Sampling and sample preparation

Analytical procedures for mycotoxin determination ordinarily commence with sampling and sample comminution to obtain a suitable size that would allow the extraction of analytes and subsequent measurements. The aim of sampling is to collect representative sample(s), e.g. from a lot of an agricultural product, of a workable size in the laboratory (Köppen et al., 2010). Due to the heterogeneous distribution of mycotoxins in agricultural commodities, adequate sampling is a very crucial step for reliable analytical determination. Generally, the most effective approaches to reduce the variability of mycotoxin test procedures are to increase the laboratory sample size, the degree of sample comminution and the number of aliquots quantified (Whitaker, 2006). The methods of sampling and analysis for official control of the levels of mycotoxins in foodstuffs are laid down in Commission Regulation (EC) No 401/2006 (EC, 2006a). Strategies applied for sampling of parent mycotoxins are also suitable for conjugated mycotoxins (EFSA, 2014). For research purposes, sampling approaches may have to be considerably adjusted in order to embrace the experimental hypothesis and conditions (sample amount, type, chemical stability of analytes, etc.). An additional complication in certain metabolomics studies is that during sampling, immediate quenching of all

metabolic processes should be performed without alteration of the metabolic state (Kluger et al., 2015b). Storage conditions must also not permit alterations of the metabolic state or cause changes to metabolite species and levels.

Sample preparation is the next key step for a successful protocol in the chemical analysis of mycotoxins. The majority of LC–MS methods rely on the extraction of analytes from the biological matrices into a liquid phase and optional clean-up to remove undesirable matrix components (Turner et al., 2009). The extraction methods applied depend on the chemistry of the target analytes and on the type of matrix for which extraction is performed. Because of the wide range of physico-chemical properties of mycotoxins, particular consideration needs to be given to the extraction efficiency of the solvent mixture. Conjugated mycotoxin derivatives are usually more polar than their precursors. Therefore, extraction schemes should facilitate the removal of both forms from sample matrices with comparable recoveries. Several different extraction procedures are used, of which the oldest, but still frequently used, is solvent extraction (Kralj Cigić and Prosen, 2009). Typically, the extraction of polar analytes is favoured by the presence of water in the extraction solution mixture, whilst the extraction of hydrophobic mycotoxins is enhanced by organic solvents. The most widely used extraction solvents are mixtures of methanol:water and acetonitrile:water in varying ratios, regularly containing modifiers such as acids or bases. (Meneely et al., 2011). According to the literature, the extraction of DON, NIV, HT2, T2 and ZEN, as well as many of their derivatives, has been successfully achieved using various ratios of acetonitrile and water, acidified with formic or acetic acid (Zachariasova et al., 2010; De Boevre et al., 2012; Malachová et al., 2014). In a metabolomics study, Cajka et al. (2014) accomplished optimal extraction of polar/medium polar barley metabolites employing either acetonitrile:water (84:16, v/v) or a mixture of methanol:water (50:50, v/v). The authors eventually chose the latter solution due to its broadest representation of metabolites based on polarity. Other extraction techniques that have been successfully incorporated in mycotoxin analysis include accelerated solvent extraction, pressurised liquid extraction, microwave-assisted extraction and supercritical fluid extraction (Krska et al., 2008).

For further purification and analyte enrichment, liquid extracts can be submitted to solid phase extraction (SPE), specifically tailored multifunctional columns (e.g. MycoSep® _{push-through clean-up columns), immunoaffinity columns or modifications of} the “quick, easy, cheap, effective, rugged and safe” (QuEChERS) technique (Malachová et al., 2017). The purpose of sample clean-up is to enhance the signal-to-noise ratio and minimise interference by matrix components that may cause analyte signal suppression or enhancement (SSE; i.e. matrix effects) (Trufelli et al., 2011). The majority of sample preparation techniques used in mycotoxin analyses have been developed for the determination of parent mycotoxins, and more so the regulated ones (see Section 2.7). As a consequence, they may not be suitable for the determination of modified

mycotoxins. Vendl et al. (2009), evaluated the suitability of several clean-up strategies (C18-SPE, primary and secondary amines, MycoSep and immunoaffinity columns) for the determination of DON, ZEN and their major modified metabolites in cereals and cereal- derived products by liquid chromatography–tandem mass spectrometry (LC–MS/MS). Overall, none of the tested clean-up methods proved suitable to accommodate a wide range of polarities and the final method thus used no sample clean-up.

The limited availability of analytical standards for modified mycotoxins has led to the development of indirect methods as an attractive alternative to target analysis. Such approaches are based on chemical or enzymatic treatments that cleave the conjugated moieties and release the parent mycotoxin. The supposed advantage of these techniques is that measurements of all modified forms present within a sample can be performed without the need for analytical standards, because the modified mycotoxin concentration is estimated as the total precursor analyte content after subtracting the free precursor content. However, a critical assessment was conducted of three indirect methods for total DON determination, and none was found able to de-conjugate the parent mycotoxin (DON) from the modified forms DON3Glc, 3Ac-DON and 15Ac-DON (Malachová et al., 2015). The authors concluded that use of acidic or alkaline hydrolytic process was not suitable for the indirect determination of DON and efforts should be directed towards methods relying on enzymatic hydrolysis. Ideally, accurate determination of modified forms should be based on analytical standards.