Preparation of Thiol Conjugates of the Mycotoxin Deoxynivalenol and their Occurrence in Nature

Cereal crops (such as oats, wheat, barley, rye, corn etc.) are often contaminated with toxigenic fungi during field or storage conditions, which can consequently lead to food and feed contamination with mycotoxins and to crop losses. One of the most common and worldwide occurring genera of fungi infecting crops is Fusarium, including species that may produce mycotoxins like trichothecenes, fumonisins and zearalenone. Trichothecenes inhibit protein synthesis by binding to the ribosomes. Based on their chemical structure, they are further divided into Type A, B, C and D. The Type B trichothecenes, characterized by the presence of a carbonyl group at the C-8 position, is not the most acutely toxic subgroup, but it is the most prevalent in North America and Europe. The economically most relevant mycotoxin, belonging to the Type B group of trichothecenes, is 4-deoxynivalenol (DON). It contains several structural features that enable plants or animals to biotransform the molecule, thereby decreasing or eliminating its toxicity. These features are an epoxy ring, hydroxyl groups and an α,β-unsaturated carbonyl group. Some of the plant biotransformation products of DON (e.g. deoxynivalenol 3-β-Dglucoside) are well known and characterized. More recent studies used carbon isotope labelled DON in metabolomic studies in wheat and tentatively identified new conjugates such as DON–glutathione (DON–GSH), DON–cysteinylglycine (DON–CysGly) and DON–cysteine (DON–Cys). As the thiol addition may either occur irreversibly on the 12,13-epoxide group of DON or reversibly at the 9,10-conjugated double bond, it was necessary to synthesize these conjugates in order to identify the biotransformation products in real samples. The main focus of this work was to get insight into the reaction between DON and different thiols, including biologically significant compounds as L-cysteine and L-glutathione. The reaction between the model thiol 2-mercaptoethanol and DON was initially used to optimize the reaction conditions and to develop a liquid chromatography–mass spectrometry (LC–MS) method to follow the reaction and to analyze the products. The reactions were tested in a pH range of 7.3–10.7, and with several thiols (2-aminoethane thiol, sodium methanethiolate, sodium 2-mercaptoethanesulfonate, L–cysteine and L–glutathione). The pattern in the LC–MS chromatograms of the reaction between DON and tested thiols was similar – what was shown to be the C-13 (epoxide) conjugate was always the earliest eluting peak, while there were several C- 10 (Michael) adducts that were isomerizing and their ratios changing with time. The reactions occurred fastest under the basic conditions, but also occurred at physiological pH. The reaction was also carried out between 2-mercaptoethanol and trichothecenes that differentiate from DON in a way that is significant to prove the binding position of a thiol, such as T-2 tetraol that has hydroxyl group at C-8 and therefore lacking a conjugated double bond, and deepoxy-DON that contains reduced 12,13-epoxy ring. A DON-13-mercaptoethanol and 10,13-double adduct was purified and characterized using nuclear magnetic resonance (NMR). Once major reaction products were characterized, optimized conditions (basic pH and thiol excess) were applied and used to synthesize and purify epoxide and thermodynamically favored Michael adducts of DON–GSH and DON–Cys conjugates. In order to prepare analytical standards, conjugates were quantified using qNMR (ERETIC2). The effect of the synthesized conjugates (DON-13-mercaptoethanol, DON-10,13-dimercaptoethanol, DON-10-Cys and DON-13-Cys) on cell proliferation and metabolic activity, as well as the expression of proinflammatory cytokines was tested. All of the tested compounds displayed significantly lower toxicity in vitro compared to DON. The reaction mixtures containing DON and two dipeptides, CysGly and γ-GluCys, and Nacetylcysteine (NAC) were also prepared and followed over time for the comparative analysis of real samples and together with the obtained standards were used in the analysis of oats and spring wheat samples from Norway that were naturally contaminated with DON and in an extract of wheat spikelets that were artificially treated with DON. The LC–HRMS analysis showed that the main products of the artificially contaminated wheat were C-10 conjugates of DON and GSH, Cys and CysGly, while the naturally contaminated wheat and oat samples contained predominantly C-13-linked conjugates with GSH, Cys, CysGly and NAC