In the developed world, HIV is no longer a death sentence. But it still is a life sentence. Like other retroviruses, HIV physically inserts its genome into that of its host and cannot be expelled.

Or at least it couldn't until now. In the June 29, 2007 issue of Science, German researchers from the Heinrich-Pette Institute for experimental virology and immunology in Hamburg and the Max-Planck-Institute for molecular cell biology in Dresden described creating an enzyme that can recognize and cut DNA sequences flanking HIV, ridding host cells of the HIV genome.

The method the researchers used to cut the viral genome back out is essentially a variant on inducible knockouts. The scientists altered the enzyme Cre, which is a molecular biology workhorse that is specific to a so-called recognition site known as loxP, and can cut DNA sequences flanked by loxP out of the genome.

"The idea was to teach Cre to recognize HIV sites," Joachim Hauber told BioWorld Today. Hauber is head of the department of cell biology and virology at the Heinrich-Pette Institute and a co-author of the paper.

HIV, once it is integrated into the host genome, is flanked by DNA sequences known as long terminal repeats. The long terminal repeats have regions of similarity to the loxP recognition site, but they also have one important difference: The loxP site is symmetrical, and the long terminal repeats are not.

The researchers first identified a sequence in the HIV flanking regions that was similar to loxP and tested whether Cre was able to cut that sequence, but that was not the case. So they used a combination strategy of altering the target and the enzyme to develop an active version.

The target sequence was altered until Cre showed some activity against it, and Cre then was subjected to a form of directed evolution developed by co-author Frank Buchholz known as substrate-linked protein evolution.

After more than 120 cycles, the scientists met with success. A new recombinase, which they christened Tre, was able to recognize the long terminal repeat sequence of a specific HIV strain and cut it out of infected cells in culture.

The technology's most direct impact, Hauber said, is that it expands the possibilities for knock-in and knockout technologies: "It shows you can generate recombinases for any site, including asymmetric sites," he said.

It also has therapeutic potential, but whether that potential will be realized hinges on a number of factors. Hauber stressed that therapeutic victories, if any, are years away from reaching patients. Aside from the usual risks of drug development, HIV evolves rapidly, and patients also end up being infected with several different strains. That could make the method impractical in the field, as could the fact that the virus integrates into the host genome at a variety of sites.

Still, it allows researchers to start exploring the possibility of actually curing HIV, rather than managing it as a chronic disease, however successfully.

"You will never achieve a cure without excising the viral DNA," Hauber said. "This technology is the first step in that direction."