Arch. Pharm. Res. DOI 10.1007/s12272-014-0508-0
RESEARCH ARTICLE
Quantification of 4-methylimidazole in carbonated beverages by ultra-performance liquid chromatography-tandem mass spectrometry Hyo-Hyun Cho • Kyung-Oh Shin • Cho-Hee Seo • Shin-Hee Lee • Hwan-Soo Yoo • Hye-Ran Yoon • Jeong-Woo Kim • Yong-Moon Lee
Received: 11 August 2014 / Accepted: 20 October 2014 Ó The Pharmaceutical Society of Korea 2014
Abstract 2- and 4-methylimidazoles (2-MI and 4-MI) are undesired byproducts produced during the manufacture of caramel color used to darken food products such as carbonated beverages. The Office of Environmental Health Hazard Assessment in California listed 4-MI as carcinogen in January 2011 with a proposed no significant risk level at 29 lg per person per day. Thus, a quantitative analytical measurement for 2-MI and 4-MI is desired for reliable risk assessments for exposure. An ultra-performance liquid chromatography (UPLC) coupled tandem mass spectrometric (MS/MS) method was developed for the quantification of 4-MI in beverage samples. Chromatographic separation of 2-MI and 4-MI were achieved by using a PFP reversed-phase column and a stepwise gradient of methanol and distilled water containing 0.1 % formic acid. Identification and quantification of 2-MI and 4-MI were performed using electrospray ionization-tandem mass monitoring the precursor to product ion transitions for 2-MI at m/z 83.1 ? 42.2 and 4-MI at m/z 83.1 ? 56.1 with melamine at m/z 127.1 ? 85.1 as the internal standard. The performance of the method was evaluated against validation parameters such as specificity, carryover, linearity and calibration, correlation of determination (r2), H.-H. Cho K.-O. Shin C.-H. Seo S.-H. Lee H.-S. Yoo Y.-M. Lee (&) College of Pharmacy, Chungbuk National University, Cheongju 361-763, Korea e-mail: [email protected] H.-R. Yoon College of Pharmacy, Duksung Women’s University, Seoul 132-714, Korea J.-W. Kim Department of Life Science and Technology, PaiChai University, Daejeon 302-735, Korea
detection limit, precision, accuracy, and recovery. Calibration curves at 10–400 ng/mL were constructed by plotting concentration versus peak-area ratio (analyte/ internal standard) and fitting the data with a weighted 1/x. The accuracy of the assay ranged from 93.58 to 110.53 % for all analytes. Intra-assay precision for 2-MI and 4-MI were below 7.28 (relative standard deviation/RSD %) at QC samples. Here we present a new and improved method using UPLC-MS/MS to significantly simplify sample preparation and decrease chromatographic run time. This method allows accurate and reproducible quantification of 4-MI in carbonated beverages as low as sub ng/mL (ppb) levels. Keywords 2-Methylimidazole 4-Methylimidazole UPLC-tandem mass spectrometry Abbreviations 2-MI 2-Methylimidazole 4-MI 4-Methylimidazole UPLC Ultra-performance liquid chromatography PFP Pentafluorophenyl MS/MS Tandem mass spectrometry QC Quality control MeOH Methanol
Introduction Methylimidazoles (2-MI and 4-MI) were found as undesired byproducts in the browning of certain foods such as carbonated beverages and soy sources through the Maillard reaction between carbohydrates and amino-containing
123
H.-H. Cho et al. Fig. 1 Typical chromatograms of 4-MI and 2-MI. UPLC separated with gradient elution for 4-MI (1.04 min) (a) and 2-MI (0.94 min) (b)
compounds. Recently, The Office of Environment Health Hazard Assessment (OEHHA) in California listed 4-MI as carcinogen in January 2011 with a proposed no significant risk level (NSRL) at 29 lg per person per day. In January 2014, a consumer review of various beverages in the United States reported that Pepsi ONE and Malta Goya contain the chemical in excess of 29 lg per can or bottle. Although the correlation of adverse health effects of exposure to methylimidazoles is controversial, sensitive or excessive consuming populations should be precautious. Several procedures have been reported for the determination of imidazoles compounds in caramel-colored foods. Most procedures are based on HPLC-mass spectrometry (Hengel and Shibamoto 2013; Kim et al. 2013; Schlee et al. 2012; Yamaguchi and Masuda 2011), gas chromatography–mass spectrometry (GC–MS)(Fuschs and Sundell 1975; Li et al. 2013a), paper spray MS (Li et al. 2013b), and electrophoresis (Petruci et al. 2013). The International Agency for Research on Cancer (IARC) noted a general lack of occurrence data of 2-MI and 4-MI in beverages. Thus, the objective of this study was to develop an ultraperformance liquid chromatography-tandem mass spectrometry (UPLC–MS/MS) method for the quantitation of 4-MI in various carbonated beverages without sample preparation. This method was applied to overview the imidazole contents in carbonated beverages from the Korean markets so that it could provide possible risk assessments.
123
Materials and methods Chemical and reagents Chemical standards, 2-MI, 4-MI and melamine (internal standard) shown in Fig. 1 were purchased from SigmaAldrich (St. Louis, MO, USA). For mobile phase preparation, methanol and formic acid were obtained from Sigma-Aldrich (St. Louis, MO, USA). Deionized (DI) water was obtained from a Millipore water station. Standard stock solutions were prepared by dissolving each chemical to DI water at 1 mg/mL. Working standard solutions were prepared by mixing individual standard stock solutions in DI water at 10 lg/mL and then diluted subsequently to 1 lg/mL and 100 ng/mL. Internal standard (IS) stock solution was diluted in DI water to 1 lg/mL to prepare calibration standards or spike unknown samples. Sample preparation Bottled carbonated beverage samples were purchased from local grocery stores. All samples were stored at room temperature until opened for analysis. A 500 lL aliquot of each sample was transferred into Eppendorf tube and degassed for 10 min in a sonication bath at room temperature. After the tube was centrifuged at 12,000 rpm for 10 min at 4 °C, a 100 lL aliquot of each degassed sample was diluted with 800 lL of DI water and vortex mixed
Quantification of 4-methylimidazole in carbonated beverages
with 100 lL of a 100 ng/mL IS stock solution. Each prepared sample (5 lL) was subjected to UPLC-MS/MS analysis. UPLC-MS/MS conditions The UPLC-MS/MS analysis was performed using an ACQUIRTY series UPLC system coupled in-line to a Xevo TQ MS system (Waters, MA, USA). Chromatographic separation was performed using a KINETEX PFP column (4.6 9 50 mm, 2.6 lm) (Phenomenex, CA) at 35 °C. Mobile phase A used 0.1 % formic acid in deionized water. Mobile phase B used 0.1 % formic acid in MeOH. Gradient elution was performed using mobile phases A and B at a flow rate of 0.6 mL/min with the following parameters: initial conditions, A/B = 95:5 maintained for 0.2 min, linear gradient of A/B = 20:80 for 1.3 min which was maintained for 0.5 min, then the gradient was switched back to initial conditions. Column was re-equilibrated for 2 min. The total run time was 4 min per sample. The volume of injected sample was 5 lL. For optimization, a mixture of 2-MI and 4-MI standard was infused directly into the mass spectrometry. All source parameters and ionization conditions were adjusted to improve the sensitivity of the assay. The configurations for mass spectrometry were: capillary voltage at 3.5 kV; source temperature at 150 °C; desolvation temperature at 650 °C; desolvation gas flow rate at 1000 L/hr; and cone gas flow rate at 150 L/ hr. Analyses were performed using electrospray ionization in the positive-ion mode with multiple reaction monitoring to select both precursor and product ions specific to each analyte simultaneously from a single injection. The MS/ MS transitions (m/z) were at 83.1 ? 42.2 for 2-MI, 83.1 ? 56.1 for 4-MI, 127.1 ? 85.1 for melamine. The optimized collision energies for 2-MI, 4-MI, and melamine were 14, 14, and 18 V, respectively. Method validation To validate the UPLC-MS/MS method, several parameters such as the limit of detection (LOD), the limit of quantification (LOQ), linearity, recovery, stability and repeatability were evaluated. The linearity was assessed with calibration standards at 10, 40, 80, 200, and 400 ng/mL. To validate the method, standard solutions of 2-MI and 4-MI standards at 10, 40, 80, and 200 ng/mL were analyzed five times daily. The reproducibility of the proposed method was expressed by relative standard deviation percent (% RSD). Analysis of carbonated beverage containing 2-MI and 4-MI were performed in triplicates. To evaluate the intra-day precision and accuracy, we extracted quality control (QC) samples at LOQ (10 ng/mL), LQ (40 ng/mL),
MQ (80 ng/mL), and HQ (199 ng/mL). The LOQ was calculated as signal to noise ratio (S/N ratio). The matrix effect of 4-MI was calculated as % in relation to spiked amount when standard solution (0.5 lg/L) was added to a blank matrix of sample A and deionized water.
Results Method development The chemical structures of 2-MI, 4-MI, and melamine as an internal standard were shown in Table 1. 2-MI and 4-MI are isomers containing alkaline ‘‘pyridine-like’’ nitrogen and weak acidic ‘‘pyrrole-like’’ amino nitrogen. They have the same molecular weight. From a previous report, methanol was the first choice for organic solvent to escape the interference from acetonitrile (Wang and Schnute 2012). Therefore, methanol was selected as a leading mobile phase. The chromatographic separation using column depends on the character of attached functional group on the surface of the column. In order to separate high polar compounds 2-MI and 4-MI, we chose KinetexÒ PFP column bearing pentafluorphenyl group on the surface of the column with high selectivity deference to polar compounds. In these separation conditions, a standard mixture of 4-MI (40 ng/mL) and 2-MI (20 ng/mL) was completely separated in 1.2 min (Fig. 1). However, this separation condition was not recommended for the quantitation of analyte due to the occurrence of low precision and low accuracy originated from the short retention time. In practice, isocratic elution of mobile phase (A:B = 20:80, v/v) mixture at a flow rate 0.6 ml/min resulted in co-elution of 4-MI and 2-MI at 1.65 min. The observed protonated molecular ions [M?H]? were used as precursor ions. Multiple reactions monitoring (MRM) transitions were used to determine the product ions by infusing standard 4-MI and 2-MI. In this condition, the precursor ion (m/ z 83.1) of 4-MI and 2-MI was scanned in the product scan conditions at cone voltage of 35 V and collision energy of 15 V to produce fragment ions. The fragmented ions of 4-MI and 2-MI had identical m/z value. However, their intensity pattern was clearly different, with the main fragment ions of 4-MI and 2-MI at m/z 56 and m/z 42, respectively (Fig. 2). The chromatographic separation of 4-MI and 2-MI in isocratic elution was not completed with overlapped peaks easily differentiated by main fragment ion intensity. The calculated fragment ion ratio of 10.7 % for 4-MI and 7.6 % for 2-MI represented maximum permitted tolerance values of 4-MI at ±30 % and 2-MI at ±50 % based on the European Commission Council Directive 96/23/EC/SANCO/1805/2000.
123
H.-H. Cho et al. Table 1 The chemical structures, formula and exact mass for analytes
4-MI
2-MI
Melamine(I.S.)
H N
H N
Structure
N
NH2 N
N H2N
Chemical formula
C4H6N2
C4H6N2
C3H6N6
Exact mass
82.05
82.05
139.04
N N
NH2
Fig. 2 MS/MS spectrum of 4-MI (a) and 2-MI (b) using direct infusion method
Method validation The method validation was evaluated for selectivity, recovery, precision, sensitivity, linearity, and accuracy. Selectivity was confirmed by the absence of quantifiable peaks with the assay using blank samples and positive peak detection at specific retention time with 4-MI and melamine (I.S.) in MRM mode at m/z 83.1 ? m/z 56.1 for 4-MI and m/z 127.1 ? m/z 85.1 for melamine. There was no quantifiable peak in blank samples, indicating highly selective monitoring of UPLC-MS/MS conditions. Carryover was not observed when the volume of injector needle washing solution was increased to 600–1,800 lL. Recovery was evaluated by spiking 500 ppb of 4-MI in deionized water (DW) and in blank matrix, respectively. The matrix
123
effect was calculated by the following formula: matrix effect = (quantified amount in matrix)/(quantified amount in DW) 9 100 %. Results are summarized in Table 2. The average recovery was 94.7 ± 1.5 % RSD, indicating high reproducibility. Precision was addressed by injecting 4-MI at four concentrations (11, 40, 80, and 199 ng/mL, n = 3). The relative retention area (RRA) and relative retention time (RRT) using the ratio of 4-MI/melamine peaks were calculated from the UPLC-MS/MS MRM chromatogram. As shown in Table 3, excellent precision was observed overall for these experimental conditions, although 4-MI at low concentration of 11 ng/mL with relatively high variation of % RSD (\8 %) was observed. The LOQ [10 was observed. Considering the instrumental detection limit in practical usage, LOQ was determined at 2 ng/mL (S/N 13).
Quantification of 4-methylimidazole in carbonated beverages Table 2 Matrix effect of spiked 500 ng/mL of 4-MI (n = 3) MRM Transition
83.1 ? 56.1
#
500 ng/ mL in DW Peak area
500 ng/mL in matrix Peak area
Matrix effect (%)a
1
113,895
107,445
94.3
2
105,880
102,004
96.3
3
123,716
115,664
93.5
Ave. (%)
94.70
RSD (%)
Table 5 Quantitation of 4-MI in beverages purchased from Korean Markets
1.52
# Indicates sample number a
Matrix effect (%) = Quantified amount in matrix/Quantified amount in DW 9 100
Table 3 Precision tests for relative retention area (RRA) and relative retention time (RRT) with 4-MI and melamine Spike amount (ng/mL)
4-MI/IS
RSD %
RRT
RSD %
11
0.05
7.28
1.14
0.35
40
0.053
4.16
1.14
0.24
80
0.102
2.65
1.13
0.24
199
0.285
0.75
1.13
0.17
RSD(%) = standard concentration 9 100
deviation
of
the
concentration/mean
4-MI (ng/mL)
Cola 1
374.08 ± 15.55
Cola 2
1,050.27 ± 38.30
Cola 3
912.50 ± 40.51
Cola 4
1,751.6 ± 65.59
Apricot drink
166.35 ± 4.83
Beer 1
17.85 ± 1.37
Beer 2
15.84 ± 0.38
Beer 3
12.95 ± 0.51
Beer 4
14.38 ± 0.18
Beer 5
19.25 ± 0.31
The calibration curve was linear between 10 and 400 ng/ mL of 4-MI standard solution (r2 = 0.9998, Y = 0.1467X-1.0099) (Fig. 3). Accuracy test with four different concentrations of 4-MI was addressed by % accuracy calculated by (observed amount from the 4-MI/IS peak area ratio)/(specified amount) 9 100 % (Table 4). Analysis of 4-MI in food samples We analyzed food beverages of 10 different brands available in Korea markets. The concentration of 2-MI was not detectable in all tested beverages. However, variable high amount of 4-MI was detected in all dark-colored cola samples, with a range of 166.35–1,751.60 ng/mL. The 4-MI concentrations in beer samples were relatively low (\20 ng/mL) with small S.D. values (Table 5).
60 y = 0.1467x - 1.0099, r2 = 0.9998
50
Analyte / IS ratio
Sample
40 30 20
Discussion
10
The main purpose of this study is to establish a conventional method which is able to monitor the non-intentional impurity of a potential carcinogen, 4-MI during the manufacturing process (Murray 2011). To develop an optimized method, the choice of mobile phase and column was essential to the identification and quantitation of each analyte. Previously, Wang and Schnute (2012) have chosen methanol and acetonitrile as optimal organic solvents for the analysis of 2-MI and 4-MI. However, strong interference was reported by acetonitrile cluster ion [2M?H]? which had identical mass-to-charge ratio (m/z) at 83 corresponding to the m/z of 2-MI or 4-MI (Wang and Schnute 2012). Therefore, methanol was the first choice for organic solvent to escape the interference from acetonitrile. The validation parameters including selectivity, recovery, precision, sensitivity, linearity, and accuracy were predominantly evaluated compared to developed LC-MS/
0 0
50
100
150
200
250
300
350
400
4-MI Conc. (ng/mL)
Fig. 3 Calibration curve of 4-MI by UPLC-MS/MS method
Table 4 Accuracy (%) test for four concentrations of 4-MI Spike amount (ng/ mL)
Calc. value (ng/ mL)
Accuracy (%)
% Dev
11
10.45 ± 0.01
95.00 ± 0.13
40
43.48 ± 1.03
108.7 ± 2.58
-5.00 8.70
80
76.55 ± 2.39
95.69 ± 3.00
-4.31
199
200.95 ± 0.91
100.98 ± 0.45
0.98
123
H.-H. Cho et al.
MS methods (Kim et al. 2013; Li et al. 2013a; Schlee et al. 2012). In conclusion, a UPLC-MS/MS method was developed. It was demonstrated to be successful for the quantification of 4-MI in carbonated beverages. Unlike conventional methods that require a sample pretreatment step with solid phase extraction (SPE), we simplified the sample pretreatment using solvent dilution without any difficulties. Among the dark-colored samples tested, 4-MI was observed in all of them in the range of 224–693 ng/ mL(ppb). All samples appeared to be 2-MI free. A lightcolored, lime-flavored carbonated beverage with no quantifiable target analyte was assayed as comparison. Our results showed that the UPLC-MS/MS method could simplify sample preparation, shorten analysis time, and allow accurate and reproducible quantification of 4-MI in carbonated beverages at as low as sub ng/mL (ppb) levels. Acknowledgments This study was partly supported by the Research Grant of the Chungbuk National University (2011) and a NRF Grant (KRF-2010-0025271) funded by the Korean Government. Conflict of interest Authors certify that there is no conflict of interest with any financial organization regarding the material discussed in this manuscript.
References Fuschs, G., and S. Sundell. 1975. Quantitative determination of 4-methylimidazole as 1-acetyl derivative in caramel color by gas-liquid chromatography. Journal of Agriculture and Food Chemistry 23: 120–122. Hengel, M., and T. Shibamoto. 2013. Carcinogenic 4(5)-methylimidazole found in beverages, sauces, and caramel colors:
123
Chemical properties, analysis, and biological activities. Journal of Agriculture and Food Chemistry 61: 780–789. Kim, T.R., S.U. Kim, Y. Shin, J.Y. Kim, S.M. Lee, and J.H. Kim. 2013. Determination of 4-Methylimidazole and 2-Acetyl-4()tetrahydroxybutylimidazole in Caramel Color and Processed Foods by LC-MS/MS. Preventive Nutrition and Food Science 18: 263–268. Li, A., P. Wei, H.C. Hsu, and R.G. Cooks. 2013a. Direct analysis of 4-methylimidazole in foods using paper spray mass spectrometry. Analyst 138: 4624–4630. Li, K., S. Liu, Y. Tan, N. Chao, X. Tian, L. Qi, W.A. Powell, X. Jiang, and Y. Gai. 2013b. Optimized GC–MS method to simultaneously quantify acetylated aldose, ketose, and alditol for plant tissues based on derivatization in a methyl sulfoxide/1methylimidazole system. Journal of Agriculture and Food Chemistry 61: 4011–4018. Murray, F.J. 2011. Does 4-methylimidazole have tumor preventive activity in the rat? Food and Chemical Toxicology 49: 320–322. Petruci, J.F., E.A. Pereira, and A.A. Cardoso. 2013. Determination of 2-methylimidazole and 4-methylimidazole in caramel colors by capillary electrophoresis. Journal of Agriculture and Food Chemistry 61: 2263–2267. Schlee, C., M. Markova, J. Schrank, F. Laplagne, R. Schneider, and D.W. Lachenmeier. 2012. Determination of 2-methylimidazole, 4-methylimidazole and 2-acetyl-4-(1,2,3,4-tetrahydroxybutyl)imidazole in caramel colours and cola using LC/MS/MS. Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences 927: 223–226. Wang, J., and W.C. Schnute. 2012. Simultaneous quantitation of 2-acetyl-4-tetrahydroxybutylimidazole, 2- and 4-methylimidazoles, and 5-hydroxymethylfurfural in beverages by ultrahighperformance liquid chromatography-tandem mass spectrometry. Journal of Agriculture and Food Chemistry 60: 917–921. Yamaguchi, H., and T. Masuda. 2011. Determination of 4(5)methylimidazole in soy sauce and other foods by LC-MS/MS after solid-phase extraction. Journal of Agriculture and Food Chemistry 59: 9770–9775.
RESEARCH ARTICLE
Quantification of 4-methylimidazole in carbonated beverages by ultra-performance liquid chromatography-tandem mass spectrometry Hyo-Hyun Cho • Kyung-Oh Shin • Cho-Hee Seo • Shin-Hee Lee • Hwan-Soo Yoo • Hye-Ran Yoon • Jeong-Woo Kim • Yong-Moon Lee
Received: 11 August 2014 / Accepted: 20 October 2014 Ó The Pharmaceutical Society of Korea 2014
Abstract 2- and 4-methylimidazoles (2-MI and 4-MI) are undesired byproducts produced during the manufacture of caramel color used to darken food products such as carbonated beverages. The Office of Environmental Health Hazard Assessment in California listed 4-MI as carcinogen in January 2011 with a proposed no significant risk level at 29 lg per person per day. Thus, a quantitative analytical measurement for 2-MI and 4-MI is desired for reliable risk assessments for exposure. An ultra-performance liquid chromatography (UPLC) coupled tandem mass spectrometric (MS/MS) method was developed for the quantification of 4-MI in beverage samples. Chromatographic separation of 2-MI and 4-MI were achieved by using a PFP reversed-phase column and a stepwise gradient of methanol and distilled water containing 0.1 % formic acid. Identification and quantification of 2-MI and 4-MI were performed using electrospray ionization-tandem mass monitoring the precursor to product ion transitions for 2-MI at m/z 83.1 ? 42.2 and 4-MI at m/z 83.1 ? 56.1 with melamine at m/z 127.1 ? 85.1 as the internal standard. The performance of the method was evaluated against validation parameters such as specificity, carryover, linearity and calibration, correlation of determination (r2), H.-H. Cho K.-O. Shin C.-H. Seo S.-H. Lee H.-S. Yoo Y.-M. Lee (&) College of Pharmacy, Chungbuk National University, Cheongju 361-763, Korea e-mail: [email protected] H.-R. Yoon College of Pharmacy, Duksung Women’s University, Seoul 132-714, Korea J.-W. Kim Department of Life Science and Technology, PaiChai University, Daejeon 302-735, Korea
detection limit, precision, accuracy, and recovery. Calibration curves at 10–400 ng/mL were constructed by plotting concentration versus peak-area ratio (analyte/ internal standard) and fitting the data with a weighted 1/x. The accuracy of the assay ranged from 93.58 to 110.53 % for all analytes. Intra-assay precision for 2-MI and 4-MI were below 7.28 (relative standard deviation/RSD %) at QC samples. Here we present a new and improved method using UPLC-MS/MS to significantly simplify sample preparation and decrease chromatographic run time. This method allows accurate and reproducible quantification of 4-MI in carbonated beverages as low as sub ng/mL (ppb) levels. Keywords 2-Methylimidazole 4-Methylimidazole UPLC-tandem mass spectrometry Abbreviations 2-MI 2-Methylimidazole 4-MI 4-Methylimidazole UPLC Ultra-performance liquid chromatography PFP Pentafluorophenyl MS/MS Tandem mass spectrometry QC Quality control MeOH Methanol
Introduction Methylimidazoles (2-MI and 4-MI) were found as undesired byproducts in the browning of certain foods such as carbonated beverages and soy sources through the Maillard reaction between carbohydrates and amino-containing
123
H.-H. Cho et al. Fig. 1 Typical chromatograms of 4-MI and 2-MI. UPLC separated with gradient elution for 4-MI (1.04 min) (a) and 2-MI (0.94 min) (b)
compounds. Recently, The Office of Environment Health Hazard Assessment (OEHHA) in California listed 4-MI as carcinogen in January 2011 with a proposed no significant risk level (NSRL) at 29 lg per person per day. In January 2014, a consumer review of various beverages in the United States reported that Pepsi ONE and Malta Goya contain the chemical in excess of 29 lg per can or bottle. Although the correlation of adverse health effects of exposure to methylimidazoles is controversial, sensitive or excessive consuming populations should be precautious. Several procedures have been reported for the determination of imidazoles compounds in caramel-colored foods. Most procedures are based on HPLC-mass spectrometry (Hengel and Shibamoto 2013; Kim et al. 2013; Schlee et al. 2012; Yamaguchi and Masuda 2011), gas chromatography–mass spectrometry (GC–MS)(Fuschs and Sundell 1975; Li et al. 2013a), paper spray MS (Li et al. 2013b), and electrophoresis (Petruci et al. 2013). The International Agency for Research on Cancer (IARC) noted a general lack of occurrence data of 2-MI and 4-MI in beverages. Thus, the objective of this study was to develop an ultraperformance liquid chromatography-tandem mass spectrometry (UPLC–MS/MS) method for the quantitation of 4-MI in various carbonated beverages without sample preparation. This method was applied to overview the imidazole contents in carbonated beverages from the Korean markets so that it could provide possible risk assessments.
123
Materials and methods Chemical and reagents Chemical standards, 2-MI, 4-MI and melamine (internal standard) shown in Fig. 1 were purchased from SigmaAldrich (St. Louis, MO, USA). For mobile phase preparation, methanol and formic acid were obtained from Sigma-Aldrich (St. Louis, MO, USA). Deionized (DI) water was obtained from a Millipore water station. Standard stock solutions were prepared by dissolving each chemical to DI water at 1 mg/mL. Working standard solutions were prepared by mixing individual standard stock solutions in DI water at 10 lg/mL and then diluted subsequently to 1 lg/mL and 100 ng/mL. Internal standard (IS) stock solution was diluted in DI water to 1 lg/mL to prepare calibration standards or spike unknown samples. Sample preparation Bottled carbonated beverage samples were purchased from local grocery stores. All samples were stored at room temperature until opened for analysis. A 500 lL aliquot of each sample was transferred into Eppendorf tube and degassed for 10 min in a sonication bath at room temperature. After the tube was centrifuged at 12,000 rpm for 10 min at 4 °C, a 100 lL aliquot of each degassed sample was diluted with 800 lL of DI water and vortex mixed
Quantification of 4-methylimidazole in carbonated beverages
with 100 lL of a 100 ng/mL IS stock solution. Each prepared sample (5 lL) was subjected to UPLC-MS/MS analysis. UPLC-MS/MS conditions The UPLC-MS/MS analysis was performed using an ACQUIRTY series UPLC system coupled in-line to a Xevo TQ MS system (Waters, MA, USA). Chromatographic separation was performed using a KINETEX PFP column (4.6 9 50 mm, 2.6 lm) (Phenomenex, CA) at 35 °C. Mobile phase A used 0.1 % formic acid in deionized water. Mobile phase B used 0.1 % formic acid in MeOH. Gradient elution was performed using mobile phases A and B at a flow rate of 0.6 mL/min with the following parameters: initial conditions, A/B = 95:5 maintained for 0.2 min, linear gradient of A/B = 20:80 for 1.3 min which was maintained for 0.5 min, then the gradient was switched back to initial conditions. Column was re-equilibrated for 2 min. The total run time was 4 min per sample. The volume of injected sample was 5 lL. For optimization, a mixture of 2-MI and 4-MI standard was infused directly into the mass spectrometry. All source parameters and ionization conditions were adjusted to improve the sensitivity of the assay. The configurations for mass spectrometry were: capillary voltage at 3.5 kV; source temperature at 150 °C; desolvation temperature at 650 °C; desolvation gas flow rate at 1000 L/hr; and cone gas flow rate at 150 L/ hr. Analyses were performed using electrospray ionization in the positive-ion mode with multiple reaction monitoring to select both precursor and product ions specific to each analyte simultaneously from a single injection. The MS/ MS transitions (m/z) were at 83.1 ? 42.2 for 2-MI, 83.1 ? 56.1 for 4-MI, 127.1 ? 85.1 for melamine. The optimized collision energies for 2-MI, 4-MI, and melamine were 14, 14, and 18 V, respectively. Method validation To validate the UPLC-MS/MS method, several parameters such as the limit of detection (LOD), the limit of quantification (LOQ), linearity, recovery, stability and repeatability were evaluated. The linearity was assessed with calibration standards at 10, 40, 80, 200, and 400 ng/mL. To validate the method, standard solutions of 2-MI and 4-MI standards at 10, 40, 80, and 200 ng/mL were analyzed five times daily. The reproducibility of the proposed method was expressed by relative standard deviation percent (% RSD). Analysis of carbonated beverage containing 2-MI and 4-MI were performed in triplicates. To evaluate the intra-day precision and accuracy, we extracted quality control (QC) samples at LOQ (10 ng/mL), LQ (40 ng/mL),
MQ (80 ng/mL), and HQ (199 ng/mL). The LOQ was calculated as signal to noise ratio (S/N ratio). The matrix effect of 4-MI was calculated as % in relation to spiked amount when standard solution (0.5 lg/L) was added to a blank matrix of sample A and deionized water.
Results Method development The chemical structures of 2-MI, 4-MI, and melamine as an internal standard were shown in Table 1. 2-MI and 4-MI are isomers containing alkaline ‘‘pyridine-like’’ nitrogen and weak acidic ‘‘pyrrole-like’’ amino nitrogen. They have the same molecular weight. From a previous report, methanol was the first choice for organic solvent to escape the interference from acetonitrile (Wang and Schnute 2012). Therefore, methanol was selected as a leading mobile phase. The chromatographic separation using column depends on the character of attached functional group on the surface of the column. In order to separate high polar compounds 2-MI and 4-MI, we chose KinetexÒ PFP column bearing pentafluorphenyl group on the surface of the column with high selectivity deference to polar compounds. In these separation conditions, a standard mixture of 4-MI (40 ng/mL) and 2-MI (20 ng/mL) was completely separated in 1.2 min (Fig. 1). However, this separation condition was not recommended for the quantitation of analyte due to the occurrence of low precision and low accuracy originated from the short retention time. In practice, isocratic elution of mobile phase (A:B = 20:80, v/v) mixture at a flow rate 0.6 ml/min resulted in co-elution of 4-MI and 2-MI at 1.65 min. The observed protonated molecular ions [M?H]? were used as precursor ions. Multiple reactions monitoring (MRM) transitions were used to determine the product ions by infusing standard 4-MI and 2-MI. In this condition, the precursor ion (m/ z 83.1) of 4-MI and 2-MI was scanned in the product scan conditions at cone voltage of 35 V and collision energy of 15 V to produce fragment ions. The fragmented ions of 4-MI and 2-MI had identical m/z value. However, their intensity pattern was clearly different, with the main fragment ions of 4-MI and 2-MI at m/z 56 and m/z 42, respectively (Fig. 2). The chromatographic separation of 4-MI and 2-MI in isocratic elution was not completed with overlapped peaks easily differentiated by main fragment ion intensity. The calculated fragment ion ratio of 10.7 % for 4-MI and 7.6 % for 2-MI represented maximum permitted tolerance values of 4-MI at ±30 % and 2-MI at ±50 % based on the European Commission Council Directive 96/23/EC/SANCO/1805/2000.
123
H.-H. Cho et al. Table 1 The chemical structures, formula and exact mass for analytes
4-MI
2-MI
Melamine(I.S.)
H N
H N
Structure
N
NH2 N
N H2N
Chemical formula
C4H6N2
C4H6N2
C3H6N6
Exact mass
82.05
82.05
139.04
N N
NH2
Fig. 2 MS/MS spectrum of 4-MI (a) and 2-MI (b) using direct infusion method
Method validation The method validation was evaluated for selectivity, recovery, precision, sensitivity, linearity, and accuracy. Selectivity was confirmed by the absence of quantifiable peaks with the assay using blank samples and positive peak detection at specific retention time with 4-MI and melamine (I.S.) in MRM mode at m/z 83.1 ? m/z 56.1 for 4-MI and m/z 127.1 ? m/z 85.1 for melamine. There was no quantifiable peak in blank samples, indicating highly selective monitoring of UPLC-MS/MS conditions. Carryover was not observed when the volume of injector needle washing solution was increased to 600–1,800 lL. Recovery was evaluated by spiking 500 ppb of 4-MI in deionized water (DW) and in blank matrix, respectively. The matrix
123
effect was calculated by the following formula: matrix effect = (quantified amount in matrix)/(quantified amount in DW) 9 100 %. Results are summarized in Table 2. The average recovery was 94.7 ± 1.5 % RSD, indicating high reproducibility. Precision was addressed by injecting 4-MI at four concentrations (11, 40, 80, and 199 ng/mL, n = 3). The relative retention area (RRA) and relative retention time (RRT) using the ratio of 4-MI/melamine peaks were calculated from the UPLC-MS/MS MRM chromatogram. As shown in Table 3, excellent precision was observed overall for these experimental conditions, although 4-MI at low concentration of 11 ng/mL with relatively high variation of % RSD (\8 %) was observed. The LOQ [10 was observed. Considering the instrumental detection limit in practical usage, LOQ was determined at 2 ng/mL (S/N 13).
Quantification of 4-methylimidazole in carbonated beverages Table 2 Matrix effect of spiked 500 ng/mL of 4-MI (n = 3) MRM Transition
83.1 ? 56.1
#
500 ng/ mL in DW Peak area
500 ng/mL in matrix Peak area
Matrix effect (%)a
1
113,895
107,445
94.3
2
105,880
102,004
96.3
3
123,716
115,664
93.5
Ave. (%)
94.70
RSD (%)
Table 5 Quantitation of 4-MI in beverages purchased from Korean Markets
1.52
# Indicates sample number a
Matrix effect (%) = Quantified amount in matrix/Quantified amount in DW 9 100
Table 3 Precision tests for relative retention area (RRA) and relative retention time (RRT) with 4-MI and melamine Spike amount (ng/mL)
4-MI/IS
RSD %
RRT
RSD %
11
0.05
7.28
1.14
0.35
40
0.053
4.16
1.14
0.24
80
0.102
2.65
1.13
0.24
199
0.285
0.75
1.13
0.17
RSD(%) = standard concentration 9 100
deviation
of
the
concentration/mean
4-MI (ng/mL)
Cola 1
374.08 ± 15.55
Cola 2
1,050.27 ± 38.30
Cola 3
912.50 ± 40.51
Cola 4
1,751.6 ± 65.59
Apricot drink
166.35 ± 4.83
Beer 1
17.85 ± 1.37
Beer 2
15.84 ± 0.38
Beer 3
12.95 ± 0.51
Beer 4
14.38 ± 0.18
Beer 5
19.25 ± 0.31
The calibration curve was linear between 10 and 400 ng/ mL of 4-MI standard solution (r2 = 0.9998, Y = 0.1467X-1.0099) (Fig. 3). Accuracy test with four different concentrations of 4-MI was addressed by % accuracy calculated by (observed amount from the 4-MI/IS peak area ratio)/(specified amount) 9 100 % (Table 4). Analysis of 4-MI in food samples We analyzed food beverages of 10 different brands available in Korea markets. The concentration of 2-MI was not detectable in all tested beverages. However, variable high amount of 4-MI was detected in all dark-colored cola samples, with a range of 166.35–1,751.60 ng/mL. The 4-MI concentrations in beer samples were relatively low (\20 ng/mL) with small S.D. values (Table 5).
60 y = 0.1467x - 1.0099, r2 = 0.9998
50
Analyte / IS ratio
Sample
40 30 20
Discussion
10
The main purpose of this study is to establish a conventional method which is able to monitor the non-intentional impurity of a potential carcinogen, 4-MI during the manufacturing process (Murray 2011). To develop an optimized method, the choice of mobile phase and column was essential to the identification and quantitation of each analyte. Previously, Wang and Schnute (2012) have chosen methanol and acetonitrile as optimal organic solvents for the analysis of 2-MI and 4-MI. However, strong interference was reported by acetonitrile cluster ion [2M?H]? which had identical mass-to-charge ratio (m/z) at 83 corresponding to the m/z of 2-MI or 4-MI (Wang and Schnute 2012). Therefore, methanol was the first choice for organic solvent to escape the interference from acetonitrile. The validation parameters including selectivity, recovery, precision, sensitivity, linearity, and accuracy were predominantly evaluated compared to developed LC-MS/
0 0
50
100
150
200
250
300
350
400
4-MI Conc. (ng/mL)
Fig. 3 Calibration curve of 4-MI by UPLC-MS/MS method
Table 4 Accuracy (%) test for four concentrations of 4-MI Spike amount (ng/ mL)
Calc. value (ng/ mL)
Accuracy (%)
% Dev
11
10.45 ± 0.01
95.00 ± 0.13
40
43.48 ± 1.03
108.7 ± 2.58
-5.00 8.70
80
76.55 ± 2.39
95.69 ± 3.00
-4.31
199
200.95 ± 0.91
100.98 ± 0.45
0.98
123
H.-H. Cho et al.
MS methods (Kim et al. 2013; Li et al. 2013a; Schlee et al. 2012). In conclusion, a UPLC-MS/MS method was developed. It was demonstrated to be successful for the quantification of 4-MI in carbonated beverages. Unlike conventional methods that require a sample pretreatment step with solid phase extraction (SPE), we simplified the sample pretreatment using solvent dilution without any difficulties. Among the dark-colored samples tested, 4-MI was observed in all of them in the range of 224–693 ng/ mL(ppb). All samples appeared to be 2-MI free. A lightcolored, lime-flavored carbonated beverage with no quantifiable target analyte was assayed as comparison. Our results showed that the UPLC-MS/MS method could simplify sample preparation, shorten analysis time, and allow accurate and reproducible quantification of 4-MI in carbonated beverages at as low as sub ng/mL (ppb) levels. Acknowledgments This study was partly supported by the Research Grant of the Chungbuk National University (2011) and a NRF Grant (KRF-2010-0025271) funded by the Korean Government. Conflict of interest Authors certify that there is no conflict of interest with any financial organization regarding the material discussed in this manuscript.
References Fuschs, G., and S. Sundell. 1975. Quantitative determination of 4-methylimidazole as 1-acetyl derivative in caramel color by gas-liquid chromatography. Journal of Agriculture and Food Chemistry 23: 120–122. Hengel, M., and T. Shibamoto. 2013. Carcinogenic 4(5)-methylimidazole found in beverages, sauces, and caramel colors:
123
Chemical properties, analysis, and biological activities. Journal of Agriculture and Food Chemistry 61: 780–789. Kim, T.R., S.U. Kim, Y. Shin, J.Y. Kim, S.M. Lee, and J.H. Kim. 2013. Determination of 4-Methylimidazole and 2-Acetyl-4()tetrahydroxybutylimidazole in Caramel Color and Processed Foods by LC-MS/MS. Preventive Nutrition and Food Science 18: 263–268. Li, A., P. Wei, H.C. Hsu, and R.G. Cooks. 2013a. Direct analysis of 4-methylimidazole in foods using paper spray mass spectrometry. Analyst 138: 4624–4630. Li, K., S. Liu, Y. Tan, N. Chao, X. Tian, L. Qi, W.A. Powell, X. Jiang, and Y. Gai. 2013b. Optimized GC–MS method to simultaneously quantify acetylated aldose, ketose, and alditol for plant tissues based on derivatization in a methyl sulfoxide/1methylimidazole system. Journal of Agriculture and Food Chemistry 61: 4011–4018. Murray, F.J. 2011. Does 4-methylimidazole have tumor preventive activity in the rat? Food and Chemical Toxicology 49: 320–322. Petruci, J.F., E.A. Pereira, and A.A. Cardoso. 2013. Determination of 2-methylimidazole and 4-methylimidazole in caramel colors by capillary electrophoresis. Journal of Agriculture and Food Chemistry 61: 2263–2267. Schlee, C., M. Markova, J. Schrank, F. Laplagne, R. Schneider, and D.W. Lachenmeier. 2012. Determination of 2-methylimidazole, 4-methylimidazole and 2-acetyl-4-(1,2,3,4-tetrahydroxybutyl)imidazole in caramel colours and cola using LC/MS/MS. Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences 927: 223–226. Wang, J., and W.C. Schnute. 2012. Simultaneous quantitation of 2-acetyl-4-tetrahydroxybutylimidazole, 2- and 4-methylimidazoles, and 5-hydroxymethylfurfural in beverages by ultrahighperformance liquid chromatography-tandem mass spectrometry. Journal of Agriculture and Food Chemistry 60: 917–921. Yamaguchi, H., and T. Masuda. 2011. Determination of 4(5)methylimidazole in soy sauce and other foods by LC-MS/MS after solid-phase extraction. Journal of Agriculture and Food Chemistry 59: 9770–9775.
