Development and Validation of a Hybrid Screening and Quantitative

Feb 27, 2018 - European Union (EU).24 In the end, the best compromise proved to be ..... Food of Animal Origin; EU: Brussels, Belgium, 2003. (25) Wang...
1 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/JAFC

Cite This: J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Development and Validation of a Hybrid Screening and Quantitative Method for the Analysis of Eight Classes of Therapeutants in Aquaculture Products by Liquid Chromatography−Tandem Mass Spectrometry Ryan S. Gibbs,* Shauna L. Murray, Lynn V. Watson, Brandy P. Nielsen, Ross A. Potter, and Cory J. Murphy Dartmouth Laboratory, Canadian Food Inspection Agency, 1992 Agency Drive, Dartmouth, Nova Scotia B3B 1Y9, Canada S Supporting Information *

ABSTRACT: A method using reverse-phase ultra-high-performance liquid chromatography coupled with tandem mass spectrometry is described for eight classes of therapeutants that are used in marine aquaculture products. Validation studies to evaluate recovery, precision, method detection limits, and measurement uncertainty were performed at three levels, using three representative matrices [salmon (fatty fish), tilapia (lean fish), and shrimp (crustaceans)] to assess the method performance for use as a screening or determinative (quantitative and confirmatory) method. A total of 16 sulfonamides (plus 2 potentiators), 2 tetracyclines, 11 (fluoro)quinolones, 7 nitroimidazoles, 3 amphenicols, 5 steroids, and 3 stilbenes met the quantitative criteria for method validation. An additional 5 triphenylmethane dyes, 2 sulfonamides, 2 tetracyclines, and 1 amphenicol met the required performance for use as a screening method. Limits of detection (LODs) for the compounds that met the quantitative criteria ranged from 0.1 to 5 μg/kg, while LODs for compounds from the screening group ranged from 0.1 to 30 μg/kg. This method provides a comprehensive approach to the determination of different classes of compounds in aquaculture products. KEYWORDS: veterinary drug residues, growth promoters, multi-residue methods, multi-class methods, LC−MS/MS, fish, crustaceans

■

INTRODUCTION Antimicrobial drugs are an important component of overall human and animal health. Many therapeutants are used in food-producing animals for disease prevention and treatment as well as growth promotion. Many of these compounds are approved for use with proper veterinary guidance; however, there are others that are either banned or not approved for use in specific species. Despite the lack of approval, these compounds may be used “off-label” when other compounds are not available or for a financial advantage in the increasingly competitive global food economy. Abuse or overuse of both approved and unregulated compounds is one of many factors that has led to the emergence of antimicrobial-resistant (AMR) strains of microorganisms. These AMR microorganisms represent a significant emerging risk to both human and animal health. The World Health Organization (WHO) and Food and Agriculture Organization of the United Nations (FAO) have recently published guidance on preventing further spread of AMR microorganisms.1,2 WHO has published a list of critically important antimicrobials (CIA) in human medicine to assist all stakeholders in ensuring that the CIA compounds are used prudently.3 Detection of these residues in food samples is critical to keep these residues from entering the food supply and contributing to antimicrobial resistance. The Canadian Food Inspection Agency (CFIA) monitors food products in Canada for the presence of a number of potentially harmful residues to safeguard consumers. The Dartmouth Laboratory of the CFIA is responsible for the analysis of veterinary therapeutant residues in fish and fishPublished XXXX by the American Chemical Society

based products. Currently, the laboratory provides results for regulatory monitoring programs for 13 classes of therapeutants in fish products. These include tetracyclines, sulfonamides, amphenicols, quinolones (including fluoroquinolones), macrolides, nitroimidazoles, nitrofurans, triphenylmethane dyes, avermectins, benzoylureas, pyrethroids, stilbenes, and steroids. Therapeutant monitoring has historically been conducted using large numbers of complex single-class, multi-step methods to ensure the best possible detection limits and specificity for each class of compound.4−10 These methods rely on liquid chromatography−tandem mass spectrometry (LC− MS/MS) as the detection technique because it provides sufficient quantitative performance and structural information to identify a suspected residue. As testing requirements have increased, this approach has become more difficult to implement in an efficient manner. Conversion to a multiclass, multi-residue (MCMR) method for screening and quantitative purposes was required. Technological advances in the sensitivity of tandem mass spectrometers has made it possible to achieve the required detection limits, incorporating several analytes with much simpler dilution methods. The goal of this project was to incorporate as many of the therapeutant classes as possible into a simple, rapid, extraction Special Issue: 54th North American Chemical Residue Workshop Received: November 17, 2017 Revised: February 9, 2018 Accepted: February 14, 2018

A

DOI: 10.1021/acs.jafc.7b05357 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Table 1. Primary Standards Used in the Method, Including Identifying Acronym and Supplier compound

acronym

CAS Registry Number

supplier

mixed standard concentration (μg/L)

epinandrolone nandrolone 17α-methyltestosterone 17β-boldenone 17α-boldenone ormetomprim trimethoprim sulfamethizole sulfisoxazole sulfamoxole sulfamerazine sulfathiazole sulfamethoxazole sulfadiazine sulfapyridine sulfaguanidine sulfacetamide sulfanilamide sulfadoxin sulfadimethoxine sulfaquinoxaline sulfachloropyridazine sulfamethoxypyridazine sulfamonomethoxine sulfamethazine malachite green leucomalachite green crystal violet leucocrystal violet brilliant green nalidixic acid flumequin oxolinic acid ciprofloxacin danofloxacin enrofloxacin sarafloxacin marbofloxacin norfloxacin orbifloxacin difloxacin erythromycin thiamphenicol chloramphenicol florfenicol florfenicol amine hexestrol diethylstilbestrol dienestrol oxytetracycline tetracycline chlortetracycline doxycycline ronidazole metronidazole−OH ipronidazole−OH metronidazole ipronidazole HMMNI dimetridazole

epiNAN NAN MT BOLD epiBOLD OMP TMP SMZL SIX SXL SMR STZ SMX SDZ SPY SGD SAA SNL SDX SDM SQX SCP SMP SMM SMZ MG LMG CV LCV BG NLDX FLMQ OXO CIPRO DANO ENRO SARA MARBO NOR ORBI DIFLOX ERY TAP CAP FLR FLRA HEX DES DIEN OTC TC CTC DOXY RNZ MNZ−OH IPZ−OH MNZ IPZ HMMNI DMZ

4409-34-1 434-22-0 58-18-4 846-48-0 27833-18-7 6981-18-6 738-70-5 144-82-1 127-69-5 729-99-7 127-79-7 72-14-0 723-46-6 68-35-9 144-83-2 57-67-0 144-80-9 63-74-1 2447-57-6 122-11-2 59-40-5 80-32-0 80-35-3 1220-83-3 57-68-1 2437-29-8 129-73-7 547-62-9 603-48-5 633-03-4 389-08-2 42835-25-6 14698-29-4 85721-33-1 112398-08-0 93106-60-6 91296-87-6 115550-35-1 70458-96-7 113617-63-3 91296-86-5 114-07-8 15318-45-3 56-75-7 73231-34-2 76639-93-5 84-16-2 56-53-1 84-17-3 2058-46-0 64-75-5 64-72-2 24390-14-5 7681-76-7 4812-40-2 35175-14-5 443-48-1 14885-29-1 839-05-0 551-92-8

Steraloidsa Sigmab Sigmab TRCc TRCc Hoffman-Laroched Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab Sigmab WITEGAe Sigmab Sigmab

100 100 100 100 100 50 50 100 50 100 50 50 50 50 50 5000 300 9000 50 50 50 50 50 50 50 20 10 100 10 20 100 50 50 100 200 100 100 200 200 50 200 400 200 30 50 9000 250 150 250 1000 1000 2000 2000 300 300 200 200 200 500 200

B

DOI: 10.1021/acs.jafc.7b05357 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Table 1. continued a

Newport, RI, U.S.A. bSt. Louis, MO, U.S.A. cToronto, Ontario, Canada. dNutley, NJ, U.S.A. eBerlin, Germany. development and validation of this method. Finfish samples were received by the lab and stored at −20 °C. Samples were allowed to partially thaw at room temperature or fully thaw overnight at 4 °C prior to final preparation, which included removal of all skin, bones, and other organs with only the muscle tissue remaining. The muscle tissue was blended until homogeneous in a Robot Coupe Blixer food processor and stored in 8 oz. screw-capped plastic sample containers at −20 °C until analysis. All shrimp tissue had shells removed after thawing and were homogenized and stored as stated above. Sample Extraction. A total of 2 ± 0.1 g of homogeneous muscle tissue was weighed into a 50 mL polypropylene (PP) tube. For fortified samples, an aliquot of working standard was added to samples prior to continuing with the extraction procedure. Samples were extracted with 17 mL of 80% ACN containing 1% FA and homogenized using a Polytron homogenizer equipped with a 20 mm generator to ensure a maximum tissue surface area for solvent extraction. All tubes were capped and vortexed at 1500 rpm in pulsed mode for 15 min using a multi-tube vortex-mixing machine. After vortexing, sample tubes were placed in an ultrasonic bath for an additional 15 min. All tubes were spun at 5250 relative centrifugal force (RCF) for 10 min. The supernatant was transferred into a second 50 mL PP tube, made up to 20 mL with 80% ACN, and mixed. A 5 mL aliquot of this mixture was transferred to a 14 mL Falcon tube (Fisher Scientific) and evaporated under nitrogen to approximately 3 mL at 45 °C to remove some ACN from the sample. Samples were then made up to 5 mL with DIW, corresponding to a final ACN concentration of approximately 40%. This resulted in a final tissue concentration of the extract of 0.1 g of tissue/mL of extract. A 1.5 mL aliquot of the extract was transferred to a 2 mL microcentrifuge tube and centrifuged at 16 000 RCF for 5 min. Taking care to avoid disturbing the residual solids, the supernatant was transferred to a 2 mL glass autosampler vial and capped for LC−MS/MS analysis. Preparation of the Matrix-Matched Calibration Curve. A matrix-matched external calibration curve was used for quantification with this method. A portion of the blank tissue was extracted as per the procedure above and used as a diluent for the calibration curve. To prepare the matrix-matched calibration curve, a 1:1 dilution of the working standard was prepared by adding equal volumes of working standard and 80% ACN. Appropriate volumes of the 1:1 diluted working standard and 80% ACN (listed in Table S1 of the Supporting Information) were added to 2 mL volumetric tubes. All standards were made to volume with blank tissue extract. After mixing, approximately 1.5 mL was transferred to a 2 mL microcentrifuge tube and centrifuged at 16 000 RCF for 5 min. The resulting supernatant was transferred to a 2 mL glass autosampler vial. Ultra-High-Performance Liquid Chromatography−Tandem Mass Spectrometry (UHPLC−MS/MS) Conditions. UHPLC− MS/MS experiments were conducted on two different systems from different vendors to ensure that the method was suitable on multiple platforms. The first system was a Waters I-Class UPLC with a flowthrough needle autosampler coupled to a TQ-S Micro tandem mass spectrometer (Milford, MA, U.S.A.). The second system was an Agilent 1290 UHPLC (Santa Clara, CA, U.S.A.) coupled to a Sciex 5500 quadrupole linear ion trap (QTRAP) mass spectrometer (Concord, Ontario, Canada). Chromatographic separation on both systems was accomplished with a Waters 50 × 2.1 mm, 1.7 μm HSS T3 UPLC column equipped with a 5 × 2.1 mm, Vanguard HSS T3 guard column held at 30 °C. A gradient separation was employed where the system started at 0% B mobile phase at 0.5 mL/min. These conditions were held for 0.5 min after injection prior to a 3 min linear ramp to 95% B from 0.5 to 3.5 min. This condition was held for 1 min prior to the system returning to initial conditions to re-equilibrate. A sample injection volume of 3 μL was injected on the Waters I-Class system, and 4 μL was injected on the Agilent 1290 system. To optimize the detection limits for all compounds in the method, two

method that would meet the detection limit requirements for monitoring programs in Canada. Each compound/class was evaluated to determine suitability in both a screening-based approach and a quantitative/confirmatory capacity. A number of methods have been reported with a wide scope of antibiotic compounds in the muscle of terrestrial animals and other fish products;11−16 however, to our knowledge, no methods have been reported for this particular scope of analytes in fish products.

■

MATERIALS AND METHODS

Primary Standards and Reagents. All primary standards for these methods, except for ormetoprim, were purchased from commercial suppliers. Table 1 lists each standard, the identifying acronym, the CAS Registry Number, the nominal concentration, and the supplier. Distilled deionized water (DIW) was purified using a Millipore Water purification system. High-performance liquid chromatography (HPLC)-grade acetonitrile (ACN), HPLC-grade methanol (MeOH), and hexane (distilled in glass grade) were purchased from Caledon Laboratories (Georgetown, Ontario, Canada). Liquid chromatography−mass spectrometry (LC−MS)-grade ACN, MeOH, and ammonium formate (AF), glacial acetic acid (AcOH), nitric acid, and sodium hydroxide (NaOH) pellets were purchased from Fisher Scientific (Ottawa, Ontario, Canada). Formic acid (FA, 98%) was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Stock and Working Standard Solutions. (Fluoro)quinolone stock standards (500 mg/L) were prepared in 30 mM NaOH. Triphenylmethane dye (TPMD), nitroimidazole, and sulfonamide stock standards (100 mg/L) were prepared in ACN. Tetracycline, steroid, stilbene, and amphenicol stock standards (100 mg/L) were prepared in MeOH. Both MT and NAN primary standards were provided as 1 mg/mL solutions from the supplier. All stock standards were corrected for purity and stored in amber glass bottles. Table 2 lists the storage temperatures and stability information for the stock standards used with this method.

Table 2. Stock Standard Storage Conditions and Stability class/compound

storage condition (°C)

tetracyclines sulfonamides amphenicols triphenylmethane dyes CIPRO, DANO, SARA, and ORBI ENRO, FLMQ, NLDX, and MARBO NOR, DIFLOX, and OXO nitroimidazoles stilbenes steroids

−20 4 −20 4 4 4 4 4 4 4

stability 1 1 1 1 1 6 3 1 1 1

year year year year year months months year year year

For method development and validation studies, mixed standards were prepared daily, while long-term stability studies were conducted on the mixed working standards. The procedure for preparing standards with the routine method is highlighted in Figure 1. The compound concentrations in the standards were designed with the method screening limits (SLs) in mind, so that compounds were relative to their determined SLs. Sample Preparation. Samples used for these experiments were submitted to the lab as regulatory samples under various programs conducted by the CFIA. Samples that were reported as “not detected” for all classes were selected and used in spiking studies for the C

DOI: 10.1021/acs.jafc.7b05357 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Figure 1. Mixed standard preparation workflow for the MCMR analysis method. separate mobile phase systems were employed using the same gradient for different classes of compounds. The first mobile phase system, which is used for the sulfonamides, quinolones, steroids, macrolides, and triphenylmethane dyes, consisted of 5 mM AF + 0.05% FA in DIW for the “A” mobile phase, and pure ACN as the “B” mobile phase. The second mobile phase system, which is used for the amphenicols, stilbenes, tetracyclines, and nitroimidazoles, used 0.1% AcOH for “A” and 0.1% AcOH in ACN for “B”. On the 5500 QTRAP system, samples were injected with the AcOH mobile phase system twice, once for the positive electrospray ionization (ESI+) compounds and again for the negative electrospray ionization (ESI−) compounds, resulting in three total injections. This was necessary, because the settling time to switch from positive to negative electrospray during a single run was not fast enough to maintain a 10 ms dwell time for each component in the method. Source gas flows, temperatures, and other source-specific settings were optimized for each compound by either infusing the compound into a stream of mobile phase at the approximate composition where the compound elutes or performing replicate injections using the full chromatographic method while adjusting the compound-specific voltage settings to determine which settings gave the largest peak to peak signal/noise ratio (S/N). At least two ion reactions were optimized for each compound. Tables S2 and S3 of the Supporting Information show the source-specific parameters for the TQ-S Micro and 5500 QTRAP, respectively. Compound-specific parameters are listed for both instruments in Table S4 of the Supporting Information. The MS/MS methods on each instrument were designed to operate in two different workflow scenarios:

(1) As a screening method: The instrument was programmed to monitor only the most sensitive ion transition for each compound to maximize the number of compounds that could be monitored in a single injection with acceptable dwell times. (2) As a confirmatory method: Confirmatory methods were designed to look for two ion reactions per compound during a specified time window to ensure that the full duty cycle was used to obtain the best MS data quality for quantifying and confirming residue results. Both methods were programmed to acquire at least 10 data points across each peak to ensure acceptable peak definition for each compound. Mixed Standard Stability Studies. Mixed standards were prepared at the required concentrations and stored at 4 °C. At set time points (1−4 days, 1−4 weeks, and 2, 3, 4, 6, 9, and 12 months), these standards were analyzed against fresh standards prepared from stock standards, and the difference was calculated. Any compound that produced less than 90% of the fresh standard response was considered to lack sufficient stability to continue use with the final method without a fresh preparation. Method Validation. There are a number of guideline documents available that provide guidance on how to conduct validation studies for regulatory analytical methods in food products.17−19 Muscle tissue from Atlantic salmon (Salmo salar) was used to represent high-oil finfish; various tilapia species were used to represent lean finfish; and various species of raw shrimp were chosen to represent crustaceans. The validation procedure is primarily based on the guidelines set out in 2002/657/EC and was designed to evaluate linear range, matrix effects, ruggedness, limit of detection (LOD), limit of identification D

DOI: 10.1021/acs.jafc.7b05357 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Figure 2. Separation profile for the MCMR method on a 2.1 × 50 mm Waters HSS T3 column with a linear 32%/min gradient after a 30 s hold at 100% A: (A) mobile phase A, 5 mm AF + 0.05% FA; mobile phase B, ACN; and column temperature, 30 °C and (B) mobile phase A, 0.1% AcOH; mobile phase B, 0.1% AcOH in ACN; and column temperature, 30 °C. (LOI), recovery, precision, and measurement uncertainty (MU). The workflow of the method also necessitated the determination of another method limit: the SL, which was the LOD for the screening workflow MS method. The specific experimental design of how each criterion was evaluated is included in the Supporting Information. The acceptance criteria for recovery and precision are listed in Table S14 of the Supporting Information.

concentrations of modifiers gave improved responses for most compounds. Two mobile phase systems were chosen because a single set of conditions could not be established that was suitable for all of the compounds of interest. A 5 mM AF + 0.05% FA and ACN system resulted in excellent responses for the sulfonamides, quinolones, TPMDs, and steroids; however, it caused a complete suppression of the stilbenes. Because stilbenes are banned residues in Canada, low LODs are required. A 0.1% AcOH in water and ACN system provided the best response for the tetracyclines, stilbenes, amphenicols, and nitroimidazoles. AcOH resulted in reduced signals for the quinolones, which were required to be detected in the 1 μg/kg range for regulatory purposes. After evaluation of the pros and cons of each phase, the Waters HSS T3 column was chosen because the compounds that were poorly retained were compounds that were lower priority as a result of the historical lack of samples containing these residues (SNL and SGD) or a relatively high regulatory limit (FLRA). HSS T3 is also compatible with 100% aqueous mobile phases without suffering from phase collapse, which allowed for better retention of the polar early eluting compounds. Figure 2 displays the final overall separation of a mixed standard prepared in 40% ACN for each mobile phase system used in this method. Figures S5−S13 of the Supporting Information provide more detailed class-specific separation profiles. Extraction Method Development. There is no shortage of MCMR methods described in the scientific literature. It would be impossible to provide references to every method that is available; however, many methods that include similar compound classes in animal tissues were studied as the basis for developing extraction and instrument methods.11,13,15,16,21 Mastovska published an excellent review of multi-residue methods used for the analysis of foods of animal origin that includes a section on MCMR methods as well.22 Most methods that target the residues of interest use a high percentage of polar organic solvents (ACN or MeOH). Some methods contain acids, buffers,16 or other additives16,21 to assist with pH control, increase solubility, prevent compound

■

RESULTS AND DISCUSSION Initial Method Development. In examining the compounds analyzed at the laboratory, it was known from the outset that not all of the classes would be amenable to a MCMR method. The nitrofurans, for example, are typically derivatized with 2-nitrobenzaldehyde and are quantified as their nitrophenyl derivatives.20 A general screening method that does not include the derivatization step would likely not be suitable for nitrofuran analysis. Several other compounds were difficult to ionize using electrospray, even in the absence of the sample matrix; therefore, it was not possible to meet the detection limit requirements for these compounds. In the end, nine classes, tetracyclines, sulfonamides, quinolones, amphenicols, macrolides, TPMDs, nitroimidazoles, stilbenes, and steroids, were chosen to continue investigating during initial method development studies. Instrument Setup and Optimization. Initial optimization of the compounds for MS/MS detection was performed by infusing a constant stream of compound into the mobile phase. The first step in the method development process was to develop an appropriate UHPLC separation, which provided adequate separation and peak shape for most of the compounds. Most methods use some variant of C18 columns for chromatographic separation of these compounds. C18, phenyl, and pentafluorophenyl (PFP) columns from several manufacturers were chosen for evaluation. Several combinations of mobile phases and modifiers were tested with each column. A comparison of ACN and MeOH as the organic solvent was conducted, with ACN providing significantly better peak shape and responses for many compounds. Several acidic and buffered mobile phases were tested to optimize instrument response. In general, lower E

DOI: 10.1021/acs.jafc.7b05357 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Figure 3. Response of a mixed standard stored at 4 °C for various time points compared to a standard prepared from stock standards. The dashed line represents the 90% response threshold.

included in the “unstable mix” of standards. The sulfonamides SIX, SMP, and OMP all showed less than 90% response versus the fresh standard after 2 weeks. The tetracyclines OTC and CTC also showed less than 90% response after 1 week. Results of the first month of the study are displayed in Figure 3. Method Validation. Linearity and Matrix Effects. Examination of the residual plots for all standards showed good linearity for all compounds in both 40% ACN and in all three matrices when prepared as described in the linearity section of the protocol. The difference in slope of the matrixmatched curves for each representative matrix however resulted in a >10% difference for greater than 90% of the compounds on both instruments. This resulted in a requirement to use matrixmatched standards and/or recovery correction for acceptable quantitative performance using this method. The enhancement or suppression of analytes when using electrospray ionization is a common phenomenon when analyzing complex sample matrices, and there is no generally accepted approach to compensate for the results.28 It is common practice in our laboratory to fortify the specific samples that are suspected to contain residues in confirmation runs; results are corrected for recovery and matrix effects to the respective sample to ensure accurate quantification. Ruggedness. Results from the ruggedness study identified two critical control points for the method. The first was the difference in centrifuge speed used during the initial extraction. Our lab has several methods that use shared centrifuge programs. The difference between spinning samples at 5250 and 930 RCF is commonly tested during ruggedness as a result of the ease in which the programs could accidentally be selected during routine use. For this method, 55% of the compounds showed a statistically significant difference between the two tested conditions. This could be the result of less matrix components being removed from solution by gravitational force, resulting in more matrix suppression during ionization. Other methods that employ more extensive cleanup procedures (i.e., SPE) may not suffer as drastically from ion suppression effects. The other critical control point identified was the

conversion, or prevent chelation with metal ions contained in the sample matrix and extraction solvents. In addition to various ratios of ACN, MeOH, and water, several additives were tested with the nine compound classes of interest for this method. Several concentrations of formic and acetic acids were tested along with various buffers at several pH values. Sodium ethylenediaminetetracetic acid (EDTA) was evaluated as a modifier because it has been reported to help reduce losses of tetracyclines as a result of metal chelation.23 Although the addition of EDTA to the extract assisted slightly with tetracycline recovery, it also caused a significant loss of sensitivity for chloramphenicol, which, as a banned substance, has a performance requirement of 0.3 μg/kg for export to the European Union (EU).24 In the end, the best compromise proved to be an 80% ACN solution containing 1% FA. This selection was not optimal for the macrolide compound erythromycin because it is known to be unstable in acidic conditions and breakdown to erythromycin−H2O.25 Some methods also report improved results after implementing a quick cleanup step, such as defatting with hexane12,21,26 or using some form of solid-phase extraction (SPE),14,15,27 to remove interferences. One must be careful to evaluate these cleanup steps to ensure that compounds of interest are not removed along with the interferences. A common practice in many of the single-class methods used at the Dartmouth Laboratory of the CFIA is to defat with hexane prior to filtering extracts for MS/MS analysis.5,7−10 Defatting the final extract with 3 mL of hexane was evaluated as a quick cleanup step. This caused significant loss of the leuco (reduced) forms of the TPMDs and, hence, was not used in the final extraction procedure. Standard Stability Studies. Stock standard stability had been previously determined in-house during development and validation of each respective single-class method. The storage conditions and expiration times for each class are listed in Table 2. Previous in-house stability studies on TMPDs were unstable at concentrations below 1 mg/L in MeCN, and they were F

DOI: 10.1021/acs.jafc.7b05357 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry

Article

Table 3. SL, LOD, and LOI for the MCMR Method compound

SL (μg/kg)

LOD (μg/kg)

LOI (μg/kg)

compound

SL (μg/kg)

LOD (μg/kg)

LOI (μg/kg)

epiNAN NAN MT BOLD epiBOLD OMP TMP SMZL SIX SXL SMR STZ SMX SDZ SPY SGD SAA SNL SDX SDM SQX SCP SMP SMM SMZ MG LMG CV LCV BG

0.3 0.3 0.3 0.3 0.3 0.1 0.1 0.2 0.2 0.2 0.1 0.1 0.1 0.1 0.1 15 0.9 30 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.2 0.1 0.1 0.1 0.1

0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 2.0 0.1 2.6 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.03 0.03 0.03 0.03 0.03

0.3 0.4 0.1 0.1 0.3 0.1 0.1 0.2 0.1 0.2 0.1 0.1 0.1 0.1 0.1 2.8 0.1 2.7 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1

NLDX FLMQ OXO CIPRO DANO ENRO SARA MARBO NOR ORBI DIFLOX ERY TAP CAP FLR FLRA HEX DES DIEN OTC TC CTC DOXY RNZ MNZ−OH IPZ−OH MNZ IPZ HMMNI DMZ

0.2 0.1 0.1 0.3 0.6 0.3 0.3 0.4 0.6 0.2 0.4 1.3 0.6 0.1 0.2 30 0.9 0.5 0.8 2.7 2.0 5.8 5.7 1.0 1.0 0.6 0.6 0.5 1.6 0.6

0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.4 0.3 0.1 0.1 11 0.4 0.3 0.5 0.5 0.3 0.7 1.3 0.2 0.1 0.2 0.1 0.1 0.6 0.4

0.1 0.1 0.1 0.2 0.2 0.4 0.2 0.2 0.1 0.1 0.1 0.4 0.3 0.2 0.2 29 1.0 0.8 0.8 11 0.4 0.9 2.1 1.1 0.3 0.3 0.1 0.1 1.2 0.2

percentage of ACN in the extraction solution. A difference of 2% ACN resulted in a statistically significant difference in the extractability of the compound for 20% of the residues tested. There were other statistically significant differences observed for some of the other factors, but in all cases,