Corn-based human foodstuffs from retail outlets in five countries were analyzed for fumonisin Bi (FBi) and fumonisin B2 (FB2). The highest mean concentrations occurred in two Egyptian samples (2380 ng/g FBi and 595 ng/g FB2). Only one of four Peruvian samples contained 660 ng/g FBi and 68 ng/g FB2, while only one of two Canadian samples contained a detectable level of FBi. The 16 cornmeal (CM) and 10 corn grits (CG) products from the United States contained mean concentrations of 1048 ng/g FBi and 298 ng/g FB2 and 601 ng/g FBi and 375 ng/g FB2, respectively, while the mean concentrations in 52 CM and 18 CG samples from South Africa were 138 ng/g FBi and 83 ng/g FB2 and 125 ng/g FBi and 85 ng/g FB2, respectively. Only 1 of 10 cornflakes/lime-treated samples contained a low level of FBi. Of several samples obtained from a high esophageal cancer (EC) risk area in the United States 7/7 contained FBi ng/g) and 6/7 FB2 (70-460 ng/g).

As part of a comprehensive risk assessment study for fumonisins, reliable data on exposure of populations to these dietary toxins must be obtained. To assess the extent of worldwide exposure, the published literature on the contamination of food and feed supplies has been reviewed and supplemented with unpublished material from various international sources. Fumonisin contamination of corn and corn-based products occurs in many countries. Animal mycotoxicoses such as equine leukoencephalomalacia and porcine pulmonary edema are caused by heavily contaminated animal feeds. For example, as much as 330 μg/g fumonisin B1 (FB1) has been found in swine feed. Although commercially available refined corn products for human consumption are generally contaminated at levels below 1 μg/g FB1, individual products in certain countries can reach far higher levels. Health risks associated with consumption of these products depend on the extent to which they are consumed in a varied diet. Home-grown corn in certain rural areas, where it also constitutes the staple diet, can be contaminated at >100 μg/g. Consumption of corn contaminated at these high levels has: been associated with a high incidence of esophageal cancer in these areas.

The natural occurrence of the carcinogenic mycotoxin ochratoxin A (OTA) in wines sold in local retail outlets in South Africa and Italy was investigated by HPLC analysis with fluorescence detection following cleanup by immunoaffinity column. All 24 local South African wines tested (15 white and 9 red) were found to contain detectable levels (>0.01 microg/L) of OTA, with a mean of 0.16 microg/L in the white wines and a mean of 0.24 microg/L in the red wines. Results were subsequently confirmed by LC-MS analysis using positive ion electrospray ionization with collision-induced dissociation of the protonated molecular ion [M + H](+) at m/z 404 and selected reaction monitoring of the resultant product ions [M + H - H(2)O - CO](+) at m/z 358 and [M + H - H(2)O](+) at m/z 386. Comparison with the fluorescence method gave a significant correlation (r = 0.87; p < 0.01). Although OTA contamination was present in all of the South African samples analyzed, levels were well below the suggested European Union limit of 0.5 microg/kg. The highest level found in a locally purchased wine was 0.39 microg/L in a blend of local and imported Spanish red wine. Of the eight Italian wines analyzed, only two red wines were contaminated above the suggested maximum level.

No abstract

A liquid chromatographic (LC) method for simultaneous determination of fumonisins B1 (FB1), B2 (FB2), and B3 (FB3) in corn was subjected to a collaborative study involving 12 participants from 10 countries, in which the accuracy and reproducibility characteristics of the method were established. Mean analyte recoveries from corn ranged from 81.1 to 84.2% for FB1 (at a spiking range of 500 to 8000 ng/g), from 75.9 to 81.9% for FB2 (at a spiking range of 200 to 3200 ng/g), and from 75.8 to 86.8% for FB3 (at a spiking range of 100 to 1600 ng/g). The valid data were statistically evaluated after exclusion of outliers. Relative standard deviations for within-laboratory repeatability ranged from 5.8 to 13.2% for FB1, from 7.2 to 17.5% for FB2, and from 8.0 to 17.2% for FB3. Relative standard deviations for between-laboratory reproducibility varied from 13.9 to 22.2% for FB1, from 15.8 to 26.7% for FB2, and from 19.5 to 24.9% for FB3. HORRAT ratios, calculated for the individual toxin analogues, ranged from 0.75 to 1.73. The LC method for determination of fumonisins B1, B2, and B3 in corn (at concentrations of 800–12800 ng total fumonisins/g) has been adopted by AOAC INTERNATIONAL.

No abstract

Two biological species of Gibberella fujikuroi (A and F mating populations) share the Fusarium moniliforme anamorph. Twenty strains of each of these biological species were tested for the ability to produce fumonisins B 1 , B 2 , and B 3 and moniliformin and for toxicity to 1-day-old ducklings. Most of the members of the A mating population (19 of 20 strains) produced more than 60 g of total fumonisins per g, whereas only 3 of 20 members of the F mating population produced more than trace levels of these toxins and none produced more than 40 g of total fumonisins per g. In addition, only 3 of 20 members of the A mating population produced more than 1 g of moniliformin per g (and none produced more than 175 g/g), while all 20 strains of the F mating population produced more than 85 g of this toxin per g and 1 strain produced 10,345 g/g. The duckling toxicity profiles of the strains of the two mating populations were similar, however, and the level of either toxin by itself was not strongly correlated with duckling toxicity. On the basis of our data we think that it is likely that the members of both of these mating populations produce additional toxins that have yet to be chemically identified. These toxins may act singly or synergistically with other compounds to induce the observed duckling toxicity.

Traditional medicine is an important aspect of healthcare delivery in South Africa and is used by at least 70% of the country’s population. The trade in medicinal plants is a multi-million rand business which is a major driver for rural economies. However, the conditions in which these plant products are transported and stored make them prone to fungal contamination which results in economic losses to the traders and pose potential health hazards to consumers. Of major concern is the possible presence of toxigenic fungi and mycotoxins. This study assessed fungal and mycotoxin contamination of African herbal products sold in Cape Town and Tshwane (formerly Pretoria) in South Africa. Of the 16 samples analyzed, 15 were contaminated with at least one of these three fungal genera: Aspergillus, Fusarium, and Penicillium. Fumonisin B1 was present in 13 of the samples in quantities ranging from 14 to 139 μg/kg (detection limit 5 μg/kg). None of the samples was contaminated with aflatoxigenic fungi or aflatoxin (detection limit 0.5 μg/kg). This is the first study to report on mycological and mycotoxin contamination of commercial traditional African medicines in South Africa. There is a need to expand the study to other urban centers to gain enough insight into this problem and then to intervene with measures that can protect the public from potential harm.