Noncorrelation between mouse toxicity and serologically assayed toxin in Clostridium botulinum type A culture fluids.

Toxicity in culture fluids of several Clostridium botulinum type A strains was assayed in mice and converted to weight equivalent. The toxin-related antigen in the samples was quantitated by a radioimmunoassay which used standards of known antigen concentration instead of the usually used toxicity. Freshly prepared samples had reasonably similar titers of toxin and antigen. When the samples were held at room temperature for several weeks, toxicity decreased more than antigenicity, but the relative decreases of the two varied with the samples. The results are discussed as evidence that serological assays of botulinum toxin cannot always be used for accurate determination of toxicity.