Circadian dysfunction is implicated in Alzheimer’s disease, but which one comes first? Two papers from scientists at Washington University School of Medicine in St. Louis lend some clues. Yo-El Ju and colleagues report in the January 29 JAMA Neurology that cognitively healthy people with brain amyloid deposition have interrupted sleep-wake cycles. Meanwhile, researchers led by Erik Musiek report in the January 30 Journal of Experimental Medicine that knocking out the circadian clock in mice dampens daily Aβ oscillation in the hippocampus and leads to higher buildup of fibrillar plaque there. Together, the two studies hint that circadian dysfunction is a driving force that contributes to Alzheimer’s disease.
- Circadian rhythms falter in cognitively normal people with plaques.
- Mice without a central circadian clock lose their normal Aβ oscillations.
- Mice without more local circadian regulation deposit more plaques.
Daily rhythms are profoundly disturbed in people with AD; indeed, severe insomnia and nighttime activity are a major cause of institutionalization (Harper et al., 2005; Bianchetti et al., 1995). These and other circadian abnormalities, including inactivity during the day, fragmented sleep, and flattened melatonin amplitude, have been extensively described in AD. Autopsies of AD patients have uncovered degeneration of the suprachiasmatic nucleus (SCN), the tiny clutch of neurons sitting above the optic chiasm in the hypothalamus that acts as the central pacemaker for the body’s circadian clock (Zhou et al., 1995). However, only one study has reported circadian disturbances in mild cognitive impairment, and no one has described whether circadian rhythms falter in people who have brain amyloid without memory problems (Naismith et al., 2013). Ju and colleagues previously reported that sleep quality takes a nosedive when cognitively healthy people have brain amyloid (Apr 2013 conference news), but are other daily cycles disturbed in preclinical AD, as well?
To find out, Ju and colleagues, including Musiek, who was first author on the JAMA Neurology paper, studied 189 cognitively normal volunteers participating in longitudinal research at the Knight Alzheimer’s Disease Research Center. All were older than 45. For this study, each volunteer underwent seven to 14 days of actigraphy, wearing a watch-like device that recorded daily rest and activity. Of the 189, 142 had had an amyloid PET scan within three years of the actigraphy and 26 were amyloid-positive.
Among all the participants, rest/activity patterns were more fragmented in older than younger people. Rather than long periods of rest separated by long periods of activity, as is normal, these blocks were broken up; for example, a night’s sleep was interrupted by several active periods, and daytime activity alternated with more naps and sedentary phases. Circadian patterns among the 26 people who were amyloid-positive, both men and women, were even more fragmented than explained by age alone. Fragmentation also correlated with the CSF p-tau/Aβ42 ratio taken as a continuous variable. CSF cutoffs yielded too few volunteers to correlate Aβ positivity with circadian rhythms.
Taken together, the data suggest that aging and amyloid pathology independently associate with fragmentation of circadian rhythms. “These circadian problems are not just something that happens late in the disease as a consequence of dementia,” Musiek told Alzforum. “They could play a role in disease pathogenesis.”
Kristine Yaffe, University of California, San Francisco, praised the results. “This is a great model where you can look at people who are amyloid-positive and see if there are changes in circadian parameters long before cognitive symptoms develop,” Yaffe said. She previously found that cognitively normal women with irregular circadian activity were at higher risk for AD, but she did not measure Aβ pathology in that cohort (Tranah et al., 2011).
The main limitation of the current study is that it is cross-sectional, Yaffe said, so it cannot distinguish whether amyloid or circadian changes come first. Musiek said that the volunteers are being followed longitudinally, so researchers will be able to track the relationship over time. Yaffe added that scientists have not explored circadian rhythms in dementia much, though they have studied sleep parameters.
In parallel, Musiek studied mice to get a better handle on cause and effect. As reported in the JEM paper, first author Geraldine Kress and colleagues manipulated the circadian clock in APPPS1 mice, which express mutated human forms of APP and PSEN1. In mice, as in humans, a central clock driven by a circadian pacemaker in the SCN controls daily cycles, including those of body temperature and activity. Mammals also have local clocks that oscillate on the cellular level to control tissue-specific changes in gene expression and metabolism. These gene-expression cycles are coordinated by the central clock, which makes sure all the local clocks are synchronized.
In this study, the researchers wondered if the level of soluble Aβ in the brain’s interstitial fluid, which is known to fluctuate with a circadian rhythm, was controlled by the SCN master regulator or by local brain clocks. They deleted one of several core clock genes, called Bmal1, in three specific patterns: globally throughout the entire brain; throughout the brain bar the SCN; or only within the hippocampus. The last two Bmal1 deletions did not alter the central circadian pacemaker because core body temperature, one of its known outcome markers, rose and fell in its normal rhythm. In general, mice run hotter at night, when they are active, and cooler in the daytime when they sleep, and this pattern held. In contrast, mice with brain-wide Bmal1 knockout lost their normal circadian temperature rhythm, suggesting the central circadian pacemaker had gone out of whack.
For each deletion, when the mice were two months old and before plaques started to deposit, the researchers measured Aβ oscillations in their hippocampi using microdialysis.
In the global Bmal1 knockouts, the normal amplitude of soluble Aβ oscillation in the hippocampus shrunk by half. More of the peptide stuck around when mice were asleep and would normally clear it, while less new Aβ formed when the mice were awake and would normally be producing it. There was no net change in the average Aβ levels in the brain, however. The mechanism behind this dampening is not clear, and could be due to loss of sleep or body temperature rhythms, or blunted neuronal activity, Musiek said. In contrast, when Bmal1 was knocked out everywhere but the SCN, the amplitude of soluble hippocampal Aβ fluctuations remained normal, implying that this rhythm is regulated by the central circadian clock in the SCN, not locally.
Fibrillar Aβ told a different story. When APP/PS1mice expressed Bmal1 only in the SCN, they built up more fibrillar Aβ plaques in the hippocampus by four months of age than did APP/PS1 controls. ApoE expression in the cortex rose during the same time, suggesting local Bmal1, or local circadian rhythms, usually tone down this lipoprotein. Since mouse ApoE promotes plaque formation, this finding may explain how local loss of Bmal1 affects fibrillogenesis, though this idea will require direct testing, said Musiek.
The results suggest that the central circadian pacemaker controls daily oscillations of soluble Aβ in the hippocampus, while local circadian rhythms might control plaque load nearby. If the same holds true in people, it could imply that treating circadian dysfunction at its onset could help prevent AD, but this has yet to be tested, Musiek said.
“This is a very exciting and novel study,” said Marilyn Duncan, University of Kentucky, Lexington. She took it as strong evidence that central, not local, brain clocks regulate the daily Aβ oscillation in the hippocampus. Future research should address how the SCN clock does this, she said.—Gwyneth Dickey Zakaib
Research Models Citations
- Harper DG, Volicer L, Stopa EG, McKee AC, Nitta M, Satlin A. Disturbance of endogenous circadian rhythm in aging and Alzheimer disease. Am J Geriatr Psychiatry. 2005 May;13(5):359-68. PubMed.
- Bianchetti A, Scuratti A, Zanetti O, Binetti G, Frisoni GB, Magni E, Trabucchi M. Predictors of mortality and institutionalization in Alzheimer disease patients 1 year after discharge from an Alzheimer dementia unit. Dementia. 1995 Mar-Apr;6(2):108-12. PubMed.
- Zhou JN, Hofman MA, Swaab DF. VIP neurons in the human SCN in relation to sex, age, and Alzheimer's disease. Neurobiol Aging. 1995 Jul-Aug;16(4):571-6. PubMed.
- Naismith SL, Hickie IB, Terpening Z, Rajaratnam SW, Hodges JR, Bolitho S, Rogers NL, Lewis SJ. Circadian Misalignment and Sleep Disruption in Mild Cognitive Impairment. J Alzheimers Dis. 2013 Oct 7; PubMed.
- Tranah GJ, Blackwell T, Stone KL, Ancoli-Israel S, Paudel ML, Ensrud KE, Cauley JA, Redline S, Hillier TA, Cummings SR, Yaffe K, . Circadian activity rhythms and risk of incident dementia and mild cognitive impairment in older women. Ann Neurol. 2011 Nov;70(5):722-32. PubMed.
- Musiek ES. Circadian clock disruption in neurodegenerative diseases: cause and effect?. Front Pharmacol. 2015;6:29. Epub 2015 Feb 27 PubMed.
- Musiek ES, Xiong DD, Holtzman DM. Sleep, circadian rhythms, and the pathogenesis of Alzheimer disease. Exp Mol Med. 2015 Mar 13;47:e148. PubMed.
- Musiek ES, Bhimasani M, Zangrilli MA, Morris JC, Holtzman DM, Ju YS. Circadian Rest-Activity Pattern Changes in Aging and Preclinical Alzheimer Disease. JAMA Neurol. 2018 May 1;75(5):582-590. PubMed.
- Kress GJ, Liao F, Dimitry J, Cedeno MR, FitzGerald GA, Holtzman DM, Musiek ES. Regulation of amyloid-β dynamics and pathology by the circadian clock. J Exp Med. 2018 Apr 2;215(4):1059-1068. Epub 2018 Jan 30 PubMed.