Circadian gene variants in cancer

Abstract

Humans as diurnal beings are active during the day and rest at night. This daily oscillation of behavior and physiology is driven by an endogenous circadian clock not environmental cues. In modern societies, changes in lifestyle have led to a frequent disruption of the endogenous circadian homeostasis leading to increased risk of various diseases including cancer. The clock is operated by the feedback loops of circadian genes and controls daily physiology by coupling cell proliferation and metabolism, DNA damage repair, and apoptosis in peripheral tissues with physical activity, energy homeostasis, immune and neuroendocrine functions at the organismal level. Recent studies have revealed that defects in circadian genes due to targeted gene ablation in animal models or single nucleotide polymorphism, deletion, deregulation and/or epigenetic silencing in humans are closely associated with increased risk of cancer. In addition, disruption of circadian rhythm can disrupt the molecular clock in peripheral tissues in the absence of circadian gene mutations. Circadian disruption has recently been recognized as an independent cancer risk factor. Further study of the mechanism of clock-controlled tumor suppression will have a significant impact on human health by improving the efficiencies of cancer prevention and treatment.

Introduction

The evolutionary adaptation dictates that most, if not all, physiological processes in mammals follow a circadian rhythm. Circadian rhythm is generated by an endogenous circadian clock and coupled to the diurnal oscillation of environmental cues. Changes in lifestyles due to industrialization and economic globalization in the modern world has led to frequent disruption of endogenous circadian homeostasis, which is coupled with a dramatic increase in the modern-day diseases including psychological disorders, cardiovascular diseases, metabolic syndromes, aging, immune deficiencies, reproductive and neuroendocrine dysfunction, as well as cancer (1–4).

The circadian control of mammalian physiology

The mammalian circadian clock is composed of a central clock in the hypothalamic suprachiasmatic nucleus (SCN), peripheral oscillators in all peripheral tissues studied, and circadian input and output pathways. The SCN consists of multiple, self-sustained single-cell circadian oscillators that constantly synchronize to environmental cues to generate coordinated circadian outputs. The most potent environmental circadian cue is the ambient light received by a subset of melanopsin-expressing retinal ganglion neurons and transmitted directly to the SCN via the retinohypothalamic tract (RHT) (5). The SCN clock targets other hypothalamic centers to generate the circadian rhythm in the neuroendocrine and autonomic systems (NES and ANS) via direct neuronal connections and secreting diffusible molecules. The NES and ANS are circadian output pathways that generate a coupled circadian rhythm in diverse peripheral tissues by controlling cell signaling in a tissue-specific manner (Figure 1A) (5–7).

Figure 1.

The mammalian circadian clock. (A) The mammalian circadian clock is comprised of circadian input pathways, such as the light input pathway via the retinohypothalamic tract (RHT), the central clock located in the hypothalamic suprachiasmatic nucleus (SCN), the circadian output pathways including the neuroendocrine and autonomic nervous systems (grey arrows) and peripheral clocks in all tissues studied. (B) A simplified model of the molecular clock. Solid lines show direct regulation of the positive and negative feedback loops by core circadian genes Bmal1, Clock, Ck1δ and ε, Cry, Naps2, Per, Rev-erbα and β, as well as Rorα and γ. The dashed lines show indirect regulation of Rev-erbα and β by PER. The molecular clock also targets clock-controlled genes (CCGs) that regulate diverse cellular processes. The first-order CCGs are controlled by the molecular clock directly at the transcriptional level.

The molecular clock is operated by interacting feedback loops of core circadian genes in all cells in the body. At the transcriptional level, the clock is driven by heterodimers of bHLHPAS transcription factors BMAL1/CLOCK or BMAL1/NPAS2 that activate core circadian genes Cryptochrome (Cry1,2) and Period (Per1–3) via E-boxes in gene promoters at the beginning of a subjective day. The PER and CRY proteins then form a transcriptional repressor complex that enters the nucleus at the beginning of a subjective night to inhibit the heterodimer activity by protein–protein interactions and/or recruitment of transcriptional termination complexes. Bmal1 is also rhythmically regulated by its own transcriptional targets Rorα, Rev-erbα and β encoding nuclear receptors RORα, REV-ERBα and β, respectively. Upon activation, RORα stimulates Bmal1 expression, while REV-ERBα and β suppress Bmal1 transcription (8,9). The molecular feedback loops are also controlled by post-translational mechanisms. The stability of PER and CRY, controlled by casein kinase 1ε and δ (CK1ε/δ) and the Skp1-cullin-F-box protein (SCF) E3 ubiquitin ligase complexes, respectively, determines the time of the PER/CRY repressor nuclear entry (10,11). The cell-autonomous oscillation of multiple interlocked feedback loops of circadian genes defines the intrinsic circadian rhythmicity of the molecular clock (Figure 1B) (8,9).

The clock targets clock-controlled genes (CCGs) to control diverse cellular processes in peripheral tissues. System-level approaches have identified a large number of first-order CCGs controlled by the clock at the transcriptional level. The majority of CCGs encode tissue-specific expressed mRNAs to control key tissue functions. A small group of ubiquitously expressed CCGs encode proteins supporting basic cellular functions (12,13). The rhythmic expression of these CCGs is controlled by mechanisms including direct transcriptional regulation by heterodimers via E-box sequences in gene promoters, indirect regulation by clock-controlled gene-specific transcriptional regulators, and circadian oscillation in chromatin remodeling (9,12,14).

The molecular clock constantly responds to daily entrainment signals to maintain the synchrony with the environment. In SCN neurons, light stimuli phase-shifts the molecular clock via activating immediate early responsive genes such as Ap1, Per1, and Per2 in a time-dependent manner via signal transduction pathways including the calcium/calmodulin-dependent protein kinases II (CaM kinases II), c-Jun N-terminal kinase (JNK), c-AMP-protein kinase A (PKA), extracellular signal-regulated kinases (ERK), mitogen-activated protein kinases (MAPK), nitric oxide (NO)/c-GMP, or protein kinase C alpha (PKCα) (15,16). In peripheral tissues, the circadian output pathways generate cyclic changes in the levels of neurotransmitters, growth factors, cytokines, and blood-borne hormones in the tissue microenvironment, which rhythmically entrain the molecular clock via intracellular signaling controlled by pathways including those mediating the light response in SCN neurons (4,7,17).

The homeostasis of internal physiology is maintained by coordinated activities of the central and peripheral clocks. Disruption of external light cues phase-shifts the SCN clock leading to a phase-shift in circadian output pathways, which then phase-shifts peripheral clocks via phase-shifting intracellular signaling in a tissue-specific manner. The consecutive phase-shifts in the hierarchical circadian timing system temporarily disrupts the homeostasis of physiology due to various rates of phase-shifts of peripheral tissues resulting from their differential innervation by circadian output pathways. The time needed for re-establishing internal circadian homeostasis is determined by when the circadian disruption occurs during a day and how many hours of phase-shift in the SCN clock it initially induces. Therefore, human rotating working schedules or frequent rapid long-distance transmeridian travel leads to chronic misalignment of internal physiology from environmental cues, which has been shown by recent studies to increase the risk of cancer significantly (1,2,4,18).

The mechanism of clock-controlled tumor suppression

The clock plays a key role in controlling tissue-specific functions. Tissues normally supported by cellular processes frequently deregulated in cancers are most likely sensitive to circadian dysfunction-induced tumor development. Recent studies have revealed that circadian disruption specially increases the risk of cancers in the immune, skeletal, digestive and reproductive organs that need cell proliferation, metabolism, and DNA damage repair to maintain daily function and are prone to cell death, senescence, and inflammatory response induced by adverse physiological conditions (1–4,19).

Circadian control of cell proliferation

Cell cycle progression follows a circadian rhythm in all rapidly renewing mammalian tissues studied, but is arrhythmic or displays ultradian rhythms in metastatic cancers (4, 20). The ubiquitously expressed CCGs include key cell cycle and proto-oncogenes as well as tumor suppressors (12,13). The existing evidence suggests that the molecular clock likely suppresses proto-oncogenes but stimulates tumor suppressors at transcriptional level. For example, the binding of BMAL1/CLOCK and BMAL1/NPAS2 heterodimers at E-boxes in gene promoters negatively regulates proto-oncogene c-Myc but stimulates the tumor suppressor Wee1 in response to extracellular mitogenic signals (19, 21–23). Disruption of the molecular clock leads to deregulation of these CCGs, which is coupled with increased risk of neoplastic growth and cancer in various circadian gene mutant mouse models (19, 22–27).

At the post-translational level, the molecular clock may modulate both positive and negative loops of the cell cycle clock. For example, PER1 and CRY2 were reported to regulate p53-controlled checkpoint function by interacting with the tumor suppressor ataxia-telangiectasia mutated (ATM) and Rad3-related (ATR), respectively (28,29), whereas CK1ε promotes β-catenin-mediated activation of T-cell-specific transcription factor/lymphoid enhancer factor-1 (TCF/LEF) family that stimulates c-Myc and Cyclin D1-mediated cell cycle progression (30). Thus, the molecular clock acts to maintain the homeostasis but not inhibition of cell proliferation at the cellular levels (4).

The molecular clock alone is unable to generate the circadian rhythm of cell proliferation because G1 cell cycle progression is controlled by extracellular mitogenic signals (31). In vivo, these signals include neurotransmitters, steroid and peptide hormones, chemokines, growth factors, or cytokines either directly released from the circadian output pathways or rhythmically produced by peripheral tissue via paracrine and autocrine signaling. The tissue-specific interaction of these signals with their targets, such as G-protein coupled receptors, tyrosine kinase receptors, integrins, and nuclear receptors, simultaneously entrains the molecular and cell cycle clocks by activation of early responsive genes (4).

One pathway for the central clock to control cell proliferation in peripheral tissues is via the sympathetic nervous system (SNS). The SNS innervates all peripheral organs to control diverse physiological functions (32). The direct neuronal targeting of hypothalamic paraventricular pre-sympathetic neurons by the SCN clock generates a robust circadian variation in the basal level of the SNS tone, which then drives the circadian rhythm of diverse cellular processes in peripheral tissues by controlling peripheral clocks and intracellular signaling (6). The SNS signaling is deregulated or arrhythmic in human night-shift workers and jet-lagged mouse models (19,33). Surgical ablation of SNS innervation abolishes circadian oscillation of immune function via inhibition of β-adrenergic receptor (ADRβ)-mediated hematopoietic stem cell proliferation, differentiation, and trafficking, which is a key mechanism of immune suppression that promotes cancer development (4). Uncontrolled SNS signaling promotes tumor initiation by stimulating cell proliferation, while sympathectomy inhibits tumor growth. Thus, β-blockers are suggested as novel anti-cancer drugs for future anticancer therapies (34, 35). The mechanism of SNS-mediated cell cycle progression can be explained by its ability to activate Ap1, Per1 and 2, and ATM (19). AP1 activation leads to Myc-dependent G1 cell cycle progression. Per induction stimulates BMAL1/CLOCK-mediated Myc transcriptional repression, whereas the peripheral clock-dependent induction of ATM leads to p53 activation. Thus, the rhythmic SNS signaling is a circadian cue for both cell cycle and peripheral clocks in vivo. Loss of function in the peripheral clock abolishes the activation of BMAL1/CLOCK and p53 in response to SNS signaling, but has no effect on AP1-controlled Myc activation, leading to Myc oncogenic activation, uncontrolled cell proliferation, and neoplastic growth in mice (19, 23). Together, these findings described above explain why disruption of the circadian homeostasis increases the risk of cancer in humans and rodent models (19, 36–58) and highlight the role of circadian disruption as one of the independent risk factors for the dramatic increase in the rate of sporadic cancers in the modern societies.

Circadian control of DNA damage response and cellular senescence

The activities of cell proliferation and metabolism generate a large amount of lesions in genomic DNA each day, which activates DNA repair machinery and cell cycle checkpoints to repair damaged DNA or induces cell death or cellular senescence when the damage exceeds the capacity of repair (32, 59). In mammals, ATM and ATR are two master regulators of DNA damage. The clock not only couples ATM activation with daily cell cycle progression in peripheral tissues but also activates ATM and ATR in response to acute DNA damage (19,22,28), leading to protein kinases CHK1/2 and p53-mediated cell cycle arrest and/or p53-mediated apoptosis (60). The clock also controls the expression of key DNA replication, recombination, and repair enzymes (12, 13). Loss of function in the molecular clock deregulates these CCGs leading to accumulation of DNA damage and increased risk of neoplastic growth in mice. For example, the gene encoding xeroderma pigmentosum A (XPA), an essential enzyme for nucleotide excision repair, is deregulated in mice lacking both Cry1 and 2, which is coupled with a dampened circadian rhythm in nucleotide excision repair after UVB irradiation in epidermis (61, 62). Keratinocyte-specific deletion of Bmal1 in mice results in deregulation of cell proliferation, DNA repair, and oxidative phosphorylation genes, leading to increased intracellular reactive oxygen species (ROS), elevated and arrhythmic cell proliferation, dampened UVB-induced DNA damage response, and accumulation of DNA lesions in epidermis (25). The circadian regulators may also directly participate in DNA damage response. After γ-radiation, CLOCK translocates to the sites of DNA double-strand breaks (63). PER1 was reported to interact directly with ATM and CHK2, and BMAL1 is required to activate the p53-p21^WAF1/CIP1 pathway (28,64), whereas CRY2 is involved in intra-S checkpoint activation after UV radiation by interacting with ATR and CHK1 via TIMELESS (29). In mice, the activation of all circadian genes studied is required for a time-dependent γ-radiation response in thymus. Loss of function in Per2 abolishes the response of all circadian genes to γ-radiation, leading to radiation resistance and increased tumor incidence (19, 22).

The deregulation of DNA damage response and accumulation of DNA lesions are associated with premature aging phenotypes frequently observed among circadian gene mutant mice. Bmal1^−/− mice display aggressive aging phenotypes leading to a significantly reduced lifespan (65). Mice carrying a mutated Clock or lacking Per or Cry also display premature aging phenotypes that are more evident after being exposed to γ-radiation (19, 22, 66, 67).

Aging in various tissues shares a common mechanism of replicative cellular senescence, which refers to a state of permanent withdrawal from the cell cycle due to accumulation of genetic alterations beyond the capacity of repair (68, 69). Since metastasizing tumors show unlimited capacity of cell proliferation, cellular senescence has been widely considered as a mechanism of tumor suppression (70). However, recent studies have revealed that senescent cells are still metabolically active but show deregulation of chromatin organization and gene expression, and increased secretion of proinflammatory cytokines, proteases, and growth factors that stimulate tumor growth and metastasizing (68). In addition, senescent cells can regain the ability of proliferation in response to changes in internal physiology (71). The uncontrolled Myc and Ras/MAPK oncogenic signaling, activation of p53/p21^WAF1/CIP1 and pRB/p 16^INK4a tumor suppressors, and loss of function in Sirt1, which encodes the NAD-dependent deacetylase sirtuin-1, a class III histone deacetylase, are established molecular mechanisms promoting cellular senescence (68, 71–73). Especially, SIRT1 may bridge aging and cancer prone phenotypes found in circadian gene mutant mice by directly deacetylating BMAL1, PER, p53, β-catenin, and DNA repair protein KU70 (11, 72–78). In mice, the early onset of cellular senescence is, at least in part, due to AKT-dependent vascular senescence in Per2 mutants, and via a p53-independent mechanism in Bmal1 nulls (79,80). Thus, cancer and aging share a common mechanism of DNA damage accumulation in vivo.

Circadian control of metabolic homeostasis and inflammatory response

Among various DNA damaging agents, ROS can induce oxidative DNA damage including double-strand DNA breaks to promote genomic instability (81). The exogenous sources of ROS include pollutants, tobacco, xenobiotics, and radiation. The endogenous ROS are natural by-products of cell metabolism mainly produced by NADPH oxidase complexes located in cell membranes, mitochondria, peroxisomes, and endoplasmic reticulum (82). ROS normally function in various signaling pathways to maintain the homeostasis of cellular function. Their intracellular levels are tightly controlled by multiple mechanisms (83). However, under pathological conditions, such as viral infection, chronic inflammation, and metabolic dysfunction, intracellular ROS levels can increase dramatically, leading to significant oxidative damages to cells including double-strand DNA breaks (84, 85).

The clock is a master regulator of cell metabolism. It controls the expression and activities of rate-limiting enzymes for ROS production and cellular antioxidative response including the NADPH oxidase complexes that produce ROS and superoxide dismutases, glutathione peroxidases, and peroxiredoxins that mediate antioxidative response (86–88). Loss of function in the molecular clock has been shown to deregulate cell metabolism leading to excess accumulation of ROS in vivo (25). Circadian dysfunction also promotes metabolic adaptations in favor of cancer progression. One example of such adaptations is the Warburg effect (89), which describes that, in contrast to normal somatic cells metabolizing glucose to CO₂ and H₂O via a low rate of glycolysis followed by oxidative phosphorylation, cancer cells predominantly use glucose to produce energy by a high rate of anaerobic glycolysis in the cytosol even in the presence of oxygen, leading to a dramatic increase in intracellular ROS and synthesis of purines, pyrimidines, non-essential amino-acids, and free fatty acids essential for cell growth and division. Recent findings suggest that the Warburg effect is driven by oncogene activation and, therefore, is considered as the result rather than the cause of neoplastic growth (85). In vivo, the clock not only controls the expression of proto-oncogenes and tumor suppressors but also the key metabolic genes involved in the Warburg effect, such as glucose-6-phosphatase, pyruvate kinase, and glucose transporter 2 (GLUT) (86,90). Thus, disruption of circadian homeostasis may promote the Warburg effect in the absence of any gene mutations by promoting oncogenic activation and metabolic dysfunction.

Recent studies have established a strong correlation between circadian disruption and the development of metabolic syndromes (1, 91). The mechanisms linking metabolic syndromes with cancer include deregulation of extracellular signaling, induction of angiogenesis, and chronic inflammation, a key event promoting cancer (92, 93). Chronic inflammation can be induced by excess ROS that activates NF-κB-mediated proinflammatory and pro-proliferation pathways (94, 95). Together, these events lead to tissue damage and cell death as well as the following tissue regeneration supported by active cell proliferation. However, the intrinsic deficiencies in cell cycle control, metabolic balance, cancer immuno-editing (reviewed in (4)), and the accumulation of DNA damage due to chronic circadian disruption would significantly increase the risk of neoplastic growth and cancer development.

The circadian genes in cancer

Recent studies have revealed that genes operating both positive and negative loops of the molecular clock are important for tumor suppression in vivo, and that the mechanism of clock-controlled tumor suppression is conserved among humans and rodents (Tables I and II).

Table I.

Circadian genes in human diseases.

Clock genes	Cancer types	Other diseases	References
Bmal1	Breast, colorectal, head and neck, ovarian, pancreatic, and prostate cancers, B-cell lymphoma, pleural mesothelioma, ALL, AML, and CML	Aging, hypertension, immune deficiency, mood disorders, neurodegeneration, and T2D	(45, 96–108)
CK1δ/ε	Breast, colorectal, ovarian, pancreatic, and prostate cancers	Aging, chronic inflammation, hypertension, insulin resistance, neurodegeneration, and sleep disorders	(102, 120–135)
Clock	Breast, colorectal, lung, ovarian, pancreatic, prostate and skin cancers, B-cell lymphoma, and CLL	Bipolar disorder, hypertension, obesity, neurodegeneration, NAFLD, and T2D	(50, 102–104, 106, 139–144)
Cry1/2	Breast, colorectal, endometrial, head and neck, ovarian, pancreatic, prostate and skin cancers, glioma, pleural mesothelioma, CLL, CML, HCC, and NHL	Acute inflammation, hyperglycemia, mood disorders, obesity, Parkinson’s, and T2D	(102–105, 125, 126, 134, 139–140, 153–171)
Dec1/2	Breast, endometrial, gastric, head and neck, lung, pancreatic, and renal cancers, and HCC	Autoimmune disease, chronic inflammation, and sleep disorders	(182–194)
Npas2	Breast and prostate cancers, pleural mesothelioma, glioma, and NHL	Aging, chronic fatigue, chronic inflammation, hypertension, mood disorders, and neurodegeneration	(100, 102, 134, 161, 169, 197–204)
Per1/2/3	Breast, colorectal, endometrial, head and neck, lung, ovarian, pancreatic, prostate, skin and thyroid cancers, B-cell lymphoma, glioma, pleural mesothelioma, AML, CLL, CML, and HCC	Aging, cardiovascular disease, chronic inflammation, insulin resistance, neurodegeneration, obesity, eating, mood and sleep disorders, and T2D	(98, 99, 104, 105, 126, 134, 139, 140, 157, 158, 160, 161, 164, 169, 170, 213, 223, 229, 240)
Rorα/γ	Breast, colorectal, pituitary, prostate, and thyroid cancers	Immune deficiency, insulin resistance, and obesity	(122, 255–264)
Rev-erbα/β	Breast and thyroid cancers and pleural mesothelioma	Autoimmune disease, chronic inflammation, and obesity	(162, 256, 273–280)
βTrCP1/2	Breast, colorectal, gastric, pancreatic, and prostate cancers, and HCC	N/A	(284–289)

ALL = acute lymphocytic leukemia; AML = acute myeloid leukemia; CLL = chronic lymphocytic leukemia; CML = chronic myeloid leukemia; HCC = hepatocellular carcinoma; NAFLD = non-alcoholic fatty liver disease; NHL = non-Hodgkin’s lymphoma; T2D = type 2 diabetes.

Table II.

Circadian gene mutant mouse models.

Genetic model	Mouse strain	Mouse phenotype	References
Bmal1+/−	C57BL/6J	Premature aging, cancer prone	(19, 65)
Bmal1−/−	C57BL/6J	Arrhythmic, aggressive aging, hypoglycemia, redox deregulation, immune deficiency cellular senescence, cancer prone, metabolic syndromes, neurological disorders	(19, 23, 65, 111–113, 115, 116)
Bmal1−/− islets or adipocytes	C57BL/6J × 129 × ICR	Hyperglycemia, hypoinsulinemia, adipocyte hyperplasia, redox deregulation	(115, 116)
Bmal1−/− keratinocytes	C57BL/6J	Genomic instability, cellular senescence, deregulation of cell proliferation and metabolism, redox deregulation	(25, 65)
CK1δ or CK1ε	C57/BL6J	Lengthened period, autoimmune diseases, mammary gland neoplasms	(136–138)
ClockΔ19/Δ19	C57BL/6J & C57BL6 × BALB/c	Arrhythmic in constant darkness, metabolic syndromes, obesity, premature aging, NAFLD	(21, 66, 116, 145–152)
Clock−/−	C57BL/6J	Aging, chronic inflammation	(150)
Clock−/− cardiomyocytes	FVB/NJ	Cardiac dysfunction, fatty acid dysregulation	(149)
Cry1−/− or Cry2−/−	C57BL/6J	Abnormal behavioral rhythm, premature aging, hypertension, chronic inflammation, sleep disorder, cancer prone, metabolic syndromes	(21, 61, 173–175)
Cry1−/−; Cry2−/−	C57BL/6J	Arrhythmic, premature aging, chronic inflammation, hypertension, impaired tissue regeneration, sleep disorder, cancer prone, metabolic syndromes	(19, 23, 62, 67, 175–178)
Dec1−/−; Dec2−/−	C57BL/6J	Lengthened circadian period	(195, 196)
Npas2−/−	C57BL/6J × 129Sv	Metabolic syndromes, behavioral abnormalities, sleep disorder	(205–208)
Per1−/−	C57BL/6J	Immune deficiencies, cancer prone	(242, 243, 252)
Per2−/− or Per2 tm1Brd/tm1Brd	C57BL/6J & C57BL/6J × 129Sv	Arrhythmic inconstant darkness, premature aging, cancer prone, metabolic syndromes, vascular diseases, immune deficiency	(19, 22, 23, 173, 241, 250, 253)
Per2S662G or Per2S662D	C57BL/6J	Shortened period, premature aging, deregulation of apoptosis, cancer prone	(24, 27)
Per1−/−; Per2 tm1Brd/tm1Brd	C57BL/6-Tyrc-2J	Arrhythmic, premature aging and neurological disorders, cancer prone, metabolic syndromes, neurological disorder	(19, 242, 243, 251)
Per3−/−	C57BL/6 × 129Sv	Metabolic syndromes, arteriosclerotic disease	(249, 250)
Rorα −/− or Rorγ −/−	C57/BL6J	Premature aging, adipocyte hyperplasia, immune-deficient, neurodegeneration	(265–271)
Rev-Erbα−/−	C57BL/6J × 129Sv	Metabolic syndromes, immune deficiencies	(272, 281, 282)
Rev-Erbα−/−; Rev-erbβ −/−	C57BL/6J × 129Sv	Arrhythmic, metabolic syndromes	(272)
βTrCP1m/m	C57BL/6J	Tissue-specific expression of dominant negative mutants promotes tumor development	(298–300, 302)

Brain and muscle aryl hydrocarbon receptor nuclear translocator-like 1 (Bmal1)

Bmal1 variants due to single nucleotide polymorphisms (SNPs) in humans are associated with various diseases, such as type 2 diabetes, hypertension, mood and sleep disorders, aging, neurodegeneration, and immune deficiencies (96–101), as well as cancers including breast, colorectal, prostate, pancreatic, and ovarian cancers, head and neck squamous cell carcinoma, B-cell lymphoma, pleural mesothelioma, acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), and chronic myeloid leukemia (CML) (45, 102–108). Suppression of Bmal1 expression in human prostate, lung, or glioma cancer cells significantly increases their metastasizing potential (109). The role for Bmal1 in tumor suppression include but are not limited to the suppression of uncontrolled PI3K-AKT-MMP2 signaling essential for tumor invasion and the activation of p53-mediated apoptosis and cell cycle arrest (64, 110). Ablation of Bmal1 in mice abolishes circadian behavioral homeostasis even under entrained conditions (111), which is coupled with phenotypes of aggressive aging (65, 112), cognitive deficits (113), chronic inflammation (65), cancer (19), and deregulated response to anti-cancer drugs such as docetaxel, etoposide, oxaliplatin, and cyclophosphamide (109, 114). Tissue-specific ablation of Bmal1 increases the risk of abnormal adipogenesis (115), insulin resistance (116), toxic ROS accumulation, genomic instability, senescence, and uncontrolled cell proliferation in epidermis (25).

Casein kinase 1δ and ε (Ck1δ and ε)

CK1ε plays a critical role in MYC-driven cancers in humans (117). It also regulates gluconeogenic genes in response to PGC-1α activation and modulates lipoprotein metabolism by phosphorylating PGC-1β (118, 119). Mutations and/or deregulation of Ck1δ and ε are associated with colorectal, pancreatic, prostate, breast, and ovarian cancers (102, 120–126), neurodegeneration and sleep disorders (127–131), chronic inflammation and aging (132, 133), as well as metabolic syndromes (134, 135). Disruption or deregulation of Ck1δ or ε in mice deregulates circadian homeostasis (136) and increases the risk of mammary carcinogenesis (137) and autoimmune diseases (138).

The circadian locomotor output cycles kaput (Clock)

The SNPs in the human Clock gene are associated with increased susceptibility to breast, colorectal, lung, ovarian, skin, pancreatic, and prostate cancers, B-cell lymphoma, and chronic lymphocytic leukemia (CLL) (50, 102–104, 106, 139–141), as well as bipolar disorder, neurodegeneration, and metabolic syndromes including obesity and obesity-related non-alcoholic fatty liver disease (NAFLD) (142–144). Mice homozygous in dominant negative Clock mutation (^{ClockΔ19/Δ19}) are unable to maintain circadian behavior rhythm after prolonged exposure to constant darkness (145) and show increased risk of cardiac and immune malfunctions, premature aging, deregulation of DNA damage response, and various metabolic syndromes including obesity, adipocyte hypertrophy, NAFLD, diabetes, hypercholesterolemia, and hypertriglyceridemia (66, 116, 146–150). In peripheral tissues, ^{ClockΔ19/Δ19} mice show deregulation of genes controlling glycolysis, mitochondrial oxidative phosphorylation, lipid metabolism, inflammatory response, fatty acid oxidation, chromatin remodeling, redox synthesis, cellular senescence DNA damage response, and tumor suppression (74, 86, 148, 151, 152).

Cryptochrome (Cry)

The SNPs or deregulation of Cry1 and/or 2 are associated with increased susceptibility and mortality to breast, colorectal, endometrial, prostate, skin, thyroid, and prostate cancers, hepatocellular carcinoma (HCC), pancreatic ductal adenocarcinoma, head and neck squamous cell carcinoma, glioma, CLL, CML pleural mesothelioma, and non-Hodgkin’s lymphoma (NHL) (102–105, 125, 126, 139–140, 153–161), Parkinson’s disease (162), mood disorders (163), acute inflammation (164), and metabolic syndromes including obesity and type 2 diabetes (134, 165–171). Inhibition of Cry2 expression in MCF-7 cells leads to dysregulation of genes important for proliferation, apoptosis, angiogenesis, inflammation, and tumor migration and invasion (172). Cry mutant mice are deficient in tissue regeneration (21) and also show increased risk of neoplastic growth (23, 173), chronic inflammation, high-fat diet-induced metabolic syndromes (174–176), premature aging (19, 67), deregulation of DNA damage repair (61, 62), hypertension (177), and sleep disorders (178). Although previously reported as lack of evidence of increased cancer risk (67), a recent study shows that mice lacking Cry1 and/or 2 also display increased risk of spontaneous and radiation-induced tumor development (19).

Deleted in esophageal cancer 1 and 2 (Dec1 and 2)

Dec1 and 2 are activated by BMAL1/CLOCK and encode bHLH transcription factors that suppress Per and Cry transcription (179–181). In humans, SNPs, deletion and/or deregulation of Dec1 or 2 are associated with gastric, non-small-cell lung, pancreatic, endometrial, renal, and breast cancers, HCC, head and neck esophageal squamous cell carcinoma and lymph node metastasis (182–191), sleep disorders (192), and inflammation and autoimmune disease (193, 194). Mice lacking Dec1 or 2 (Dec1^−/− or Dec2 ^−/−) maintain normal circadian rhythmicity. However, mice lacking both Dec1 and 2 (Dec1^−/−; Dec2^−/−) have a lengthened circadian period (195, 196).

Neuronal PAS domain protein 2 (Npas2)

SNPs and deregulation of Npas2 in humans increase the risk of metabolic syndromes, aging, neurodegeneration, chronic fatigue syndrome, mood and sleep disorders, chronic inflammation (100, 134, 169, 197, 198), and breast and prostate cancers, pleural mesothelioma, glioma, and possibly NHL (102, 161, 199–204). The role of Npas2 in tumor suppression has not been vigorously studied using mouse models. However, Npas2 mice display a variety of pathological changes that may directly or indirectly promote tumor development, such as impaired responses to food entraining signals, metabolic syndromes (205, 206), and lack of sleep homeostasis and behavioral adaptability (207, 208). At the cellular level, Npas2 has been found to control redox levels, oxidative and DNA-damage response, and the expression of cell cycle genes and tumor suppressors (209–212).

Period (Per)

The SNPs and deregulation of Per1, 2, and/or 3 in humans are associated with increased risk of thyroid, breast, prostate, ovarian, endometrial, pancreatic, colorectal, skin, and non-small-cell lung cancers (NSCLC), HCC, colorectal carcinoma, diffusible large B-cell lymphomas, malignant pleural mesothelioma, head and neck squamous cell carcinoma, glioma, CML, AML, and CLL (104, 105, 126, 139, 140, 157, 158, 160, 161, 213–223), which often show deregulation of inflammatory cytokines, oncogenic and tumor suppressors including p38, ERα, G1 and S-phase cyclins, c-Myc, NF-κB, Bcl-XL, PKA, telomerase, ATM, p53, p21, and Wee1 (28, 106, 175, 219, 224, 228). The SNPs and deregulation of human Per genes are also linked to obesity, metabolic syndromes, type 2 diabetes, eating, mood and sleep disorders, cardiovascular diseases, aging, depression, chronic inflammation, and neurodegeneration (98, 99, 134, 164, 169, 170, 197, 229–240).

Various research teams have independently demonstrated that mouse models deficient in Per1, Per2, or both Per1 and 2 show increased risk of spontaneous and radiation-induced tumor development in immune, digestive, skeletal, and reproductive organs (19, 22, 24, 26, 27), neoplastic growth in bone (23, 173), premature aging and vascular senescence (19, 80, 241, 242), mood disorders (243), immune deficiencies (244, 255), and metabolic syndromes including diabetes, liver cholestatic diseases, hypoglycemia, and hyperinsulinemia (246–251). Per mutant mice show deregulation of key genes controlling cell proliferation, metabolism, senescence, inflammatory response, and death such as c-Myc, p53, Gadd45α, Atm, Ap1, Cyclin D1, Cyclin A, β-catenin, APC, PPARγ, Mdm2, Akt, Cyp7A1, interferon gamma (IFNγ), and nuclear receptors Fxr and Car (22, 23, 26, 27, 244, 247, 251–254).

Retinoic acid-related orphan receptor (Rorα and γ)

The deregulation and/or SNPs of Rorα and/or γ in humans have been linked to an increased risk of breast and prostate cancers, colorectal adenocarcinomas, pituitary and thyroid tumors (122, 255–260), immune deficiency (261), obesity, insulin resistance, and adipogenesis (262–264). The cancer risk of mice lacking Rorα or γ (Rorα^−/− or Rorγ^−/−) has not been carefully studied. However, these mice are immune-deficient (265, 266) and show adipocyte hyperplasia (267–269). Deregulation of Rorα, and/or γ is also associated with accelerated aging and neurodegeneration in mice (270, 271).

Reverse viral erythroblastosis oncogene products alpha and beta (Rev-erbα and β)

Rev-erbα and β are also known as the nuclear receptor subfamily 1, group D, member 1 and 2 (Nr1d1 and 2). The direct targets of REV-ERBα and β transcription also include key metabolic genes controlling lipid and energy homoeostasis (272). Rev-erbα is the only nuclear receptor gene frequently amplified in human breast cancers, which is associated with poor clinical outcomes and survival (273–276). The SNPs and/or deregulation of human Reverbα are also associated with thyroid tumors, pleural mesothelioma, obesity and metabolic syndromes, chronic inflammation, and autoimmune diseases (162, 256, 277–280). Mice lacking Rev-erbα (Nr1d1^−/−) do not have circadian homeostasis and show immune deficiencies, metabolic syndromes, and diet-induced obesity (272, 281, 282).

Beta-transducin repeat-containing protein1 and 2 (βTrCP1 and 2)

βTrCP is also known as F-box/WD repeat containing protein (Fbxw) that belongs to one of the four components of the ubiquitin protein ligase complex. βTrCP-mediated PER cytoplasmic degradation controls the activity of the major negative loop in the molecular clock (283). SNPs, mutations, and/or deregulation of βTrCP in humans are associated with prostate, breast, colorectal, gastric, and pancreatic cancers as well as HCC (284–289). βTrCP plays important roles in cell cycle checkpoints and DNA damage response (290, 291), protein synthesis, cell growth, survival, and metabolism (292–294). It also regulates the pro-apoptotic protein BimEL, NF-κB, and WNT pathways to control immune response and cell survival (77, 295, 296). In vitro, loss of βTrCP stimulates angiogenesis and migration of human cervical and thyroid cancer cells (297).

Mice lacking βTrCP (βTrCP1^−/− and βTrCP2^−/−) do not show overt abnormalities. However, tissue-specific expression of a dominant negative mutant βTrCP1 in mice promotes tumor development in targeted tissues (298, 299). The βTrCP transgenic mice show deregulation of cell adhesion, migration, and proliferation, as well as suppression of p53 via activation of oncogene Mdm2 (290, 300–302).

Conclusion

The compelling evidence provided in this review supports the notion that the mammalian circadian clock suppresses tumor development in vivo. The fact that circadian disruption induces similar pathophysiological changes in the same organ systems in humans and animal models via deregulation of the same molecular pathways suggests that the mechanism of clock-controlled tumor suppression is conserved during evolution. The observation that circadian dysfunction induces a coupled increase in the risk of cancer and other modern-day diseases in affected individuals, such as accelerated aging, neurodegeneration, neuroendocrine dysfunction, metabolic syndromes, and immune deficiencies, indicates that the manifestations of various abnormal physiological conditions but not a single molecular pathway promote cancer development in vivo. These findings have opened an exciting new research direction to investigate the molecular mechanisms of cancer initiation and progression as a consequence of chronic malfunction in mammalian physiology. Such study will have important impact on human health by developing novel strategies for cancer prevention and treatment as, in industrialized societies, the majority of the population experiences chronic misalignment of endogenous circadian systems throughout their lifespan due to personal, social, and professional demands (1, 2, 303).

Key messages.

• Circadian dysfunction is an independent risk factor for cancers in modern societies.

• The circadian clock suppresses tumor development by maintaining homeostasis of physiology.

• The mechanism of circadian control of tumor suppression is conserved during evolution.