A mixture of three drugs that block adrenergic receptors, applied directly to the exposed cortex of awake mice, widened the space between brain cells from 14.3 percent of the tissue volume to 22.6 percent. That is the range measured in mice that were asleep or under anesthesia. The figures come from a paper by Lulu Xie and colleagues, published in Science on 18 October 2013.
All the experimental data in the study came from male C57/BL6 mice supplied by Charles River, 10 to 12 weeks old, according to the supplementary methods. The corresponding author was Maiken Nedergaard, of the Center for Translational Neuromedicine at the University of Rochester Medical Center.
The space was measured with real-time iontophoretic tetramethylammonium, shortened to TMA in the paper. One microelectrode releases TMA ions into the cortex and a second electrode, about 150 micrometres away, records how much arrives. A smaller extracellular space dilutes the ions less, so more TMA is detected. Both electrodes sat 150 micrometres below the cortical surface, in head-fixed mice.
So what makes those gaps change size with the state of the brain?
The background to the question is in the paper's introduction. The brain has no conventional lymphatic system, and instead cerebrospinal fluid recirculates through it and exchanges with the interstitial fluid around cells, carrying off interstitial proteins including beta-amyloid. The authors call these pathways the glymphatic system, named for their dependence on astrocytic aquaporin-4 water channels. Why we sleep at all has been argued over for a long time, and the paper's discussion notes theories going back to the ancient Greek philosophers.
Two-photon imaging came first. Fluorescent tracer infused into the cisterna magna of sleeping mice moved into the cortex along periarterial spaces. When the same animal was woken by gentle handling of its tail, periarterial and parenchymal tracer influx fell by about 95 percent over the 30 minute imaging session.
The TMA recordings were the test of why. In sleeping mice recorded at midday, the interstitial volume fraction averaged 23.4 plus or minus 1.9 percent across six animals. In awake mice recorded between 8 and 10 p.m., it was 14.1 plus or minus 1.8 percent in four animals. In a separate paired design, the same awake mice went from 13.6 plus or minus 1.6 percent to 22.7 plus or minus 1.3 percent after ketamine and xylazine, an increase of more than 60 percent across ten animals.
The paper summarises the set as 13 to 15 percent awake against 22 to 24 percent in sleeping or anesthetized mice.
Two checks sit alongside those numbers. Tortuosity, which describes how obstructed diffusion through the space is, did not differ significantly with the state of brain activity, and all three states gave values between 1.3 and 1.8. Recordings taken 300 micrometres down did not differ significantly from those at 150 micrometres, which the authors read as a sign the cranial window had not injured the tissue.
From there the paper turns to noradrenaline. Arousal is driven by a concerted release of neuromodulators, and the authors single out signaling derived from the locus coeruleus, which they say "appears critical for driving cortical networks into the awake state of processing". They also note that in peripheral tissues such as kidney and heart, noradrenaline regulates membrane transporters and channels that control cell volume. The step they take from that is stated plainly in the text: "We hypothesized that adrenergic signaling in the awake state modifies cell volume and thus the size of the interstitial space."
The receptor block was the test of that hypothesis. To keep it confined to the cortex, the antagonists were applied directly to the exposed cortical surface rather than delivered into the cisterna magna. The supplementary methods name the mixture as prazosin, atipamezole and propranolol, all at 25 micromolar, dissolved in artificial cerebrospinal fluid.
TMA recordings under that treatment gave the headline figures, 14.3 plus or minus 5.2 percent rising to 22.6 plus or minus 1.2 percent, in four to eight mice. The value was significantly greater than in awake littermates exposed to vehicle alone, and comparable with sleeping and anesthetized animals, with P values of 0.77 and 0.95 for those two comparisons.
The electrical recording moved with it. Cortical ECoG showed an increase in slow wave power under the antagonists in seven mice, and the authors describe the inhibition as having turned the awake trace into "a more sleep-like, albeit less regular, profile".
The same three drugs were used by a different route in the imaging work. Infused into the cisterna magna at 2 millimolar, they raised CSF tracer influx in awake mice to rates the authors call more comparable with sleep or anesthesia than with the awake state. One control addressed the stress of being held at the microscope. Microdialysis of cortical interstitial fluid found that noradrenaline concentration did not increase in trained mice while restrained, and fell after ketamine and xylazine.
The clearance measurements were done separately. Radiolabelled 125I-beta-amyloid 1-40, injected into the frontal cortex, was cleared twofold faster in sleeping mice than in awake mice across 23 to 29 animals per group. An inert tracer injected at the same time, 14C-inulin, was also cleared more efficiently, by more than twofold, in sleeping and anesthetized mice. Inulin is neither transported across the blood brain barrier nor retained by the brain, so the speed-up was not specific to amyloid.
The popular line is that sleep washes the brain. What the 2013 mouse experiments actually showed is more specific: a noradrenaline signal tied to being awake also keeps the gaps between brain cells tight, and when those receptors are blocked those gaps open to a size comparable with sleep. That is a mechanism you can picture, and it is also why the Alzheimer's-toxin version of the story still outruns what was measured, which was injected protein in young male mice, and which another lab later contested.
Some of the figures do not say what they are often used to say. The 60 percent in the abstract is attached to "natural sleep or anesthesia" together, and the paired comparison that runs from 13.6 to 22.7 percent was anesthesia in the same animals rather than natural sleep. The amyloid was injected radiolabelled peptide, not plaque that had accumulated, and the animals were young adult males of a single strain. The authors' own framing stays conditional, saying the restorative function of sleep "may be a consequence of" enhanced removal of waste products.
The receptor block was a test of adrenergic receptors in the cortex, and the locus coeruleus itself was not manipulated in these experiments. The authors are also explicit that the cell-level account is unfinished, writing that "additional analysis is clearly required to determine which cell types contribute to expansion of the interstitial space volume" during sleep, anesthesia or blockade of noradrenaline receptors.
The clearance side of the 2013 paper has since been contested. In Nature Neuroscience on 13 May 2024, Miao and colleagues reported measurements of fluorescent molecules in the brains of male mice, and their abstract states that brain clearance is markedly reduced, rather than increased, during sleep and anesthesia.
What the TMA electrodes recorded in 2013 is narrower than any of that. The space between cortical cells measured about 14 percent of the tissue in awake mice, and about 22 to 24 percent in sleeping animals, in anesthetized animals and in awake animals whose cortical adrenergic receptors had been blocked. Which cell types contribute to that expansion was not determined.
Sources:
- Xie L, Kang H, Xu Q, Chen MJ, Liao Y, Thiyagarajan M, O'Donnell J, Christensen DJ, Nicholson C, Iliff JJ, Takano T, Deane R, Nedergaard M. "Sleep Drives Metabolite Clearance from the Adult Brain." Science 342(6156):373 to 377, 18 October 2013. DOI 10.1126/science.1241224. https://doi.org/10.1126/science.1241224
- Xie et al. 2013, author manuscript and figure legends, PubMed Central PMC3880190. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3880190/
- Xie et al. 2013, Supplementary Materials and Methods (animals, TMA method, drug concentrations and routes), Science 342, 373 (2013). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3880190/
- Miao A, Luo T, Hsieh B, Edge CJ, Gridley M, Wong RTC, Constandinou TG, Wisden W, Franks NP. "Brain clearance is reduced during sleep and anesthesia." Nature Neuroscience, 13 May 2024. DOI 10.1038/s41593-024-01638-y. https://www.nature.com/articles/s41593-024-01638-y
- Neuroscientifically Challenged, "Why do we sleep?" (2014). https://neuroscientificallychallenged.com/posts/why-do-we-sleep