A plasmid coding for the anthrax toxin has been found in a sample of Bacillus cereus taken from a person who survived a life-threatening pneumonia symptomatic of inhaled anthrax. The finding, reported in the advanced online edition of PNAS, is the first time the complete plasmid has been found in any naturally occurring microbe other than Bacillus anthracis.

The study grew out of a Centers for Disease Control and Prevention (CDC) review of reported cases of unusually severe illnesses caused by these or other Bacillus species. The researchers did not study cases of anthrax of the skin. They note that 2 years ago, other scientists found one of the genes for the anthrax toxin in two fatal cases of pneumonia caused by B. cereus. But those samples were not further characterized, so if the plasmid was present, it was not found.

B. cereus is known to cause food poisoning, but rarely causes pneumonia. However, when scientists at the CDC and at the Institute for Genomic Research (TIGR) quickly sequenced the DNA of their pneumonia-causing sample, they found a plasmid outside the bacterium's single chromosome that is 99.6% genetically similar to the plasmid in anthracis, pX01. In anthrax, that codes for the three-part toxin.

Although the researchers did not find a homologue to pX02—the anthracis plasmid–containing genes that create an impenetrable capsule surrounding and protecting the bacterium—they found a different plasmid coding for a similar capsule, which may serve the same function.

Researchers at the US Naval Medical Research Center showed that when they injected mice known to be sensitive to B. anthracis peritoneally with the plasmid-containing B. cereus, all the mice died within 14 days. “It may not be appropriate to consider B. anthracis, as currently defined, as the only species capable of causing inhalation anthrax-like disease,” the study concludes.

The study's authors say their research proves that scientists could respond rapidly to newly emerging diseases, both those caused naturally or others manufactured by bioterrorists by altering existing diseases. “These findings show that genomics can rapidly assist public health experts in responding to novel pathogens,” Claire Fraser, one of the authors and the president of the Institute for Genetic Research, said in a statement.

But for public health agencies to be able to respond quickly and reliably, the authors point out, the agencies must revise the way they monitor for anthrax cases, shifting from analysis of isolates to evaluation of the similarity of the patient's symptoms to anthrax. Now, when a Bacillus sample is found to be able to move and to destroy red blood cells, it is assumed not to cause inhalation anthrax because anthrax cannot do that.

“Normally, when we look for anthrax, we say that it has to be non-hemolytic and non-motile, whereas this one [B. cereus] is motile and hemolytic,” said Nammalwar Sriranganathan, an associate professor of biomedical sciences and pathobiology at Virginia Tech who is developing an anthrax vaccine, but who was not involved in the present study.

No one knows how this particular isolate of B. cereus acquired its plasmids, but the feat doesn't surprise Sriranganathan or Arthur Aronson, a professor of biological sciences at Purdue University who studies the B. cereus variant strain ATCC 4342. They said the bacillus family of bacteria is known to be capable of “gene jumping”—transferring genes laterally from one species to another.

However, Aronson said he is not convinced that the B. cereus isolate actually caused anthrax in the human patient or in the mice that were injected with it. He said the descriptions of their symptoms do not uniquely describe inhalation anthrax.

Aronson also said that mice are not a good animal model for anthrax and that rabbits are better. Moreover, injecting mice in their peritoneal cavities may not produce the same results as making them inhale the isolate.

“So what they really needed to do,” Aronson said, “is get rid of that plasmid from that B. cereus isolate to see if it was still pathogenic or disrupt the gene that encodes one of the principal toxins. That would have been definitive. In the absence of that, I don't think it's fully convincing evidence.”

“We totally agree,” said Alex Hoffmaster, the lead author of the paper. “It's very intriguing that this organism caused the very severe disease and it had these plasmids. But we have not shown that these plasmids contributed to the disease, so we don't know that this organism produced anthrax toxin and the patient suffered from anthrax toxin.”

Editor's Note: See The Scientist's special issue on biosecurity here.

Correction (posted May 25): When originally posted, this story incorrectly stated that the capsule surrounding B. cereus protects it even while the organism is dormant in soil. The Scientist regrets the error.