Box Extension 15.3 Sleep

David S. Garbe

Why animals sleep continues to be one of the most elusive and mysterious questions in biology. Sleep, nonetheless, is found widely among animals—from relatively simple phyla such as worms (e.g., Caenorhabditis elegans) to higher-order groups such as humans—suggesting that sleep is a required, evolutionarily conserved behavior. Yet from certain perspectives, sleep could be considered disadvantageous to an organism’s overall survival and fitness because while animals are sleeping, they are not immediately able to eat, mate, or protect themselves from predation. Still we as humans sleep for approximately one-third of our entire lives. Most scientists conclude that there must be an overriding benefit of sleep that compensates for the lack of interaction with the surrounding environment. In fact, sleep is crucial for survival. Long-term sleep deprivation is lethal to rodents and fruit flies (Drosophila melanogaster). Moreover, disturbances in sleep and its regulation are associated with several chronic human disease states, including insulin resistance and diabetes, fatal familial insomnia, shift-work disorder, certain neurodegenerative diseases, and major depressive disorder. Box Extension 15.3 discusses functions of sleep and mechanisms of its regulation.

In the broadest behavioral sense, sleep can be defined simply as a quiescent state of inactivity. A more elaborate and specific definition of sleep is that it is a homeostatically regulated period of reduced movement and sensory responsiveness (i.e., increased arousal threshold) accompanied by a stereotypical body posture. In mammals, sleep is commonly identified through the use of electrophysiological techniques such as electroculography (EOG; a technique that records electrical activity from the eyes), electromyography (EMG; a technique that records electrical activity from muscles), and electroencephalography (EEG; a technique that records overall electrical activity from the brain). Over the past 60 years, using mainly these electrophysiological, as well as anatomical, techniques, research in mammals has identified areas of the brain involved in sleep–wake transitions. These studies have revealed the midbrain, forebrain, and brainstem—which include noradrenergic, serotonergic, and cholinergic neurons—as the main regions controlling sleep and wakefulness. EEG studies in mammals have also revealed that sleep is not a homogeneous state. On a gross level, human sleep is divided into two types: Rapid Eye Movement (REM) sleep and non-REM (NREM) sleep. NREM sleep, also called slow wave sleep (SWS), is further divided into four stages, each characterized by distinct stereotypical EEG patterns of brain activity (Table A, Figure A). Throughout a night of sleep, humans transition between and cycle through these types and stages of sleep, each cycle lasting roughly 90–110 min (see Figure A). Unfortunately, the biological functions of REM and NREM sleep remain obscure.

Figure A Human sleep cycles during a typical night. (1) A subject cycles through sleep stages roughly every 90-110 minutes (vertical dotted lines). Stage 1 sleep is associated with rapid eye movements (red). (2) Expansion of the first hour of sleep shown in (1), with sample EEG recordings.

Table A General characteristics of human sleep stages

Sleep Stage

EEG* actvity


REM (Rapid Eye Movement) sleep

Small-amplitude, high-frequency oscillations occur as a result of desynchrony in firing between cortical neurons. EEG pattern resembles that of waking EEG but is accompanied by loss of muscle tone and the presence of rapid eye movements.

Muscle tone is weak, but twitches in facial and digital muscles are observed. Pronounced fluctuations occur in cardiac and respiratory rhythms and core body temperature. Penile erections and clitoral engorgement are common.

Human non-REM (NREM) I, II sleep (collectively known as SWS-I** in other animals)

Larger-amplitude, lower-frequency oscillations occur and are thought to reflect the transition of cortical neurons to a more synchronous and bursting firing pattern.

Decreases occur in body temperature, blood pressure, and heart and respiratory rates. Increased release of growth and sex hormones occurs.

Human NREM III, IV sleep (collectively known as SWS-II** in other animals)

Largest amplitude, lowest frequency oscillations (termed delta waves) are observed. Synchrony in neuronal firing is thought to be at its highest.

This is considered the deepest part of sleep. Also known as delta sleep. Muscle tone progressively decreases.

*EEG =Electroencephalogram.
**SWS = Slow wave sleep.

Members of the general public sometimes think that sleep is a passive, inactive process that simply allows animals to become refreshed. However, research reveals that this is far from the truth. Sleep is a very dynamic physiological state, and many cellular processes become active while animals sleep.

Transcriptomic (gene-expression) studies performed in several model organisms have revealed discrete functional gene groups associated with either the state of wakefulness or sleep. For example, many genes involved in detoxification, stress and immune responses, and synaptic plasticity are expressed principally during wakefulness. By contrast, genes involved in the metabolism of macromolecules (carbohydrates, proteins, lipids) and energy production are expressed principally during sleep. Additionally, “sleep-on” neurons (active while animals sleep; versus “wake-on”) have been identified. These neurons are often GABAergic in nature and function as inhibitory neurons that dampen the arousal system. Taken together, these data suggest that sleep is an active state with specific functions that probably often benefit the organism.

Over the years, several hypotheses have been proposed and tested to address “why animals sleep” (each of these hypotheses is by no means mutually exclusive to the others). One hypothesis posits that organisms sleep to conserve energy and to reduce metabolic load. In this way, the body “shuts down” to prevent the buildup of cellular metabolites and reactive oxygen species (ROS; see Box 8.1 in the book), which could damage cells. Increasing evidence also suggests that sleep is involved in the consolidation of memory and the maintenance of synaptic homeostasis. As animals encounter novel experiences and form complex memories, their neurons are constantly making new connections. It has been suggested that sleep helps consolidate the most important occurrences and homeostatically prune back saturated synapses, thereby “making room” for the following day’s events. In support of this hypothesis, it has been demonstrated that as waking experiences become more vigorous (i.e., with increasingly intense learning and adaptation), the need to sleep also increases (see the discussion of the homeostatic mechanism below). Lastly, it has been proposed that sleep functions to maintain and/or enhance the immune system. Whether sleep functions to serve any or all of these functions is still being debated and studied.

So how does an animal’s regulatory apparatus determine when and for how long the animal should sleep? It has been proposed that the transition from wakefulness to a quiescent state of sleep is tightly regulated by two mechanisms: (1) a circadian mechanism (Process C) and (2) a homeostatic mechanism (Process S). The circadian mechanism functions to restrict sleep to the time of day that is ecologically appropriate, whereas the homeostatic mechanism provides input on how the need for sleep accumulates during the day (Figure B). It has been proposed that Factor S (currently unknown) accumulates while animals are awake and intensifies sleep need. As an organism sleeps, this factor is cleared from the body, thereby homeostatically reducing the impetus to sleep. Although the circadian and homeostatic processes may develop independently, their interaction ultimately determines the timing, duration, and quality of sleep and wakefulness. In humans, the combined action of the circadian and homeostatic mechanisms generates a feeling of fatigue late in the day, resulting in sleep during the night.

Figure B Two process models of sleep regulation Process C (blue line), the circadian mechanism, functions to restrict sleep to the time of day that is ecologically appropriate. This mechanism oscillates with a period of approximately 24 h and is controlled by the animal’s internal clock. Process S (green line), the homeostatic mechanism, reflects accumulation of sleep need during the animal’s awake period and reduction of sleep need during sleep. It has been hypothesized that while an animal is awake, an unknown Factor S accumulates and intensifies sleep need. Factor S is proposed to continue to accumulate during sleep deprivation. (After Borbely 1982.)

As stated earlier, sleep could simply be defined as a period of inactivity. Said another way, sleep could be described as the absence of wakefulness. Therefore it is also critical to define wakefulness and provide a brief overview of the wake-promoting systems in our brains. Wakefulness is a complicated and complex behavior that can be defined as a state of consciousness in which an animal engages in coherent and cognitive responses to ever-changing external and internal environments. As might be expected, there are many different neuronal circuits and systems that promote wakefulness, each contributing in a different way to this overall state of consciousness. However, none of these circuits or systems appears absolutely required for the generation or maintenance of a wakeful state. Studies from the 1990s showed that hypocretin/orexin, a neuropeptide produced by a group of cells in the lateral hypothalamus, plays a central role in the mechanisms that promote and maintain wakefulness. In mammals, mutations in the hypocretin/orexin gene (or its receptor) have been shown to cause narcolepsy-cataplexy (i.e., brief episodes of muscle weakness).

Although previous research has defined the regions of mammalian neural anatomy important for controlling sleep and wakefulness, less is known about the molecular and genetic factors that govern and regulate sleep. Recently, genomic microarray studies (see Chapter 3 in the book) have helped define the genome-wide benefits and consequences of sleep and sleep loss, respectively. Moreover, twin studies in humans as well as forward genetic screens in simpler animals have clearly indicated that sleep, like many other behaviors, can be affected by mutations in just a single gene. Nevertheless, the fact that only a few unique “sleep” genes have been identified is consistent with sleep being a complex trait controlled by many genetic factors.

To better understand the genetic and molecular mechanisms of sleep and its regulation, model organisms such as the fruit fly (Drosophila) are being used (Figure C). Interestingly, evidence indicates that mammalian and fruit fly sleep share many behavioral and genetic similarities. Thus candidate genes for sleep identified in fruit flies are likely to have relevance to humans and human disease. However, since the basic electrophysiological measures used in mammalian EEG studies are not relevant in fruit flies, behavioral criteria must be used to identify and study sleep in flies. For example, the following standards, as defined by Scott Campbell and Irene Tobler in 1984, are used to establish a sleeplike state in fruit flies. Sleep should be (1) controlled by the circadian clock (see earlier), (2) homeostatically controlled (see above), (3) characterized by a reversible absence of voluntary movement, and (4) accompanied by an increase in arousal threshold (i.e., a stronger stimulus is required to obtain an organismal response). Sleep experiments in fruit flies are commonly done using Drosophila Activity Monitors (DAMs).

Figure C Experimental setup for studying sleep in fruit flies (Drosophila) Individual experimental flies are placed in small glass tubes and loaded into a Drosophila Activity Monitor (DAM). In a standard experiment, multiple monitors are placed in an incubator containing a strict light–dark cycle and kept at constant temperature. Activity is monitored through the use of an infrared (IR) beam. When a fly crosses the beam, the IR light is interrupted, and these interruptions are recorded by a computer. A fly is considered to be in a restlike (sleep) state when the IR beam remains unbroken for more than 5 min at a time. Generally, both overall activity and sleep are monitored over 3–5 consecutive days. Data are then averaged over a 24-h period and plotted on a computer using dedicated software packages. More recent technology monitors fly movement (and thus sleep) in real time using time-lapse video (data not shown).

This system uses an infrared (IR) beam to record activity and movements of flies housed in individual tubes (see Figure C). A fly is considered to be in a sleeplike state when the IR beam remains unbroken for greater than 5 min. Fruit flies offer many advantages in sleep research, including ease of handling, the ability to use large numbers of individuals in each experiment, and the ability to perform genetic screens in hopes of identifying novel genes that regulate sleep at a basic biological level. However, fruit flies do not appear to have different types of sleep (e.g., NREM vs. REM) and therefore cannot be used to identify the relative advantages or disadvantages of the various types or stages of sleep.

Current human genetic studies using techniques such as Quantitative Trait Locus (QTL) and Genome-Wide Association Studies (GWAS) are making strides in our understanding of a select number of neurological sleep disorders and studies. Simultaneously, model systems such as fruit flies and zebrafish (Danio rerio) are being used to identify genetic loci important for sleep regulation. Whereas scientists are beginning to unravel the mystery of why animals sleep, as well as the molecular and genetic factors that control entry into and maintenance of this behavioral state, many unanswered questions in the field of sleep physiology remain. These include: Why do mammals have different types and stages of sleep that proceed through temporally defined and restricted cycles throughout the night? What are the benefits of these various types and stages? Why don’t invertebrates have them? How does the body know how much sleep is required? How is the homeostatic mechanism of sleep controlled? Future genetic and physiological studies are likely to answer many of these questions in coming years.


Abel, T., R. Havekes, J. M. Saletin, and M. P. Walker.2013. Sleep, plasticity, and memory from molecules to whole-brain networks. Curr. Biol. 23: R774–R788.

Allada, R., and J. M. Siegle. 2008. Unearthing the phylogenetic roots of sleep. Curr. Biol. 18: R670–R679.

Borbely, A. A. 1982. A two process model of sleep regulation. Hum. Neurobiol. 1: 195–204.

Campbell, S. S., and I. Tobler. 1984. Animal sleep: a review of sleep duration across phylogeny. Neurosci. Biobehav. Rev. 8: 269–300.

Crocker, A., and A. Sehgal. 2010. Genetic analysis of sleep. Genes Devel. 24: 1220–1235.

Datta, S. 2010. Cellular and chemical neuroscience of mammalian sleep. Sleep Med. 11: 431–440.

Kirszenblat, L. and B. van Swinderan. 2015. The yin and yang of sleep and attention. Trends Neurosci. In press (doi: 10. 1016/j.tins2015. 10.001)

Pace-Schott, E. F., and J. A. Hobson. 2002. The neurobiology of sleep: genetics, cellular physiology and subcortical networks. Nat. Rev. Neurosci. 3: 591–605.