Thinking about sleep: Maiken Nedergaard and the function of deep sleep

All animals sleep, even flies (research with the fruit fly Drosophila revealed the presence of clock genes).  All mammals have REM (rapid eye movement) and non-REM sleep. Human infants sleep 16 hours/day (if their parents are lucky); adults sleep 8 hours/day; elderly adults 5.5 hours/day. (S Lockley and R Foster: Sleep: A Very Short Introduction, Oxford University Press, 2012, pages 48-49).

Continuous sleep deprivation kills rodents and flies within days to weeks.

But why do we sleep? It would seem to be an evolutionary disadvantage. We are more vulnerable to predators when asleep. We could be working instead of sleeping. Korean secondary-school students attend hagwons (private cram academies) for five hours after school, then go home at 10 PM to study until past midnight (The Economist, September 19-25, 2015). If they didn’t have to sleep, they could study all night.

So sleep must perform some very important functions. But what?

Maiken Nedergaard, a neuroscientist at the University of Rochester (more specifically, an astroglial biologist), appears to have answered the non-REM portion of this question, in her paper “Sleep Drives Metabolic Clearance from the Adult Brain” (Science Vol 342, 10/18/2013, p. 373-377)

Other bodily organs rid themselves of waste protein products by using bulk flow of fluid between cells to wash them into the blood or lymphatic system, which carry them to the liver, where they are metabolized.

The brain uses 20% of the body’s energy supply. Yet it has no lymphatic system. How does it get rid of its waste products, such as beta-amyloid (Alzheimer’s), alpha-synuclein (Parkinson’s disease, Lewy body dementia), and tau (Alzheimer’s), to name a few, all of which are present in the interstitial fluid surrounding brain cells.

This question becomes more interesting when one considers that “essentially all neurodegenerative diseases are associated with misaccumulation of cellular waste products. Of these, misfolded or hyperphosphorylated proteins are among the most difficult for the brain to dispose of. For example tau and beta-amyloid can accumulate as stable aggregates that are neurotoxic in conditions such as Alzheimer’s disease.”

Nedergaard describes the glymphatic system, or, less politely, the “Garbage Truck of the Brain”.

Astrocytes express a water channel, the aquaporin 4 water channel (AQP4). Penetrating arteries which end in the brain are covered by astrocytic endfeet which express AQP4. This perivascular space around the arteries is a “highway for fast influx”, which can be observed with radiolabeled tracer.

In a three-step process, first cerebrospinal fluid (CSF) passes from the para-arterial space, through the aquaporin 4 (AQP4) water channels, into the interstitial space, where “vectorial convective fluxes drive waste products away from the arteries and toward the veins”, and CSF exchanges with interstitial fluid (ISF). Then the ISF and its waste products enter the paravenous space, eventually reaching lymphatic vessels in the neck, and later the systemic circulation, where the proteins travel to the liver, where they are metabolized. The brain has better things to do than chopping up the garbage.

The AQP4 water channels in the astroglial endfeet are crucial to this process (AQP4- knockout mice have a 65% reduction in beta amyloid clearance). In traumatic brain injury and stroke, AQP4 gets “mislocated to the cell body of astrocytes or to astrocytic processes that do not abut the vasculature”, and protein clearance “declines substantially”.

The interstitial concentration of beta amyloid is higher in awake rodents and humans than it is in sleeping ones. One possibility is that “wakefulness is associated with increased beta amyloid production”.

Nedergaard tested “the alternative hypothesis that beta amyloid clearance is increased during sleep and that the sleep-wake cycle regulates glymphatic clearance”.

Her group found that CSF influx into the brain was decreased by 95% in awake mice, compared to  sleeping mice or mice anaesthetized with ketamine/xylazine. (Ketamine has recently been discovered to be a very rapidly acting treatment for bipolar depression, but its antidepressant effect lasts only 1-2 weeks. One wonders if its short duration of action suggests that it clears the brain of garbage, which soon returns, leading to relapse.)

CSF influx into the brain  is “in part driven by arterial pulse waves that propel the movement of CSF inward along periarterial spaces”.

“Convective glymphatic fluxes of cerebrospinal fluid (CSF) and interstitial fluid propel the waste products of neuronal metabolism (proteins, peptides, lactate, ammonia, amyloid) into the paravenous space from which they are directed into lymphatic vessels and ultimately returned to the general circulation for clearance by the kidney and the liver.” This is analogous to garbage removal by “street sweeping” with liquid.

To manipulate the glymphatic system (the garbage highway from the glial endfeet lining the arteries, through the extracellular space, to the paravenous spaces, and the lymph channels), 4 methods were employed, all of which decreased efflux:

1.        AQP40 knockout mice: decreased fluid influx

2.        Cisternotomy: this opening eliminated the low-pressure system

3.        Acetazolamide: blocked CSF fluid production

4.        Sleep deprivation

After traumatic brain injury, increased protein markers are noted in plasma; but all 4 of the above interventions blocked the increase.

(Footnote: This glymphatic process calls to mind both the first sentence of James Joyce’s novel of sleep, Finnegans Wake:  “riverrun, past Eve and Adam’s, from swerve of shore to bend of bay, brings us by a commodius vicus of recirculation back to Howth Castle and Environs.”, and Steven Dedalus’s comment in Ulysses: “All Ireland is washed by the gulfstream.”)

Nedergaard points out that this process was first discovered by Patricia Grady in the early 1980’s, when she observed the movement of horseradish peroxidase into the brain. Attempts by others to replicate her work failed because the replicators used a “cranial window” cut into the skull, which destroyed the low pressure which drives the system. She left science and is now director of nursing at the NIH.

If pulse were the driving force behind the diurnal variation, one would expect to see greater influx during the day, when arterial blood pressure is higher than when asleep. Instead, one sees the opposite.

This system, Nedegaard reasoned, is like plumbing: all that matters is pressure and resistance. Pulse pressure is bigger when one is awake and alert. Yet influx is less. So therefore there must be less resistance while asleep.

“An alternative possibility is that the awake brain state is linked to a reduction in the volume of the interstitial space because a constricted interstitial space would increase resistance to convective fluid movement and suppress CSF influx”.

Nedegaard used techniques developed by Charles Nicholson to assess the volume and tortuosity of the interstitial space in awake, sleeping, and anaesthetized  mice, and found that the interstitial space volume fraction averaged 23.4% in sleeping mice, and 14.1% in awake mice. Both sleeping and anaesthetized mice had  higher levels of slow (delta) wave sleep. “Thus, the cortical interstitial volume fraction is 13  to 15% in the awake state as compared to 22 to 24% in sleeping or anaesthetized mice.” There was no change in tortuosity.

Interestingly, other studies have shown that the interstitial volume declines by 1/3 in aged mice compared to young mice. “The smaller space during wakefulness increases tissue resistance to interstitial fluid flux and inward movement of CSF.” The smaller space in aged animals (and humans) would make it harder to clear out the garbage/amyloid; neurodegenerative diseases are more common in the elderly.

Beta amyloid was cleared twice as fast in sleeping as in awake mice. Before the streets of the brain are swept, they are widened, not by removal of parked cars, as in Santa Monica, but by shrinkage of brain cells into their quiet, resting state.

Nedergaard next asked “what drives the brain state-dependent changes of the interstitial space volume?” Her observation that anesthesia increases glymphatic influx and efflux led her to hypothesize that it is not circadian rhythm but the sleep wake state itself.

Since noradrenergic neurons in the locus coeruleus drive cortical networks into the awake state, Nedergaard administered a cocktail of adrenergic antagonists (prazosin (α1 adrenergic receptor antagonist, which improves sleep in patients with PTSD), atipamezole (α2 adrenergic receptor antagonist), and propranolol (non-selective beta adrenergic receptor blocker) and found that they induced an increase in CSF tracer influx comparable to the sleep state, and increased the interstitial volume fraction from 14.3 to 22.6%.  The antagonists also increased the prevalence of slow, delta waves. “Norepinephrine is the master regulator of ISS volume”, she notes.

Nedergaard can be seen talking about “The Nightlife of the Brain” at the National Institute of Health (2/11/2015) at the NIH website (534 views ( or, more efficiently, on YouTube (the NIH website frequently freezes; 821 views)

She also spoke at Cold Spring Harbor Laboratory on the Glymphatic System on 12/12/2014. . (1043 views).

So, clearance of brain garbage occurs primarily during slow wave, or deep (N3, formerly called stage 3-4) sleep.

Now, sleep problems occur very early in the course of Alzheimer’s disease (AD), even during mild cognitive impairment, often an Alzheimer’s precursor, with less slow wave sleep (SWS). (Ju, Lucey, Holtzman: Sleep and Alzheimer disease pathology-a bidirectional relationship. Nature Reviews Neurology 10: February 2014, 115-119; ). Alzheimer’s pathology begins 10-15 years before cognitive symptoms appear, when soluble beta amyloid becomes insoluble and aggregates into amyloid plaques.  Amyloid accumulation disrupts sleep; disrupted sleep “increases the risk of beta amyloid accumulation in mice, as well as dementia due to Alzheimer’s disease in humans”.

While chronic sleep deprivation “accelerates beta amyloid deposition into insoluble amyloid plaques”, improving sleep “through treatment with an orexin inhibitor antagonist decreased beta amyloid plaque deposition” in mice. “One obvious approach is to investigate whether improving the quality of sleep in humans can either reduce the risk of AD or delay the progression of preclinical to symptomatic AD.”

In a more recent paper, Holtzman suggests that “considering the profound protective effect of almorexant on beta amyloid plaque burden in mice, the orexin system is a high priority target. The recent approval of suvorexant, the first Food and Drug Administration approved orexin receptor antagonist, provides an excellent opportunity to evaluate orexin-targeted therapeutics on Aβ dynamics and cognitive endpoints in early-stage or presymptomatic AD.”

So, in summary:

  1. The brain rids itself of garbage during deep sleep (n3, non-REM, slow wave), by expanding interstitial fluid volume (due to neuronal cell shrinkage?), and washing away detritus in a slow moving stream of extracellular fluid. Failure to remove amyloid leads to amyloid plaque deposition and cognitive impairment.
  2. Treatments that can enhance stage N3 sleep (deep, non-REM, slow wave), whether medication or behavioral ( ), might help clear garbage such as amyloid and prevent or delay the onset of neurodegenerative disease. Medication possibilities that increase slow wave sleep include gabapentin, trazodone, prazosin, and suvorexant. Interestingly, very low carbohydrate diets increased the percentage of deep, slow wave, stage 4 (n3) sleep, in young males.
  3. Things that decrease slow wave sleep (alcohol, medications, poor sleep habits, sleep apnea, benzodiazepines, opiates) would impair garbage clearance, and increase the risk of neurodegenerative disease.
This entry was posted in First Comments. Bookmark the permalink.

Comments are closed.