When bacteria team up in sticky, drug-resistant communities called biofilms, they can be nearly impossible to eradicate using conventional antibiotics. Commonly found on medical devices such as catheters, biofilms can cause gingivitis and chronic ear infections. They can also cause environmental and industrial problems, clogging up water pipes.

Boston University biomedical engineers have designed a new, highly effective means of dispersing and killing the bacteria living in biofilms. Led by synthetic biologist James Collins, the team has engineered viruses that attack biofilms on two fronts: by killing the bacteria that live in them, and by dissolving the carbohydrates that hold them together. If such bacteria-attacking viruses are proved safe for industrial and clinical use, says Collins, researchers could develop stocks of different kinds of viruses, each tailored to attack a different kind of biofilm.

Collins has designed a virus that can disperse more than 99 percent of the E. coli in a model biofilm. Helen Blackwell, a chemist at the University of Wisconsin-Madison, believes that this is an "enormous" achievement: "I haven't seen anything as effective as this approach." Collins's engineered virus is described online in the journal Proceedings of the National Academy of Sciences.

Bacteria living communally in biofilms are one thousand times more resistant to antibiotics than free-swimming bacteria are, says Collins. They are protected by a sticky carbohydrate scaffold called a matrix. The matrix blocks antibiotics and cells from the human immune system, and even provides something like a primitive circulatory system for the bacteria.

In a few cases, including some chronic ear infections in children and chronic lung infections in cystic-fibrosis patients, the tissue harboring a biofilm must simply be cut out. (See "Biofilms to Blame for Chronic Ear Infections.") Large doses of antibiotics can usually eradicate these infections, says Blackwell. But she notes that there is some worry that drug-resistant biofilm infections are becoming more common, and that the use of antibiotics seems to induce biofilm formation.

"One thing I like about [Collins's] approach is that it is two-pronged," says Philip Stewart, director of the Center for Biofilm Engineering at Montana State University. "The [viruses] kill the bacteria, but they also target the biofilm matrix."

Collins's approach is to select a virus that already targets the bacteria of interest, such as E. coli or Staphylococcus. Then he introduces into the virus a gene for an enzyme that dissolves the main carbohydrate component of the biofilm matrix protecting the bacteria. There are viruses specialized to infect every bacterial species. These viruses replicate inside bacterial cells, then burst them open, killing the bacteria, and spread to other bacterial cells. But they do not harm animal cells or bacteria other than the kind to which they are targeted.

Naturally occurring viruses can attack biofilms. But Collins showed that giving a virus a gene for dissolving the matrix increased the virus's effectiveness by 4.5 orders of magnitude.