The wrinkling of time: Aging, inflammation, oxidative stress, and the circadian clock in neurodegeneration

Abstract

A substantial body of research now implicates the circadian clock in the regulation of an array of diverse biological processes including glial function, metabolism, peripheral immune responses, and redox homeostasis. Sleep abnormalities and other forms of circadian disruption are common symptoms of aging and neurodegeneration. Circadian clock disruption may also influence the aging processes and the pathogenesis of neurodegenerative diseases. The specific mechanisms governing the interaction between circadian systems, aging, and the immune system are still being uncovered. Here, we will review the evidence supporting a bidirectional relationship between aging and the circadian system. Further, we explore the hypothesis that age-related circadian deterioration may exacerbate multiple pathogenic processes, priming the brain for neurodegeneration.

Introduction

The myriad correlations between aging, aging-related disease, and circadian rhythms^1–3 provide ample justification for investigation into potential causative relationships between these phenomena. The progressively increasing prevalence of circadian dysfunction with increasing age suggests that aging drives circadian dysfunction. However, disruption of the circadian clock – either behaviorally or through genetic manipulation - can also drive aging-like phenotypes, suggesting that the relationship between aging and circadian rhythm dysfunction is bidirectional¹. More recently, evidence has accumulated documenting changes in circadian systems preceding or being predictive of the development of neurodegenerativo diseases, suggesting that circadian dysfunction could increase dementia risk³. However, this possibility as well as the implication that aging and circadian dysfunction could represent concomitant, positively reinforcing cycles of deterioration remain active areas of investigation. Additionally, while the mechanisms by which these cycles may lead to increased risk for dementia remain unknown, immune dysregulation and oxidative stress have been identified as prime candidates³. Initial studies suggest the circadian clock as a potentially viable therapeutic target for the treatment of both neurodegeneration and other age-related diseases. In connecting these concepts, it is helpful to contextualize newer investigations exploring circadian clock regulation of the immune system by considering what is known linking circadian dysfunction with aging (Also see Hood et al., 2017¹).

Overview of the mammalian circadian system

Circadian rhythms are a fundamental part of biology, as most organisms have a circadian clock that allows behavioral and physiological adaptation to the 24-hour light-dark cycle of earth. In mammals, the “master clock” of the body resides in the suprachiasmatic nucleus (SCN) of the hypothalamus. The SCN receives synaptic input from the retina and the cellular clocks within neurons of the SCN are thus entrained to the external light-dark cycle. These cellular clocks then keep 24-hour time and the SCN has specific neural circuitry to ensure timekeeping that is both robust and flexible⁴. The SCN provides synchronizing cues through regulation of endocrine and autonomic nervous system function to cellular clocks throughout the body, including in neurons and glia in the brain^5,6. The core molecular clock found in each cell is comprised of a positive transcriptional limb and negative feedback limb. The positive limb is composed of the bHLH-PAS transcription factor BMAL1 (aka Arntl), which forms hetereodimers with CLOCK or NPAS to drive circadian transcription via binding to Ebox motifs. The negative limb consists of the PERIOD and CRYPTOMCHROME families of proteins, which are direct transcriptional targets of BMAL1 and which in turn inhibit BMAL1 function⁵. The ROR and REV-ERB proteins, positive and negative regulators of Bmal1 transcription, respectively, are also transcriptional targets of BMAL1 and further modulate clock timing. This core clock is tuned to a 24-hour period through the concerted actions of numerous post-translational mechanisms carried out by a network of secondary clock proteins⁷. These core clock genes are expressed in nearly every cell in the body and can generate circadian rhythms in transcription and cellular function in the absence of any external cues. The circadian clock regulates between 10–50% of all transcripts in a cell, depending on tissue type, and influences critical processes such as cell cycle, redox homeostasis, inflammation, and metabolism⁸. This breadth of clock-controlled genes may partially explain the wide ranging consequences of circadian clock disruption for aging as well as in the pathogenesis of many chronic diseases⁹.

Behavioral circadian disruptions in aging

On a behavioral level, circadian disruption is a widely-studied characteristic of both aging^10–13 and neurodegeneration^14–19. Specifically, age-associated sleep changes, including sleep fragmentation, represent perhaps the most consistent and clear evidence linking behavioral circadian disruption to aging¹. Sleep disturbances such as difficulties with falling and staying asleep¹², increased sleep to wake transitions (sleep fragmentation)^10,17, and increased daytime drowsiness and napping^10,12 are all characteristic of elderly populations. Sleep structure is also altered²⁰ with a particularly prominent age-associated decrease in slow wave sleep^11,21–24, which is deemed important for protein clearance^25–28, maintaining metabolic health²⁹, and potentially in memory consolidation³⁰. Interestingly, a recent report details dampening of rhythms in cortical excitability with age, which correlates with sleep changes and potentially contributes to age-related cognitive decline³¹. Older populations tend to display earlier chronotypes^20,32–34 while, at least in men, an individual’s chronotype shifts earlier as age increases³⁵. Somewhat paradoxically, in a Dutch population aged 18–65, a later sleep onset was correlated with shorter telomere length³⁶, a feature associated with cellular aging and senescence³⁷. While the robustness of an individual’s sleep rhythm declines with age, their ability to adapt to an environmentally imposed phase shift, as with jet-lag, also declines with age in humans^38,39 and in mice^40,41. Increasing fragmentation of circadian activity rhythms is also specifically noted in aging men and is independent of preclinical Alzheimer Disease pathology¹⁷. However, further research is required to disentangle whether these changes reflect alterations to the circadian system itself, independent from aberrant regulation of sleep homeostasis. The incorporation of other circadian readouts in addition to sleep may help facilitate this endeavor.

Other systemic circadian changes with aging

Outside of sleep, alterations in several other systemic circadian processes have been shown with age. For instance, body temperature normally exhibits peak temperature in the evening while the trough occurs in the early morning before waking⁴². In aged humans there is a phase advance in body temperature rhythm such that the nadir occurs earlier. The relationship between sleep and body temperature rhythms may also be altered, with age being associated with a later body temperature nadir relative to time of awakening^32,43. At least in men^44,45, there may also be a reduction in amplitude⁴³ and variability⁴⁶ of the temperature rhythm in aged adults (60s or older).

Melatonin, a hormone regulated by the SCN and secreted by the pineal gland, normally induces sleep, possibly by acting on BMAL1⁴⁷ and regulates body temperature⁴⁸. A potential decrease in melatonin secretion with age^53–55 has been inconsistently documented^56,57 and may be specific to women⁵⁴. Additionally, it is possible that a decrease in melatonin secretion could be indicative of pathological instead of healthy aging^57,58. In the SCN, the expression of melatonin receptor declines with age, which may contribute to the dispersion of behavioral rhythms^59,60. This decrease may be at least partially responsible for the loss of both sleep and body temperature rhythm robustness in advanced age.

Glucocorticoids, steroid hormones of which cortisol is the primary form in humans, have a complex relationship to stress, the immune system, and the regulation of plasma glucose homeostasis⁶¹. Glucocorticoids are regulated by the SCN, follow a circadian pattern of secretion, and are potent synchronizers of a number of peripheral molecular clocks^61–63. Van Cauter et al. as well as several others^64–66 found that the rhythm in circulating cortisol is dampened in aging, enabled by a progressive rising of the nadir and accompanied by an overall increase in levels⁶⁷. However, others found a phase shift in elderly subjects, but no change in amplitude⁶⁸, while still others found no major changes in cortisol with age⁶⁹. In mice, an age-related decline in glucocorticoid signaling in the hippocampus, potentially due to a decrease in glucocorticoid receptor expression, may play a role in the depletion of the neural stem cell pool⁷⁰. Despite disagreement on specific cortisol rhythm abnormalities (possibly due to high variability or vastly different sample sizes), a remaining diurnal rhythm in circulating cortisol in aged humans is consistent between studies^64–69. This persistence of a cortisol rhythm in aging, which in contrast to the case of melatonin is retained during pathological aging^15,58, adds another level of complexity to the series of interconnected feedback loops that experience age-associated alterations or losses in robustness.

Age-related changes in the SCN

In addition to the impairments and alterations observed in SCN-regulated rhythms, age-associated changes in the SCN itself have been documented. An impairment in the rhythm of neuronal firing in mice^71,72 and flies⁷³, as well as decreases or altered rhythms⁷⁴ in the expression of neuropeptides arginine vasopressin (AVP) and vasoactive intestinal peptide (VIP) in the SCN have been observed in humans, especially in men⁷⁵, and in rodents^76,77 with advanced age. A loss in GABAergic synapses in the SCN has also been reported in aged mice⁷⁸. GABA-mediated neuronal activity, as well as the expression of VIP are critically important for the cohesiveness of SCN neuronal firing rhythms^79,80 and the maintenance of behavioral rhythms depends on the coordination of SCN neuron firing^81,82. In a small sample of elderly people and Alzheimer patients, fragmented sleep-wake rhythms during life were associated with loss of VIP-ergic neurons in the SCN on post-mortem examination⁸³. Thus, current data support the hypothesis that changes in SCN neurons could contribute to age-associated behavioral rhythm desynchrony. However, this possibility requires more thorough evaluation to show that these changes in neuronal populations directly influence organismal rhythms.

Aging and circadian clock gene expression

Reports documenting age-induced alterations in the molecular clock have been more controversial. Some have shown dampening or dispersion in the SCN expression rhythms of Bmal1, Clock^84,85, and Per2^86,87 while others report normal Per1 and Per2 rhythms^88,89 with advanced age. Altered molecular rhythms, including an impaired ability to phase reset, have also been observed in the mouse liver⁹⁰ (although to a lesser extent in some reports^91,92), heart⁹³, kidney, lung⁸⁸, thymus³⁸, and pancreas⁹² among others. However, intact molecular rhythms have been observed in muscle and epidermal stem cells of aged mice⁹⁴. Interestingly, the induction of replicative cellular senescence has been found to impair entrainment of the molecular clock⁹⁵, suggesting that perhaps the accumulation of senescent cells in a given tissue with age could play a role in the dispersion of circadian phases between cells. Outside of genes directly involved in the core molecular clock, a large number (more than 1000) of clock-controlled genes display altered rhythmicity, some even gaining rhythms with age in the human prefrontal cortex⁹⁶. The liver⁹¹ as well as muscle and epidermal stem cells⁹⁴ also undergo substantial circadian reprogramming in aged mice. However, more data is needed to solidify the physiological relevance of altered molecular rhythms, especially in the aged brain.

The circadian system, healthspan, and lifespan

Changes in the circadian system can be predictive of, while inducing circadian disruption can reduce, healthspan and lifespan. For instance, the degree of deviation from a 24-hour circadian period was found to be negatively correlated with lifespan in both rodent and primate species^97,98. Conversely, implantation of young SCN tissue improved the molecular⁹⁹ and behavioral rhythms¹⁰⁰ of rats and the longevity of aged hamsters after surgery compared with cortex- and mock-implanted controls¹⁰¹. Additionally, inducing weekly phase shifts, especially phase advances, can shorten the lifespan of aged, but not young, mice⁴⁰ while phase shifting can also increase the vulnerability of mice to an lipopolysaccharide (LPS) challenge^102,103. Genetically, ablating the clock via a global knockout of Bmal1 shortens lifespan and induces a number of other “aging-like” pathologies, such as cataracts and sarcopenia in mice¹⁰⁴. Moreover, deficiencies in either Per2¹⁰⁵ or Clock/Bmal1¹⁰⁶ mediated transcription has been shown to exacerbate cancer or drive age-dependent insulin dysfunction and diabetes, respectively. The pathologies in the Bmal1 KO model have since been partially attributed to loss of Bmal1 during development/early life¹⁰⁷ and exhibit tissue specificity¹⁰⁸. However, these and further studies utilizing macrophage/monocyte^{102,109–112}, muscle¹¹³, liver¹¹⁴, brain-specific (sparing the SCN)¹¹⁵, and other tissue-specific circadian mutants have recapitulated components of aging-like phenotypes and vulnerabilities, including insulin resistance (recently reviewed¹¹⁶). Interestingly, a contingent of these metabolic abnormalities, including lipid accumulation and glucose intolerance, can be mitigated by time restricted feeding, highlighting the importance of the clock in maintaining metabolic homeostasis^117–119. These studies also suggest time restricted feeding as a potentially viable behavioral intervention for age-related metabolic dysregulation.

Maintaining the integrity of circadian rhythms is crucial for optimizing a large host of physiological outputs including, but not limited to long-term potentiation^120,121 and associated cognition^122,123, metabolic health¹²⁴, reaction time^125,126, and muscle performance^{113,127–129}, age-associated deteriorations of which have been extensively documented. Accordingly, the perturbation of rhythms, for instance with nighttime light exposure¹³⁰, circadian misalignment (shift work)¹³¹, or jet-lag¹³² can impair these functions. Circadian disruption also negatively impacts insulin sensitivity as well as increases risk factors^133–136 and worsens outcomes¹³⁷ for acute neurological and cardiovascular events, which already display daily rhythms in occurrence^138–140. These data suggest that at the very least, disruption of the circadian clock is detrimental in the context of aging. The intriguing possibility that such disruptions could be driving the aging process itself, negatively impacting healthspan and lifespan should be of particular interest to future studies. Additionally, recent studies suggest that the efficacy and toxicity of drug therapies for age-related diseases such as cancer can be dramatically affected by the circadian phase in which they are administered^141–143. Circadian regulation of treatment efficacy may become even more complicated and warrants further investigation in the context of aging, given the age-related alterations in phase and dispersion of rhythms discussed here. The intimate interaction between the immune system and the circadian clock, discussed in the next section, adds yet another layer of complexity to be considered, and perhaps leveraged, in the development of therapeutics to treat age-related disease.

Mechanistically, the circadian clock is linked to the mammalian target of rapamycin (mTOR) and Sirtuin 1 (SIRT1)^87,144–147. These factors are closely tied to the regulation of aging with mTOR negatively and SIRT1 positively impacting healthspan and lifespan^148–152. SIRT1 interacts with the BMAL1/CLOCK complex and may impact circadian transcription directly by deacetylating BMAL1¹⁵³, PER2¹⁵⁴, and histone H3, acting counter to the histone acetyltransferase functions of the CLOCK protein itself¹⁵³. Additionally, levels of NAD+, an essential metabolite and necessary substrate for SIRT1 deacetylase activity^151,155, as well as the expression of NAMPT, the rate-limiting enzyme in the NAD+ salvage pathway¹⁵⁶, have been shown to oscillate in the mouse liver^157,158 and human red blood cells (NADH)¹⁵⁹. This circadian clock regulation of NAD+ through NAMPT is important for maintaining homeostatic levels of mitochondrial oxidative phosphorylation¹⁶⁰ and for feeding back into SIRTT^153,157,158 (as well as mitochondrial SIRT3¹⁶⁰) activity. Although through a modestly different mechanism, modulation of the circadian clock by SIRT1 is also present in the SON⁸⁷. In concordance with decreased expression of Sirt1 in aged animals, this control of the clock by SIRT1 wanes with age⁸⁷. An age-associated systemic decline in NAD+ (recently reviewed^161–163), possibly due to decreased levels of clock-regulated NAMPT in several tissues^164–166 has been thoroughly documented. Taken together, these data suggest that a deficit in the interaction between SIRT1 and circadian signaling could bear some responsibility for the connection between circadian dysfunction and the aging process^1,87. In support of this idea, aged mice experience a substantial dampening of the protein acetylation rhythms under dual regulation by NAD+/SIRT1 and the circadian clock in the liver⁹¹. These rhythms are restored by caloric restriction⁹¹, currently the most robust lifespan extension intervention¹⁶⁷. Caloric restriction can also induce circadian reprogramming in both young¹⁶⁸ and old animals as well as greatly enhance NAD+ levels and SIRT1 activity⁹¹.

Circadian physiology is also inextricably linked with a number of other metabolic pathways¹⁶⁹, including the insulin signaling¹¹⁶ and mTOR pathways^{144–146,170}, the suppression of which have both been shown to extend lifespan and healthspan^{148,149,152,171–173}. Insulin induces phosphorylation of BMAL1 via AKT, thereby inhibiting BMAL1 transcriptional activity¹⁷⁴. On the other hand, downstream insulin signaling target mTOR can also induce BMAL1 phosphorylation via S6K1, a modification that enables BMAL1 to play a critical role in mTOR-regulated translation¹⁷⁵. Additionally, activation or inhibition of mTOR results in acceleration or dampening of the circadian clock, respectively^144,145. Calorie restriction, which extends lifespan, impairs insulin signaling, and inhibits mTOR, also increases Bmal1 expression and BMAL1 mediated transcription¹⁷⁶. Finally, loss of Bmal1 has been found to increase mTOR activity (although not in all reports⁴⁷), while inhibition of mTOR extends the lifespan of Bmal1 KO mice by 50%¹⁴⁶. Taken together, these data suggest a bidirectional relationship whereby maintaining a metabolic equilibrium that favors longevity also promotes robustness of the circadian clock, while maintaining the integrity of the clock may promote longevity by sustaining metabolic homeostasis.

Glial clocks and aging

In addition to in the SCN and throughout the body¹⁷⁷, oscillating molecular clocks have been documented in a variety of extra-SCN brain regions¹⁷⁸ as well as in astrocytes⁶ and microglia^179,180. In the SCN, astrocytic clocks are synchronized by VIP¹⁸¹ and can be altered by immune factors such as TNFα¹⁸². Astrocytic extracellular ATP release¹⁸³, which has potential implications for allodynia¹⁸⁴, gliotransmission¹⁸⁵, and glutamate uptake¹⁸⁶ are regulated by the clock. Additionally, astrocytes play a substantial role in maintaining behavioral circadian rhythms. For instance, under certain conditions, glial clock dysfunction can cause behavioral arrhythmicity in flies¹⁸⁷. Several recent studies have independently documented an even more impressive role for the clock in SCN astrocytes in determining the phase and period of mouse circadian rhythms^188–190. Surprisingly, it was also shown that SCN astrocytes are capable of generating population-wide circadian clock oscillations and mouse activity rhythms in the absence of intact neuronal clocks¹⁹¹. Despite the prominence of the astrocyte clock in the SCN, relatively little is known about its function elsewhere in the brain and outside of behavioral rhythm maintenance. However, recent evidence suggests that glial clocks may play a substantial role in regulating the neuroimmune system – discussed in more detail in the next section - with potential implications for neurodegeneration¹⁹². Notably, glia regulate blood-brain barrier permeability, which has been shown to exhibit circadian oscillation in flies¹⁹³. Additionally, multiple groups have reported marked aging-induced changes to the astrocytic^194,195 and microglial¹⁹⁶ transcriptomes that may substantially overshadow those in neuronal populations¹⁹⁷. Together, these data suggest that glial clocks may represent a fresh perspective from which to consider the ballooning interest in the role of both astrocytes and microglia in the pathogenesis of neurodegenerative diseases.

The clock and the immune system

Recent studies have convincingly demonstrated circadian regulation of the immune system in the periphery¹⁹⁸, while emerging evidence links the clock to regulation of the immune response in the CNS^3,192. Indeed, the circadian clock regulates inflammatory and oxidative stress responses. For example, both lesions of the SCN¹⁹⁹ and light induced rhythm disruption²⁰⁰ can exacerbate release of cytokines TNFα¹⁹⁹ and IL-6^199,200 in response to LPS, while LPS can differentially activate SCN neurons based on time of day¹⁹⁹. Chronic circadian phase shifts (chronic jet-lag)²⁰¹ or merely varying the time of day^102,103 can heighten both inflammation and LPS-induced endotoxemic death in mice. In addition to the aging-related pathologies previously discussed, global and brain-specific Bmal1 KO as well as global Clock/Npas2 double KO mice have age-dependent increases in ROS damage, chronic inflammation^104,115 including increased Tnfa, microglia and astrocyte activation, and synapse degeneration¹¹⁵. In monkeys, Bmal1 KO can also induce immune system activation and depression-like symptoms²⁰².

Importantly, clock genes including Clock, Per2²⁰³, Bmal1, and the BMAL1 target Nr1d1 oscillate in peripheral macrophages^112,203 and lymphocytes²⁰⁴. In humans, the LPS-induced blood levels of cytokines Interferon-γ (IFN-γ), Interleukin-10 (IL-10)²⁰⁵, and TNFα vary consistently based on time of day while IL-6 levels vary inconsistently^206,207. In mice, lymphocyte trafficking²⁰⁴, LPS-induced monocyte recruitment, cytokine levels including TNFα, IL-6²⁰³, IL-12¹¹¹, inducible nitric oxide synthase (iNOS - reactive NO-producing enzyme)¹¹², chemokines including CCL5¹¹¹, and mortality²⁰⁸ exhibit time of day dependence with a reduction during late wake/early rest periods^112,203. This reduction can be abolished upon monocyte Bmal1¹¹² or Nr1d1¹¹¹ KO indicating an immune-suppressive role for these proteins. Accordingly, Bmal1 KO in monocytes reduces survival in response to infection and exacerbates chronic inflammation and glucose intolerance in a mouse model of diet-induced obesity¹¹². Deletion of Bmal1 also induces an Nrf2-dependent increase in ROS and IL-1β in macrophages¹¹⁰. Disruption of BMAL1-regulated neutrophil aging can impair immune defense and vascular protection in mice¹⁰⁹.

In vitro, Bmal1 KO can cause increased neuronal degeneration, death, and susceptibility to oxidative damage¹¹⁵. Additionally, it was found that the BMAL1/CLOCK complex binds chemokine Ccl2 and Ccl8 promoters¹¹² while BMAL1 binds the promoters of genes protective against oxidative stress, which are also downregulated in global Bmal1 KO mice¹¹⁵. Macrophages from global Nr1d1 KO increase IL-6 secretion while REV-ERBα (Nr1d1) agonist GSK4112^111,209 and Nr1d1 overexpression in culture²⁰⁹ suppresses IL-6 release. In further support of BMAL1-mediated immune suppression, global KO of two repressors of BMAL1 activity, Per2²¹⁰ and microRNA miR-155, can reduce TNFα¹⁰², IL-1β, and IFN-γ¹¹⁰ secretion upon LPS treatment. Taken together, these and similar studies make a strong case for the circadian clock as an important immune regulator, providing a limiting check on immune over activation in the periphery.

The neuroimmune system, primarily under the purview of glia, may also be subject to regulation by the molecular clock. In addition to the astrocytic clock discussed previously, a few recent reports have documented oscillating clock gene expression including Bmal1, Per2, and Nr1d1 in microglia^179,180,211. Cytokine levels including Il-6, Tnfa, and the critical inflammasome component Nlrp3 (only measured after LPS), among others, show circadian variation in unstimulated and LPS-stimulated whole hippocampus and microglia¹⁷⁹. Aging abolishes these differences, clamping the microglial inflammatory response to LPS at its highest level in younger mice²¹². Little is known about astrocyte clock function in the immune system. However, we have shown that astrocyte clock dysfunction induces astrogliosis and can impair neuronal survival²¹³. Additionally, BDNF and Nrf2-dependent oxidative stress protection provided by astrocytes to neurons²¹⁴ and NF-κB-mediated inflammation may both be regulated by the astrocytic clock²¹⁵. Nr1d1 KO induces microgliosis and astrogliosis in vivo and exacerbates the neuroinflammatory response to LPS treatment, including NF-κB signaling, in vivo as well as in cultured microglia²¹⁶. However, one study demonstrated a surprising depression of Il-6 expression in microglia and a mitigation of stroke damage in vivo after deleting microglial Bmal1²¹⁷. The varied results from glial clock manipulations suggest a more nuanced clock regulation of the glial immune response and underscore the need for further investigation. Such efforts may be especially relevant in the context of neurodegeneration where glial cells play an increasingly appreciated and crucial role in disease progression.

The general finding of a more active immune system at the rest to wake transition^111,112,179 is likely preemptive, preparing the body for increased possibilities of infection exposure during “morning” foraging and conservationist, minimizing both energy expenditure at unneeded times and collateral damage induced by a constitutively active immune system²¹⁸. These studies in the peripheral and central immune systems as well as the pro-inflammatory, pro-ROS phenotypes of Bmal1 mutants support the possibility that the circadian clock could regulate the CNS immune response in microglia and astrocytes. In total, these data suggest that alterations to circadian systems documented both in the SCN and in tissue-specific clocks could play a substantial role in immune hyperactivation with aging. Thus, these alterations may generate tissue environments susceptible to the overproduction of oxidative stress, and prime the body for the development of neurodegenerative disease.

Circadian clocks and oxidative stress

Considerable evidence supports a bidirectional relationship between the circadian clock and oxidative stress, as changes in redox status can influence core clock function, while clock proteins themselves regulate redox homeostasis of cells²¹⁹. Binding of BMAL1/CLOCK to DNA is dependent on the NAD(H)/NADP(H) ratio, a barometer of cellular redox status, with increased binding occurring under reducing conditions²²⁰. Circadian rhythms in hydrogen peroxide levels are observed in cultured cells²²¹ and mouse liver, and can directly regulate rhythms in CLOCK function via cysteine oxidation²²². Deletion of the redox-responsive protein p66^shc, which is itself rhythmic, disrupts these oxidation rhythms in CLOCK and alters mouse behavioral and transcriptional rhythms²²². Inhibition of pentose phosphate pathway (PPP), which is critical for generation of NADPH to fuel ROS generation by the NADPH oxidase enzymes, as well as for glutathione production, can also alter clock function²²³. PPP inhibition with resultant loss of NADPH leads to oxidative stress, activation of the NRF2 redox response pathway, increased BMAL1/CLOCK DNA binding, altered clock gene expression, and lengthened circadian period²²³. NRF2 itself appears to regulate clock function, as Nrf2^−/− cells have diminished clock gene expression and blunted circadian rhythms in Per2 expression²²⁴. In the SCN, circadian rhythms in redox tone regulate rhythmic neuronal activity via regulation of potassium currents²²⁵. Thus, oxidative stress can regulate clock function by multiple pathways.

Conversely, the core clock controls expression of redox response genes and dictates cellular responses to oxidative stress. Drosophila exhibit circadian rhythms in ROS sensitivity which are lost in arrhythmic Per⁰¹ mutant flies. Per⁰¹ flies have shortened lifespans, increased oxidative damage, and age-related neuronal degeneration^226,227. Glutathione, a critical small molecule antioxidant present in all cells, is regulated by the clock at several levels. Glutathione levels, glutathione-producing enzymes, and glutathione transferase enzymes all show circadian oscillation in drosophila and mice^228,229. NRF2, which strongly regulates glutathione synthesis, is a transcriptional target of BMAL1¹¹⁰. BMAL1 controls Nrf2 in pancreatic beta cells²³⁰, macrophages¹¹⁰, and lung cells²³¹, with loss of BMAL1 causing blunted NRF2-mediated antioxidant responses and enhancing ROS levels. Accordingly, deletion of BMAL1 leads to increased oxidative damage in multiple organs, including the brain^104,115. The clock protein REV-ERBα can be induced by oxidative stress and can in turn regulate expression of the antioxidant transcription factor F0X01 as well as stimulate autophagy and mitochondrial biogenesis^232–234. Overexpression of REV-ERBα provides protection against oxidative stressors and improves mitochondrial function^232–234. Oxidative stress in flies can even reprogram genome-wide circadian transcription toward a redox stress response²³⁵. Taken together, these data suggest that the circadian clock responds to changes in cellular redox tone and regulates expression of redox response pathways. As oxidative stress is strongly implicated in many aspects of aging, age-related changes in clock function could promote oxidative damage.

Circadian dysfunction in Alzheimer Disease

As detailed above, the circadian clock has a panoply of effects on cellular aging, inflammation, and oxidative stress and is impacted by all of these processes. Thus, it is perhaps not surprising that circadian disruption is a common symptom of multiple neurodegenerative diseases (as reviewed in detail elsewhere^3,18,236). Among these, Alzheimer Disease (AD) is the most common age-related neurodegenerative condition and is associated with considerable circadian dysfunction²³⁷. The sequence of pathogenic events in AD is thought to begin with the accumulation of amyloid plaques, composed of aggregated amyloid-beta (Aβ) peptide, followed by the formation of hyperphosphorylated tau protein aggregates (neurofibrillary tangles) within neurons. Plaques appear 10–20 years before disease onset, while tau pathology is apparent within 5 years of first symptom and is closely associated with inflammation and neurodegeneration²³⁸. Several human studies using actigraphy show that circadian and sleep fragmentation occur during the presymptomatic phase of the disease and worsens with disease progression^17,239–241. This pattern is similar to that seen in normal aging, but more severe. Degeneration of the SCN, with subsequent blunting of rhythmic melatonin release, may provide a mechanistic explanation for this exacerbation^83,242–244. However, alterations in BMAL1 methylation²⁴⁵ or direct effects of Aβ on BMAL1 degradation have also been proposed²⁴⁶. In the mouse brain, interstitial fluid Aβ levels exhibit a clear circadian rhythm which is driven by the sleep–wake cycle²⁴⁷. Deletion of Bmal1 causes severe circadian fragmentation, significantly blunts Aβ rhythms, and increases amyloid plaque deposition in a transgenic mouse model of AD²⁴⁸. The exact mechanisms underlying this effect of clock disruption on plaques is unclear. One potential mechanism is through dysregulation of sleep, as sleep deprivation can increase Aβ plaque deposition in mice ²⁴⁷, perhaps by increasing neuronal activity-dependent Aβ production²⁴⁹ or by impairing Aβ clearance by the glymphatic system²⁵⁰. Humans also have diurnal rhythms in cerebrospinal fluid (CSF) Aβ levels²⁵¹. Moreover, sleep deprivation in healthy adults acutely increases CSF Aβ²⁵² and may increase amyloid deposition²⁷. However, kinetic labeling studies suggest this effect occurs via increased Aβ production, rather than impaired clearance²⁵². Extracellular levels of tau also increase during wakefulness and are exacerbated by sleep deprivation in both mice and humans²⁵, while sleep deprivation in mice increases tau pathology^25,253,254. Thus, the clock may influence amyloid deposition and tau pathology in part through effects on sleep.

Aside from sleep regulation, circadian disruption could potentially influence AD or other neurodegenerative disease by any of the previously mentioned mechanisms, including alterations in inflammation, glial function, NAD+/SIRT1 signaling, mitochondrial function, or redox homeostasis. Circadian regulation of protein misfolding and proteostasis in the brain is also relatively unexplored, though the clock has been linked to regulation of autophagy and the proteasome^234,255,256. Accordingly, the core clock could potentially be leveraged as a therapeutic mechanism to optimize these factors in the aging brain and prevent degeneration. Attempts at improving circadian function indirectly through light and/or melatonin supplementation have yielded modest or mixed results on sleep integrity and cognition^257–259, but may offer increased benefit when used in combination^260,261. A variety of drugs which directly target the circadian system are currently being developed and tested^262–264, suggesting that a circadian therapy may be an option in the future.

Conclusions

Impaired circadian function and immune dysfunction, including altered redox homeostasis, coexist consistently across aging and pathological conditions, including neurodegenerative disease^1,3,198. Despite documented regulatory overlap between these areas, the idea that the age-worn circadian system could represent a common link between these phenomena has not been thoroughly explored. Circadian rhythm integrity, including sleep and metabolic cycle competency, is crucial for maintaining brain homeostasis, but breaks down in aging and during neurodegeneration (Fig. 1). At the same time, the circadian clock is vital in optimizing immune function, which is also compromised in aging and neurodegeneration. Elucidation of unifying threads that directly link these observations has the potential to address unanswered questions in several fields simultaneously. These efforts may reveal both innovative therapeutic strategies for tempering the ravages of time and age-related disease while also establishing intriguing avenues for future study.

Figure 1.

Interaction of aging, circadian rhythms, and neurodegeneration. Age-related dampening and dispersion of circadian rhythms (imprecise light red oscillation depicting aged vs robust and precise dark red oscillation depicting young), can promote various pathogenic changes in the brain, including oxidative stress, inflammation, glial activation, and metabolic dysfunction. Disruption of normal sleep-wake patterns can also contribute to these pathologies. Loss of peripheral circadian synchronization can also promote systemic inflammation and impact the immune system, potentially contributing to brain dysfunction. Thus, the circadian system orchestrates brain homeostasis through multiple emerging mechanisms disruption of which may prime the brain for neurodegeneration.