A commonly used drug-screening method that relies on the enzyme that gives fireflies their glow may be flawed. Researchers at the U.S. National Institutes of Health (NIH) have found that for some compounds--including one now in clinical trials for cystic fibrosis and other conditions--the luminescence is just an artifact.
For decades, researchers have used firefly luciferase--the protein that makes the insects light up--as a so-called reporter enzyme. Take the example of cystic fibrosis: Some forms of the disease result when a genetic "stop sign" prevents the body's cells from making a complete version of the cystic fibrosis protein. An effective drug would allow cells to ignore this stop sign so that the complete protein can be produced. To screen for such drugs, scientists attach firefly luciferase DNA to the end of the cystic fibrosis gene. If a drug works, the complete cystic fibrosis protein will be made--and scientists will see a glow.
But Douglas Auld, an expert on high-throughput drug screening at NIH's Chemical Genomics Center, has found a problem with this approach. His team looked at an experimental drug called PTC124, which the firefly luciferase assay had identified as a promising candidate for treating stop-sign forms of cystic fibrosis and which is now being tested in clinical trials. The researchers found that PTC124, and many compounds related to it, bind to the firefly luciferase protein, causing it to glow brighter than normal. And that makes the drug appear more effective than it actually is, says Auld. When the researchers tested PTC124 using another luminescent enzyme with a different structure, they saw no glow at all.
Co-author James Inglese, the deputy director of the Chemical Genomics Center, says the work emphasizes the importance of using more than one assay when searching for drugs. Most compounds identified by the firefly luciferase screen do not interact in this way with luciferase, he says, but it is important to make sure.
John Babiak, senior vice president of drug discovery technologies at PTC Therapeutics in South Plainfield, New Jersey, which is developing and testing PTC124, shrugs off the NIH work. "You get false positives and artifacts" in any assay, he says. Babiak says his company has found that PTC124 looks promising in a number of other assays that don't use firefly luciferase, as well as in animal tests and clinical trials. The fact that PTC124 binds to luciferase, he says, "has nothing to do with whether it's a good drug." (The NIH researchers do not question the clinical and animal results on PTC124.)
Keith Wood, head of research at Promega Corp. in Madison, Wisconsin, and an expert in bioluminescent assays, agrees that the NIH work should not undermine PTC124. But the new study, he says, provides an important caution for researchers using any enzyme, not just firefly luciferase, to report on biological activity. "No tool for screening is perfect," he says.