Scientists at the National Institutes of Health have begun raising litters of mice that carry a copy of genes for the AIDS virus in every cell in their bodies, the first time the complete genetic code for an organism causing a lethal disease in humans has been introduced into another animal species.

Concern that some of the more than 100 mice that may be involved in the project over the next few months could escape and mate with wild mice, establishing a natural animal reservoir for the AIDS virus, has prompted unprecedented security precautions at NIH's Bethesda campus.

The experiments are being carried out in a maximum-security laboratory where the amimals, which could have infectious AIDS particles in their blood or secretions, are confined in cages inside glove boxes -- enclosed cabinets with thick gloves built into their walls to allow laboratory workers to handle the animals -- behind moats filled with Clorox.

When the project was being considered for approval by a biosafety committee last summer, some scientists within the NIH community voiced such serious doubts about its safety that Dr. Malcolm Martin, a virologist at the National Institute of Allergy and Infectious Diseases, eventually agreed to limit its duration, promising to kill all of the mice carrying the viral genes by next April.

Martin said the safety procedures are more than adequate.

"This is basically a mouse jail," he said. "There is no way the mice can get out. I think the safety precautions are in one sense overkill, but that they are appropriate considering both the science and the psychology of this experiment."

Scientists said they believe that introduction of the virus into mice could produce major gains in the battle against acquired immune deficiency syndrome by providing a much-needed way to study the virus in a small, inexpensive research animal. Besides humans, the AIDS virus naturally infects only certain varieties of monkeys, which are expensive and difficult for scientists to work with.

The mice experiment raises questions about how such tests should be regulated. No federal guidelines govern such so-called transgenic research, in which a piece of the genetic material called DNA is injected into the fertilized egg of an animal, according to

Dr. Dinah Singer, chairman of NIH's Institutional Biosafety Committee.

Nor do guidelines govern introduction of the genes of human disease-producing organisms into animals, she said.

"I have heard through the rumor mill that other institutions are planning this experiment" or similar ones with less stringent safety precautions, she said, noting that there are only a few other maximum-containment laboratories in the country. "I'd hate to see that," she added.

Martin said scientists who initially expressed concerns about the project concluded that it was safe after he showed them the special laboratory and consented to limits on the project's size and duration and to additional precautions. Singer said her committee approved

the experiment unanimously last

July.

"People have created scenarios where mice are going to escape, fornicate outside the National Library of Medicine and there will be some new . . . epidemic that we are creating," Martin said. "It's rather far-fetched, given the lab we're doing it in."

In the experiment, researchers inject a DNA copy of AIDS virus genes into the nucleus of a fertilized mouse egg. If the injection is successful, the viral DNA becomes part of one of the mouse cell's chromosomes, the gene-containing structures in the nucleus. Every time the cell divides, copies of the viral genes are carried into the daughter cells.

The fertilized mouse eggs are then implanted into the uteri of adult female mice, where they develop into embryos. Martin said that about 30 mice have been born as part of the experiment so far and that DNA from their cells is being analyzed to determine how many carry the viral genes.

Martin said he hopes that, in at least some of a carrier mouse's cells, viral genes will become active, manufacturing viral proteins or complete AIDS virus particles. The mouse may then develop an AIDS-like illness or other evidence of disease.

He said such animals could be used to study how the viral genes are activated and to test drugs that might block ill effects.

He said the experiment will fail if the presence of viral genes either kills the mice during embryonic development or produces no ill effects.

If it succeeds, causing a virus-related illness, he said he hopes to duplicate the research using a mutant form of the AIDS virus that is genetically defective and cannot infect humans. He said such a variant could not be used initially because he wanted to test the effect of all of the virus' genes.

Martin said animal waste, instruments, tissue specimens and other materials must be sterilized in

an autoclave or sealed in plastic bags and passed through a Clorox-filled "moat" before they can be removed from cages inside the glove boxes.

The live animals remain in the boxes at all times, he said. In addition, the laboratory is equipped with mouse traps and extra doors, and the locks were changed before the experiment to prevent possible sabotage, he said.

Martin said he could not rule

out the slight risk that a mouse might infect a laboratory worker by biting through one of the thick gloves. "That's a possibility, though we think it's very, very slim," he said.

Experts in transgenic research at other institutions said the experiment might provide valuable knowledge about the AIDS virus and said it poses no serious risks.

"I think the chance of the mice actually expressing whole virus is practically zero," said Dr. Ann Kiessling, an associate professor of obstetrics, gynecology and reproductive biology at Harvard Medical School.

But others expressed concern about the implications of such research and said it should be more strictly regulated. The proposal for the experiment was not formally considered by the NIH's Recombinant DNA Advisory Committee (RAC), although six individual committee members were asked to comment on it.

"The issue is, what were the {other} available options?" said Sheldon Krimsky, an associate professor of urban and environmental policy at Tufts University and a former RAC member. "When you begin taking a virus and putting it in a different species, you can't guarantee that the methods of spread and infection will always be the same as they have been."