Analysis of mycotoxins in grains

Brian Kinsella explores a method to prepare samples using polymeric SPE and  LC-MS-MS.

Mycotoxins are naturally occurring secondary metabolites produced by several species of fungi. They pose a health risk to humans and animals due to their harmful biological properties and common occurrence in food and feed. Analysis of mycotoxins is challenging due to their large number and varied physicochemical properties. Additionally, the food samples tested are complex in nature and are often contaminated with several mycotoxins at low concentrations.

This application details a method for the sample preparation of grain-based food via polymeric SPE and analysis by LC-MS/MS. HPLC separation of 16 mycotoxins and three internal standards was successfully conducted in < 16 minutes using a Selectra DA column – 100 × 2.1mm, 3 µm from UCT. This polyaromatic column provides greater retention of the polar trichothecenes as compared to a C18 column. Compounds included are representative of a wide range of mycotoxins, including type A- and B-trichothecenes, ochratoxin A, alternariol, zearalenone, α- & β-zearalanol and aflatoxins. 

2g of sample was hydrated with 2ml of DI water (≥15 min). 10ml of ACN was added to each sample and vortex mixed for 10 minutes followed by centrifugation for 10 minutes (≥3,000 × g, 4°C). The supernatant was transferred to a clean glass tube and evaporated to dryness at 50°C under a gentle stream of nitrogen. 10ml of DI water was added to each sample and vortex mixed.

SPE procedure

Polymeric SPE cartridges (EnviroClean HL DVB 200 mg/6ml cartridge from UCT) were conditioned with 3ml MeOH and 3ml DI water. The sample supernatants were loaded onto each cartridge and allowed to percolate through the cartridges under gravity. The cartridges were then washed sequentially with 3ml of DI water followed by 3ml of 10% MeOH. The cartridges were dried under vacuum for 10 minutes (≥10inHg). Then 3ml of hexane was added to each cartridge and slowly drawn through. The cartridges were dried under vacuum (≥10 inHg) for five minutes. The mycotoxins were eluted from the SPE cartridges using 4ml ACN. The extracts were evaporated to dryness at 40°C under a gentle stream of nitrogen, then reconstituted in 1ml of MeOH:H2O (50:50,v/v).

The instrumentation used for this experiment was a Thermo Scientific Dionex Ultimate 3000 LC system coupled to a Thermo Scientific TSQ Vantage tandem mass spectrometer. Various LC-MS/MS conditions were examined prior to settling on the use of APCI ionisation and MeOH/ammonium formate (10mm) as the mobile phase. This set-up provided the best combination to effectively detect and quantitate all the mycotoxins. Parameters investigated included ESI versus APCI; ACN versus MeOH as the organic solvent, as well as a variety of mobile phase buffers.

Owing to their polarity, type-B trichothecenes typically elute early in chromatographic runs, and are known to be prone to matrix effects in the ion source. To reduce this possibility a Selectra DA column containing a polyaromatic phase was employed that suitably retained the polar trichothecenes. Using this column, the first compound (nivalenol) does not elute until 3.65 minutes. Additionally, the first two minutes of flow was diverted to waste minimising ion source contamination. The separation of the mycotoxins, including α- and β-zearalanol, was achieved within 16 minutes. The use of rapid polarity switching allowed all target analytes to be detected in a single run.

The experiment produced several clear conclusions. Firstly, SPE extraction of the mycotoxins, including the polar trichothecenes, was successfully achieved using an EnviroClean hypercrosslinked divinylbenzene SPE cartridge. The SPE wash step was optimised to remove matrix interferences without analyte loss. Next, separation of the mycotoxins, including baseline resolution of α- and β-zearalanol, was achieved within 16 min on a polyaromatic Selectra DA HPLC column. APCI ionisation was selected as it provided the best overall results, including better signal response for problematic trichothecenes. Using rapid polarity switching allows all target analytes to be detected in a single run. Overall, good accuracy and precision data were obtained. Finally, for best results, matrix-matched calibration curves and isotopically labelled internal standards are recommended.

For more information visit www.scientistlive.com/eurolab

Brian Kinsella is an application chemist at UCT, in Philadelphia, USA.

Recent Issues