Vaccine experts say they may be closer to understanding why some experimental AIDS vaccines have been proven ineffective. They say the viruses being used in the vaccines may damage and exhaust immune cells.

Some of the vaccines that had recently failed in human testing trials use adeno-associated viruses, a family of usually harmless viruses used as so-called "vectors" to carry genetic material from HIV into the body. The hope had been that the immune system would notice the HIV genes and mount a response against them.

But the adeno-associated viruses themselves may be doing harm, reports Dr. Hildegund Ertl in a study released Thursday.

Ertl, the director of the Wistar Institute Vaccine Center in Philadelphia, says in his study, conducted on mice, adeno-associated viruses directly interfered with immune cells called CD8 T-cells. These are the so-called "killer" T-cells that a vaccine is supposed to stimulate to fight HIV.

His team's findings are reported in the Journal of Clinical Investigation.

Ertl says he doesn't know the mechanisms at work, but notes that there is still much that science doesn't understand about viral vectors and how they interact with the immune system.

He adds that vaccines using the viruses should not be further tested on humans until more studies are done.

Earlier this year, two international studies on an experimental AIDS vaccine manufactured by Merck & Co. were halted, after researchers found worrying indications that not only did the vaccine not prevent infection with HIV, it may have actually raised the risk of infection.

During the course of the study, 49 of 914 vaccinated men became infected with HIV, compared with 33 of the 922 men who got a placebo shot.

On Tuesday, Merck & Co. and academic researchers announced they would "unblind" their study, conducted on 3,000 people in the United States, Peru, Brazil, Dominican Republic, Haiti, Jamaica, Australia and South Africa, so that everyone involved could find out who got the real shot and who got a placebo injection.

Officials at Merck said it's unclear why more of the vaccinated volunteers wound up getting HIV than those who got the placebo. But if Ertl's theory is correct, it is possible that the adeno-associated viruses exhausted the immune system, causing "killer" T-cells "to turn themselves off." An immune system with turned-off T-cells would then be more vulnerable than usual to infection.

Seth Berkley, the president and chief executive of the nonprofit International AIDS Vaccine Initiative, says her group is not discouraged by the results of the discontinued Merck trial. She says another 30 or so vaccine candidates are in various stages of testing, as are other therapies that target the antibody arm of the immune system rather than T-cells.

"Research suggests that the human immune system is better at fighting HIV than we'd originally thought," she noted in an editorial in the Los Angeles Times earlier this year. "Most people suppress the virus for many years before developing AIDS. A small number never contract the virus despite repeated exposure. Others are infected but haven't developed AIDS.

"If we can figure out the mechanism that makes this so, we'll have clues about how to engineer a vaccine."