Retroviruses pose a major threat to the fitness and survival of many animals, including humans. Researchers have now identified a factor that keeps one type of the most notorious human retrovirus, HIV-1, stuck to the surface of human cells and unable to spread.

HIV-1 encodes a slew of proteins, and relies on host cells' machinery to make them. Among these is a set of accessory proteins that for some time didn't seem to be important for virus replication. One of these, Vpu, allows the virus to escape from the external surface of cells. HIV-1 viruses that lack Vpu get stuck, and cannot leave cells to invade others.

Paul Bieniasz, a virologist at the Aaron Diamond AIDS Research Center and Laboratory of Retrovirology at The Rockefeller University in New York, and his group began scouring the literature for cell-surface proteins that might hold on to HIV-1 in the absence of Vpu. They thought that Vpu might function by antagonizing such a protein — which they dubbed tetherin — allowing the virus to escape.

The search led the team to an adhesion molecule known as ICAM-1. “We knew that tetherin made the cell surface sticky, so a good candidate for that would be a cell-adhesion molecule,” says Bieniasz. Earlier research had shown that ICAM-1 could be incorporated into HIV particles.

Stuart Neil, then a postdoctoral researcher in Bieniasz's laboratory, was keen to get to work. He quickly cloned ICAM-1 and introduced it and HIV-1 to cells. ICAM-1 seemed to inhibit the release of the viral particles, an effect that the addition of Vpu could counteract.

But, says Bieniasz, when Neil, now a lecturer at King's College London, repeated the same experiment using ICAM-1 DNA that had been more thoroughly purified, “the effect completely went away”. Although initially disappointing, the results ultimately provided a key piece of information. When cells are transfected with 'dirty' DNA, they sometimes produce interferon and other cytokines as part of an immune response. Bieniasz and Neil realized that the protein they were looking for was not ICAM-1, but one that could be induced by interferon.

After a series of experiments failed to isolate the mystery protein, the researchers reluctantly shifted strategy. They decided to compare gene expression in various cell lines, treating some of them with interferon-α. But they were pessimistic: “We thought there would be too many candidates to weed through,” says Bieniasz. Instead, the results were “quite stunning”, with fewer than 10 candidate genes meeting their selection criteria (see page 425).

Once the gene-expression data were in hand, the story unfolded very quickly. The team identified a favourite candidate within an hour, and the protein a week later. “It was very satisfying,” Bieniasz says.

There is still much to learn about tetherin, however. There are hints that the protein has broad antiviral activity, yet few viruses make Vpu, which suggests that they have some other way of skirting tetherin's stickiness. Bieniasz also wonders whether genetic variation in tetherin might affect susceptibility to HIV infection or other viral diseases. So there's plenty of work to keep him tethered to the lab bench.