Emmanuel Mignot's recent finding might be compared to the discovery of insulin.

The Parisian-born associate professor of psychiatry, who directs the Stanford Center for Narcolepsy, and scientists in his lab spent 10 years mapping a specific gene in narcoleptic Dobermans. They took another year to apply those findings to studies of brain tissue from deceased human narcoleptics. They found thousands of hypocretin cells--a small peptide--in healthy brains. But diseased tissue had virtually none.

Now that the cause of some forms of narcolepsy has been identified, Mignot predicts that help soon will be on the way for an estimated 125,000 to 200,000 affected Americans, many of whom suffer paralyzing bouts of weakness and hallucinations.

Designing a drug to treat narcolepsy by replacing the missing hypocretin is "totally doable," Mignot says. But delivering it will be a challenge. Unlike insulin, which can be taken intravenously, hypocretin must go directly into the brain. So, while a substitute drug will be the immediate remedy, brain-cell transplantation likely will be the best long-term treatment.

Mignot's team reported its findings in Nature Medicine in September, the same month a UCLA team published similar findings in the competing journal Neutron. To reward his lab, Mignot took all 30 scientists to a hotel on Hawaii's Big Island, where they swam with turtles and considered the future direction their research might take.

"When you get a discovery like this, it's important to try to reflect," he says.