killing cells laden with the hiv may form a key part of future aids vaccination strategies. A new aids vaccine, which does exactly that holds good promise, say researchers. Vaccinated monkeys survived a usually lethal infection with a monkey-human hybrid virus, shiv. Their primed immune system kept virus levels below detection, shows research conducted by Emilio Emini and his team at the Merck Research Laboratories in Pennsylvania, usa (Nature, Vol. 415, p331 - 339).

But doubts are being voiced about the efficacy of the treatment. Low-lying virus can change to elude the immune response, argues Dan Barouch of Harvard Medical School in Boston. A year after his team administered eight animals a similar vaccination, a mutant form of the virus emerged, killing one of the monkeys. "These findings should be a reality check rather than a death knell,' Barouch says, adding that the technique can be modified to block the emergence of mutant viruses.

Future attempts could hit several different proteins to prevent the virus escaping. Both the latest vaccines targeted a single protein. Emini is already pursuing this goal. But like multi-drug-resistant bacteria, hiv could conceivably outwit even a broadly targeted vaccine, although this is less likely. Whether the vaccines tested in the monkeys will be as effective in humans, however, remains unknown as yet.

aids is notoriously difficult to vaccinate against. Initial attempts took the conventional approach of injecting a single hiv surface protein to trigger the animals' immune system into producing antibodies that would attack the virus during a subsequent infection. But hiv evades antibodies by hiding the proteins that the antibodies latch onto as well as by evolving new strains. In the past few years, vaccine hunters have switched to simulating the way the body's immune system attempts to deal with the infection. Emini's team achieved this by using a harmless virus to deliver siv dna direct to specific immune cells.