Trichothecenes: reference materials and method validation

The frequent contamination of food and feed with trichothecene mycotoxins, the high consumption of these products, and the potential risk associated herewith, has led to an increasing public awareness and therefore to the establishment of measures to control trichothecene contamination. The analytical difficulty and the economic importance of controlling trichothecenes in food and feed support the need for certified reference materials (CRMs) and validated methods. They form invaluable tools to ensure comparability and traceability in analytical measurements and are very useful for the implementation of written standards, legislation/regulations and laboratory accreditation. The present paper provides an overview of previous work, current strategies and prospectives for the production of CRMs and validation of analytical methods in the field of trichothecene analysis. Additional information is given on methodological demands, normative frameworks and commonly accepted procedures.

Quantification of deoxynivalenol in wheat using an immunoaffinity column and liquid chromatography

A simple and accurate method to quantify the mycotoxin deoxynivalenol (DON) in wheat is described. The method uses immunoaffinity chromatography for DON isolation and liquid chromatography (LC) for toxin detection and quantification. Wheat samples are extracted in water, filtered twice and applied to an immunoaffinity column. Following a water wash, DON is eluted from the column with methanol and injected onto an LC system with a UV detector for quantification. Test performance was evaluated in terms of antibody specificity, limit of detection, percentage recovery, precision, column capacity, assay linearity and comparison with the GC-electron-capture detection (ECD) method of Tacke and Casper. Specificity of the immunoaffinity column cleanup procedure was confirmed with only DON (>80%) and its 15-C derivatives (40-50%) being recognized by the antibody while 3-C DON derivatives, nivalenol, T-2 and fusarenon-X did not bind. The limit of detection is at least 0.10 microg/g. Percentage recovery for the entire assay range averages 90% with an average relative standard deviation of 8.3%. Naturally contaminated samples showed comparable precision. Column capacity was determined to be 3.3 microg. The assay showed a high degree of linearity (r2=0.999) and an optimum assay range of 0.10 to 10.0 microg/g. Comparative analysis of 28 naturally or artificially contaminated wheat samples using DONtest-HPLC and the GC-ECD method of Tacke and Casper showed that DONtest-HPLC is a statistically significant predictor of the GC-ECD method (r2=0.982).