- Email: [email protected]
Multi-residue determination of phenolic and salicylanilide anthelmintics and related compounds in bovine kidney by liquid chromatography–tandem mass spectrometry
Journal of Chromatography A, 1216 (2009) 8200–8205
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Multi-residue determination of phenolic and salicylanilide anthelmintics and related compounds in bovine kidney by liquid chromatography–tandem mass spectrometry M. Caldow, M. Sharman ∗ , M. Kelly, J. Day, S. Hird, J.A. Tarbin Central Science Laboratory, Sand Hutton YO41 1LZ, UK
a r t i c l e
i n f o
Article history: Available online 9 April 2009 Keywords: Nitroxinil Oxyclozanide Rafoxanide Closantel LC–MS/MS Method validation
a b s t r a c t This paper describes an analytical method for four phenolic and salicylanilide anthelmintics authorised for use within the EU (nitroxinil, oxyclozanide, rafoxanide and closantel) in bovine kidney, and the extension of this procedure to include a number of related compounds; ioxynil, niclosamide, salicylanide and 3trifluoromethyl-4-nitrophenol (TFM). The method comprises a solvent extraction with 1% acetic acid in acetone and clean-up using a mixed-mode anion-exchange solid phase extraction column. Determination is by reversed phase LC–MS/MS. The method was validated to the latest EU requirements (Commission Decision 2002/657/EC) using both spiked and incurred tissues and was subject to second laboratory evaluation. Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved.
1. Introduction Compounds contained within the chemical grouping of phenolic and salicylanilide anthelmintics include nitroxinil, oxyclozanide, rafoxanide and closantel. All of these compounds are active against liver fluke and are used extensively to treat against fasciolosis and haemonchosis in sheep and cattle. Oxyclozanide is principally active against adult flukes, whereas nitroxinil, rafoxanide and closantel have activity against both adult and immature flukes [1]. The chemical structures and EU Maximum Residue Limits (MRLs) [2] for these compounds in bovine tissues are shown in Table 1. In each case, the marker residue for monitoring purposes is the parent compound. For nitroxinil, literature states that the most persistent residues are found in the kidneys and at the injection site [3]. High residue concentrations for the other three compounds have been found in the kidneys of animals used in controlled depletion studies [4–6], hence kidney is the matrix of choice when monitoring for their usage. Historically, single-residue methods have been mainly used to detect these compounds. For example, there are a number of methods available for the determination of nitroxinil in bovine milk and/or tissues utilising determination by HPLC-UV [7], HPLC-electrochemical detection [8], LC–MS [9] and GC [10]. LC-electrochemical detection has been used to determine five fasciolides, including nitroxinil and oxyclozanide, in bovine milk
∗ Corresponding author. E-mail address: [email protected] (M. Sharman).
[11]. Closantel determination has been undertaken using HPLCfluorescence in bovine milk [12] and in plasma and tissues [13] and a method has been reported for the analysis of closantel and rafoxanide in ovine plasma using HPLC-UV [14]. In addition LC–MS has been used for the determination of rafoxanide in various tissues [15]. Few, if any of these methods meet the method validation criteria of EU Legislation for confirmatory methods (Commission Decision 2002/657/EC [16]) and, since all pre-date 2002, none has been validated to this standard. Bound residues are also of particular concern; nitroxinil has been shown to undergo binding to plasma proteins [17]. Protein binding is also an issue with the salicylanilides [1]. A number of structurally related compounds also exist, which have a variety of uses, i.e. herbicides, anthelmintics, moluscicides, fungicides and piscicides. In order to test the applicability of this method to monitor for other potential contaminants, four such compounds were selected (see Table 1 for structures). Those compounds chosen included the herbicide ioxynil (structurally related to nitroxinil), the piscicide 3-trifluoromethyl-4-nitrophenol (TFM, structurally related to nitroxinil), and the salicylanilides niclosamide (a molluscicide and anthelmintic, used in conjunction with TFM) and salicylanilide (a fungicide). Methods for determination of these compounds in animal tissues, although somewhat limited, are available; for example, both TFM [18] and niclosamide [19] have been determined in trout and catfish tissues using HPLCUV. The aim of this work was therefore to: • Develop a multi-residue method for phenolic and salicylanilide anthelmintics and related compounds in bovine kidney.
0021-9673/$ – see front matter. Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2009.04.008
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
8201
Table 1 Compound structures and MS/MS transitions monitored. Bovine MRLsa (g kg−1 )
Quantification (Q) transition
Confirmation (C) transition
Typical ion ratio (Q/C)
Nitroxinil
400 muscle 400 kidney
288.9 > 126.9
288.9 > 162.0
2.1
Oxyclozanide
20 muscle 100 kidney
397.9 > 175.8
399.9 > 363.9
1.2
Closantel
1000 muscle 3000 kidney
660.8 > 126.9
662.8 > 126.9
1.6
Rafoxanide
30 muscle 40 kidney
623.8 > 126.9
625.8 > 126.9
2.1
Ioxynil
None
369.8 > 126.9
369.8 > 214.9
5.7
TFM
None
206.0 > 176.0
206.0 > 160.0
1.8
Niclosamide
None
325.0 > 170.9
325.0 > 289.0
2.6
Salicylanide
None
212.1 > 92.0
212.1 > 93.0
3.6
Compound
a
Structure
European Commission, Regulation 2377/90, as amended.
• Optimise the extraction of bound residues via the use of incurred tissues. • Validate the developed method to the standard of Commission Decision 2002/657/EC [16].
Sigma (Dorset, UK). 3-Trifluoromethyl-4-nitrophenol (TFM) was purchased from Aldrich (Dorset, UK). Solvents were HPLC grade or equivalent. Oasis MAX cartridges (150 mg/6 cm3 ) were purchased from Waters (Elstree, UK). All other chemicals were of analytical grade.
2. Materials and methods 2.2. Standard solutions 2.1. Chemicals and reagents Closantel, ioxynil, nitroxinil, oxyclozanide and rafoxanide were purchased from Riedel-de-Haan (Dorset, UK). Salicylanilide was purchased from Fluka (Dorset, UK). Niclosamide was obtained from
Stock standards were prepared in methanol for all compounds (except closantel which was prepared in acetone) at 1000 g mL−1 . These stock solutions were diluted in methanol to give a ‘mixed spiking standard solution’ containing nitroxinil (8 g mL−1 ), rafox-
8202
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
anide (0.8 g mL−1 ), closantel (60 g mL−1 ) and oxyclozanide, ioxynil, TFM, niclosamide and salicylinide (2 g mL−1 each).
dance with Commission Decision 2002/657/EC [16]. The validation exercise included determination of:
2.3. Sample preparation
(i) CC␣ (Decision limit)/CC (Detection capability) and repeatability. (ii) Method performance (both inter- and intra-laboratory). (iii) Specificity and selectivity.
2.3.1. Extraction 20 mL of either (a) 1% triethylamine in acetonitrile or (b) 1% acetic acid in acetone was added to sample (2 g). Following homogenisation (30 s) the extract was centrifuged at 3900 × g. Ammonia (5 mL, 0.88 sp. gr.) was added to the supernatant of procedure (b). 2.3.2. Clean-up The extract from Section 2.3.1 was applied to a Waters Oasis MAX (150 mg, 6 mL) pre-conditioned with ammonia (0.88 sp. Gr.) in acetonitrile (5 mL, 25% v/v). After washing the cartridge with acetonitrile (5 mL), the cartridge was dried under vacuum prior to elution with formic acid in acetonitrile (5 mL, 5% v/v). The eluate was evaporated to dryness under a stream of nitrogen at a temperature of <50 ◦ C and the residue re-dissolved in methanol (10 mL) prior to analysis by LC–MS/MS. 2.4. LC-ESI–MS/MS LC–MS/MS analysis was carried out on a Micromass Quattro Ultima® Pt triple quadrupole instrument in ES −ve mode. MS/MS detection was via the selected reaction monitoring (SRM) acquisition mode with two transitions being used for each analyte (see Table 1). The chromatographic separation employed a Hypersil Hypurity C18 3 m 50 × 2.1 mm column fitted with a C18 4 × 2.0 mm guard column. Mobile Phase (A) comprised 10 mM ammonium acetate solution in water and Mobile Phase B was 100% methanol. Initial conditions were 10% B at 0 min followed by a linear gradient to 90% B at 5 min, followed by a further 7 min at 90% B. Flow rate was 0.2 mL min−1 . 2.5. Calibration Matrix-matched calibrations, typically covering the range 0.25–4 ×MRL, were prepared immediately prior to LC–MS/MS by spiking known blank kidney extracts prepared according to Sections 2.3.1 and 2.3.2.
2.8. Analyte stability Commission Decision 2002/657/EC also requires an investigation of analyte stability in sample, sample extract and standards in solvent. To meet this requirement stability studies were conducted whereby analyte in solvent, cleaned-up sample extract, and spiked animal tissue were tested at intervals of 1, 2, 3, 4, and 10 weeks. 2.8.1. Solvent stability At week zero, freshly prepared aliquots of both the stock standards and the ‘mixed spiking solution’ (Section 2.2) were taken and stored under four conditions—room temperature in daylight, room temperature in darkness, +4 ◦ C (darkness) and −20 ◦ C (darkness). At each time point an aliquot was removed from each storage condition, diluted onto the calibration range and measured against freshly prepared solvent standards. 2.8.2. Extract stability At week zero blank samples of bovine kidney were extracted and the final extract spiked (post clean-up) with analytes at concentrations equivalent to their respective MRLs or, where MRLs were not available, a nominal concentration of 100 g kg−1 . These fortified extracts were stored under the same conditions listed in Section 2.8.1. At each time point an aliquot was removed from each condition and measured against freshly prepared matrix-matched standards. 2.8.3. Matrix stability At week zero replicate aliquots of blank bovine kidney were spiked at the MRL/nominal concentration and then stored in the dark at −20 ◦ C. At each time point six aliquots were extracted and analysed against a matrix-matched calibration curve (prepared at the time of extraction). 3. Results and discussion
2.6. Incurred samples 3.1. Method development Two sets of incurred bovine tissue were produced, one containing nitroxinil, the second containing oxyclozanide. One cow was dosed with Trodax® (active ingredient nitroxinil 34%) by subcutaneous injection at a rate of 10 mg active ingredient kg−1 bodyweight. Tissues were harvested five days post-treatment. Kidneys were homogenised, sub-sampled and stored at −20 ◦ C. One cow was dosed orally with Zanil® (active ingredient oxyclozanide 3.4%) at a rate of 10 mg active ingredient kg−1 bodyweight. Tissues were harvested five days post-treatment. The kidneys were homogenised, sub-sampled and stored at −20 ◦ C. Control tissue from a non-dosed animal was also produced and harvested in a similar manner. These sub-samples were used for two purposes: (i) in an investigation of the extractability of bound residues and (ii) as a Quality Control measure by inclusion with every batch of samples analysed by the developed method (see Sections 2.3.1 and 2.3.2). 2.7. Method validation procedure The method was validated in bovine kidney as a quantitative LC–MS/MS-based screening and confirmatory method in accor-
This laboratory previously developed a method for nitroxinil utilising an anion-exchange clean-up column in an “on-line” mode [7]. Triethylamine (1%, v/v, TEA) in acetonitrile was used as extraction solvent. This type of approach was also used during the initial development of the multi-residue method during which a number of “off-line” anion-exchange columns were investigated. Of those tested the Waters Oasis MAX (a mixed-mode polymeric anionexchange material) was shown to offer the best compromise for the range of compounds under investigation and a procedure consisting of solvent extraction, centrifugation followed by anion-change chromatography with retention under basic conditions and elution in acid, and evaporation followed by redissolution in methanol was developed. Chromatographic separation was achieved on a Hypersil HyPURITY C18 column prior to MS/MS analysis. Two transitions were monitored for each compound (Table 1). CC␣ and CC data for bovine kidney were generated using this procedure. As previously stated, protein binding is a recognised issue with these analytes; nitroxinil, for example, shows greater than 90% binding to proteins in plasma. In order to optimise the extraction efficiency of the bound residues, ‘real’ samples containing incurred
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
8203
Table 2 Effect of solvent composition on extraction of nitroxynil and oxyclozanide from incurred bovine kidney. Extraction solvent
1% TEA in MeCN 5% TEA in MeCN 1% acetic acid in acetone
Nitroxynil
Oxyclozanide
Spike recovery (as %)
Incurred sample concentration (as ×MRL)
Spike recovery (as %)
Incurred sample concentration (as ×MRL)
95 77 75
2.81 2.41 2.31
68 55 66
1.64 1.95 2.31
residues were needed, thus bovine tissues incurred with nitroxinil and with oxyclozanide were produced by dosing animals at therapeutic concentrations and harvesting tissues after slaughter (Section 2.6). In this second development phase a number of alternative extraction solvents were investigated by comparing data obtained from both spiked and incurred samples (Table 2). Using TEA (1%, v/v) in acetonitrile gave optimal extraction of incurred nitroxinil, but was less suitable for oxyclozanide. Conversely acetic acid (1%, v/v) in acetone proved to be the best solvent for extraction of incurred oxyclozanide. Although the spiked analyte recovery obtained for nitroxinil was less than that obtained using the TEAbased extractant, it was decided that this represented an acceptable compromise. Therefore acetic acid in acetone was used for all further work and the TEA extraction was only used for comparison during the determination of CC␣ and CC. The mean analyte recoveries for samples spiked at the MRL/nominal concentration with rafoxanide, closantel, ioxynil, and niclosamide, and extracted with acetic acid (1%, v/v) in acetone, were between 77% and 81% (n = 21 per compound). In the case of salicylanilide and TFM (both n = 21) the mean recoveries were lower at 61% and 33%, respectively. Typical ion chromatograms for each of the analytes (using the acetic acid in acetone extraction method) are shown in Fig. 1.
calibration curve procedure according to ISO 11843 was employed. In this case known blank material was fortified around the MRL (where available) in equidistant steps. Using this approach CC␣ and CC measurements for the licensed anthelmintics in bovine kidney were undertaken by extracting seven known blank samples spiked at three concentrations (0.5, 1.0 and 1.5 times MRL) on each of three separate days. For the additional compounds, where no veterinary medicine derived MRL exists, nominal concentrations of 50, 100 and 150 g kg−1 were used during validation. Data were obtained using both TEA (1%, v/v) in acetonitrile and acetic acid (1% v/v) in acetone as the extraction solvent. These data are presented in Table 3 for all compounds. Where veterinary medicine MRLs exist these data demonstrated that the method using an acetic acid in acetone extraction was capable of quantifying and confirming all target analytes and that the CC␣ values were typically 3.5–10% higher than the MRL concentration for each analyte. These data proved that the developed method was a robust procedure.
3.2. Method validation
3.2.2.1. An intra-laboratory assessment of the method by a second analyst. In this experiment a second analyst extracted a single batch containing blank bovine kidney (spiked at the three different concentrations (see Section 3.2.1). Data generated by the second analyst were comparable, in terms of precision and analyte recovery, to that of the primary analyst demonstrating that method could easily be transferred to another analyst within the same laboratory (data not shown).
3.2.1. Determination of CC˛/CCˇ and repeatability According to EU legislation (Commission Decision 2002/657/EC [16]) two key parameters must be measured during method validation for a confirmatory method: CC␣ (Decision limit) and CC (Detection capability). This legislation permits a number of approaches for determining these parameters and in this study the
3.2.2. Method performance To check method performance and identify any critical points within the developed method several additional experiments were undertaken including.
Fig. 1. Typical ion chromatograms for the each of the target compounds (quantification channel only).
8204
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
Table 3 Comparison of CC␣ and CC for bovine kidney obtained using different extraction solvents. TEA = 1%, v/v, TEA in MeCN; AcOH = 1%, v/v, AcOH in acetone (n = 21). MRLa (g kg−1 )
1st laboratory −1
TEA (g kg
Nitroxynil Oxyclozanide Closantel Rafoxanide Ioxynil Niclosamide TFM Salicylamide
400 100 3000 40 100a 100a 100a 100a
2nd laboratory −1
)
AcOH (g kg
AcOH (g kg−1 )
)
CC␣
CC
CC␣
CC
CC␣
CC
438 116 3672 49 11 15 12 23
477 133 4345 59 19 25 20 38
427 107 3106 44 5 7 3 7
454 114 3211 48 8 12 5 12
454 111 3425 43 NT NT NT NT
508 123 3851 46 NT NT NT NT
NT = not tested. a Nominal target concentration.
3.2.2.2. An inter-laboratory assessment of the method at a second laboratory. The method was transferred and set-up at a second laboratory. Once again CC␣/CC measurements in bovine kidney (for the original four compounds only) were obtained by extracting seven known blank samples spiked at three concentrations (0.5, 1.0 and 1.5 times MRL) on each of three separate days. A summary of the second laboratory CC␣/CC data, using the acetic acid in acetone extraction, is shown in Table 3. Data produced by this laboratory compare well to the primary data (Section 3.2.1) thus demonstrating the method to be both robust and transferable. 3.2.2.3. Applicability of the method to an additional matrix—bovine muscle. In addition to bovine kidney, the applicability of the method was also tested in bovine muscle for the four original compounds. The MRLs in bovine muscle are as follows: nitroxinil (400 g kg−1 ), oxyclozanide (20 g kg−1 ), closantel (1000 g kg−1 ) and rafoxanide (30 g kg−1 ). Seven blank samples were spiked at MRL concentrations. Due to the much lower MRL for oxyclozanide the final volume of the extract (prior to LC–MS/MS) was reduced from 10 to 5 mL to aid detection. Despite this change there was insufficient sensitivity in the method to allow reliable detection and quantification of oxyclozanide. In addition the chromatography and data were more variable than had previously been demonstrated during the analysis of kidney. Although the transfer of this method to muscle shows promise some further refinement of the method will be necessary before it could be employed for this additional matrix. 3.2.3. Specificity and selectivity Commission Decision 2002/657/EC [16] states that it is of prime importance that, “interference, which might arise from matrix components, is investigated”. In this study twenty different untreated samples of bovine or ovine kidney were extracted and analysed using the optimised method based on extraction with acetic acid (1%, v/v) in acetone. No measurable matrix interferences at the retention times of the eight analytes were observed in each of the twenty different samples—thus demonstrating the applicability and specificity of the method across each matrix type. The selectivity (discriminatory power) of the method was measured against the four structurally related compounds—ioxynil, niclosamide, salicylanilide and 3-trifluoromethyl-4-nitrophenol (TFM). Sets of blank samples of bovine kidney (three replicates per set) were spiked with one of these additional analytes at 100 g kg−1 , and analysed using the validated method. No LC–MS/MS responses were measured in these samples at the specific MRM transitions for nitroxinil, oxyclozanide, rafoxanide and closantel – thus demonstrating the selectivity of the method.
Fig. 2. Residue concentrations found in stored incurred tissue over time.
3.2.4. Analyte stability in solvent solutions and spiked samples These data showed that all eight stock standards (1000 g mL−1 ) are stable for at least 10 weeks under all of the storage conditions tested. Deterioration (i.e. reduction in measured concentration) was noted for oxyclozanide, closantel and rafoxanide after 1 week in those spiking solutions, which were stored at room temperature in daylight. Spiked sample extracts were stable under all conditions over the 10 weeks measured, except closantel and rafoxanide, which showed deterioration when stored at room temperature in the light. All eight compounds were stable when spiked onto bovine kidney and stored at −20 ◦ C over a period of 10 weeks prior to extraction. 3.2.5. Analyte stability in incurred material Although not a direct EU requirement governing method validation [16] the stability of residues in both incurred materials was measured over the timescale of this project. To achieve this a subsample of the material stored at −20 ◦ C was analysed with each analytical batch, including those undertaken by the second laboratory. The analyte concentrations in these materials were shown to remain stable for at least 3 months (Fig. 2). 4. Conclusions A robust screening and confirmatory method has been developed for all target analytes in bovine kidney. The final method was validated according to the requirements of Commission Decision 2002/657/EC [16] and was successfully transferred and validated at a second laboratory. Experiments using incurred tissues demonstrate that, of the solvents tested, the best solvent for extracting incurred residues
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
of nitroxinil and oxyclozanide in bovine kidney is 1% acetic acid in acetone. The consistency of data obtained from the repeated analysis of incurred tissues over a period of many months demonstrates both: (i) the reliability of the method and (ii) the stability of residues in stored tissues. Three of the eight analytes measured are unstable when stored in solvent at room temperature in the light. Analysts should take care to store sample extracts and spiking solutions in the dark and/or at lower (i.e. refrigerator or freezer) temperatures. Acknowledgements The financial support of the Veterinary Medicines Directorate of the UK Department for Environment, Food and Rural Affairs is gratefully acknowledged (Project No. VMO2146). The production of second laboratory validation data by HFL Sport Science (Fordham Cambridgeshire, UK) is gratefully acknowledged. References [1] Y. Bishop (Ed.), The Veterinary Formulary 6th Edition, London, 2005, p. 199. [2] European Commission, Official Journal of European Communities, 26 June 1990 (Regulation 2377/90, as amended).
8205
[3] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Nitroxinil Summary Report, EMEA/MRL/452/98-FINAL, 1998. [4] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Oxyclozanide Summary Report (3), EMEA/MRL/889/03-FINAL, 2004. [5] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Closantel Summary Report, http://www.emea. europa.eu/pdfs/vet/mrls/closantel.pdf. [6] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Rafoxanide Summary Report (2), EMEA/MRL/775/01-FINAL, 2001. [7] J.A. Tarbin, G. Shearer, J. Chromatogr. Biomed. Appl. 613 (1993) 347. [8] K. Takeba, M. Matsumoto, H. Nakazawa, J. Chromatogr. 596 (1992) 67. [9] W.J. Blanchflower, D.G. Kennedy, Analyst 114 (1989) 1013. [10] M. Kazacos, V. Mok, Aust. J. Dairy Technol. 41 (1986) 82. [11] K. Takeba, T. Itoh, M. Matsumoto, H. Nakazawa, S. Tanabe, J. AOAC Int. 79 (1996) 848. [12] G. Stoev, T. Dakova, A. Michailova, J. Chromatogr. A 846 (1999) 383. [13] G. Stoev, J. Chromatogr. B 710 (1998) 234. [14] H.A. Benchaoui, Q.A. McKellar, Biomed. Chromatogr. 7 (1993) 181. [15] W.J. Blanchflower, D.G. Kennedy, S.M. Taylor, J. Liquid Chromatogr. 13 (1990) 1595. [16] European Commission, Official Journal of European Communities, 12 August, 2002 (2002/657/EC). [17] M. Alvinerie, R. Floch, P. Galtier, J. Vet. Pharmacol. Therapeut. 14 (1991) 170. [18] T.D. Hubert, C. Vue, J.A. Bernardy, D.L. Van Horsen, M.I. Rossulek, J. AOAC Int. 84 (2001) 392. [19] T.M. Schreier, V.K. Dawson, Y. Choi, N.J. Spanjers, M.A. Boogard, J. Agric. Food Chem. 48 (2000) 2212.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Multi-residue determination of phenolic and salicylanilide anthelmintics and related compounds in bovine kidney by liquid chromatography–tandem mass spectrometry M. Caldow, M. Sharman ∗ , M. Kelly, J. Day, S. Hird, J.A. Tarbin Central Science Laboratory, Sand Hutton YO41 1LZ, UK
a r t i c l e
i n f o
Article history: Available online 9 April 2009 Keywords: Nitroxinil Oxyclozanide Rafoxanide Closantel LC–MS/MS Method validation
a b s t r a c t This paper describes an analytical method for four phenolic and salicylanilide anthelmintics authorised for use within the EU (nitroxinil, oxyclozanide, rafoxanide and closantel) in bovine kidney, and the extension of this procedure to include a number of related compounds; ioxynil, niclosamide, salicylanide and 3trifluoromethyl-4-nitrophenol (TFM). The method comprises a solvent extraction with 1% acetic acid in acetone and clean-up using a mixed-mode anion-exchange solid phase extraction column. Determination is by reversed phase LC–MS/MS. The method was validated to the latest EU requirements (Commission Decision 2002/657/EC) using both spiked and incurred tissues and was subject to second laboratory evaluation. Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved.
1. Introduction Compounds contained within the chemical grouping of phenolic and salicylanilide anthelmintics include nitroxinil, oxyclozanide, rafoxanide and closantel. All of these compounds are active against liver fluke and are used extensively to treat against fasciolosis and haemonchosis in sheep and cattle. Oxyclozanide is principally active against adult flukes, whereas nitroxinil, rafoxanide and closantel have activity against both adult and immature flukes [1]. The chemical structures and EU Maximum Residue Limits (MRLs) [2] for these compounds in bovine tissues are shown in Table 1. In each case, the marker residue for monitoring purposes is the parent compound. For nitroxinil, literature states that the most persistent residues are found in the kidneys and at the injection site [3]. High residue concentrations for the other three compounds have been found in the kidneys of animals used in controlled depletion studies [4–6], hence kidney is the matrix of choice when monitoring for their usage. Historically, single-residue methods have been mainly used to detect these compounds. For example, there are a number of methods available for the determination of nitroxinil in bovine milk and/or tissues utilising determination by HPLC-UV [7], HPLC-electrochemical detection [8], LC–MS [9] and GC [10]. LC-electrochemical detection has been used to determine five fasciolides, including nitroxinil and oxyclozanide, in bovine milk
∗ Corresponding author. E-mail address: [email protected] (M. Sharman).
[11]. Closantel determination has been undertaken using HPLCfluorescence in bovine milk [12] and in plasma and tissues [13] and a method has been reported for the analysis of closantel and rafoxanide in ovine plasma using HPLC-UV [14]. In addition LC–MS has been used for the determination of rafoxanide in various tissues [15]. Few, if any of these methods meet the method validation criteria of EU Legislation for confirmatory methods (Commission Decision 2002/657/EC [16]) and, since all pre-date 2002, none has been validated to this standard. Bound residues are also of particular concern; nitroxinil has been shown to undergo binding to plasma proteins [17]. Protein binding is also an issue with the salicylanilides [1]. A number of structurally related compounds also exist, which have a variety of uses, i.e. herbicides, anthelmintics, moluscicides, fungicides and piscicides. In order to test the applicability of this method to monitor for other potential contaminants, four such compounds were selected (see Table 1 for structures). Those compounds chosen included the herbicide ioxynil (structurally related to nitroxinil), the piscicide 3-trifluoromethyl-4-nitrophenol (TFM, structurally related to nitroxinil), and the salicylanilides niclosamide (a molluscicide and anthelmintic, used in conjunction with TFM) and salicylanilide (a fungicide). Methods for determination of these compounds in animal tissues, although somewhat limited, are available; for example, both TFM [18] and niclosamide [19] have been determined in trout and catfish tissues using HPLCUV. The aim of this work was therefore to: • Develop a multi-residue method for phenolic and salicylanilide anthelmintics and related compounds in bovine kidney.
0021-9673/$ – see front matter. Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2009.04.008
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
8201
Table 1 Compound structures and MS/MS transitions monitored. Bovine MRLsa (g kg−1 )
Quantification (Q) transition
Confirmation (C) transition
Typical ion ratio (Q/C)
Nitroxinil
400 muscle 400 kidney
288.9 > 126.9
288.9 > 162.0
2.1
Oxyclozanide
20 muscle 100 kidney
397.9 > 175.8
399.9 > 363.9
1.2
Closantel
1000 muscle 3000 kidney
660.8 > 126.9
662.8 > 126.9
1.6
Rafoxanide
30 muscle 40 kidney
623.8 > 126.9
625.8 > 126.9
2.1
Ioxynil
None
369.8 > 126.9
369.8 > 214.9
5.7
TFM
None
206.0 > 176.0
206.0 > 160.0
1.8
Niclosamide
None
325.0 > 170.9
325.0 > 289.0
2.6
Salicylanide
None
212.1 > 92.0
212.1 > 93.0
3.6
Compound
a
Structure
European Commission, Regulation 2377/90, as amended.
• Optimise the extraction of bound residues via the use of incurred tissues. • Validate the developed method to the standard of Commission Decision 2002/657/EC [16].
Sigma (Dorset, UK). 3-Trifluoromethyl-4-nitrophenol (TFM) was purchased from Aldrich (Dorset, UK). Solvents were HPLC grade or equivalent. Oasis MAX cartridges (150 mg/6 cm3 ) were purchased from Waters (Elstree, UK). All other chemicals were of analytical grade.
2. Materials and methods 2.2. Standard solutions 2.1. Chemicals and reagents Closantel, ioxynil, nitroxinil, oxyclozanide and rafoxanide were purchased from Riedel-de-Haan (Dorset, UK). Salicylanilide was purchased from Fluka (Dorset, UK). Niclosamide was obtained from
Stock standards were prepared in methanol for all compounds (except closantel which was prepared in acetone) at 1000 g mL−1 . These stock solutions were diluted in methanol to give a ‘mixed spiking standard solution’ containing nitroxinil (8 g mL−1 ), rafox-
8202
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
anide (0.8 g mL−1 ), closantel (60 g mL−1 ) and oxyclozanide, ioxynil, TFM, niclosamide and salicylinide (2 g mL−1 each).
dance with Commission Decision 2002/657/EC [16]. The validation exercise included determination of:
2.3. Sample preparation
(i) CC␣ (Decision limit)/CC (Detection capability) and repeatability. (ii) Method performance (both inter- and intra-laboratory). (iii) Specificity and selectivity.
2.3.1. Extraction 20 mL of either (a) 1% triethylamine in acetonitrile or (b) 1% acetic acid in acetone was added to sample (2 g). Following homogenisation (30 s) the extract was centrifuged at 3900 × g. Ammonia (5 mL, 0.88 sp. gr.) was added to the supernatant of procedure (b). 2.3.2. Clean-up The extract from Section 2.3.1 was applied to a Waters Oasis MAX (150 mg, 6 mL) pre-conditioned with ammonia (0.88 sp. Gr.) in acetonitrile (5 mL, 25% v/v). After washing the cartridge with acetonitrile (5 mL), the cartridge was dried under vacuum prior to elution with formic acid in acetonitrile (5 mL, 5% v/v). The eluate was evaporated to dryness under a stream of nitrogen at a temperature of <50 ◦ C and the residue re-dissolved in methanol (10 mL) prior to analysis by LC–MS/MS. 2.4. LC-ESI–MS/MS LC–MS/MS analysis was carried out on a Micromass Quattro Ultima® Pt triple quadrupole instrument in ES −ve mode. MS/MS detection was via the selected reaction monitoring (SRM) acquisition mode with two transitions being used for each analyte (see Table 1). The chromatographic separation employed a Hypersil Hypurity C18 3 m 50 × 2.1 mm column fitted with a C18 4 × 2.0 mm guard column. Mobile Phase (A) comprised 10 mM ammonium acetate solution in water and Mobile Phase B was 100% methanol. Initial conditions were 10% B at 0 min followed by a linear gradient to 90% B at 5 min, followed by a further 7 min at 90% B. Flow rate was 0.2 mL min−1 . 2.5. Calibration Matrix-matched calibrations, typically covering the range 0.25–4 ×MRL, were prepared immediately prior to LC–MS/MS by spiking known blank kidney extracts prepared according to Sections 2.3.1 and 2.3.2.
2.8. Analyte stability Commission Decision 2002/657/EC also requires an investigation of analyte stability in sample, sample extract and standards in solvent. To meet this requirement stability studies were conducted whereby analyte in solvent, cleaned-up sample extract, and spiked animal tissue were tested at intervals of 1, 2, 3, 4, and 10 weeks. 2.8.1. Solvent stability At week zero, freshly prepared aliquots of both the stock standards and the ‘mixed spiking solution’ (Section 2.2) were taken and stored under four conditions—room temperature in daylight, room temperature in darkness, +4 ◦ C (darkness) and −20 ◦ C (darkness). At each time point an aliquot was removed from each storage condition, diluted onto the calibration range and measured against freshly prepared solvent standards. 2.8.2. Extract stability At week zero blank samples of bovine kidney were extracted and the final extract spiked (post clean-up) with analytes at concentrations equivalent to their respective MRLs or, where MRLs were not available, a nominal concentration of 100 g kg−1 . These fortified extracts were stored under the same conditions listed in Section 2.8.1. At each time point an aliquot was removed from each condition and measured against freshly prepared matrix-matched standards. 2.8.3. Matrix stability At week zero replicate aliquots of blank bovine kidney were spiked at the MRL/nominal concentration and then stored in the dark at −20 ◦ C. At each time point six aliquots were extracted and analysed against a matrix-matched calibration curve (prepared at the time of extraction). 3. Results and discussion
2.6. Incurred samples 3.1. Method development Two sets of incurred bovine tissue were produced, one containing nitroxinil, the second containing oxyclozanide. One cow was dosed with Trodax® (active ingredient nitroxinil 34%) by subcutaneous injection at a rate of 10 mg active ingredient kg−1 bodyweight. Tissues were harvested five days post-treatment. Kidneys were homogenised, sub-sampled and stored at −20 ◦ C. One cow was dosed orally with Zanil® (active ingredient oxyclozanide 3.4%) at a rate of 10 mg active ingredient kg−1 bodyweight. Tissues were harvested five days post-treatment. The kidneys were homogenised, sub-sampled and stored at −20 ◦ C. Control tissue from a non-dosed animal was also produced and harvested in a similar manner. These sub-samples were used for two purposes: (i) in an investigation of the extractability of bound residues and (ii) as a Quality Control measure by inclusion with every batch of samples analysed by the developed method (see Sections 2.3.1 and 2.3.2). 2.7. Method validation procedure The method was validated in bovine kidney as a quantitative LC–MS/MS-based screening and confirmatory method in accor-
This laboratory previously developed a method for nitroxinil utilising an anion-exchange clean-up column in an “on-line” mode [7]. Triethylamine (1%, v/v, TEA) in acetonitrile was used as extraction solvent. This type of approach was also used during the initial development of the multi-residue method during which a number of “off-line” anion-exchange columns were investigated. Of those tested the Waters Oasis MAX (a mixed-mode polymeric anionexchange material) was shown to offer the best compromise for the range of compounds under investigation and a procedure consisting of solvent extraction, centrifugation followed by anion-change chromatography with retention under basic conditions and elution in acid, and evaporation followed by redissolution in methanol was developed. Chromatographic separation was achieved on a Hypersil HyPURITY C18 column prior to MS/MS analysis. Two transitions were monitored for each compound (Table 1). CC␣ and CC data for bovine kidney were generated using this procedure. As previously stated, protein binding is a recognised issue with these analytes; nitroxinil, for example, shows greater than 90% binding to proteins in plasma. In order to optimise the extraction efficiency of the bound residues, ‘real’ samples containing incurred
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
8203
Table 2 Effect of solvent composition on extraction of nitroxynil and oxyclozanide from incurred bovine kidney. Extraction solvent
1% TEA in MeCN 5% TEA in MeCN 1% acetic acid in acetone
Nitroxynil
Oxyclozanide
Spike recovery (as %)
Incurred sample concentration (as ×MRL)
Spike recovery (as %)
Incurred sample concentration (as ×MRL)
95 77 75
2.81 2.41 2.31
68 55 66
1.64 1.95 2.31
residues were needed, thus bovine tissues incurred with nitroxinil and with oxyclozanide were produced by dosing animals at therapeutic concentrations and harvesting tissues after slaughter (Section 2.6). In this second development phase a number of alternative extraction solvents were investigated by comparing data obtained from both spiked and incurred samples (Table 2). Using TEA (1%, v/v) in acetonitrile gave optimal extraction of incurred nitroxinil, but was less suitable for oxyclozanide. Conversely acetic acid (1%, v/v) in acetone proved to be the best solvent for extraction of incurred oxyclozanide. Although the spiked analyte recovery obtained for nitroxinil was less than that obtained using the TEAbased extractant, it was decided that this represented an acceptable compromise. Therefore acetic acid in acetone was used for all further work and the TEA extraction was only used for comparison during the determination of CC␣ and CC. The mean analyte recoveries for samples spiked at the MRL/nominal concentration with rafoxanide, closantel, ioxynil, and niclosamide, and extracted with acetic acid (1%, v/v) in acetone, were between 77% and 81% (n = 21 per compound). In the case of salicylanilide and TFM (both n = 21) the mean recoveries were lower at 61% and 33%, respectively. Typical ion chromatograms for each of the analytes (using the acetic acid in acetone extraction method) are shown in Fig. 1.
calibration curve procedure according to ISO 11843 was employed. In this case known blank material was fortified around the MRL (where available) in equidistant steps. Using this approach CC␣ and CC measurements for the licensed anthelmintics in bovine kidney were undertaken by extracting seven known blank samples spiked at three concentrations (0.5, 1.0 and 1.5 times MRL) on each of three separate days. For the additional compounds, where no veterinary medicine derived MRL exists, nominal concentrations of 50, 100 and 150 g kg−1 were used during validation. Data were obtained using both TEA (1%, v/v) in acetonitrile and acetic acid (1% v/v) in acetone as the extraction solvent. These data are presented in Table 3 for all compounds. Where veterinary medicine MRLs exist these data demonstrated that the method using an acetic acid in acetone extraction was capable of quantifying and confirming all target analytes and that the CC␣ values were typically 3.5–10% higher than the MRL concentration for each analyte. These data proved that the developed method was a robust procedure.
3.2. Method validation
3.2.2.1. An intra-laboratory assessment of the method by a second analyst. In this experiment a second analyst extracted a single batch containing blank bovine kidney (spiked at the three different concentrations (see Section 3.2.1). Data generated by the second analyst were comparable, in terms of precision and analyte recovery, to that of the primary analyst demonstrating that method could easily be transferred to another analyst within the same laboratory (data not shown).
3.2.1. Determination of CC˛/CCˇ and repeatability According to EU legislation (Commission Decision 2002/657/EC [16]) two key parameters must be measured during method validation for a confirmatory method: CC␣ (Decision limit) and CC (Detection capability). This legislation permits a number of approaches for determining these parameters and in this study the
3.2.2. Method performance To check method performance and identify any critical points within the developed method several additional experiments were undertaken including.
Fig. 1. Typical ion chromatograms for the each of the target compounds (quantification channel only).
8204
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
Table 3 Comparison of CC␣ and CC for bovine kidney obtained using different extraction solvents. TEA = 1%, v/v, TEA in MeCN; AcOH = 1%, v/v, AcOH in acetone (n = 21). MRLa (g kg−1 )
1st laboratory −1
TEA (g kg
Nitroxynil Oxyclozanide Closantel Rafoxanide Ioxynil Niclosamide TFM Salicylamide
400 100 3000 40 100a 100a 100a 100a
2nd laboratory −1
)
AcOH (g kg
AcOH (g kg−1 )
)
CC␣
CC
CC␣
CC
CC␣
CC
438 116 3672 49 11 15 12 23
477 133 4345 59 19 25 20 38
427 107 3106 44 5 7 3 7
454 114 3211 48 8 12 5 12
454 111 3425 43 NT NT NT NT
508 123 3851 46 NT NT NT NT
NT = not tested. a Nominal target concentration.
3.2.2.2. An inter-laboratory assessment of the method at a second laboratory. The method was transferred and set-up at a second laboratory. Once again CC␣/CC measurements in bovine kidney (for the original four compounds only) were obtained by extracting seven known blank samples spiked at three concentrations (0.5, 1.0 and 1.5 times MRL) on each of three separate days. A summary of the second laboratory CC␣/CC data, using the acetic acid in acetone extraction, is shown in Table 3. Data produced by this laboratory compare well to the primary data (Section 3.2.1) thus demonstrating the method to be both robust and transferable. 3.2.2.3. Applicability of the method to an additional matrix—bovine muscle. In addition to bovine kidney, the applicability of the method was also tested in bovine muscle for the four original compounds. The MRLs in bovine muscle are as follows: nitroxinil (400 g kg−1 ), oxyclozanide (20 g kg−1 ), closantel (1000 g kg−1 ) and rafoxanide (30 g kg−1 ). Seven blank samples were spiked at MRL concentrations. Due to the much lower MRL for oxyclozanide the final volume of the extract (prior to LC–MS/MS) was reduced from 10 to 5 mL to aid detection. Despite this change there was insufficient sensitivity in the method to allow reliable detection and quantification of oxyclozanide. In addition the chromatography and data were more variable than had previously been demonstrated during the analysis of kidney. Although the transfer of this method to muscle shows promise some further refinement of the method will be necessary before it could be employed for this additional matrix. 3.2.3. Specificity and selectivity Commission Decision 2002/657/EC [16] states that it is of prime importance that, “interference, which might arise from matrix components, is investigated”. In this study twenty different untreated samples of bovine or ovine kidney were extracted and analysed using the optimised method based on extraction with acetic acid (1%, v/v) in acetone. No measurable matrix interferences at the retention times of the eight analytes were observed in each of the twenty different samples—thus demonstrating the applicability and specificity of the method across each matrix type. The selectivity (discriminatory power) of the method was measured against the four structurally related compounds—ioxynil, niclosamide, salicylanilide and 3-trifluoromethyl-4-nitrophenol (TFM). Sets of blank samples of bovine kidney (three replicates per set) were spiked with one of these additional analytes at 100 g kg−1 , and analysed using the validated method. No LC–MS/MS responses were measured in these samples at the specific MRM transitions for nitroxinil, oxyclozanide, rafoxanide and closantel – thus demonstrating the selectivity of the method.
Fig. 2. Residue concentrations found in stored incurred tissue over time.
3.2.4. Analyte stability in solvent solutions and spiked samples These data showed that all eight stock standards (1000 g mL−1 ) are stable for at least 10 weeks under all of the storage conditions tested. Deterioration (i.e. reduction in measured concentration) was noted for oxyclozanide, closantel and rafoxanide after 1 week in those spiking solutions, which were stored at room temperature in daylight. Spiked sample extracts were stable under all conditions over the 10 weeks measured, except closantel and rafoxanide, which showed deterioration when stored at room temperature in the light. All eight compounds were stable when spiked onto bovine kidney and stored at −20 ◦ C over a period of 10 weeks prior to extraction. 3.2.5. Analyte stability in incurred material Although not a direct EU requirement governing method validation [16] the stability of residues in both incurred materials was measured over the timescale of this project. To achieve this a subsample of the material stored at −20 ◦ C was analysed with each analytical batch, including those undertaken by the second laboratory. The analyte concentrations in these materials were shown to remain stable for at least 3 months (Fig. 2). 4. Conclusions A robust screening and confirmatory method has been developed for all target analytes in bovine kidney. The final method was validated according to the requirements of Commission Decision 2002/657/EC [16] and was successfully transferred and validated at a second laboratory. Experiments using incurred tissues demonstrate that, of the solvents tested, the best solvent for extracting incurred residues
M. Caldow et al. / J. Chromatogr. A 1216 (2009) 8200–8205
of nitroxinil and oxyclozanide in bovine kidney is 1% acetic acid in acetone. The consistency of data obtained from the repeated analysis of incurred tissues over a period of many months demonstrates both: (i) the reliability of the method and (ii) the stability of residues in stored tissues. Three of the eight analytes measured are unstable when stored in solvent at room temperature in the light. Analysts should take care to store sample extracts and spiking solutions in the dark and/or at lower (i.e. refrigerator or freezer) temperatures. Acknowledgements The financial support of the Veterinary Medicines Directorate of the UK Department for Environment, Food and Rural Affairs is gratefully acknowledged (Project No. VMO2146). The production of second laboratory validation data by HFL Sport Science (Fordham Cambridgeshire, UK) is gratefully acknowledged. References [1] Y. Bishop (Ed.), The Veterinary Formulary 6th Edition, London, 2005, p. 199. [2] European Commission, Official Journal of European Communities, 26 June 1990 (Regulation 2377/90, as amended).
8205
[3] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Nitroxinil Summary Report, EMEA/MRL/452/98-FINAL, 1998. [4] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Oxyclozanide Summary Report (3), EMEA/MRL/889/03-FINAL, 2004. [5] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Closantel Summary Report, http://www.emea. europa.eu/pdfs/vet/mrls/closantel.pdf. [6] The European Agency for the Evaluation of Medicinal Products, Committee for Veterinary Medicinal Products, Rafoxanide Summary Report (2), EMEA/MRL/775/01-FINAL, 2001. [7] J.A. Tarbin, G. Shearer, J. Chromatogr. Biomed. Appl. 613 (1993) 347. [8] K. Takeba, M. Matsumoto, H. Nakazawa, J. Chromatogr. 596 (1992) 67. [9] W.J. Blanchflower, D.G. Kennedy, Analyst 114 (1989) 1013. [10] M. Kazacos, V. Mok, Aust. J. Dairy Technol. 41 (1986) 82. [11] K. Takeba, T. Itoh, M. Matsumoto, H. Nakazawa, S. Tanabe, J. AOAC Int. 79 (1996) 848. [12] G. Stoev, T. Dakova, A. Michailova, J. Chromatogr. A 846 (1999) 383. [13] G. Stoev, J. Chromatogr. B 710 (1998) 234. [14] H.A. Benchaoui, Q.A. McKellar, Biomed. Chromatogr. 7 (1993) 181. [15] W.J. Blanchflower, D.G. Kennedy, S.M. Taylor, J. Liquid Chromatogr. 13 (1990) 1595. [16] European Commission, Official Journal of European Communities, 12 August, 2002 (2002/657/EC). [17] M. Alvinerie, R. Floch, P. Galtier, J. Vet. Pharmacol. Therapeut. 14 (1991) 170. [18] T.D. Hubert, C. Vue, J.A. Bernardy, D.L. Van Horsen, M.I. Rossulek, J. AOAC Int. 84 (2001) 392. [19] T.M. Schreier, V.K. Dawson, Y. Choi, N.J. Spanjers, M.A. Boogard, J. Agric. Food Chem. 48 (2000) 2212.