Scientists don't yet know exactly how the Rhodococcus strain acquired the ability to make this new toxin. Only one out of a number of flasks of Rhodococcus growing with enemy Streptomyces produced the antibiotic. Kurosawa and his colleagues discovered that the drug-producing strain contains a large chunk of DNA from the other organism. While previous research suggests that DNA swapping between bacteria is quite common--it's thought to underlie bacteria's ability to quickly evolve drug resistance--the exchange has been difficult to observe firsthand. "In this case, the process is caught in the act, and you can see the consequences," says Jon Clardy, a chemist at Harvard Medical School, in Boston.

The work has elicited excitement from scientists developing novel antibiotics because the method could provide a new way to uncover the hidden antibiotic-producing abilities of different kinds of bacterium. "Advances in sequencing technology are now making it possible to see how the diversity of known antibiotics has come from gene swapping," says Michael Fischbach, a microbial geneticist at the Broad Institute, in Cambridge, MA. Fischbach is overseeing a project to sequence 16 strains of the Streptomyces, in which scientists will try similar methods to coax out new drugs.

Previous sequencing research suggests that some strains have the genetic ability to produce 20 to 30 different antibiotics, but when grown on their own in comfortable lab conditions, they produce only two or three. "Where are the other 90 percent?" asks Fischbach. "I think [Kurosawa's] approach is the right way to explore this."

It's not yet clear if the swapped piece of DNA contains genes for the antibiotic itself or if it triggers a regulatory mechanism that warns Rhodococcus of encroaching bacteria, turning on an inherent but often silent mechanismto make toxins. So far, the researchers have sequenced only half of the DNA insert; they expect to sequence the other half soon.