Search

Various liquid chromatographic methods used in the analysis of mycotoxins (zearalenone, trichothecenes and fumonisins) produced by Fusarium species were compared in this work. The results demonstrate the suitability of modern clean-up procedures employing multifunctional MycoSep and immunoaffinity columns although these methods are more expensive than conventional methodologies for clean-up. HPLC with both fluorescence and photodiode array detection is a suitable technique for the analysis of toxic secondary metabolites produced by Fusarium species; different derivatisation strategies have been studied to improve the sensitivity of the technique because of the low concentration of these metabolites in contaminated food. The utility of the proposed methodology was assessed in cereal cultures of various Fusarium strains.

Toxigenic isolates of Fusarium sporotrichioides were tested for the production of type A trichothecenes (T-2 toxin, HT-2 toxin, diacetoxyscirpenol and neosolaniol) when grown on three substrates (maize, rice and wheat) under various conditions of temperature and water activity in the laboratory for 3 weeks. Trichothecenes were determined by high-performance liquid chromatography with fluorescence detection, after derivatisation with coumarin-3-carbonyl chloride. This is the first time this analytical method has been applied to an extensive study of trichothecene accumulation. With minor exceptions, greater trichothecene production occurred when samples were incubated at 20 degrees C and moistened with 35% water (water activity 0.990) although incubation conditions affected the substrates studied in different ways. No correlation between the different pairs of trichothecenes was found except for neosolaniol and diacetoxyscirpenol (r=0.56). Principal component analysis results show that the data points can be grouped in three rough clusters related to cereal type, which points out that the composition of these cereals can influence the production of type A trichothecenes.

Twenty nine isolates of Fusarium spp. (twenty four of them belonging to the Gibberella fujikuroi complex) isolated from banana and corn from different geographical regions were analyzed for their ability to produce fumonisins B1 and B2 and for genetic relatedness using random amplified polymorphic DNA (RAPD) and restriction analysis of PCR amplification products of the 5.8s ribosomal DNA-intervening internal transcribed spacer regions (ITS I-5.8S-ITS II). For RAPD analysis, six of twenty oligonucleotide primers were selected after testing with five Fusarium spp. isolates and used to characterize 24 additional isolates. DNA fragments from the 29 isolates of Fusarium spp., which were approximately 560 bp, were amplified with the universal primers ITS1 and ITS4. The restriction enzymes HaeIII, MboI, HpaII and MspI were useful for distinguishing the isolates. The RAPD analysis permitted to find interspecific differences among the isolates of Fusarium spp., between isolates with low and high capacity of fumonisin production and among isolates from different hosts. The restriction fragment length polymorphism (RFLP-PCR) analysis permitted to distinguish among different species of Fusarium. In combination with morphological analysis, the results of this research may find an application for the diagnosis of unknown Fusarium spp. and, particularly, for the characterization of fumonisin-producing isolates, which may be very useful in the food technology field.

The aim of this work was the optimization of some procedures usually used in the analysis of zearalenone (ZEA) in corn and other cereals by high-performance liquid chromatography (HPLC) with photodiode array and/or fluorescence detection. The comparison of five extraction solvents is presented. Three solid-phase extraction cartridges (C-18, silica, Florisil) and immuno-affinity columns were also compared to obtain the best recovery of the mycotoxin with the minimal presence of co-extractives in the chromatograms. Mixtures of methanol-1% aqueous NaCl (80:20 or 60:40 v/v) were the best extraction solvents. Florisil provided higher recovery of ZEA than C-18, and silica proved unsuitable. The immuno-affinity column was very efficient in cleaning the extracts, but its sample capacity was lower than that of SPE columns due to saturation. The mobile phase (methanol-water 80:20 v/v) gave a low retention time for ZEA (~5 min), high sensitivity and acceptable separation between this mycotoxin and α-zearalenol. The optimized protocol is straightforward, provides high ZEA recoveries in spiked corn (mean 102.4%), has an acceptable sensitivity and has a lack of interference with fluorescence detection (detection limit 4 ng ZEAg[sup -1] corn). The photodiode array detector was useful, except at very low ZEA levels, to confirm the identity of the mycotoxin. The method was applied to search for ZEA accumulation in corn, wheat and rice grains inoculated with selected strains of Fusarium graminearum, F. oxysporum and F. culmorum. [ABSTRACT FROM AUTHOR]