A Unique and Simple Matrix Preparation Roadmap for Aflatoxin B1, B2, G1, G2 Food Safety Testing Prior to Automated SmartPrep Extractor Solid Phase Extraction Toni Hofhine, Horizon Technology Inc., Salem, NH USA Elizabeth Krantz, Dr. Cheri A. Barta, and Dr. Pamela Doolittle, University of Wisconsin, Madison, WI, USA Robert Buco, Richard Koeritz, and Zachary Lilla, Shimadzu Scientific Instruments, Marlborough, MA, USA Jennifer Claus, Kenneth Espenschied, and Michael Ye, Sigma-Aldrich, Bellefonte, PA, USA Key Words Aflatoxin, Mycotoxins, Derivatization, Solid Phase Extraction (SPE), Liquid-Liquid Extraction (LLE), Food and Drug Administration (FDA), European Union (EU), ultra High Performance Liquid Chromatograph (uhplc) Introduction Mycotoxin testing awareness has increased as countries involved in world trade of raw agriculture and processed consumer products rely on a safe global food supply. Several mycotoxins are naturally produced and monitored; however, aflatoxins, produced mainly by the Aspergillus species, are considered to be the most important mycotoxin group in the world s food supply to monitor 1. The frequency of occurrence and negative health impacts to both animals and humans has spiked interest in the aflatoxins B1, B2, G1, and G2 levels present within foods commonly exported to determine if these levels meet the current acceptable regulatory limits specified by each country. The health impacts from exposure to the four common metabolites B1, B2, G1, and G2 vary greatly based on the consumption level and exposure time with the aflatoxin metabolite, ongoing research goals are to identify if there is an impact on specific animal species or human age, sex, and health 2. Acceptable limits for aflatoxins B1, B2, G1, and G2 in raw agriculture and processed consumer products vary by country, which can negatively impact the fair trade of goods 1. Currently, the United States has an advisory limit for total aflatoxins of 20 μg/kg in finished foodstuffs 3 (Table 1). The European Union legislation sets the maximum level of both the B1 strain and total aflatoxins in foodstuffs depending on the product and level of processing 4 (Table 1). The EU also has performance criteria in place for methods testing for mycotoxins in foodstuffs 5. With a lack of harmonization between countries, the methodology for identification of aflatoxins B1, B2, G1, and G2 in traded products requires efficiency and accuracy.
Table 1: European Union and United States Recommended Maximum Levels for Aflatoxins in Foodstuffs Foodstuffs Maximum B1 levels (μg/kg) Maximum sum of B1, B2, G1, G2 levels (μg/kg) Matrix United States European Union United States European Union Almonds (almond milk) - 8.0 20.0 10.0 Corn - 2.0 20.0 4.0 Groundnuts (peanut paste) - 2.0 20.0 4.0 Capsicum spp. (chili powder, cayenne, paprika), Piper spp. (pepper), Myristica fragrans (nutmeg), Zingiber officinale (ginger), Curcuma longa (turmeric) Curry (Indian curry spice mix) - 5.0 20.0 10.0 Preparation of samples for aflatoxins B1, B2, G1, and G2 testing generally requires liquid-liquid extraction (LLE) as a sample preparation step and solid phase extraction (SPE) as a sample cleanup step to first remove as much as possible, any sample matrix effects, such as color, protein, fat, etc. Proper cleanup of samples allows aflatoxins B1, B2, G1, and G2 to be easily detected and quantified by HPLC analysis. The historical standard practiced for the sample cleanup step is to use immunoaffinity columns (IAC) for the SPE process. Immunoaffinity uses a gel sorbent to bind the aflatoxins with an antibody, allowing interfering compounds to be effectively washed away prior to elution of the aflatoxins. An alternative methodology for sample cleanup is to bind the interfering compounds and immediately elute the aflatoxins from a non-gel, standard solid phase extraction (SPE) cartridge. The effectiveness of the Supel Tox AflaZea SPE Cartridges in combination with the SmartPrep Extractor System from Horizon Technology efficiently prepared a variety of foodstuffs samples representing a quadrant of moisture and viscosity levels prior to analysis using HPLC with fluorescence detection. Sample matrices chosen to represent the quandrant were almond milk, fresh corn, peanut paste, and curry powder were selected. Data presented concludes that aflatoxins B1, B2, G1, and G2 are effectively recovered using the novel and simple sample matrix preparation roadmap to meet both United States and European Union limits. P a ge 2
Experimental Each sample type was drawn from at least two sources (e.g., two containers of peanut paste or 12 fresh ears of corn). Each sample was ensured to have a homogeneous mix of product by a combination of grinding and mixing. Fresh corn was husked, and kernels manually removed with a sharp knife prior to blending to create a baby food consistency with no visible kernels. All samples were stored in plastic reusable storage containers. Wherever possible, plastic or silanized glass materials were used throughout the method. The multi-step procedure used for these samples is outlined in Figure 1. Details for each step (pre-preparation, liquid-liquid extraction, solid phase extraction, evaporation, and analysis) are outlined below. Figure 1: General steps for sample preparation and analysis for aflatoxin B1, B2, G1, and G2 Instrumentation Solvents Purified water Milli-Q Acetonitrile laboratory grade Standard Sigma-Aldrich Aflatoxin mix 4-solution in acetonitrile (2 µg/ml B1 G1; 0.5 µg/ml B2 G2) Samples purchased from local USA suppliers and specialty Asian Markets Almond Milk Peanut Paste Fresh Corn (purchased on the cob, unhusked) Curry Powder SPE SmartPrep Extractor automated SPE cartridge system Supelco Supel Tox AflaZea SPE Cartridge, 6 ml Evaporation XcelVap Automated Evaporation/Concentration System Purified compressed nitrogen HPLC Shimadzu, Nexera XR UHPLC System with UV detection Titan C18 UHPLC Column, 10 cm x 2.1 mm I.D., 1.9 µm particle size Titan C18 HPLC Guard Cartridge, 5 mm x 2.1 mm. I.D., 1.9 µm particle size Titan Guard Cartridge Holder P a ge 3
Sample Preparation Liquid-Liquid Extraction Procedure There is currently no common method for the determination of aflatoxin B1, B2, G1, and G2 in multiple sample matrices. Separate published AOAC methods detail the specific LLE and SPE procedure for each sample matrix. Simplifying the sample preparation steps by providing a single method to use as a roadmap for multiple matrices was developed, starting with the LLE procedure. The matrices chosen for this application varied in physical properties for both viscosity and moisture. The LLE procedure was generalized for multiple matrices by factoring the % water for each matrix into the quantity of water used in the LLE solvent mixture, in order to maintain the same ratio of sample weight to extraction solvent. Figure 2: Sample matrix viscosity and moisture properties The extraction solution for the four matrices was developed by keeping the acetonitrile content fixed, enabling consistent extraction of the aflatoxins from each matrix. Understanding each sample s natural water content allowed for this content to be subtracted from the water content used in the extraction solution, customizing the extraction solution ratio for each sample matrix (Figure 3). Aflatoxin Extraction Solution Formula for Water Volume: Water (ml): 16 ml [(matrix water content/100)*15 g matrix] ACN (ml): 84 ml Figure 3: Aflatoxin extraction solution formula for volume P a ge 4
Each sample matrix was run in triplicate at the appropriate US and EU spike level, with a control sample to account for any aflatoxins B1, B2, G1, and G2 naturally present in the sample when calculating percent recoveries. Into 100-mL glass beakers, 15 g of sample was weighed. For spiked samples, the neat aflatoxin mixture was added (Table 1 and Table 2) and the sample then mixed thoroughly. The sample was allowed to sit undisturbed and uncovered for 2 minutes. Water was then added according to the formula in Figure 3, accounting for the water content of each sample matrix. Samples were mixed well following the addition of water. Into a graduated cylinder, 84 ml acetonitrile was measured. Approximately 4 aliquots of 10 ml of acetonitrile were added to the sample beaker, with stirring in between. The contents of the beaker were transferred to a 500-mL Erlenmeyer flask, with the remaining acetonitrile used for quantitative transfer. The sample was covered with laboratory parafilm and then shaken on a vortex mixer @ 1750 RPM for 15 minutes (Figures 4 7). After extraction, the sample was filtered through Whatman No. 2 filters, and the filtrate was collected (Procedure 1). Procedure 1: Liquid-Liquid Extraction Procedure The aflatoxin neat mixture contained B1, B2, G1, and G2 levels of 2 μg/ml (B1 and G1) and 0.5 μg/ml (B2 and G2). Spiked samples were at the level or below the maximum levels in Table 2. Table 2: European Union and United States Aflatoxins B1, B2, G1, and G2 Spike Levels Foodstuffs Matrix Amount of neat solution for sample spikes (μl) United European States Union spikes spikes Spiked B1 levels (μg/kg) United States spikes European Union spikes Spiked sum of B1, B2, G1, G2 levels (μg/kg) United European States Union spikes spikes Almonds (almond milk) 60 30 8 4 20 10 Corn 60 12 8 1.6 20 4 Groundnuts (peanut paste) 60 12 8 1.6 20 4 Capsicum spp. (chili powder, cayenne, paprika), Piper spp. (pepper), Myristica fragrans (nutmeg), Zingiber officinale (ginger), Curcuma longa (turmeric), Curry (Indian curry spice mix) 60 30 8 4 20 10 P a ge 5
Figure 4: Almond Milk Liquid- Liquid Extraction Figure 5: Fresh Corn Liquid- Liquid Extraction Figure 6: Peanut Paste Liquid- Liquid Extraction Figure 7: Curry Powder Liquid- Liquid Extraction Sample Preparation Automated SmartPrep Extractor SPE Procedure & XcelVap Evaporation Procedure There is currently no common method for the determination of Aflatoxin B1, B2, G1, and G2 in multiple sample matrices. Separate published AOAC methods detail the specific LLE and SPE procedure for each sample matrix. Simplifying the sample preparation steps by providing a single method to use as a roadmap for multiple matrices was developed, starting with the LLE procedure. The SmartPrep Extractor automated the solid phase extraction process. From the extracted filtrate, 2 ml was automatically loaded onto the Supel Tox AflaZea SPE Cartridge (Procedure 2). The SPE process was simply a load and collect function. No condition, wash, or elute steps were required. Procedure 2: Automated SmartPrep Extractor SPE P a ge 6
Pre-Column Derivatization Using the XcelVap Evaporator/Concentrator System Following the automated SPE process, 220 μl of eluent was pipetted into a 20-mL vial. The derivatizing agent (7:1:2 MilliQ water:glacial acetic acid:tetrafluoric acid) was added (300 μl), and the vial was sealed with a cap. The sample was placed into the closed XcelVap water bath heated to 65 C. The sample was allowed to derivatize for 25 minutes prior to cooling to room temperature (cap remained on vial and in the dark with covered foil until room temperature was reached). Derivatized samples are known to have light sensitivity. Extra precaution was taken during the heating and cooling processes to have as little light exposure as possible. To the cooled vial 580 μl MilliQ water was added. Preparation of Standards A seven point standard curve was prepared to bracket samples for analysis on the uhplc system. The standard curve contained μg/kg levels of aflatoxin B1 of: 0, 0.8, 1.6, 3.2, 4, 8, and 12. Calculated for total aflatoxin concentration (B1+B2+G1+G2), this equates to total μg/kg levels of 0, 2, 4, 8, 10, 20, and 30. To prepare these levels, a 0.1 μg/ml aflatoxin B1 stock solution was made using 5 μl of the 2 μg/ml neat solution and 95 μl 16:84 water:acetonitrile solution for a total volume of 100 μl. From this, the creation of the other standards were diluted according to the volumes described in Table 3 to generate a total volume of 1 ml for each standard concentration. Table 3: Aflatoxin Standard Dilution Schema for the Seven Point Calibration Curve Level Conc. B1 (μg/kg) Total conc. (μg/kg) Stock needed (μl) 16:84 water:acn Blank Blank Blank n/a n/a 2 0.8 2 1.2 998.8 3 1.6 4 2.4 997.6 4 3.2 8 4.8 995.2 5 4 10 6 994 6 8 20 12 988 7 12 30 18 982 P a ge 7
All aflatoxin standards followed the same derivatization steps as the samples. Standards were found to be unstable after 24 hours, prompting a new set of standards to be made for every day the sample matrices were tested. HPLC Analysis Conditions Aflatoxin samples were injected onto a Shimadzu Nexera XR uhplc system with fluorescence detection using a 50 µl injection. The new uhplc column technology of 1.9 µm particle size increased sensitivity and peak height, allowing for optimal separation of the four aflatoxin peaks between 7.5 and 13.5 minutes. Retention time consistency was achieved within each run. Variation of retention time was due to the preparation of a new mobile phase from run to run. Run time length of 25 minutes was optimized to elute any matrix effects prior to the next injection. Table 4: HPLC conditions for aflatoxin analysis Flow Rate HPLC Conditions 0.4 ml/min Column Titan C18 UHPLC Column, 10 cm x 2.1 mm I.D., 1.9 μm particle size Guard Column Column Temperature 45 C Mobile Phase A Mobile Phase B Titan C18 HPLC Guard Cartridge, 5 mm x 2.1 mm I.D., 1.9 μm particle size 100% acetonitrile 5% acetonitrile in water Gradient 0 min: 100% B 3.75 min: 85% B 15.5 min: 85% B 18 min: 100% B Injection Volume 50 µl Run Time 25 minutes Wavelength Excitation: 360 nm/emission: 440 nm P a ge 8
Results and Discussions Samples were spiked to their corresponding US and EU advisory limits. Any natural aflatoxins detected in the controls were subtracted from spiked recoveries. Within a run, the samples were bracketed by standards. Linearity of all four aflatoxin standards was calculated for each run, with typical R2 values at 0.995 or higher (Figures 8-11). Figure 8: Aflatoxin B1 Linearity Figure 9: Aflatoxin B2 Linearity Figure 10: Aflatoxin G1 Linearity Figure 11: Aflatoxin G2 Linearity Sample chromatograms representing each of the four matrices, controls, and spike levels are shown in Figures 12 23. Sample recoveries for each matrix at each US and EU spike level were calculated for aflatoxin B1 and for total aflatoxins (Table 5). P a ge 9
Fresh Corn Almond Milk Figure 12: Control Sample Figure 13: 8 µg/kg B1, 20 µg/kg Figure 14: 4 µg/kg B1, 10 µg/kg Figure 15: Control Sample Figure 16: 8 µg/kg B1, 20 µg/kg Figure 17: 1.6 µg/kg B1, 4 µg/kg P a ge 10
Curry Powder Peanut Paste Figure 18: Control Sample Figure 19: 8 µg/kg B1, 20 µg/kg Figure 20: 1.6 µg/kg B1, 4 µg/kg Figure 21: Control Sample Figure 22: 8 µg/kg B1, 20 µg/kg Figure 23: 4 µg/kg B1, 10 µg/kg P a ge 11
Sufficient resolution of aflatoxin peaks enabled effective baseline quantitation with minimal sample matrix interference. The US does not specify acceptance ranges for either recovery or %RSD. Across the four sample matrices, calculated recoveries for aflatoxin B1 and total aflatoxins were 85.6 to 117%. Reproducibility of sample spikes ranged from 1.91 to 5.10%. Calculated recoveries for aflatoxin B1 and total aflatoxins passed EU performance criteria of 70 to 110% for all sample matrices. Across the four sample matrices, calculated recoveries were 73.2 to 110%. Sample reproducibility (%RSD) also passed regulatory criteria for all sample matrices (Table 5), with results ranging from 1.91% to 9.14% across all matrices. Table 5: Calculated Recoveries and %RSD for US and EU Performance Criteria of Aflatoxins in Sample Matrices Automated SmartPrep Extractor SPE Results (n=3) Spiked Recovery Results US Advisory Limits % Recovery and %RSD Values (No Acceptance Ranges Stated) European Advisory Limits % Recovery and % RSD Values (70 110% Recovery & 20% RSD Acceptance) Almond Milk Aflatoxin B1 RSD Total Aflatoxins RSD Aflatoxin B1 RSD Total Aflatoxins RSD (8 µg/kg B1, 20 µg/kg total US; 4 µg/kg B1, 10 µg/kg EU) Fresh Corn (8 µg/kg B1, 20 µg/kg total US; 1.6 µg/kg B1, 4 µg/kg EU) Peanut Paste (8 µg/kg B1, 20 µg/kg total US; 1.6 µg/kg B1, 4 µg/kg EU) Curry Powder (8 µg/kg B1, 20 µg/kg total US; 4 µg/kg B1, 10 µg/kg EU) 93.7 5.10 93.1 4.90 97.0 7.94 97.0 7.96 86.5 4.78 88.0 4.87 81.0 3.83 85.3 1.46 85.6 4.35 86.2 5.05 81.0 9.14 73.2 5.97 117 1.91 97.1 2.61 110 6.44 101 5.83 P a ge 12
Conclusion Data demonstrated that one sample preparation method, accounting for matrix moisture, is a viable and reliable solution for a wide range of matrices varying in viscosity and moisture content. The single sample preparation extraction formula developed for LLE followed by automated SmartPrep Extractor SPE with Supel Tox AflaZea cartridges using four matrices of almond milk, fresh corn, peanut paste, and curry powder generated acceptable recoveries and %RSD values for sample matrix spikes. Derivatization using the XcelVap offered a consistent moist and dark environment, and in combination with the efficiency of the uhplc system and column technology, sharp aflatoxin peaks at the lower EU spike levels were generated and easily quantified. Better resolution of the four aflatoxin peaks from any matrix peaks was obtained for high moisture matrices, with sufficient peak height generated by the gradient to quantitate low moisture matrices, as demonstrated by the low %RSD. Using a single sample preparation method for multiple sample matrices reduces error associated with referencing several methods and preparation procedures. References 1. Joint FAO/WHO Food Standards Programme Codex Committee on Contaminants in Foods. ftp://ftp.fao.org/codex/ meetings/cccf/cccf8/cf08_20e.pdf (accessed September 2014) 2. International Scientific Cooperation Projects (INCO) on Aflatoxin http://www.icrisat.org/aflatoxin/ anamika_effects_aflatoxins.asp (accessed September 2014) 3. US Food and Drug Administration Site. http://www.fda.gov/food/guidanceregulation/guidancedocumentsregulatoryinformation/chemicalcontaminantsmetalsnatu raltoxinspesticides/ucm077969.htm (accessed August 2014) 4. European Union Site. http://eur-lex.europa.eu/legal-content/en/all/?uri=celex:32010r0165 (accessed August 2014) 5. European Union Site. http://eur-lex.europa.eu/legal-content/en/all/?uri=celex:32006r0401 (accessed August 2014) www.horizontechinc.com AN0981410_1 16 Northwestern Drive, Salem, NH 03079 USA Tel: (603) 893-3663 Email: Support-Service@horizontechinc.com