Talk:Neurobiology of sleep and wakefulness

Reviewer B

The review concerns Sleep and Wakefulness, yet begins directly with EEG, which is only one measure of sleep and wakefulness. It would be preferable to begin with a behavioral definition of sleep and then proceed to describe EEG and other polygraphic correlates.

EEG Frequency bands should include Sigma 11 or 12 - 14 Hz, corresponding importantly to the frequency of sleep spindles.

“In the human, EEG data are combined with those from concurrent…” in the way it is written suggests that this is unique to humans and not other mammals, which of course is not the case, so that ‘in the human’ should be removed. Or “In the animal, EEG and EMG signals are sufficient…” should be changed to ‘In animals, EEG and EMG can be sufficient…’ In this paragraph, spindles should be described with Stage 2.

“Non-rapid eye movement (NREM) sleep is…” would more appropriately be called SWS in humans and then introduced as also called NREM sleep in animals (by some not all investigators). NREM sleep for rodents is particularly inadequate since as stated above eye movements are not recorded and in fact do not really define the state, which accompanied more by rapid whisker movements or rapid muzzle movements in this group of animals. For this reason, paradoxical sleep (or active) sleep is a more appropriate term than REM sleep. Perhaps this concept and term could be more fully introduced on the sections on REM sleep. In this paragraph, more emphasis could be given to muscle atonia, not simply reduction in muscle tone.

“As sleep begins, a stepwise descent from wakefulness to Stage 1 through to Stage 4 sleep occurs, followed by an abrupt ascent back towards Stage 1 (Figure 3)” is inaccurate. Would be better phrased, ‘as sleep progresses..’ and then, ‘following a period of Stage 3-4, sleep spindles reappear with Stage 2-like EEG preceding a return to Stage 1 like EEG with irregular theta activity”.

“Later studies determined that these cells were cholinergic and localized them to the laterodorsal tegmentum/pedunculopontine (LDT/PPT) region.” Stimulation of the reticular formation in the region of the LDT/PPT in fact likely activates more fibers of passage than cholinergic cells in this region. Nonetheless, it is not contested that the cholinergic LDT/PPT neurons play an important role in cortical activation particularly through their thalamic projections, however they represent a small proportion of the neurons projecting to the thalamus and comprising the ascending activating system. Within the pontomesencephalic tegmentum, including the reticular formation proper, there are likely 10 times the number of cholinergic cells which are noncholinergic/nonGABAergic and likely glutamatergic (with evidence now of high concentrations of glutamate, content of phosphate activated glutaminase, the synthetic enzyme for glutamate and one of the vesicular transporter proteins for glutamate). Particularly since the authors go on in the next paragraph to mention glutamate, they should rephrase the above sentence to read, ‘Later studies established that many of these cells or an important contingent of these (thalamic projecting) cells…’

“The importance of orexin to the control of sleep and wakefulness was discovered because its pathology is the cause of narcolepsy.” This sentence could be better developed to give the Reader a better idea of what ‘pathology’ represents, ie loss of orexin, its neurons or receptors.

“Evidence points to adenosine, a neurotransmitter with extracellular levels coupled to cellular metabolism, as a mediator of the sleep homeostat (McCarley, 2007). As it accumulates during wakefulness, adenosine inhibits components of the AAS, in particular the BF cholinergic neurons, and after very prolonged wakefulness, the cortex itself.” Given recent negative evidence for any key role of adenosine and its action via BF cholinergic neurons {Blanco-Centurion, 2006 #2706}, this statement should be softened if not deleted from this section. Whereas evidence is very strong for a role of adenosine in blocking excitation under conditions of hypoxia or hypoglycemia in both periphery and brain, it is weak for such a role during the normal or natural sleep-wake cycle.

“The importance of this region was later confirmed by stimulation, electrophysiological and histological studies, which identified the critical area as being within the hypothalamic preoptic area of the BF. This area, localized to the ventrolateral preoptic area (VLPO) (Figure 5) contains a group of cells that are specifically active during sleep and contain the inhibitory neurotransmitters, γ-amino butyric acid (GABA) and galanin.” That neurons in the POA play a role in sleep is generally accepted, that those neurons are “localized to the” VLPO is an oversimplification, which has unfortunately become very popular due to its simplicity. Perhaps the statement here could be rephrased to state, ‘… critical area within the hypothalamic preoptic area of the BF. Cells concentrated within the ventrolateral preoptic area (VLPO) have been shown to discharge during sleep and to contain …’

“The various components of REM sleep are each under the control of different groups of effector neurons, including:

• Eye movements, controlled by a group of pontine RF neurons;
• Muscle atonia by a group of neurons in the dorsolateral pontine RF. These send inhibitory projections to the motor neurons in the spinal cord (SC);
• EEG desynchronization by neurons in the midbrain RF, in addition to the LDT/PPT and BF cholinergic neurons.”

It is difficult to identify separate regions for control of the various components of REM sleep, particularly since they are all more or less centered in the pontine or ponto-mesencephalic tegmentum. Perhaps medial pontine RF for eye movements, lateral (which should include the ventrolateral as well as the dorsolateral according to many lines of evidence) pontine tegmentum and the more extensive pontomesencephalic tegmentum including the LDT/PPT cholinergic neurons.

“A subset of the brainstem LDT/PPT cholinergic neurons are not active during wakefulness but discharge preferentially just before and during REM sleep (Steriade et al., 1990).” Since noone has recorded in anaesthetized animals from identified cholinergic neurons in the LDT/PPT, where they are outnumbered by noncholinergic neurons by >4x, this statement should be softened and also recognize Sakai with Jouvet {El Mansari, 1989 #700}, who along with Koyama in the rat, identify two ‘possibly’ cholinergic cell types, one that is W/PS active and one that is as suggested here, PS active. In contrast, Steriade et al actually found all presumed cholinergic cells to discharge in both W and PS.

For example, brainstem GABAergic neurons were recently found to play an important role in the generation of this state, with mutually inhibitory REM-on and REM-off localized populations of these cells acting in concert to gate the appearance of REM sleep (Lu et al., 2006). It is inappropriate to cite this recent article for a role of GABAergic pontine neurons in REM sleep and ignore the preceding work by the groups of Chase {Xi, 1999 #1697}, Jones {Maloney, 1999 #1652; Maloney, 2000 #1743}and Luppi {Boissard, 2003 #2204}.