Cover Story

The head bobs, eyes flutter closed, muscles slump: It's the classic sign that someone is falling asleep. And, it turns out, it's also a sign that a bird, a dog, an elephant, even a fly are dozing off. In fact, sleep is so universal that most researchers assume it must serve some primordial purpose so vital that it's been preserved across species since the dawn of complex life.

Research backs up that notion. For example, rats deprived of sleep die faster than those deprived of food. Humans, too, have an undeniable drive to sleep-so much so that when we miss it, we make up for the loss by sleeping deeper and longer the next chance we get. Add to that the discovery just six years ago that fruit flies sleep up to 12 hours a day-head bobs, muscle slumps and all.

"It adds fuel to the idea that sleep is something universal that began very early in animal evolution," says University of Pennsylvania sleep researcher and dean of veterinary medicine Joan Hendricks, VMD, PhD, who published the first study demonstrating that the flies have a rest state in a 2000 issue of Neuron (Vol. 25, No. 1, pages 129-138).

That's certainly the most parsimonious conclusion, agrees sleep expert Jerry Siegel, PhD, a professor in the department of psychiatry at the University of California, Los Angeles School of Medicine. But he's not convinced it's correct.

"It might be more a case of convergent evolution," he says, referring to the process whereby different species develop a similar trait independently often for different functions, such as how dolphins and bats separately developed echolocation. "I think that if you look at sleep more as an adaptation to an animal's environment, it makes a lot more sense."

What both camps agree on is that studying sleep in nonhuman animals-particularly fruit flies-will be key to answering the question that has eluded the keenest minds: Why do we sleep?

A theory to fit the data

There are no shortage of theories about why we sleep-from the idea that sleep helps consolidate memories (see article page 58) to theories that it conserves energy or keeps organisms out of harm's way when they are most vulnerable.

Siegel discusses all these in a recent review in Nature (Vol. 437, No. 7063, pages 1,264-1,271), where he argues that any theory with lasting power must account for the incredible variation in sleep patterns across the animal world. There's even a huge range in the amount and type of sleep: Bats and opossums sleep for 18 to 20 hours a day, while elephants and giraffes sleep only three to four hours a day. In general, herbivores sleep less then omnivores, who sleep less than carnivores.

"This makes sense if you think about how these animals live," says Siegel. "If a cat in the wild kills and eats an antelope, what use is spending more time prowling around? It might as well save its energy. For grazing animals, on the other hand, it's dangerous to not be alert for long periods of time. If sheep slept really deeply, they would be extinct."

Add to that Siegel's recent data on marine mammals and, he argues, it's hard to make a case that sleep serves the same vital function for, say, cattle, that it does for rats. Indeed, in an article published in Nature (Vol. 435, No. 1177, pages 424-435), he found that right after a dolphin mother gives birth, neither she nor her newborn sleep for substantial amounts. The babies then gradually increase their amount of sleep to adult levels over a period of months, and the mothers go back to their regular sleeping pattern-one half of the brain at a time. Amazingly, it's the opposite of what researchers see in land mammals, which sleep the most as newborns and gradually decrease sleep time as they age. This finding is a challenge for any theory that argues sleep is necessary for proper brain function and development, says Siegel.

"What we're seeing in marine mammals, that have brains as big and bigger than ours, is that they can go without sleep or much sleep for very long periods of time," he says. "It's a perfectly normal part of their life cycle."

His conclusion, after looking at the comparative data, is that animals adapted a system of sleep to fit their environment and then co-opted that rest time for functions more efficiently accomplished during rest.

However, University of Iowa psychologist and sleep researcher Mark Blumberg, PhD, cautions that researchers need to delve much deeper into comparative research before they can come to any conclusions.

While he agrees that sleep's functions may vary among species, he notes that the data in marine mammals, for example, don't prove that sleep doesn't contribute to brain development in dolphins and whales. Odds are, he says, they've evolved to take care of the needed developmental issues related to sleep before the infant is born. Only a study of these mammals in utero can fully answer the question, he notes.

The comparative data certainly raises endless possibilities in the search for sleep's function, admits Paul Shaw, PhD, a sleep researcher at Washington University in St. Louis. It could be a case of convergent evolution or, that through the course of evolution, some species have dropped sleep's original function. He believes that studying the comparatively simple fruit fly will allow researchers to tease apart some of the possibilities.

Hendricks agrees: "The leap that the flies let us make is that we finally can look at [sleep] not as a mysterious black box, but we can break it down into simple components that we can examine at a molecular level."

That's because researchers have mapped all the genes of Drosophila melanogaster-the fruit fly's scientific name. Now, they can look for genes related to sleep mutations and then find out what those genes do. In addition, the quick breeding and short life span of Drosophila allows researchers to create new strains with different traits-such as sleeping too much or too little-and then replicate their findings in a matter of days.

"My view right now is that there are so many bad theories out there, the world doesn't need another one," he says. "I've decided to collect data in the meantime and hope that something will become apparent."

Others are more cautious about the work in Drosophila. Blumberg, for one, thinks it's going too far to think that work in flies will lead to the discovery of a single, universal function of sleep.

"They have a very different nervous system, anatomy, physiology, you name it," says Blumberg. That's not to say that the work isn't worthwhile, he adds. Theories coming out of the fruit fly work will certainly help researchers focus their work in new ways.

The price of plasticity

One of those theories comes from University of Wisconsin psychiatrist Guilio Tononi, MD, PhD, a co-author with Shaw on the fruit fly sleep paper. He and his colleague Chiara Cirelli, PhD, believe that sleep originally evolved to help the brain recover from a long day of learning.

"Sleep is the price we pay for plasticity," explains Tononi. While we're awake, the connections between our nerve cells, known as synapses, are constantly changing-strengthening and weakening through active and passive learning. The same goes for any creature with a nervous system. The net effect is a brain full of stronger synapses that not only use more energy but also take up more physical space.

"By the end of the day we've saturated our ability to learn by maxing out the brain's energy and space capacity," says Tononi. "That's when sleep kicks in."

He argues that slow-wave sleep-the nondreaming part of sleep that we spend the most time in-weakens all the brain's synapses, preserving the relative strength of connections but freeing up space and using less energy.

Based on theories of how synapses form, slow waves look like the ideal paradigm to reduce synaptic strength, says Tononi, who created a computer simulation that supports this idea.

So why, then, does a fruit fly with its arguably smaller brain and learning capacity need more sleep than we humans? Tononi argues that you have to think in relative terms. A fruit fly likely strengthens a larger fraction of its neurons in a day than a person and therefore requires more sleep.

Right now, this is all speculation. However, Tononi and his colleagues are looking for molecular mechanisms that control sleep in the fruit fly that could support their theory. Even Siegel admits he's interested to see what the work in flies produces.

It's currently more a matter of faith than science whether researchers lean toward Siegel's claim that sleep is a common solution serving multiple functions or the theory that there's some underlying role of sleep common to all species.

Shaw, for one, can't shake the idea there's some universal and vital function of slumber.

"What sold me," says Shaw, "is that sleep is costly. When an animal sleeps it's not taking care of its young, it's not protecting itself, it's not eating, it's not procreating." That vulnerability, he says, once led sleep research pioneer Alan Rechtschaffen, PhD, to say "If sleep doesn't serve an absolutely vital function, it is the biggest mistake evolution ever made."

Beth Azar is a write in Portland, Ore.