To the editor:

In a Perspective in the September issue (Nat. Biotechnol. 22, 1105–1109, 2004), Heinemann and Traavik suggested that horizontal gene transfer from Bt crops (transgenic plants expressing a cry gene from Bacillus thuringiensis) may pose a food safety or other environmental hazard because “it is noteworthy that B. thuringiensis has “a significant history of mammalian pathogenicity” and is thus not irrelevant to food safety or other environmental issues.” The reference is to a review from us, but misconstrues our original intent because the original text in our review actually runs: “Bt does not have a significant history of mammalian pathogenicity...”1 [emphasis added].

In the April issue, the authors make a correction indicating that they wrongly cited our reference but persist in their opinion by now referring to the close relationship between B. thuringiensis, Bacillus cereus and Bacillus anthracis, of which strains of the latter two do have significant pathogenicity. In their corrigendum2, to support their allegation that plants containing a cry gene may constitute a hazard, they state: “Members of this group are so closely related that they may be considered members of the same species, often differing only by the presence or absence of certain plasmids”. Thus, they damn B. thuringiensis, and the use of its genes, by association with B. cereus, or worse, with B. anthracis, the causal agent of anthrax.

The authors are correct in noting that B. thuringiensis, B. cereus and B. anthracis are closely related; they are also correct in noting that the main differences are often due to the presence or absence of certain plasmids3. What they fail to note is that in this group of bacteria, as in many others, host specificity is largely determined by plasmid-encoded factors. Thus, the distinctive feature of B. thuringiensis is the production of insecticidal crystal proteins, and it is only the genes encoding these proteins (which are mostly plasmid-borne) that are expressed in transgenic plants. Moreover, these genes are derived from strains of B. thuringiensis that have been used as safe and environment-friendly sprayable pesticides for decades, including by many organic farmers.

Furthermore, the focus of Heinemann and Traavik's original Perspective is on transgenic plants expressing Bt toxin proteins from the bacterium, not on the bacterium itself. Clearly, the safety of a single gene product does not inherently reflect that of the complex organism from which it originates. We are not suggesting that safety issues should not be considered, but to date there have been many years of safe Bt toxin use. The overwhelming evidence is that B. thuringiensis, and transgenic crops expressing cry genes, are not a threat to mammalian species. The authors of this paper are using a poor interpretation of good science to imply a potentially serious risk.