As we have discussed in previous lectures, the brain is the organ in the body that is most affected and at the same time responsible for sleep and changes during sleep. It is therefore fundamental to understand what goes on inside the brain while we sleep. In particular, we would like to know which neuronal networks and chemical systems are activated during the different stages of sleep and what neuronal commands are responsible for sleep-wake transitions and sleep stage changes during the sleep period.

This is a very active area of sleep research that has its roots in the paradigm-shifting work of Bremer and Magoun and Moruzzi discussed previously. By making lesions in specific areas of the cat brainstem they managed to convincingly show that sleep is not a passive "shut-down" mechanism but that there are neuronal mechanisms that actively induce sleep.

Bremer (1937)

cerveau isolé : transection through the brainstem between the superior and inferior colliculi. The cat survived for a few days ans showed permanent synchronized EEG and pupillary constriction (comatose state).

encéphale isolé : transection at the caudal end of the medula, just above the spinal cord. The cat showed normal sleep and waking cycles. The EEG showed periodic episodes of synchronized and desynchronized activity. It was able to track objects with the eyes.

Moruzzi and Magoun (1959)

"Midpontine cats". Transection just caudal to the cerveau isole preparation. The cat was awake 70-90 % of the time (as opposed to 30-40 % in normal cats). Cats show pupillary dilation, can track visual stimuli and show accomodation. When cortical synchrony is observed, visual stimuli have no effect.

Another reason to study the chemistry of sleep is that we may gain insights into the evolution of sleep. Evolution is natural selection acting over mutations. Enourmous changes in behavior or sleep patterns are ultimately dependent on spatial and temporal changes in the concentrations of different molecules, in particular of specific neurotransmitter systems as described in more detail below. Studying the evolution of sleep-relevant molecules can give important insights into how sleep evolved.

A general problem in neuroanatomy is that we are still talking at the level of hypothalamus, amygdala, hippocampus, visual cortex. This is equivalent to describing american people, russians and french. Saying that neurons in the hippocampus are active during sleep is similar to saying that american people like horror movies or fries. It is true for some but not necessarily true for everyone. There are approximately 10¹¹ neurons in the human brain (actually quite more than the population in the world). It is most certain that each neuron is not unique as each person is but still, talking at the level of nuclei is an enormous generalization. Much more work is still needed to understand which specific molecules, being released from which specific neurons and reaching which specific receptors in certain neurons are relevant for the changes during sleep. There are lots of fascinating questions to be addressed.

One of the oldest theories about the induction of sleep states that a fatigue or toxin substance gets accumulated and induces the sleep state. Initially, it was thought that this substance should reside in the blood. However, the fact that Siamese twins with a common circulatory system sleep independently argues against a common sleep-inducing factor in the blood (Alekseyeva, 1958).

But maybe this putative sleep factor is not in the blood but in the brain. In order to test this idea, Legrende and Pieron (1913 ) kept dogs awake for several days. Then, they extracted cerebrospinal fluid from these animals and were able to induce sleep by injecting the fluid into the ventricular system of non-sleep deprived dogs.

It should be mentioned, however, that from subjective reports, sleepiness and tiredness have a circadian rhythmicity which cannot be explained by the simple accumulation of a sleep-inducing factor (Kleitman).

 Melatonin has gained considerable attention recently as non-hazardous sleep-inducing pills. It is also widely used to combat the effects of jet-lag. It is a natural hormone produced by the pineal gland (top of the midbrain, between the superior colliculi). It only affects the latency to sleep and not the sleep structure. It seems to have a powerful hypnotic effect on birds but, interestingly, it seems to induce wakefulness when applied to rats during the daytime.

Several other substances have been suggested to have hypnogenic properties:

No one has found neurons that "lack energy and need to sleep".

No one has found neurons that run-out of neurotransmitters during the awake state and need to sleep to replenish them.

There is no universal decline in firing rate during sleep. Neurons in some areas decrease their activity during sleep while neurons in other areas actually increase their firing. This is valid both for NREM as well as REM sleep. Furthermore, this has been observed both electrophysiologically and by functional imaging studies of the human brain (see below).

It's not the absence of sensory stimulation that causes sleep. The body sensory stimulation can be severed and the animal still shows wake-sleep cycles.

As discussed previously, making lesions in the brainstem reticular formation in cats produced slow-wave activity in the EEG and inactivity as in the coma state in humans. Some of the usual concerns about lesion experiments include the fact that the experiment may be unwillingly and without knowing damaging fibers of passage. That if a lesion in area X produces a given deficit or state, it is possible that there are actually axons going from area Y to area Z that go through area X and are being concomitantly damaged and responsible for the deficits. On the other hand, if there is no deficit, it could be argued that area X is indeed important but that recovery and plasticity are helping the brain cope with the damage.

An elegant and landmark experiment was carried out by Moruzzi and Magoun (1949) to further establish the importance of activity in the brainstem in regulating sleep/wake cycles. They showed that electrical stimulation of the brainstem reticular formation yields a state of arousal in cats.

Furthermore, single unit recording studies show that there is a correlation between the firing rate of neurons in the brainstem reticular formation and arousal. These neurons project to non-specific thalamic neurons that show similar properties and direct their output all over the cortex.

The hypothalamus is an interesting structure. It is involved in regulating heart rate, blood pressure, thirst, hunger, sex, wakefulness and sleep!

The hypothalamus is important in regulating the circadian rhythm. Damage to the hypothalamus in rats disrupts otherwise organized daily rhythms. A particular area of the hypothalamus, called the suprachiasmatic nucleus (because it is above the optic chiasm which is where the information from the two eyes cross) seems to be particularly important for the circadian regulation. This nucleus receives direct input from the retina, thus providing a mechanism to entrain the circadian rhythm dependent on the light in the environment.

If an area of the hypothalamus, called the basal forebrain is damaged, animals become hypervigilant. Furthermore, electrical stimulation of this region can induce slow wave sleep.

Dopamine is a neurotransmitter released by the substantia nigra. Lesions in this region cause a comatose state in cats but not in rats (the reason for this difference is unclear). Patients with Parkinson's disease show damage in the substantia nigra neurons. They also display signs of progressive immobility. Antidopaminergic drugs (used to combat schizophrenia) also induce hypoactivity and lack of facial expressivity. The dopaminergic system seems to be more involved in the initiation of movement than on conciousness or the lack thereof. There are also dopamine releasing neurons in the posterior hypothalamus. Lesions in the substantia nigra produces unresponsiveness and immobility.

Several lines of evidence suggest an important role for noradrenalline and monoamines in general in being responsible for the arousal state.

Noradrenaline (NA) is released (among others) by the locus coeruleus in the pons. Amphetamine facilitates the release of dopamine and NA and retards re-uptake. The effects of amphetamine dissapear after lesions in the pontine-midbrain reticular formation. Cocaine inhibits reuptake of catecholamines. There is also prolonged and intense arousal by MAO inhibitors (mono-amine oxidase inhibitors). Conversely, arousal is decreased by inhibition of catecholamine synthesis.

The electrodes in the stimulation experiment of Moruzzi and Magoun were very close to this noradrenergic system.

Reserpine, which produces a state of inactivityy and tranquilization acts by depletion of monoamines. The effect is reversed by administration of L-dopa.

Clinical studies have also provided evidence for the involvement of dopamine and norepinephrine neurons in conditions of akinesia and coma.

Raphe nuclei (Greek seam, suture)

Disconnecting it from the rest of the cerebrum causes almost constant arousal. Lesions cause 3-4 day insomnia (in cats); NREM sleep slowly recovers (why this is so or what exactly happens during this recovery is unclear). Anesthetic injections have anti-sleep effects. Electrical stimulation (within the caudal solitary tract) induces sleep.

Functional imaging provides a fascinating non-invasive tool to look at the changes in blood levels in the human brain while subjects are sleeping. For an introductory article on the physical and biological basis of functional imaging, see Raichle (1994). For an overview of studies of functional imaging during sleep, see Hobson (1998)

There seems to be an overall decline in oxygen consumption during NREM sleep.

A fascinating question that people have studied for decades is whether the brain activation during REM is similar to that during the wake state. Using EEG measurements it is not easy to distinguish the awake state and REM sleep. The EOG (electro-oculogram) is thus required. Differences can also be found at the single-neuron or pharmacological level but many researchers had suggested that REM is the same as the waking state except for closed input-output. The imaging studies provide the first clear evidence to the contrary.

Limbic and paralimbic regions of the brain are suggested to be involved in processing emotions as evidenced by neurophysiological, lesion and molecular studies in animals and humans. These areas are more activated during REM than during the awake state. This fits with our subjective recall of most dreams as being rich in emotional content. In particular, the amygdala is strongly activated during REM sleep.

There seems to be a de-activation of pre-frontal cortex during REM sleep. It was suggested that dream-amnesia could be due to prefrontal deactivation.

Visual areas are activated during REM sleep. This broadly correlates with the visuospatial vividness of dreams.

It was suggested that the fictive movement depends on the activation of the basal ganglia or cerebellar nuclei.

Scientists and philosophers have always discussed about the problem of consciousness. Recently, the neuroscience community has become particularly interested in this fascinating problem. Specific hypothesis and experiments have been suggested in order to address rigorously what the neuronal correlates of visual consciousness are. Crick and Koch have claimed that there must be a specific group of neurons that explicitly represent the contents of our moment-to-moment aware perception.

In this regard, studies of sleep and dreams in particular may provide particularly relevant data. From a subjective point of view, it seems evident that we are perfectly conscious during our dreams. In particular, since most dreams have strong visual content (in non-congenitally blind people) we are visually aware during our dreams. We could therefore include one more constraint for the neurons that correlate with our visual perception. The neurons that represent our visual awareness should be activated when we see something, when we close our eyes and imagine it or when we are asleep and dream about the same thing.

Crick and Koch have argued based on neurological and electrophysiological evidence that the activity in area V1 does not correlate with our visual awareness. It is therefore interesting to note that V1 does not seem to be activated during dreams. It is also quite fascinating to note that higher visual areas, particularly the temporal lobe is activated during sleep.

See also the Special Note on Sleep and Consciousness

As fMRI gets better and better in spatial and temporal resolution it will be more and more fascinating to look more carefully at what happens in different brain areas during sleep.

Sleep stages and brain activity (requires simultaneously recording EEG and fMRI; work in progress).

Correlation between dreams and brain activity (fMRI and wake-up subjects; complications with this experiment).

Molecular biology of sleep. Gene knock-outs and sleep patterns.