Development of ion trap HPLC-MS detection methods for the determination of prevalent mycotoxins in grain and application to real samples

The prevalent trichothecenes deoxynivalenol, nivalenol and HT-2 as well as zearalenone, zearalenol, ochratoxin A and citrinin are typical widespread mycotoxins in temperate climate zones mainly produced by the genus Fusarium, Aspergillus and Penicillium. Consequently, they co-occur frequently in the same agricultural commodities, especially in grain. Their simultaneous determination is highly desirable for screening method devices, due to the continuous need to protect the health of humans and animals by limiting their exposure to mycotoxins.
A rapid quantitative method for the simultaneous determination of the above mentioned mycotoxins was developed using HPLC coupled to an ion trap mass spectrometer equipped with an atmospheric pressure chemical ionization (APCI) source. It has been specifically developed for the rapid primary screening of cereal samples. The fast HPLC separation and APCI(+) detection which allows to detect mycotoxins of highly different polarity (lipophilic to ionic), make this method suitable for high sample throughput and routine analysis. Additionally, a fast and easy sample cleanup method was developed applying C18 and SAX SPE cartridges for the elution of mycotoxins with a wide range of polarity. Recoveries were in the range of 46-106 % and the limits of detection were between 25 and 125 ppb for different mycotoxins. Wheat and corn was chosen for the analysis of real samples.
Fumonisins belong to a substance class of mycotoxins, which was discovered recently (1988) and consists of long-chain methyl branched amino alcohols. Even nowadays the analysis of fumonisins is a big challenge not only because of the risk of the already known fumonisins in food, but also due to the fact that they form degradation and reaction products during food processing, which are not sufficiently characterized yet.
A method was developed for the quantification of fumonisins in corn based on reversed phase HPLC/MS and multiple MS. This lead for FB1 to a improved LOD of 0.6 ppb and a LOQ of 8 ppb, which is a factor of five better compared to the former fluorescence method. Furthermore, a reduced chromatographic run time of up to 50 % could be applied successfully even for the analysis of real samples. Various corn products were examined for fumonisin contaminations. Products from the Swiss market were all below the tolerance level of 1 ppm.
To overcome matrix effects and guarantee to a reliable quantification of analytes an adequate ISTD was searched. Several approaches using uncomplicated and easy to perform synthesis was carried out adding protection groups to fumonisins. Unfortunately, the obtained products had either a too low stability under acidic conditions or the synthesis was not complete.
A total of 48 winter wheat samples from 1998 and 2000 were investigated by HPLC-MS for the presence of the trichothecenes nivalenol (NIV), deoxynivalenol (DON), neosolaniol (NEO), fusarenon-X (F-X), diacetoxyscirpenol (DAS), 3-acetyldeoxynivalenol (3-ADON), 15-acetyldeoxynivalenol (15-ADON), HT-2-toxin (HT-2) and T-2-toxin (T-2). Robustness of the applied technique under routine conditions was good and corresponding quality control information is given. The wheat samples were cultivated on a long-term field trial in
Switzerland with three different farming systems (biodynamic, bioorganic and conventional). DON was detected in 100%, NIV in 67% and HT-2 in 25% of all wheat samples. Only 48% of the detectable DON concentrations were above the limit of quantification (LOQ, 50 μg/kg). The range varied between 10-206 μg/kg. NIV and HT-2 were detected at concentrations around or below their LOQ (NIV: 100 μg/kg; HT-2: 10 μg/kg).
Statistically significant differences between the three farming systems could not be found although there were some indications that wheat from organic farming had lower DON contaminations than that from conventional farming. However, the results indicated that the presence and levels of (selected) trichothecenes may not be suitable as marker to differentiate between the three given farming systems. Furthermore, the results showed that even in years with low or nearly no fungal infection (warm and dry summers) low contaminations of DON in wheat are inevitable.