HIV hangouts. The virus (lollipops) inserts itself in many different chromosomal spots.

HIV, at Home in the Chromosome

A most unwelcome guest, HIV insinuates itself into the chromosomes of the cells it infects. Now, researchers have found that the virus is like a party crasher who arrives early but doesn't help with the gala preparations: A new study shows HIV prefers to settle into sections of the genome with active genes, and then makes them work harder. The results will help scientists better understand how HIV replicates itself and causes disease.

HIV propagates by squeezing a DNA duplicate of its genome into the chromosomes of infected cells--usually immune system T cells. Researchers have long wondered how the virus chooses where to settle. Early on, they looked for short, specific sequences of chromosomal DNA. The guess was that HIV might head for a predictable site every time. No luck. With the availability of the human genome sequence, however, virologist Frederic Bushman at the Salk Institute in San Diego, California, decided to do a sweeping assessment of the chromosomal sites into which HIV integrates, hoping to pin down the virus's favorite places.

The team infected cultured T cells with enough HIV to get one or two viruses into each cell. After 48 hours, the researchers ground up the DNA and isolated 524 DNA regions into which HIV had integrated. Then they sequenced the DNA and mapped the sites on the human genome. HIV landed in genes 69% of the time. Those genes produced twice as much RNA as did clean genes, indicating that integration spurs genes to action. In addition, the team found short chromosomal regions into which HIV settled significantly more often than others. These integration "hotspots" were all in areas dense with genes, they report in the 23 August issue of Cell.

Calling the results definitive and straightforward, virologist Michael Malim at King's College London, U.K., says the work raises many interesting questions, such as how HIV selects and cranks up active genes, which "is very important for understanding pathogenesis." Retroviruses have to integrate to replicate efficiently, notes retrovirologist John Coffin at Tufts University School of Medicine in Boston, but no one knows why. If researchers find out, perhaps they can give the party crasher the boot.

Related sites
Bushman Lab home page
Frederic Bushman's "Lego Series of Integration"
Co-author Berry home page
Co-author Ecker home page