31 May 2024

In recent years, several studies have suggested that the brain rinses itself of harmful waste, such as Aβ, during sleep. In the May 13 Nature Neuroscience, researchers led by Nicholas Franks and William Wisden at Imperial College London cast doubt on those findings. They reported that instead of speeding clearance, sleep slowed the removal of a tracer dye from the brains of wild-type mice. They also found that neither sleeping, waking, nor anesthesia had any effect on how quickly dyes diffused throughout the brain. “Our results challenge the idea that the core function of sleep is to clear toxins from the brain,” the authors wrote.

In mice, injected dyes diffused through the brain at the same rate during sleep and wakefulness.
Clearance of solutes from the parenchyma slowed during sleep.
The findings belie the view that the brain clears toxins during sleep.

Alan Verkman at the University of California, San Francisco, was impressed by the data. “I find the results convincing and important. They strongly refute the ‘glymphatic’ mechanism, which I and many other scientists do not believe,” he wrote to Alzforum. Others had concerns about methodology. “We found the results intriguing, but we believe the study has technical issues and potential for data misinterpretation,” wrote Benjamin Plog and Jonathan Kipnis, Washington University, St. Louis (comment below). A news piece in Science also noted that the methodology had drawn criticism.

Faster Cleanup While Awake? In the frontal cortices of mice anesthetized with ketamine-xylazine (top) or sleeping (middle), an injected dye clears more slowly than in awake mice (bottom). Quantification at right: anesthesia curve in red, sleep in blue, awake in black. [Courtesy of Miao et al., Nature Communications.]

The glymphatic hypothesis, first proposed by Maiken Nedergaard at the University of Rochester Medical School in New York 12 years ago, posits that astrocytes help drive the flow of cerebrospinal fluid through the brain parenchyma, washing away waste (Aug 2012 news). Subsequently, Nedergaard and others reported that this mechanism works best during sleep (Oct 2013 news; May 2014 conference news; Nov 2019 news). Not everyone was convinced, however, with some studies questioning the hypothesis in favor of clearance by diffusion (Asgari et al., 2016; Smith et al., 2017; Holter et al., 2017).

To investigate further, joint first authors Andawei Miao and Tianyuan Luo measured the diffusion rate of molecules in wild-type mouse brain. They injected fluorescently labeled 4 kilodalton dextran into caudate putamen and allowed the dye to diffuse throughout the brain until it reached a steady state. Then they bleached the dye in a small region of the frontal cortex, and measured how quickly new dye moved into that region. This is an established way to determine diffusion rate in tissues (Thiagarajah et al., 2006). The dye dispersed at about 32 square meters per second, in good agreement with a previous report in rodent brain (Syková and Nicholson, 2008). Importantly, this rate did not change whether the mice were awake, sleeping, or anesthetized.

The authors next homed in on brain clearance, i.e., whether solutes are being removed from the brain as they circulate. For this experiment, they used the 0.5 kD dye AF488, which is small enough to move freely through parenchyma. Two to three hours after injection, the dye concentration in frontal cortices of awake mice was less than a third of what would be expected based on the diffusion rate, indicating that much of it had been cleared from the brain. In sleeping or anesthetized mice, on the other hand, about half the dye remained, implying that clearance ebbs during sleep and anesthesia. The authors confirmed the findings by sacrificing mice after injection and directly measuring the amount of dye in brain sections at set time points.

Why have so many previous studies reported higher clearance during sleep? The authors noted that most studies injected tracers into the CSF and measured how far into the cortex they penetrated, using that as a proxy for how fast solutes were moving out of the brain. However, because dye concentration in a brain region results from a balance between its influx and efflux rates, deeper penetration during sleep could be due to slower clearance, rather than faster movement, the authors suggested. “Under these circumstances, entry, exit, and redistribution of the marker are all occurring simultaneously, greatly confounding any quantification of clearance,” they wrote.

Jurgen Claassen, Radboud University Medical Center, Nijmegen, The Netherland, was not surprised by the findings. Ten years ago, he reported that as little as one night of sleep deprivation slowed clearance of Aβ from people’s brains (Ooms et al., 2014). But subsequent work from his lab seemed to disprove that idea. “These conflicting studies in mice and people show that we still do not know enough to say whether sleep has an important role to play in clearance,” he wrote to Alzforum (comment below).

Kaoru Yamada and Takeshi Iwatsubo, University of Tokyo, think clearance from the brain is more complicated than can be teased out with tracers. “Unlike the small dyes used in this study, many proteins present in the brain parenchyma are cleared not only by glymphatic clearance but also by a combination of mechanisms, including proteolytic degradation, phagocytosis by glial cells, and blood-brain barrier transport. Therefore, whether sleep and anesthesia have similar effects on other CNS proteins may warrant careful consideration,” they wrote to Alzforum. At the time of writing, Nedergaard had not responded to a request for comment.

If solute clearance does not speed up during sleep, then what explains the observed association between poor sleep and increased brain Aβ? One previous study suggested this was due to higher production of the peptide during a person’s wakeful hours, rather than to clearance during sleep (Jan 2018 news).

Claassen still thinks clearance plays a role. “Recent studies from different groups all clearly show that CSF flows dynamically, driven by hemodynamic and respiratory pressure oscillations, during wakefulness. I would still say it is likely that this CSF flow supports clearance beyond passive diffusion,” he wrote.—Madolyn Bowman Rogers and Tom Fagan

Comments

- Benjamin Plog
  Washington University School of Medicine in St. Louis
- Jonathan Kipnis
  Washington University in St. Louis, School of Medicine
- Posted: 31 May 2024
We read this article with great interest. We found the results intriguing but believe the study has technical issues and potential for data misinterpretation. The authors suggest that brain solute clearance is independent of consciousness state and that clearance is faster during wakefulness. However, their experimental approach raises concerns. For instance, the infusion volumes used in Figure 1 to generate the model, 10 μl injected into striatum, are much larger than typical, likely distorting normal brain tissue architecture and influencing tracer movement.

The authors' interpretation of photometric clearance data may be flawed, as it does not account for differences in tracer influx rates. For their histological assessment, the authors do not provide data for t=0 to ensure that all groups start at the same baseline fluorescence.

Finally, the exclusion of data that did not fit theoretical models may introduce bias. The sentence in the data exclusion paragraph states: “For the diffusion coefficient measurements, bleaching recordings that could not be fitted by the custom curve-fitting algorithm were excluded. For the photometry recordings, poor fits to the theoretical curves were excluded ...” All told, the authors are using an interesting model, which we believe needs to be further calibrated.

View all comments by Benjamin Plog View all comments by Jonathan Kipnis

- Jurgen Claassen
  Radboud University Nijmegen Medical Center
- Posted: 31 May 2024
Andawei Miao and colleagues, in an elegant mouse model, scientifically challenge the concept of “enhanced glymphatic brain clearance during deep sleep,” which was born a decade ago, also in a mouse model.

Over this decade, the original mouse model inspired human studies, and one from my own group in 2014 (Ooms et al.) was the first to translate the animal findings to humans (Ooms et al., 2014). It was interesting to see that results of these human studies seemed to fit in with the mouse models. Most of these “clearance in sleep” studies focused on clearance of Aβ, making sleep a new hot topic in Alzheimer’s research. Could we prevent Alzheimer’s just by sleeping better?

The question on the table now is if this study by Miao set us back and disproves the notion that sleep enhances clearance, or if their findings are based on error. Scientific debate will undoubtedly settle the latter.

However, I am not surprised by the findings, because my own group has published papers that appear to disprove our original 2014 findings. We explored evidence of increased amyloid accumulation in people with long-term (20 years) partial sleep deprivation, in what you could call a “natural experiment.” Harbor pilots in The Netherlands have a job that causes highly irregular sleep. In contrast with our own hypothesis, we found no evidence for increased brain Aβ (Thomas et al., 2020). These conflicting studies in mice and people show that we still do not know enough to say whether sleep has an important role to play in clearance.

Meanwhile, recent studies from different groups all clearly show that CSF flows dynamically, driven by hemodynamic and respiratory pressure oscillations, during wakefulness. I would still say it is likely that this CSF flow supports clearance beyond passive diffusion.

References:

Ooms S, Overeem S, Besse K, Rikkert MO, Verbeek M, Claassen JA. Effect of 1 night of total sleep deprivation on cerebrospinal fluid β-amyloid 42 in healthy middle-aged men: a randomized clinical trial. JAMA Neurol. 2014 Aug;71(8):971-7. PubMed.

Thomas J, Ooms SJ, Mentink LJ, Booij J, Olde Rikkert MG, Overeem S, Kessels RP, Claassen JA. Effects of long-term sleep disruption on cognitive function and brain amyloid-β burden: a case-control study. Alzheimers Res Ther. 2020 Aug 26;12(1):101. PubMed.

View all comments by Jurgen Claassen

- Kaoru Yamada
  The University of Tokyo
- Takeshi Iwatsubo
  University of Tokyo
- Posted: 03 Jun 2024
In this paper, Miao et al. present data that contradicts the findings of Xie et al., Science 2013, by showing that clearance is inhibited during sleep or under anesthesia. As the authors discuss, sleep and anesthesia may have different effects on the penetration of the substrate from CSF into the brain, and its clearance from the brain. Another factor that could affect the results is the difference in the substrates themselves. Unlike the small dyes used in this study, many proteins present in the brain parenchyma are cleared not only by glymphatic clearance but also by a combination of mechanisms, including proteolytic degradation, phagocytosis by glial cells, and blood-brain barrier transport. Therefore, whether sleep and anesthesia have similar effects on other CNS proteins may warrant careful consideration.

View all comments by Kaoru Yamada View all comments by Takeshi Iwatsubo

- Steven T. Proulx
  University of Bern
- Adrian Madarasz
  University of Bern
- Posted: 05 Jun 2024
We found the results in Miao et al. very intriguing. The authors have utilized an elegant approach to explore how fluorescent solutes are transported between different brain regions, by infusing the solutes into the striatum and detecting them using implanted sensors in the cortex. Using a photobleaching approach, the study has produced data that has demonstrated that the movement of a 4kDa fluorescent solute within the brain parenchyma occurs predominantly by diffusion. This conclusion is consistent with previous studies (Hladky and Barrand, 2014; Smith et al., 2017).

The authors have also shown with their approach that reduced levels of a low-molecular-weight fluorescent dye are apparent in the cortex under awake conditions, when compared to different anesthesia regimens and during sleep. These reduced levels are attributed to an increased clearance out of the brain (through undefined pathways) while the mice are awake, which would be the opposite conclusion reached by the Nedergaard group in 2013 (Xie et al., 2013).

Overall, we believe the authors’ approach is technically sound, with a built-in recovery time incorporated after the implantation of the parenchymal injection cannula and the use of more conservative values for injection rates than utilized in previous studies (Aspelund et al., 2015; Plá et al., 2022).

The results are consistent with a previous report from our group. In Ma et al., 2019, we utilized infusions of tracers into the CSF of the lateral ventricle and evaluated the distribution of the tracers between anesthetized and awake conditions. When the mice were awake, there was significantly faster clearance of tracers from the CSF to the CNS-draining lymphatic vessels and, ultimately, to the systemic circulation. This resulted in much less tracer apparent in the paravascular subarachnoid space on the cortical surface of mice in the awake group. As the surface of the cortex is where microscopic measurements of fluorescence intensity of tracers were made in Xie et al., 2013, and alternative destinations (i.e., to the lymphatic system) for CSF tracer flow were not considered, we suggested that the measurements of CSF influx in this influential paper may not have been correctly interpreted.

Taken together, one can envision a model in which an increased turnover of CSF during waking conditions would enhance the diffusive clearance of low-molecular-weight solutes from the interstitial fluid of the brain. This would be consistent with a historical concept of CSF acting as a "sink" for brain clearance that was originally suggested by the eminent British physiologist Hugh Davson in the 1960s (Bradbury and Davson, 1964; Oldendorf and Davson, 1967). We eagerly await new data, including measurements of CSF production under different vigilance conditions and further advancements to methods to evaluate protein clearance from the brain parenchyma, which will allow rigorous testing and refinement of the various models of brain clearance.

References:

Aspelund A, Antila S, Proulx ST, Karlsen TV, Karaman S, Detmar M, Wiig H, Alitalo K. A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J Exp Med. 2015 Jun 29;212(7):991-9. Epub 2015 Jun 15 PubMed.

BRADBURY MW, DAVSON H. THE TRANSPORT OF UREA, CREATININE AND CERTAIN MONOSACCHARIDES BETWEEN BLOOD AND FLUID PERFUSING THE CEREBRAL VENTRICULAR SYSTEM OF RABBITS. J Physiol. 1964 Jan;170(1):195-211. PubMed.

Hladky SB, Barrand MA. Mechanisms of fluid movement into, through and out of the brain: evaluation of the evidence. Fluids Barriers CNS. 2014;11(1):26. Epub 2014 Dec 2 PubMed.

Ma Q, Ries M, Decker Y, Müller A, Riner C, Bücker A, Fassbender K, Detmar M, Proulx ST. Rapid lymphatic efflux limits cerebrospinal fluid flow to the brain. Acta Neuropathol. 2019 Jan;137(1):151-165. Epub 2018 Oct 10 PubMed.

Oldendorf WH, Davson H. Brain extracellular space and the sink action of cerebrospinal fluid. Measurement of rabbit brain extracellular space using sucrose labeled with carbon 14. Arch Neurol. 1967 Aug;17(2):196-205. PubMed.

Plá V, Bork P, Harnpramukkul A, Olveda G, Ladrón-de-Guevara A, Giannetto MJ, Hussain R, Wang W, Kelley DH, Hablitz LM, Nedergaard M. A real-time in vivo clearance assay for quantification of glymphatic efflux. Cell Rep. 2022 Sep 13;40(11):111320. PubMed.

Smith AJ, Yao X, Dix JA, Jin BJ, Verkman AS. Test of the 'glymphatic' hypothesis demonstrates diffusive and aquaporin-4-independent solute transport in rodent brain parenchyma. Elife. 2017 Aug 21;6 PubMed.

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. 2013 Oct 18;342(6156):373-7. PubMed.

View all comments by Steven T. Proulx View all comments by Adrian Madarasz