The idea could have come from the ghoulish children of TV's Addams family: In the 1970s, scientists seeking antihypertensive drugs would stretch rabbit muscle and pour random chemicals on it. Mechanical levers attached to each end recorded on paper if the muscle constricted around the blood vessels or relaxed--signaling a potential treatment. When researchers weren't torturing bunny muscles, they filled test tubes with other "screening targets," such as cancer cells, and then mashed up chemicals from plants, soil, or wherever to see if they killed the targets. Of the ones that did, many would bomb in animal or human trials. It took perhaps 10,000 labor-intensive tests over several years--and lots of luck--to find an effective drug.
Today, such medieval methods have given way to screening techniques that are far likelier to produce more effective drugs with fewer side effects. Thanks to biotech tools, drugmakers are deep-sixing crude targets such as muscles in favor of finely engineered targets--cells and key parts of cells--that make it possible to tell more precisely how a potential drug will interfere with the disease process. Buoyed by that progress, companies are also inventing new ways of making and testing billions of these matchups in a matter of weeks.
It's too soon to tell how big a boost these changes will give drug-lab productivity. But both startups and giants such as Merck & Co. are working to find out--and in doing so, are widening their search for new compounds. They're reexamining existing drug libraries, using novel techniques to synthesize drug candidates, and foraging for new substances in nature. Genentech Inc., among others, finds that new screening methods are "fundamentally changing the way we do research," says Arthur D. Levinson, the company's vice-president for research.
What's driving all this are new insights biotech has opened into disease--and into how drugs can best halt it. Drugs may kill bad cells, protect good ones, stop cells from making harmful chemicals, or prompt them to make desirable substances, such as the body's own disease-fighting antibodies. But drugs don't do anything unless their chemical structure matches up with structures on cell surfaces called receptors. When a drug binds to a receptor, like a key fitting a lock, it triggers reactions inside the cell's chemical pathways that tell the cell's DNA, or genetic machinery, to make a protein, to change, or to self-destruct. Understanding those triggers and pathways is key to turning drug-screening from a black art into a predictor of how well a drug may work. It's also crucial to designing drugs that act precisely on specific receptors, and thus have fewer side effects.
Gene-splicing is a critical technology in this effort. It lets scientists make copies of virtually any protein in the body, including receptors. So researchers can now clone lots of key receptors to use as screening targets: Genentech already has very specific receptors that have been implicated in cancer, inflammatory diseases, and other ills.
REPORTER GENES. Biotech also makes it possible to fiddle with the DNA of drug-screening target cells. At Oncogene Science Inc. in Uniondale, N.Y., researchers are splicing a so-called reporter gene into the DNA of target cells. When these cells are exposed to a molecule that triggers specific activity--such as the instruction for the cell to make a hormone--the reporter gene sends up a chemical flare that tells scientists they may have found a drug. At Cephalon Inc., in West Chester, Pa., researchers are using clones of receptors to find ways to regenerate damaged nerves. Cephalon has found that when a chemical pathway in a nerve cell is activated by a compound that binds to one specific receptor, the cell makes a protein that enhances its survival. Cephalon has already licensed chemicals that activate production of the crucial protein from Japanese drugmaker Kyowa Hakko Kogyo Co. By 1994, Cephalon hopes to test a drug that keeps nerve cells alive after a stroke.
Meanwhile, researchers are devising more efficient approaches to the "supply side" of screening--finding chemicals that may work as drugs. One is called rational drug design. Companies such as Agouron, Arris, and Vertex try to design drugs from scratch by studying the structure of targets. Using this approach, both Agouron and Vertex have designed molecules that may gum up the way the AIDS virus replicates. Both now have drugs in clinical trials.
Rational drug design is hard to do, however. So Gilead, Nexagen, Selectide, Isis, and Affymax, among others, are adopting what Ron Tuttle, executive vice-president of Houghten Pharmaceuticals Inc. in San Diego, calls "the accelerated needle-in-a-haystack" method. Each company's technology differs, but they all use tricks of chemistry and biotech to make huge numbers of drug candidates, each with a unique structure, in a single test tube. Put a cloned receptor or other screening target in that brew, and one of those molecular structures will likely bind to the target. Gilead Sciences Inc. in Foster City, Calif., can screen a trillion such molecules in a few weeks. Still, Michael L. Riordan, Gilead's chief executive, says this would be for naught if the company didn't use precisely the right screening targets. "Choosing targets is the most agonizing decision we make," he says.
As the quality of targets improves, some drugmakers are looking backward. Kyowa Hakko, Ajinomoto, and Takeda, all Japanese, want to exploit libraries of microbial organisms they have developed since World War II for making antibiotics. Lacking good screening methods, they're scouting U.S. biotech companies for partners.
ARSENALS. U.S. drugmakers are reviewing their own libraries, too. And they're looking to nature as well. Many modern drugs come from plants: morphine from poppies, aspirin from willow bark, oral contraceptives from the Mexican yam. Nature "produces chemicals chemists would never dream of creating," says Gordon Cragg, director of the National Cancer Institute's natural-products screening program. In rain forests, for example, competition among plants for space, light, and nutrients is fierce. To survive, many species have evolved chemicals that combat insects, viruses, or fungi. Merck has signed a two-year, $1 million deal with Costa Rica's Institute for Biodiversity to get access to these goodies. Eli Lilly, SmithKline Beecham, and Syntex, among others, are also checking out exotic locales and chemical libraries.
Increasingly, the natural chemicals they find will enter automated screening systems such as Oncogene's, where several robots each run more than 100,000 tests a year--a tenfold leap over what entire companies once did. With the right targets to screen, these machines just may help find miracle drugs.