n the early days of bio-tech, all eyes focused on the techniques of "recombinant DNA": splicing together bits of DNA from different sources. These 30-year-old genetic engineering methods are now the basis of a multibillion-dollar market for protein-based drugs. Today, another class of biotech drugs is emerging from the lab, but the technology for these "antisense therapies" isn't new-it dates back to 1978, just a few years after the first gene-splicing experi-ments. A few persistent researchers have shepherded it down a long, bumpy road.

In the early 1970s, Paul Zamecnik (pronounced ZAM-es-nick) was studying a cancer-causing chicken virus that transmits its genetic information via RNA, a chemical cousin of DNA. Zamecnik and his colleagues at Massachusetts General Hospital found that, as the virus replicated, its RNA looped around on itself. They speculated that if they could block this step, they could stop the bug in its tracks. So they constructed a short piece of DNA designed to stick to the virus's single strand of RNA and thereby gum up its works. The RNA encoded the virus's proteins; functionally, it made sense, so the researchers called it the "sense" strand. The DNA molecule (called an oligo-nucleotide) was its chemical opposite-the "antisense." Zamecnik mixed the designer DNA snippet with infected chicken cells, and voil-no cancer. He and colleague Mary L. Stephenson suggested that antisense molecules could be used to treat all sorts of infections-as well as cancer-by preventing RNA from being translated into the proteins the invaders need to live.