bout one in 2,000 adults suffers from the debilitating disease called narcolepsy. It usually begins in adolescence, and includes extreme daytime sleepiness and muscle weakness, as well as vivid hallucinations just before or after falling asleep. In humans there is no link to simple genetics; identical twins are no more likely to share the condition than unrelated strangers. Dogs, however, can inherit narcolepsy as a simple recessive disease. And that has enabled two teams independently to identify part of the cellular pathway that controls sleep.

Emmanuel Mignot, of the Center for Narcolepsy at Stanford University School of Medicine, led a team that cloned and sequenced a cellular receptor that is defective in narcoleptic dogs (see L. Lin et al., Cell, 98[3]:365-76, 6 August 1999; with 32 citations this period, the paper is currently ranked at #12). Masashi Yanagisawa, at the Howard Hughes Medical Institute, and his group showed that mice in which the receptor's ligand is missing show symptoms identical to human (and canine) narcolepsy (paper #9). Together these reports and the research they have triggered hold out hope for new treatments for narcolepsy and, perhaps paradoxically, insomnia.

Mignot's group worked with two families of dogs, Doberman pinschers and Labradors, that inherit narcolepsy. They adopted a strategy called positional cloning, which has been of enormous help in mapping human genes. Man's best friend, however, is practically unknown genetically. So having found a marker linked to the condition, the team had to go through all the work of creating a library of canine DNA with which to localize the marker. After a ten-year search, using information from human chromosome 6, they located the sequence on canine chromosome 12.

The only gene known to lie within that region was Hcrtr2, a G protein-coupled receptor with an affinity for small neuropeptides called orexins that were associated with eating rather than sleeping. Nevertheless, the presence of Hcrtr2 in the right place made it worthwhile sequencing the gene. Every narcoleptic Doberman shared the same mutation, which affects the way the gene is transcribed and spliced. The Labradors had a different mutation. But in all cases the receptor molecule would be extremely abnormal.

Yanagisawa came at the problem by a completely different route. He specializes in finding the signals of G protein-coupled receptors by inserting so-called orphan human receptors—for which no known ligand is known—into cells and then exposing those cells to tissue extracts. That was how he discovered orexins (from the Greek for appetite), which initially seemed to be involved only in eating and satiety. Later studies showed that orexins also seem to affect arousal and waking.

Yanagisawa's team destroyed the function of the orexin gene in mice, and then watched them round the clock, using infrared cameras to see the mice at night, when they are normally active. Just like narcoleptic dogs, and people, the knockout mice showed unexpected brief periods of rest when they should have been awake. Further studies of their brain and muscle activity revealed that these were not the result of seizures or anything similar. They were narcoleptic episodes.

A year ago Mignot's group published their results of the search for a narcolepsy gene in humans (see C. Peyron, et al., Nature Medicine, 6[9]:991-7, September 2000). This showed, somewhat surprisingly, that there was nothing wrong with the orexin gene in narcoleptic patients, but for some reason the cells that normally make it in the brain were not doing so."We think something specifically kills the cells that make hypocretin," said Mignot. "We don't know how or why, but it's most likely an autoimmune disease."

On the prospects for benefits from this research, Mignot is clearly optimistic. "You are happy when you make a discovery," he said, "but you are really happy when you make a discovery with therapeutic possibilities.". Current treatments for narcolepsy include stimulants that keep people alert during the day, but these do not deal with other symptoms and do nothing about the underlying cause of the disease. With these discoveries in hand it should be possible to design drugs that will replace the missing orexin and help narcoleptics stay awake. Orexin-antagonists might likewise make better sleeping pills.

One intriguing aspect of these studies is that they suggest a link between eating and sleeping. It is reasonably common to feel sleepy after a meal, and wide awake when hungry. Could both states be related to the level of orexins? And what are the implications for drug companies seeking to control patterns of sleeping and eating?