Technical Guidance on the assessment of the toxigenic potential of Bacillus species used in animal nutrition

Applicants seeking regulatory approval for strains of Bacillus species have relied for guidance on how to assess any toxigenic potential on a SCAN opinion published in 2000. Subsequent to this opinion there have been significant developments in the knowledge of the nature of the toxins found in Bacillus species and their prevalence. In addition, a number of Bacillus species to which the SCAN opinion would have applied have now been transferred to other genera (Geobacillus, Aneurinibacillus and Paenibacillus). In view of the need to update the guidance contained in the SCAN opinion, the FEEDAP Panel takes the opportunity to adopt the revision as part of its technical guidance. The SCAN guidance took as its basis that toxins produced by species other than members of the B. cereus taxonomic group would have sufficiently similar properties to the known B. cereus toxins to be detected by the methods developed for the B. cereus group. It now seems unlikely that B. cereus-like enterotoxins are produced in species other than the B. cereus group. Any toxigenic potential in other species appears far more likely to arise from the production of surfactins. A test for haemolysis coupled with PCR detection of non-ribosomal peptide synthase genes seems adequate to identify surfactin-positive strains. However, a single cytotoxicity assay is retained to confirm the expected absence of B. cereus-like enterotoxins. The selection of strains belonging to the B. cereus taxonomic group for use in animal production is considered inadvisable. If, however, they are proposed then it is recommended that the full genome should be sequenced and a bioinformatic analysis made. If there is evidence of homology with genes encoding toxins then their nonfunctionality (mutation, deletion, etc.) should be demonstrated. Strains harbouring a toxigenic potential should not be used as feed additives.