Clock Genes and Altered Sleep–Wake Rhythms: Their Role in the Development of Psychiatric Disorders

There are different categories of physiological rhythms with different time scales, such as ultradian, tidal, circadian, lunar, and seasonal rhythms. A daily rhythm is a regular and predictable phenomenon which is defined as a series of significant physiological changes over 24 h. The biological rhythm consists of two components: the first is exogenous and modulated by environmental factors such as light–dark alternations, sleep–wake, hot–cold, and seasonal change [28,29,30,31]; the second is endogenous and linked to genetic factors (studies in identical twins showed that they had identical biological rhythms [32]). Homologous genes involved in the animal’s activity–rest cycle have been described in humans. It has been observed that the individual tendency to get up and go to bed more or less early was associated with a polymorphism of the circadian locomotor output cycles kaput (Clock) gene. Socio-ecological exogenous factors modulate rhythms and are called synchronizers. Endogenous factors, genetic by nature, underpin the internal biological clock which is responsible for an internal time synchronization coordinating the circadian variations of biochemical, physiological and behavioral parameters.

Circadian rhythms are the best known biological rhythms at the molecular level. The alternation of activity and rest over a period of 24 h (or rest-activity cycle) was observed in drosophila, fish and mammals including rodents (rats, mice, hamsters) and humans. In mammals, this circadian rhythm is generated by a master central clock, mainly reset by ambient light, and is located in the suprachiasmatic nuclei (SCN) of the hypothalamus [33,34]. Circadian regulation of all biological functions is processed through direct or indirect signals (cyclic hormone production) between the suprachiasmatic nuclei and different body structures (brain regions, organs). The body has peripheral clocks located in each organ (heart, lung, liver, muscles, kidneys, retina, etc.) that optimize the function of each organ according to the environmental context allowing therefore adaptation of the organism to environmental changes. Thus, the circadian clocks network has an adaptive function [35]. Both central and peripheral clocks are detectable though both cyclic gene expression and rhythmic physiological processes. These peripheral clocks work independently but must be re-synchronized continuously through the brain’s master clock acting as a real conductor. Furthermore, the neurohormone melatonin is involved in the synchronization (i.e., adjustment of the timing of existing oscillations) of peripheral oscillators; the nocturnal synthesis and release of melatonin by the pineal gland are controlled by the SCN master clock and inhibited by light exposure [36].

The first gene known as being responsible for the circadian rhythm was found in drosophila; it is the period (Per) gene [37]. Subsequent research has highlighted another gene involved in the circadian rhythm: the timeless (Tim) gene [38,39,40,41].

Chemical mutagenesis performed in mammals such as mice allowed identification of the first clock genes: the Clock gene [42] (or Npas2 in neuronal tissue), whose mutation was responsible for a lengthened rest–activity cycle. Other clock genes were then identified in the circadian regulation of the mouse: the brain and muscle ARNT-like protein 1 (Bmal1) gene as a Clock partner, Per1 and Per2 genes, and cryptochrome-1 (Cry1) and Cry2 genes (genes encoding proteins involved in blue light reception in non-mammalian species).

The role of melatonin on the circadian oscillatory rhythms, and in particular as a neuroendocrine synchronizer of molecular oscillatory systems has been documented [36,43,44,45,46]. Studies in animal models showed that melatonin is involved in the regulation of circadian expression of several clock genes, such as Per1, Per2, Bmal1, reverse erythroblastosis virus (REV-ERBα), Clock and Cry1, in both central and peripheral melatonin target tissues. For example, in rat and mouse models, it has been shown that melatonin induces rhythmic expression of Per1, Bmal1, Clock and Cry1 in the pituitary (pars tuberalis) [47,48]. It is noteworthy that Per1 expression is undetectable in melatonin-deficient mice [49,50] and pinealectomy abolishes rhythmic expression of Per1 in the Pars tuberalis (PT) and desynchronizes Per1 and Per2 expression in the SCN [51,52]. Moreover, melatonin synchronizes circadian oscillations in the cardiovascular system by influencing circadian rhythmic expression of both Per1 and Bmal1 in the rat heart [53]. In the pituitary of sheep, melatonin stimulates Cry1 expression and suppresses other clock genes expression [54]. These effects are probably mediated by melatonin (MT) receptors, such as the MT1 receptors. Indeed, in the pars tuberalis of MT1 knockout mice, expression of Per1, Bmal1, Clock and Cry1 was dramatically reduced but not changed in MT2 knockout mice [54]. Furthermore, in the hypothalamic SCN of rats, melatonin treatment induces phase advance of the nuclear receptor REV-ERBα expression [55]. In adipose tissue, melatonin synchronizes metabolic and hormonal function [56] by regulating Per2, Clock and REV-ERBα [57] (REV-ERBα is required for the daily balance of carbohydrate and lipid metabolism [58]). Finally, melatonin regulates oscillation of Clock genes in healthy and cancerous human breast epithelial cells [59] and induces a shift in the 24-h oscillatory expression of Per2 and Bmal1 in cultured fetal adrenal gland [60]. Taken together, these studies underline the major role of melatonin in the regulation of clock genes expression allowing the synchronization of central and peripheral oscillators.

In humans, homologues of Clock, Bmal1, Per and Cry were identified. The Clock gene, the only cloned circadian rhythm gene, is located on chromosome 4 [61]. There are other clock genes involved in circadian regulation, such as retinoic acid receptor-related orphan receptor (ROR)-A and ROR-B, REV-ERB, and casein kinase-1 (Ck1)ε and Ck1δ, which control transcription of Bmal1 gene. Additionally, the albumin-D-site-binding protein (Dbp) is also a circadian clock-controlled gene, involved in the circadian transcriptional regulation of several metabolic enzymes and some transcription factors. Some studies [25,26,27] report altered expression of certain clock genes in cases of shift workers (Per1 and Per2), and free-running/constant conditions (Per1). However, most of the clock genes are affected in case of sleep-deprivation studies [62,63], highlighting the interaction between clock genes and sleep reported by Franken et al. [64,65]. Indeed, clock genes contribute to the homeostatic aspect of sleep regulation and mutations in some clock genes modify the markers of sleep homeostasis and an increase in homeostatic sleep drive alters clock genes expression in the forebrain [64]. It is noteworthy that the expression of clock genes is ubiquitous in humans (central nervous system, spleen, thymus, intestine, heart, lung, etc.), but also in animals (clock genes are found in most cell types among vertebrates).

The discoveries of the different clock genes led to the proposal of a molecular clock model based on a feedback loop running over 24 h [66]. The molecular mechanisms underlying the circadian rhythms are rather similar regardless of the species. They include enhancer elements, repressor elements and control loops involving phosphorylation-dephosphorylation, methylation, acetylation reactions and the specific protein dimerization [67,68]. In mammals, this molecular circadian system is present in the hypothalamic central clock, the suprachiasmatic nuclei, and in secondary clocks within the brain and peripheral organs. Light perceived by the retina leads to changes in transcription of certain genes of the circadian rhythm in the master clock and shows the ability of the body to adjust to a change in the cycle of the photoperiodic environment. This master clock synchronizes multiple peripheral circadian oscillators via mechanisms that remain to be better ascertained.

The molecular loop of the circadian clock seems to imply in a general way two types of mechanisms: transcriptional (transcriptional regulation of genes at the DNA level, i.e., their copy in the form of RNA messenger) and post-transcriptional (regulating steps downstream of the transcription) (see Figure 1 [69]).

Transcriptional mechanisms: these mechanisms can be summarized by autoregulatory feedback loops. The first is a primary negative feedback loop: they rely on a pair of positive elements and a pair of negative elements. In mammals, the positive elements are two proteins, CLOCK and BMAL1 transcription factors which heterodimerize each other. The two negative elements are also two proteins, PERIOD (PER) and CRYPTOCHROME (CRY). In a neuron of the suprachiasmatic nucleus, there is little PER and CRY in the morning. The CLOCK–BMAL1 complex activates maximally the transcription of Per and Cry genes. However, the PER and CRY proteins do not accumulate immediately due to their instability, which allows the cycle not to be interrupted prematurely. Following interactions with other proteins they are gradually stabilized during the day, then they heterodimerize and finally migrate into the nucleus [70], where they inhibit the transcriptional activity of the CLOCK–BMAL1 heterodimer and thus their own transcription [71,72]. The Per and Cry genes being less and less active, the PER and CRY proteins are less and less produced. Their amount reaches a maximum at the beginning of night, and then decreases. At the same time, the CLOCK–BMAL1 heterodimer gradually regains its activity during the night. Thus, the molecular mechanism of the clock consists principally in a feedback loop comprising a positive component (with the CLOCK–BMAL1 heterodimer) and a negative component (with the PER–CRY heterodimer). A complete cycle of this loop lasts about 24 h. CLOCK and BMAL1 heterodimers are also involved in a daily transcription of many clock-controlled genes (CCGs) in different peripheral tissues [73,74,75]. In addition to the primary feedback loop, another regulatory feedback loop is formed by the orphan nuclear receptors REV-ERBα and RORα. In the nucleus, REV-ERBα competes with RORα for binding to the ROR-responsive element (RORE) in the Bmal1 promoter. Whereas RORα activates transcription of Bmal1, REV-ERBα represses it. Consequently, the cyclic expression of Bmal1 is achieved by both positive and negative regulation of RORs and REV-ERBs, respectively. This secondary feedback loop is called the “stabilizing loop”.

Post-transcriptional mechanisms: the proper functioning of the circadian loop requires that PER and CRY proteins disappear in due course, once they have served their purpose. As long as they inhibit the activity of positive elements (CLOCK and BMAL1), a new cycle cannot start. Several post-transcriptional processes thus affect the ability of PER and CRY to act on CLOCK and BMAL1. The most studied aspect is the modification of these proteins by phosphorylation and dephosphorylation. Several kinase proteins include PER and/or CRY among their targets, and each one of the two proteins can be phosphorylated on many distinct sites. They accelerate or slow down their transfer to the proteasome. Proteins to be degraded are “marked” by ubiquitin. Post-transcriptional regulations of the clock proteins of the main molecular loop make it possible to ensure intracellular traffic, functionality and degradation of clock proteins which are crucial for the functioning of the molecular loop over 24 h [76].

The mechanisms underlying the association between alterations in circadian rhythms and mood disorders are still unclear. However, current scientific studies provide food for thought. According to Jackson et al. [107], sleep disorder is the most common prodrome of mania and one of the six most common prodromes of depression in bipolar disorder (bipolar type I disorder is defined as an isolated manic episode, and bipolar type II disorder is defined as manic and depressive episodes) [108]. Between mood episodes, the sleep–wake cycle remains disrupted in patients with bipolar disorder who have some difficulty falling asleep and frequent night awakenings [109]. These patients have a significantly more eventide sleep pattern than control subjects, in other words, they go to bed later and have more difficulty getting up early compared to healthy controls [110].

McClung [111] studied the involvement of the Clock gene in manic episodes in mice. She compared Clock mutant mice to mice with a normal Clock gene. Mice in which the gene had been modified had mania-like behavior. Those mice were hyperactive, less anxious and less depressed. They slept less and showed greater brain activity in response to sugar water, cocaine and a mild electrical stimulation of the brain. When researchers added lithium in their water, these mice did not show their manic behavior anymore and started to act like the mice with a normal Clock gene. A lithium salt treatment could affect circadian rhythms by modulating the expression of clock genes [112,113,114,115]. Indeed, lithium salts might inhibit the expression of glycogen synthase kinase 3β (GSK-3β) gene [116], a gene involved in the brain’s biological clock [117]. Kaladchibachi demonstrated that both genetic and pharmacological reduction of GSK-3 activity in mice have a specific effect on the circadian transcriptional oscillation consisting of mPer2 period lengthening, indicating a delay in phase [116]. Moreover, it has been shown that GSK-3β is rhythmically expressed in the SCN and liver of mice, and that it undergoes a daily cycle in phosphorylation in vivo. Lithium chloride treatment inhibits the GSK-3β expression and results in a phase delay of Clock gene expression in fibroblasts [118]. Bipolar spectrum disorder has been associated with variations in the Clock gene. A single-nucleotide polymorphism (SNP) in the 3-flanking region of the Clock gene (3111 T to C) is associated with a higher recurrence rate of bipolar episodes [16]. This SNP was also associated in bipolar disorder and/or antidepressant treatment with sleep problems (insomnia and decreased need for sleep) [17,119]. Some studies reported other clock genes associated with bipolar disorder (such as Bmal1 and Per3) [96]. Furthermore, a SNP in Bmal1 and a SNP in Tim have also been identified as having a link with bipolar disorder [99]. One South Indian study showed that the occurrence of the five repeat alleles of Per3 may be a risk factor for bipolar disorder onset in this ethnic group [120]. The coding region of Per3 gene contains a variable number tandem-repeat (VNTR) polymorphisms which has been associated with diurnal preference, sleep structure and sleep homeostasis in healthy individuals. In a homogeneous sample of patients with bipolar type I disorder (occurrence of one or more manic episodes or mixed episodes), they observed that Per3 VNTR influenced age of onset: earlier onset in homozygote carriers of Per35 variant, later onset in homozygotes for Per34, and intermediate onset in heterozygotes. Sjöholm et al. [121] studied four single-nucleotide polymorphisms (SNPs) of Cry2 gene in a cohort of bipolar patients, some with fast cycles. They observed that Cry2 is associated with rapid cycling in bipolar disorder patients. Rapid cycling in bipolar disorder is defined according to the 5th version of Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as four or more mood episodes in any combination or order within any year in the course of the illness. The A allele of rs10838524 (Cry2 SNP) was significantly overrepresented among bipolar disorder cases with rapid cycling compared to controls. There was a significant trend in the association between the A allele of rs10838524 and rapid cycling (p = 0.0076) and this allele increased the risk for rapid cycling both in a homozygote and a heterozygote form (that is dominant model) compared to controls. Two closely related clock genes, retinoid-related orphan receptors α (RORA) and β (RORB) are involved in a number of pathways including neurogenesis, stress response, and modulation of circadian rhythms. A study reports that four intronic RORB SNPs showed positive associations with the pediatric bipolar phenotype and suggests that clock genes in general and RORB in particular may be important candidates for further investigation in the search for the molecular basis of bipolar disorder [122]. Further studies are necessary to better assess the role of RORB in bipolar spectrum disorder and explore if this gene is particularly relevant for this disorder compared to other clock genes.

Several genetic studies suggest that certain polymorphisms of clock genes are more common in bipolar disorder than in the general population [123,124]. The clock genes might therefore be viewed as genetic vulnerability factors to bipolar disorder [125]. These results need to be examined further.

The litterature review on clock genes and bipolar spectrum disorder is presented in Table 1.

Many sleep disorders are associated with attention deficit hyperactivity disorder (ADHD), both in adults and children. In particular, the following sleep problems occurring at sleep–wake transition were observed: bed-time refusal, delayed sleep-onset, and early awakenings [148]. Differences in sleep problems were found as a function of ADHD subtype: children with ADHD inattentive type (ADHD-I) had the fewest sleep problems and did not differ from controls, children with ADHD combined type (ADHD-C) had more sleep problems than controls and children with ADHD-I. Daytime sleepiness was greatest in ADHD-I and was associated with sleeping more (not less) than normal [149].

Several studies reported genetic associations between polymorphism (rs1801260) at the 3′-untranslated region (3′-UTR) of the Clock gene and psychiatric disorders [17,90] (see Table 3). Concerning the association between the rs1801260 polymorphism and ADHD, Kissling et al. [90] found at least one T-mutation being the risk allele in Caucasians of western European origin and German background.

Xu et al. [91] conducted a study which investigated a previously reported discovery of an association between a single nucleotide polymorphism in the 3′-UTR region of the Clock gene (rs1801260) in two independent samples of ADHD probands from UK and Taiwan (aged 5 to 15). They found a significant over-transmission of the T allele of rs1801260 SNP, which is associated as a risk allele for delayed sleep phase syndrome [145] in ADHD cases in the Taiwanese population. No association was observed between this polymorphism and ADHD in the UK sample. However, they did find evidence for increased transmission of the T allele of rs1801260 in the Taiwanese samples. Their findings support the hypothesis that genetic variation in the 3′-UTR region of Clock gene might be a risk factor for the development of ADHD, particularly in the Taiwanese sample studied. Therefore, more functional polymorphisms of this region should be investigated in other independent studies using larger samples.

Schizophrenia is often associated with sleep alterations [88,150], observed also in untreated patients [151]. Patients with schizophrenia have a desynchronization between the sleep–wake rhythms and melatonin profiles [88,152,153], the rhythms of body temperature [154], and serum levels of tryptophan and prolactin [155]. These anomalies suggest strongly disturbed circadian rhythms in these patients [156]. Studies of polymorphisms of clock genes or alterations in the regulation of these genes in schizophrenia are rare. The litterature review on clock genes and schizophrenia is presented in Table 4.

The expression of Per1 mRNA in the temporal lobe in individuals with schizophrenia decreases significantly compared with age-matched healthy controls [98]. Another study suggested that Per3 but not Per2 abnormalities were associated with schizophrenia [99]. In a sample of 145 Japanese individuals with schizophrenia compared with healthy controls, it was reported that the T3111C polymorphism of the Clock gene presented a transmission bias. The T3111C polymorphism of the Clock gene might be associated with aberrant dopaminergic transmission in the suprachiasmatic nucleus, which is presumably involved in the pathophysiology of schizophrenia [157]. More recently, two studies focused on white blood cells and fibroblast from smaller samples: Sun et al. [100] reported altered expressions of Per1, Per2, Per3 and Npas2 in white blood cells in individuals with schizophrenia. Indeed, compared with healthy controls, schizophrenia patients presented disruptions in diurnal rhythms of the expression of Per1, Per3, and Npas2, accompanied by a delayed phase in the expression of Per2 and by a decreasing in Per3 and Npas2 expression. Johansson et al. [101] reported a loss of rhythmic expression of Cry1 and Per2 in fibroblasts from individuals with schizophrenia compared to cells from healthy controls.

To date, the associations between clock genes and schizophrenia are not clear.

Surveys of parents show that the prevalence of sleep problems in ASD is 50%–80% compared to 9%–50% in age-matched typically developing children [7,8,9,10,11,12,13,14]. Recent results showing abnormal melatonin secretion in ASD children may change the initial disregard of these disorders and suggest a possible key role of the clock and circadian regulations in ASD. Several independent groups detected abnormal melatonin levels in ASD [158,159,160,161,162,163]. These studies conducted on independent autism samples and using different methodologies indicate that abnormally low melatonin level is a frequent trait in ASD. More precisely decreased nocturnal as well as diurnal levels have been reported in individuals with ASD; trial studies support therapeutic benefits of melatonin use in ASD [83,84,161]. Nevertheless, both the underlying cause of this anomaly and its relationship with ASD (cause or consequence?) remain still unexplained.

Many studies have advocated a genetic etiology for autism [164] involving in particular synaptic genes related to synaptic cell adhesion molecules NLGN3, NLGN4, and NRXN1 and a postsynaptic scaffolding protein SHANK3. This protein complex is crucial for the maintenance of functional synapses as well as the adequate balance between neuronal excitation and inhibition. Among the factors that could modulate this pathway there are genes controlling circadian rhythms. However, no direct association has been reported between clock genes and NLGN3, NLGN4, NRXN1 and SHANK3 in humans, but epistasis mechanisms cannot be ruled out. SHANK3 expression is modulated by melatonin concentration and the modulation is brain structure-dependent [165]. Concerning relationships between sleep disorders and NLGN3, NLGN4, NRXN1 and SHANK3 could be a pleiotropic effect on sleep of genes associated with brain disorganization. It is noteworthy that possible interplay of synaptic and clock genes may increase the risk of ASD [166]. The involvement of clock genes in ASD was first suggested by Wimpory et al. [167] who stated the hypothesis that anomalies in clock genes operating as timing genes in high frequency oscillator systems may underline timing deficits that could be important in the development of autism spectrum disorder, notably in autistic communication impairment. To test this hypothesis, Nicholas et al. [89] screened single-nucleotide polymorphisms in 11 clock/clock-related genes in 110 individuals with ASD and their parents. A significant allelic association was detected for Per1 and Npas2. It should be noted that it was a small population and results were not significant after correction for multiple testing. However, the association between clock genes and ASD was confirmed by a more recent study [168] reporting also mutations in other circadian clock genes (Per2, Per3, Clock, Bmal1, Tim, Cry1, Cry2, Dbp and Ck1ε) in ASD patients. Taken together, these findings suggest that the circadian rhythm abnormalities observed in ASD may be linked to abnormalities of the circadian clock genes. Finally, it is noteworthy that clock genes are related to the protein association network described in ASD by Roubertoux and Tordjman [169] either through direct gene × gene interactions or protein × protein interactions.

The litterature review on clock genes and autism spectrum disorder is presented in Table 5.