The drug (blue) jams a key component of the SARS virus. Credit: source:

Researchers have made a three-dimensional image of a crucial part of the virus that causes severe acute respiratory syndrome (SARS). The model suggests that a drug being tested as a cold remedy is a good starting point in treating the killer disease.

Another group has already tested the molecule, called AG7088, against the SARS virus in the test tube. Results have been negative, says virologist Robert Baker of the US Army Medical Research Institute of Infectious Diseases (USAMRIID) in Fort Detrick, Maryland. "Yet that result is preliminary and we are re-testing now," he says.

But many very similar chemicals show activity against SARS, suggesting that researchers are closing in on the right target, says Baker.

The drug blocks a component of the virus called a proteinase. All coronaviruses, the family to which the SARS virus belongs, rely on these molecules to reproduce.

"It's the Achilles' heel of the virus," says biochemist Rolf Hilgenfeld of the University of Lübeck in Germany, who led the team behind the model. "We're convinced that hitting this target will be enough to stop infection."

To sculpt their SARS target, Hilgenfeld's team used models of two coronavirus proteinases that they were studying before the outbreak. One virus causes diarrhoea in pigs, the other triggers cold-like symptoms in humans. The team's computer model is based on these proteinase structures and the SARS genome1.

In the groove

AG7088 fits a cleft in the proteinase, the analysis revealed. Like a chemical spanner in the virus' works, the molecule should prevent the proteinase from working.

"It doesn't fit very well," admits Hilgenfeld. "But it will be a very good starting point."

AG7088 is in human tests now - the drug company Pfizer created it to block the proteinase of an unrelated cold virus. The chemical has so far proved to be safe, and a modified version would probably take less time to develop into a SARS drug.

The drug is one of about 100 candidate compounds loaned by drugs companies to the USAMRIID. If, as is likely, none is a perfect match, the German team's structure, combined with the molecules, "should quickly lead to a wealth of good leads we can follow up", says Baker.

It doesn't fit very well, but it will be a very good starting point Rolf Hilgenfeld , University of Lübeck

If a ready-made drug to treat SARS cannot be found, Hilgenfeld and his colleagues advocate using their model to design a better SARS blocker. This strategy was used to design the latest HIV drugs.

The other, and some would argue quicker, approach is to screen chemicals. In the past month, USAMRIID scientists have tested around 120,000 compounds for activity against SARS. Given the urgency of finding a treatment, both approaches should continue. The rational approach, says Baker, "will find leads that we wouldn't necessarily run across using brute force".