Fingerprinting fugitive microbes

Even in a struggling economy, the job market is booming for genetically engineered bacteria.

These microscopic machines are being put to work making everything from pharmaceuticals to fuels, raising the question of how to track the invisible critters if they ever got loose—or worse, if engineered pathogens were ever released as an act of bioterrorism.

Scientists have developed a software tool that finds characteristic “fingerprints” in the microbes’ DNA that can distinguish altered bacteria from natural ones.

Typically, scientists deliver foreign genes to bacteria on plasmids, small rings of DNA that bacteria naturally swap back and forth. Researchers have designed many kinds of artificial plasmids for various uses, but because new designs are usually based on older ones, artificial plasmids typically share many of the same segments of DNA.

Jonathan Allen and his colleagues at Lawrence Livermore National Laboratory in Livermore, Calif., reasoned that they might be able to use these shared segments to identify artificial microbes.

“The biggest question in our minds was how hard it would be to distinguish these [artificial plasmids] from just natural plasmids,” says Allen, a computational biologist in the pathogen bioinformatics group.

The new software tool automates the process of finding the optimal set of genetic fingerprints. The team input the genetic code for 3,799 known artificial plasmids into the software, which compared the sequences and found hundreds of matching stretches, each with about 20 “letters” of genetic code. The program then computed the smallest set of these shared snippets that can accurately distinguish artificial plasmids from natural ones.

Applying the test to another group of artificial plasmids identified 98 percent of them with no false positives, the team reports in the March Genome Biology.