180 imported to the domestic market. Later altrenogest and clostebolacetate were included. Analytes 34 35 181 were quantified using spiked standards, i.e. muscle samples spiked with analytes and deuterated 36 37 38 182 internal standards prior to sample preparation. For the endogenous steroids (testosterone, 39 40 183 epitestosterone and androstenedione) external calibration curve was used. The calibration standards 41 42 184 were measured before and after the test samples and determination coefficients (R 2) were larger 43 44 45 185 than 0.95 for all analytes. For the determination of specificity 20 different blank muscle samples 46 47 186 were analysed. No interferences were observed in the region of the retention times for any of the 48 49 -1 50 187 steroids except melengestrolacetate where interferences occurred in some samples (< 0.2 g kg ) 51 52 188 and since the ion ratio was more than 10 times higher than the target value it was concluded that it 53 54 189 was not due to carry over or possible treatment of the animal but “noise”. MRM chromatograms of 55 56 57 190 a blank muscle sample fortified with all analytes are shown in figure 1. 58 59 60

8 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 9 of 23 Food Additives and Contaminants

1 2 3 4 5 191 Precision (repeatability and intra-laboratory reproducibility) and trueness were determined by 6 7 192 spiking blank muscle tissue samples with the steroids at four concentration levels. Repeatability and 8 9 193 intra-laboratory reproducibility are expressed as relative standard deviations (%RSD r and %RSD IR ) 10 11 12 194 in Table 2. For a few steroids the precision was slightly less at the lowest spike level than for the 13 14 195 other three concentration levels. Precision and trueness for the endogenous steroids were calculated 15 16 196 by subtracting theFor quantitative Peer result of unspiked Review samples from theOnly result of spiked sample probably 17 18 19 197 leading to a slight overestimation of the precision. 20 21 198 Since no certified reference material was available, en estimate of trueness was determined by 22 23 199 calculating the recovery (corrected with internal standard) of fortified blank samples. Trueness was 24 25 26 200 determined by analysing 78 spiked samples and ranged from 97 – 107%. In our laboratory, matrix 27 28 201 effect, is now a mandatory parameter to calculate when developing and validating a method, based 29 30 31 202 on ESI detection. In order to calculate matrix effect, samples of muscle tissue were extracted and 32 33 203 the final extracts were spiked at 1 µg/kg and analysed. A standard solution in solvent with the 5 34 35 204 µg/L was likewise prepared and analysed. Matrix effect, or ion suppression, is calculated as the 36 37 38 205 relative difference between the area of the standard solution and the extracted sample. Matrix effect 39 40 206 was less than 20% except for the acetylgestagens where matrix effect was 20 – 34%, and it was 41 42 207 decided not to try to reduce matrix effect further. 43 44 45 208 Decision limit (CC α) means the limit at and above which it can be concluded with an error 46 47 209 probability of α that a sample is non-compliant [3]. CC α values were estimated from the noise level 48 49 50 210 for the primary and secondary transition as well as from the standard deviation on uspiked samples 51 52 211 or samples spiked at low level. Final results are presented in Table 3. Detection capability (CC β) 53 54 55 212 means the smallest amount of analyte that may be detected, identified and quantified in a sample 56 57 213 with an error probability of β. For banned substances with no permitted limit, the detection 58 59 60 214 capability is the lowest concentration at which the method is able to detect truly contaminated

9 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 10 of 23

1 2 3 4 5 215 samples with a statistical certainty of 1- β . CC β was calculated as the value of the decision limit 6 7 216 plus 1.64 times the intra-laboratory standard deviation (S R) of the measured concentrations in 8 9 10 217 samples spiked at a level close to the CC β [3] . For the analytes, where the calculated CC β is below 11 12 218 the spiking level, CC β is given as the spike level in Table 3. For practical reasons the decision limit 13 14 219 for the endogenous steroids is estimated by analysis of samples where a quantitative result, and a 15 16 For Peer Review Only 17 220 S/N ratio, is given at low level. The S/N ratio and the quantitative result, for these steroids, is used 18 19 221 for estimating the CC α. 20 21 22 222 According to Commission Decision 2002/657/EC, an analyte is identified by assessing the 23 24 223 retention time and the abundance of specific ions. When using LC-MS/MS in the MRM mode for 25 26 224 confirmation, two transition ions have to be detected, both with a signal-to-noise ratio > 3. In 27 28 29 225 addition, the maximum permitted tolerances for relative ion intensities also have to be fulfilled. 30 31 226 Results from the determination of precision are used for calculating ion ratio. For 13 analytes (out 32 33 -1 34 227 of 17) the criteria on ion ratio was fulfilled in the concentration range 0.5 – 2.0 g kg when only 35 36 228 one injection is made. The results are presented in table 4. 37 38 229 In addition it is possible to confirm the mass of the substance with accurate mass spectrometry. 39 40 41 230 As can be seen in figure 2 the compound eluting at retention time 8.35 min. is identified by the 42 43 231 retention time and the two transitions when using LC-MSMS and by using LC-TOFMS 44 45 46 232 identification is supported by correct mass detection. In figure 3 the mass of α-nortestosterone (RT 47 48 233 = 4.39 min.) is measured with less than 5 ppm precision in a sample with 10 ppb and the signal to 49 50 234 noise ratio is 50 for the extracted TOFMS chromatogram. 51 52 53 235 Factorial effect experiment 54 55 236 During method development a factorial experiment was done to evaluate the performance of the 56 57 58 237 method, when introducing changes in possible critical factors. When running the method on a 59 60 238 routine basis, it is likewise practical to know whether treatment of the meat sample before analysis

10 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 11 of 23 Food Additives and Contaminants

1 2 3 4 5 239 can influence the analytical result (ex. freeze-thaw cycles). Five factors were evaluated: a) amount 6 7 240 of sample (0.5g or 1.0 g); b) species (bovine or poultry); c) temperature during 8 9 241 cleanup/centrifugation (4 oC or 20 oC); d) freeze-thaw cycles of sample before analysis (once or 10 11 12 242 twice); e) duration of hydrolysis (two hours or three hours), and a factorial design was used for 13 14 243 estimating the factors. For practical purposes a ¼ × 2 5 fractional design was used and to evaluate the 15 16 244 influence of the factorsFor each Peerfactor setup wasReview done in replicate. TheOnly number of samples was hereby 17 18 19 245 reduced from 64 samples to 16 samples and the statistical evaluation was done in accordance with 20 21 246 Miller and Miller [19] using an F-test for testing the factors. In figure 4 each diagram represent one 22 23 247 factor (the factor “time of incubation for hydrolysis” is not shown since no effect was observed 24 25 26 248 when changing this factor). The relative difference between the two levels is shown and on the left 27 28 249 side in the diagram, the response is used for calculation and on the right side the area is used for 29 30 31 250 calculating the difference. For clostebolacetate it can be seen that the relative difference between 32 33 251 the two levels is exceeding 10% for two factors, “species” and “sample amount” (using response 34 35 252 and area), but since the precision for this compound is large and not accepted for confirmation this 36 37 38 253 steroid will not be used for testing the factors. It was demonstrated that none of the factors had an 39 40 254 effect on the analytical performance when testing the responses (taking internal standard into 41 42 255 account) for most steroids. When testing the areas, the factors “species” and “amount of sample” 43 44 45 256 had an effect on the analytical performance but the other factors had no significant effect. The 46 47 257 factor “species” included bovine and poultry and it was shown that the absolute recovery was 48 49 258 higher when testing bovine compared to poultry. Likewise the absolute recovery was higher when 50 51 52 259 using ½ g of meat sample compared to 1 g. 53 54 260 Real sample analysis 55 56 57 261 The developed method has been used for the past two years, primarily, for import control. 58 59 60

11 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 12 of 23

1 2 3 4 5 262 Nearly 100 samples of beef have been analysed and in many samples endogenous steroids have 6 7 263 been found. In 22%, 72% and 54% of the samples testosterone (max. 0.6 g kg -1), epitestosterone 8 9 -1 -1 10 264 (max. 1.6 g kg ) and androstenedione (max. 2.8 g kg ) were found. The findings of testosterone, 11 12 265 epitestosterone and androstenedione are in line with already published data [20-22]. In eight 13 14 266 samples, both testosterone and the epimer was found with a testosterone/epitestosterone (T/E) ratio 15 16 For Peer Review Only 17 267 of 0.1 – 3.2. In human sport, misuse of testosterone has been detected by measurement of the T/E 18 19 268 ratio in urine [23]. A T/E-ratio in urine exceeding 6 indicated misuse of testosterone. Since the 20 21 269 interconversion of testosterone and androstenedione to epitestosteron is species specific the same 22 23 24 270 T/E ratio cannot be used, according to Angeletti and coworkers [21] and their results found that a 25 26 271 difference of T/E-ratio in urine from treated and untreated calves could not immediately be done. 27 28 29 30 272 Conclusion 31 32 273 A method has been presented for confirmation of steroids in muscle tissue. The method is fully 33 34 35 274 validated and the intra-laboratory reproducibility is acceptable for all substances and the CC α and 36 37 275 CC β is below 1 g kg -1 which is the recommended concentration set by the European Commission. 38 39 40 276 The MRM transition ratios are fulfilling the criterias for confirmation and the consumption of 41 42 277 solvents is low due to the small sample size (1 g) and use of LC-TOF is possible for confirming the 43 44 278 mass of the substance. Inclusion of new substances, because of new information, is presented, and a 45 46 47 279 fractional factorial design is shown to be useful during method development and method validation. 48 49 280 The described procedure was successfully used in the residue control and the findings of 50 51 281 endogenous steroids are in line with already published data. 52 53 54 55 282 Acknowledgements 56 57 283 We thank Lis Abildgaard Andersen, Maud Bering Andersen and Lene Gram Hansen for their 58 59 60 284 skilful technical assistance with the chemical analyses.

12 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 13 of 23 Food Additives and Contaminants

1 2 3 4 5 285 6 7 8 286 References 9 10 287 [1] Council Directive 96/22/EC. Off. J. Eur. Commun. L 125 (1996) 3. 11 12 288 [2] Community Reference Laboratories for Residues (2007): CRL guidance paper, December 13 14 289 2007: CRLs view on the state of the art analytical methods for national residue plans. 15 290 [3] Commission Decision 2002/657/EC. Off. J. Eur. Commun. L 221 (2001) 8. 16 For Peer Review Only 17 291 [4] Database on accreditated methods in the National Food Institute 18 19 292 (http://published.danak.dk/register.asp?nohead=y&lang=d&akk=350) 20 21 293 [5] Sharmann M., Thomas M. Scientific/Technical Report submitted to EFSA. 22 294 CFP/EFSA/CONTAM/2008/02. Published on 10 December 2009. 23 24 295 [6] Blasco C. et al. Analysis of meat samples for anabolic steroids residues by liquid 25 26 296 chromatography/tandem mass spectrometry. J. Chrom. A 1154 (2007) 230-239. 27 28 297 [7] Hartmann S., Steinhart H. Simultaneous determination of anabolic and catabolic steroid 29 30 298 hormones in meat by gas-chromatography-mass spectrometry. J. Chrom. B 707 (1997) 105- 117. 31 299 [8] Maume D. et al. Modification of 17 - metabolite profile in steer edible tissues after 32 33 300 estradiol implant administration.. Anal. Chim. Acta 483 (2003) 289-297. 34 35 301 [9] Marchand P. et al. Ultra trace detection of a wide range of anabolic steroids in meat by gas 36 37 302 chromatography coupled to mass spectrometry. J. Chrom. A 867 (2000) 219-233. 38 303 [10] Schmidt KS. et al. In-house validation and factorial effect analysis of a liquid 39 40 304 chromatography-tandem mass spectrometry method for the determination of steroids in bovine 41 42 305 muscle. Anal. Chim. Acta 637 (2009) 156-164. 43 44 306 [11] Malone EM. et al. Development of a rapid method for the analysis of synthetic growth 45 307 promoters in bovine muscle using liquid chromatography tandem mass spectrometry. Anal. Chim. 46 47 308 Acta 637 (2009) 112-120. 48 49 309 [12] Shao B. et al. Simultaneous determination of residual hormonal chemicals in meat, kidney, 50 51 310 tissues and milk by liquid chromatography-tandem mass spectrometry. Anal. Chim. Acta 548 52 53 311 (2005) 41-50. 54 312 [13] Kaklamanos G. et al. Determination of anabolic steroids in muscle tissue by liquid 55 56 313 chromatography-tandem mass spectrometry. J. Chrom. A 1216 (2009) 8072-8079. 57 58 314 [14] Van Poucke C. Determination of anabolic steroids in dietary supplements by liquid 59 60 315 chromatography-tandem mass spectrometry. Anal. Chim. Acta 586 (2007) 35-42.

13 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 14 of 23

1 2 3 4 5 316 [15] Mortensen S., Pedersen M. Confirmatory analysis of acetylgestagens in plasma using liquid 6 317 chromatography- tandem mass spectrometry. Anal. Chim. Acta 586 (2007) 217-222. 7 8 318 [16] Andersen JH. et al. Optimization of soild phase extraction clean up and validation of 9 10 319 quantitative determination of corticosteroids in urine by liquid chromatography-tandem mass 11 12 320 spectrometry. Anal. Chim. Acta 617 (2008) 216-224. 13 321 [17] H. Mei, Matrix effects: causes and solutions, in: W.A. Korfmacher (Ed.), Using Mass 14 15 322 Spectrometry for Drug Studies, CRC Press, 2004, ISBN 0849319633. 16 For Peer Review Only 17 323 [18] J.-P. Antignac, K. De Wasch, F. Monteau, H. De Brabander, F. Andre, B. Le Bizec. Anal. 18 19 324 Chim. Acta 529 (2005) 129-136. 20 21 325 [19] J. N. Miller, J. C. Miller. Statistics and chemometrics for analytical chemistry. 4 ed. (2000), 22 326 Prentice Hall. 23 24 327 [20] Gaiani R., Chiesa F. Physiological levels of androstenedione and testosterone in some edible 25 26 328 tissues from calves, bulls and heifers . Meat Sci. 17 (1986) 177-185. 27 28 329 [21] Angeletti R. et al. The urinary ratio of testosterone to epitestosterone: a good marker of illegal 29 330 treatment also in cattle?. Vet. Res. Comm. 30(Suppl. 1) (2006) 127-131. 30 31 331 [22] Hartmann S. et al. Natural occurence of steroid hormones in food. Food Chem. 62 (1) (1998) 32 33 332 7-20. 34 35 333 [23] Linnet K. Effect of the biological matrix on the urinary testosterone/epitestosterone ratio 36 334 measured by gas chromatography/mass spectrometry in doping analysis. Biol. Mass Spectr. 22 37 38 335 (1993) 412-418. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

14 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 15 of 23 Food Additives and Contaminants

1 2 3 4 5 336 Figure captions: 6 337 7 338 Figure 1. Bovine muscle fortified with all analytes at a concentration of 1 g kg -1. 8 339 9 340 Figure 2. LC-MS/MS of bovine muscle fortified with all analytes at a concentration of 10 g kg -1. 10 11 341 (a) Total ion chromatogram. (b) Extracted ion chromatogram corresponding to α-nortestosteron 12 342 (RT = 8.35). 13 343 14 344 Figure 3. LC-TOFMS of bovine muscle fortified with α- and β-nortestosteron at a concentration of 15 -1 16 345 10 g kg . (a) ExtractedFor ion Peerchromatogram. Review (b) Scan of α-nortestosteron Only (down) and isotopic 17 346 spectrum corresponding to α-nortestosteron (upper) 18 347 19 348 Figure 4. Graphical presentation of four factors “species” (a), “sample amount” (b), “temperature 20 349 during sample cleanup” (c) and “treatment of sample before analysis” (d). 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

15 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 16 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 For Peer Review Only 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 147x125mm (96 x 96 DPI) 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 17 of 23 Food Additives and Contaminants

1 2 3 4 S5 S5 090429b15 2: MRM of 10 Channels ES+ 090429b15 2: MRM of 10 Channels ES+ 5 8.35 7.37 TIC 100 275.2 > 239.2 100 1.61e5 6 7.21e5 8.41 7 8 9 10

11 7.77 % 12 13 14 For8.05 Peer Review Only 15 16 0 8.35 7.50 8.00 8.50 9.00 17 % 090429b15 2: MRM of 10 Channels ES+ 8.39 18 100 8.35 275.2 > 109.1 1.90e5 19 20 21 22 23 24 % 25 26 27 28 0 Time 0 Time 29 7.50 8.00 8.50 9.00 7.50 8.00 8.50 9.00

30 31 32 (a) (b) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 18 of 23

1

2 100426a 15 1: TOF MS ES+ 100426a 15 (0.008) Is (1.00,1.00) C18H26O2 1: TOF MS ES+ 3.41 275.2011 8.12e12 100 275.201 0.01Da 100 3 860 4 3.40 4.39 5 4.38 6 % 7 276.2045 277.2075 8 0

9 % 100426a 15 749 (4.383) Cm (745:759-(680:720+782:820)) 1: TOF MS ES+ 10 100 275.2003 8.11e3 11 12 % 13 276.2046 297.1837 14 5.12 329.2085 For Peer Review255.1562 Only 313.1574 338.2109 244.2329 260.1670 282.1467 338.9519 15 0 Time 0 m/z 1.00 2.00 3.00 4.00 5.00 6.00 7.00 240 250 260 270 280 290 300 310 320 330 340 16 17 (a) (b) 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 19 of 23 Food Additives and Contaminants

1 2 3 4 5 6 7 8 9 10 11 12 13 14 For Peer Review Only 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

36 275x234mm (96 x 96 DPI) 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 20 of 23

1 2 3 4 Table 1. Data acquisition method for analysis by LC-MS/MS. 5 6 m/z a 7 Name (abbreviation) Rt (min.) MRM transitions ( ) Collision energy (eV) 8 9 Medroxyprogesterone acetate (MPA) 19.1 387.3>327.2 14 10 (387.3>285.1) 18 11 12 (MGA) 18.4 385.3>325.2 12 13 (385.3>267.1) 18 14 (CMA) 18.8 405.2>309.1 14 15 16 For Peer Review (405.2>345.1) Only 12 17 (MLA) 19.0 397.2>337.1 14 18 19 (397.2>279.1) 18 20 Testosterone/epi-Testosterone (TES/ETE) 8.4/10.7 289.1>97.1 18 21 22 (289.1>109.1) 20 23 α-Boldenone/ β-Boldenone (ABO/BBO) 7.8/6.4 287.2>121.1 18 24 25 (287.2>135.1) 15 26 Methylboldenone (MBO) 7.7 301.2>149.1 15 27 28 (301.2>121.1) 20 29 Methyltestosterone (MT) 10.4 303.2>97.0 25 30 (303.2>109.1) 30 31 32 α-Trenbolon/ β-Trenbolon (ATR/BTR) 6.3/5.9 271.2>199.1 20 33 (271.2>253.2) 20 34 35 α-Nortestosterone/ β-Nortestosterone (ANO/BNO) 8.7/7.0 275.2>109.1 22 36 (275.2>239.2) 17 37 38 Androstenedione (ASD) 10.3 287.0>97.0 20 39 (287.0>109.0) 20 40 41 Altrenogest (ALTR) 13.0 311.1>227.0 23 42 (311.1>269.0) 15 43 44 Clostebolacetat (CLAc) 22.9 365.1>142.9 22 45 (365.1>305.1) 15 46 47 Medroxyprogesterone acetate-d3 (MPA-d3) 19.1 390.3>330.2 14 48 Megestrol acetate-d3 (MGA-d3) 18.4 388.4>270.2 18 49 Melengestrol acetate-d (MLA-d ) 19.0 400.1>340.2 14 50 3 3 51 Methyltestosterone- d 2 (MT-d2) 10.4 306.2>97.0 25 52 Testosterone- d (TES-d ) 8.4 291.1>99.1 18 53 2 2 54 β-Boldenone-d3 (BBO-d3) 6.4 290.2>121.1 15 55 β-Nortestosterone-d (BNO-d ) 7.0 278.2>109.1 22 56 3 3 57 Methylboldenone- d 3 (MBO-d3) 7.7 304.2>152.0 12 58 β-Trenbolon-d2 (BTR-d2) 5.9 273.2>199.1 20 59 60

a Primary MRM transitions used for quantitative purposes are listed first, and secondary MRM transitions for

confirmation are in parentheses.

http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 21 of 23 Food Additives and Contaminants

1 2 3 4 Table 2. Repeatability, intra-laboratory reproducibility and corrected recovery in spiked samples of muscle 5 6 7 tissue. 8 9 Compound Nominal Repeatability Intra-laboratory N Corrected 10 concentration (%RSD ) Reproducibility recovery (%) 11 r -1 12 (g kg ) (%RSD IR ) 13 14 Medroxyprogesterone acetate 0.13-1.0 5 7 78 99 15 Megestrol acetate 0.13 13 16 16 For Peer Review6 Only 101 17 0.25-1.0 7 62 18 Chlormadinone acetate 0.13 12 16 19 5 100 20 0.25-1.0 7 62 21 Melengestrol acetate 0.13 43 16 22 5 105 23 0.25-1.0 7 62 24 Testosterone 0.13 19 16 25 10 107 26 0.25-1.0 8 62 27 epi-Testosterone 0.25-2.0 4 16 78 103 28 0.13 12 16 29 α-Boldenone 5 98 30 0.25-1.0 5 62 31 β-Boldenone 0.13-1.0 4 7 78 98 32 33 Methylboldenone 0.13-1.0 4 5 78 101 34 Methyltestosterone 0.13-1.0 8 9 78 99 35 36 α-Trenbolon 0.25-2.0 15 22 78 98 37 β-Trenbolon 0.5-4.0 8 18 78 98 38 39 α-Nortestosterone 0.25-2.0 5 8 78 101 40 β-Nortestosterone 0.25-2.0 7 8 78 107 41 42 Androstenedione 0.25-2.0 6 18 78 107 43 Altrenogest 0.5-1.0 9 20 28 98 44 Clostebolacetat 0.5-1.0 38 49 25 86 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

http://mc.manuscriptcentral.com/tfac Email: [email protected] Food Additives and Contaminants Page 22 of 23

1 2 3 4 5 Table 3. CC α and CC β obtained in muscle tissue. 6 7 Compound CC α CC β 8 (g kg-1) (g kg-1) 9 10 Medroxyprogesterone acetate 0.04 0.13 11 12 Megestrol acetate 0.03 0.13 13 14 Chlormadinone acetate 0.10 0.13 15 16 Melengestrol acetateFor Peer 0.19 Review 0.20 Only 17 18 Testosterone 0.07 0.13 19 20 epi-Testosterone 0.08 0.25 21 22 α-Boldenone 0.02 0.13 23 24 β-Boldenone 0.04 0.13 25 26 Methylboldenone 0.02 0.13 27 28 Methyltestosterone 0.07 0.13 29 30 α-Trenbolon 0.16 0.33 31 32 β-Trenbolon 0.33 0.50 33 34 α-Nortestosterone 0.09 0.25 35 36 β-Nortestosterone 0.10 0.25 37 38 Androstenedione 0.12 0.25 39 40 Altrenogest 0.06 0.50 41 42 Clostebolacetate 0.30 0.50 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

http://mc.manuscriptcentral.com/tfac Email: [email protected] Page 23 of 23 Food Additives and Contaminants

1 2 3 4 -1 5 Table 4. Ion ratio data (0.5-2.0 g kg ). 6 7 Compound Ion ration Ion ratio Samples 8 average CV% fulfilling 9 10 criteria 11 Medroxyprogesterone acetate 2.86 7 100 12 13 Megestrol acetate 0.71 7 100 14 15 Chlormadinone acetate 1.16 10 88 16 For Peer Review Only 17 Melengestrol acetate 1.45 11 94 18 19 Testosterone 1.25 7 98 20 21 epi-Testosterone 1.18 7 99 22 23 α-Boldenone 1.34 5 100 24 25 β-Boldenone 1.16 8 100 26 27 Methylboldenone 1.53 6 100 28 29 Methyltestosterone 1.48 7 98 30 31 α-Trenbolon 0.25 13 95 32 33 β-Trenbolon 0.85 13 86 34 35 1.28 12 96 36 α-Nortestosterone 37 38 β-Nortestosterone 1.50 13 83 39 40 Androstenedione 1.40 6 95 41 42 Altrenogest 0.82 5 100 43 44 Clostebolacetate 0.95 15 70 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

http://mc.manuscriptcentral.com/tfac Email: [email protected]