As bacteria have grown increasingly resistant to standard antibiotics, scientists have begun a desperate search for alternatives to the drugs. In one promising approach, they are trying to harness viruses that naturally evolved to prey on harmful bacteria and to use them as weapons for staving off intruders.
That may sound like a new idea, but it is a revival of an ancient remedy. Therapies that use bacteria-eating viruses -- bacteriophages, or just phage for short -- have been part of traditional cures for centuries in India, and they have been practiced in the former Soviet Union for decades.
They have yet to be properly tested on people, and some researchers are skeptical about unleashing a self-replicating virus into a patient's bloodstream.
As drug companies abandon antibiotics for more reliable products, a handful of start-up companies is developing precision-engineered viruses for medical use. One of the youngest, GangaGen, founded in 2000 and based in San Francisco and Bangalore, uses genetic engineering to control the healing viruses.
GangaGen's name refers to the moment when Western science discovered phage therapies in 1896, when a British chemist noticed that water from the Ganges River in India, traditionally known for its curative properties, was lethal to the cholera bacterium. After gaining attention in Sinclair Lewis's ''Arrowsmith'' in 1925, phage cocktails were marketed by American drug companies for a brief period before being overshadowed by powerful antibiotics developed in World War II.
Phage treatments in the form of oral or topical medications are now poised for a comeback. But they are riskier than drugs. For good results, many strains of virus are often needed; the viruses can multiply quickly in a host; and they may in some cases produce toxic chemicals.
But GangaGen, which will conduct human trials in hospitals in India this year for a virus that attacks staph bacteria, is using genetic engineering to reduce the risks.
Unlike broad-spectrum antibiotics, phages evolved to attack a narrow range of bacteria, so they have traditionally been applied in combinations of three or more. It would be safer to find a single virus that did the job. To that end, GangaGen is pursuing ways to increasing the number of a virus's potential hosts by inserting new genes into its tail.
There is, however, a far more serious problem. The viruses, although they may save lives, can also kill people, ''not just because of the fact that they might be contaminated with something, but also because the viruses themselves can carry toxin genes,'' said Dr. Carl R. Merril, a senior investigator at the National Institutes of Health.
A phage uses its host to make hundreds of copies of itself that then burst out to infect other cells, growing exponentially in dosage. If a phage is allowed to replicate unimpeded, it may acquire a toxin-coding gene from its host, a potentially lethal turn of events.
To eliminate that problem, Dr. Ryland F. Young, a professor of biochemistry at Texas A&M who is also on the GangaGen advisory board, has engineered the first virus that does not burst open its host. [Correction: see below.] This phage kills bacteria like any other, the founder of GangaGen, Dr. Janakiraman Ramachandran, said, but it does not risk unchecked replication or the buildup of toxins.
Some experts said research to modify the viruses was not necessary for effective phage treatments.
''Why would you want to spend a lot of time and money fine tuning phage that already may be out there?'' said Dr. Alexander Sulakvelidze, head of research at Intralytix in Baltimore. The company was an early leader in phage therapy and is developing phage-based food additives to kill salmonella and listeria. ''The beauty is that they're so ubiquitous and natural. Mother Nature has the best genetic engineering lab.''
Although phage therapy has longterm potential as a treatment for illnesses like diarrhea and tuberculosis, Mr. Ramachandran said, GangaGen will focus on medical applications that do not require inserting viruses in the bloodstream.
The companies developing phage products seek money for research and trials. If officials decide that the antibiotics shortage has reached a crisis level or that phage therapy should be integral to biodefense, Dr. Merril said, anything is possible.
CORRECTION: An article in Science Times on May 18 about bacteriaphage therapy, which harnesses viruses to fight bacteria, misstated the role of Dr. Ryland F. Young in researching the treatment. Dr. Young studied the way the viruses replicate and burst out of the host cell to invade other cells. His research group was also the first to delete genes in a bacteriphage, preventing the virus from bursting out of the cell after replication. But Dr. Young was not the first researcher to manipulate a phage genetically so that it killed bacteria without bursting out and infecting other cells. Scientists isolated mutant viruses with that property more than four decades ago.