A.M. and P.M. clocks: fruit fly brain has double timekeepers.

A fruit fly relies on a different group of cells to tick out the rhythm to perk up in the morning than it does to boost evening activity after daytime doldrums, report two research teams.

Both teams performed experiments that altered the functions of cell clusters in each fly-brain hemisphere. Although the investigators, one team in France and the other in the United States, took different approaches, both groups pinpointed the same clusters of neurons for the morning-and the evening-activity increases. The two teams' papers appear in the Oct. 14 Nature.

"This is the first assignment of morning-ness and evening-ness to specific cells," comments cluck researcher William J. Schwartz of the University of Massachusetts Medical School in Worcester.

The basic notion of dual control has been around for years, he says. Fruit flies, mice, and plenty of other organisms bustle about in the morning and then slow down until a second peak of activity in the evening. Fruit flies in a lab can anticipate the usual turning on of lights in the morning, starting to move about even in the dark. In the evening, the lab flies likewise get more active shortly before lights-out.

In the 1970s, researchers proposed that such patterns come from dual built-in pace-makers, rather than just one. Each pace-maker would control one of the daily activity peaks.

In each hemisphere of the fly brain, clock genes are active in six clusters of neurons. Francois Rouyer's lab at Alfred Fessard Institute of Neurobiology in Gif-sur-Yvette examined three clusters situated toward the outside of the hemisphere. The researchers worked with mutant flies with no daily rhythms because one of their clock genes, per, doesn't work.

When researchers introduced a working per gene into the lower two of the outside clusters, designated the ventral lateral neurons, the morning activity peak showed up. The evening peak, however. didn't. When the researchers restored gene function in both the upper and lower clusters, both morning and evening activity reappeared.

The same neuron clusters attracted the attention of Michael Rosbash of Brandeis University in Waltham, Mass., and his colleagues. They took a different approach, disabling the upper or lower brain clusters. Those flies with disabled lower clusters didn't show a strong morning peak, and those with disabled upper clusters, called dorsal lateral neurons, didn't behave normally in the evening.

The team also found evidence that the two clocks interact.

"These are really, really satisfying papers," says Steve Kay of the Scripps Research Institute in La Jolla, Calif. He comments that a view of circadian rhythm as a single feedback loop is giving way to a more elaborate vision that includes many interlocking cycles.