Recent advances in modulators of circadian rhythms: an update and perspective

Abstract Circadian rhythm is a universal life phenomenon that plays an important role in maintaining the multiple physiological functions and regulating the adaptability to internal and external environments of flora and fauna. Circadian alignment in humans has the greatest effect on human health, and circadian misalignment is closely associated with increased risk for metabolic syndrome, cardiovascular diseases, neurological diseases, immune diseases, cancer, sleep disorders, and ophthalmic diseases. The recent description of clock proteins and related post-modification targets was involved in several diseases, and numerous lines of evidence are emerging that small molecule modulators of circadian rhythms can be used to rectify circadian disorder. Herein, we attempt to update the disclosures about the modulators targeting core clock proteins and related post-modification targets, as well as the relationship between circadian rhythm disorders and human health as well as the therapeutic role and prospect of these small molecule modulators in circadian rhythm related disease.


Introduction
Circadian rhythm is the result of natural selection during the longterm evolution of organisms, enabling organisms to better adapt to changes in the external environment 1,2 . Various behaviours and physiological functions of the body show obvious circadian rhythms, such as the sleep-wake cycle 3,4 , food intake and other autonomous activities 5 , as well as physiological activities including blood pressure 6 , blood lipids, coagulation-fibrinolysis balance, heart rate 7,8 , body temperature 9 , locomotor activity 10,11 , hormone levels 12 , cell metabolism 13 , and cell proliferation 14,15 . The generation, maintenance, and regulation of circadian rhythms depend on the synergy of the circadian clock system, circadian input system, and circadian output system at the overall level ( Figure 1) and at the cellular level, relying on the precise regulation of the endogenous circadian clock gene network ( Figure 2). Any abnormalities in these intrinsic rhythms can cause disturbances in the circadian rhythm.
The physiological basis for the generation and maintenance of circadian rhythms comprises the central and peripheral circadian clock systems, rhythm input systems, and rhythm output systems. The rhythm input system senses and transmits environmental synchronisation signals represented by light signals to the central circadian clock system. The central biological clock system acts as the circadian rhythm pacemaker through the output system to transmit the generated rhythm signals to the periphery, and cooperates with the endogenous biological clock system of the peripheral organs to maintain the physiological activity of the body 16 ( Figure 1). The circadian clock system is composed of the central circadian clock and the peripheral circadian clock. In mammals, the apex of this system is the suprachiasmatic nuclei (SCN) master pacemaker, which is considered the central or master clock 17 . The SCN integrates the environmental time information (primarily light) via the retina to revamp or entrain its phase, and then mastermind other oscillators in extra-SCN brain regions and peripheral organs 18,19 . The rhythm output system is regulated by SCN, which can regulate gene expression, cellular function, metabolism, physiology, activity, behaviour, and sleep-wake cycles 20 . Additionally, the rhythm output systems in turn can affect the SCN master pacemaker 19,20 . For example, the arrhythmic food intake, excessive exercises, and sleep/circadian disorders affect SCN by remodel clock-controlled circuit [21][22][23][24] .
Circadian rhythm production and maintenance are regulated by circadian clock genes. The molecular mechanism of the mammalian circadian clock is produced by a cell-autonomous feedback loop [25][26][27] . The periodic oscillation of circadian rhythm depends on the precise regulation of the circadian clock gene and the clock-controlled gene regulatory network, including transcriptional-translational feedback loops 28,29 and the non-transcription mechanism of post-translational modification 30,31 . As shown in Figure 2, the transcriptional-translational feedback loops include a core loop and a secondary stabilisation loop.
In mammals, the transcription factors circadian locomotor output cycles kaput (CLOCK) and brain and muscle ARNT-like 1 (BMAL1) form a heterodimer, which binds to E-box enhancers to activate the target gene transcription of circadian clock gene Period (including Per1 and Per2) and Cryptochrome (including Cry1 and Cry2). When PER and CRY proteins accumulate to a certain extent, they could be further transferred from the cytoplasm to the nucleus, and the PER/CRY heterodimer as a negative regulator directly interacts with CLOCK/BMAL1 to inhibit its transcriptional activity 32 . In the stabilisation loop, the CLOCK/BMAL1 heterodimer can also induce the expression of nuclear receptors REV-ERBa and RORa. As a negative regulator, REV-ERBa can bind to the retinoic acid receptor-related orphan receptor binding element (RRE) Figure 1. The physiological basis for the generation and maintenance of mammalian circadian rhythm. Reproduced from Chen et al. 19 Figure 2. Molecular clock loops and their potential targets with representative small molecule modulators. CLOCK: circadian locomotor output cycles kaput; BMAL1: brain and muscle ARNT-like 1; CRY: cryptochrome; PER: period; ROR: RAR-related orphan receptor; RRE: retinoic acid receptor-related orphan receptor binding element; CCGs: clock-controlled genes; CK1: casein kinase 1; CDKs: cyclin-dependent kinases; GSK3b: glycogen synthase kinase 3b; SIRT1: silent information regulator 1; PPARc: peroxisome proliferator-activated receptor c; DNA TOPs: DNA topoisomerases. Reproduced from He and Chen 49 . Copyright 2016 American Chemical Society.
(sequence AGGTCA) in the BMAL1 promoter region and block the transcription of BMAL1 33 . Conversely, RORa can be used as a positive regulator to bind to the RRE of the BMAL1 promoter region to promote the transcription of BMAL1, thereby forming an auxiliary loop for the transcription and translation oscillations of the circadian clock gene 34 . However, beyond that, post-translational modifications (phosphorylation/dephosphorylation, acetylation/ deacetylation, etc.) and degradation (ubiquitination/proteasome pathway) of various circadian proteins enable fine-tuning of the transcriptional-translational feedback loops (such as adjusting the expression phase and the period of oscillation), so it can also play an important role in the cyclical cycle of circadian rhythms 30,31 . For example, PER and CRY proteins can be phosphorylated by casein kinase 1e (CK1e)/casein kinase 1e (CK1d), which affects the increase in the continuous length of the cycle 35 . Silent information regulator 1 (SIRT1) regulates the expression of the clock gene BMAL1, Cry1, and Per2 by interacting with the CLOCK/BMALI complex and catalysing the deacetylation and degradation of the PER protein 36 .
Circadian alignment in humans has great effect on human health, and circadian misalignment has been involved in metabolic syndrome 37,38 , cardiovascular diseases 39,40 , acute lung injury and inflammation 41 , cancer [42][43][44] , neurological diseases 45,46 , and immune diseases 47,48 . While accumulating evidence indicates that small molecule modulators of circadian rhythms can be used to rectify circadian disorder 18,[49][50][51] , in this review, we pay attention to the recent progress of small molecule modulators targeting core clock proteins (such as CRYs, REV-ERBs, and RORs) and related post-modification targets (such as casein kinase 1 (CK1), cyclin-dependent kinases (CDKs), glycogen synthase kinase 3 (GSK3), cdc2-like kinase 1 (CLK1), breakpoint cluster region-Abelson tyrosine kinase (BCR-ABL), and silent information regulator 1 (SIRT1)), as well as the relationship between circadian rhythm disorders and human health and the therapeutic role and prospect of these small molecules in circadian rhythm related disease.

Overview of modulators targeting circadian rhythms
As mentioned above, circadian rhythms are associated with a variety of biological functions and biological dysfunctions. Efforts to develop initial modulators have focussed on the circadian clock, and modulators including endogenous and synthetic compounds have been discovered. The identified modulators can be classified into two broad categories, which are targeting core clock proteins and other or unknown targets.  (Table 1 and Figure 3), the first-in-class small molecules, comprise carbazole derivative and an activator of cryptochromes (CRYs) 52 . The carbazole derivatives, such as compound 1-3, can specifically interact with CRY1 and CRY2 and cause period lengthening and amplitude reduction in a dose-dependent manner in stable U2OS reporter cell lines harbouring Bmal1-dLuc or Per2-dLuc. Compound 1 can inhibit glucagon-induced gluconeogenesis by stabilising the CRYs. The co-crystal structure of murine CRY2 PHR core domain (1-512) with compound 1 has been reported, and shows that compound 1 can be readily located in the compound 4 (Flavin adenine dinucleotide, FAD)binding pocket of CRY2 53 . Compound 4 was also proven to be an endogenous ligand which can stabilise CRY proteins by competing with F-Box and leucine rich repeat protein 3 (FBXL3), thus lengthening the circadian period 54 . The complex of small molecule and protein is vital in understanding the binding mode and further improving the potency for acting as a modulator against protein. Therefore, the highly active compound 2-(9H-carbazol-9-yl)-N-(2chloro-6-cyanophenyl)acetamide (compound 5) was disclosed under structure-activity relationship analysis and CRY2-compound 1 complex structure 55 . Compound 5 can lengthen the circadian period, repress Per2 activity, and stabilise CRY better than compound 1. More interestingly, another group discovered a series of Compound 9 can enhance E-box-mediated transcription and attenuate the rhythm without affecting the period. Recently, the potent compound 1-(3-(3,6-difluoro-9H-carbazol-9-yl)-2-hydroxypropyl)imidazolidin-2-one (compound 10) significantly enhanced glucose clearance at 100 mg/kg in an oral glucose tolerance test 58 . Furthermore, the compound N-(2-(2,4-dimethylphenyl)-2,6-dihydro-4H-thieno[3,4-c]pyrazol-3-yl)-3,4-dimethylbenzamide (compound 11) as a selective agonist for CRY1 and 1-(4-chlorophenyl)-N-(2-(4-methoxyphenyl)-5,5-dioxido-2,6-dihydro-4H-thieno [3,4-c] pyrazol-3-yl)cyclopentane-1-carboxamide (compound 12) as moderately selective agonist for CRY2 than CRY1 were reported by using human U2OS cells with a Bmal1 promoter-luciferase (Bmal1-dLuc) reporter 59 . The X-ray crystal structures of CRY1 in complex with compound 11 and compound 12 show that these molecules were located in the FAD-binding pocket. As a useful tool for high selectivity against CRY isoform, the compound 11 and compound 12 were proved to facilitate brown adipocyte differentiation. Altogether, the modulators including the agonist or inhibitor of CRYs may be useful tools to treat circadian clockrelated diseases through its action on CRY (see Figure 3 and Table 1).

Modulators for REV-ERBs
Endogenous ligands for REV-ERBs. In 2007, compound 13 was confirmed as a physiological ligand of nuclear receptors REV-ERBa (encoded by nuclear receptor subfamily 1, group D, member 1 (NR1D1)) and REV-ERBb (Nuclear receptor subfamily 1, group D, member 2 (NR1D2)) by two research groups, Rastinejad et al. and Lazar et al 60,61 . Multiple biochemical and biophysical methods were used to demonstrate the association of compound 13 with ligand-binding domains of REV-ERB receptors, including mutation studies, transcriptional repressor function and repression of target gene transcription, ultraviolet-visible spectroscopy, mass spectrometry, isothermal titration calorimetry (ITC), and circular dichroism. Soon afterward, the crystal structure of REV-ERBb in complex with compound 13 was also reported 62,63 . All the results disclosed suggest that compound 13 can bind the REV-ERBs and is indeed a physiological ligand of nuclear receptors REV-ERBs. In mammalian cells, compound 13 can cause the recruitment of the co-repressor nuclear receptor corepressor-1 (NCoR) by targeting REV-ERB, giving rise to the repression of target genes including BMAL1 (also known as ARNTL) 60 . Moreover, by targeting the REV-ERBa, compound 13 can suppress the expression of hepatic gluconeogenic gene and the output of glucose 61 . These findings would facilitate the development of small molecule modulators against REV-ERBs to treat diseases related to the dysfunctional disorder of metabolism and the mammalian clock.
Synthetic ligands for REV-ERBs. In 2008, the compound 1,1-dimethylethyl N-[(4-chlorophenyl)methyl]-N-[(5-nitro-2-thienyl)methyl]glycinate was reported by using REV-ERBa-NCoR fluorescence resonance energy transfer (FRET) assay, which showed an EC 50 value of 250 nM 64 . This compound was the first agonist of REV-ERBa and was competitive with compound 13. In subsequent studies, this compound was successively named SR6452 or GSK4112 (compound 14) ( Table 2 and Figure 4) 65,66 . Compound 14 can induce adipocyte differentiation in 3T3-L1 cells, enhance the recruitment of nuclear receptor co-repressor (NCoR) to REV-ERBa, and inhibit expression of the circadian target gene Bmal1. In addition, similar to compound 13, compound 14 also repressed the expression of gluconeogenic genes in liver cells and reduced glucose output in primary hepatocytes 66 . These studies suggest that compound 14 may be used to treat diabetes or to modulate the circadian rhythm.
Although compound 14 was used as a probe to investigate the pharmacological effects in in vitro, it has a poor pharmacokinetic profile with rapid clearance (C int > 1.0 ml min À1 mg À1 protein) in rat liver microsomes and lower oral bioavailability (F 1% in mice) 66 . Therefore, a series of analogues of compound 14 were synthesised by medicinal chemists to explore the applicable pharmacokinetics and pharmacodynamics used in in vivo studies.
The analogues of compound 14, the potent compounds 15-16, were disclosed by Burris et al. 67 , which were the first REV-ERB agonists with in vivo activity. Compounds 15-16 can generate loss of locomotor activity during the subject dark phase and 1-3h delay in the onset of nocturnal locomotor activity. The two compounds can alter the expression of the core clock genes, including Per2, Bmal1, Clock, Cry2, and Npas2. The ability of REV-ERB agonists in modulating the circadian behaviour of C57BL/6 mice may be used as a drug to treat sleep disorders and jet lag. Indeed, compound 15 was found to be able to induce wakefulness and reduce paradoxical sleep-rapid eye movement (REM) and slowwave sleep [77][78][79] .
As previously reported, the double-knockout REV-ERBa and REV-ERBb mice can also markedly alter metabolic effects 80 . The administration of the agonist of REV-ERBa and REV-ERBb, compound 16, gives rise to increase in energy expenditure and weight loss 67 . In obese mice, including diet-induced obese mice and genetic model of obesity (OB/OB mice), REV-ERB agonist treatment results in a decrease in fat mass and plasma lipids. Recently, a study investigated further the metabolic profile of the nuclear receptor REV-ERB agonist. The results of the experiment show that the enzymatic isoforms mainly involved in the compound 15 phase I biotransformation pathways are cytochrome P450 3A4 (CYP3A4), cytochrome P450 3A5 (CYP3A5), cytochrome P450 2C19 (CYP2C19), and cytochrome P450 2D6 (CYP2D6) 81 .
With the further study of REV-ERB agonist, compound 15 was associated with heart failure 82 , cancer 83,84 , atherosclerosis 85 , chikungunya and O'nyong'nyong virus 86 , and autoimmune disease 87 . However, Lazar et al. 88 discovered that compound 15 can decrease cell viability, rewire cellular metabolism, and alter gene transcription in hepatocytes and embryonic stem cells lacking both REV-ERBa and REV-ERBb, which means that the effects of compound 15 cannot be used solely as surrogate for REV-ERB activity. Therefore, more efforts are needed to explore its mechanism of action. Highly selective compounds also need to be developed urgently.
According to published papers in the same period as compound 15, Kamenecka et al. also conducted structure-activity relationship analysis on compound 14. Compounds 18-19 show slightly better plasma and brain exposure as compound 14, but they displayed the best CNS exposure with brain penetration of 100% or 67%, respectively 69,70 . The analogue of compound 14, compound 17, was reported by Tomkinson et al. 68 Compound 17 shows > 1000-fold selectivity over liver X receptor a (LXRa) and is a potent agonist with REV-ERBa activity (EC 50 ¼ 0.05 lM), which may be the best compound with high selectivity and may serve as a pharmacological toolbox to investigate the biology of REV-ERBa. Recently, the novel small molecular compound 20 was disclosed, which can reinforce REV-ERBa activity by acting in a RORE-dependent manner, though not by the same mechanism as known REV-ERB agonists. It may also provide a new way of exploring the REV-ERB modulator 71 .
The first antagonist of REV-ERBa is compound 21 (Table 2 and Figure 4). Compound 21 is derived from compound 14 based on the tertiary amine scaffold. In HepG2 cells, compound 21 could increase the expression of either glucose 6-phosphatase (G6Pase) or phosphoenolpyruvate carboxykinase (PEPCK) mRNA expression by blocking the action of the endogenous agonist 72 . Compound 21 also caused significant increases in the expression levels of growth/differentiation factor Growth and differentiation factor 10 (GDF10) and Growth and differentiation factor 15 (GDF15) in uterine endometrial stromal cells (UESCs). These results show that cellular oscillators may serve an important role of regulating the expression of downstream genes during the differentiation of UESCs 89 .
Although the pharmacokinetic properties of small molecular compound 21 is poor 72 , which has also been confirmed by our group 90 , compound 21 serves as a useful probe to explore the REV-ERB function by others. In vesicular stomatitis virus (VSV)induced encephalitis model, administration of compound 21 increased C-C motif chemokine ligand 2 (CCL2) mRNA expression and decreased mice survival, which is associated with neuroprotective effects and lifetime 91 . The molecular connection between the circadian timing system and mood regulation was identified by Kim et al. 92 The circadian nuclear receptor REV-ERBa is associated with bipolar disorder, as it influences midbrain dopamine production and mood-related behaviour in mice. Treatment with compound 21 induced mania-like behaviour in association with a central hyperdopaminergic state. The evidence suggests that targeting REV-ERBa may be beneficial to the treatment of circadian rhythm-related affective disorders. Compound 21 could slow the progression of muscular dystrophy by increasing lean mass and muscle function and decreasing muscle fibrosis and muscle protein degradation in C57BL/10ScSn-Dmd mdx /J (mdx) mice 93 . This research suggests that the antagonist compound 21 of REV-ERB may be a profound agent for the treatment of Duchenne muscular dystrophy (DMD). In conclusion, these results suggest that compound 21 is a unique chemical tool. However, it must be clearly recognised that poor pharmacokinetic properties of Recently, a novel oxazole inverse agonist of REV-ERB, compound 24, was discovered by Ray et al. 75 based on fluorescence resonance energy transfer (FRET) assay. Compound 24 showed a high selectivity over 20 nuclear receptors, which can reverse the degradation of REV-ERBa protein mediated by inflammatory stimuli. Subsequently, Gul et al. 76 established a mammalian cell-based two-hybrid assay system and found compound 25 as a novel agonist of REV-ERB. In addition, three other compounds against REV-ERB, compounds 26-28 (Figure 4), were found using this method. Compound 28 was confirmed as an antagonist, and compounds 26-27 were confirmed as agonists. Although the three compounds showed a poor selectivity against other targets, these compounds present a new kind of scaffold and can be used as a profound hit to reveal a drug-like compound.

Modulators for RORs
Natural ligands for RORs. In 2002, the first ligand of RORa, compound 29, was proved by X-ray structure (PDB entry 1N83). It is present in the ligand-binding pocket (LBP) and is important in designing the ligand targeting RORs 94 . The analogue of compound 29, compound 30, can also bind to RORa as confirmed by the crystal structure (PDB entry 1S0X) 95 . Other sterols including oxysterols as ROR inverse agonists and neoruscogenin as RORa agonist were found and reviewed in other papers 96,97 . The representative structure of sterols (compounds 31-37) is presented in Figure 5 to analyse the structure for researchers. In 2001, the first ligand of RORb, compound 38, was proved by X-ray structure (PDB entry 1K4W) 98 . This crystal structure of compound 38 and the ligand-binding domain (LBD) of the rat RORb shed new light on the development of ligands against RORs. Subsequently, the crystal structure of the complex between compound 38 and RORb (PDB entry 1N4H) was solved by Sch€ ule group 99 . They also solved the crystal structure of the complex between synthetic analog compound 40 and RORc (PDB entry 1NQ7). All these two-crystal structures present similar results, namely, the compound 39 and analogs were binding to the RORb ligand-binding domain (LBD). Hydroxycholesterols (compounds 41-43) were binding to the RORc LBD using the same method in 2010, with accession codes 3KYT (RORc/Compound 41), 3L0J (RORc/Compound 42), and 3L0L (RORc/Compound 43), respectively 100 . Recently, the natural compound 44 as an agonist for the ROR was reported by using Clock D19/þ cells with PER2::Luc reporter 101,102 . The potent natural compound and all these crystal structures of the complex between natural ligand and ROR have inspired researchers to search for potent and selective small molecule modulators targeting RORs ( Figure 5).
Synthetic ligands for RORs. In 2010, using cell-based GAL4-NR LBD cotransfection assay, Griffin et al. 103 found the first RORa/c inverse agonist compound 45, which was also the agonist of the liver X receptor (LXR) 104 . Compound 45 was binding to RORa/c but not to RORb. This compound provided the scaffold to further exploit the potent and selective ligands targeting ROR. A compound with multiple targets is not an ideal tool to disclose the function of protein. Therefore, the core scaffold of compound 45 was optimised, and a round of agonists or inverse agonists against RORa/c, RORa, and RORc were reported. These compounds have been reviewed elsewhere 96,97 . The representative compounds can be found in Figure 6 and Table 3 to systematically review the research studies.
The first synthetic RORa-selective partial inverse agonist compound 48 based on the core scaffold of compound 45 103

Small molecule modulators with other or unknown targets
Compounds targeting other proteins including kinase, epigenetic proteins, and others can also alter circadian characteristics. All these compounds are summarised as follows.

Modulators for kinases
Casein kinase 1 (CK1). The casein kinase family comprises seven distinct genes encoding CK1 isoforms (a, a2, c1, c2, c3, d, and e) in mammals 113 . CK1d and CK1e have been discovered to regulate the circadian clock, and their substrates are proved to be PER1, PER2, BMAL1, and CRYs 114 . CK1e-selective inhibitor compound 53 can increase in period length, leading to about 1.2-h in synchronised Rat-1 (mPer1::luc) cells 115 . Afterward, compounds 54-60 (Table 4 and Figure 7) were also proven to lengthen the period in cultured cells and were reviewed in other papers 49,116 . Recently, compound 61 was identified as a regulator to increase period length in mammalian cells and larval zebrafish assay 117 . Compound 62 lengthens the period through CK1 inhibition 118 . All these studies reveal that the role of CK1 is important in the regulation of circadian rhythm 119 .
Other kinases. As other studies in the literature have reported, glycogen synthase kinase 3b (GSK3b) can also regulate the circadian clock, which can phosphorylate CLOCK, PER, REV-ERB, and CRY proteins 114 . The selective GSK3b inhibitors, compounds 68-70, have been reported to shorten the circadian period 123 . Other kinase inhibitors including compounds 71-72 targeting p38 and compound 73 targeting CLK1 have been reported to increase period length 125 . Recently, compounds 74-76 selective BCR-ABL tyrosine kinase inhibitor were found to shorten the circadian period 51 (Table 4 and Figure 7).

Modulators for epigenetic proteins and others
Silent information regulator 1 (SIRT1) has been found to contribute to circadian control, which regulates circadian clock gene expression through PER2 deacetylation 36,127 . SIRT1 activator compound 77 is involved in physiological rhythms and clock gene expression 128 . The potent SIRT1 activators, compounds 78-81, also show that they can reduce circadian expression, lengthen period, and reduce amplitude 128 . Recently, SIRT6 was also found to regulate circadian rhythms via Per2 129 . However, the small molecules of SIRT6 have not been tested by researchers.
In a recent study, peroxisome proliferator-activated receptor c (PPARc) was involved in regulating the expression of Bmal1 and REV-ERBa, and its agonist compound 82 can induce expression of Bmal1 131 . Compounds 83-84, DNA topoisomerase (TOP) inhibitors, were also found to enhance the circadian expression and lengthen the circadian period 132 . Recently, the androgen antagonist and oestrogen activator compound 85 was found to shorten the circadian period 51 (Table 5 and Figure 8).

Implications in circadian rhythm-related diseases
Circadian rhythm plays a very important role in the normal maintenance of organisms, but physical and psychological influences including jet lag, shift work, and diseases can cause a misalignment of the intrinsic oscillators. Jet lag occurs in individuals travelling across multiple time zones, who may suffer from some symptoms including disruption of sleep, gastrointestinal disturbances, decreased vigilance and attention span, a general feeling of malaise, and an increased risk of cancer and heart disease 133,134 . Shift work is apparent among people employed in factories or social event firms and work from 7 pm to 9 am 135 . Shift work has become a common phenomenon in society, and was found to be involved in cancer, cardiovascular disease, depression, and infertility. Jet lag and shift work induce rhythm disorder, which can cause a mass of psychological, nervous system, mental health, and physical health problems 135 . Beyond all that, diseases are closely related to circadian rhythms. Diseases can cause disturbances in circadian rhythms, and disorder in circadian rhythms, in turn, further aggravate the severity of the disease 136,137 . This section will focus on the relationship between disease and circadian rhythm disorders (Figure 9).

Metabolic diseases
Circadian rhythm has been associated with homeostasis and physiology, which is closely related to physical health 138 . Numerous lines of evidence are emerging that circadian dysfunctions are closely associated with increased risk for metabolic disease such as obesity and diabetes 136,139 . Evidence that the circadian rhythm is associated with energy homeostasis, glucose homeostasis, and lipid homeostasis has been found. Homozygous Clock mutant mice can lead to type 2 diabetes mellitus, with metabolic syndromes of hyperleptinemia, hyperlipidaemia, hepatic steatosis, and hyperglycaemia, with insufficient compensatory insulin production. Clock mutant animals can induce obesity, hyperphagia, reduced energy expenditure, adiposity, as well as dysregulation of glucose and lipid metabolism 140 . The core clock genes Clock mut or Bmal1 À/À depress and abolish   gluconeogenesis 141 . Clock mut also induced hypertriglyceridaemia in animal models 140 . REV-ERBa knockout mice also displayed altered lipid and bile metabolism 142 . Subsequent studies have shown that double knockout mice (REV-ERBa and REV-ERBb) have disorganised lipid homeostatic gene networks 80 . The other core circadian rhythm gene ROR also turns out to be related to the regulation of energy homeostasis and several lipid and glucose metabolic genes 143 . Mutant RORa mice (also known as staggerer mice) display hypo-a-lipoproteinemia 144 . Recent studies have shown that RORa accommodates peripheral glucose tolerance, torpor, and hepatic lipid metabolism by regulating the expression of fibroblast growth factor 21 (FGF21) 145,146 . All of these pieces of evidence suggest that the circadian rhythm is associated with metabolism and that clock proteins can be as drug targets to treat metabolic diseases.
Many small molecule modulators of circadian proteins have been found to be useful in metabolic diseases. The CRYs activator compound 1 has been shown to inhibit glucagon-induced gluconeogenesis, which may provide a foundation for the treatment of diabetes 53,147 . Aside from the ligands of CRYs, the ligands of circadian nuclear receptors REV-ERB and RORs also demonstrated that they can be conducive to regulate metabolism in vivo. Compounds 15-16 as agonists of REV-ERBa and REV-ERBb proved highly effective in the improvement of the metabolic profile in obese mice 67 . Recently, Chen et al. 101 identified that compound 44 as an agonist for ROR can potently protect against metabolic syndrome and remodel the circadian and metabolic gene expression in diet-induced obese mice. Subsequently, they demonstrate that compound 44 can serve as a potential drug to treat the metabolic disorders and age-related decline by regulating cholesterol and bile acid metabolism 148 and overcome the metabolic challenge by enhancing mitochondrial respiration in skeletal muscle 149 . Therefore, with an in-depth study of the mechanism for clock proteins and the discovery of selective and potent small molecule modulators, it is believed that in the near future, the ligands of CRYs, REV-ERBs, or RORs will provide first-class treatment for metabolic diseases such as obesity and diabetes.

Sleep disorders
Sleep plays a very important role in the biological process of all creatures; it is regulated by circadian rhythm and homeostatic mechanisms 150 153 . Recent studies indicate that the core clock gene expression has a close association with sleep apnoea (SA). Canales et al. 154 identified that the Per3 expression of SA was lower than that in the normal group. Pharmacological treatment targeting the mammalian clock has been shown to have beneficial effects on sleep architecture 78 . Compound 16 as an agonist of REV-ERBa and REV-ERBb displays increase in wakefulness and reduction of paradoxical sleep-rapid eye movement (REM) sleep and slow-wave sleep in vivo 67,78 . Therefore, the REV-ERB ligands may be beneficial in treating sleep disorders.

Ophthalmic diseases
As widely appreciated, light has profoundly influenced the mammalian circadian rhythm. Light is mainly received by intrinsically photosensitive retinal ganglion cells (ipRGCs) 155 . A large number of studies show that the knockout of the rhythm gene affects retinal processing of light information [156][157][158][159][160] . The circadian rhythm is involved in ophthalmic diseases including glaucoma, macular degeneration, cataract, retinitis pigmentosa, diabetic retinopathy, and optic nerve atrophy. Evidence is accumulating that glaucoma directly damages the light input into the circadian system and causes optic nerve dysfunction 161,162 . Recently, a mass of transcripts of nocturnal rodents and diurnal primates with daily and circadian oscillations were presented by RNA Sequencing (RNA-Seq) technology [163][164][165] . Panda et al. 163 identified that around 4-12% of the transcripts are rhythmic in the cornea, optic nerve head, retina, and retinal pigment epithelium for young male baboons (Papio anubis). Recently, we also disclosed that 3% and 24% of the transcripts are rhythmic in the murine extraorbital lacrimal glands and murine cornea 164,165 . In addition, FitzGerald et al. identified that structural modification of the cornea and the lens was observed in Bmal1 knockout mice 157,158 . Moreover, rhythm disorders can further aggravate diabetic retinopathy in per2 knockout mice 160 . In previous studies, we found that compound 21 as an antagonist of REV-ERBa can enhance corneal wound healing 166 . Therefore, the small molecule modulators of circadian proteins provide a potential solution for the treatment of ophthalmic diseases.

Other diseases
The impact of the circadian system on immune diseases 166,167 , mood disorders 168,169 , neuropsychiatric diseases 18,170 , aging 171,172 , renal diseases (such as hypertension, chronic kidney disease, renal fibrosis, and kidney stones) 173,174 , and cancer 175,176 has been reviewed by others. As described in section 2, small molecule modulators of circadian proteins supply pharmacological tools to treat these diseases. For example, the REV-ERB ligand compound 14 can regulate innate immune responses by repressing interleukin 6 (il6) 92 . Interestingly, Kim et al. identified that the pharmacological inhibition of REV-ERBa activity produces mania-like behaviour. The mice showed more hyperactive behaviour after the administration of REV-ERBa antagonist compound 21 177 . The REV-ERBa agonist may be useful for mood regulation.

Perspectives and concluding remarks
In this review, we detailed all aspects of the physiological basis, molecular clock loops, biological function, potential targets, and small molecule modulators of circadian rhythm. The generation, maintenance, and regulation of circadian rhythms depend on the synergy of the circadian clock system, circadian input system, and circadian output system at the overall level. In particular, the circadian clock system is composed of the central circadian clock and the peripheral circadian clock. The apex of this system is the SCN master pacemaker in mammals. The periodic oscillation of circadian rhythm depends on the precise regulation of the circadian clock gene and the clock-controlled gene regulatory network, including transcriptional-translational feedback loops and the nontranscription mechanism of post-translational modification. Extensive research has been performed on the relationship between circadian clock disorder and disease. Circadian clock genes knockout has confirmed that circadian misalignment is involved in metabolic syndrome, cardiovascular diseases, acute lung injury and inflammation, neurological diseases, immune diseases, cancer, mood disorders, sleep disorders, and ophthalmic diseases. As summarised in this article, circadian rhythms are important for human health, which suggests that the development of small molecules is imminent and could be used to treat circadian rhythm related diseases.
More importantly, a large number of small molecule modulators of circadian rhythm have been discovered, and most modulators have potential therapeutic effects on disease. In order to identify hits of the circadian clock, hundreds of thousands of compounds have been filtered by cell-based high-throughput circadian assays. The effectiveness of chemical biology approaches contributed to the discovery of the small molecule modulators of circadian rhythm 178 . In recent years, with the emergence and popularisation of some new technologies, biophysical methods (such as differential scanning fluorimetry, differential scanning calorimetry, isothermal titration calorimetry, and surface plasmon resonance) and computer-aided drug design will help in the discovery of more modulators targeting clock proteins. It is believed that in the near future, small molecule modulators will be a useful tool in the treatment of circadian rhythm related diseases.

Author contributions
All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

Disclosure statement
No potential conflict of interest was reported by the author(s).