In a darkened room, a scientist watches a blurred computer image. Chemist Michael D. Varney hands a visitor a $1,500 set of goggles, and the image on the monitor is suddenly in focus--a multicolored 3-D spiderweb flung over invisible round bulges. With the push of a few buttons, Varney rotates the image, zooms in, then back. He is checking out the "contour map" of an enzyme that cancer cells need to grow. A pocket-like section of the enzyme could be its Achilles' heel: A drug placed there just right could be a cancer fighter.
Varney is a researcher at Agouron Pharmaceuticals in La Jolla, Calif., where he and others collect clues from biology that tell them what causes disease. Then, they try to use these insights to create synthetic drugs that will stop a disease in its tracks. This approach--an amalgam of biotech and traditional drug development techniques--is still controversial and, critics say, unproven. Yet on Apr. 5, Agouron gave the critics pause. It said that it had unraveled the structure of an enzyme the AIDS virus needs to replicate. Now, Agouron is testing drugs designed to disable the enzyme. If those work, they might stop the virus cold.
The story of Agouron's work with the enzyme, RNase H, may foretell the future of drug research. Scientists have long known about RNase H, and have spent years trying to map it. But there was a snag: It was hard to get enough of the substance to study.
Then, starting in late 1989, Agouron's scientists used a battery of biotech tricks to discern the genetic makeup of RNase H. That gave them the recipe for the chemicals that form the enzyme. They spliced that into bacteria, which grew the larger quantities needed to produce a crystal--a rigid, highly structured representation of a molecule. Next, Agouron's biologists handed off the protein to crystallographers, who placed droplets of it in special wells that encourage a crystal to grow.
FAST FORWARD. After a few weeks, crystals appeared, each smaller than a grain of salt. A single crystal was placed in a two-inch-long sealed glass cylinder and bombarded from all angles by X-rays. The pattern of light diffractions, interpreted by a computer, helped sketch a 3-D image of how the molecule exists in nature.
Once the structure was known, the scientists examined it on their computers and found the critical junctures where the enzyme helps the HIV virus to replicate. In the case of RNase H, just getting the structure and identifying its "active site," which suggests the best place to disable it, was a dramatic advance. The day after word of this discovery was published in the journal Science, the company's stock shot up more than $8. On April 24, Agouron signed a $6.5 million deal with Schering-Plough Corp. to use the approach on anticancer agents.
Just 15 months after starting the project, Agouron's chemists have several promising synthetic chemicals in hand that might disrupt RNase H at its active site. The next stage will require many rounds of screening, to see if these computer-designed drugs actually disable RNase H.
If one of these drugs works, it will still take years of testing on animals and people before it can be sold. But the method Agouron took to get this far is "unlike anything in the traditional pharmaceutical industry," says Peter Johnson, the company's president. And it may turn out to be a shortcut to saving many lives.