Funding for this site is provided by readers like you.
Sleep and dreams
Our Biological Clocks

HelpLes mécanismes du sommeilLES MÉCANISMES DU SOMMEILSérotonine et sommeil
Mécanismes du cycle veille-sommeil-rêveLa théorie sérotoninergique du sommeilA neural mechanism of sleep and wakefulness

Brain Rhythms: The Oscillations That Bind

The solitary tract nucleus, in the medulla, also plays a role in the process of falling asleep, through its projections to the preoptic area of the hypothalamus. For example, stimulation of the pneumogastric nerves is known to cause subjects to fall asleep. Also, martial arts practitioners know how their alertness can be affected by blows to the neck. Likewise, Balinese massage practitioners are familiar with the relaxing effects of massaging the carotid area. The word “carotid” itself comes from a Greek word that means “causing deep sleep”.

Given the great complexity of the circuits involved in sleep, we can see that the various forms of insomnia can have many different causes: persistent stimuli that maintain wakefulness, or underfunctioning of the antiwaking system, or a phase lag in the body’s biological clock.


Once the brain’s wakefulness network has been activated, its activity is maintained by internal and external stimuli. How then does the desire to sleep come about?

First, of course, these stimuli must subside. But after that, the process of falling asleep is actually generated by the wakefulness network itself!

This antiwaking system is set in motion by one of the neurotransmitters secreted during waking periods: serotonin. This molecule is produced during these periods by the neurons of the anterior raphe nucleus, which project directly to the hypothalamus and the cortex. Some of these neurons’ axon terminals stimulate the preoptic area of the anterior hypothalamus, which in turn inhibits the entire wakefulness network, most likely by means of GABAergic neurons.


These GABAergic neurons, sometimes described as non-REM (or NREM) sleep-on neurons, are at their most active during deep (non-REM) sleep and are inactive during wakefulness and REM sleep. Electrical stimulation of these neurons quickly causes sleep, and their destruction causes insomnia. This insomnia can be interrupted, however, by the injection of muscimol (a GABA analogue) into the posterior hypothalamus, where several components of the wakefulness system converge. 

The preoptic area of the hypothalamus is an ideal location for this sleep-inducing system, because this area is a strategic crossroads that controls vital functions such as thermoregulation, hunger, and reproduction. This area can therefore monitor and analyze the body’s functional status and trigger sleep before the body becomes too fatigued, at the ideal time indicated by its biological clock.

The suprachiasmatic nucleus, which is the main component of this biological clock, is thus also involved in triggering sleep. When its neurons are damaged, the normally long periods of wakefulness shorten and become randomly distributed across the day. These neurons influence wakefulness through one of their neuropeptides: vasopressin. (Note that the effects that the vasopressin synthesized by the suprachiasmatic nucleus has on the brain are completely different from those of the vasopressin produced by the posterior pituitary gland, which acts mainly on kidney function and blood pressure.)


Back to serotonin, however. This neurotransmitter plays a specific dual role. On the one hand, it is produced in large amounts during wakefulness and contributes importantly to this state. But on the other hand, serotonin also plays a fundamental role in the process of falling from wakefulness into non-REM sleep.The explanation for this contradiction took scientists a while to find. They had even long regarded serotonin as the “sleep hormone”, because in animal experiments, destroying the neurons that synthesized it, or inhibiting its synthesis in other ways, caused periods of sleeplessness that lasted several days, but the same animals could sleep again if the immediate precursor of serotonin was then injected into the preoptic area of their anterior hypothalamus.

Since, then, scientists have obtained a far better understanding of why a lack of serotonin in the anterior hypothalamus prevents the onset of sleep. This improved understanding led to the hypothesis that the preoptic area of the anterior hypothalamus did not operate as a “sleep centre”, but rather as an area that imposed an inhibition on wakefulness. This hypothesis was subsequently confirmed electrophysiologically. It was found that the measured unit activity of the serotonergic raphe neurons is at its greatest during wakefulness, declines at the onset of non-REM sleep, and ceases during REM sleep. This gradual onset of electrical silence as someone moves from non-REM sleep into REM sleep thus indicates that the raphe neurons stop releasing serotonin into the synapses because it has done its work of inhibiting wakefulness, and its levels can therefore be allowed to decline.

Once the wakefulness network has been thus inhibited by the antiwaking system, the pacemaker cells for non-REM sleep can be expressed. At the same time, the disinhibition of the thalamic pacemaker also contributes to the onset of sleep. The rhythmic activity that then becomes established in the thalamus prevents the cortex from performing cognitive processes that require rapid communication between the thalamus and the cortex, such as that which takes place during waking and dreaming.



Sleep and DreamingBrain Rhythmic Activities
Mircea STERIADESteriade, Mircea

The rhythmic activity recorded in an electroencephalogram (EEG) trace is an overall measurement of the sum of the currents generated by the activity of the neurons in the cortex (see box below).

Each of the brain’s different states of alertness (wakefulness, non-REM sleep, and REM sleep) has its own distinct oscillation pattern, or rhythm. This rhythm is the result of interactions between the thalamus and the cortex, which in turn depend on modulations in the brainstem and the hypothalamus.

For example, the neurons of the mesopontine cholinergic nuclei of the ascending pathway, which are located in the rostral part of the pons, project their axons to the thalamus. There they make cholinergic connections not only in its sensory areas, but also in its reticular nucleus, a layer of neurons that surrounds the thalamus like a skin and exerts a general inhibiting effect on it by means of the neurotransmitter GABA. (By the way, despite its name, the thalamic reticular nucleus has nothing to do with the reticular formation.)

The cholinergic neurons from the pons sensitize the sensory thalamus but inhibit the reticular nucleus. How can this be so? The answer is that these two structures have different kinds of receptors for acetylcholine and hence respond to it differently. The sensory thalamus is sensitized by the activation of its nicotinic receptors for acetylcholine, whereas the reticular thalamus is inhibited by the activation of its muscarinic receptors for this same neurotransmitter. 

When the brain is awake, its cholinergic, histaminergic, and noradrenergic networks thus activate the thalamus in two ways: directly, by facilitating the sensory thalamus, and indirectly, by inhibiting the reticular nucleus and thus suppressing its general inhibiting effects on the thalamus.

It is important to note that acetylcholine does not excite the neurons of the sensory thalamus directly. Instead, it simply sensitizes them by depolarizing them slightly, so that instead of firing action potentials in bursts, they begin doing so at regular intervals. In this state, the thalamic neurons are more sensitive to sensory inputs. As a result, the activity of the pyramidal cortical neurons, which receive major connections from these thalamocortical neurons, is desynchronized, and the EEG trace becomes typical of the waking state: low amplitude but high frequency. Note that these pyramidal cells receive direct nicotinic cholinergic excitation directly from the basal nucleus of Meynert as well as from the thalamocortical neurons.

During the minutes when an individual is falling asleep (Stage 1 non-REM sleep), the firing frequency of the noradrenergic, cholinergic, and serotonergic neurons of the activating system in the brainstem decreases, so that the thalamus is less activated.

Concurrently, the inhibition of the neurons of the thalamic reticular nucleus is correspondingly released, so that they can resume their spontaneous oscillating activity, which makes their inhibiting effect on the thalamocortical neurons grow stronger and stronger. The rhythmic action potentials that are then generated by the GABAergic pacemaker neurons of the reticular nucleus cause a cyclical hyperpolarization of the thalamocortical neurons, thus helping to generate the rhythmic activity of the thalamus. The thalamus then becomes less and less sensitive to stimuli from the environment, which is the distinctive feature of the deepest phase of sleep.

In Stage 2 of non-REM sleep, the cortex goes into an automatic activity pattern of thalamic origin, characterized by sleep spindles on the EEG. These spindles are caused by the process just described: the rhythmic firing of the reticular neurons produces cyclical hyperpolarizations in the thalamocortical neurons, followed by bursts of action potentials. These potentials are received by the cortical cells, where they generate the sleep spindles.

The slow, high-amplitude waves produced in stages 3 and 4 of non-REM sleep result from the hyperpolarization of the pyramidal cells of the neocortex, which is triggered by local GABAergic interneurons, most likely under the influence of the preoptic neurons of the anterior hypothalamus. The thalamic neurons, whose membrane potential is then even more negative than during sleep spindles (seen mainly in Stage 2) probably also contribute to these slow cortical waves.

Lastly, during REM sleep, one cause of the desynchronized EEG traces characteristic of this stage is the influence of the cholinergic neurons on the thalamic cells, which prevents the expression of their rhythmic oscillatory activity by the same mechanisms described above with regard to wakefulness.


There are various types of neuronal connectivity that encourage rhythmic bursts of action potentials within neural networks. One of the simplest is a reciprocal connection between an excitatory neuron and an inhibitory neuron, both of which are activated by a third neuron whose own activation pattern may be steady, with no rhythmic bursts. As long as this continuous activation of the excitatory neuron by the third neuron persists, the activity of this excitatory neuron will be interrupted regularly, because this neuron activates the inhibitory neuron, which inhibits it in return. But then the temporary cessation of the excitatory neuron’s activity immediately causes the inhibitory neuron’s activity to cease. Hence the excitatory neuron is once again receptive to the continuous input from the third neuron. The excitatory neuron is therefore activated again, which quickly activates the inhibitory neuron, and the oscillatory cycle continues.

Lien : Circuits of Neurons Process Information

  Presentations | Credits | Contact | Copyleft