TORONTO - Canadian and U.S. scientists have identified a potential new "Trojan Horse" method for creating a vaccine against AIDS, even as repeated efforts by researchers to prevent HIV infection using traditional immunization approaches continue to fail.

Instead of trying to directly target the rapidly mutating human immunodeficiency virus (HIV), the scientists suggest a vaccine could take aim at what's known as fossil DNA - genetic material from ancient viruses that has inserted itself into every human cell over our evolutionary history.

In a study published Friday in the journal PLoS Pathogens, the researchers say it appears that HIV reactivates this usually dormant DNA - called human endogenous retroviruses, or HERVs - by disrupting the normal controls that keep it in check.

The study found that in some HIV-positive individuals, infection-fighting T cells are able to target HERV-enabled cells, said co-principal author Brad Jones, a PhD candidate in immunology at the University of Toronto.

Jones said a huge stumbling block for scientists and drug companies seeking an effective vaccine is that HIV is like a moving target - it exists in many variations and constantly mutates.

"If we can find other ways for the immune system to target HIV-infected cells, we can overcome this problem in making an HIV vaccine," co-author Dr. Keith Garrison, a postdoctoral fellow in experimental medicine at the University of California, San Francisco, said in a statement. "HERV may provide us with a good target to test."

In the latest setback, drugmaker Merck and Co. announced Wednesday that an experimental AIDS vaccine not only failed to work, but volunteers who got the shots were more likely to get infected with the virus through sex or other risky behaviours than those who got dummy inoculations.

Merck had already announced in late September that it was stopping its trial because the vaccine did not work, begging the question whether that failure is a harbinger of a similar fate for a number of other AIDS vaccines now being tested.

Traditional vaccines work by stimulating a response by key immune system cells to seek out and destroy foreign invaders like bacteria and viruses. But because HIV is an extremely adaptable retrovirus, it seems able to foil attempts by the immune system to shut it down.

In fact, it uses the body's own immune T cells as an incubator to make copies of itself, before spreading its progeny to other host cells.

Even if a traditional vaccine does elicit an immune response, "the virus may still be able to escape," said Jones. "So it may not matter how hard we hit it."

But because HERVs are already part of our genetic makeup, they are virtually unable to mutate, he said. "So there's a great advantage in that in targeting HERV."

That's where the Trojan Horse idea comes in: HIV activates HERVs within the cell it enters, so a vaccine that takes aim to destroy HERVs will incidentally kill HIV and stop it from jumping to other cells.

"We think this is an opportunity where we can actually use the fact that HIV has to disrupt these control mechanisms, has to therefore result in expression of these HERVs, use that back against HIV as a way of targeting HIV-infected cells," Jones explained.

"So we basically think they're a proxy for HIV infected cells."

In their study, the researchers looked at 29 people who were recently infected with HIV and compared them to 13 HIV-negative individuals and three others infected with hepatitis C but not the AIDS virus. In the group recently infected with HIV, they found a relationship between the degrees of T-cell response to HERVs and the levels of HIV virus present in their blood.

Dr. Rafick-Pierre Sekaly, scientific director and program leader of the Canadian Network for Vaccines and Immunotherapeutics (CANVAC), said targeting HERV "certainly would not be my approach."

"But right now even the most rational approach has not yielded any good results," said Sekaly, a professor of immunology at the University of Montreal who was not involved in the study.

"To try something which is totally outside of the box certainly merits some consideration," he said, while cautioning that the Toronto-UCSF results need to be confirmed by other researchers.

Jones agreed that although the evidence his group found is encouraging, it's still early days.

"This is fairly preliminary and even though we think it's a really fascinating idea, there's a long way to go before it's a vaccine."