Development and validation of a multi-residue ...

Microchemical Journal 115 (2014) 70–77

Contents lists available at ScienceDirect

Microchemical Journal journal homepage: www.elsevier.com/locate/microc

Development and validation of a multi-residue method for determination of 18 β-agonists in bovine urine by UPLC–MS/MS D. Mauro a, S. Ciardullo a, C. Civitareale a, M. Fiori a, A.A. Pastorelli a, P. Stacchini a,⁎, G. Palleschi b a b

National Reference Laboratory for Residues of Veterinary Drugs, Department of Food Safety and Veterinary Public Health, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy Università di Roma Tor Vergata, Dipartimento di Scienze e Tecnologie Chimiche, Via della Ricerca Scientifica 00133, Rome, Italy

a r t i c l e

i n f o

Article history: Received 20 January 2014 Received in revised form 27 February 2014 Accepted 27 February 2014 Available online 11 March 2014 Keywords: β-agonists Bovine urine UPLC-MS/MS Solid phase extraction Validation

a b s t r a c t Ultra performance liquid chromatography (UPLC) hyphenated to tandem mass spectrometry (MS/MS) was used for the development of an analytical method capable of simultaneous identification and quantification of 18 β-agonist compounds namely, brombuterol, chlorbrombuterol, cimaterol, cimbuterol, clenbuterol, clencyclohexerol, clenisopenterol, clenpenterol, clenproperol, hydroxymetylclenbuterol, isoxsuprine, mabuterol, mapenterol (clenbuterol-like compounds), ractopamine, ritodrine, salbutamol, salmeterol (salbutamol-like compounds) and zilpaterol in bovine urine. In compliance with the Commission Decision 2002/657/EC (CD 2002/657/EC), the method was validated applying a matrix-comprehensive in-house validation approach based on a fractional factorial design. Six experimental factors varied on two levels were selected for this purpose. The relevant validation parameters such as decision limit CCα (range, 0.24–0.51 μg L−1) and detection capability CCβ (range, 0.27–0.71 μg L−1), within-laboratory reproducibility (b20%) as well as recovery (range, 92–109%) were in agreement with the performance criteria set in the CD 2002/657/EC. Overall, the proposed method enabled both screening and confirmatory detection of the β-agonist compounds in the framework of official monitoring plans. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Veterinary drugs are therapeutically used to treat and prevent diseases in farm animals and in some cases, as growth-promoting agents to stimulate animal growth enhancement. As a consequence of veterinary drug administration to food-producing animals, the presence of their residues in animal food products is a critical public health concern. In order to ascertain the high quality of food products of animal origin and to protect consumers by reducing their exposure to veterinary drug residues, several Directives and Regulations are stipulated by the European Union (EU). The EU introduced the procedure to establish maximum residue limits (MRLs) for veterinary drugs in the Council Regulation 1990/2377/EEC [1]. In the Annex of the subsequent Council Regulation 2010/37/EU, which repealed the Regulation 1990/2377/EC, veterinary drugs are distinguished in allowed and prohibited veterinary substances [2]. As reported by the EU, the residues of these prohibited veterinary drugs in food animal products constitute a risk for consumer health at any level of concentration and consequently, these substances do not require MRL values. The prohibition on the use of veterinary drugs having a hormonal or thyrostatic action as well as β-agonist compounds improperly exploited as growth-promoting agents has been stipulated in the Council ⁎ Corresponding author. Tel.: +39 06 49902533; fax: +39 06 49387101. E-mail address: [email protected] (P. Stacchini).

http://dx.doi.org/10.1016/j.microc.2014.02.012 0026-265X/© 2014 Elsevier B.V. All rights reserved.

Directive 1996/22/EC, subsequently amended by Directive 2008/97/EC of the European Parliament and the Council [3]. As previously mentioned, these veterinary substances may be dangerous for consumers and in addition, they negatively affect the foodstuff quality, since the legal tolerance by the EU for their residues in animal food products is “zero”. Community legislation on veterinary drugs also involves guidelines for the official monitoring of their residues in animals and animal products in each State Member of the EU. The Council Directive 1996/23/EC classifies the veterinary drugs included in the official monitoring plans into two main groups namely, substances having anabolic effects and prohibited substances (Group A) as well as veterinary drugs and contaminants (Group B) [4]. Among prohibited substances included in the European Union monitoring plans, β-agonist compounds are one of the focuses of the scientific community. Their association with human poisoning cases from consumption of contaminated meat and liver animal products illicitly treated with clenbuterol are highlighted by other authors [5–9]. Clenbuterol and compounds belonging to the family of β-agonists are therapeutically administered to farm animals as bronchodilatants and tocolitic agents to accomplish relaxation of smooth muscle cells. Nevertheless, β-agonists are well-known as repartitioning agents illicitly employed in veterinary framework as growth promoters. On the one hand, these compounds are capable of reducing fat deposition stimulating lipolysis and decreasing at the same time, fatty acid synthesis. On the other hand, β-agonist

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77

administration to animals has been reported to increase muscle accretion modifying the rates of protein degradation and synthesis [10–12]. A high number of analytical methods are developed in order to quantify and confirm β-agonists in animals and animal products. Screening determinations of β-agonist residues are achieved using high sample throughput approaches such as enzyme and radio immunoassay, immunoaffinity chromatography, surface plasma resonance (SPR) and high performance thin layer chromatography (HPTLC) [13–19]. In addition, more sensitive and specific mass spectrometry techniques hyphenated to gas and liquid chromatography are performed for screening investigations especially by official laboratories involved in both screening as well as confirmatory analysis. In the framework of confirmatory purpose, that is, in the case of a positive response after a screening procedure, mass spectrometry is the only analytical technique taken into account by the European legislation [20]. Confirmatory methods for detection of β-agonists in animals and their products are carried out using gas chromatography–mass spectrometry (GC–MS) or tandem mass spectrometry (GC–MS/MS) and liquid chromatography–mass spectrometry (LC–MS) or tandem mass spectrometry (LC–MS/MS) [21–26]. MS/MS instrumentation coupled with a liquid chromatographic separation that does not require a derivatization procedure (i.e. gas chromatography) recently made LC–MS/MS the main instrumental method performed by official laboratories for confirmatory investigation of non volatile and polar compounds such as β-agonists [27–33]. The aim of this study was to develop an analytical method capable of simultaneous identification and quantification of 18 β-agonists namely, brombuterol, chlorbrombuterol, cimaterol, cimbuterol, clenbuterol, clencyclohexerol, clenisopenterol, clenpenterol, clenproperol, hydroxymetylclenbuterol, isoxsuprine, mabuterol, mapenterol (clenbuterol-like compounds), ractopamine, ritodrine, salbutamol, salmeterol (salbutamol-like compounds) and zilpaterol in bovine urine by means of ultra performance liquid chromatography (UPLC) hyphenated to tandem mass spectrometry. The proposed procedure based on an advanced chromatographic technology with high peak resolution and rapid separation properties could be successfully applied for both screening and confirmatory detection of β-agonists in the framework of official monitoring plans [34–36]. In order to estimate the performance level of this method, an in-house validation approach was undertaken in compliance with the “validation according to alternative models” section of the CD 2002/657/EC [20]. A factorial design was performed to evaluate the possible influence of selected experimental factors (e.g. operator, SPE cartridge size, NH 3 percentage in elution, sample pH value during application on cartridge, final volume in sample extraction, sample storage duration before measurement) on analytical results of each β-agonist compound. 2. Experimental 2.1. Reagents and materials Hydrochloride salts of bromclorbuterol, brombuterol, clencyclohexerol, clenisopenterol, clenpenterol, hydroximethylclenbuterol, mabuterol and mapenterol as well as cimaterol, cimbuterol, and clenproperol were purchased from Witega (Berlin, Germany). Clenbuterol, isoxsuprine, ractopamine, ritodrine and salbutamol as hydrochloride salts were purchased from Sigma Aldrich (Germany) and zilpaterol was obtained from TRC (Toronto, Canada). Internal standards namely brombuterol-d9, clenbuterol-d9, clenpenterol-d5, mabuterol-d9 and mapenterol-d11 as hydrochloride salts and cimaterol-d7, cimbuterold9 and clenproperol-d7 were purchased from Witega (Berlin, Germany). Finally, hydrochloride salts of isoxsuprine-d5, ractopamined5 were obtained from Rikilt (Wageningen, The Nederlands) as well as hydrochloride salt of salbutamol-d9 was purchased from Sigma Aldrich (Germany). The stock standard solutions (from 1 g L− 1 to

71

10 mg L− 1) were prepared in methanol and stored at − 20 °C. The working mixed standard solutions (at 100 μg L−1 and 10 μg L−1) of β-agonist compound were prepared by dilution of the stock standard solutions and subsequently kept at 4 °C. As regards internal standards, mixed standard solutions including all deuterated compounds were prepared as well. Deionized water employed in all solutions was obtained by a Milli-Q system (Millipore, USA). Methanol, acetonitrile, 2-propanol of high performance liquid chromatography (HPLC) grade, and β-glucoronidase/arylsulfatase (from Helix pomatia, Merck, Germany) were obtained from Merck (Germany). Glacial acetic acid, formic acid, sodium acetate anhydrous, sodium hydroxide and dichloromethane were purchased from Carlo Erba (Italy). As far as SPE cartridges are concerned, Bond Elute Certify 300 mg (Agilent, USA) were applied for the cleanup procedure. 2.2. Analytical procedure Samples of bovine urine were collected from a local slaughterhouse, frozen at − 20 °C and kept at this temperature until analysis. Pooled urine obtained from twenty calves previously analyzed for each βagonist under investigation was used as blank urine to quantify the validation samples. As regards the sample preparation procedure, 2.5 ml of urine was placed in 15 ml centrifuge tubes and subsequently spiked with 5 μg L− 1 of each internal standard and β-agonists in relation to the level of fortification required. After an addition of 30 μL of βglucoronidase/arysulfatase as well as 1 ml of acetate buffer (pH 4.8; 2 M), urine samples were placed in a thermoregulated bath (GFL, Germany) at 37 °C for 12 h. Afterwards, the urine samples were adjusted to a pH between 7.5 and 8 using a solution of NaOH (10 N and 1 N) for the subsequent SPE purification process. Samples were applied on SPE cartridges previously activated with 2 ml of methanol and 2 ml of water and placed on a vacuum manifold device (Supelco, Germany). Subsequently, SPE cartridges were washed with 1 ml of water and a following addition of 2 ml of acetate buffer (pH 4.0; 0.1 M) was employed. SPE cartridges were washed with 2 ml of methanol and finally β-agonists were eluted with 4 mL of dichloromethane and 2propanol in a ratio of 80:20, v/v with 3% of NH3. After the elution process, the obtained solutions were evaporated at 40 °C to dryness under a gentle stream of nitrogen by means of a Thermo Scientific (USA) evaporator. Finally, the samples were dissolved in 400 μL of a water/methanol mixture (60:40, v/v) for UPLC–MS/MS analyses. 2.3. LC–MS/MS system A Waters UPLC system was used to perform a reverse phase chromatography separation of β-agonists. An Acquity ethylene bridged hybrid (BEH) C18 column (100 mm × 2.1 mm and 1.7 μm particle size) was used as stationary phase and a mobile phase consisting in deionized water with 0.1% of formic acid and acetonitrile with 0.1% of formic acid were employed. An injection volume of 5 μL, a flow rate of 0.35 ml/min and a linear gradient from 2% to 95% of acetonitrile in 5.5 min were applied for chromatographic separation. The LC system was coupled with a Waters Xevo TQ MS mass spectrometer equipped with an ESI source operating in positive ionization mode (ESI+). The MassLynx 4.11 software (Waters, USA) was used in order to control the UPLC–MS/MS system. Capillary voltage, source temperature, desolvatation gas flow rate and its temperature were set at 1 kV, 150 °C, 900 L/h and 600 °C, respectively. Cone voltage (CV) in the range of 16–28 eV as well as collision energy (CE) in the range of 12– 40 eV were optimized for each protonated molecular ions [M + H]+ and different product ions, respectively. In Table 1 are summarized [M + H]+, product ions, and β-agonist-specific MS/MS parameters. For the quantification purpose, six concentration levels were prepared by spiking blank urine samples at 0, 0.2, 0.5, 1, 2, 5 μg L−1 with a mixture of 18 β-agonists and 11 internal standards (5 μg L− 1).

72

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77

Table 1 [M + H]+, product ions, and β-agonist-specific MS/MS parameters. Number

Compound

RT (min)

CV (V)

Transition MRM

1

Zilpaterol

2.46

20

2 3

Salbutamol-d9 Salbutamol

2.47 2.48

18 20

4 5

Cimaterol-d7 Cimaterol

2.49 2.51

16 16

6 7

Cimbuterol-d9 Cimbuterol

2.72 2.74

18 18

8

Ritodrine

2.82

22

9

Clencyclohexerol

2.96

20

10

Hydroxymetylclenbuterol

3.09

20

11

Ractopamine

3.10

20

12 13 14

Ractopamine-d5 Clenproperol-d7 Clenproperol

3.10 3.16 3.17

22 16 16

15 16

Clenbuterol-d9 Clenbuterol

3.36 3.37

20 20

17

Chlorbrombuterol

3.45

20

18 19

Brombuterol-d9 Brombuterol

3.51 3.52

20 20

20

Isoxsuprine

3.57

22

21 22

Clenpenterol-d5 Clenpenterol

3.60 3.60

22 20

23 24 25

Isoxsuprine-d5 Mabuterol-d9 Mabuterol

3.60 3.63 3.64

18 22 22

26

Clenisopenterol

3.79

18

27 28

Mapenterol-d11 Mapenterol

3.84 3.85

22 22

29

Salmeterol

4.49

28

262.2 262.2 249.2 240.2 240.2 227.2 220.2 220.2 243.2 234.2 234.2 288.2 288.2 319.1 319.1 293.2 293.2 302.2 302.2 307.2 270.1 263.1 263.1 286.2 277.1 277.1 323.1 323.1 376.1 367.0 367.0 302.2 302.2 296.2 291.1 291.1 307.3 320.2 311.2 311.2 291.2 291.2 336.2 325.2 325.2 416.4 416.4

→ → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → →

244.1a 185.0b 148.7 222.1b 148.0a 209.1 202.1a 143.0b 161.0 160.1a 143.0b 270.1b 121.0a 203.0a 131.9b 275.1b 203.0a 164.2a 107.0b 167.1 252.1 245.0a 131.9b 204.0 259.0b 202.9a 249.0a 168.0b 293.9 349.0b 292.9a 284.1a 107.0b 278.1 203.0a 131.9b 289.2 238.0 237.0a 217.0b 273.1a 188.0b 238.0 237.0a 217.0b 398.2a 91.0b

CE (eV)

IS

12 24 20 10 18 10 12 28 16 14 24 14 22 20 30 12 20 16 32 16 12 12 26 16 10 16 18 28 20 12 18 14 28 16 16 28 14 18 18 26 14 26 18 18 30 14 40

4 n/a 2 n/a 4 n/a 6 12 12 12 12 n/a n/a 13 n/a 15 18 n/a 18 23 n/a 21 n/a n/a 24 21 n/a 27 23

RT, retention time; CV, cone voltage; CE, collision energy; IS, internal standard. n/a, not applicable. a Quantitative product ions b Identification product ions.

Quantification was performed by means of TargetLynx 4.11 software (Waters, USA). 2.4. Validation procedure According to the CD 2002/657/EC, the method was validated applying an “alternative validation” namely a matrix-comprehensive in-house validation approach based on a fractional factorial design [37,38]. Each β-agonist compound selected for this study was validated close to zero spiking 2.5 mL of a matrix validation sample, consisting of different bovine urine samples, at five levels of fortification, namely 0, 0.2, 0.5, 1, 2 and 5 μg L − 1 . For the quantification purpose, pooled calf urine was employed to obtain a matrix calibration curve. The software “InterVal Plus” (quodata GmbH, Dresden, Germany) was applied for the establishment of factorial study design as well as calculation of validation parameters [39]. Six experimental factors were selected so as to determine their possible influence on analytical results. Each factor such as operator (1 and 2), SPE cartridge size (3 and 6 mL), NH3 percentage in elution (3 and 4%),

sample pH value during application on cartridge (7.5 and 8), final volume in sample extraction (200 and 400 μL), and sample storage duration before measurement (0 and 48 h) was varied on two levels. The experimental design accomplished by means of the “InterVal Plus” software is elucidated in Table 2. Each of the eight runs included in the experimental design consisted of one matrix validation sample as well as its five levels of fortification as previously reported with regard to the employing of validation curves. For each experimental run a matrix validation sample obtained from eight different calves was used. In compliance with the CD 2002/657/EC, the following validation parameters were estimated using the “InterVal Plus” software: decision limit CCα and detection capability CCβ, withinlaboratory reproducibility, recovery (in terms of trueness), validation curves with prediction interval and power curves [20]. The method specificity, that is, the presence of possible matrix interferences was evaluated in the framework of validation process analyzing matrix samples fortified with different internal standards. As far as uncertainty is concerned, this parameter was investigated during the validation of analytical method as well.

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77 Table 2 Experimental design of the in-house validation procedure obtained by varying the six influencing factors at two levels. Run

Operator

SPE size (mL)

NH3 percentage (%)

SPE pH

Final volume (μL)

Storage (h)

1 2 3 4 5 6 7 8

1 1 1 1 2 2 2 2

6 6 3 3 6 6 3 3

3 4 3 4 3 4 3 4

8.0 7.5 7.5 8.0 8.0 7.5 7.5 8.0

400 400 200 200 200 200 400 400

0 48 0 48 48 0 48 0

3. Results and discussion 3.1. Method development β-Agonists were detected in the positive electrospray ionization mode (ESI +) owing to the high proton affinity of these compounds [32,33,36]. The [M + H]+ was obtained using formic acid as the protonation agent. Full scan mode for precursor ion and multiple reaction monitoring (MRM) mode for product ions were used for the characterization of each β-agonist compound and for the subsequent optimization of MS working parameters. The parameters such as cone voltage (CV) and collision energy (CE) were optimized in order to achieve the maximum signal of each [M + H]+ and the greatest response for different product ions, respectively. The characterization of β-agonist compounds and the optimization of previously mentioned MS parameters were obtained using a direct infusion of individual standard solution of both β-agonists and their deuterated compounds (1 mg L−1) using nitrogen and argon as desolvatation and collision gasses, respectively. Finally, the data acquisition was performed in MRM mode after optimization of the collision energies capable of identifying at least three diagnostic product ions. According to the CD 2002/657/EC one precursor and two product ions are required for the confirmation of Group A substances listed in the Council Directive 1996/23/EC [20,4]. Consequently, two product ions characterized by a higher intensity were selected for this purpose. In detail, the most intensive product ion was used for the quantification and the second product ion for the subsequent confirmation of the compound. As a result of previous analytical determinations, the BEH C18 column and a mobile phase consisting in a solution of water and acetonitrile were selected for the determination of β-agonists. The following parameters were optimized so as to improve the chromatographic method: flow rate (0.35 mL min−1), injection volume (5 μL), and gradient elution programme (see Experimental). The possible effect of different concentrations of formic acid in mobile phase was evaluated as well. In order to obtain satisfactory results in chromatography, a formic acid percentage of 0.1 was added on water and acetonitrile. Finally, the optimized ratio of water and methanol employed as reconstituted solvent was 60:40, v/v. Under optimized conditions, the retention times of each β-agonist investigated in this survey were included in a range between 2.49 min (zilpaterol) and 4.49 min (salmeterol) with a run time of 9 min including the column equilibration time. In Fig. 1a and b is reported the MRM chromatograms of each β-agonist obtained for urine blank samples and urine blank samples spiked at 0.5 μg L−1 (identification product ions). In addition, MRM transitions of a bovine urine sample positive for clenbuterol (4 μg L−1) are shown in Fig. 2. Since the biotransformation pathway of β-agonists includes conjugation with glucuronic acid or sulfate, enzymatic hydrolysis using glucuronidase, arylsulfatase or both has been previously exploited to release their conjugates in urine [31,32,41–44]. In this study, a solution of 30 μL containing 15 U/ml and 40 U/ml of glucuronidasi and arylsulfatase, respectively, yielded better results in the enzymatic deconjugation process. In addition to the enzymatic hydrolysis, several other parameters,

73

namely pH of sodium acetate buffer during extraction (pH 4.8), cartridge washing after sample loading (2 mL of water followed by addition of 2 mL of sodium acetate buffer at pH 4 and 2 mL of methanol), and elution solvent (4 mL of dichloromethane and 2-propanol in a ratio of 80:20, v/v with 3% of NH3) were optimized in the extraction and cleanup procedure in order to reduce the signal suppression as well as the presence of possible matrix interferences (see Fig. 1a/b). 3.2. Validation results As previously shown under Experimental, the method was validated using a matrix-comprehensive in-house validation approach. As reported in Table 2, the validation procedure consisted of eight runs with various combinations of experimental factors selected for this purpose. Each of the eight runs included in the experimental design encompassed a matrix validation sample and its five spiked levels of concentration (0, 0.2, 0.5, 1, 2 and 5 μg L−1), resulting in a total number of forty-eight analytical measurements. The selection of matrix validation curves for each β-agonist under investigation was performed according to the CD 2002/657/EC, that is, spike concentration values as low as possible for substances without MRL [20]. A matrix calibration curve consisting in pooled urine obtained from different calves was used to quantify the validation curves so as to achieve an high level of accuracy in the quantification of validation samples. The choice of influencing factors such as operator, SPE cartridge size, NH3 percentage in elution, sample pH value during application on cartridge, final volume in sample extraction, and sample storage duration before measurement was suggested by method requirements as well as laboratory conditions. The results acquired by varying the previously mentioned influencing factors at two levels showed that the determination of each β-agonist was not significantly affected by various combinations of these factors. Consequently, both levels of the experimental factors can be used in the framework of β-agonist quantification in bovine urine. In addition, the use of eight different bovine urine samples as the matrix validation sample for each of the eight runs employed into experimental design allowed one to achieve further information about the robustness of the analytical method. Altogether, the obtained results are indicative of a robust method for each β-agonist compound under study. These data gained by using the “InterVal Plus” and graphically expressed by the software are not shown in this paper. The values of the validation parameters, namely, CCα and CCβ, within laboratory reproducibility (at CCα) and recovery (at CCα), obtained for each β-agonist are summarized in Table 3. As far as critical concentrations are concerned, the results of CCα and CCβ were defined by a narrow range of values, that is, from 0.24 μg L−1 for cimbuterol to 0.51 μg L−1 for salmeterol and from 0.27 μg L−1 for cimbuterol to 0.71 μg L−1 for salmeterol, respectively for CCα and CCβ. The values of withinlaboratory reproducibility ranged from 5.5% for cimbuterol to 18.9% for salmeterol and the recoveries were estimated to be in the range of 92–109% for isoxsuprine and hydroxymetylclenbuterol, respectively. Each result of the validation parameters were characterized by acceptable values and they were in agreement with the performance criteria set in the CD 2002/657/EC. On the basis of these data, it is apparent that the proposed method could be applied to all investigated β-agonists in concentration values at least more than 0.24 μg L−1 for cimbuterol, 0.25 μg L−1 for chlorbrombuterol, clenbuterol, clenisopenterol, isoxsuprine, mapenterol and ractopamine, 0.26 μg L−1 for brombuterol, clenpenterol and zilpaterol, 0.27 μg L−1 for clencyclohexerol, ritodrine and salbutamol, 0.28 μg L−1 for cimaterol, 0.29 μg L−1 for clenproperol, 0.30 μg L−1 for hydroxymetylclenbuterol, 0.31 μg L−1 for mabuterol and 0.51 μg L−1 for salmeterol. As previously elucidated, the results of CCα were comparable for each detected β-agonist (range from 0.24 μg L − 1 to 0.31 μg L − 1 ) except for salmeterol (0.51 μg L−1) which was characterized by a higher value of decision limit than the other β-agonist compounds. This behavior can be explained by the lack of a deuterated internal standard with an

74

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77

Fig. 1. a/b. MRM chromatograms of each β-agonist obtained for urine blank samples and urine blank samples spiked at 0.5 μg L−1 (identification product ions).

identical molecular structure with respect to salmeterol. However, the use of a matrix calibration curve for the quantitative purpose allowed one to achieve a satisfactory value of CCα for salmeterol as well. The validation curves with prediction interval and power curves graphically obtained by means of the “InterVal Plus” software are not reported in this paper.

As far as the method specificity is concerned, this validation parameter was satisfactorily tested by means of eleven deuterated internal standards added to the matrix samples. The measurement of uncertainty was obtained by means of the “InterVal Plus” software which allowed one to determine both the uncertainty of single components such as the uncertainty of the

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77

75

Fig. 1 (continued).

run, the matrix, the repeatability and the calibration as well as combined uncertainty [40]. The uncertainty data were graphically estimated as a function of the matrix fortification levels for each βagonist by the “InterVal Plus” software, but they are not reported in this paper. Even though the estimation of combined uncertainty is

already included in the CCα value, particularly interesting is to evaluate the effects of single uncertainty components on the total uncertainty of each β-agonist as well as the possible differences among the β-agonists behavior. With a few exceptions the percentage of combined uncertainty decreased with increasing of concentration levels,

76

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77

Fig. 2. MRM transitions of a bovine urine sample positive for clenbuterol (4 μg L−1).

although the uncertainty related to the matrix in some cases increased. In details, the combined uncertainty decreased from the zero level (0 μg L− 1) to the second level (1 μg L− 1) of the validation curves and this decrease has been approximately estimated to be 5% on average. From the second level to the end of the validation curves the percentage of combined uncertainty was roughly the same. As regards salmeterol, the percentage reached higher values, that is, approximately a combined uncertainty value of 40% with respect to the first and second levels of the validation curves. This outcome corroborates the previously discussed data about CCα value found for salmeterol compound.

4. Conclusions An analytical approach for the sensitive and selective detection of 18 β-agonists in bovine urine was developed and validated by means of UPLC–MS/MS. The method validation was undertaken in compliance with the CD 2002/657/EC using a matrix-comprehensive in-house validation approach based on a fractional factorial design. On the basis of the satisfactory results gained by the validation procedure, the proposed method has been showed to enable both screening and confirmatory detection of the β-agonist compounds in the framework of official monitoring plans.

Table 3 Critical concentrations, within laboratory reproducibility (at CCα) and recovery (at CCα) gained by the in-house validation procedure for each detected β-agonist. Analyte

Calibration interval (μg L−1)

Total number of Measurements

CCα (μg L−1)

CCβ (μg L−1)

RSD (%) at CCα

Recovery (%) at CCα

Brombuterol Chlorbrombuterol Cimaterol Cimbuterol Clenbuterol Clencyclohexerol Clenisopenterol Clenpenterol Clenproperol Hydroxymetylclenbuterol Isoxsuprine Mabuterol Mapenterol Ractopamine Ritodrine Salbutamol Salmeterol Zilpaterol

0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5

40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40

0.26 0.25 0.28 0.24 0.25 0.27 0.25 0.26 0.29 0.30 0.25 0.31 0.25 0.25 0.27 0.27 0.51 0.26

0.30 0.29 0.33 0.27 0.29 0.33 0.30 0.30 0.35 0.40 0.28 0.39 0.30 0.29 0.35 0.32 0.71 0.30

8.0 7.9 10.7 5.5 7.3 9.1 8.8 9.2 10.1 16.8 6.8 12.8 8.4 8.1 10.7 8.8 18.9 8.5

96 100 102 98 100 94 95 100 106 109 92 107 106 100 95 97 96 100

D. Mauro et al. / Microchemical Journal 115 (2014) 70–77

References [1] Council of the European Union, Council Regulation 1990/2377/EEC of 26 June 1990 laying down a community procedure for the establishment of maximum residue limits of veterinary medicinal products in foodstuffs of animal origin, Off. J. Eur. Communities L224 (1990) 1–136. [2] Council of the European Union, Commission Regulation 2010/37/EC of 22 December 2009 on pharmacologically active substances and their classification regarding maximum residue limits in foodstuffs of animal origin, Off. J. Eur. Communities L15 (2010) 1–72. [3] Council of the European Union, Council Directive 1996/22/EC of 29 April 1996 concerning the prohibition on the use in stockfarming of certain substances having a hormonal or thyrostatic action and of beta-agonists, Off. J. Eur. Communities L125 (1996) 1–14. [4] Council of the European Union, Council Directive 1996/23/EC of 29 April 1996 on measures to monitor certain substances and residues thereof in live animals and animal products, Off. J. Eur. Communities L125 (1996) 10–32. [5] J.F. Martinez-Navarro, Food poisoning related to consumption of illicit beta-agonist in liver, Lancet 336 (1990) 1311. [6] G. Brambilla, A. Loizzo, L. Fontana, M. Strozzi, A. Guarino, V. Soprano, Food poisoning following consumption of clenbuterol-treated veal in Italy, J. Am. Med. Assoc. 278 (1997) 635–640. [7] V. Sporano, L. Grasso, M. Esposito, G. Oliviero, G. Brambilla, A. Loizzo, Clenbuterol residues in non-liver containing meat as a cause of collective food poisoning, Vet. Hum. Toxicol. 40 (1998) 141–143. [8] G. Brambilla, T. Cenci, F. Franconi, R. Galarini, A. Macri, F. Rondoni, M. Strozzi, A. Loizzo, Clinical and pharmacological profile in a clenbuterol epidemic poisoning of contaminated beef meat in Italy, Toxicol. Lett. 114 (2000) 47–53. [9] F. Ramos, A. Cristino, P. Carrola, T. Eloy, J.M. Silva, A.C. Castilho, M.I. Noronha da Silveira, Clenbuterol food poisoning diagnosis by gas chromatography–mass spectrometric serum analysis, Anal. Chim. Acta. 483 (2003) 207–213. [10] C.A. Ricks, R.H. Dalrymple, P.K. Baker, D.L. Ingle, Use of a β-agonist to alter fat and muscle deposition in steers, J. Anim. Sci. 59 (1984) 1247–1255. [11] H.J. Mersmann, Overview of the effects of β-adrenergic receptor agonists on animal growth including mechanisms of action, J. Anim. Sci. 76 (1998) 160–172. [12] M. Salem, H. Levesque, T.W. Moon, C.E. Rexroad, J. Yao, Anabolic effects of feeding β2-adrenergic agonists on rainbow trout muscle proteases and proteins, Comp. Biochem. Phys. A 144 (2006) 145–154. [13] R. Angeletti, M. Paleologo Oriundi, R. Piro, R. Bagnati, Application of an enzymelinked immunosorbent assay kit for β-agonist screening of bovine urines in northeastern Italy, Anal. Chim. Acta. 275 (1993) 215–219. [14] W. Haasnoot, L. Streppel, G. Cazemier, M. Salden, P. Stouten, M. Essers, P. van Wichen, Development of a tube enzyme immunoassay for “on-site” screening of urine samples in the presence of beta-agonists, Analyst 121 (1996) 1111–1114. [15] P. Delahaut, M. Dubois, O. Pri-Bar, I. Buchman, G. Degand, F. Ectors, Development of a specific radioimmunoassay for the detection of clenbuterol residues in treated cattle, Food Addit. Contam. 8 (1991) 43–53. [16] B.A. Rashid, P. Kwasowski, D. Stevenson, Solid phase extraction of clenbuterol from plasma using immunoaffinity followed by HPLC, J. Pharm. Biomed. Anal. 21 (1999) 635–639. [17] W. Haasnoot, M.E. Ploum, R.J. Paulussen, R. Schilt, F.A. Huf, Rapid determination of clenbuterol residues in urine by high-performance liquid chromatography with on-line automated sample processing using immunoaffinity chromatography, J. Chromatogr. 519 (1990) 323–335. [18] I.M. Traynor, S.R.H. Crooks, J. Bowers, C.T. Elliott, Detection of multi-β-agonist residues in liver matrix by use of a surface plasma resonance biosensor, Anal. Chim. Acta. 483 (2003) 187–191. [19] J.M. Degroodt, B. Wyhowski de Bukanski, J. De Groof, H. Beernaert, Cimaterol and clenbuterol residue analysis by HPLC–HPTLC in liver, Z. Lebensm. Unters. Forsch. 192 (1991) 430–432. [20] B.F. Spisso, C.C. Lopes, M.A. Marques, F.R. Neto, Determination of β2-agonists in bovine urine: comparison of two extraction/clean-up procedures for high-resolution gas chromatography–mass spectrometry analysis, J. Anal. Toxicol. 24 (2000) 146–152. [21] Council of the European Union, Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive 96/23/EC concerning the performance of analytical methods and the interpretation of results, Off. J. Eur. Communities L221 (2002) 8–36. [22] T. Bernsmann, P. Furst, M. Godula, Quick screening of priority beta-agonists in urine using automated TurboFlow™ LC/Exactive mass spectrometry, Food Addit. Contam. A 28 (2011) 1352–1363.

77

[23] D.R. Doerge, M.I. Churchwell, C.L. Holder, L. Rowe, S. Bajic, Detection and confirmation of β-agonists in bovine retina using LC/APCI-MS, Anal. Chem. 68 (1996) 1918–1923. [24] H. Hooijerink, R. Schilt, E.O. van Bennekom, F.A. Huf, Determination of beta sympathomimetics in liver and urine by immunoaffinity chromatography and gas chromatography–mass-selective detection, J. Chromatogr. B 660 (1994) 303–313. [25] B. Bocca, M. Fiori, C. Cartoni, G. Brambilla, Simultaneous determination of Zilpaterol and other beta agonists in calf eye by gas chromatography/tandem mass spectrometry, J. AOAC Int. 86 (2003) 8–14. [26] F. Sai, M. Hong, Z. Yunfeng, C. Huijing, W. Yongning, Simultaneous detection of residues of 25 beta-agonists and 23 beta-blockers in animal foods by high-performance liquid chromatography coupled with linear ion trap mass spectrometry, J. Agric. Food Chem. 60 (2012) 1898–1905. [27] W. Radeck, Beta-agonists, anthelmintics, anticoccidials, EURL Workshop 2012: Technical, Analytical and Statistical Issues, Berlin, Germany, 2012, p. 10. [28] C.T. Elliott, C.S. Thompson, C.J.M. Arts, S.R.H. Crooks, M.J. van Baak, E.R. Verheij, G.A. Baxter, Screening and confirmatory determination of ractopamine residues in calves treated with growth promoting doses of the β-agonist, Analyst 123 (1998) 1103–1107. [29] L.D. Williams, M.I. Churchwell, D.R. Doerge, Multiresidue quantification and confirmation of beta-agonists in bovine retina and liver using LC–ES/MS/MS, J. Chromatogr. B 813 (2004) 35–45. [30] N. Van Hoof, D. Courtheyn, J.P. Antignac, M. Van de Wiele, S. Poelmans, H. Noppe, H. De Brabander, Multi-residue liquid chromatography/tandem mass spectrometric analysis of beta-agonists in urine using molecular imprinted polymers, Rapid Commun. Mass Sp. 19 (2005) 2801–2808. [31] J. Blanca, P. Muñoz, M. Morgado, N. Méndez, A. Aranda, T. Reuvers, H. Hooghuis, Determination of clenbuterol, ractopamine and zilpaterol in liver and urine by liquid chromatography tandem mass spectrometry, Anal. Chim. Acta. 529 (2005) 199–205. [32] M.W.F. Nielen, J.J.P. Lasaroms, M.L. Essers, J.E. Oosterink, T. Meijer, M.B. Sanders, T. Zuidema, A.A.M. Stolker, Multiresidue analysis of beta-agonists in bovine and porcine urine, feed and hair using liquid chromatography electrospray ionisation tandem mass spectrometry, Anal. Bioanal. Chem. 391 (2008) 199–210. [33] B. Shao, X. Jia, J. Zhang, J. Meng, Y. Wu, H. Duan, X. Tu, Multi-residual analysis of 16 β-agonists in pig liver, kidney and muscle by ultra performance liquid chromatography tandem mass spectrometry, Food Chem. 114 (2009) 1115–1121. [34] M.I. Churchwell, N.C. Twaddle, L.R. Meeker, D.R. Doerge, Improving LC–MS sensitivity through increases in chromatographic performance: comparisons of UPLC–ES/ MS/MS to HPLC ES/MS/MS, J. Chromatogr. B 825 (2005) 134–143. [35] H. Zheng, L.G. Deng, X. Lu, S.C. Zhao, C.Y. Guo, J.S. Mao, Y.T. Wang, G.S. Yang, H.Y. Aboul-Enein, UPLC–ESI–MS–MS determination of three β2-agonists in pork, Chromatographia 72 (2010) 79–84. [36] A. Vulić, J. Pleadina, N. Perši, D. Milić, W. Radeck, UPLC–MS/MS determination of ractopamine residues in retinal tissue of treated food-producing pigs, J. Chromatogr. B 895–896 (2012) 102–107. [37] B. Jülicher, P. Gowik, S. Uhlig, Assessment of detection methods in trace analysis by means of a statistically based in-house validation concept, Analyst 123 (1998) 173–179. [38] B. Jülicher, P. Gowik, S. Uhlig, A top-down in-house validation based approach for the investigation of the measurement uncertainty using fractional factorial experiments, Analyst 124 (1999) 537–545. [39] InterValplus, Software for In-house Validation, quo data GmbH, Dresden, Germany, 2012. (http://www.quodata.de, accessed 7 November). [40] Document SANCO 2726/2004 rev. 4, Guidelines for the Implementation of Decision 2002/657/EC, http://ec.europa.eu/food/food/chemicalsafety/residues/cons_20042726rev4_en.pdf 2008. [41] D.J. Smith, The pharmacokinetics, metabolism, and tissue residues of betaadrenergic agonists in livestock, J. Anim. Sci. 76 (1998) 173–194. [42] P. Garcia, A.C. Paris, J. Gil, M.A. Popot, Y. Bonnaire, Analysis of β-agonists by HPLC/ ESI-MSn in horse doping control, Biomed. Chromatogr. 25 (2011) 147–154. [43] N. León, M. Roca, C. Igualada, C.P.B. Martins, A. Pastor, V. Yusá, Wide-range screening of banned veterinary drugs in urine by ultra high liquid chromatography coupled to high-resolution mass spectrometry, J. Chromatogr. A 1258 (2012) 55–65. [44] X.-D. Du, Y.-L. Wu, H.-J. Yang, T. Yang, Simultaneous determination of 10 _2-agonists in swine urine using liquid chromatography–tandem mass spectrometry and multiwalled carbon nanotubes as a reversed dispersive solid phase extraction sorbent, J. Chromatogr. A 1260 (2012) 25–32.