Animal models for Francisella tularensis and Burkholderia species: scientific and regulatory gaps toward approval of antibiotics under the FDA Animal Rule.

The development and regulatory approval of medical countermeasures (MCMs) for the treatment and prevention of bacterial threat agent infections will require the evaluation of products in animal models. To obtain regulatory approval, these models must accurately recapitulate aspects of human disease, including, but not necessarily limited to, route of exposure, time to disease onset, pathology, immune response, and mortality. This article focuses on the state of animal model development for 3 agents for which models are largely immature: Francisella tularensis, Burkholderia mallei, and Burkholderia pseudomallei. An overview of available models and a description of scientific and regulatory gaps are provided.

Evidence for plasmid-mediated toxin production in Bacillus anthracis.

Large-molecular-weight plasmids were isolated from virulent and avirulent strains of Bacillus anthracis. Each strain contained a single plasmid species unique from the others with respect to molecular weight. Bacterial strains were cured of their resident extrachromosomal gene pools by sequential passage of cultures at 42.5 degrees C. Coincidental to the curing of plasmids was a loss of detectable lethal toxin and edema-producing activities and a dramatic decrease in lethal factor and protective antigen serological activities. The involvement of these plasmids in the production of toxin was firmly established by transformation of heat-passaged cells with plasmid DNA purified from the parent strain. The ability to produce parent strain levels of toxin was restored, and the plasmid DNA similar in molecular weight to that isolated from the parent was reisolated in all transformants examined. The exact role these plasmids play in the production of toxin remains to be elucidated. Two additional strains of B. anthracis, designated Pasteur vaccine strains, were examined for the ability to produce toxin and for the presence of plasmid DNA. Both strains were found to be nontoxigenic and contained no detectable plasmid elements. It is therefore likely that we, like Pasteur, cured B. anthracis strains of temperature-sensitive plasmids which code for toxin structural or regulatory proteins.