Circadian disruption, clock genes, and metabolic health

Early investigations into circadian rhythms relied on easily measured physiological aspects of circadian rhythms, such as body temperature and locomotor activity (,). In animals, genetic drivers of circadian rhythms were first revealed through forward genetic screening inwhich identified mutants with short, long, and arrhythmic circadian locomotor activity patterns (,). These studies revealed that thegene and its protein product PER, a founding member of the PER-ARNT-SIM (PAS) superfamily, was an essential component of the circadian clock. Subsequent studies in mammals confirmed thesefindings and led to the discovery of additional CCGs,,, andas well as three orthologs of:(,). Notably, CLOCK, BMAL1, and PER were all identified as members of the PAS family of proteins, suggesting hetero- and homodimeric interactions driven by similar domain structures lie at the mechanistic heart of the molecular clock () (,,,–). ^{40 41 40 41 37 38} Figure 1 ^{5 36 37 42 45} Drosophila melanogaster, per PER CLOCK BMAL1 CRY1/CRY2, Drosophila PER (PER1/PER2/PER3)

As mutant models were generated in mice, the circadian clock’s role in metabolic health–related pathologies emerged.shows an overview of select studies providing evidence that clock genes are linked to metabolic health pathologies in animal models. For example, in an experiment with mice mutant for thegene, mutants displayed altered food intake timing with ad libitum feeding, consuming more calories outside the active phase (). These animals also had dampened activity rhythms and developed obesity and MetS, including high cholesterol, high triglycerides, high blood sugar, hypoinsulinemia, and elevated leptin during the rest phase. Similar metabolic perturbations were seen in other mouse models with induced mutations in molecular clock components. For example,-null mice displayed blunted postprandial insulin responses, decreased gluconeogenesis, and loss of typical glucose and triglyceride rhythms (,). Likewise-mutant mice developed without normal glucocorticoid rhythms () anddouble mutants displayed altered liver metabolism and altered patterns of circulating growth hormone (). Finally, when compared with WT, transgenic mice generated to overexpressexhibited treater hyperglycemia without increased weight gain (). Table 1 ^{37 36 45 46 44 47} Clock Bmal1 , Per2 Cry1/Cry2 Cry1

Further examinations designed for understanding the relationship between clock and glucose homeostasis inmutant mice have revealed altered expression of genes involved in pancreatic islet cell development and insulin signaling (). At eight months, these mutants displayed elevated serum glucose compared with WT counterparts, seemingly consequent to defective glucose-stimulated insulin release from the-mutant islets. Interestingly, findings from young-mutant mice aged two to three months demonstrate that a compensatory component exists early on, driven by themutation in other organ systems such as the liver, leading to age-related emergence of insulin resistance and underlying deficits in insulin secretion manifested as pathogenic hyperglycemia (i.e., diabetes) (). An explanation for the age-dependent hyperglycemia phenotype may come from the observation that-mutant andnull mice exhibit loss of typical glucose and triglyceride rhythms, with impaired and abolished gluconeogenesis, respectively (). Because gluconeogenesis occurs predominantly in the liver, the contrasting effects between the paired metabolic organs regulating glucose homeostasis in-mutant andnull mice caused a “masking” of the phenotype when the animals were young. Examples such as this highlight the challenges in disentangling the role of the circadian clock when using global mutant/null models (). Clock- Clock Clock Clock Clock Bmal1- Clock Bmal1- ^{45 45 36 36}

To address this issue, pancreas-specific-null mice () were employed to knock out the circadian clock in pancreatic islet cells. These experiments revealed that, even when mice are at a young age, the islet cell clock influences insulin secretion, glucose levels, and glucose tolerance (i.e., hypoinsulinemic diabetes) (). Thus, the discrepancy in young versus old-mutant mice may be due to the circadian clock yielding different tissue-specific effects in metabolic processes such as glucose metabolism, which is particularly relevant for the pathologic processes of diabetes and MetS. Further, many metabolic conditions emerge in humans with aging, e.g., type 2 diabetes (T2D), nonalcoholic steatohepatitis, etc. Concordantly, disruption of the circadian clock in humans results from global disruption/desynchrony. Therefore, it is plausible that the early impact of circadian desynchrony on MetS in humans is difficult to completely understand given the organ-specific functions of the clock. In fact, disease states may only become apparent after long-standing desynchrony, which is supported by studies demonstrating increased risk of developing diseases such as diabetes with increasing duration of shift work (–). Bmal1 Bmal1; Pdx1-Cre fl/fl Clock ^{45 48 51}

Glucose homeostasis is influenced by cellular-specific gene signaling mechanisms controlled by specific CCGs. For example, a 2015 study by Perelis et al. examined pancreatic β cells from mice with intact or disrupted BMAL1 expression. For intact cells, CLOCK/BMAL1 dimers were shown to bind to regulatory sites (CCGs) in islet cells to drive transcription of genomic targets (COGs) in these cells that were distinct from those of other cells in the liver. They further found that mice with disrupted BMAL1 expression developed glucose intolerance, suggesting a direct genetic mechanism controlling diabetes as one measure of metabolic health (). Further, the majority of CLOCK/BMAL1 binding sites identified within β cells are not commonly identified in the liver or other tissues — supporting the tissue-specific role of the clock in altering metabolic health through active enhancer regions and epigenetic chromatin regulation of unique genes within cells and suggesting that polymorphisms or alterations contribute to metabolic disruption. Tissue-specific subsets of circadian genes reinforce this idea. Despite the substantial number of cycling genes in each tissue (), only a small number of common genes are rhythmic in all tissues (). ^{52 5 11}

A large body of work has been undertaken for understanding the consequences of external/environmental circadian disruptors in mice (). These protocols are meant to mimic circadian disruptors commonly experienced by humans, such as altered timing of light exposure, activity, sleep, or food intake (,,–). Experimental studies aiming to mimic shift-work exposures through phase-shift and time-restricted feeding protocols suggest that exposure to these various circadian disruptors can alter metabolic health and CCG expression (–). A 2011 study found that phase-shift protocols among Sprague-Dawley rats were associated with changes in the acceleration of multiple indicators of T2D, particularly for animals with altered β cells (). Similarly, a 2021 study of circadian disruption via chronic jet lag investigated transcriptional changes in mice and found that 5% of the transcriptome in the pancreas is regulated by CCGs and that external phase shifting in mice alters regular rhythmic control of genes in the pancreas associated with insulin and enzyme regulation (). ^{53 17 30 54 56 56 58 58 56}

Emerging evidence from experimental food-intake and timing models illustrates the multiple mechanisms by which circadian disruption affects metabolic health. For example, WT mice subjected to misaligned food intake relative to active and inactive periods display accelerated weight gain similar to that of humans. Along with lowered amplitude of clock genes, mice consuming a high-fat diet during their inactive period gained substantially greater weight and had a higher body fat percentage than mice consuming the same diet during their active period (). In contrast, when a time-restricted feeding (TRF) protocol restricted feeding to the active period, mice on high-fat diets were protected from weight gain and the increased markers of adverse metabolic health experienced with ad libitum feeding. In several studies, TRF prevented obesity and impaired glucose tolerance, restored insulin sensitivity, and protected against inflammation and hepatic steatosis (–). Moreover, during the active phase, TRF restored normal hepatic glucose metabolism elicited by the ad libitum high-fat diet. TRF also restored the oscillation of metabolic regulators in the liver that were dysregulated with an ad libitum high-fat diet. Moreover, it restored CCG expression amplitude. Thus, eating at times misaligned to circadian rhythms leads to altered metabolic health, while TRF ameliorates this pathology. ^{18 59 61}

A combination of TRF and diet quality affects metabolic health by regulating CCGs, as seen in a series of critical mouse experiments. Animals were subjected to high-fat diets with either ad libitum feeding, TRF confined to an active period, or TRF confined to an inactive period (). Despite similar activity and timing of food intake overall, a high-fat diet during the inactive period caused increased body mass and lower energy expenditure. Carbon-labeling studies demonstrated that the high-fat diet during the inactive period resulted in decreased glycolysis in adipocytes and dampened oscillation of CCGs compared with the high-fat diet during the active period. A series of genetically engineered mouse models demonstrated that the adverse metabolic consequences of a high-fat diet during the inactive period were related to impaired adipocyte thermogenesis. The mediators of thermogenesis in the adipocytes were found to be regulated by the core molecular clock and responsible for maintaining metabolic health in the high-fat diet/active-phase TRF group. This study supports the idea that optimal metabolic health can depend on the alignment of food intake and the biological rhythms of cellular thermogenesis controlled by the molecular clock. Moreover, these data help us to understand the contributions of high-risk dietary regimens and circadian misalignment. These results also underscore important considerations for humans experiencing circadian disruption, leading to the suggestion that mitigation of risk for adverse metabolic health could include strategies to align specific nutrient intake with the internal clock and highlighting the importance of developing simple and reproducible methods of monitoring organ-specific reentrainment that may facilitate optimizing such an approach. ⁵⁹

While human studies often lack the specificity to advance mechanistic insights, coupled with animal-based findings, they provide crucial foundational evidence linking circadian disruption to metabolic health, as summarized in. Some of the first studies to suggest circadian disruption alters metabolic health come from epidemiological investigations of shift workers. For example, an examination of 54,724 participants in the Nurses’ Health Study II found that individuals exposed to any duration of night-shift work had increased odds of obesity, higher total calorie intake, and shorter sleep durations than those who had never worked night shifts, after adjusting for age and socioeconomic status (). In a separate study, shift workers displayed increased odds of being overweight or obese and were more likely to report insufficient sleep than individuals working traditional schedules. Moreover, there was a stronger association between shift work and overweight conditions among individuals reporting insufficient sleep, suggesting a protective effect of adequate sleep duration during shift work (). A much smaller study of 24 women (12 night-shift and 12 day-shift workers) found that night-shift workers had greater fat mass, larger energy intake, impaired sleep, lower insulin sensitivity, and higher triglycerides compared with their day-worker counterparts (). Night-shift workers also had higher postprandial ghrelin levels and lower bloodstream levels of xenin, a gut-derived hormone, offering preliminary evidence of shift work and metabolic health in the form of appetite regulation (). Table 2 ^{62 63 64 64}

Another significant circadian disruptor linked to metabolic health is “social jet lag,” measured as the difference in midsleep time on nights before work or school and those before work- or school-free days (). One study of 815 nonshift workers born between 1972 and 1973 in New Zealand identified an association between social jet lag and numerous metabolic health indicators, including BMI, fat mass, and waist circumference (). In independent studies, individuals with social jet lag displayed greater adiposity, lower HDL cholesterol levels, higher triglycerides, increased insulin resistance, and higher fasting plasma insulin even after controlling for behavior and sleep quality. Moreover, individuals with a tendency to be most active in the evenings and delay sleep onset, known as “evening chronotypes,” had lower HDL cholesterol levels, consistent with similar findings that evening types have a heightened risk for cardiovascular disease, given their proclivity to circadian-disrupted schedules and social jet lag in particular (,). ^{14 65 66 67}

Additional insights come from several studies among non–shift-work female populations. A cross-sectional analysis of non–shift-working middle-aged women revealed a positive association between bedtime variability and bedtime delay with increased insulin resistance (). After more than five years of follow-up, greater bedtime delay predicted higher insulin resistance, suggesting that both acute- and chronic-inconsistent sleep timing induce metabolic dysfunction (). These observations were echoed among a study of older women over the age of 80 that found intraindividual variation in objectively measured wake time, sleep duration, sleep delays (social jet lag), and the midpoint of sleep were associated with alterations in body composition, including percentages of increased fat mass and lower lean mass as metabolic health indicators (). ^{68 68 69}

Experimental phase-shift studies among adults requiring short-term alterations of regular sleep patterns provide additional evidence for links between circadian disruptors and markers of metabolic health. In one such study, 21 healthy adults (10 men and 11 women) of varying ages were subjected to circadian disruption, achieved via imposed 28-hour days and 5.6 hours per night sleep restriction. After three weeks, circadian-disrupted participants had increased fasting and postprandial glucose levels and significantly decreased resting metabolic rates compared with baseline, a pattern observed in both young and old subjects. These changes normalized after nine days of return to standard sleep/wake patterns (). In a separate study, two experimental groups of healthy young adults were exposed to two five-day weeks of five hours per night of sleep restriction, with one group permitted an interim two-day weekend of ad libitum “catch-up” sleep, while the other group continued sleep restriction for the entire study period. A control group with normal sleep (nine-hour daily sleep windows) was also included. Despite sleeping an extra three hours each weekend day, the weekend catch-up group experienced reduced insulin sensitivity relative to controls. While all three groups increased their energy intake compared with baseline calorie-controlled meals, only the control group did not experience weight gain. Moreover, both disrupted groups increased after-dinner snacks on days following sleep restriction. These findings suggest that a weekend catch-up sleep after sleep debt, akin to social jet lag, is not sufficient to correct the metabolic dysfunction accrued during sleep restriction (). ^{70 71}

There is mounting evidence that TRF, particularly TRF during the active phase and aligned with circadian rhythms, is also beneficial for human metabolic health, which is similar to observations in mice. A meta-analysis of 19 TRF clinical trials found TRF diets significantly reduced body weight, decreased fat mass, preserved fat-free mass, and reduced systolic blood pressure, triglycerides, and fasting glucose (). TRF is thus seen as a promising treatment for changing metabolic health, even without reducing the total calorie intake, hence emphasizing the importance of circadian timing in metabolic processes (). Further, in a disrupted sleep and TRF study, participants with three days of 28-hour day simulated night-shift protocols were divided into two groups and fed meals in sync with their shifted/disrupted schedule or their usual dietary intake schedule. The group fed in sync with their shifted/disrupted, 28-hour cycle (eating meals between midnight and 4 am, when they would typically be asleep) showed impaired glucose tolerance (). In parallel with experimental animal findings, human study results further establish the role of food intake as a strong entraining cue for the circadian clocks in the periphery and indicate metabolic outcomes can be improved by aligning the food-intake window with the active phase while removing windows of food intake later in the day or night. ^{72 72 73}

Circadian disruption is also associated with altered clock gene expression in human studies, offering clues to the biological mechanisms linking circadian disruption with metabolic health. A small 2019 study including 18 female nurses revealed fewer rhythmic genes observed in the PBMCs of rotating night shift workers (= 9) compared with nurses working day shift (= 9). Moreover, phase desynchrony of core body temperature, peak cortisol, and dim-light melatonin onset were also observed ().A more extensive investigation of 60 nurses using a single measurement time point identified differential expression of nearly all CCGs, including transcripts of the CCGs, andgenes. These CCGs were dysregulated in morning blood draws taken among rotating shift nurses and compared with those day-shift nurses (). A small study of eight participants exposed to a simulated night-shift work protocol, including three days of ten-hour sleep period delay, repeated blood draws over two 24-hour periods, and transcriptome-wide analysis of PBMCs, showed a significant reduction of rhythmic transcripts, includingpostdisruption compared with baseline (). n n BMAL1, CLOCK, NPAS2, PER1, PER2, PER3, REVERBA, CRY1 CRY2 PER3 ^{74 75 76}

Interestingly, similarly to what was found in the 2019 nurses’ study (), some CCGs, includingandmaintained rhythmicity. Genes that remained rhythmic after disruption displayed dampened oscillation patterns that matched habitual sleep and wake times rather than the newly disrupted schedule. These results show that, while circadian function may remain largely intact in a simulated night shift, many COGs lost expression amplitude, influencing several important metabolic pathways, including immune-system regulation (). ^{74 76} PER1 BMAL1,

In a separate forced desynchrony experiment, 22 participants underwent baseline assessment followed by several 28-hour days of forced desynchrony, with sleep onset pushed back four hours further each night. During disruption, 24-hour melatonin rhythms remained largely preserved as compared with baseline rhythms. In contrast, there was a six-fold reduction in PBMC-measured circadian transcripts, including, and(). Experimental sleep-deprivation studies also aimed at identifying resultant transcriptome alterations have identified similar reductions in the circadian rhythmicity of genes and changes in the expression of genes from chromatin-remodeling immune and stress-response pathways. In a study of 26 sleep-restricted individuals who were limited to six hours per night, gene expression in thousands of transcripts was altered, and the number of genes with detectable circadian patterns was reduced by 20%. Again,, andwere significantly impacted. Additionally, several metabolic and oxidative stress–related genes were altered after the sleep-restriction protocol (). It is important to note that these transcriptome-wide analyses were conducted on relatively small numbers of individuals with short desynchrony and sleep-deprivation protocols. Therefore, the longer-term impacts of these exposures are unclear. Nonetheless, these findings collectively support the idea that circadian misalignment has widespread effects on the transcriptome and has a differential impact on central versus peripheral clocks, highlighting the potential for adverse metabolic health from disruption of the habitual sleep/wake cycle (see a summary of findings in). BMAL1, CLOCK PER3 CLOCK, PER1, PER2, PER3, CRY2 RORA ^{77 78} Table 3

The connection between clock gene expression and metabolic health may be bidirectional. Evidence from various human studies suggests that a state of adverse metabolic health itself, including elevated BMI, can alter the expression of CCGs in a tissue-specific fashion. In one study of 21 lean and 28 morbidly obese female nonshift workers without diabetes, examination of 24-hour adipose gene expression revealed that obese individuals displayed altered circadian expression of many CCGs, includingand, compared with healthy, lean subjects. Positive correlations were found among all subjects betweenand BMI/waist circumference, CLOCK and LDL cholesterol, andwith HDL cholesterol. An interesting conclusion from this study was thatis an important gene associated with metabolic health (). CLOCK, BMAL1, PER1, CRY2, REVERBA REVERBA RORA REVERBA ⁷⁹

Further evidence of connections between metabolic phenotypes driving circadian disruption comes from studies showing that weight loss alters CCG expression patterns. A 2020 study examined differential mRNA levels and expression of CCGs in skeletal muscle among 23 obese patients (5 women and 18 men) undergoing gastric bypass surgery and 14 normal-weight controls (6 women and 8 men). Males in the obese group had significantly lower, andexpression than lean male controls (). Obese women exhibited downregulatedmRNA levels compared with lean female controls, butexpression was restored to lean-control levels following gastric bypass–induced weight loss (). Interestingly, while changes in CCG expression varied by sex, additional research is needed to replicate these findings. A second study examined participants’ expression of CCGs in adipose tissue before and after hypocaloric diet–induced weight loss (). After eight weeks, 50 subjects who lost 8% or more of their body weight saw significant increases inexpression compared with baseline, with similar changes to genes regulating fat metabolism, autophagy, and inflammatory responses (). CLOCK, CRY1 CRY2 CRY1 CRY1 PER2 ^{80 74 81 81}

Human studies also suggest that even limited, short-term exposure to circadian disruptors can alter metabolic pathways and clock gene expression. One such study, in 2018, subjected 14 healthy men to three days of normal sleep, followed by three days of reversed day/night schedules. After three days of disruption, fasting glucose and free fatty acids were significantly elevated compared with what occurred with normal sleep conditions (). In addition, a significant transcriptional alteration in PPAR signaling was observed, leading to the hypothesis that misalignment promotes a preference for intramuscular fatty acid metabolism over glucose metabolism. Interestingly, after the three days of misalignment, CCGs had not reentrained to the reversed schedule and remained aligned to the regular day/night schedule (). In another 2015 investigation, 15 healthy male participants were exposed to acute 24-hour sleep deprivation, and increased methylation in theandgenes in adipose tissue was observed, suggesting that methylation is also a mechanism for the downregulation of CCG expression (). Furthermore, after sleep deprivation, expression ofandin skeletal muscle was decreased, and postprandial plasma glucose concentrations were increased (). These studies also provide evidence that even short-term misalignment of the circadian clock from standard behavior patterns, from a single night of wakefulness to a few days of misalignment, can be linked to metabolic changes in humans. ^{82 82 83 83} CRY1 PER1 BMAL1 CRY1

Studies of CCG expression have also shown that TRF studies can help combat circadian misalignment’s negative consequences, suggesting new opportunities for preventing and treating adverse metabolic health outcomes. In a 2019 crossover study of 11 obese participants, the efficacy of TRF was investigated (). Comparison of four-day ad libitum feeding (8 am–8 pm) with an early daytime feeding window (8 am–2 pm) revealed that TRF resulted not only in increased expression of CCGs (, and), but also elevated ketones, elevated cholesterol levels, reduced mean blood glucose levels, and reduced glucose spikes throughout the day. This was despite equal calories consumed between conditions. While many studies have supported using TRF to improve metabolic health, this experiment provided key insights by measuring and associating CCG expression with improvements in lipid metabolism and glucose regulation (). A similar randomized case-crossover study of TRF examined 11 men who were either overweight or obese and found that TRF improved daytime insulin profiles and reduced night-time glucose levels. The oscillation patterns of CCGs, including,, -, and -, andand -were unchanged between 15-hour free-feeding and 8-hour TRF conditions (). However, the authors identified an increase in the amplitude of oscillating muscle transcripts related to amino acid transport, suggesting that TRF has multiple health benefits (). ^{84 84 85 85} BMAL1, CRY1, CRY2 RORA CLOCK, BMAL1, CRY1 PER1 2 3 REVERBA B,

As in mice, genetic variation in CCGs and genes that modify the clock can influence rhythms and metabolic health in humans. For example, in a meta-analysis of cohort studies from the Cohorts for Heart and Aging Research in Genomic Epidemiology Consortium, associations among sleep duration, genetic variants of core clock and other circadian genes, and cardiometabolic traits were identified (). In carriers of the T allele of the melatonin receptor 1B, long sleep duration (9+ hours per night) was associated with increased BMI. Additionally, in carriers of the A allele of(SNP rs11605924), sleep duration was positively associated with HDL cholesterol level (). Additional evidence supporting the connection can be seen in investigations of SNP mutations and their role in metabolic health, reviewed in Škrlec et al. (). The growing evidence of CCG SNPs underlines the complex relationships between CCGs and metabolic health.summarizes several examples and study findings. ^{86 86 87} Table 4 CRY2