1784 ZhengCai Liu1 Fang Yang1 Minna Yao2 YongHui Lin1 ZhiJiao Su1 1 Fujian
Entry-Exit Inspection & Quarantine Bureau, Fujian Provincial Key Laboratory of Inspection and Quarantine Technology Research, Fuzhou, China 2 College of Food Science, Fujian Agriculture and Forestry University, Fuzhou, China Received January 7, 2015 Revised February 14, 2015 Accepted February 15, 2015
J. Sep. Sci. 2015, 38, 1784–1793
Simultaneous determination of antiviral drugs in chicken tissues by ultra high performance liquid chromatography with tandem mass spectrometry An ultra high performance liquid chromatography with tandem mass spectrometry method was established for the rapid and simultaneous analysis of seven antiviral drugs, amantadine, rimantadine, memantine, moroxydine, imiquimod, oseltamivir, and acyclovir, in chicken liver, muscle, and egg. Homogenized samples were extracted with trichloroacetic acid and acetonitrile solutions and then purified by cation-exchange solid-phase extraction. The target drugs were analyzed by liquid chromatography with a UPLC BEH Amide column (2.1 mm × 100 mm, 1.7 m) coupled with a tandem mass spectrometer operating in the positive multiple-reaction mode. A perfectly linear relationship was obtained within the concentration ranges of 0.5–20 g/L for acyclovir and 0.1–10 g/L for the other six antiviral drugs. The average recoveries of the seven antiviral drugs using four addition levels in chicken liver, muscle, and eggs were 82.67–90.10, 82.30–92.27, and 81.98–93.77%, respectively, and the acceptable coefficients of variation were 5.18–9.88, 4.84–11.2, and 42.8–9.95%, respectively. The detection limits and detection capabilities of the analysis method for the seven antiviral drugs were in the ranges of 0.04–0.64 and 0.11–0.78 g/kg, respectively. Additionally, an inter-laboratory study among five laboratories further validated the method. Keywords: Antiviral drugs / Chicken tissues / Multi-residue detection / Tandem mass spectrometry / Ultra high performance liquid chromatography DOI 10.1002/jssc.201401461
1 Introduction Antiviral drugs are active against influenza type A viruses, including avian influenza (H5N1) . The consumption of antiviral drugs has increased to treat viral diseases due to the outbreak of avian influenza and swine influenza viruses, thus enhancing the drug resistance of the target viruses [2–4] and aggravating antiviral drug residues . As a matter of precaution, considering the threat of a human influenza pandemic and the scarcity of influenza drugs, WHO, FAO, and OIE urge their member states not to use antiviral drugs in animals to preserve the efficacy of these drugs for the treatment of influenza infections in humans. To prevent the excessive use of antiviral drugs, the Food and Drug Administration prohibited the inconsistent use of the following two types of antiviral drugs in chickens, turkeys, and ducks: antiinfluenza adamantanes (amantadine and rimantadine) and neuraminidase inhibitors (oseltamivir and zanamivir). Antiviral drugs, including amantadine, rimantadine, moroxydine, and acyclovir (Fig. 1), are administered to poultry and other livestock in China. Nevertheless, these compounds are Correspondence: Dr Minna Yao, College of Food Science, Fujian Agriculture and Forestry University, Fuzhou 350002, China E-mail: [email protected]
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
still illegally used to combat the avian influenza virus in chickens. It was reported that several Shandong chicken farms fed their chickens antibiotics, hormones, and antiviral drugs every day to reduce their death rate and shorten their growth period. Therefore, to control their illegal use and ensure the safety of consumers, it is necessary to develop a sensitive high-throughput analytical method for the detection of antiviral drug residues in chicken tissues. In the past two decades, various analytical methods have been reported for the determination of the residues of these compounds in animal foods and body fluids. Immunoassays  have been used as a screening tool due to their high sensitivity and throughput, but the stability of immunoassays is poor. A GC–MS method  can provide definite qualitative and quantitative results but requires time-consuming derivatization steps, which limit its application scope. LC [8, 9] and CE  allow the direct determination of polar drugs, and MS/MS provides a high detectability and sensitivity. Recently, the application of LC–MS/MS in the detection of antiviral drug residues in poultry muscle has been reported [11–14]. However, the analysis of antiviral drug residues in chicken tissues was seldom reported. In this study, based on previous studies [12–14], we increased the number of residues determined, optimized the sample pre-treatment procedure, and developed a UHPLC–MS/MS method for the determination of the following seven antiviral drug residues in chicken liver, www.jss-journal.com
J. Sep. Sci. 2015, 38, 1784–1793
Figure 1. Chemical structures and pKa values of the anti-viral compounds.
muscle, and egg: amantadine, rimantadine, memantine, moroxydine, imiquimod, acyclovir, and oseltamivir. The method reported herein is characterized by high sensitivity and selectivity and can provide the basis for monitoring the use and residues of these antiviral drugs. This method consists of a one-step extraction and a HR-XC SPE cleanup. The final extract was separated by UHPLC within 10 min and was then injected into an ESI mass spectrometer for the determination of seven antiviral drugs in a single run. Validation parameters including the decision limit (CC␣), detection capability (CC␤), selectivity, linearity, precision, accuracy, and robustness were established. To enhance the quality and accreditation of the method, the validation procedure was complied with acceptable guidelines [15 and 16] and statistical techniques were adopted. Furthermore, an inter-laboratory study was performed in five independent labs to validate the authenticity and accuracy of the procedure.
Individual stock standards (100 g/mL) were prepared by dissolving 10 mg (corrected for the salt and its purity) of each compound in 100 mL of the corresponding solvent. Methanol was used as the solvent for amantadine, rimantadine, memantine, imiquimod, moroxydine, and oseltamivir, while DMSO was used as the solvent for acyclovir. Intermediate stock standards (1.0 mg/L) were prepared by adding 100 L of the stock standard solutions into 10 mL of methanol. Working mixture standards (0, 0.5, 1.0, 2.0 5.0, 10.0, and 20.0 ng/mL) and an internal standard solution (10 ng/mL) were obtained by combining the intermediate stock standards and then diluting them with solvents. All of the standard solutions were stored at –18⬚C and new working solutions were prepared weekly. The SPE strong mixed-mode polymer-based cationexchange (Chromabond HR-XC) columns (60 mg, 3 mL) were obtained from the Beijing Zhenxiang Industry and Trading. Blank samples and test samples were obtained from the local market (Fuzhou, China).
2 Materials and methods 2.1 Instrumentation
2.3 Sample preparation
An ACQUITY UPLC system (Waters, Milford, MA, USA) with a Triple Quadrupole API 5500 (AB SCIEX) was used.
The samples were homogenized in a homogenizer at 3000 rpm firstly. Portions (5 g) of well-homogenized samples were then weighed in 50 mL centrifuge tubes, and 100 L of the internal standard solution was added. Then, 15 mL of a trichloroacetic acid solution (20 g/L)/acetonitrile (9:1, v/v) was added to the sample. The mixture was vortexed for 2 min, sonicated for 15 min, and then centrifuged at 15 000 rpm for 5 min at 4⬚C. The supernatant was transferred into a 25 mL flask, and then the extraction was repeated with another 10 mL of extraction solution. The extraction solution was collected and diluted to 25 mL with acetonitrile. Next, 5 mL of the supernatant was loaded onto an HR-XC cartridge (60 mg, 3 mL), which had been preconditioned sequentially with 3 mL of methanol, 3 mL of water, and 3 mL of 0.2% formic acid. The cartridge was then washed with 5 mL of a formic acid solution (0.2%) and 3 mL of a formic acid/acetonitrile solution (1:99, v/v). Subsequent elution occurred with 4.0 mL of a methanol/ammonium acetate solution (5 mol/L)/ammonia (95:2.5:2.5, v/v/v). The eluate was dried under a gentle nitrogen stream at 50⬚C and then dissolved in 1.0 mL of acetonitrile/water/methanol (70:20:10,
2.2 Reagents and materials All of the chemicals used were of analytical grade. Trichloroacetic acid (TCA), acetic acid, and ammonium acetate were obtained from Sinoreagent (Shanghai, China). Formic acid, acetonitrile, methanol, DMSO, and ammonia (25% solution in water) were obtained from Merck (Darmstadt, Germany). Standard compounds of amantadine (CAS No. 768-94-5), rimantadine (CAS No. 1501-84-4), memantine (CAS No. 41100-52-1), acyclovir (CAS No. 59277-89-3), imiquimod (CAS No. 99011-02-6), moroxydine (CAS No. 3731-59-7), and oseltamivir (CAS No. 196618-13-0) as well as isotope-labeled internal standards (amantadine _D6 , memantine _D6 , acyclovir_D4 with the purity > 99%) were obtained from Sigma (Ronkonkoma, USA). Milli-Q water was prepared by using a Milli-Q system at a resistivity of at least 18.2 M⍀/cm (Millipore, Billerica, MA, USA). C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 1784–1793
Z. Liu et al.
Table 1. LC–MS/MS acquisition parameters for the seven anti-viral drugs
Parent ion (m/z)
Daughter ion (m/z)
Declustering potential (v)
Entrance potential (v)
Collision energy (v)
Collision cell exit potential (v)
Acyclovir _D4 Oseltamivir
135.1a) 93.1 141.1 163.3 121.2a) 163.1 107.1a) 169.3 130.1 113.1a) 185.3a) 168.2 152.1a) 130.0 152.1 166.0a) 225.1
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
8 8 10 9 9 10 10 10 10 10 10 10 10 10 10 10 10
22 39 26 22 37 22 34 20 24 27 36 50 24 27 17 30 20
12 12 12 12 12 12 12 12 12 12 15 15 12 12 12 15 15
a) indicates the quantified ion.
v/v/v). The solution was filtered with a cellulose acetate membrane filter (0.22 m pore size, Millipore, Billerica, MA, USA) before UHPLC–MS/MS analysis. This solution shall be analyzed within 24 h after preparation.
2.4 UPLC–MS/MS conditions The LC separation used an Acquity UPLC BEH Amide column (100 mm × 2.1 mm, 1.7 m) at 25⬚C; a linear gradient elution of 0.2% formic acid in a 5 mM ammonium acetate solution (A) and acetonitrile (B) was used as the mobile phase. The gradient elution was linearly programmed as follows: 95% B and 5% A; 4 min, 5–30% A; 1 min, 30–60% A; and 4 min, 60–5% A. The flow rate was 0.4 mL/min and the injection volume was 2 L. An ACQUITY UPLC system coupled to an API 5500 tandem mass spectrometer (AB Sciex) with an ESI interface was utilized in the analysis. The entire column effluent was directed into the mass spectrometer interface. Positive ions were acquired in the multiple reaction monitoring (MRM) mode. The optimal conditions for the sample analysis were as follows: electrospray voltage (IS) of 5500 V; curtain gas (CUR) of 25.00 psi (nitrogen); ion source temperature of 500⬚C; and gas settings of 7 mL/min collision gas, 25 mL/min curtain gas, ion source gas 1:55, and ion source gas 2:55. In the operating parameters shown in Table 1, the MRM transition parameters for the quantification are presented in bold font.
2.5 Method validation In the paper, the specific retention time for every compound is expressed as the capacity factor (k ). The capacity factor is C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
calculated as: k = (TR –T0 )/T0 , where T0 is the retention time of the undelayed compound and TR is the retention time of the compound. The matrix effect  was evaluated according to the following equation: E = (Sb –Sm )/Sb * 100%, where E represents the matrix effect; Sm represents the slope of the matrixmatched calibration curve; Sb represents the slope of the blank solvent calibration curve. The validation process was carried out according to the criteria of the Commission Decision 2002/657/EC (Commission Decision 2002/657/EC 2002) for banned compounds . The validation included the determination of the stability of mixed standard solutions, specificity, linearity, decision limit (CC␣), detection capability (CC␤), accuracy, and precision (repeatability and reproducibility). The decision limit (CC␣) and the detection capability (CC␤) of the method were calculated according to the European Commission Decision 2002/657/EC. For non-defined MRLs substances, CC␣ = 3S/N20representative blank samples ; and CC␤ = CC␣ + 1.64 SD20 representative samples spiked at the CC␣ level.
3 Results and discussion 3.1 Optimization of the UHPLC–MS/MS conditions The fixed conditions of MS/MS, such as the ion spray voltage and vaporizer temperature, were optimized using the FIA mode through a syringe infusion pump. According to the structures of the analytes, the positive ion scanning mode provided better results compared with the negative ion scanning mode. The precursor ion [M+H] + and two product ions were subsequently selected. The analysis conditions for each www.jss-journal.com
J. Sep. Sci. 2015, 38, 1784–1793
drug were optimized through multiple reaction monitoring scans (Table 1).
3.2 Sample preparation Considering the weakly basic nature and polarity of the target compounds, methanol, acetonitrile, and ethyl acetate were first used as extractants in the optimization. Acetonitrile allowed higher average recoveries (64–86%) than methanol and ethyl acetate for the compounds amantadine, rimantadine, memantine, and oseltamivir, but the extraction recoveries for imiquimod, moroxydine, and acyclovir were relatively low. Because all seven drugs contain amino groups, belong to the organic alkali family, and have pKa values (Fig. 1) over 7.0, their solubilities in water were higher than in acidic solutions and their solubilities in organic solvents were higher than in alkaline solutions. In the traditional extraction procedure for these drugs, these drugs were firstly adjusted to a pH > 7, extracted with methanol or water, and then extracted with organic solvents. Moreover, the matrix to be analyzed was generally blood, plasma, or other matrices and required a simple purification procedure. The purification process is relatively simple because the common substrates to be determined include blood, blood plasma, and other common liquid substrates. Considering that the analyzed matrix in this study was complex and contained a high fat content, a trichloroacetic acid and acetonitrile solution was used as the extraction solution, which was conducive to the next steps of protein precipitation and SPE purification. The highest extraction recoveries of the seven drugs were obtained with a trichloroacetic acid solution (20 g/L)/acetonitrile (9:1, v/v). In the analysis of drug residues in food matrices, the purification step is extremely important because of the interference from the matrix. For example, fat could affect the behavior of the chromatograms and signal intensities. In this study, SPE was used to purify and concentrate the sample. According to the weakly basic natures and acidities of the target compounds, three cation ion exchange cartridges (Oasis MCX, HR-XC, and LCX) were screened based on their recoveries. As shown in Fig. 2, the recoveries obtained with MCX, HR-XC and LCX were between 80 and 100% for all the drugs except acyclovir. The recovery of acyclovir obtained with HRXC was between 80 and 90%, but the recoveries of acyclovir obtained with MCX and LCX were between 40 and 70%. Moreover, the elution with 5% ammonia methanol yielded a low recovery of moroxydine (10%) because moroxydine was a strong alkaline compound (pKa = 13.6) and could be firmly bound with the packing material of the HR-XC column. Therefore, salts, such as ammonium acetate, should be added into the elution solution. A better purification of moroxydine was achieved with 4.0 mL of ammonium acetate (5 mol/L, pH = 7.0)/ammonia/methanol (volume ratio of 2.5:2.5:95).
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Mean recoveries obtained in the solvent extraction of the antiviral drugs in chicken samples spiked at the concentrations of 5 g/kg (n = 6).
3.3 Hydrophilic effect on the chromatographic column separation of antiviral drugs R Two HILIC columns (NUCLEOSHELL HILIC and Amide) were used to separate the target compounds. The separation effects were shown in Fig. 3. Good retentions of the seven drugs were achieved on both the Amide and NUCLEOSHELL columns. However, the response value for acyclovir by NUCLEOSHELL was better than that achieved on Amide. One possible explanation for these results is that the silica gel packing of the Amide column is bonded with the strong polar amide group as there is a strong hydrophilic hydroxyl on the acyclovir structure. Due to the high proportion of the water phase of the ion source at the peak, the response signal of the compound was inhibited by the salt flow phase, and the response values were lower than before. The Amide column bonded with neutral amide groups was selected as the separation column. Ammonium acetate solutions with different pH values (2.5, 3.0, 3.5, 6.0, 6.5, and 6.8) prepared with formic acid and ammonia were added into the mobile phase as the brine solution. The influence of the pH of the mobile phase on the chromatographic behavior of the antiviral drugs was analyzed. The analysis results showed that pH variations basically had the same influence on the retention time of all the seven drugs. A better retention effect would theoretically be obtained when the pH of the mobile phase was lower than the pKa of the drugs. When the pH < pKa, the change in the capacity factor (K) was stable. The change in the capacity factor (K) obviously declined when the pH was close to the pKa values of the drugs. An ion exchange mechanism existed between the seven types of drugs and the stationary phase of the chromatographic column. When the pH value increased, the drugs were converted from ions into molecules with weak adsorption capacities and their retentions were reduced. The capacity factors (K) of the seven drugs were detected when different concentrations (2, 5, and 10 mmol/L) of ammonium
Z. Liu et al.
J. Sep. Sci. 2015, 38, 1784–1793
Figure 4. Matrix effects for the analysis of seven antiviral drugs in different hydrophilic interaction liquids with various injection volumes (1. Amantadine; 2. Rimantadine; 3. Memantine; 4. Moroxydine; 5. Imiquimod; 6. Acyclovir; 7. Oseltamivir).
R HILIC Figure 3. Total ion chromatograms of (A) NUCLEOSHELL and (B) Amide for the seven antiviral compound standards in HILIC–MS/MS (1. Acyclovir; 2. Imiquimod; 3. Rimantadine; 4. Memantine; 5. Amantadine; 6. Oseltamivir; 7. Moroxydine).
acetate solutions were used as the aqueous solution in the mobile phase at the same pH. The K values of the seven drugs decreased when the concentration of the buffer solution increased because the ion intensity increased. Therefore, the ion exchange capacity was correspondingly enhanced.
3.4 Matrix effect It is necessary to estimate the matrix effects (MEs) during the development of a LC–MS method. In this paper, the MEs were evaluated according to the slopes of the standard working curve and the matrix curve. In this experiment, the following two aspects of the matrix effects were investigated: the choice of the chromatographic column and the sample volume. As shown in Fig. 4, the inhibition effects of C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
different substrates were shown in different columns in the same matrix curve, and the MEs increased with an increase of the injection volume. The ME values of seven drugs obtained using NUCLEOSHELL were higher than those obtained using Amide. Serious MS substrate inhibition was observed for acyclovir and moroxydine with high ME values. The mechanism of the matrix effects is not clear yet. It is generally believed that the competitive ionization between the extracted interfering matrix components and the analytes is responsible for matrix effects. Further research is required to confirm the specific reason for the matrix effects. Isotope-labeled internal standard calibration is suitable for trace drug residue analysis in a complex matrix. However, it is not usually possible to obtain isotope-labeled standards for all of the target compounds in multi-component analysis. In this study, the following isotope-labeled internal standards were used: amantadine_D6, memantine_D6, and acyclovir_D4. According to their chromatographic retention time, ionization responses in ESI MS, and structural similarities with the studied compounds, these three compounds were chosen as internal standards for plotting calibration curves in this study. Amantadine_D6 was used for www.jss-journal.com
J. Sep. Sci. 2015, 38, 1784–1793 Table 2. CC␣ and CC␤ values of seven antiviral drugs in spiked chicken muscle, liver, and eggs
Egg Liver Muscle Egg Liver Muscle Egg Liver Muscle Egg Liver Muscle Egg Liver Muscle Egg Liver Muscle Egg Liver Muscle
0.12 0.15 0.11 0.23 0.25 0.22 0.09 0.12 0.09 0.21 0.34 0.19 0.04 0.06 0.04 0.52 0.64 0.50 0.05 0.08 0.05
0.25 0.27 0.21 0.32 0.37 0.35 0.17 0.24 0.18 0.35 0.48 0.29 0.12 0.16 0.11 0.67 0.78 0.63 0.15 0.21 0.18
the quantitation of amantadine, oseltamivir, and moroxydine; memantine_D6 was used for the quantitation of rimantadine, memantine, and imiquimod; acyclovir_D4 was used for the quantitation of acyclovir.
3.5 The influence of the dissolution solution The constant-volume solvent selection tests indicated that different dilution solvents led to significant differences in the chromatographic peak shapes and the mass spectrometric responses. When the initial proportion of the organic solvent in the mobile phase was high and the proportion of the organic phase in the constant-volume solvent was