Why do we sleep?


Tim Steury

Go without sleep and the effects will be obvious and progressively dramatic. Your attention and learning abilities will dwindle quickly. Your metabolic rate will shoot up. Your body will be unable to regulate its temperature. At some point you will start hallucinating. We don't know how long a human can go without sleep, but a rat drops dead, of infection, after three sleepless weeks.

In spite of these dramatic symptoms, however, why we sleep is far from clear, particularly when the question is framed by our evolutionary history. Why would an animal with few innate defense skills to start with further place itself at risk by descending into unconsciousness for better than one-third of every day? Under such circumstances, sleep makes no sense--unless it serves a greater need.

Many reasons have been proposed, but most are no more satisfying than one offered by Aristotle, that sleep served to cool the brain. Many basic physiological needs can be met simply with rest. And besides, sleep does not seem to stem immediately from bodily needs, since the first symptoms of sleep loss show up in the brain.

In short, the average person spends 27 years asleep, so unless all our ideas about evolutionary expediency are wrong, sleep must serve some very important function! James Krueger, a sleep researcher at Washington State University, has an explanation that might persuade you to set your alarm just a little later.

About an hour and a half after you fall asleep, you enter a phase called REM

sleep, so called because your eyes dart rapidly during this phase, reflecting the activity of your brain. French sleep researcher Michel Jouvet called REM "paradoxical" sleep, for even though you are deep asleep, your brain waves are hardly distinguishable from the preceding light transitional state.

Until the sleep research pioneer Nathaniel Kleitman published Sleep and Wakefulness in 1939, the general scientific perception was that nothing much happened during sleep. Sleep was a passive state, forced on the nervous system by the separation of the brain from the rest of body. However, Kleitman's notion that the brain attended to very important business during sleep was strengthened by the discovery of REM sleep 1953 by Eugene Aserinsky, a doctoral student in his lab.

But the discovery of REM sleep simply complicated things, raising a question parallel to the big "why"--what in the world is the brain doing during REM? Even though the body is limp with sleep paralysis, the brain is just as busy as it is while awake.

Krueger believes that REM sleep is when the brain does its housekeeping. All day long the brain is busy processing information, synapses firing on and off, neurons constantly exchanging information and rearranging relationships with their neighbors. Here comes a new thought or sensation, and the appropriate neurons hurriedly form new networks. Information in, information out. In, out. On, off. Hello, goodbye. Thousands of times a day.

Experience actually alters the anatomy, or "microcircuitry," of the brain. Memories, in other words, take on an actual form. If this process goes unchecked, says Krueger, the brain will eventually completely transform itself.

Think: How many memories did you form today? How many fleeting impressions? What did you read, hear, touch, taste, feel? Your brain is much different in the evening than it was 16 hours earlier, its circuitry thoroughly reworked. Somehow, reasons Krueger, it must put itself back in order. For much of who you are, physiologically as well as behaviorally, is genetically patterned. If your brain never gets a chance to step back and ask itself "who am I?" well, then, who IS it after a few days?

Sleep, in other words, affects the "synaptic plasticity" of the brain. It reorders and restores the brain's "synaptic superstructure." It keeps you who you are. It serves to preserve both acquired and inherited behaviors.

The most intriguing thing about Krueger's hypothesis is that it is testable. Only one other lab in the world has such a tangible theory; it is at Stanford and uses a hibernating hamster as a model system. Krueger, on the other hand, uses a rat with half its whiskers shaved.

Rats rely on their whiskers for spatial location. Shaving one side of those whiskers throws their orientation way off.

But immediately, the somatosensory cortex on the side of the brain opposite to the shaved whiskers begins reorganizing itself. This process begins within a few hours of the shave and continues for several days.

What Krueger and the rest of the lab watch for is a pattern of molecular markers, specified molecules involved in the formation of synapses. But what they're particularly interested in is the effect of sleep on that reorganization.

If the semi-whiskered rat is kept from sleeping, the changes seem to be different from the changes in a well-rested rat. If the lack of sleep blocks the synaptic reorganization, then the role of sleep has been further defined. This indeed seems to be the case as the results of the experiment start to materialize.

There's another way to test this idea that sleep tidies and reorganizes. This involves the relation between memory and the hippocampus. Among its other roles, the hippocampus is a short-term memory structure as well as being vital to our understanding where we are.

Gina Poe trains her rats on a simple rectangular runway, with six food stations around the perimeter. Even though they are otherwise identical, only three of the stations actually offer food. The rats' task is to memorize where the food is, which they do after three or four days.

Poe is an electrophysiologist, which means she is very good at measuring electrical impulses in the brain, the sound of synapses, of the brain at work. What sets her lab apart is that it is one of the few in the world capable of monitoring and analyzing groups of over 30 neurons at a time.

Most such monitoring systems use single electrodes, which are sufficient for studying the cortex, where the neurons are relatively far apart. The signals of individual neurons are teased apart by cutting off the amplitude of the signal, and thus losing the fine-tuning necessary for telling one neuron from another--as is necessary in the hippocampus, where the neurons are more tightly packed.

Poe works with an electrode comprising four wires, providing four channels, which are then analyzed with a massive bank of eight Sun System computers, an extraordinarily sophisticated process that Poe learned as a post doc in the lab of Bruce McNaughton at the University of Arizona.

Poe and Krueger are working together with her system to elaborate the synaptic superstructure hypothesis. His expertise in brain biochemistry combined with her electrochemical talents should paint a convincing picture of REM's reorganizational role.

But ideas Poe has pursued lend further elegance to the housekeeping model. Somewhere along the way, she observed that when a rat learns to orient itself, which can take several days of training, immediately before everything clicks for it, just before the "aha," its REM sleep increases.

Closer inspection revealed that the rat's hippocampal neurons were firing in a manner consistent with long-term potentiation, the strengthening of memory. Also, they were firing at its amplitudal peak.

Neurons firing at their amplitudal trough, on the other hand, result in de-potentiation, or paring of memory. So what happens during REM sleep? she wondered. Do cells fire mostly at their peak, strengthening memory? Or do they fire mostly at their trough , breaking apart synapses and perhaps refreshing the hippocampus for learning something new?

The answer, it appears, is both. Neurons associated with memories already well consolidated no longer need the short-term capability of the hippocampus, so they de-potentiate. Somehow those memories are transferred to the cortex. Cells associated with new memories, however, fire during REM at their peak, strengthening, potentiating.

In other words, again, the brain during REM both cleans out and puts things in order. It both strengthens and pares.

Naturally, consolidation and paring of memory is likely only part of sleep's function. Given the severe physiological effects of sleep deprivation, sleep surely has a restorative effect, says Poe. What exactly it is, we're still far from understanding, and much remains for the curious sleep researcher to explore.