Biology 202
1999 Second Web Reports
On Serendip

On the Function and Evolution of Sleep

Jason Bernstein

If physiologists devoted the most research time to behaviors humans engaged in the most, we would probably have a full understanding of the biological purpose of sleep. After all, humans, with the exception of most college students, spend one third of their lives in a somnolent state. Despite its fundamental role in human and animal life, sleep is, even in an age when neuroscience has reduced many behaviors to neurological mechanisms, still quite mysterious. What processes are taking place during sleep that benefit the organism? Why spend so much time in an unresponsive and vulnerable state? That these questions haven't been definitively answered is really not a function of a lack of effort on the part of scientists, but rather of the difficulties inherent in studying sleep. It is easy to observe the harm that is done to a human or animal deprived of sleep. A rat prevented from sleeping will lose the ability to maintain body temperature and die in about three weeks, showing no evidence of physiological damage (1). In humans, sleep deprivation impairs thinking and suppresses the immune system. But why this deterioration take place is less clear, and the object of disagreeing theories. This essay will try to explain the function of sleep based on what is now known, and attempt to shed light on the reasons and mechanisms for its evolution.

An investigation of the control, and perhaps the origin, of sleep might best begin with the hypothalamus, a flat, horizontal structure in the brain that is known to play an important role. Damage to the back portion of the hypothalamus causes somnolence, indicating that when intact, it sustains alertness (1). In contrast, damage to the front part induces insomnia, suggesting that it provides the stimulus for sleep. The hypothalamus is also involved in temperature regulation, and it has been hypothesized that sleep evolved from a more primitive thermostat-like mechanism. In 1995 researchers found the first evidence of neurons involved in both temperature regulation and sleep (1). A team discovered neurons in the front part of the hypothalamus of cats which, when warmed by two degrees Celsius, fire more rapidly. These neurons also increase their firing frequency when the cats sleep (1). The researchers attributed a dual role to the neurons in this region. These dual functioning neurons may be living evidence of an ancient transition from mere temperature maintenance to actual sleep.

The origin of sleep seems to be closely related to the evolution of mechanisms of enhancement and maintenance of synaptic capability. Roffwarg, Musio, and Dement conjectured over thirty years ago that the function of spontaneous, repetitive excitations of neural circuits during REM sleep in human embryos is to facilitate circuit development and maintenance (2). This concept provided the foundation for the "dynamic stabilization" (DS) paradigm of neural circuitry. According to this model, frequent synaptic activation enhances synaptic strength in neural circuits storing inherited information, or "phylogenic memories," and information acquired through experience, "ontogenetic memories." (2) DS can occur either through regular functional use or by way of spontaneous oscillatory neural activity. The spontaneous activations do not actually trigger the performance of a function , and are hence referred to as "non-utilitarian." (2) This is the type of DS that is thought to be the main purpose of spontaneous oscillations in brain regions during sleep, and is the type that will be discussed here.

DS is required for circuit development, maturation, fine-tuning, and maintenance in the embryos of warm-blooded animals (2). In the nervous systems of human fetuses, almost all DS occurs during sleep states, primarily during REM sleep. Similar processes have been attributed to infant, child, and adult sleep. Human infancy and childhood are periods of intense new learning and cerebral development, and are accompanied by large amounts of sleep, and thus extensive self-activations in brain circuitry. While there is no conclusive proof, a large body of evidence suggests that DS during sleep maintains the phylogenetic and ontogenetic memories in all humans (2). Recently, researchers at the University of Chicago have found strong evidence of dynamic stabilization in birds. In a study of young male songbirds, Daniel Margoliash and his colleagues concluded that at least some song learning and refinement occurs while the birds sleep (4). They based this claim on measurements comparing the activity of song-specific neurons in the brains of waking and sleeping zebra finches. They found that in sleeping finches, a recording of each bird's own song caused auditory signals to flow freely between the brain areas that control singing. When the birds awakened, however, the signal flow ceased completely. Margoliash proposed that during sleep the wide-open signal gate allows the birds' brains to refine the firing patterns that yield the song, "an 'off line' learning similar to the memory strengthening that some neuroscientists think may occur during sleep when rats learn mazes and humans learn motor tasks." (4)

According to J.L. Kavanau of the University of California, DS during sleep may have evolved from DS during rest (most cold-blooded vertebrates do not sleep--they rest, despite having fairly complex brains) (2). As brains became more complex, the need for DS of expanding memory stores also increased, conflicting with the neural activities of rest. The need to resolve the conflict between extensive sensory input processing and increasing dynamic stabilization needs provided selective pressure for the evolution of sleep. Just as sleep probably evolved from rest, REM (rapid eye movement) sleep probably has its roots in NREM (non-REM) sleep, which still accounts for the majority of adult sleep (2). Sleeping cold-blooded animals, such as reptiles, do not exhibit REM sleep, which exists only in warm-blooded animals and corresponds with a highly developed forebrain and cortex. During NREM sleep, skeletal muscle tone is merely reduced, while during REM sleep it is absent. The reduction in muscle tone is believed to prevent sleep-disruptive movements during motor circuit reinforcement. As higher metabolic rates and more intense DS evolved, more powerful inhibitory mechanisms were needed to maintain sleep, thus favoring the more complete muscular suppression of REM sleep (2). It is likely that the eye movements of REM are unrelated to the psychology of this state. While REM discharges of motor neurons controlling large muscles would awaken the sleeper if they were not inhibited, eye movement does not cause arousal from sleep. "Nature, being frugal, did not develop an unneeded inhibitory pathway to block eye movements. The biological significance of eye movement activity during REM probably lies elsewhere." (3)

Sleep is not only a benefit to neural functioning. Carol A. Everson of the University of Tennessee has found evidence of something that caring mothers have known all along: that sleep benefits the immune system (1). She conducted experiments showing that rats deprived of sleep have high numbers of bacterial pathogens that are normally suppressed by the immune system. Everson is fairly certain that the bacteria eventually kill the rats. A deficient immune system causes the exhausted rats to fail to develop fever, which would be the normal response to infection. Another non-neural force that may have driven the evolution of sleep may have been its capacity to allow an organism to "lay low" during periods when waking activity was unlikely to yield much survival benefit. In humans, for whom vision is the dominant sensory window to the world, nighttime was, during the period of evolutionary adaptation, a necessarily unproductive time. It was dark, so we couldn't effectively hunt, make things, or travel. For any organism, nighttime activity burns energy and hence requires extra fuel consumption. If the animal isn't productive during this period, that activity is wasted. For many animals, safety is another concern-sleep allows them to remain in the same hiding place for the whole night, where they are much less accessible to predators than if they were roaming around. Of course, some organisms have adapted to daytime sleep/rest, taking advantage of sensory abilities that give them an advantage in the dark.

From what we know, sleep, essentially, is a state organisms assume that prevents their brains from being diverted from the task at hand: enhancing and reinforcing synaptic efficacy. More complex brains require more of this maintenance, and hence larger-brained organisms with higher metabolic rates have been forced to develop more potent inhibitory mechanisms. What needs to be inhibited? Both distracting sensory input and sleep-disruptive motor output during DS of motor circuits are blocked. Interestingly, the reasoning behind sleep function is somewhat circular--one of its main duties is to assure its own continuance . The notion that waking processing of sensory input is not fully compatible with simultaneous neural activities that serve to maintain memory circuits should not be interpreted as maladaptive, according to Kavanau (2). Rather, it suggests that some of the circuits employed in sensory reception and processing also function to establish and maintain memory, and that both functions cannot be achieved simultaneously. Sleep circumvents this limitation, and is thus a well adapted behavior.

WWW Sources

1)Scientific American "In Focus"

2)Sleep and Memory: Evolutionary Perspectives

3)REM Sleep = Dreaming: Only a Dream

4)Birds May Refine Their Songs While Sleeping" ,Science Magazine article

Additional Resources

Discussion on Sleep Evolution ,Scientific American article

The Journal "Sleep"




| Course Home Page | Back to Brain and Behavior | Back to Serendip |

Send us your comments at Serendip
© by Serendip 1994- - Last Modified: Monday, 07-Jan-2002 14:39:48 EST

This paper reflects the research and thoughts of a student at the time the paper was written for a course at Bryn Mawr College. Like other materials on Serendip, it is not intended to be "authoritative" but rather to help others further develop their own explorations. Web links were active as of the time the paper was posted but are not updated.

Contribute Thoughts | Search Serendip for Other Papers | Serendip Home Page