Biological clocks are intrinsic time-keeping systems that regulate behavior and physiological functions in most living organisms. Recent works in this area have addressed possible molecular links between the endogenous circadian clock and cell cycle regulation. In this review, by addressing how circadian clocks can interfere with the cell cycle and how the disruption of the circadian rhythm may cause defects in regulation of cell proliferation, we highlight this potential connection between circadian rhythm and cell cycle. We also discuss how the acquisition of recent data in circadian clock mechanism may help chronotherapy, which takes into account the biological time to improve cancer treatments, and may open new therapeutic avenues for treating circadian-related diseases.
The existence in all living organisms of numerous biological and physiological processes that occur at well-defined and controlled time intervals is now fully recognized. Among these processes ticking on periodic timing mechanisms, the cell cycle and the circadian rhythm are intensively studied. As the 24 h periodicity of biological clocks (1 , 2) , the mammalian cell division period is around a day, since in human mucosa and skin, a greater proportion of cells tend to divide in the evening between 18 and 24 h (3 , 4) . It is also clearly demonstrated that genes associated with these periodic processes exhibit themselves in rhythmic patterns of expression (3 , 5) . Are these similarities between circadian rhythms and cell cycle clocks entirely coincident or do they suggest an intrinsic relationship between them?
In this review we focus on clinical and experimental evidences for circadian synchronization of cell division up to the establishment of a molecular connection between cell and circadian cycles. The important implications of a circadian control in the cell cycle are also discussed for the optimization of cancer chemotherapy in circadian-related diseases.
Circadian Rhythm: Organization of Biological Clocks.
Circadian rhythms enable organisms to adapt to daily environmental changes such as light, temperature, and social communication, and serve to synchronize multiple molecular, biochemical, physiological, and behavioral processes with each other (5, 6, 7, 8, 9) . As shown Fig. 1 ⇓ , a biological circadian clock is made up of three components: a central circadian oscillator generating autonomously a rhythmic signal, an input pathway that connects the oscillator to external time cues, and an output pathway allowing synchronization of physiology and behavior. In mammals, it is known that the master pacemaker controlling circadian rhythm is located in the SCN 1 of the hypothalamus (10) . The light provided by the day/night cycle is the major factor that resets the activity of the SCN in animals. Importantly, recent works show that, in addition to the central pacemaker of the SCN, other endogenous oscillators are present in peripheral tissues such as the heart (7) , the liver (6 , 11) , and the kidney (8) , as well as in embryos (12) and isolated cells (4 , 6 , 11 , 13 , 14) , suggesting that a cell autonomous circadian clock exists in each cell of the organism. Interestingly, these peripheral clocks can be reset by alternative routes independently of the SCN, for example by changing the feeding time (15 , 16) . Several lines of evidence strongly suggest that the peripheral circadian clocks are not SCN-independent and require inputs from the SCN to drive the rhythmicity (17) . Nevertheless, the mechanisms used for the control of the peripheral clocks by the SCN, which are thought to occur via a combination of neural and humoral signals, are still poorly understood (18) . The importance of humoral pathways is underlined by a well-defined daily secretion of many hormones, such as melatonin, thyroid hormones, or glucocorticoids.
Genetic and biochemical experiments in Drosophila and mice have allowed for the identification of several clock genes, which contribute to entrain and to maintain a circadian rhythm: a clock gene, a gene encoding brain-muscle Arnt-like protein 1 (bmal1), three period genes (per1, per2, and per3), and two cryptochrome genes (cry1 and cry2; Refs. 2 , 19 , 20 ). In addition, there is evidence that the NPAS2 gene, which encodes a functional analogue of the clock, plays a role in the functioning of the circadian clock in the forebrain where it is expressed (21) . The current model for the circadian oscillator is based on two interacting self-sustained transcriptional-translational feedback loops (Fig. 2) ⇓ , which drive the rhythmic expression of clock- controlled output genes (19) . Two transcriptional activators, CLOCK and BMAL1 (or NPAS2-BMAL1 in the cortex and in the vasculature), form a heterodimer, which regulates gene expression by direct interaction of a bHLH motif with a CACGTG response element on the DNA called E-box. Target genes of CLOCK and BMAL1 include several repressor proteins such as PER1, 2, and 3, and CRY1 and 2, which dimerize in the cytoplasm to enter the nucleus where they function as inhibitors of CLOCK/BMAL1 complex, thereby generating a circadian oscillation of their own transcription. In a second regulatory loop, PER provides the positive drive on bmal1 transcription, independently of the BMAL1 protein regulation. Whereas BMAL1 activates transcription via the E-box, it down-regulates its own transcription (2 , 19 , 20) . Additionally, Rev-erbα, a gene encoding an orphan nuclear receptor with a repressing activity on transcription, was shown to play an important role in generating circadian rhythm. Indeed, the Rev-erbα gene is a target of the molecular oscillator, because its expression is activated by the CLOCK/BMAL1 heterodimer and repressed by PER and CRY proteins. The Rev-erbα protein acts as a molecular part of the regulatory loop by controlling the expression of bmal1 (22) . Of note, its paralogue Rev-erbβ exhibits a circadian expression in most tissues of vertebrates, suggesting that it might play an important role in circadian rhythm. Interestingly, the ROR proteins are transcriptional activators of genes that are negatively regulated by Rev-erbs (23 , 24) . The role of Rev-erb/ROR in the circadian pacemaker suggests that the current model of two feedback loops is far too simplistic (19) and that, in fact, several interlocked loops are performing a fine tuning of circadian gene expression as illustrated in Fig. 2 ⇓ . Other couples of positive and negative factors being part of circadian oscillators, such as DBP-related factors and E4BP4, are known (25 , 26) . Dec proteins were found recently to suppress CLOCK/BMAL1-induced activation of the per1 promoter (27) , which suggests that these proteins are important components of an additional regulatory feedback loop.
Circadian Rhythm and Cancer.
Another well-known biological oscillator is the cell cycle. In the eukaryotic cell division cycle, cells oscillate between DNA synthesis and mitosis. A few decades ago, the CDKs, their associated proteins, the cyclins, and CKIs emerged as the main regulatory proteins in the control of cell cycle events (28 , 29) . The tumor suppressor gene products p53 and pRb are also important regulators of the cell cycle (30 , 31) . Molecular mechanisms that ensure the correct progression of the cell cycle have been described, and interlocked feedback loops constitute an essential tool for the regulation of molecular components of the cell cycle machinery (32 , 33) . As a result of this tight regulation, defects in cell cycle checkpoints may be responsible for an uncontrolled and continuous proliferation leading to cancer development (34 , 35) .
From Experimental and Clinical Data Showing Circadian Coordination of Cell Cycle Regulation.
Several in vivo studies have strongly supported the concept that circadian time coordinates cell-cycle progression. At the tissue level, experimental and clinical data suggest a circadian coordination of cells progressing through the cell cycle. In the 70s, Scheving et al. (36) were the first to report a significant circadian variation in mitotic activity in the murine duodenum. In addition, they have confirmed a circadian variation of cell proliferation along the alimentary tract from tongue to rectum epithelia (37) . Analogous observations have been reported from human rectal mucosa biopsies performed over 24 h (38) and human oral mucosa (39) . In humans and rodents, circadian variations were also documented in progenitor cells of bone marrow (40) , in biopsy samples of intact human oral mucosa (3) , and in neuronal cells (41) . Moreover, correlation of the timing of clock gene expression with the timing of specific cell cycle stages largely suggests that the circadian clock might be involved in the ordering of cell cycle events. For example, in humans, the major peak of hper1 expression coincides with the G1 phase, whereas the peak of hbmal1 expression coincides with M phase (3 , 4) . However, active tissue proliferation (as reported in intestinal mucosa, skin, and bone marrow) is not a prerequisite for synchronous clock gene expression, because many mouse and rat organs that are not actively proliferating (e.g., muscle and kidney) have been shown to have a circadian expression of per, bmal1, and cry genes (7 , 8) . Similarly, most adult neurons do not divide and manifest potent molecular rhythm. Molecular rhythms persist in rat fibroblast cells even if cell division is blocked (13) , suggesting that rhythms are controlled separately from cell division. In nondividing tissues the clock genes might provide a pathway to control nonproliferative metabolic and physiological processes by the regulated expression of clock controlled genes.
Of special interest was the demonstration that the cell proliferation and apoptosis in rapidly renewing tissues are circadianly synchronized (42 , 43) . Equally interesting is the demonstration that circadian rhythm of cell division remains evident at all of the stages of tumor growth (44) . The proliferation of tumor cells follows autonomous circadian patterns that are not in phase with normal cells (45) , and DNA synthesis in tumor cells shows a clear diurnal rhythm (46, 47, 48) .
Other clinical observations support the fact that growth control might be influenced by circadian rhythm. Indeed, it is now well documented that alterations in circadian rhythm can be associated with cancers in both animal and human tumors (49) , and that chronobiology largely improves cancer therapy (50 , 51) . It has also been reported that the irregular circadian cycles due to night-shift work in humans or a constant exposure to light in rodents increase mammary tumorigenesis, although a mechanistic explanation for this observation is probably extremely complex (52, 53, 54) . Patients with advanced cancers sometimes present a disturbance of their sleep-wake cycle (55 , 56) . In such cases, the link between the circadian clock and cancer may also be very indirect. It has been observed that phase shifting of mice leads to a corresponding shift in the timing of the cell cycle events in both gut and bone marrow (57) . Such phase shifting is associated with a shift in the circadian expression of per1 in the rat SCN and in peripheral tissues (14) . In addition, circadian oscillation of hormone levels may be involved in the development of hormone-sensitive cancers such as breast cancer (58) . Suppression of nocturnally released melatonin levels by light increases the development of spontaneous mammary tumors in rats and humans (52 , 59 , 60) . Pinealectomy also produces an acceleration of aging in normal strains of mice, which is a result of the progressive age-dependent decrease of normal circadian thyroid gland activity (61) . Aging has been shown to be associated with complex alterations of rhythms (62) . All of these data thus suggest the existence of complex indirect influences of circadian rhythms on the mechanisms controlling cell proliferation.
… to the Identification of Circadian Variation of Cell Cycle Genes.
A detailed study of Bjarnason et al. (3) has examined circadian variations in the levels of nuclear expression of cell cycle-associated proteins (cyclins) in biopsy samples of intact human oral mucosa over a 24-h period. A significant circadian variation of p53, and cyclins E, A, and B1 was found with a normal physiological progression over time. It can be hypothesized that p53-dependent apoptosis is rhythmic as is the expression of p53. The fact that the activity of thymidylate synthase, an enzyme involved in DNA synthesis and used as S phase marker, shows a significant circadian rhythm with a peak in the early afternoon additionally reinforces the existence of a circadian coordination of cell cycle events in human tissues (4) .
Circadian variation in proliferative activity has also been associated recently with a circadian variation in the expression of cell cycle-associated proteins in serum shock fibroblasts. Using high density oligonucleotide microarrays, our laboratory detected four cycling transcripts coding for proteins known to play a role in the regulation of the cell cycle, such as the replication-dependent histone H2A1, the cyclin D3, the cell division protein kinase CDK4, and the nuclear serine/threonine kinase RING3 (5) . Cyclin D3, CDK4, and RING3 show similar patterns of expression between 12 and 56 h with two peaks, one at 24 and one at 48 h. Cyclin D3 and CDK4 are known to interact directly and to play an important role in the G1-S transition of the cell cycle (64) . RING3 is a serine/threonine kinase, which is able to activate cyclin D11, A, and E, genes, and is implicated in the cell cycle (63) . The discovery that many genes encoding important cytoskeletal elements (tuba4, tubb3, tubb4, and tubb5) strongly oscillate in phase and that genes involved in cell proliferation and apoptosis such as cyclin D1, cyclin A, mdm-2, or gadd45a are deregulated in per2-deficient mice reinforces the link between circadian rhythm and cell-cycle regulation (65) .
Glucose was characterized as a direct resetting signal for peripheral clocks, because it is able to induce circadian rhythm in confluent rat-1 fibroblasts by down-regulation of per genes (66) . At the same time, glucose also up-regulates four genes involved in DNA replication peaking at the G1-S transition such as thymidylate synthase and gadd45α, and down-regulates two genes peaking in G2 and M phases, such as cyclin B1. By microarray studies it has also been demonstrated that after 24 h of fasting, genes involved in apoptosis and cell cycle (faf1, ddx1, tnfr-I, dapk1, and c-myc) show diurnal variation in the liver (8) . Thus, it is likely that the circadian control of cell proliferation takes into account the physiological context, such as the glucose level. Clinical studies have also established that circadian variation in bone resorption occurs due to food intake and is acutely reduced during fasting (67) . All of these results suggest that both peripheral clocks and the cell cycle respond to a change in feeding time, and that there is an intricate cross-coupling among physiology, biological rhythm, and cell proliferation.
Defects in Circadian Clock Genes and Cancer Development.
Interestingly, the disruption of circadian rhythm in SCN-lesioned mice is related to an accelerated growth of implanted malignant tumors, suggesting that the circadian clock of the host might play an important role in the endogenous control of tumor progression (68) . Fu et al. (65) have recently published strong evidence that the genes of the circadian clock play an important role in tumor suppression and DNA damage through the control of cell proliferation and death. Indeed, they observed that per2 mutant mice developed more salivary gland hyperplasia and more lymphoma than wild-type mice. The authors clearly demonstrated that per2 mutant mice were more sensitive to γ-radiation as they showed premature hair graying, hair loss, a high frequency of lymphoma formation, and a reduced apoptotic response in the thymus compared with the wild-type mice. Whereas the expression of major clock genes was induced in wild-type murine liver in response to γ-radiation, the mutation in per2 severely diminished or abolished the induction of clock genes, indicating that the core circadian genes respond to γ-radiation in a coordinated manner in vivo. Surprisingly, whereas the expression of the c-myc gene is circadian in wild-type liver, in per2 mutant mice the expression of this gene is shifted and dramatically increased. The expression of two c-myc regulated genes, cyclin D1 and gadd45α, is circadian in wild-type mice but is altered in mutant mice, confirming that these two c-myc target genes are under circadian control in vivo. Importantly, per2-deficient mice also displayed aberrant temporal expression of other genes involved in cell proliferation and apoptosis such as cyclin A and mdm-2. Because the mdm-2 gene is involved in the negative post-transcriptional regulation of p53, which might still be controlled at the posttranscriptional level by circadian clock in vivo (3) , the per2 gene probably functions in tumor suppression by regulating the DNA damage response pathway. Some experiments on embryonic fibroblasts in cell culture have ultimately demonstrated that the proto-oncogene c-myc is a direct target of the circadian regulators NPAS2/BMAL1 through E-box-mediated reactions (65) . As an additional support, N-myc gene transcription was shown previously to be differentially controlled by RORα and Rev-erbβ in transient transfection assays (69) . Indeed, whereas the oncogenic potential of N-myc in primary rat embryonic fibroblasts is enhanced by the cotransfection of RORα, which activates N-myc expression, the oncogenic potential of N-myc is decreased by Rev-erbβ, which is a N-myc repressor. The role of other clock gene products such as the CLOCK/BMAL1 heterodimer on N-myc expression is still unknown. Myc family proteins clearly play a key role in cell proliferation and apoptosis (70) . Deregulated expression of myc genes is a frequent observation in tumors (71 , 72) . Inactivation of the per2 gene results in deregulation of bmal1 expression, leading to c-myc overexpression, which subsequently has a high incidence of tumor development and genomic DNA stability (65) . This clear mechanistic link between the circadian clock and cell proliferation strongly suggests that the circadian clock suppresses cancer development in vivo by regulating the expression of clock-controlled genes implicated in growth control (Fig. 3) ⇓ . The fact that an essential clock gene per2 is associated with cell proliferation control is in agreement with previous observations on the circadian synchronization of rapidly renewing tissues (42 , 43) .
SOME COMMON CONTROL ELEMENTS OF THE TWO PATHWAYS
The intracellular clock mechanism is driven by interlocked positive and negative transcriptional feedback loops that lead to recurrent rhythms of key clock transcription factors on the RNA and protein levels. Progression through the cell cycle is controlled by the regulated expression of specific protein complexes, which regulate cell cycle activity. Together, these proteins regulate the cellular decision to proliferate, to differentiate, or to arrest the cell cycle. Despite these mechanistic differences, several studies allow us to point up some biochemical and molecular similarities shared by both the circadian rhythm and the cell cycle as reported in Fig. 4 ⇓ .
Hormones and Growth Factors.
The circadian and cell cycles are regulated by common factors like metabolites, hormones, or nutrients, which are affected by food intake (8 , 67 , 73) . As said previously, glucose is able to both reset peripheral clocks and regulate genes involved directly in cell cycle progression (66) . As another example, melatonin is a chronobiotic hormonal regulator that relays the circadian rhythm to the peripheral organs for physiological regulations (74) . Furthermore, this molecule has a clear role in the sleep/wake cycle regulation (75) . Melatonin also controls cell proliferation (60 , 76) . At physiological concentrations, it inhibits cancer cell proliferation in vitro through specific cell cycle effects. It also acts as a differentiating agent in cancer cells, and lowers their invasive and metastatic status through alterations in adhesion molecules and maintenance of gap junctions (59) . In some cancer cell types, melatonin can induce apoptotic cell death alone or in association with other drugs (77) . Moreover, other molecules that act as synchronizers of peripheral clocks can also regulate expression of proto-oncogenes. For example, retinoids were shown both to play an important role for the synchronization of the peripheral clock (78) and also to be negative regulators of N-myc expression (79 , 80) . Nitric oxide release is reported to play an important role in tumor progression (81) , and to also have a role in circadian clock and sleep/wake cycle (82 , 83) . Excitingly, the nitric oxide synthase activity was reported to be both increased in tumors (84) and to follow circadian variation in mouse tissue (85) , implying the central role of clock control in cancer.
Cell cycle and circadian clock machineries share mechanisms to control the oscillation of phosphorylated proteins. In mammals, Casein Kinase I-mediated phosphorylation of PER proteins promotes their degradation via the ubiquitin-proteasome pathway (86 , 87) . A circadian rhythm of Casein Kinase I subcellular distribution was also shown, because it only accumulates in the nucleus during the time of transcriptional inhibition and results in prominent circadian variations of the phosphorylation status of nuclear PER, CLOCK, and BMAL1 proteins (88 , 89) . In hamsters, the spontaneous, semidominant mutation in the tau locus, which encodes a Casein Kinase I, leads to an inability to phosphorylate the PER proteins and to a marked shortening of the circadian period (90) . In humans, familial advanced sleep phase syndrome is associated with a mutation of human per2 at a Casein Kinase I phosphorylation site, again indicating the crucial function of this enzyme in establishing a circadian period (91) . In addition, Casein Kinase I has been linked to the regulation of DNA repair and to certain aspects of cell division (92) . A study has also reported that Casein Kinase I is essential for proper and timely cell cycle progression during early development in the mouse (93) . It has been shown recently that a specific inhibitor of Casein Kinase I has profound effects on the progression of the cell cycle in a p53-dependent manner (94) . It was also demonstrated that the phosphorylation of β-catenin by Casein Kinase I is promoted by the activity of the kinase GSK-3 (95) , which is shown to be a regulator of circadian rhythmicity (96) . Interestingly, β-catenin interacts with TCF/LEF transcription factors to activate c-myc and promote tumorigenesis. The phosphatidylinositol 3′-kinase/Akt/glycogen synthase kinase-3- dependent signaling pathway also plays a major role in regulating the proliferative and apoptotic response of malignant cells (97) . Equally, the phosphorylation activity of Casein Kinase II was shown to be important for the normal function of the circadian oscillator, probably by acting on the timing of nuclear entry and/or the stability of clock proteins (98, 99, 100) . It is also involved in the control of cell cycle progression as a critical regulator of c-Myc protein (101) . Additionally, Casein Kinase II phosphorylates p53 in the DNA damage response (102) and is frequently up-regulated in human cancers (103) . Similarly, CDK inhibitors are not only regulators in cell cycle progression but were also reported to perturb the circadian rhythm of the eye probably due to their inhibitory effects on transcription (104) . 5,6-Dichlorobenzimidazole, a well-known CKI, has been described as a potent inhibitor of Casein Kinase II, another casein kinase involved in circadian timing (100 , 105) .
Phosphorylation and Proteolysis.
Numerous studies suggest that cAMP and cyclic GMP mediate their effects by regulation of kinases. One important aspect of cAMP is its ability to stimulate cell proliferation in many cell types, whereas inhibiting cell growth in others by acting differentially through the MAP kinase cascade (106) . Circadian changes of cAMP and cyclic GMP levels have been evocated to be essential for the control of the cell cycle by the circadian clock (107 , 108) . MAP kinase pathway is also essential for light response of transcriptional activation of clock genes (109) . A role for the cAMP-responsive element binding protein CREB has been demonstrated in the resetting of the circadian clock because light-induced phase shifts are accompanied by a cAMP-responsive element-dependent transcription in the SCN (110 , 111) . The phosphorylation of CREB is involved in the molecular mechanism that synchronizes the circadian clock to the day/night cycle. Furthermore, the induction of the per1 promoter by activation of cAMP and MAP kinase signaling pathways has been shown to involve CREB and to be distinct from CLOCK/BMAL1-driven transcription required (112) . Moreover, CREB interacts directly with CDK/cyclin complexes and is shown both to be a central regulator in the response to DNA damage (113 , 114) , and to be involved in cellular transformation and development of cancer (115) .
Phosphorylation determines the cellular localization and stability of clock and cell cycle proteins. Even if a tight regulation of the expression level of clock components is a critical process for maintaining a 24-h period, it has been shown that protein products are progressively phosphorylated before degradation (89 , 116) . The proteolysis of proteins during specific phases of the cell cycle also appears to be important for regulation and control of the cell cycle (117 , 118) . Moreover, the intracellular concentration of cyclin is regulated by a cyclic, selective, and rapid proteolysis. This periodic degradation provides a powerful way to ensure correct progression of the cell cycle. The control of nucleo-cytoplasmic localization has been described for the cell cycle regulators, and it is tempting to speculate that a similar mechanism is involved in the biological clock, as CRY and PER proteins are redistributed from the cytoplasm to the nucleus (119) . Finally, F-box proteins were shown to control the levels of clock proteins through a ubiquitin proteasome pathway (120) . F-box proteins also play an important role at the G1-S transition phase of the cell cycle by targeting phosphorylated cyclins and CKIs for degradation by proteasome (121) .
E-Box and bHLH Transcription Factors.
The binding of a bHLH transcription factors on the DNA response element E-box allows the coupling at the level of target promoters of clock-controlled and cell-cycle genes. The involvement of the E-box in the regulation of gene expression by the circadian clock is now fully recognized (1 , 19 , 26) , although the molecular features required for circadian control are yet to be defined. One important characteristic of circadian E-boxes is their ability to specifically recruit a BMAL1/CLOCK complex through bHLH motifs. The E-box is also known to affect proliferation, differentiation, and cell death. For example, c-Myc heterodimerizes with the Max protein to activate transcription through a DNA E-box. Interestingly, one mechanism of c-Myc inactivation is a direct competition of c-Myc with the Mad protein, which can also dimerize with Max and act through an E-box. It has been shown that c-Myc interacts through a bHLH motif with the E-box sequences in the promoter of target genes (70) . By recruiting the proto-oncogene c-Myc, an E-box can drive cells to become growth factor independent, to speed through the G1 phase of the cell cycle, or to undergo apoptotic death (122 , 123) . Similarly, circadian regulators may target genes that are controlled by c-Myc. c-myc itself might be controlled by the circadian clock because it contains multiple consensus E-box motifs. The molecular basis for differential use of the E-box-containing promoters of the vasopressin and cyclin B1 genes has been investigated recently (124) . These two genes can respond both to the transcriptional oscillators driving the circadian clock and the cell cycle suggesting that the cellular environment has a significant role in maintaining circadian and noncircadian regulation of transcription. These and other factors could mediate the use of an E-box either as an element of cell cycle-related events or as a support of circadian gene expression.
Resetting by Light and Serum Response.
Many of the light-induced genes in the SCN are also immediate early genes after serum stimulation of growth-arrested cells such as c-fos, fosB, and junB (13) . Therefore, these genes may not only play a role in the resetting of circadian clocks but may also be implicated in cell cycle regulation (125) . Interestingly, the photolyase/cryptochrome family of proteins has been shown to both repair damaged DNA and reset the circadian clock as sensors of environmental light (126) .
The conservation of essential elements for the molecular regulation of circadian and cell cycles underlies similarities in the delicate molecular mechanisms, which allow both the ordered progression along the cell cycle and the maintenance of the circadian rhythm. The fact that circadian and cell cycles probably evolved independently and are controlled by different players reinforces the idea of a very tight and complex relationship between these two processes. The main goal of future research in this area is to understand in detail the molecular nature of these links.
MEDICAL PERSPECTIVES: IMPROVING CANCER THERAPY THROUGH TIMING
For most cancer therapies, the toxic:therapeutic ratio, the balance between harm to the host and harm to the tumor, is usually not favorable. Additionally, the response rate of established tumors to therapy is often not optimal. Therefore, approaches that might decrease toxicity whereas still maintaining or increasing the antitumor effect have been largely studied, including new drug development, use of biochemical or biological modulators to lessen host toxicity, and improvement of drug delivery and of combinatorial treatment.
Given the importance of circadian coordination in normal and tumoral physiology (49 , 52 , 53 , 55 , 56 , 58 , 60 , 68) , circadian drug timing may be one additional approach to additionally enhance the efficacy of cancer therapy. The first question to address is how drug administration at selected time of day can impact the efficacy of therapeutic strategy. There is now experimental and clinical evidence that the tolerability and the efficacy of radiotherapy and antitumor drugs are dependent on both circadian rhythm and cell cycle phase (44 , 49 , 50) . Studies have documented a time-dependent toxicity of radiations both for several normal tissues and for whole animal (127 , 128) . When mice are irradiated at different times of day a clear circadian rhythm was observed in apoptotic cells without change in the circadian variation of cell proliferation (128 , 129) . It is also well established that antitumor drugs show large differences in toxicity and antitumor response depending on the time of day of the therapy as reported for fluoropyrimidines, methotrexate, or platin complexes (50 , 51 , 130) . Generally, most cancer drugs target cells that are engaged in DNA synthesis more severely than cells that are not dividing. It is clearly reported that DNA synthesis rhythm in healthy cells is in phase opposition with that of tumor cells (45, 46, 47, 48) . It is also shown that the best treatment time to induce apoptosis is in the morning coinciding with the time when most of the target cells are in G2-M transition. Rhythmic changes in cell division can in part explain the circadian variation in sensitivity of proliferating tissues to chemotherapy. The development of drugs that modulate the regulation of cell cycle may represent a relevant approach to treat circadian-related cancers. For example, as a clock-controlled gene, thymidilate synthase gene, could constitute a specific target of circadian-related cancers. Another interesting approach would be to consider circadian characteristic of cancer cells to target the time of their highest proliferative activity. Applications of these strategies to cancer therapy will enhance the ability to control cancer at the same time minimizing the damage done to normal tissues.
The possibility that abnormalities in cellular circadian time keeping could contribute or even cause malignant cell transformation is another interesting approach. Cellular rhythms can modulate the generation or the catabolism of intracellular cytotoxic substances, their interaction with the molecular targets leading to cell dysfunction or death, and the repair of cytotoxic damage (57 , 131 , 132) . It is also well shown that the majority of tumors produce cytokines and growth factors, which can modify the central and the peripheral circadian systems (133) . For patients suffering from circadian-related cancers, specific therapies to restore a normal circadian rhythm should result in improvement in end points due to circadian disruptions (depression, performance, and sleep quality). As used to cure transient disturbance of the circadian rhythms due to jet lag, rotational shift work, seasonal depression, insomnia, and other sleep disorders, therapy combining chronobiotics such as melatonin, light therapy, or sleep regulation could better coordinate the circadian rhythms of cancer patients to their environments. Such treatment may largely help to improve the quality of life of cancer patients and contribute to improve the therapeutic efficacy of cancer chronotherapy.
More basically, experimental research of cancer-associated circadian system dysfunctions require models, which allow the examination of circadian system in pathological situations. From patients suffering from familial advanced sleep syndrome (91 , 134) , it should be of interest to examine their predisposition to develop cancer, how the expression of clock genes is affected in the case of tumors, and also to decipher the synchronization mechanisms of biological clocks in tumors compared with other healthy peripheral clocks. The collection of systematic biopsies from cancer patients, even restricted, will also allow for the validation of data observed in current circadian rhythm animal models (65 , 90 , 135) . Moreover the development of noninvasive techniques to precisely follow the circadian system in animals and humans would be a huge advance. Systematic analyses of all of the relevant clock genes and clock-controlled genes involved in cell cycle regulation, in healthy and tumoral tissues, both in experimental and human models, should help to better understand the mechanisms between cell and circadian cycles. Similarly, such understanding should lead to a new strategy for pharmacological manipulation of the human clock to improve the treatment of circadian-related cancers and other clock-associated disorders.
The impressive amount of information accumulated over the last decades regarding circadian clocks illustrates the complexity of the links among environment, circadian clock, and the basic cellular functions. Recent studies suggest that circadian system is important for proper growth control, and is consistent with the apparent circadian regulation of cell proliferation and apoptosis. Also, in certain conditions, cancer can be a direct consequence of the absence of the circadian regulation. The elucidation of the signals involved in these complex interactions and of the mechanisms that govern the interplay between cell and circadian cycles has just begun. In the near future, it would be important to better understand the exact contribution of the circadian clock in the cell cycle regulation and to decipher the role of each molecular actor. Additional studies addressing the mechanisms through which the cellular circadian clock regulates cell-cycle checkpoints will be greatly facilitated by the possibility of investigating cell cycle regulation in models of circadian rhythms. Moreover, the time-dependent response of animals and humans to radiotherapy or to anticancer drugs reinforces the potential importance of circadian principles in cancer therapies. At long term, the development of new therapies that harmonize the natural circadian rhythm in cancer patients should largely improve the quality of life and the chronotherapeutic treatment of cancer patients.
We thank Juliette Rambaud and Michael Schubert for critical reading of the manuscript and helpful suggestions.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Grant support: Centre National de la Recherche Scientifique, MENRT, Association pour la Recherche contre le Cancer, and Région Rhône-Alpes.
Requests for reprints: Vincent Laudet, Ecole Normale Supérieure de Lyon, Structure and Evolution of Nuclear Hormone Receptors, UMR 5665 Centre National de la Recherche Scientifique, 46, allée d’Italie, 69364 Lyon Cedex 07, France. Phone: 33-4-72-72-81-90; Fax: 33-4-72-72-80-80; E-mail:
↵1 The abbreviations used are: SCN, suprachiasmatic nucleus; bHLH, basic helix-loop-helix; CDK, cyclin-dependent kinase; CKI, cyclin-dependent kinase inhibitor; cAMP, cyclic AMP; MAP, mitogen-activated protein.
- Received May 14, 2003.
- Revision received August 1, 2003.
- Accepted August 5, 2003.
- ©2003 American Association for Cancer Research.