Glycidyl fatty acid esters in food by LC-MS/MS: method development.

An improved method based on liquid chromatography-tandem mass spectrometry (LC-MS/MS) for the analysis of glycidyl fatty acid esters in oils was developed. The method incorporates stable isotope dilution analysis (SIDA) for quantifying the five target analytes: glycidyl esters of palmitic (C16:0), stearic (C18:0), oleic (C18:1), linoleic (C18:2) and linolenic acid (C18:3). For the analysis, 10 mg sample of edible oil or fat is dissolved in acetone, spiked with deuterium labelled analogs of glycidyl esters and purified by a two-step chromatography on C18 and normal silica solid phase extraction (SPE) cartridges using methanol and 5% ethyl acetate in hexane, respectively. If the concentration of analytes is expected to be below 0.5 mg/kg, 0.5 g sample of oil is pre-concentrated first using a silica column. The dried final extract is re-dissolved in 250 µL of a mixture of methanol/isopropanol (1:1, v/v), 15 µL is injected on the analytical C18 LC column and analytes are eluted with 100% methanol. Detection of target glycidyl fatty acid esters is accomplished by LC-MS/MS using positive ion atmospheric pressure chemical ionization operating in Multiple Reaction Monitoring mode monitoring 2 ion transitions for each analyte. The method was tested on replicates of a virgin olive oil which was free of glycidyl esters. The method detection limit was calculated to be in the range of 70-150 µg/kg for each analyte using 10 mg sample and 1-3 µg/kg using 0.5 g sample of oil. Average recoveries of 5 glycidyl esters spiked at 10, 1 and 0.1 mg/kg were in the range 84% to 108%. The major advantage of our method is use of SIDA for all analytes using commercially available internal standards and detection limits that are lower by a factor of 5-10 from published methods when 0.5 g sample of oil is used. Additionally, MS/MS mass chromatograms offer greater specificity than liquid chromatography-mass spectrometry operated in selected ion monitoring mode. The method will be applied to the survey of glycidyl fatty acid esters in food products on the Canadian market.

Determination of melamine, ammeline, ammelide and cyanuric acid in infant formula purchased in Canada by liquid chromatography-tandem mass spectrometry.

A liquid chromatography-tandem mass spectrometry-based isotope dilution method was developed for the analysis of the triazine compounds melamine (MEL), ammeline (AMN), ammelide (AMD) and cyanuric acid (CYA) in infant formula samples purchased in Canada in 2008 for the purpose of a combined exposure and risk assessment. Infant formula samples were extracted with 1:1 acetonitrile-water, cleaned up on disposable ion-exchange solid-phase extraction cartridges, and analysed by ultra-high-performance liquid chromatography-tandem mass spectrometry. MEL and CYA were detected in almost all infant formula products: the highest concentrations observed were 0.32 mg kg(-1) MEL and 0.45 mg kg(-1) CYA. Samples that were relatively high in MEL in this survey tended to be low in CYA, and vice versa. Concentrations of AMN and AMD were very low in all samples. The total of MEL-related compounds (sum of all four analytes) in all samples was below the interim standard of 0.5 mg kg(-1) for infant formula products established by Health Canada.

Surveys of rice sold in Canada for aflatoxins, ochratoxin A and fumonisins.

Approximately 200 samples of rice (including white, brown, red, black, basmati and jasmine, as well as wild rice) from several different countries, including the United States, Canada, Pakistan, India and Thailand, were analysed for aflatoxins, ochratoxin A (OTA) and fumonisins by separate liquid chromatographic methods in two different years. The mean concentrations for aflatoxin B(1) (AFB(1)) were 0.19 and 0.17 ng g(-1) with respective positive incidences of 56% and 43% (≥ the limit of detection (LOD) of 0.002 ng g(-1)). Twenty-three samples analysed in the second year also contained aflatoxin B(2) (AFB(2)) at levels ≥LOD of 0.002 ng g(-1). The five most contaminated samples in each year contained 1.44-7.14 ng AFB(1) g(-1) (year 1) and 1.45-3.48 ng AFB(1) g(-1) (year 2); they were mostly basmati rice from India and Pakistan and black and red rice from Thailand. The average concentrations of ochratoxin A (OTA) were 0.05 and 0.005 ng g(-1) in year 1 and year 2, respectively; incidences of samples containing ≥LOD of 0.05 ng g(-1) were 43% and 1%, respectively, in the 2 years. All positive OTA results were confirmed by LC-MS/MS. For fumonisins, concentrations of fumonisin B(1) (FB(1)) averaged 4.5 ng g(-1) in 15 positive samples (≥0.7 ng g(-1)) from year 1 (n = 99); fumonisin B(2) (FB(2)) and fumonisin B(3) (FB(3)) were also present (≥1 ng g(-1)). In the second year there was only one positive sample (14 ng g(-1) FB(1)) out of 100 analysed. All positive FB(1) results were confirmed by LC-MS/MS.

Determination of perchlorate in infant formula by isotope dilution ion chromatography/tandem mass spectrometry.

A sensitive and selective isotope dilution ion chromatography/tandem mass spectrometry (ID IC-MS/MS) method was developed and validated for the determination of perchlorate in infant formula. The perchlorate was extracted from infant formula by using 20 ml of methanol and 5 ml of 1% acetic acid. All samples were spiked with (18)O(4) isotope-labelled perchlorate internal standard prior to extraction. After purification on a graphitised carbon solid-phase extraction column, the extracts were injected into an ion chromatography system equipped with an Ionpac AS20 column for separation of perchlorate from other anions. The presence of perchlorate in samples was quantified by isotope dilution mass spectrometry. Analysis of both perchlorate and its isotope-labelled internal standard was carried out on a Waters Quattro Ultima triple quadrupole mass spectrometer operating in a multiple reaction monitoring (MRM) negative ionisation mode. The method was validated for linearity and range, accuracy, precision, sensitivity, and matrix effects. The limit of quantification (LOQ) was 0.4 µg l(-1) for liquid infant formula and 0.95 µg kg(-1) for powdered infant formula. The recovery ranged from 94% to 110% with an average of 98%. This method was used to analyse 39 infant formula, and perchlorate concentrations ranging from

Brominated flame retardants in Canadian chicken egg yolks.

Chicken eggs categorised as conventional, omega-3 enriched, free range and organic were collected at grading stations in three regions of Canada between 2005 and 2006. Free run eggs, which were only available for collection from two regions, were also sampled during this time frame. Egg yolks from each of these egg types (n = 162) were analysed to determine brominated flame retardant levels, specifically polybrominated diphenyl ethers (PBDEs) and hexabromocyclododecane (HBCD). PBDEs were detected in 100% of the 162 samples tested, while HBCD was observed in 85% of the egg yolks. Total PBDE concentrations in egg yolks ranged from 0.018 to 20.9 ng g(-1) lipid (median = 3.03 ng g(-1) lipid), with PBDE 209 identified as being the major contributor to ΣPBDE concentrations. In addition to PBDE 209, PBDE 99, 47, 100, 183 and 153 were important contributors to ΣPBDE concentrations. Total HBCD concentrations ranged from below the limit of detection to a maximum concentration of 71.9 ng g(-1) lipid (median = 0.053 ng g(-1) lipid). The α-isomer was the dominant contributor to ΣHBCD levels in Canadian egg yolks and was the most frequently detected HBCD isomer. ΣPBDE levels exhibited large differences in variability between combinations of region and type. ΣHBCD concentrations were not significantly different among regions, although differences were observed between the different types of egg yolks analysed in the present study.

Antioxidant capacity of potato chips and snapshot trends in acrylamide content in potato chips and cereals on the Canadian market.

The concentration of acrylamide was measured in selected varieties of five brands of potato chips and breakfast cereals over a 5-year period. Most of the products were purchased in one locality in Canada. Samples were analysed by an isotope dilution ((13)C(3)) acrylamide method. They were extracted with water, partitioned with dichloromethane, filtered through a 5 kDa centrifuge filter, cleaned-up on HLB Oasis polymeric and Accucat mixed mode anion and cation exchange SPE columns, and analysed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The acrylamide concentration in potato chips varied from 106 to 4630 ng g(-1), while values in cereals varied from 50 to 347 ng g(-1). Wide variations were observed between brands, within brands over time, and between lots of the same brand. A subset of potato chip samples was analysed for in vitro antioxidant activity. No relationship was found between antioxidative capacity of potato chips and their acrylamide content.

Formation of iodoacetic acids during cooking: interaction of iodized table salt with chlorinated drinking water.

Iodoacetic and chloroiodoacetic acids were formed when municipal chlorinated tap water was allowed to react with iodized (with potassium iodide) table salt or with potassium iodide itself. Iodoacetic acid was recently shown to be a potent cytotoxic and genotoxic agent. For analysis, samples were extracted with t-amyl methyl ether and converted to the corresponding methyl esters using methanol and sulfuric acid. The concentration of iodoacetic acid was determined by gas chromatography-mass spectrometry (GC-MS) using an authentic standard. The identities of iodoacetic and chloroiodoacetic acids were further confirmed by gas chromatography-high-resolution mass spectrometry (GC-HRMS). Certain influences of sodium hypochlorite and humic acid as well as the concentration of potassium iodide on the yields of these acids were investigated. The concentration of iodoacetic acid in tap water samples boiled with 2 g l-1 of iodized table salt was found to be in the 1.5 microg l-1 range, whilst the concentration of chloroiodoacetic acid was estimated to be three to five times lower.

Semicarbazide in Canadian bakery products.

Levels of semicarbazide were determined by liquid chromatography-tandem mass spectrometry using isotope dilution ((13)C(15)N(2)-semicarbazide) methodology, and they were measured, after hydrolysis in 0.125 M hydrochloric acid and derivatization with 2-nitrobenzaldehyde, as a sum of free and bound semicarbazide. Levels of semicarbazide in 11 bakery products, which were sampled at three time intervals from the same source, varied from not detected (<1ng g(-1)) to 560 ng g(-1). In some instances, concentrations of semicarbazide varied between batches of the same product, at times more than tenfold, suggesting that the addition of azodicarbonamide to the same product is not standardized in many baking establishments.

Analysis of wines, grape juices and cranberry juices forAlternaria toxins.

Sixty six samples of red and white wine from Ontario (VQA), British Columbia (VQA), Québec ("vins artisanaux"), imported wines (from Italy, South America and USA) and Canadian and US grape and cranberry juices were analysed for theAlternaria mycotoxins alternariol (AOH) and alternariol monomethyl ether (AME). After cleanup on aminopropyl SPE columns, AOH and AME were initially determined by reversed phase LC with UV detection. Positive sample extracts were re-analysed by LC-tandem negative ion electrospray mass spectrometry (MS/MS) in multiple reaction mode. Overall mean method recoveries measured by LC-UV were 93% for AOH and 81% for AME. Limits of detection in wine (and juice) by LC-UV for AOH were 0.8 (0.4) ng/ml and for AME were 0.5 (0.4) ng/ml; they were below 0.01 ng/ml by LC-MS/MS. As determined by LC-MS/MS, AOH was found in 13/17 Canadian red wines at levels of 0.03 to 5.02 ng/ml and in 7/7 imported red wines at 0.27-19.4 ng/ml, usually accompanied by lower concentrations of AME. Red grape juices (5 positive/10 samples) contained only sub ng/ml levels of AOH or AME except for one sample (39 ng AME/ml). White wines (3/23 samples), white grape juices (0/4 samples) and cranberry juices (1/5 samples) contained little AOH/AME (≤1.5 ng/ml).

Development and validation of a headspace method for determination of furan in food.

In the past furan had been found to form in foods during thermal processing. These findings and a recent classification of furan as a possible human carcinogen prompted us to develop a simple isotope dilution method for its determination in foods. We also investigated effects of furan volatility, sample matrix and partitioning of furan between water and fat constituents of sample on the analytical determination of furan. The method is based on headspace sampling of a 2 ml vial containing 1 g of sample. For analysis, samples were spiked with d(4)-furan, homogenized in a blender at 0 degree C, with water if required, and sub-sampled to vials containing sodium sulphate. After equilibration at 30 degrees C, 50 microl of headspace was injected into the split/splitless injection port of a GC/MS (EI, SIM). The method is linear in the 0.4-1000 ng/g range with a limit of detection of 0.1 ng/g.

Analysis of heat-processed corn foods for fumonisins and bound fumonisins.

Thirty retail samples of heat-processed corn foods, i.e. corn flakes, corn-based breakfast cereals, tortilla chips and corn chips, were analysed for fumonisins--fumonisin B1 (FB1), fumonisin B2 (FB2) and hydrolysed FB1 (HFB1)--as well as for protein- and total-bound FB1. Bound (hidden) fumonisins cannot be detected by conventional analysis. Improved methods for the determination of bound FB1 were developed. The protein-bound FB1 was extracted with 1% sodium dodecylsulfate (SDS) solution. The SDS, which interfered with high-performance liquid chromatography (HPLC) analysis, was then separated from protein-bound FB1 by complexing with methylene blue followed by solvent extraction and hydrolysis with 2 N KOH. To measure total-bound FB1, the sample itself was hydrolysed with KOH. In both cases, clean-up was accomplished on an OASIS polymeric solid-phase extraction column and the bound fumonisins were determined by HPLC measurement of HFB1. Fourteen of 15 samples of corn flakes and other corn-based breakfast cereals analysed contained detectable levels of FB1 with a mean in positive samples of 67ng g(-1) (13-237 ng g(-1)). Two samples also had detectable levels of FB2 (21-23ng g(-1)). Bound FB1 was found in all samples; the mean protein-bound FB1 measured was 58 ng g(-1) (22-176 ng g(-1)) and the mean total-bound FB1 measured was 106 ng g(-1) (28-418 ng g(-1)), reported as FB1 equivalents after correction for recoveries of HFB1. There was an average of about 1.3 times more FB1 in the bound form compared with extractable FB1, and this was about twice as much as protein-bound FB1. Seven of the 15 samples of alkali-processed corn-based foods, such as tortilla chips and corn chips, contained FB1 and three contained HFB1 with means in measurable positive samples of 78 (48-134) and 29 (13-47) ng g(-1), respectively. Five of these alkali-processed corn foods contained bound FB1; the mean measurable protein-bound FB1 was 42 ng g(-1) (39-46 ng g(-1)) and the mean measurable total-bound FB1 was 100 ng g(-1) (54-209 ng g(-1)). HFB1 derived from bound FB1 in selected samples was confirmed by HPLC with mass spectrometry (MS).

Hidden fumonisin in corn flakes.

Twenty-five samples of retail corn flakes (from 15 lots) were analysed for fumonisin B(1) (FB(1)) and fumonisin B(2) (FB(2)). They were detected in 22 and 12 samples, respectively, at respective mean concentrations 68 and 8 ng g(-1). Samples were extracted with methanol-acetonitrile-water (25:25:50) and there was an excellent correlation for FB(1) between results obtained with C(18) clean-up and those obtained with the immunoaffinity column (IAC) clean-up. After extraction of the corn flakes' residue with 1% sodium dodecyl sulphate (SDS) solution and hydrolysis with 2 N potassium hydroxide, hidden (protein bound) fumonisin was determined as HFB(1), which was found in residues from all the corn flakes samples, even those containing no detectable FB(1); the average concentration of HFB(1) was 101 ng g(-1), equivalent to 180 ng FB(1) g(-1). Thus, our results showed an average of 2.6 times more FB(1) present in bound form as was determined by conventional analysis. We found a correlation coefficient of -0.5034 for a logarithmic relationship between the FB(1) (C(18) clean-up) and HFB(1) concentrations The highest concentration of HFB(1) formed was 288 ng g(-1) from a sample containing only 12-15 ng FB(1) g(-1), while the lowest concentration of HFB(1) was 26 ng g(-1) from a sample with 152-155 ng FB(1) g(-1). This low degree of correlation should be taken into account by food safety authorities in estimates of human exposure to protein bound fumonisin.

Survey of Canadian retail coffees for ochratoxin A.

One-hundred and one specimens of coffee were gathered from retail outlets across Canada and analysed for ochratoxin A. Seventy-one specimens were roasted beans or roasted ground coffee, and 30 were instant (or 'soluble') coffees. All samples were extracted with methanol-sodium bicarbonate. The extracts were cleaned up either by immunoaffinity column chromatography or by a combination of solid-phase extraction and immunoaffinity column chromatography. Ochratoxin A was quantified by liquid chromatography (LC) with fluorescence detection. The minimum quantifiable level was 0.1 ng g(-1). Ochratoxin A was present, above the minimum quantifiable level, in 42 (59%) of 71 beans and ground coffee and in 20 (67%) of 30 instant coffees. The mean ochratoxin A level in the positive samples of beans and ground coffee was 0.6 ng g(-1), and the mean level in the positive samples of instant coffee was 1.1 ng g(-1).