- Email: [email protected]
Three liquid chromatographic methods for the analysis of aflatoxins in for different corn (Zea mays) matrices
Journal of Food Composition and Analysis 54 (2016) 20–26
Contents lists available at ScienceDirect
Journal of Food Composition and Analysis journal homepage: www.elsevier.com/locate/jfca
Original research article
Three liquid chromatographic methods for the analysis of aflatoxins in for different corn (Zea mays) matrices Hyun Ee Oka , Hyelee Junga , Sung-Eun Leeb , Ockjin Peakc , Hyang Sook Chuna,* a
Advanced Food Safety Research Group, BK21 Plus, School of Food Science and Technology, Chung-Ang University, Anseong 17546, Republic of Korea School of Applied Biosciences, Kyungpook National University, Daegu 702-701, Republic of Korea c Food Contaminants Division, National Institute of Food & Drug Safety Evaluation, Osong 28159,Republic of Korea b
A R T I C L E I N F O
Article history: Received 2 May 2016 Received in revised form 18 September 2016 Accepted 22 September 2016 Available online 29 September 2016 Keywords: Aflatoxins analyses Corn Derivatization method Photochemical reactor for enhanced detection Kobra electrochemical cell Trifluoroacetic acid Zea mays
A B S T R A C T
Liquid chromatographic analyses of aflatoxins (AFs) in corn, with post-column derivatization using a photochemical reactor for enhanced detection (PHRED) and a Kobra electrochemical cell system were compared with the pre-column derivatization method using trifluoroacetic acid (TFA). AFs in four different corn matrices were analyzed and validated in terms of the limit of detection (LOD), limit of quantification (LOQ), linearity, accuracy, and precision. The LOD and LOQ for the PHRED, Kobra, and TFA methods were 0.004–0.03, 0.01–0.05, and 0.03–0.17 ng/g, respectively, and 0.01–0.10, 0.02–0.14, and 0.11–0.51 ng/g, respectively. Accuracy expressed as average recoveries was 79–110% for PHRED, 70–109% for Kobra, and 77–133% for TFA. In the three derivatization methods, the mean recoveries of AFs were significantly different, at some but not all concentrations, between matrices of dehulled corn and corn with hull (p < 0.05). For dehulled yellow corn, the TFA method consistently gave slightly poor recovery values for AFs B1 and G1 than did the PHRED and Kobra methods. The values for the TFA methods were improved by using a modified cleanup procedure. These results indicate that PHRED and Kobra derivatization methods as well as TFA method comply with the analytical requirements for AF analyses in corns. ã 2016 Elsevier Inc. All rights reserved.
1. Introduction Aflatoxins (AFs) are natural secondary metabolites produced by some molds, mainly Aspergillus flavus and Aspergillus parasiticus. They are contaminants of agricultural commodities, particularly under critical temperature and humidity conditions, before or during harvest, or because of inappropriate storage (Rustom, 1997; Sweeney and Dobson, 1998). AFs B1 (AFB1), B2 (AFB2), G1 (AFG1) and G2 (AFG2) can contaminate corn, wheat, rice, groundnuts, pistachios, cottonseed, copra, and spices. Corn (Zea mays L.) and its associated by-products are traditional staple foods and feed ingredients for humans and animals. Mycotoxins such as AFs, fumonisins, deoxynivalenol, and zearalenone are frequently detected in corn and corn products. In particular, AF contamination in corn probably generates the most concern since the AFs are potent liver toxins and carcinogens.
* Corresponding author. E-mail addresses: [email protected], [email protected] (H.S. Chun). http://dx.doi.org/10.1016/j.jfca.2016.09.010 0889-1575/ã 2016 Elsevier Inc. All rights reserved.
National and international institutions and organizations, such as the European Commission (EC), the US Food and Drug Administration (FDA), the World Health Organization (WHO), and the Food and Agriculture Organization (FAO), have recognized the potential health risks to animals and humans posed by consuming AF contaminated food and feed (Manetta, 2011). To protect human and animal health, many countries have enacted specific regulations for mycotoxins in food and animal feed. In 2013, more than 100 countries worldwide had enacted regulations or detailed guidelines for the control of mycotoxins in food and animal feed. Most of these countries regulate the sum of the levels of the four most prominent types of AFs B1, B2, G1, and G2, or in combination with a specific limit for AFB1 in food. Currently, the European Commission set maximum levels for AFB1 (5.0 mg/kg) and total AFs (10.0 mg/kg) in all cereals and their products including corn (maize) and corn products. In contrast, for corn to be subjected to sorting or other physical treatment before human consumption or use as an ingredient in foodstuffs, maximum levels for AFB1 and total AFs are 5.0 mg/kg and 10.0 mg/kg, respectively (European Commission, 2006a, 2006b).
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
South Korea set regulatory limits for AFB1 (10.0 mg/kg) and total AFs (20.0 mg/kg) in all cereals and their products including corn and corn products (MFDS, 2015). Considering the toxicity of AFs and the maximum acceptable limit set in food and feedstuffs, the analytical identification and quantification of AFs at very low levels must be carried out using reliable methods. The literature includes many reports on AF detection. There are many sensitive and specific methods, but simple and rapid methods are now also available. As new analytical technologies have developed, they have been quickly incorporated into mycotoxin testing strategies. Sometimes reports reflect advances in analytical science, but often modifications of existing methods to improve the analytical process are published. Because AFs are naturally strongly fluorescent compounds, the high-performance liquid chromatography (HPLC) analysis of these molecules is most often achieved by fluorescence detection. Reversed-phase (RP) eluents quench the fluorescence of AFB1 and AFG1 (Kok, 1994). It is for this reason, to enhance the response of these two analytes, that chemical derivatization is commonly required; use is made of pre- or post-column derivatization with suitable fluorophores to improve detectability. The pre-column approach uses trifluoroacetic acid (TFA) with the formation of the corresponding hemiacetals (Akiyama et al., 2001), which are relatively unstable derivatives. Post-column derivatization is based on the reaction of the 8,9double bond of AFs with halogens. Initially, the post-column reaction used iodination (Shepherd and Gilbert, 1984); however, it has several disadvantages, such as peak broadening and the risk of iodine crystallization. An alternative method involves bromination by an electrochemical cell (Kobra cell) with potassium bromide (KBr) dissolved in an acidified mobile phase, or by the addition of bromide or pyridinium hydrobromide perbromide to a mobile phase, and using a short reaction coil at ambient temperature (Stroka et al., 2001; Senyuva and Gilbert, 2005; Brera et al., 2007). The bromination methods offer the advantages of being simple, rapid, and easy to automate, which improve reproducibility and ruggedness and reduce analysis time. A post-column derivatization method that is supposed to be analytically equivalent to iodination and bromination is the photochemical method. It is based on the formation of hemiacetals of AFB1 and AFG1, due to the UV radiation of the HPLC column eluate (Joshua, 1993; Waltking and Wilson, 2006). In the Food Code published by the Ministry of Food and Drug Safety in Korea (MFDS, 2015), TFA derivatization is the only method adopted as an analytical method for AFs using HPLC. Hence, the aim of the present study was to compare the official Korean precolumn derivatization method using TFA, as well as the widely used reference method in many countries, with two post-column derivatization methods—a photochemical reactor for enhanced detection (PHRED) and a Kobra electrochemical cell system—for the determination of AFs in four different corn matrices. The HPLC methods coupled with fluorescence detection (HPLC-FLD) with immunoaffinity cleanup were validated in terms of the limit of detection (LOD), limit of quantification (LOQ), linearity, accuracy, and precision.
20 sieve. All samples were stored at bags.
21
18 C in aluminum zipper
2.2. Materials A standard stock solution for total AFs was used, specifically, an ‘Aflatoxin Mix’ from Sigma-Aldrich (St. Louis, MO). This is a ‘readyto-use’ product, supplied in a vial. It contains the following: AF 2600 ng/mL (1000 ng/mL of aflatoxin B1 and G1, and 300 ng/mL of aflatoxin B2 and G2). Methanol (MeOH, HPLC grade), acetonitrile (ACN, LC grade), and HPLC-grade water for extraction were purchased from Burdick & Jackson (Muskegon, MI). Sodium chloride (NaCl) and Tween 20 were obtained from Junsei (Tokyo, Japan). KBr, TFA, and nitric acid (HNO3) were purchased from Sigma-Aldrich. Cellulose filter paper No.4, glass microfiber filters (GF/A and GF/B) and 0.22-mm PVDF syringe filters were obtained from Whatman (Maidstone, UK). AflaTest1 WB immunoaffinity columns (IACs) were purchased from VICAM (Watertown, MA). 2.3. Sample preparation Samples were analyzed using HPLC-FLD, according to the Korean Food Code (MFDS, 2015). Briefly, a 25-g sample was placed in a 250-mL beaker with 100 mL of MeOH:water (70:30, v/v) containing 1% NaCl and then blended for 5 min in a high-speed blender (ULTRA-TURRAX1; IKA Werke, Staufen, Germany). After extraction, the sample was filtered through a filter paper (Whatman No. 4). A 10-mL volume of filtrate extract was then diluted with 30 mL of water containing 1% Tween 20. After filtration through a GF/A filter, 20 mL of the filtrate were passed through an IAC at a flow rate of one drop per second. The IAC was washed with 10 mL of water and dried by rapidly passing air through. The toxins were eluted into screw-cap amber glass vials with 3 mL of ACN for TFA derivatization. The eluate was evaporated in a heated aluminum block at 50 C using a gentle stream of nitrogen. The dried residues were derivatized as follows: 200 mL of TFA were added, residues were left to stand for 15 min in a place where they were protected from direct UV light, and then they were diluted with 800 mL of ACN:water (20:80, v/v). The derivatized sample was vortexed for 30 s and then filtered through a 0.22-mm membrane into HPLC vials for autoinjection. The filtered solution (10 mL) was injected into the HPLC. Because AFs are subject to light degradation, all analytical work must be adequately protected from daylight. Therefore, all the procedures were carried out under subdued light and protected from direct UV light. For derivatization using the Kobra cell and the PHRED instrument, the eluate was first dried, redissolved with 1 mL of MeOH:water (1:1, v/v), and then filtered through a 0.22-mm PVDF syringe filter. 2.4. Instrumentation and conditions
2.1. Samples
2.4.1. TFA derivatization Chromatographic analysis was performed with an Agilent 1260 Infinity Quaternary LC system (Agilent Technologies, Santa Clara, CA), comprising pumps, an autosampler, a column oven, and a fluorescence detector. AFs were monitored at lex = 360 nm and lem = 450 nm. The RP column of Synergi Hydro-RP (250 mm 4.6 mm, 4 mm; Phenomenex, Torrance, CA) was operated at 40 C. The mobile phase comprised ACN:water (25:75, v/v) and the flow rate was 1.0 mL/min. The injection volume of sample was 10 mL.
The types of corn samples analyzed were yellow corn and waxy corn, dehulled yellow corn and dehulled waxy corn, depending on the variety of corn and the processing conditions. A minimum sample size of 1 kg was used. All samples were finely ground using a high-speed blender; a ground sample would pass through a No.
2.4.2. PHRED cell (post-column photochemical derivatization cell) Chromatographic analysis was performed on an Agilent 1260 Quaternary LC system (Agilent Technologies) with a post-column photochemical derivatization reactor. This reactor (Aura Industries, New York, NY) with a UV lamp (l = 254 nm) and a knitted
2. Materials and methods
22
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
reactor coil of 0.74 mL (15 m 0.25 mm) was added between the chromatographic column and fluorescence detector. AFs were monitored at lex = 360 nm and lem = 455 nm. The RP Symmetry C18 column (150 mm 4.6 mm, 3.5 mm; Waters, Santry, Ireland) was operated at 35 C. The mobile phase comprised ACN:water: MeOH (14:59:27, v/v/v) and the flow rate was 1.0 mL/min. The injection volume of the sample was 50 mL. 2.4.3. Kobra cell (electrochemical cell, post-column bromination derivatization cell) Chromatographic analysis was performed with an Agilent 1260 Infinity Quaternary LC system (Agilent Technologies). The signal enhancement was performed using the Kobra cell from RBiopharm Rhône Inc. (Glasgow, Scotland) and the detection was obtained with an Agilent 1260 fluorescence detector (Agilent Technologies) with an excitation wavelength of 362 nm and emission wavelength of 440 nm. The RP Symmetry C18 column (150 mm 4.6 mm, 3.5 mm; Waters, Santry, Ireland) was operated at 35 C. The injection volume of the sample was 50 mL. The isocratic mobile phase comprised water:MeOH:ACN (60:25:17, v/v/ v); 120 mg of KBr and 350 mL of 4 M HNO3 were added to 1 L of the mobile phase. The flow rate was 0.8 mL/min. Using this mobile phase, retention times of 6.1, 7.3, 8.2, and 10.1 min were registered for AFs G2, G1, B2, and B1, respectively. 2.5. Single laboratory validation The method was validated according to IUPAC guidelines (Thompson et al., 2002). The LOD was calculated by the “b + 3s” approach. For this purpose, a blank sample standard deviation “s”
(10 injections) was calculated, where “b” was the blank signal. Similarly, the LOQ was calculated by applying the formula “b + 10s”. After carrying out the theoretical LOQ calculation, the value was verified by injecting 10 times a blank sample fortified at the LOQ level and the level of precision was determined. Applicability and linearity of the method were assessed from the calibration curves. Selectivity was tested by adding toxins to positive samples and then observing the increase of toxin peaks. For the determination of accuracy and precision, spiking experiments were carried out at two concentration levels (2 and 10 ng/g for AFB1 and AFG1, and 0.6 and 3 ng/g for AFB2 and AFG2) for three different runs, each in three repetitions. The accuracy was expressed in terms of recovery rate of spiked quantity of AFs. Additionally, a reference material (T04105; FAPAS, York, UK) with assigned values for AFB1 (5.21 ng/ g), AFB2 (2.26 ng/g), AF G1 (2.32 ng/g) and AFG2 (0.98 ng/g) was also analyzed in the validation study. The precision was expressed as the relative standard deviation of repeatability (RSD). 3. Results and discussion 3.1. Performance characteristics of method 3.1.1. Selectivity The three derivatization methods were determined to be selective for AFs even in the presence of interfering or co-eluting compounds, because no overlapping of matrix compounds was observed. All AF peaks emerging from the three derivatizations were detected without the presence of any interfering peaks, as was the case with the patterns of the chromatograms in the four matrices (Fig. 1). Use of the immunoaffinity column clean-up
Fig. 1. Chromatograms of aflatoxins using TFA, PHRED and Kobra cell derivatization in yellow corn (left) and dehulled yellow corn (right) based standards (AFB1/G1, 5 ng/g; AFB2/G2, 1.5 ng/g). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
permitted an efficient elimination of possible endogenous matrixinterfering substances. These compounds showed retention times significantly different from those of AFs under the adopted experimental conditions. 3.1.2. Linearity For all derivatization methods, the calibration curves were linear in the range 0.16–16 ng/g for AFB1 and AFG1, and in the range 0.048–4.8 ng/g for AFB2 and AFG2. Correlation coefficients (R2) of each method ranged from 0.9966 to 1.000 in the four matrices. Calibration sensitivity is the slope of the calibration curves at the concentration of interest; the greater the slope the higher the sensitivity. The slope of the calibration curve for the Kobra cell derivatization was higher than the slopes for the other methods. However, LOD levels for the Kobra cell and PHRED derivatizations were similar (Table 1).
23
3.1.3. LOD and LOQ LOD and LOQ of AFs in the four matrices were determined for the three derivatization methods (Table 1). For the TFA derivatization, the LODs and LOQs in the four matrices were 0.03–0.17 and 0.11–0.44 ng/g, respectively. The LODs and LOQs of AFs for the PHRED derivatization were 0.004–0.03 ng/g and 0.01–0.10, respectively. In case of the Kobra cell derivatization, LODs and LOQs in the four matrices were 0.01–0.05 and 0.02–0.14 ng/g, respectively. Among the three derivatization methods, the LODs and LOQs for the PHRED method of derivatization were slightly lower than those for the TFA and Kobra cell methods in four corn matrices. In a recent study of AF contamination levels in maize grain, the LODs and LOQs for Kobra cell derivatization were determined to be 0.013 and 0.038 ng/g for AFB1, 0.007 and 0.022 ng/g for AFB2, 0.021 and 0.063 ng/g for AFG1, and 0.016 and 0.049 ng/g for AFG2, respectively (Atmaca et al., 2015). A study carried out in Iran
Table 1 Analytical parameters of aflatoxins using TFA, PHRED and Kobra cell derivatization in each matrix. Derivati-zation
Mycotoxins
Matrix
LOD
LOQ
Slope aa
Intercept ba
R2
(ng/g) TFA
AFB1
AFB2
AFG1
AFG2
PHRED
AFB1
AFB2
AFG1
AFG2
Kobra cell
AFB1
AFB2
AFG1
AFG2
Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled
0.08 0.08 0.06 0.08 0.05 0.03 0.06 0.09 0.12 0.13 0.08 0.11 0.12 0.09 0.17 0.15
0.23 0.25 0.17 0.24 0.15 0.11 0.17 0.27 0.36 0.39 0.23 0.33 0.36 0.27 0.51 0.44
0.52 0.55 0.52 0.53 0.63 0.67 0.52 0.64 0.20 0.22 0.21 0.21 0.33 0.35 0.34 0.34
0.04 0.00 0.06 0.05 0.02 0.01 0.06 0.00 0.04 0.00 0.01 0.01 0.03 0.02 0.01 0.00
1.0000 1.0000 0.9998 0.9998 1.0000 1.0000 0.9998 0.9998 0.9998 0.9999 0.9999 0.9999 1.0000 0.9999 0.9996 0.9999
Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled
0.02 0.02 0.02 0.02 0.01 0.01 0.01 0.004 0.03 0.02 0.03 0.03 0.01 0.01 0.01 0.01
0.05 0.06 0.06 0.05 0.02 0.02 0.02 0.01 0.10 0.05 0.08 0.08 0.03 0.02 0.04 0.03
30.5 40.2 34.2 40.2 70.1 95.1 78.9 95.5 12.3 22.0 17.5 20.1 32.9 61.8 49.0 56.7
12.44 3.44 4.88 0.77 8.01 2.76 2.47 0.34 1.88 0.48 0.98 0.84 2.10 1.15 10.68 0.08
0.9987 0.9999 0.9999 1.0000 0.9989 0.9998 0.9998 1.0000 0.9999 1.0000 0.9966 1.0000 0.9997 1.0000 0.9959 1.0000
Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled
0.03 0.04 0.02 0.03 0.01 0.02 0.02 0.01 0.03 0.05 0.04 0.03 0.02 0.01 0.03 0.02
0.08 0.11 0.06 0.09 0.02 0.05 0.05 0.03 0.10 0.14 0.13 0.10 0.06 0.04 0.10 0.07
106 93.6 109 88.9 167 146 162 147 59.8 51.8 56.4 46.9 84.8 74.4 62.3 53.1
7.26 2.25 20.87 7.64 1.21 1.68 5.99 8.53 2.02 1.16 5.83 7.68 1.35 0.29 1.84 3.31
1.0000 1.0000 0.9998 0.9997 1.0000 1.0000 0.9999 0.9995 1.0000 1.0000 1.0000 0.9996 1.0000 1.0000 1.0000 0.9992
a Values are the mean of six calibration curves prepared and plotted on separate occasions; slope and intercept refer to the regression y = ax + b. The range of the standard solution was 0.16–16 ng/g for AFB1/B2 and 0.048–4.8 ng/g for AFB2/G2.
24
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
revealed that the LODs of AFB1 and total AFs were 0.1 and 0.4 ng/g, respectively (Hadiani et al., 2009). For the PHRED derivatization, the LODs were 0.08, 0.02, 0.09, and 0.02 ng/g for AFB1, AFB2, AFG1, and AFG2, respectively, and the LOQs were 0.17, 0.04, 0.22, and 0.04 ng/g for AFB1, AFB2, AFG1, and AFG2, respectively (Imperato et al., 2011). Furthermore, the LODs for some cereal samples are reported to be 0.5 ng/g for AFB1 and AFG1, and 1.0 ng/g for AFB2 and AFG2 (Lutfullah and Hussain, 2012). 3.1.4. Accuracy and precision Accuracy and precision studies were conducted with four corn samples using the three derivatization methods after spiking 2 and 10 ng/g for AFB1 and AFG1, and 0.6 and 3 ng/g for AFB2 and AFG2, respectively. Results of accuracy and precision, expressed as% recovery and RSD, for TFA derivatization for each AF at two contamination levels spiked to four corn samples are tabulated in Table 2. Accuracy and precision, expressed as recovery and RSD percentages, respectively, at two tested levels were 77–133% and 0.4–12.3% for AFs, respectively. This result meets the performance criteria for EU regulations (European Commission, 2006a, 2006b) for the recommended precision, except for dehulled yellow corn. The RSD should be derived from the Horwitz equation and the method performance should not exceed twice of this value (European Commission, 2006a, 2006b). These values for RSD are 30% and 21% for AFB1 and AFG1 levels at 2.0 and 10.0 ng/g, and 33% and 26% for AFB2 and AFG2 levels at 0.6 and 3.0 ng/g, respectively. Furthermore, Regulation (EC) No. 401/2006 (European Commission, 2006a, 2006b) established that recoveries should be in the range 70–110% and 50–120% for concentrations in the ranges 1– 10 ng/g and <1 ng/g, respectively. From these results, to meet analytical guidelines for the determination of AFs in dehulled yellow corn, the analytical method involving the use of TFA derivatization needs to be improved. The unsatisfactory recovery value (>120%) obtained for the dehulled yellow corn might be partly explained by interfering substances from dehulled yellow corn. The main constituents of the corn kernel include starch, fat, protein, fiber, sugar, ash and pigments (chlorophyll, xanthophylls,
and flavonoids) that can serve as interfering substances for the aflatoxin analysis. Because the hull is the protective outer covering of the corn kernel, which is mainly composed of fiber, interfering substances such as pigments are more easily extracted from the dehulled yellow corn than other corn matrices. Table 3 shows the accuracy and precision for the determination of AFs using the PHRED derivatization system in four corn samples. The results for all samples were 79–110% for accuracy (% recovery) and 0.3–6.5% for RSD. Here, the precision values were twice lower than in the case of TFA derivatization. In the case of AF analysis using the Kobra cell derivatization system, recovery values were 70–109% and RSD values were 0.3–7.4%, which are in agreement with the recommended values (Table 4). In the three derivatization methods, the mean recoveries of AFs were significantly different, at some but not all of concentration, between matrices of dehulled corn and corn with hull (p < 0.05). In particular, recovery values of AFs B1 and G1 (10 ng/g) obtained with the TFA method were significantly higher in dehulled yellow corn than those in yellow corn (p < 0.05). Overall, the values of accuracy and precision obtained with the PHRED derivatization method were slightly more satisfactory compared with those obtained with the TFA method and Kobra cell derivatization. To further evaluate the accuracy of the three derivatization methods, the levels of AFs in a corn-based reference material were determined. As shown in Table 5, the levels of AFs B1, B2, G1 and G2 were 3.95–4.92 ng/g, 1.97–2.33 ng/g, 1.58–2.45 ng/g and 0.67– 1.07 ng/g, respectively, which are within the acceptable range of FAPAS (2.92–7.50 for AFB1, 1.27–3.26 for AFB2, 1.30–3.35 for AFG1 and 0.55–1.41 for AFG2). These results confirmed the accuracy of the analytical methods using three different derivatization methods. 3.2. Modified cleanup procedure for TFA derivatization in yellow dehulled corn To achieve good accuracy with the TFA derivatization method, the dehulled yellow corn was analyzed using a modified method;
Table 2 Recoveries and repeatability for the determination of aflatoxins in spiked corn samples using TFA derivatization. Mycotoxins
Spiked concentration (ng/g)
Yellow corn
AFB1
2 10 0.6 3 2 10 0.6 3
87.5 3.57 90.7 2.95 80.6 6.37 82.8 2.08 90.2 5.89 90.6 3.95 82.0 12.3 78.6 0.42
Yellow corn, dehulled
Waxy corn
Waxy corn, dehulled
94.6 3.11* 122.2 0.89* 87.7 6.56 107.6 2.07* 96.0 5.88 133.3 2.08* 77.4 8.20 94.7 6.83*
78.4 3.97 83.6 0.94 79.5 11.7 83.6 3.22 85.9 4.54 91.9 1.70 83.1 9.34 79.8 5.03
90.9 4.28* 91.2 1.93* 89.0 8.35* 86.5 3.62* 89.3 5.97 95.3 2.11* 84.7 4.57 85.6 3.76*
Recoveries RSD (%)
AFB2 AFG1 AFG2
*
Student t-test (p < 0.05) between dehulled corns and corns with hull (yellow and waxy corns).
Table 3 Recoveries and repeatability for the determination of aflatoxins in spiked corn samples using PHRED derivatization. Mycotoxins
AFB1 AFB2 AFG1 AFG2
*
Spiked concentration (ng/g)
Yellow corn
2 10 0.6 3 2 10 0.6 3
85.9 6.53 95.0 3.59 88.0 1.53 95.0 3.18 84.6 5.67 101.7 3.02 107.3 3.61 106.9 1.17
Yellow corn, dehulled
Waxy corn
Waxy corn, dehulled
79.2 0.36 95.2 1.34 91.1 1.88 95.8 0.29 101.4 1.69* 108.8 1.01 109.5 1.92 103.7 0.81
89.4 5.72 88.9 0.69 91.6 1.33 91.2 1.29 91.9 2.02 97.1 0.92 108.5 0.93* 108.9 0.83*
80.3 1.08 103.6 1.76* 89.5 1.14 103.7 3.33* 81.9 1.28 93.4 5.20 81.9 1.23 85.9 2.52
Recoveries RSD (%)
Student t-test (p < 0.05) between dehulled corns and corns with hull (yellow and waxy corns).
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
25
Table 4 Recoveries and repeatability for the determination of aflatoxins in spiked corn samples using Kobra cell derivatization. Mycotoxins
AFB1 AFB2 AFG1 AFG2
*
Spiked concentration (ng/g)
Yellow corn
2 10 0.6 3 2 10 0.6 3
71.7 6.43 86.0 3.32 70.1 0.41 82.7 2.85 75.6 4.06 92.6 2.94 72.3 3.59 85.1 5.16
Yellow corn, dehulled
Waxy corn
Waxy corn, dehulled
83.2 7.42* 98.5 2.18* 98.2 4.63* 103.4 0.77* 89.2 6.40* 100.9 6.36 73.4 3.44 71.7 3.04
79.4 5.31 76.7 0.38 79.7 1.16 79.5 0.33 89.1 0.60 90.9 0.36 109.2 1.26 108.9 0.94
76.9 1.25 84.5 1.03* 95.8 2.49* 92.0 0.76* 95.8 1.50* 96.9 2.57 106.8 2.31 99.8 2.32
Recoveries RSD (%)
Student t-test (p < 0.05) between dehulled corns and corns with hull (yellow and waxy corns).
Table 5 Analytical data of aflatoxins using three derivatization methods in reference material. Analyte
Assigned value (ng/g)
Satisfactory range (ng/g)
AFB1 AFB2 AFG1 AFG2
5.21 2.26 2.32 0.98
2.92–7.50 1.27–3.26 1.30–3.35 0.55–1.41
TFA
PHRED
Kobra cell
4.92 0.05 2.33 0.11 2.36 0.16 0.93 0.06
3.95 0.04 2.10 0.02 2.45 0.14 1.07 0.02
Mean SD (ng/g)
Table 6 Recoveries and repeatability for the determination of aflatoxins in spiked dehulled yellow corn using TFA derivatization, after the modified cleanup procedure. Mycotoxins
Spiked concentration (ng/g)
Recoveries RSD (%)a
AFB1
2 10 0.6 3 2 10 0.6 3
97.46 3.67 97.73 1.11 90.68 6.07 84.98 2.43 98.13 5.66 104.42 3.65 85.82 11.63 81.83 7.11
AFB2 AFG1 AFG2
a
GF/B filter.
the glass microfiber filter was changed to a GF/B filter. Recovery and RSD were 82–104% and 1.1–11.6%, respectively (Table 6). The results completely fulfilled the performance criteria required by Regulation (EC) 401/2006 (European Commission, 2006a, 2006b). Brera et al. (2007) reported that derivatization techniques used with pre-column derivatization (TFA) and post-column derivatization (pyridinium hydrobromide perbromide or an electrochemical cell by the addition of bromide to the mobile phase) were not significantly different when compared by the Student’s t test, although they only determined AFB1 (Brera et al., 2007). Our results are in good agreement with this. 4. Conclusion Three derivatization methods comply with the analytical requirements for AF analyses in the different corn matrices. The PHRED and Kobra derivatization methods are equivalent to the TFA method, but have the advantage of being relatively simple and more rapid. It can be expected that these two post-column derivatization methods will be extensively used for the determination of AFs in corn and corn products. Acknowledgments This research was supported by a grant (10162KFDA995) from the National Institute of Food and Drug Safety Evaluation in 2010,
4.19 0.00 1.97 0.06 1.58 0.15 0.67 0.08
by the Brain Korea 21 Plus Project, and by the Chung-Ang University Excellent Student Scholarship, Republic of Korea. References Akiyama, H., Goda, Y., Tanaka, T., Toyoda, M., 2001. Determination of aflatoxins B1, B2, G1 and G2 in spices using a multifunctional column clean-up. J. Chromatogr. A 932, 153–157. Atmaca, E., Güvenç, D., Aksoy, A., 2015. Determination of aflatoxin levels in maize grain by high performance liquid chromatography using an immunoaffinity column cleanup. Kafkas Universitesi Veteriner Fakultesi Dergisi 21, 619–622. doi:http://dx.doi.org/10.9775/kvfd.2015.13081. Brera, C., Debegnach, F., Minardi, F., Pannunzi, E., De Santis, B., Miraglia, M., 2007. Immunoaffinity column cleanup with liquid chromatography for determination of aflatoxin B1 in corn samples: interlaboratory study. J. AOAC Int. 90, 765–772. European Commission, 2006a. Commission regulation (EC) 401/2006 of 23 february 2006 laying down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs. Off. J. Eur. Union L70, 12–34. European Commission, 2006b. Commission regulation (EC) 1881/2006 of 19 december 2006 setting maximum levels for certain contaminants in foodstuffs. Off. J. Eur. Union L364, 5–23. Hadiani, M.R., Yazdanpanah, H., Amirahmadi, M., Soleimani, H., Shoeibi, S.H., Khosrokhavar, R., 2009. Evaluation of aflatoxin contamination in maize from Mazandaran province in Iran. J. Med. Plants 8, 109–114. Imperato, R., Campone, L., Piccinelli, A.L., Veneziano, A., Rastrelli, L., 2011. Survey of aflatoxins and ochratoxin A contamination in food products imported in Italy. Food Control 22, 1905–1910. Joshua, H., 1993. Determination of aflatoxins by reversed-phase high performance liquid chromatography with post-column in-line photochemical derivatization and fluorescence detection. J. Chromatogr. 654, 247–254. Kok, W.Th., 1994. Derivatization reactions for the determination of aflatoxins by liquid chromatography with fluorescence detection. J. Chromatogr. B 659, 127– 137. Lutfullah, G., Hussain, A., 2012. Studies on contamination level of aflatoxins in some cereals and beans of Pakistan. Food Control 23, 32–36. Ministry of Food and Drug Safety (MFDS), 2015. Food code. Available in http://fse. foodnara.go.kr/residue/RS/jsp/menu_02_01_03.jsp?idx=726 (accessed January 2016). Manetta, A.C., 2011. Aflatoxins: their measure and analysis detection. In: Irineo Torres-Pacheco (Ed.), Measurement and Control, pp. 93–108. InTech. http:// www.intechopen.com/books/aflatoxins-detection-measurement-and-control/ aflatoxins-their-measureand-analysis. Rustom, I.Y.S., 1997. Aflatoxin in food and feed: occurrence, legislation and inactivation by physical methods. Food Chem. 59, 57–67. Senyuva, H.Z., Gilbert, J., 2005. Immunoaffinity column cleanup with liquid chromatography using post-column bromination for determination of aflatoxins in hazelnut paste: interlaboratory study. J. AOAC Int. 88, 526–535. Shepherd, M.J., Gilbert, J., 1984. An investigation of HPLC post-column iodination conditions for the enhancement of aflatoxin B1 fluorescence. Food Addit. Contam. 1, 325–335.
26
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
Stroka, J., Anklam, E., Joerissen, U., Gilbert, J., 2001. Determination of aflatoxin B1 in baby food (infant formula) by immunoaffinity column cleanup liquid chromatography with post-column bromination: collaborative study. J. AOAC Int. 84, 1116–1123. Sweeney, M.J., Dobson, A.D.W., 1998. Mycotoxin production by Aspergillus, Fusarium and Penicillium species. Int. J. Food Microbiol. 43, 141–158.
Thompson, M., Ellison, S.L.R., Wood, R., 2002. Harmonized guidelines for singlelaboratory validation of methods of analysis (IUPAC technical report). Pure Appl. Chem. 74 (5), 835–855. Waltking, A.E., Wilson, D., 2006. Liquid chromatographic analysis of aflatoxin using post-column photochemical derivatization: collaborative study. J. AOAC Int. 89, 678–692.
Contents lists available at ScienceDirect
Journal of Food Composition and Analysis journal homepage: www.elsevier.com/locate/jfca
Original research article
Three liquid chromatographic methods for the analysis of aflatoxins in for different corn (Zea mays) matrices Hyun Ee Oka , Hyelee Junga , Sung-Eun Leeb , Ockjin Peakc , Hyang Sook Chuna,* a
Advanced Food Safety Research Group, BK21 Plus, School of Food Science and Technology, Chung-Ang University, Anseong 17546, Republic of Korea School of Applied Biosciences, Kyungpook National University, Daegu 702-701, Republic of Korea c Food Contaminants Division, National Institute of Food & Drug Safety Evaluation, Osong 28159,Republic of Korea b
A R T I C L E I N F O
Article history: Received 2 May 2016 Received in revised form 18 September 2016 Accepted 22 September 2016 Available online 29 September 2016 Keywords: Aflatoxins analyses Corn Derivatization method Photochemical reactor for enhanced detection Kobra electrochemical cell Trifluoroacetic acid Zea mays
A B S T R A C T
Liquid chromatographic analyses of aflatoxins (AFs) in corn, with post-column derivatization using a photochemical reactor for enhanced detection (PHRED) and a Kobra electrochemical cell system were compared with the pre-column derivatization method using trifluoroacetic acid (TFA). AFs in four different corn matrices were analyzed and validated in terms of the limit of detection (LOD), limit of quantification (LOQ), linearity, accuracy, and precision. The LOD and LOQ for the PHRED, Kobra, and TFA methods were 0.004–0.03, 0.01–0.05, and 0.03–0.17 ng/g, respectively, and 0.01–0.10, 0.02–0.14, and 0.11–0.51 ng/g, respectively. Accuracy expressed as average recoveries was 79–110% for PHRED, 70–109% for Kobra, and 77–133% for TFA. In the three derivatization methods, the mean recoveries of AFs were significantly different, at some but not all concentrations, between matrices of dehulled corn and corn with hull (p < 0.05). For dehulled yellow corn, the TFA method consistently gave slightly poor recovery values for AFs B1 and G1 than did the PHRED and Kobra methods. The values for the TFA methods were improved by using a modified cleanup procedure. These results indicate that PHRED and Kobra derivatization methods as well as TFA method comply with the analytical requirements for AF analyses in corns. ã 2016 Elsevier Inc. All rights reserved.
1. Introduction Aflatoxins (AFs) are natural secondary metabolites produced by some molds, mainly Aspergillus flavus and Aspergillus parasiticus. They are contaminants of agricultural commodities, particularly under critical temperature and humidity conditions, before or during harvest, or because of inappropriate storage (Rustom, 1997; Sweeney and Dobson, 1998). AFs B1 (AFB1), B2 (AFB2), G1 (AFG1) and G2 (AFG2) can contaminate corn, wheat, rice, groundnuts, pistachios, cottonseed, copra, and spices. Corn (Zea mays L.) and its associated by-products are traditional staple foods and feed ingredients for humans and animals. Mycotoxins such as AFs, fumonisins, deoxynivalenol, and zearalenone are frequently detected in corn and corn products. In particular, AF contamination in corn probably generates the most concern since the AFs are potent liver toxins and carcinogens.
* Corresponding author. E-mail addresses: [email protected], [email protected] (H.S. Chun). http://dx.doi.org/10.1016/j.jfca.2016.09.010 0889-1575/ã 2016 Elsevier Inc. All rights reserved.
National and international institutions and organizations, such as the European Commission (EC), the US Food and Drug Administration (FDA), the World Health Organization (WHO), and the Food and Agriculture Organization (FAO), have recognized the potential health risks to animals and humans posed by consuming AF contaminated food and feed (Manetta, 2011). To protect human and animal health, many countries have enacted specific regulations for mycotoxins in food and animal feed. In 2013, more than 100 countries worldwide had enacted regulations or detailed guidelines for the control of mycotoxins in food and animal feed. Most of these countries regulate the sum of the levels of the four most prominent types of AFs B1, B2, G1, and G2, or in combination with a specific limit for AFB1 in food. Currently, the European Commission set maximum levels for AFB1 (5.0 mg/kg) and total AFs (10.0 mg/kg) in all cereals and their products including corn (maize) and corn products. In contrast, for corn to be subjected to sorting or other physical treatment before human consumption or use as an ingredient in foodstuffs, maximum levels for AFB1 and total AFs are 5.0 mg/kg and 10.0 mg/kg, respectively (European Commission, 2006a, 2006b).
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
South Korea set regulatory limits for AFB1 (10.0 mg/kg) and total AFs (20.0 mg/kg) in all cereals and their products including corn and corn products (MFDS, 2015). Considering the toxicity of AFs and the maximum acceptable limit set in food and feedstuffs, the analytical identification and quantification of AFs at very low levels must be carried out using reliable methods. The literature includes many reports on AF detection. There are many sensitive and specific methods, but simple and rapid methods are now also available. As new analytical technologies have developed, they have been quickly incorporated into mycotoxin testing strategies. Sometimes reports reflect advances in analytical science, but often modifications of existing methods to improve the analytical process are published. Because AFs are naturally strongly fluorescent compounds, the high-performance liquid chromatography (HPLC) analysis of these molecules is most often achieved by fluorescence detection. Reversed-phase (RP) eluents quench the fluorescence of AFB1 and AFG1 (Kok, 1994). It is for this reason, to enhance the response of these two analytes, that chemical derivatization is commonly required; use is made of pre- or post-column derivatization with suitable fluorophores to improve detectability. The pre-column approach uses trifluoroacetic acid (TFA) with the formation of the corresponding hemiacetals (Akiyama et al., 2001), which are relatively unstable derivatives. Post-column derivatization is based on the reaction of the 8,9double bond of AFs with halogens. Initially, the post-column reaction used iodination (Shepherd and Gilbert, 1984); however, it has several disadvantages, such as peak broadening and the risk of iodine crystallization. An alternative method involves bromination by an electrochemical cell (Kobra cell) with potassium bromide (KBr) dissolved in an acidified mobile phase, or by the addition of bromide or pyridinium hydrobromide perbromide to a mobile phase, and using a short reaction coil at ambient temperature (Stroka et al., 2001; Senyuva and Gilbert, 2005; Brera et al., 2007). The bromination methods offer the advantages of being simple, rapid, and easy to automate, which improve reproducibility and ruggedness and reduce analysis time. A post-column derivatization method that is supposed to be analytically equivalent to iodination and bromination is the photochemical method. It is based on the formation of hemiacetals of AFB1 and AFG1, due to the UV radiation of the HPLC column eluate (Joshua, 1993; Waltking and Wilson, 2006). In the Food Code published by the Ministry of Food and Drug Safety in Korea (MFDS, 2015), TFA derivatization is the only method adopted as an analytical method for AFs using HPLC. Hence, the aim of the present study was to compare the official Korean precolumn derivatization method using TFA, as well as the widely used reference method in many countries, with two post-column derivatization methods—a photochemical reactor for enhanced detection (PHRED) and a Kobra electrochemical cell system—for the determination of AFs in four different corn matrices. The HPLC methods coupled with fluorescence detection (HPLC-FLD) with immunoaffinity cleanup were validated in terms of the limit of detection (LOD), limit of quantification (LOQ), linearity, accuracy, and precision.
20 sieve. All samples were stored at bags.
21
18 C in aluminum zipper
2.2. Materials A standard stock solution for total AFs was used, specifically, an ‘Aflatoxin Mix’ from Sigma-Aldrich (St. Louis, MO). This is a ‘readyto-use’ product, supplied in a vial. It contains the following: AF 2600 ng/mL (1000 ng/mL of aflatoxin B1 and G1, and 300 ng/mL of aflatoxin B2 and G2). Methanol (MeOH, HPLC grade), acetonitrile (ACN, LC grade), and HPLC-grade water for extraction were purchased from Burdick & Jackson (Muskegon, MI). Sodium chloride (NaCl) and Tween 20 were obtained from Junsei (Tokyo, Japan). KBr, TFA, and nitric acid (HNO3) were purchased from Sigma-Aldrich. Cellulose filter paper No.4, glass microfiber filters (GF/A and GF/B) and 0.22-mm PVDF syringe filters were obtained from Whatman (Maidstone, UK). AflaTest1 WB immunoaffinity columns (IACs) were purchased from VICAM (Watertown, MA). 2.3. Sample preparation Samples were analyzed using HPLC-FLD, according to the Korean Food Code (MFDS, 2015). Briefly, a 25-g sample was placed in a 250-mL beaker with 100 mL of MeOH:water (70:30, v/v) containing 1% NaCl and then blended for 5 min in a high-speed blender (ULTRA-TURRAX1; IKA Werke, Staufen, Germany). After extraction, the sample was filtered through a filter paper (Whatman No. 4). A 10-mL volume of filtrate extract was then diluted with 30 mL of water containing 1% Tween 20. After filtration through a GF/A filter, 20 mL of the filtrate were passed through an IAC at a flow rate of one drop per second. The IAC was washed with 10 mL of water and dried by rapidly passing air through. The toxins were eluted into screw-cap amber glass vials with 3 mL of ACN for TFA derivatization. The eluate was evaporated in a heated aluminum block at 50 C using a gentle stream of nitrogen. The dried residues were derivatized as follows: 200 mL of TFA were added, residues were left to stand for 15 min in a place where they were protected from direct UV light, and then they were diluted with 800 mL of ACN:water (20:80, v/v). The derivatized sample was vortexed for 30 s and then filtered through a 0.22-mm membrane into HPLC vials for autoinjection. The filtered solution (10 mL) was injected into the HPLC. Because AFs are subject to light degradation, all analytical work must be adequately protected from daylight. Therefore, all the procedures were carried out under subdued light and protected from direct UV light. For derivatization using the Kobra cell and the PHRED instrument, the eluate was first dried, redissolved with 1 mL of MeOH:water (1:1, v/v), and then filtered through a 0.22-mm PVDF syringe filter. 2.4. Instrumentation and conditions
2.1. Samples
2.4.1. TFA derivatization Chromatographic analysis was performed with an Agilent 1260 Infinity Quaternary LC system (Agilent Technologies, Santa Clara, CA), comprising pumps, an autosampler, a column oven, and a fluorescence detector. AFs were monitored at lex = 360 nm and lem = 450 nm. The RP column of Synergi Hydro-RP (250 mm 4.6 mm, 4 mm; Phenomenex, Torrance, CA) was operated at 40 C. The mobile phase comprised ACN:water (25:75, v/v) and the flow rate was 1.0 mL/min. The injection volume of sample was 10 mL.
The types of corn samples analyzed were yellow corn and waxy corn, dehulled yellow corn and dehulled waxy corn, depending on the variety of corn and the processing conditions. A minimum sample size of 1 kg was used. All samples were finely ground using a high-speed blender; a ground sample would pass through a No.
2.4.2. PHRED cell (post-column photochemical derivatization cell) Chromatographic analysis was performed on an Agilent 1260 Quaternary LC system (Agilent Technologies) with a post-column photochemical derivatization reactor. This reactor (Aura Industries, New York, NY) with a UV lamp (l = 254 nm) and a knitted
2. Materials and methods
22
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
reactor coil of 0.74 mL (15 m 0.25 mm) was added between the chromatographic column and fluorescence detector. AFs were monitored at lex = 360 nm and lem = 455 nm. The RP Symmetry C18 column (150 mm 4.6 mm, 3.5 mm; Waters, Santry, Ireland) was operated at 35 C. The mobile phase comprised ACN:water: MeOH (14:59:27, v/v/v) and the flow rate was 1.0 mL/min. The injection volume of the sample was 50 mL. 2.4.3. Kobra cell (electrochemical cell, post-column bromination derivatization cell) Chromatographic analysis was performed with an Agilent 1260 Infinity Quaternary LC system (Agilent Technologies). The signal enhancement was performed using the Kobra cell from RBiopharm Rhône Inc. (Glasgow, Scotland) and the detection was obtained with an Agilent 1260 fluorescence detector (Agilent Technologies) with an excitation wavelength of 362 nm and emission wavelength of 440 nm. The RP Symmetry C18 column (150 mm 4.6 mm, 3.5 mm; Waters, Santry, Ireland) was operated at 35 C. The injection volume of the sample was 50 mL. The isocratic mobile phase comprised water:MeOH:ACN (60:25:17, v/v/ v); 120 mg of KBr and 350 mL of 4 M HNO3 were added to 1 L of the mobile phase. The flow rate was 0.8 mL/min. Using this mobile phase, retention times of 6.1, 7.3, 8.2, and 10.1 min were registered for AFs G2, G1, B2, and B1, respectively. 2.5. Single laboratory validation The method was validated according to IUPAC guidelines (Thompson et al., 2002). The LOD was calculated by the “b + 3s” approach. For this purpose, a blank sample standard deviation “s”
(10 injections) was calculated, where “b” was the blank signal. Similarly, the LOQ was calculated by applying the formula “b + 10s”. After carrying out the theoretical LOQ calculation, the value was verified by injecting 10 times a blank sample fortified at the LOQ level and the level of precision was determined. Applicability and linearity of the method were assessed from the calibration curves. Selectivity was tested by adding toxins to positive samples and then observing the increase of toxin peaks. For the determination of accuracy and precision, spiking experiments were carried out at two concentration levels (2 and 10 ng/g for AFB1 and AFG1, and 0.6 and 3 ng/g for AFB2 and AFG2) for three different runs, each in three repetitions. The accuracy was expressed in terms of recovery rate of spiked quantity of AFs. Additionally, a reference material (T04105; FAPAS, York, UK) with assigned values for AFB1 (5.21 ng/ g), AFB2 (2.26 ng/g), AF G1 (2.32 ng/g) and AFG2 (0.98 ng/g) was also analyzed in the validation study. The precision was expressed as the relative standard deviation of repeatability (RSD). 3. Results and discussion 3.1. Performance characteristics of method 3.1.1. Selectivity The three derivatization methods were determined to be selective for AFs even in the presence of interfering or co-eluting compounds, because no overlapping of matrix compounds was observed. All AF peaks emerging from the three derivatizations were detected without the presence of any interfering peaks, as was the case with the patterns of the chromatograms in the four matrices (Fig. 1). Use of the immunoaffinity column clean-up
Fig. 1. Chromatograms of aflatoxins using TFA, PHRED and Kobra cell derivatization in yellow corn (left) and dehulled yellow corn (right) based standards (AFB1/G1, 5 ng/g; AFB2/G2, 1.5 ng/g). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
permitted an efficient elimination of possible endogenous matrixinterfering substances. These compounds showed retention times significantly different from those of AFs under the adopted experimental conditions. 3.1.2. Linearity For all derivatization methods, the calibration curves were linear in the range 0.16–16 ng/g for AFB1 and AFG1, and in the range 0.048–4.8 ng/g for AFB2 and AFG2. Correlation coefficients (R2) of each method ranged from 0.9966 to 1.000 in the four matrices. Calibration sensitivity is the slope of the calibration curves at the concentration of interest; the greater the slope the higher the sensitivity. The slope of the calibration curve for the Kobra cell derivatization was higher than the slopes for the other methods. However, LOD levels for the Kobra cell and PHRED derivatizations were similar (Table 1).
23
3.1.3. LOD and LOQ LOD and LOQ of AFs in the four matrices were determined for the three derivatization methods (Table 1). For the TFA derivatization, the LODs and LOQs in the four matrices were 0.03–0.17 and 0.11–0.44 ng/g, respectively. The LODs and LOQs of AFs for the PHRED derivatization were 0.004–0.03 ng/g and 0.01–0.10, respectively. In case of the Kobra cell derivatization, LODs and LOQs in the four matrices were 0.01–0.05 and 0.02–0.14 ng/g, respectively. Among the three derivatization methods, the LODs and LOQs for the PHRED method of derivatization were slightly lower than those for the TFA and Kobra cell methods in four corn matrices. In a recent study of AF contamination levels in maize grain, the LODs and LOQs for Kobra cell derivatization were determined to be 0.013 and 0.038 ng/g for AFB1, 0.007 and 0.022 ng/g for AFB2, 0.021 and 0.063 ng/g for AFG1, and 0.016 and 0.049 ng/g for AFG2, respectively (Atmaca et al., 2015). A study carried out in Iran
Table 1 Analytical parameters of aflatoxins using TFA, PHRED and Kobra cell derivatization in each matrix. Derivati-zation
Mycotoxins
Matrix
LOD
LOQ
Slope aa
Intercept ba
R2
(ng/g) TFA
AFB1
AFB2
AFG1
AFG2
PHRED
AFB1
AFB2
AFG1
AFG2
Kobra cell
AFB1
AFB2
AFG1
AFG2
Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled
0.08 0.08 0.06 0.08 0.05 0.03 0.06 0.09 0.12 0.13 0.08 0.11 0.12 0.09 0.17 0.15
0.23 0.25 0.17 0.24 0.15 0.11 0.17 0.27 0.36 0.39 0.23 0.33 0.36 0.27 0.51 0.44
0.52 0.55 0.52 0.53 0.63 0.67 0.52 0.64 0.20 0.22 0.21 0.21 0.33 0.35 0.34 0.34
0.04 0.00 0.06 0.05 0.02 0.01 0.06 0.00 0.04 0.00 0.01 0.01 0.03 0.02 0.01 0.00
1.0000 1.0000 0.9998 0.9998 1.0000 1.0000 0.9998 0.9998 0.9998 0.9999 0.9999 0.9999 1.0000 0.9999 0.9996 0.9999
Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled
0.02 0.02 0.02 0.02 0.01 0.01 0.01 0.004 0.03 0.02 0.03 0.03 0.01 0.01 0.01 0.01
0.05 0.06 0.06 0.05 0.02 0.02 0.02 0.01 0.10 0.05 0.08 0.08 0.03 0.02 0.04 0.03
30.5 40.2 34.2 40.2 70.1 95.1 78.9 95.5 12.3 22.0 17.5 20.1 32.9 61.8 49.0 56.7
12.44 3.44 4.88 0.77 8.01 2.76 2.47 0.34 1.88 0.48 0.98 0.84 2.10 1.15 10.68 0.08
0.9987 0.9999 0.9999 1.0000 0.9989 0.9998 0.9998 1.0000 0.9999 1.0000 0.9966 1.0000 0.9997 1.0000 0.9959 1.0000
Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled Yellow corn Yellow corn, dehulled Waxy corn Waxy corn, dehulled
0.03 0.04 0.02 0.03 0.01 0.02 0.02 0.01 0.03 0.05 0.04 0.03 0.02 0.01 0.03 0.02
0.08 0.11 0.06 0.09 0.02 0.05 0.05 0.03 0.10 0.14 0.13 0.10 0.06 0.04 0.10 0.07
106 93.6 109 88.9 167 146 162 147 59.8 51.8 56.4 46.9 84.8 74.4 62.3 53.1
7.26 2.25 20.87 7.64 1.21 1.68 5.99 8.53 2.02 1.16 5.83 7.68 1.35 0.29 1.84 3.31
1.0000 1.0000 0.9998 0.9997 1.0000 1.0000 0.9999 0.9995 1.0000 1.0000 1.0000 0.9996 1.0000 1.0000 1.0000 0.9992
a Values are the mean of six calibration curves prepared and plotted on separate occasions; slope and intercept refer to the regression y = ax + b. The range of the standard solution was 0.16–16 ng/g for AFB1/B2 and 0.048–4.8 ng/g for AFB2/G2.
24
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
revealed that the LODs of AFB1 and total AFs were 0.1 and 0.4 ng/g, respectively (Hadiani et al., 2009). For the PHRED derivatization, the LODs were 0.08, 0.02, 0.09, and 0.02 ng/g for AFB1, AFB2, AFG1, and AFG2, respectively, and the LOQs were 0.17, 0.04, 0.22, and 0.04 ng/g for AFB1, AFB2, AFG1, and AFG2, respectively (Imperato et al., 2011). Furthermore, the LODs for some cereal samples are reported to be 0.5 ng/g for AFB1 and AFG1, and 1.0 ng/g for AFB2 and AFG2 (Lutfullah and Hussain, 2012). 3.1.4. Accuracy and precision Accuracy and precision studies were conducted with four corn samples using the three derivatization methods after spiking 2 and 10 ng/g for AFB1 and AFG1, and 0.6 and 3 ng/g for AFB2 and AFG2, respectively. Results of accuracy and precision, expressed as% recovery and RSD, for TFA derivatization for each AF at two contamination levels spiked to four corn samples are tabulated in Table 2. Accuracy and precision, expressed as recovery and RSD percentages, respectively, at two tested levels were 77–133% and 0.4–12.3% for AFs, respectively. This result meets the performance criteria for EU regulations (European Commission, 2006a, 2006b) for the recommended precision, except for dehulled yellow corn. The RSD should be derived from the Horwitz equation and the method performance should not exceed twice of this value (European Commission, 2006a, 2006b). These values for RSD are 30% and 21% for AFB1 and AFG1 levels at 2.0 and 10.0 ng/g, and 33% and 26% for AFB2 and AFG2 levels at 0.6 and 3.0 ng/g, respectively. Furthermore, Regulation (EC) No. 401/2006 (European Commission, 2006a, 2006b) established that recoveries should be in the range 70–110% and 50–120% for concentrations in the ranges 1– 10 ng/g and <1 ng/g, respectively. From these results, to meet analytical guidelines for the determination of AFs in dehulled yellow corn, the analytical method involving the use of TFA derivatization needs to be improved. The unsatisfactory recovery value (>120%) obtained for the dehulled yellow corn might be partly explained by interfering substances from dehulled yellow corn. The main constituents of the corn kernel include starch, fat, protein, fiber, sugar, ash and pigments (chlorophyll, xanthophylls,
and flavonoids) that can serve as interfering substances for the aflatoxin analysis. Because the hull is the protective outer covering of the corn kernel, which is mainly composed of fiber, interfering substances such as pigments are more easily extracted from the dehulled yellow corn than other corn matrices. Table 3 shows the accuracy and precision for the determination of AFs using the PHRED derivatization system in four corn samples. The results for all samples were 79–110% for accuracy (% recovery) and 0.3–6.5% for RSD. Here, the precision values were twice lower than in the case of TFA derivatization. In the case of AF analysis using the Kobra cell derivatization system, recovery values were 70–109% and RSD values were 0.3–7.4%, which are in agreement with the recommended values (Table 4). In the three derivatization methods, the mean recoveries of AFs were significantly different, at some but not all of concentration, between matrices of dehulled corn and corn with hull (p < 0.05). In particular, recovery values of AFs B1 and G1 (10 ng/g) obtained with the TFA method were significantly higher in dehulled yellow corn than those in yellow corn (p < 0.05). Overall, the values of accuracy and precision obtained with the PHRED derivatization method were slightly more satisfactory compared with those obtained with the TFA method and Kobra cell derivatization. To further evaluate the accuracy of the three derivatization methods, the levels of AFs in a corn-based reference material were determined. As shown in Table 5, the levels of AFs B1, B2, G1 and G2 were 3.95–4.92 ng/g, 1.97–2.33 ng/g, 1.58–2.45 ng/g and 0.67– 1.07 ng/g, respectively, which are within the acceptable range of FAPAS (2.92–7.50 for AFB1, 1.27–3.26 for AFB2, 1.30–3.35 for AFG1 and 0.55–1.41 for AFG2). These results confirmed the accuracy of the analytical methods using three different derivatization methods. 3.2. Modified cleanup procedure for TFA derivatization in yellow dehulled corn To achieve good accuracy with the TFA derivatization method, the dehulled yellow corn was analyzed using a modified method;
Table 2 Recoveries and repeatability for the determination of aflatoxins in spiked corn samples using TFA derivatization. Mycotoxins
Spiked concentration (ng/g)
Yellow corn
AFB1
2 10 0.6 3 2 10 0.6 3
87.5 3.57 90.7 2.95 80.6 6.37 82.8 2.08 90.2 5.89 90.6 3.95 82.0 12.3 78.6 0.42
Yellow corn, dehulled
Waxy corn
Waxy corn, dehulled
94.6 3.11* 122.2 0.89* 87.7 6.56 107.6 2.07* 96.0 5.88 133.3 2.08* 77.4 8.20 94.7 6.83*
78.4 3.97 83.6 0.94 79.5 11.7 83.6 3.22 85.9 4.54 91.9 1.70 83.1 9.34 79.8 5.03
90.9 4.28* 91.2 1.93* 89.0 8.35* 86.5 3.62* 89.3 5.97 95.3 2.11* 84.7 4.57 85.6 3.76*
Recoveries RSD (%)
AFB2 AFG1 AFG2
*
Student t-test (p < 0.05) between dehulled corns and corns with hull (yellow and waxy corns).
Table 3 Recoveries and repeatability for the determination of aflatoxins in spiked corn samples using PHRED derivatization. Mycotoxins
AFB1 AFB2 AFG1 AFG2
*
Spiked concentration (ng/g)
Yellow corn
2 10 0.6 3 2 10 0.6 3
85.9 6.53 95.0 3.59 88.0 1.53 95.0 3.18 84.6 5.67 101.7 3.02 107.3 3.61 106.9 1.17
Yellow corn, dehulled
Waxy corn
Waxy corn, dehulled
79.2 0.36 95.2 1.34 91.1 1.88 95.8 0.29 101.4 1.69* 108.8 1.01 109.5 1.92 103.7 0.81
89.4 5.72 88.9 0.69 91.6 1.33 91.2 1.29 91.9 2.02 97.1 0.92 108.5 0.93* 108.9 0.83*
80.3 1.08 103.6 1.76* 89.5 1.14 103.7 3.33* 81.9 1.28 93.4 5.20 81.9 1.23 85.9 2.52
Recoveries RSD (%)
Student t-test (p < 0.05) between dehulled corns and corns with hull (yellow and waxy corns).
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
25
Table 4 Recoveries and repeatability for the determination of aflatoxins in spiked corn samples using Kobra cell derivatization. Mycotoxins
AFB1 AFB2 AFG1 AFG2
*
Spiked concentration (ng/g)
Yellow corn
2 10 0.6 3 2 10 0.6 3
71.7 6.43 86.0 3.32 70.1 0.41 82.7 2.85 75.6 4.06 92.6 2.94 72.3 3.59 85.1 5.16
Yellow corn, dehulled
Waxy corn
Waxy corn, dehulled
83.2 7.42* 98.5 2.18* 98.2 4.63* 103.4 0.77* 89.2 6.40* 100.9 6.36 73.4 3.44 71.7 3.04
79.4 5.31 76.7 0.38 79.7 1.16 79.5 0.33 89.1 0.60 90.9 0.36 109.2 1.26 108.9 0.94
76.9 1.25 84.5 1.03* 95.8 2.49* 92.0 0.76* 95.8 1.50* 96.9 2.57 106.8 2.31 99.8 2.32
Recoveries RSD (%)
Student t-test (p < 0.05) between dehulled corns and corns with hull (yellow and waxy corns).
Table 5 Analytical data of aflatoxins using three derivatization methods in reference material. Analyte
Assigned value (ng/g)
Satisfactory range (ng/g)
AFB1 AFB2 AFG1 AFG2
5.21 2.26 2.32 0.98
2.92–7.50 1.27–3.26 1.30–3.35 0.55–1.41
TFA
PHRED
Kobra cell
4.92 0.05 2.33 0.11 2.36 0.16 0.93 0.06
3.95 0.04 2.10 0.02 2.45 0.14 1.07 0.02
Mean SD (ng/g)
Table 6 Recoveries and repeatability for the determination of aflatoxins in spiked dehulled yellow corn using TFA derivatization, after the modified cleanup procedure. Mycotoxins
Spiked concentration (ng/g)
Recoveries RSD (%)a
AFB1
2 10 0.6 3 2 10 0.6 3
97.46 3.67 97.73 1.11 90.68 6.07 84.98 2.43 98.13 5.66 104.42 3.65 85.82 11.63 81.83 7.11
AFB2 AFG1 AFG2
a
GF/B filter.
the glass microfiber filter was changed to a GF/B filter. Recovery and RSD were 82–104% and 1.1–11.6%, respectively (Table 6). The results completely fulfilled the performance criteria required by Regulation (EC) 401/2006 (European Commission, 2006a, 2006b). Brera et al. (2007) reported that derivatization techniques used with pre-column derivatization (TFA) and post-column derivatization (pyridinium hydrobromide perbromide or an electrochemical cell by the addition of bromide to the mobile phase) were not significantly different when compared by the Student’s t test, although they only determined AFB1 (Brera et al., 2007). Our results are in good agreement with this. 4. Conclusion Three derivatization methods comply with the analytical requirements for AF analyses in the different corn matrices. The PHRED and Kobra derivatization methods are equivalent to the TFA method, but have the advantage of being relatively simple and more rapid. It can be expected that these two post-column derivatization methods will be extensively used for the determination of AFs in corn and corn products. Acknowledgments This research was supported by a grant (10162KFDA995) from the National Institute of Food and Drug Safety Evaluation in 2010,
4.19 0.00 1.97 0.06 1.58 0.15 0.67 0.08
by the Brain Korea 21 Plus Project, and by the Chung-Ang University Excellent Student Scholarship, Republic of Korea. References Akiyama, H., Goda, Y., Tanaka, T., Toyoda, M., 2001. Determination of aflatoxins B1, B2, G1 and G2 in spices using a multifunctional column clean-up. J. Chromatogr. A 932, 153–157. Atmaca, E., Güvenç, D., Aksoy, A., 2015. Determination of aflatoxin levels in maize grain by high performance liquid chromatography using an immunoaffinity column cleanup. Kafkas Universitesi Veteriner Fakultesi Dergisi 21, 619–622. doi:http://dx.doi.org/10.9775/kvfd.2015.13081. Brera, C., Debegnach, F., Minardi, F., Pannunzi, E., De Santis, B., Miraglia, M., 2007. Immunoaffinity column cleanup with liquid chromatography for determination of aflatoxin B1 in corn samples: interlaboratory study. J. AOAC Int. 90, 765–772. European Commission, 2006a. Commission regulation (EC) 401/2006 of 23 february 2006 laying down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs. Off. J. Eur. Union L70, 12–34. European Commission, 2006b. Commission regulation (EC) 1881/2006 of 19 december 2006 setting maximum levels for certain contaminants in foodstuffs. Off. J. Eur. Union L364, 5–23. Hadiani, M.R., Yazdanpanah, H., Amirahmadi, M., Soleimani, H., Shoeibi, S.H., Khosrokhavar, R., 2009. Evaluation of aflatoxin contamination in maize from Mazandaran province in Iran. J. Med. Plants 8, 109–114. Imperato, R., Campone, L., Piccinelli, A.L., Veneziano, A., Rastrelli, L., 2011. Survey of aflatoxins and ochratoxin A contamination in food products imported in Italy. Food Control 22, 1905–1910. Joshua, H., 1993. Determination of aflatoxins by reversed-phase high performance liquid chromatography with post-column in-line photochemical derivatization and fluorescence detection. J. Chromatogr. 654, 247–254. Kok, W.Th., 1994. Derivatization reactions for the determination of aflatoxins by liquid chromatography with fluorescence detection. J. Chromatogr. B 659, 127– 137. Lutfullah, G., Hussain, A., 2012. Studies on contamination level of aflatoxins in some cereals and beans of Pakistan. Food Control 23, 32–36. Ministry of Food and Drug Safety (MFDS), 2015. Food code. Available in http://fse. foodnara.go.kr/residue/RS/jsp/menu_02_01_03.jsp?idx=726 (accessed January 2016). Manetta, A.C., 2011. Aflatoxins: their measure and analysis detection. In: Irineo Torres-Pacheco (Ed.), Measurement and Control, pp. 93–108. InTech. http:// www.intechopen.com/books/aflatoxins-detection-measurement-and-control/ aflatoxins-their-measureand-analysis. Rustom, I.Y.S., 1997. Aflatoxin in food and feed: occurrence, legislation and inactivation by physical methods. Food Chem. 59, 57–67. Senyuva, H.Z., Gilbert, J., 2005. Immunoaffinity column cleanup with liquid chromatography using post-column bromination for determination of aflatoxins in hazelnut paste: interlaboratory study. J. AOAC Int. 88, 526–535. Shepherd, M.J., Gilbert, J., 1984. An investigation of HPLC post-column iodination conditions for the enhancement of aflatoxin B1 fluorescence. Food Addit. Contam. 1, 325–335.
26
H.E. Ok et al. / Journal of Food Composition and Analysis 54 (2016) 20–26
Stroka, J., Anklam, E., Joerissen, U., Gilbert, J., 2001. Determination of aflatoxin B1 in baby food (infant formula) by immunoaffinity column cleanup liquid chromatography with post-column bromination: collaborative study. J. AOAC Int. 84, 1116–1123. Sweeney, M.J., Dobson, A.D.W., 1998. Mycotoxin production by Aspergillus, Fusarium and Penicillium species. Int. J. Food Microbiol. 43, 141–158.
Thompson, M., Ellison, S.L.R., Wood, R., 2002. Harmonized guidelines for singlelaboratory validation of methods of analysis (IUPAC technical report). Pure Appl. Chem. 74 (5), 835–855. Waltking, A.E., Wilson, D., 2006. Liquid chromatographic analysis of aflatoxin using post-column photochemical derivatization: collaborative study. J. AOAC Int. 89, 678–692.