Atmospheric oxygen and other conditions affecting the production of cereulide by Bacillus cereus in food.

Factors influencing the production of cereulide, the emetic toxin of Bacillus cereus in food and laboratory media were investigated, using liquid chromatography-ion trap mass spectrometry and sperm motility inhibition bioassay for detection and quantitation. Oxygen was essential for production of the emetic toxin by B. cereus. When beans, rice or tryptic soy broth were inoculated with cereulide producing strains B203, B116 (recent food isolates) or the strain F-4810/72, high amounts (2 to 7 microg ml(-1) or g(-1) wet wt) of cereulide accumulated during 4-day storage at room temperature. In parallel cultures and foods, stored under nitrogen atmosphere (> 99.5% N2), less than 0.05 microg of cereulide ml(-1) or g(-1) wet wt accumulated. The outcome of the bioassay matched that of the chemical assay, with no indication of interference by substances in the rice or beans. Boiling for 20 to 30 min did not inactivate cereulide or cereulide producing strains in rice or the beans. Adding l-leucine and l-valine (0.3 g l(-1)) stimulated cereulide production 10- to 20-fold in R2A and in rice water agar. When the B. cereus strains were grown on agar media under permissive conditions (air, room temperature), cereulide was produced overnight with little or no increase when the incubation was extended to 4 days. In broth culture, the production of cereulide started later than 16-24 h. Anoxic storage prevented cereulide production also when the amino acids had been supplied. Packaging with modified atmosphere low in oxygen may thus be used to reduce the risk of cereulide formation during storage of food.

In vitro assay for human toxicity of cereulide, the emetic mitochondrial toxin produced by food poisoning Bacillus cereus.

The in vitro boar spermatozoon test was compared with the LC ion trap MS analysis for measuring the cereulide content of a pasta dish, implemented in serious emetic food poisoning caused by Bacillus cereus. Both assays showed that the poisonous food contained approximately 1.6 microg of cereulide g(-1) implying the toxic dose in human as < or =8 microg kg(-1) body weight. The threshold concentration of cereulide provoking visible mitochondrial damage in boar sperm exposed in vitro was 2 ng of cereulide ml(-1) of extended boar sperm. The same threshold value was found for cereulide extracted from the food and from the cultured bacteria. This shows that other constituents of the food did not enhance or mask the effects of cereulide. Exposure of four human cell lines (HeLa, Caco-2, Calu-3 and Paju) to cereulide showed that the threshold concentration for the loss of mitochondrial membrane potential in human cells was similar to that observed in boar sperm. Human cells and boar sperm were equally sensitive to cereulide. The results show that boar spermatozoan assay is useful for detecting cereulide concentrations toxic to humans. Spermatozoa in commercially available extended fresh boar and cryopreserved bull semen were compared, boar sperms were 100 times more sensitive to cereulide than bull sperms.