Researchers in Germany reported Thursday that they had altered the DNA of a disease-causing bacterium so that it can infect a species it cannot normally sicken.

Experts called the development a double-edged advance. Although the research could deepen scientists' understanding of human diseases, it also could speed development of novel bioterror agents.

The change in infectiousness - the first of its kind ever engineered from scratch - poses no direct threat to human health, scientists said, because the microbe already causes a human disease - the food-borne illness called listeriosis.

The change allows that microbe to sicken mice, a species that it has no natural capacity to infect.

Still, the work has biosecurity implications because it could, in theory, be applied in reverse, endowing a bacterium that causes a serious animal disease with an unprecedented ability to sicken people.

Several experts said they were disappointed that the report, in today's issue of the journal Cell, does not mention those implications.

Also worrisome to some is that Cell's editors did not seek outside advice on whether publication of the study posed a security threat. Although in this case the consensus appears to be that it would easily pass muster, several U.S. organizations have called for such reviews when "dual use" microbiological advances are submitted for publication.

"What this really points out is the difficulty of dealing with all these issues," said Claire Fraser-Liggett, director of the Institute of Genome Sciences at the University of Maryland School of Medicine, who has participated in the development of standards for publishing such research. "It's hard to come up with guidelines that are absolute or anticipate everything that scientists are going to do."

Aside from such worries, scientists said that the new work marked a remarkable achievement in protein biophysics, a quickly maturing field that is revealing how proteins - the workhorses of living cells - interact with one another on the atomic scale.

Study leaders Wolf-Dieter Schubert and Andreas Lengeling, of the Helmholtz Centre for Infection Research in Braunschweig, knew that a particular protein on the surface of the bacterium Listeria monocytogenes was crucial to its ability to infect human intestinal cells. And they knew how the mouse version of that protein differs slightly from the human version - which accounts for the microbe's inability to infect mice.

Based on knowledge about how differently charged proteins interact with each other, the researchers predicted that the Listeria protein would settle reasonably well onto the mouse intestinal receptor if they could change just two of the bacterial protein's amino acids, which are the building blocks of proteins. In their work, they successfully re-engineered Listeria's DNA code so the microbe would start making surface proteins containing the two substitute amino acids.

That made the difference. When the researchers fed the altered bacteria to mice, the animals got sick with listeriosis.