THE MOLECULAR ORGANIZATION OF THE CIRCADIAN CLOCK
Transcriptional–translational feedback loops constitute the organizational building blocks of the circadian clock. These are common to a number of organisms, including mammals (Young and Kay 2001; Bell-Pedersen et al. 2005), and the following core circadian clock genes have been identified in mammals: Clock, Bmal1, casein kinase I epsilon (CKIε), cryptochromes 1 and 2 (Cry1, Cry2), period 1, 2, and 3 (Per1, Per2, Per3), and Rev-erb-α. Interaction of clock proteins occurs via the PAS domains (named after the proteins PER–ARNT–SIM), which provide heterodimerization surfaces. The Clock and Bmal1 genes encode basic helix-loop-helix (bHLH)–PAS transcription activators that heterodimerize and induce the expression of Per and Cry genes via binding to E-box elements (CACGTG) present in their promoters (Reppert and Weaver 2001; Bell-Pedersen et al. 2005). Heterodimers of PER and CRY proteins translocate to the nucleus and inhibit CLOCK–BMAL1-mediated transcription through direct protein–protein interactions (Young and Kay 2001; Bellet and Sassone-Corsi 2010). Importantly, E-boxes are very frequent in the mammalian genome, hinting at the wide potential of possible regulation by the circadian machinery. Indeed, the CLOCK–BMAL1 heterodimer regulates the transcription of many CCGs, which in turn influence a wide array of physiological functions external to the oscillatory mechanism. This mediates the output function of the clock, thereby controlling, for example, food intake, hormonal synthesis and release, body temperature, and metabolism. Indeed, many mammalian transcripts undergo circadian fluctuations in their expression levels (Akhtar et al. 2002; Duffield et al. 2002; Panda et al. 2002).
How can the clock have such a profound impact on physiology? Genome-wide array analyses indicate that at least 10% of all expressed genes in any tissue are under circadian regulation (Masri and Sassone-Corsi 2010; Hogenesch and Ueda 2011). This high proportion of circadian transcripts suggests that the clock machinery may direct widespread events of cyclic chromatin remodeling and consequent transcriptional activation/repression. Furthermore, genome-wide studies comparing the central SCN pacemaker and peripheral tissues, such as the liver, revealed that between 5% and 10% of cycling genes were identical in both tissue types (Akhtar et al. 2002; Panda et al. 2002). A recent analysis covering 14 mouse tissues identified approximately 10,000 known genes showing circadian oscillations in at least one tissue (Yan et al. 2008). Not surprisingly, the number of common genes showing circadian oscillation in multiple tissues decreased drastically as the number of tissues included in the comparative analysis increased, with only 41 genes displaying circadian oscillation in at least eight out of 14 tissues (Yan et al. 2008). These findings suggest that the core clockwork, which can be assumed to be common to all tissues, interplays with cell-specific transcriptional systems (Masri and Sassone-Corsi 2010).
Finally, in addition to the transcripts that show the approximately 24-h-based oscillation, hundreds of CCG transcripts exist in the liver cycle at the second and third harmonic of circadian oscillations (periods of 12 and 8 h). Both circadian and ultradian transcripts are severely dampened in ex vivo cellular cultures, whereas in vivo their expression is shifted if the animals are under restricted feeding (Hughes et al. 2009). Thus, both circadian and ultradian transcripts are responsive to systemic cues in vivo, highlighting the critical position of peripheral clocks. These operate indeed as the natural interfaces between nutritional cues and physiology to ensure plasticity and accurate responsiveness of organisms.
CIRCADIAN METABOLISM AND PERIPHERAL CLOCKS
One of the critical discoveries in the field, made about a decade ago, was the finding of independent clocks in peripheral tissues of various organisms (Whitmore et al. 1998; Giebultowicz et al. 2000; Stokkan et al. 2001; Schibler and Sassone-Corsi 2002). This finding had several implications. First, what signal links the central pacemaker to the peripheral clocks, and how independent are the peripheral tissues? It is generally accepted that there are specific humoral signals involved in synchronizing the circadian system, but a comprehensive understanding is still lacking. Another key question has been, How different is the molecular organization of the peripheral clocks versus the central SCN clock? These, and several other questions of fundamental endocrine and physiological significance, remain open. Some important features have, however, been established: The oscillatory function of peripheral clocks in mammals is orchestrated by the SCN (Schibler and Sassone-Corsi 2002). Differently from the SCN, peripheral tissues appear to require physiological stimuli to sustain their circadian rhythms. Thus, it is likely that peripheral clocks are affected by physiological stimuli that may originate from the SCN and/or may be the result of SCN-mediated messages. Indeed, growth factors, some steroids such as glucocorticoids, and retinoic acid have been shown to induce oscillations of clock genes and clock-controlled genes in cultured fibroblasts or peripheral tissues (Le Minh et al. 2001; Preitner et al. 2003). These observations are likely to have physiological relevance because restricted access to food has an effect on peripheral rhythms without affecting the central pacemaker function of the SCN (Schibler and Sassone-Corsi 2002; Preitner et al. 2003). The discovery of independent clocks in peripheral tissues revealed the intrinsic metabolic nature of the circadian system and its homeostatic control.
CIRCADIAN REMODELING OF CHROMATIN
The transcriptional control of a significant proportion of the genome by the clock invokes genome-wide mechanisms of chromatin remodeling. Indeed, the highly specialized, temporally based regulation of transcription that characterizes circadian rhythms elects the cellular clock as a prominent model for the study of dynamic regulations of chromatin remodeling (Crosio et al. 2000; Borrelli et al. 2008; Zocchi and Sassone-Corsi 2010). Finally, because circadian rhythms are tightly coupled to physiological and metabolic control (Rutter et al. 2002), clock-controlled chromatin reorganization is likely to reveal as yet unexplored pathways linking histone modifications to cellular metabolism.
The first evidence linking circadian gene expression to dynamic events of chromatin remodeling was obtained by studying histone modifications in SCN neurons in response to a light stimulus (Crosio et al. 2000). Subsequent studies indicated that activation of CCGs by CLOCK–BMAL1 is associated with circadian changes in histone modifications at their promoters (Etchegaray et al. 2003; Curtis et al. 2004; Naruse et al. 2004), showing that transcription-permissive chromatin states are dynamically established in a circadian-time-specific manner. Specifically, histone H3 is acetylated in chromatin, which encompasses the Per1, Per2, and Cry1 promoters when these genes are actively transcribed (Doi et al. 2006; Ripperger and Schibler 2006). In addition, H3 Ser-10 phosphorylation is involved in the transcriptional response to light in the SCN. Thus light-mediated signaling leads to nuclear responses that in turn influence the state of higher chromatin organization (Crosio et al. 2000). Moreover, the coupled modification of Ser10 phosphorylation and Lys14 acetylation on H3 has been shown to be a hallmark of transcriptional activation (Cheung et al. 2000; Lo et al. 2000), a notion that underscores the implication of histone acetyltransferases (HATs) in circadian gene expression (Masri and Sassone-Corsi 2010). Finally, histone methylation may be important for circadian clock function, as indicated by RNA interference (RNAi) experiments in cultured fibroblasts. The Polycomb group protein EZH2 and methylation at H3 Lys27 are associated with Per gene promoters, extending the activity of the Polycomb group proteins to the core clockwork mechanism of mammals (Etchegaray et al. 2006).
Another study of the protein CLOCK revealed that, in addition to operating as a transcripton factor in combination with BMAL1, it has the enzymatic properties of an acetyltransferase (Doi et al. 2006). The carboxy-terminal region of CLOCK displays a significant sequence homology with the carboxy-terminal domain of ACTR, a domain previously described to have intrinsic HAT activity (Chen et al. 1997). In this region, at least six independent amino acid regions are found to share significant sequence similarity between the two proteins. Importantly, the amino acid residues common to CLOCK and ACTR are evolutionarily conserved in both proteins. It is also noteworthy that CLOCK and ACTR share a number of other structural features outside of the carboxy-terminal glutamine-rich region, including a putative NRID (nuclear receptor interaction domain). Using HAT-in-gel assays, it was demonstrated that CLOCK indeed has intrinsic HAT activity directed toward K9/K14 of histone H3 (Doi et al. 2006) and that it is also able to acetylate nonhistone proteins, specifically its partner BMAL1 (Hirayama et al. 2007). The question of whether CLOCK may be able to acetylate other nonhistone proteins is important, as it may connect its enzymatic, circadian function to a number of cellular pathways. In this respect, the finding that CLOCK acetylates the glucocorticoid receptor the (GR) could be highly relevant (Nader et al. 2009). CLOCK acetylates the GR at multiple lysines in the hinge region, resulting in inhibition of GR binding to DNA. Because glucocorticoids have been implicated in a variety of circadian responses, and are good candidates as possible mediators of the SCN control on the peripheral clocks, these observations could have remarkable physiological importance. Indeed, glucocorticoid secretion peaks just before the onset of the activity period and it is classically regulated by the hypothalamus–pituitary-gland–adrenal axis, which in turn is regulated by the SCN.
The HAT domain of CLOCK contains a stretch of amino acids that shares significant similarity to the so-called “motif A” in the HAT family, denominated MYST (for its founding members MOZ, Ybf/2Sas3, Sas2, and Tip60) (Doi et al. 2006). This is important because highly specific interactions have been described between nuclear receptors (NRs) and proteins with a MYST-type of “motif A,” such as the androgen receptor and the HBO1 protein. Thus, it is conceivable that, in addition to the GR, CLOCK may acetylate other NRs.
SIRT1, A DEACETYLASE WITH CIRCADIAN ACTIVITY
Circadian rhythms and changes in cellular energy state seem to be tightly linked, as suggested by the abnormal metabolic phenotypes displayed by mice mutant for clock genes (Eckel-Mahan and Sassone-Corsi 2009). A genome-wide RNAi approach has revealed that reduced circadian amplitude can be caused by individual knock-down of approximately 1000 genes (Zhang et al. 2009). Importantly, pathway analysis indicates that these genes encode components of the insulin and hedgehog signaling, cell cycle, and folate metabolism. Thus, the clock machinery is interconnected with a number of cellular functions. In this respect, the search for the HDAC (histone deacetylase), that would physiologically counterbalance the enzymatic activity elicited by CLOCK, revealed an intriguing twist.
HDAC-mediated deacetylation of histones correlates with gene silencing (Berger 2007), and several HDACs have been implicated in the reversible acetylation of nonhistone proteins, including p53, Hsp90, MyoD, and E2F1 (Glozak et al. 2005). Four classes of mammalian HDACs exist, based on their structure and regulation (Yang and Seto 2008). Class III is composed of seven mammalian enzyme homologs of yeast Sir2 (silencing information regulator) and are known as SIRT1 to SIRT7. These proteins are structurally distinct from the other HDACs and have the property of dynamically sensing energy cellular metabolism using NAD+ (nicotinamide adenine dinucleotide) as a coenzyme (Bordone and Guarente 2005). In this reaction, nicotinamide is liberated from NAD+ and the acetyl group of the substrate is transferred to cleaved NAD+, generating the metabolite O-acetyl-ADP ribose (Sauve et al. 2006). Owing to the NAD+ dependency, SIRTs are thought to constitute a functional link between metabolic activity and genome stability and, possibly, aging (Bordone and Guarente 2005).
SIRT1, the mammalian ortholog of Sir2, is a nuclear protein that occupies a privileged position in the cell and governs critical metabolic and physiological processes. SIRT1 helps cells to be more resistant to oxidative or radiation-induced stress, promotes mobilization of fat from white adipose tissues, and mediates the metabolism of energy sources in metabolically active tissues (Bordone and Guarente 2005). Thus, the finding that SIRT1 acts as a “rheostat” to modulate CLOCK-mediated acetylase activity and circadian function established an intriguing molecular link (Nakahata et al. 2008). SIRT1 was found to associate with CLOCK and to be recruited at circadian promoters, whereas Sirt1 genetic ablation or pharmacological inhibition of SIRT1 activity leads to significant disturbances in the circadian cycle. Importantly, whereas the protein levels of SIRT1 do not oscillate, as analyzed in several tissues and under various experimental conditions (Fig. 2A) (Nakahata et al. 2008), its enzymatic activity oscillates in a circadian manner (Fig. 2B), with a peak that corresponds to the lowest level of H3 acetylation at various CCGs (Nakahata et al. 2008, 2009). This may also be due to a role of SIRT1 in controlling the stability of some clock proteins, such as PER2 (Asher et al. 2008). Thus, it is possible that SIRT1 contributes to transduce signals originated by cellular metabolites to the circadian machinery (Nakahata et al. 2008). Some further evidence in this direction was obtained when it was found that the levels of NAD+ oscillate in a circadian manner in serum-entrained mouse embryonic fibroblasts (MEFs) and in mice livers (see below; Nakahata et al. 2009).