Clock


REGULATION

Translational Regulation

The Clock gene plays an essential role in the manifestation of 24 h circadian rhythms in mice and is a member of the basic helix-loop-helix (bHLH) PER-ARNT-SIM (PAS) superfamily of transcription factors. A novel Drosophila bHLH-PAS protein that is highly homologous to mammalian CLOCK has been characterized. Transcripts from this putative Clock ortholog (designated dClock) undergo daily rhythms in abundance that are antiphase to the cycling observed for the RNA products from the Drosophila melanogaster circadian clock genes period (per) and timeless (tim). Furthermore, dClock RNA cycling is abolished and the levels are at trough values in the absence of either PER or TIM, suggesting that these two proteins can function as transcriptional activators, a possibility which is in stark contrast to their previously characterized role in transcriptional autoinhibition. Finally, the temporal regulation of dClock expression is quickly perturbed by shifts in light-dark cycles, indicating that this molecular rhythm is closely connected to the photic entrainment pathway. The isolation of a Drosophila homolog of Clock together with the recent discovery of mammalian homologs of per indicate that there is high structural conservation in the integral components underlying circadian oscillators in Drosophila and mammals. Nevertheless, because mammalian Clock mRNA is constitutively expressed, these findings are a further example of striking differences in the regulation of putative circadian clock orthologs in different species (Bae, 1998).

vrille, Pdp1, and Clock form a second feedback loop in the Drosophila circadian clock

The Drosophila circadian clock consists of two interlocked transcriptional feedback loops. In one loop, Clock/Cycle activates period expression, and Period protein then inhibits Clock/Cycle activity. Clock is also rhythmically transcribed, but its regulators are unknown. vrille (vri) and Par Domain Protein 1 (Pdp1) encode related PAR family bZIP transcription factors whose expression is directly activated by Clock/Cycle. Vri and Pdp1 proteins are shown to feed back and directly regulate Clock expression. Repression of Clock by Vri is separated from activation by Pdp1 since Vri levels peak 3-6 hours before Pdp1. Rhythmic vri transcription is required for molecular rhythms, and the clock stops in a Pdp1 null mutant, identifying Pdp1 as an essential clock gene. Thus, Vri and Pdp1, together with Clock itself, comprise a second feedback loop in the Drosophila clock that gives rhythmic expression of Clock, and probably of other genes, to generate accurate circadian rhythms (Cyran, 2003).

vri and Pdp1 encode basic zipper transcription factors with highly conserved basic DNA binding domains, suggesting they bind the same set of target genes. vri and Pdp1 are both direct targets of Clk/Cyc. First, a test was performed to see which Pdp1 isoform(s) are clock-controlled since four alternative promoters and alternative splicing generate six Pdp1 isoforms in vivo. RNase protection probes specific for the different isoforms revealed that only Pdp1∊ RNA levels oscillate in adult fly heads (Cyran, 2003).

Taking time points every three hours during a light-dark (LD) cycle revealed that vri and Pdp1∊ RNA levels oscillate with similar phases to one another, but peak levels of Pdp1∊ are not reached until 3-6 hr after the peak of vri RNA levels. Oscillating Pdp1∊ RNA levels are also seen in constant darkness. Pdp1∊ RNA levels were high at both ZT2 and ZT14 in per0 and tim01 mutants. Pdp1∊ RNA is low at both ZT2 and ZT14 in ClkJrk and cyc0 mutants at levels close to the Pdp1∊ RNA levels at ZT2 in wild-type flies. The phase of Pdp1∊ RNA expression in wild-type flies, and the loss of rhythms in clock mutants, are both consistent with Pdp1∊ transcription being regulated in a similar manner to per, tim, and vri transcription. Indeed, analysis of the first 4 kb of sequence upstream of the start site of Pdp1∊ transcription reveals six perfect CACGTG E boxes, which are potential Clk/Cyc binding sites. This is similar to the vri promoter, which has 4 E boxes in 2.4 kb. Thus, Pdp1∊ is the clock-regulated Pdp1 transcript (Cyran, 2003).

The different phases of vri and Pdp1∊ RNAs may reflect subtly different transcriptional activities of their promoters and/or different mRNA half-lives. Thus, the vri promoter could be stronger than the Pdp1∊ promoter, and vri RNA may have a shorter half-life than Pdp1∊ RNA. Indeed, the vri 3′ UTR contains seven copies of an AATAA element, likely to be associated with mRNA instability (Cyran, 2003).

Direct regulation of Pdp1∊ expression by Clk/Cyc make it likely that Pdp1∊ protein would be found in clock cells as shown for vri. Pdp1 protein is detected at night (ZT21) but not by day (ZT10) in larval pacemaker cells, marked by the neuropeptide pigment dispersing factor (PDF). Oscillation of Pdp1 protein continues in constant darkness in wild-type pacemaker cells but is blocked by null or dominant-negative mutations in the per, tim, Clk, and cyc clock genes (Cyran, 2003).

A robust oscillation in Pdp1 levels is also visible in photoreceptor cells of the adult eye, which contain functional clocks. Low Pdp1 levels are seen during the day at ZT9, and high levels in the middle of the night at ZT18. The oscillation is especially clear in the outer photoreceptor cell nuclei. Pdp1 at ZT18 colocalizes with ELAV, which marks the nuclei of neurons. Although antibodies to Pdp1 do not distinguish between the different Pdp1 isoforms, RNase protection data and Western blots detect rhythmic expression of only Pdp1∊ in fly heads. Pdp1 protein is thus rhythmically detectable in both central and peripheral clock cells and it is a nuclear protein, as predicted by its ability to activate transcription (Cyran, 2003).

Current models of the Drosophila circadian oscillator are based upon rhythmic activation of per/tim transcription by cycling levels of Clk/Cyc, and rhythmic repression of per/tim transcription by cycling levels of Per/Tim. While these models explain Per and Tim oscillations, the molecular mechanisms underlying cycling of Clk/Cyc have been unknown. This study identifies Vri as a rhythmically expressed Clk repressor and Pdp1∊ as a rhythmically expressed Clk activator. Vri and Pdp1∊ are shown to directly regulate Clk transcription by binding the same site in the Clk promoter. Pdp1 is required for circadian clock oscillation and for Clk expression, thus establishing it as a novel and essential clock gene. Vri and Pdp1∊ proteins accumulate with a phase delay that presumably underlies sequential repression and activation of Clk transcription. Thus, Vri, Pdp1∊, and Clk form a second feedback loop in the circadian oscillator responsible for regulating rhythms in Clk/Cyc levels (Cyran, 2003).

A second feedback loop in the Drosophila clock, interlocked to the first feedback loop, has been predicted to explain antiphase rhythms of Clk and per expression. Direct regulation of vri and Pdp1∊ transcription by Clk/Cyc, and direct regulation of Clk expression by Vri and Pdp1∊ proteins establishes the existence of this second loop and identifies its components (Cyran, 2003).

The first loop of this model starts with activation of per and tim expression by Clk/Cyc at about noon. Per/Tim then feeds back to inhibit Clk/Cyc activity during the second half of the night. In the second loop, Clk/Cyc also activates vri and Pdp1∊ transcription at about noon. vri and Pdp1∊ RNAs and proteins accumulate with different kinetics such that Vri protein accumulates first and represses Clk expression. Pdp1∊ protein then accumulates and activates Clk transcription after Vri-mediated repression ends in the middle of the night. However, newly produced Clk protein is inactive due to the presence of Per repressor. Once Per is degraded, Clk/Cyc reactivates per/tim and vri/Pdp1∊ transcription to start a new cycle. The two loops are linked together by Clk/Cyc and restart simultaneously (Cyran, 2003).

Conceptually, a molecular clock must separate the phases of clock gene transcription and repression otherwise clock components reach a stable steady state. The delay separating active transcription and repression of per/tim is controlled by the Double-time and Shaggy/GSK3 protein kinases that regulate Per/Tim accumulation and nuclear transport. The phases of Clk transcription and repression are separated by two mechanisms: (1) accumulation of Vri protein before Pdp1∊, which ensures that repression of Clk precedes activation; and (2) Per inhibition of Clk/Cyc activity in the early morning which prevents reactivation of vri and Pdp1∊ transcription even when Clk levels are high (Cyran, 2003).

Does the model fit the data? The model explains the observation that Vri represses Clk independently of nuclear Per/Tim. It also suggests that in a per0 background, Clk expression is repressed because of high Vri protein levels. High levels of Vri must therefore dominate over high Pdp1∊ levels and suppress Clk expression in per0 flies. Indeed, overexpression of vri is dominant and stops the clock in an otherwise wild-type background with constantly low Clk expression (Cyran, 2003).

However, this model does not immediately explain why Clk RNA levels are high in ClkJrk and cyc0 mutants. In the absence of Clk/Cyc function, vri RNA levels are low, and the consequently low levels of Vri protein would not be sufficient to repress Clk expression. But how can expression of Clk RNA be maximal with very low Pdp1∊ levels in ClkJrk and cyc0 mutants? This question is especially relevant given the very low levels of Clk in Pdp1P205 homozygous mutant larvae in constant darkness, which makes the existence of additional factors that positively regulate Clk expression in constant conditions unlikely. The simplest explanation is that the very low levels of Pdp1∊ RNA present in ClkJrk and cyc0 mutants are still sufficient to give enough Pdp1∊ protein to activate Clk when competition from Vri is minimal due to very low Vri protein levels. Indeed, Clk RNA levels are close to their peak at ZT3 and ZT6 in wild-type flies when both Vri and Pdp1∊ levels are very low. In contrast, Pdp1∊ protein is totally absent in Pdp1P205 null mutants because the Pdp1 gene is deleted and thus, Clk is at low levels. However, further work is required to test this hypothesis (Cyran, 2003).

The model can also be used to explain how clock-controlled genes are expressed with different phases. Genes activated by Clk/Cyc will reach maximum RNA levels at ∼ZT14 and these include per, tim, vri, and Pdp1∊. Genes regulated by Vri and Pdp1∊ will peak at ∼ZT2 and Clk is one example. Another candidate Vri/Pdp1∊ regulated gene is cryptochrome (cry), whose RNA levels oscillate in phase with Clk RNA and follow the same pattern as Clk in clock mutants. Indeed, overexpression of Vri represses cry expression, and the cry promoter contains functional Vri (and therefore probably also Pdp1∊) binding sites (Cyran, 2003).

It is also conceivable that certain DNA sequences bind Vri with higher affinity than Pdp1∊ or vice versa. One could then imagine two promoters, one with 5 optimal Vri and another with 5 optimal Pdp1∊ binding sites, that would give RNA expression profiles differing by ∼2-4 hr. Such a mechanism may help to explain the multiple peaks of rhythmically expressed genes found in Drosophila (Cyran, 2003).

Most clock genes are conserved between Drosophila and mammals, and they function in a broadly similar mechanism. For example, peak levels of Bmal1 and Clock RNAs are antiphase to mPer1 and mPer2 in mice just as Clk RNA peaks in antiphase to Drosophila per. A recent study identified the clock-controlled Bmal1 repressor, which parallels the Vri repression of Clk data presented in this study. The data extend the similarities of the Drosophila and mammalian clocks and suggest the existence of a rhythmically transcribed Bmal1 transcriptional activator that plays an analogous role to Drosophila Pdp1 in the second mammalian feedback loop (Cyran, 2003).

However, the Bmal1 repressor is REV-ERBα, an orphan nuclear receptor, which is unrelated to Vri. Perhaps even more surprising is that REV-ERBα is dispensable for rhythmicity in mice, although it adds robustness and precision to the circadian clock. Posttranscriptional regulation of clock proteins in the first loop presumably compensates for the loss of rhythmic Bmal1 expression in the second loop in rev-erbα−/− mice. Posttranscriptional regulation of Clk protein also plays an important part in the Clk protein cycle. However, the magnitude of the period alterations in vri and Pdp1 heterozygous flies are comparable to those seen in mice homozygous for a rev-erbα knockout. Therefore, the Drosophila clock may rely more heavily on transcriptional control than the mammalian clock, especially in the second loop (Cyran, 2003).

Homologs of Vri and Pdp1 do exist in mammals and are even expressed with a circadian rhythm in pacemaker cells. However, genetic loss-of-function experiments suggest that none of the three mammalian Pdp1 homologs, either alone or in combination, affects the period length of circadian locomotor activity by more than 30 min. Similar loss-of-function experiments have yet to be performed for E4BP4, the mammalian homolog of Vri. The mammalian homologs of vri and Pdp1 may thus play only an ancillary regulatory role in the mammalian central clock, with their primary role being the regulation of rhythmic clock outputs (Cyran, 2003).

Tightly regulated and interconnected feedback loops are conserved in the circadian clocks of all the model organisms so far studied. A second interconnected loop adds robustness to oscillators. Two transcription loops also provide the potential for multiple inputs to the clock such as light, temperature, membrane potential, and redox state. Additionally, a second transcriptional loop provides a mechanism to regulate a novel phase of rhythmic expression of clock output genes. Such downstream genes presumably allow an organism to anticipate a constantly changing, but relatively predictable, environmental cycle, and adjust its behavior and physiology accordingly. The identification of downstream genes that link the molecular ticking of a central clock to changes in whole animal behavior and physiology is clearly the next major challenge in circadian biology (Cyran, 2003).

Drosophila Cryptochrome acts as a circadian transcriptional repressor to repress Clk and Cyc

Although most circadian clock components are conserved between Drosophila and mammals, the roles assigned to the Cryptochrome (Cry) proteins are very different: Drosophila Cry functions as a circadian photoreceptor, whereas mammalian Cry proteins (mCry1 and 2) are transcriptional repressors essential for molecular clock oscillations. This study demonstrates that Drosophila Cry also functions as a transcriptional repressor. RNA levels of genes directly activated by the transcription factors Clock (Clk) and Cycle (Cyc) are derepressed in cryb mutant eyes. Conversely, while overexpression of Cry and Period (Per) in the eye repressed Clk/Cyc activity, neither Per nor Cry repressed individually. Drosophila Cry also represses Clk/Cyc activity in cell culture. Repression by Cry appears confined to peripheral clocks, since neither cryb mutants nor overexpression of Per and Cry together in pacemaker neurons significantly affected molecular or behavioral rhythms. Increasing Clk/Cyc activity by removing two repressors, Per and Cry, leads to ectopic expression of the timeless clock gene, similar to overexpression of Clk itself. It is concluded that Drosophila Cry functions as a transcriptional repressor required for the oscillation of peripheral circadian clocks and for the correct specification of clock cells (Collins, 2006).

Several pieces of evidence point to Drosophila Cry, like its mammalian counterparts, functioning as a repressor of Clk/Cyc-activated transcription: (1) expression of four Clk/Cyc target genes is derepressed in cryb mutants; (2) overexpression of cry together with per is sufficient to repress tim and vri expression in the eye, and this is supported by Cry repressing Clk/Cyc-activated transcription in transfected cells, either alone or in conjunction with Per; and (3) removing both Cry and Per leads to ectopic tim expression in the brain (Collins, 2006).

Although Cry and Per seem to function together to repress Clk/Cyc activity, the results do not imply a direct interaction between Cry and Per proteins. Drosophila Cry-Per interactions have been detected in yeast, but Cry and Per appear to interact only via Tim in vivo. Furthermore, Per continues to repress Clk/Cyc activity in vivo during the first half of the day, presumably after Cry has been degraded by light. Thus, Cry and Per seem to control distinct steps in repression of Clk/Cyc activity, with Cry probably initiating, and Per maintaining, repression. Further experiments will be required to test whether Tim also facilitates repression. While in vitro studies indicated that Tim helps remove Clk/Cyc from DNA, in vivo studies of the timUL mutant suggests that Tim does not participate in repression of per and tim transcription and instead stabilizes Per and facilitates its nuclear entry. Given that Drosophila Tim interacts with both Per and Cry in vivo, it will be interesting to test whether the Per-Cry interactions detected in mammalian clock cells are mediated via mTim (Collins, 2006).

Very little is known about the developmental specification of clock neurons. Per and Cry normally restrict tim expression to cells that adopt a circadian cell fate. The results complement experiments in which overexpression of Clk led to ectopic tim expression, since they reveal that cells not normally destined to develop as clock cells repress Clk/Cyc activity during development. However, there must be additional factors that contribute to clock cell fate, since the ectopic Tim+ve cells in per01; cryb double mutant larvae did not produce PDP1. Similarly, there must be unidentified factors that maintain repression of tim in nonclock cells, since repression of Clk/Cyc activity will prevent further per expression. The presence of extra Tim-expressing cells may also explain the Tim-dependent rhythmic behavior of per01; cryb in LD cycles, since ectopic Tim expression influences LD behavior (Collins, 2006).

The findings that Cry functions as a repressor in Drosophila are supported by the high conservation across species of the “core” photolyase-like domain of Cry, which is sufficient for repression in Xenopus. TheDrosophila crym mutation removes most of the Cry C terminus and interferes with Cry's response to light. However, CryM still supports a functional clock in the eyes, suggesting that the remaining core of CryM functions as a transcriptional repressor (Collins, 2006).

Cry's homology with DNA photolyases has led to the suggestion that Cry was the original molecule that allowed organisms to respond to light -- primitive organisms could detect light and regulate gene expression with one molecule (Cry) to avoid damage by sunlight during light-sensitive processes such as DNA replication. While ancestral Cry may have acted as both a light sensor and repressor, non-Drosophilid insects such as the monarch butterfly Danaus plexippus have two cry genes and divide repressor/light sensor function between them. Thus, circadian clocks may well have their origins in rapid responses to light, and the anticipatory clock gene networks could have subsequently been built around Cry, a light-responsive protein and a transcriptional repressor, the function of which has gradually become specialized (Collins, 2006).

Targets of Activity

Drosophila Clock protein induces transcription of the circadian rhythm genes period and timeless. dClock functions as a heterodimer with a Drosophila homolog of BMAL1 termed Cycle. These proteins act through an E-box sequence in the period promoter. The timeless promoter contains an 18-base pair element encompassing an E-box, which is sufficient to confer Clock responsiveness to a reporter gene. Period and Timeless proteins block Clock's ability to transactivate their promoters via the E-box. Thus, Clock drives expression of period and timeless, which in turn inhibit Clock's activity and close the circadian loop. It is likely that either Per or Tim binds either Clock or Cycle, giving rise to a nonfunctional complex (Darlington, 1998).

The low Per and Tim levels in Jrk flies, mutant for Clock, could be due to reduced protein stability or to reduced protein synthesis in the mutant strains. To distinguish between these possibilities, PER and TIM mRNA levels were measured. Low and noncycling RNA levels were revealed, suggesting reduced synthesis rather than stability. Consistent with this notion, Jrk heterozygotes have a low amplitude of RNA cycling, which parallels the reduced amplitude of the protein rhythms and semidominance of the behavioral rhythm defect. To measure transcription rates directly, nuclear run-on assays were performed in homozygous Jrk flies. per and tim transcription rates are found to be temporally constant and approximately equal to the very low trough levels of wild-type flies. It is concluded that the behavioral arrhythmicity of Jrk mutants is largely due to a defect in the transcription of clock genes, including per and tim (Allada, 1998).

The basic region, involved in sequence-specific DNA contacts, is remarkably conserved between Drosophila and Mouse Clock proteins, with 11 out of 13 amino acids being identical; this suggests that the two proteins bind to similar if not identical DNA targets. In fact, 6 out of 9 amino acids are identical to a consensus generated for bHLH proteins that bind the CAC/GTG E box half-site, including the critical R residue at position 15; this is consistent with the dramatic effect of the Clk mutant on per E box-mediated transcription. As expected, the tim gene also has an E box in its 5' noncoding region. In-vitro experiments indicate that human Clock preferentially binds and activates transcription from DNA targets very similar to the Drosophila per E box (Allada, 1998).

The period (per) gene is an essential component of the circadian timekeeping mechanism in Drosophila. This gene is expressed in a circadian manner, giving rise to a protein that feeds-back to regulate its own transcription. A 69 bp clock regulatory sequence (CRS) has been identified previously, upstream of the period gene. Drosophila Clock and Cycle encode proteins that activate per and tim transcription via E-boxes located in per and tim upstream sequences. One of the E-boxes targeted by dCLK and CYC is located within the 69 bp CRS upstream of per; this CRS is required for rhythmic transcription. The CRS confers wild-type mRNA cycling when used to drive a lacZ reporter gene in transgenic flies. To determine whether the CRS also mediates proper developmental and spatial expression and behavioral rescue, the ability of CRS to drive either lacZ or per was tested in transgenic flies. The results show that the CRS is able to activate expression in pacemaker neuron precursors in larvae and essentially all tissues that normally express per in pupae and adults. The CRS is sufficient to rescue circadian feedback loop function and behavioral rhythms in per01 flies. However, the period of locomotor activity rhythms shortens if a stronger basal promoter is used. This study shows that regulatory elements sufficient for clock-dependent and tissue-specific per expression in larvae, pupae, and adults are present in the CRS and that the period of adult locomotor activity rhythms is dependent, in part, on the overall level of per transcripts (Hao, 1999).

Vrille is a bZIP transcription factor related to other leucine zipper proteins found in all vertebrate species. vrille plays a role in two complex Drosophila pathways. Initially described as a maternal effect lethal mutation, various alleles produce impaired zygotic development resulting in ventralization. vrille was later found in a screen for genes whose expression is controlled by the biological clock of adult flies.Since the timing and strength of the VRILLE mRNA oscillations are almost identical to those of timeless, it seemed possible that vri and tim transcription would be regulated by the same transcription factors. VRI mRNA levels were tested in adult heads of ClkJrk and cyc0 mutant flies, which show constitutively low transcription of the per and tim genes. VRI mRNA, like TIM mRNA, is produced at low levels in these mutants. Like TIM, VRI mRNA levels are high or intermediate in per01 and tim01 mutants, indicating that VRI levels are linked to the activities of period, tim, Clock, and cycle. The correspondence of vri, per, and tim regulation in wild-type and clock mutants suggests that the dCLK/Cyc complex might directly regulate vri transcription (Blau, 1999).

The vri promoter sequence was searched and four CACGTG motifs, that are potential Clk/Cyc-binding sites, were found. One of these is closely related to the functional Clk/Cyc-binding site in the per promoter. vri promoter sequence (2.8 kb), including all four sites, was fused to a luciferase reporter and transfected into Drosophila S2 cells either with or without an expression vector for Clk. This assay has been used to show that Clk can activate the per and tim promoters in cultured cells, and it relies on the endogenous production of Cyc in S2 cells. The vri, per, and tim promoters were all strongly activated by Clk (658-, 49-, and 271-fold, respectively). To test whether the most conserved of the potential Clk/Cyc-binding sites in the vri promoter is sufficient for activation by this complex, reporter constructs composed of four copies of either this wild-type sequence, or a mutant sequence in which the central CG is reversed, upstream of a basal promoter and a luciferase reporter gene, were generated. The wild-type vri sequence is strongly activated (190-fold) by Clk, while the mutant E box is activated only 3-fold. Therefore, Clk can bind and activate the vri promoter. A promoter fragment containing the remaining CACGTG sites also responds to Clk expression, suggesting further Clk/Cyc binding to one or more of these sites. The results indicate that the transcriptional loop that causes per and tim transcription to cycle with a 24 hr period also regulates other genes (Blau, 1999).

takeout (to) is a Drosophila circadian clock-regulated output gene, a transcriptional target of the central clock. The Takeout amino acid sequence shows similarity to two ligand binding proteins, including juvenile hormone binding protein. Takeout mRNA is expressed in the head and the cardia, crop, and antennae - structures related to feeding. to expression is induced by starvation, which is blocked in all arrhythmic central clock mutants, suggesting a direct molecular link between the circadian clock and the feeding/starvation response. A takeout mutant has aberrant locomotor activity and dies rapidly in response to starvation, indicating a link between locomotor activity, survival, and food status. It is proposed that takeout participates in a novel circadian pathway target that conveys temporal and food status information to feeding-relevant metabolisms and activities (Sarov-Blat, 2000).

TO mRNA expression is down-regulated in cyc01 flies and in all other circadian mutants tested. Its level is undetectable in cyc01 and Clkjrk mutants, as measured by RNase protection and Northern blotting. In contrast, there is detectable TO mRNA in all other genotypes tested, though it is substantially lower than that in wild-type flies. Since there is little or no functional CLK-CYC heterodimer in the cyc01 and Clkjrk backgrounds, the simplest way to explain this observation is that to is directly regulated by CLK and CYC. The higher to transcription in per01, tim01, and per01;tim01 double-mutant flies is presumably due to residual functional CLK-CYC heterodimer in these backgrounds. per01 flies reproducibly show a higher level of to expression than tim01, indicating that Per and Tim may differentially regulate to expression. However, the mechanism underlying this difference is still unknown. When mRNA levels at different time points are measured, to does not show a significant cycling pattern in the clock mutants tested (So, 2000).

The to promoter sequence reveals a remarkable sequence identity with the E-box region of the per and tim promoters. In particular, there is a 9-bp sequence identity around this E-box sequence. The other E-box sequences known in circadian genes usually share the 6-bp core sequence or the core sequence with an additional A (CACGTGA), which has been shown to be strongly preferred by the mammalian BMAL1-MOP4 bHLH-PAS transcription factor heterodimer. In fact, the to and per promoters share 13 out of the 18 bp shown to be sufficient to drive transcriptional activation in S2 cells. This is also consistent with the fact that TO mRNA is undetectable in cyc01 and Clkjrk mutants, suggesting that Clk-Cyc regulates to transcription directly. This would be similar to per and tim transcriptional regulation, despite the phase difference. Neverless, extensive investigatio of the to promoter suggests that to transcription requires factors other than Clk and Cyc (So, 2000).

Transcriptional regulation plays an important role in Drosophila circadian rhythms. The period promoter has been well studied, but the timeless promoter has not been analyzed in detail. Mutagenesis of the canonical E box in the timeless promoter reduces but does not eliminate timeless mRNA cycling or locomotor activity rhythms. This is because there are at least two other cis-acting elements close to the canonical E box, which can also be transactivated by the circadian transcription factor Clock. These E-box-like sequences cooperate with the canonical E-box element to promote high-amplitude transcription, which is necessary for wild-type rhythmicity (McDonald, 2001).

Noncanonical E-box-dependent transcription has been described extensively for the Myc family of binding proteins, whose canonical high-affinity binding site has been determined as CACGTG by sequential selection and amplification of binding sites. These studies have led to the identification of lower-affinity, noncanonical MYC-MAX binding sites, such as CATGTG, CACGCG, CATGCG, CACGAC, and CAACGTG. MYC-MAX dimers are able to bind a similar set of sequences in vivo, in a tumorigenic cell line. A similar observation has been reported for mCLOCK/CYC. Four E boxes have been identified in the first and second introns of the mammalian dbp gene (coding for PAR leucine zipper transcription factor family member that binds to mammalian Per promoter), whose gene product is important in generating the cycling transcription of several circadian genes in the liver. All four E boxes were shown to activate a luciferase reporter in cell culture assays upon transfection with mCLOCK and BMAL1, but only two of the E-box regions show circadian differences in DNase I-hypersensitive sites: one with a canonical CACGTG motif and the other with a noncanonical CACATG motif. This study provides definitive evidence that a noncanonical E box contributes to circadian transcription in Drosophila (McDonald, 2001).

The Drosophila circadian oscillator consists of interlocked period/timeless and Clock (Clk) transcriptional/translational feedback loops. Within these feedback loops, Clk and Cycle (Cyc) activate per and tim transcription at the same time as they repress Clk transcription, thus controlling the opposite cycling phases of these transcripts. Clk-Cyc directly bind E box elements to activate transcription, but the mechanism of Clk-Cyc-dependent repression is not known. This study shows that Clk-Cyc-activated gene, vrille (vri), encodes a repressor of Clk transcription, thereby identifying vri as a key negative component of the Clk feedback loop in Drosophila's circadian oscillator. The blue light photoreceptor encoding cryptochrome (cry) gene is also a target for Vri repression, suggesting a broader role for Vri in the rhythmic repression of output genes that cycle in phase with Clk (Glossop, 2003).

Overexpression of Vri in larval oscillator cells leads to low/absent levels of per and tim mRNA. The simplest explanation of this result is that Vri binds per and tim regulatory elements and represses transcription directly. However, given that per and tim share a common mode of activation, by CLK-CYC, low levels of per and tim mRNA would also result if Vri repressed Clk. Given these alternatives, it is of paramount importance to determine the daily phase of cycling, since Vri, which acts as a repressor, is predicted to cycle in the opposite phase as the mRNAs of the gene(s) it represses (Glossop, 2003).

Polyclonal antisera were generated against full-length Vri and used for Western blot analysis. In wild-type flies, two Vri-specific bands are detected: a weak band of ~98 kDa and a strong broad band of ~82-89 kDa. Phosphatase treatment reduces the molecular weight of the weak band to ~95 kDa and collapses the strong broad band to a tight band of ~80 kDa, which matches the predicted molecular weight of Vri. As expected, Vri levels are low in cyc01 mutant flies since vri mRNA expression is CLK-CYC dependent. In addition, Vri levels are lower at ZT 3 than at ZT 15 in wild-type flies, suggesting that Vri levels cycle (Glossop, 2003).

To more precisely determine the phase of Vri cycling, Vri was measured in wild-type flies collected every 2 hr in LD. Vri levels peak during the early night (ZT 13-17) and reach trough levels during the early day (ZT 01-05;. A similar cycling phase for Vri persists in constant darkness. This cycling profile closely follows that of per, tim, and vri mRNA transcripts and is opposite that of Clk mRNA, which peaks between ZT 22 and 04 and falls to low levels between ZT 10 and 16. Vri protein therefore cycles in opposite phase to Clk mRNA; this supports a role for Vri as the Clk repressor (Glossop, 2003).

Transcription from the Clk gene is initiated at two sites: a minor site, which has been detected only through 5' RACE and RT-PCR, and a major start site that accounts for the vast majority of Clk transcription in heads. Based on the available Clk cDNAs, the minor transcript includes an additional 5' exon, plus the entire major transcript due to the removal of a ~5 kb intron between the exons initiated at the minor (first) and major (second) transcription start sites. To identify a region that mediates Clk circadian transcription, a genomic DNA fragment from -8000 to +40 (the major Clk transcription start = +1) was used to drive Gal4 mRNA transcription in transgenic flies containing a functional clock. This fragment mediates rhythmic Gal4 mRNA transcription having the same phase and amplitude as Clk transcripts in wild-type flies, thus demonstrating that all the necessary circadian regulatory elements are present (Glossop, 2003).

For Vri to be a direct repressor of Clk, the 8.0 kb Clk genomic fragment should contain binding sites for Vri. The consensus binding site for Vri has not been determined; however, the basic (DNA binding) domain shares >85% homology with that of mammalian E4BP4, and hence, Vri should bind similar DNA sequences as E4BP4. The optimal binding site for E4BP4 has been determined as 5'-(A/G)TTAC: (A/G)T(A/C)A(A/T/C)-3'. A search for E4BP4-consensus sites at the Clk locus identified multiple sites on both DNA strands upstream of the major transcription start site. This density of E4BP4 sites is much higher than predicted by chance. Searches of the per and tim loci (i.e., 4 kb upstream through intron 1) identified no 10 E4BP4 sites. This evidence is in line with the number of sites predicted by chance. The lack of E4BP4 binding sites and the phase of Vri cycling are inconsistent with Vri directly repressing per and tim, but do support the possibility that Vri directly represses Clk (Glossop, 2003).

Thus the b-ZIP transcription factor Vri feeds back to control circadian transcription of Clk within the oscillator mechanism. Clk-Cyc heterodimers activate per and tim transcription at the same time that they repress Clk transcription, thus providing a bidirectional switch that mediates the opposing cycling phases of these transcripts within the circadian oscillator. Since there are no canonical (CACGTG) E boxes within the region known to mediate Clk mRNA cycling, the simplest interpretation of these results is that Clk-Cyc also activate an intermediate that feeds back to repress Clk transcription. A model has been developed to explain this regulatory mechanism: it proposes that Vri functions to repress Clk transcription. Positive drive by Clk-Cyc and negative drive by Per and Tim confer vri mRNA and protein rhythms, and rhythms in Vri accumulation, in turn, mediate the rhythmic repression of Clk (Glossop, 2003).

Several lines of evidence support this model. (1) vri transcription is activated by Clk-Cyc via E box elements in its upstream sequence, showing that vri could act as a Clk-Cyc-dependent intermediate factor. (2) Vri overexpression leads to the repression of per and tim. Since per and tim both rely upon Clk for their activation, Vri-dependent inhibition of Clk could readily account for their coordinate repression. That Vri acts as a repressor within the oscillator mechanism is further supported by genetic analysis: Vri overexpression leads to long period activity rhythms (via reductions in per and tim expression), and reduced vri copy number leads to short period activity rhythms (presumably through increases in per and tim expression). (3) vri mRNA and protein cycle in the same phase. With this phase relationship, Vri accumulates to high levels as Clk mRNA drops to low levels. After a substantial (~6 hr) delay, Per and Tim inhibit Clk-Cyc activation of vri, consequently reducing Vri to low levels as Clk mRNA accumulates to high levels. Once Per and Tim are degraded, the next cycle of vri expression is initiated. (4) Vri binds to sequence elements within the Clk circadian control region in vitro, which suggests that Vri action is direct. (5) Vri overexpression reduces Clk mRNA levels in vivo. This Vri-dependent repression preferentially affects the peak levels of Clk mRNA in wild-type animals, which suggests that normal peak levels of Vri are maximally active, since additional Vri cannot further reduce trough levels of Clk. Vri also represses the peak levels of Clk mRNA present in cyc01 animals. Since cyc01 flies lack a functional feedback mechanism due to the absence of Cyc, Per, Tim, and Vri, this result indicates that Vri represses Clk transcription directly in vivo rather than through other components of the feedback loop. The repression of Clk by Vri indicates that rhythmic Clk transcription occurs through the circadian repression of Clk. This is different from the situation with per, tim, and vri, where circadian transcription is mediated by rhythmic Clk-Cyc-dependent activation and PER-TIM-dependent repression (Glossop, 2003).

Vri overexpression in wild-type or cyc01 flies produces more Vri than the wild-type peak, yet Clk and cry mRNAs are not fully repressed to wild-type trough levels. The inability of high Vri levels to fully repress Clk and cry suggests that Vri may act in concert with another factor to repress transcription. Constant high levels of Vri do not fully repress Clk and cry expression in wild-type flies during the late evening and early morning. This temporal difference in the ability of Vri to repress implies that another repressor is present in limiting amounts at these times of the circadian cycle, i.e., a second rhythmically expressed repressor. Likewise, the inability of high levels of Vri to repress Clk completely in cyc01 flies suggests that the complimentary repressor is only present in limiting amounts in this genotype. Given that Vri is a bZIP transcription factor, it is tempting to speculate that Vri forms a heterodimer with another bZIP transcription factor to fully repress Clk and cry transcription. However, no such factor has been identified and it is also possible that another Clk and cry repressor acts independently of Vri. Although Vri alone cannot fully repress Clk, the ability of Vri to repress Clk activation by two-thirds indicates that Vri is the major Clk repressor (Glossop, 2003).

With the identification of Vri as a repressor of Clk transcription, it is apparent that Clk and Cyc function to activate a set of repressors that act on different targets at different times in the circadian cycle. Activation of vri by Clk-Cyc leads to the immediate production of Vri, which represses Clk transcription and consequently reduces the levels of Clk (and necessarily Clk-Cyc). Activation of per and tim by Clk-Cyc leads to the delayed accumulation of Per-Tim heterodimers. This delayed PER-TIM accumulation allows Vri to repress Clk from midday to early evening (ZT4 to ZT16) and inhibits the ability of newly generated Clk to activate per and tim expression until early morning (ZT4). This difference in accumulation between Vri and Per/Tim is therefore critical for controlling the opposite cycling phases of Clk and per/tim/vri within the interlocked feedback loop mechanism (Glossop, 2003).

The mammalian circadian oscillator is also comprised of interacting transcriptional/translational feedback loops that share many components with their Drosophila counterparts. In mammals, the mPer/mCry feedback loop is analogous to the per/tim feedback loop in flies in that CLOCK and BMAL1 (a Cyc homolog) activate transcription of the mPers (i.e., mPer1, mPer2, and mPer3) and mCrys (i.e., mCry1 and mCry2), which feed back to inhibit CLOCK-BMAL1 activation. Likewise, the mammalian Bmal1 loop is analogous to the Clk loop in flies: BMAL1 and CLOCK lead to the repression of Bmal1 transcription and the mPERs and mCRYs activate Bmal1 transcription. As in flies, CLOCK-BMAL1-dependent activation occurs via E box binding, but CLOCK-BMAL1-dependent repression has not been characterized (Glossop, 2003).

Kadener, S., Stoleru, D., McDonald, M., Nawathean, P. and Rosbash. M. (2007). Clockwork Orange is a transcriptional repressor and a new Drosophila circadian pacemaker component. Genes Dev. 21(13): 1675-86. Medline abstract: 17578907

Lim, C., Chung, B. Y., Pitman, J. L., McGill, J. J., Pradhan, S., Lee, J., Keegan, K. P., Choe, J. and Allada, R. (2007). Clockwork orange encodes a transcriptional repressor important for circadian-clock amplitude in Drosophila. Curr. Biol. 17(12): 1082-9. Medline abstract: 17555964

Matsumoto, A., et al. (2007). A functional genomics strategy reveals clockwork orange as a transcriptional regulator in the Drosophila circadian clock. Genes Dev. 21(13): 1687-700. Medline abstract: 17578908

Clockwork Orange is a transcriptional repressor and a new Drosophila circadian pacemaker component

Many organisms use circadian clocks to keep temporal order and anticipate daily environmental changes. In Drosophila, the master clock gene Clock promotes the transcription of several key target genes. Two of these gene products, Per and Tim, repress Clk-Cyc-mediated transcription. To recognize additional direct Clk target genes, a genome-wide approach was designed and clockwork orange (cwo) was identified as a new core clock component. cwo encodes a transcriptional repressor functioning downstream of Clk that synergizes with Per and inhibits Clk-mediated activation. Consistent with this function, the mRNA profiles of Clk direct target genes in cwo mutant flies manifest high trough values and low amplitude oscillations. Impaired activity of Cwo leads to an elevated trough of per, tim, vri, and Pdp1mRNA at ZT3 (three hours into the morning) in cwo RNAi transgenic flies compared with those of wild-type flies. Because behavioral rhythmicity fails to persist in constant darkness (DD) with little or no effect on average mRNA levels in flies lacking cwo, transcriptional oscillation amplitude appears to be linked to rhythmicity. Moreover, the mutant flies are long period, consistent with the late repression indicated by the RNA profiles. These findings suggest that Cwo acts preferentially in the late night to help terminate Clk-Cyc-mediated transcription of direct target genes including cwo itself. The presence of mammalian homologs with circadian expression features (Dec1 and Dec2) suggests that a similar feedback mechanism exists in mammalian clocks (Kadener, 2007). To other studies similarly identified Clockwork orange an a transcriptional repressor that inhibits Clk-mediated activation (Matsumoto, 2007; Lim, 2007).

Protein Interactions: Cyc and Clk heterodimerization

A searched for novel Drosophila circadian rhythm genes was carried out by chemically mutagenizing flies and screening for altered or aberrant locomotor activity rhythms. A recessive arrhythmic mutant fly line was identified, which has been named cycle (cyc0). These flies show arrhythmic locomotor activity patterns when they carry two mutant third chromosomes. cyc0/cyc0 flies also manifest arrhythmic eclosion. Cyc is a potential dimerization partner for Clock. The homozygous mutant flies are always arrhythmic, regardless of the per genetic background. Unlike the semidominant Clk mutant, heterozygous cyc0/+ flies manifest robust rhythms with no hint of arrhythmicity. But they have altered periods, with rhythms 1 hr longer than their per genotype would otherwise dictate. This indicates that either cyc has a dominant phenotype or that the locus has dosage sensitive effects on period. The data suggest that cyc might identify a circadian clock component (Rutila, 1998).

Homozygous mutant cyc0/cyc0 flies also have difficulties under 12:12 light/dark (LD) conditions, since only 64% of them give discernable rhythms by chi2 analysis. Greater than 90% of arrhythmic per0 and tim0 flies are light responsive and therefore show robust 24 hr rhythms under these conditions. Since the cyc0/cyc0 flies show no visual difficulties in optomotor behavior, the data suggest that the flies are specifically impaired in circadian light perception. A similar phenotype is seen with Clk mutant flies (Rutila, 1998).

To examine clock function more directly, the fluctuations of the clock proteins Period and Timeless were examined in wild-type, heterozygous, and homozygous cyc flies under LD conditions. Western analysis with an anti-Per antibody reveals very little protein in cyc0/cyc0 fly heads at any time of day. As predicted from the robust rhythms, cyc0/+ heterozygotes show normal Per cycling, with normal levels and a normal temporal phosphorylation program. For all genotypes, similar results were obtained for Tim. The low Per and Tim levels could be due to reduced protein stability or to reduced protein synthesis in the homozygous mutant strain. To distinguish between these possibilities, per and tim mRNA levels were measured. Low RNA levels and little or no cycling are found in the cyc0/cyc0 head extracts, suggesting reduced synthesis rather than reduced stability. As expected, RNA levels and cycling in cyc0/+ heterozygotes are indistinguishable from those in wild-type extracts (Rutila, 1998).

The cyc effect on per and tim RNA levels and cycling could be transcriptional or posttranscriptional. To directly measure transcription rates, nuclear run-on assays were performed in homozygous cyc flies. In this genotype, per and tim transcription rates show no evidence of cycling and are approximately equal to the very low trough levels of wild-type flies observed at ZT1. The result is essentially identical to that observed in homozygous Clk flies (Allada, 1998). It is interesting to note that the transcription rates in both genotypes are much lower than those observed in either per0 or tim0 head extracts, which also manifest little or no transcriptional oscillations. This suggests that cyc as well as Clk are epistatic to and upstream of per0 and tim0. Taken together, the data suggest that cyc, like Clk, affects the transcription of the clock genes per and tim (Rutila, 1998).

To identify specific sequence elements mediating the mutant effects on transcription, the effects of the cyc mutation were examined on a minimal per promoter element. This 69 bp enhancer contains a critical per-derived E box and drives rhythmic expression of a reporter gene (lacZ). To this end, the E box/lacZ construct was crossed into cyc0/cyc0 mutant flies and lacZ RNA levels were assayed for cycling by RNase protection. The results are dramatic and indicate that there is little or no cycling lacZ RNA transcription in the homozygous mutant flies, suggesting that Cyc affects the transcriptional activity of the per circadian transcriptional enhancer. These features of Cyc are similar to those of Clk, which probably binds to and activates transcription at the per E box (Rutila, 1998).

The mammalian protein BMAL1 was cloned as an "orphan" protein of the bHLH-PAS transcription factor family with no known biological function (Ikeda, 1997 and Hogenesch, 1997). However, there are recent biochemical experiments indicating that it may play a role in circadian rhythm-relevant transcription in mammals: it can act as a heterodimeric partner of mouse Clock (mClock) in DNA binding and transcriptional activation. The BMAL1:mClock heterodimer selects a DNA-binding sequence that resembles the critical E box sequence within the cycling element in the Drosophila per upstream region, and there are features of this enhancer in addition to the central CACGTG hexamer that provide specificity for the BMAL1:mCLOCK heterodimer. As transcripts from mouse per genes undergo circadian oscillations in level, these genes may contain a similar target cycling element to that of Drosophila per. The BMAL1:mCLOCK heterodimer could be the heterodimeric factor that binds to this cycling element and activates clock-relevant transcription (Hogenesch, 1998 and Rutila, 1998 and references).

By analogy, it is proposed that Cyc and Clk heterodimerize, bind to Drosophila clock gene E boxes, and function to drive the circadian-regulated transcription of these genes. This makes cyc and Clk the first Drosophila circadian rhythm genes with a known biochemical role and a defined place in the clock circuit; it also places them upstream of per and tim. RNA and transcription experiments in cyc0 and in Clk mutant flies (Allada, 1998) fully support such an assignment (Rutila, 1998).

Cyc is approximately half as big as Clk; the difference appears to be largely the extensive glutamine-rich C-terminal half of Clk (Allada, 1998). This may indicate that Clk brings the transcriptional activation domain to the complex. The dominant phenotype of the Clk mutant and the elimination of the Q-rich region by the mutation (Allada, 1998) are consistent with this notion. The mutant protein would then be able to dimerize with Cyc and bind DNA but would be unable to activate transcription. This would explain its recessive as well as its dominant features (Rutila, 1998).

Differential regulation of circadian pacemaker output by separate clock genes in Drosophila

Regulation of the Drosophila pigment-dispersing factor (pdf) gene products was analyzed in wild-type and clock mutants. Mutations in the transcription factors Clock and Cycle severely diminish pdf RNA and neuropeptide (PDF) levels in a single cluster of clock-gene-expressing brain cells, called small ventrolateral neurons (s-LNvs). This clock-gene regulation of specific cells does not operate through an E-box found within pdf regulatory sequences. PDF immunoreactivity exhibits daily cycling, but only within terminals of axons projecting from the s-LNvs. This posttranslational rhythm is eliminated by period or timeless null mutations, which do not affect PDF staining in cell bodies or pdf mRNA levels. Therefore, within these chronobiologically important neurons, separate elements of the central pacemaking machinery regulate pdf or its product in novel and different ways. Coupled with contemporary results showing a pdf-null mutant to be severely defective in its behavioral rhythmicity, the present results reveal PDF as an important circadian mediator whose expression and function are downstream of the clockworks (Park, 2000).

To assess the effects of clock mutations on pdf expression, the normal cellular distribution of the Drosophila gene's native products were examined. By in situ hybridization, the expression pattern of pdf mRNA has been shown to be similar to that determined with anti-crab-PDH. There are four positive cells in each larval brain hemisphere; these persist into adulthood and become the small ventrolateral neurons (s-LNvs), whose neurites project into a dorsal region of the adult brain. Four large ventrolateral neurons (l-LNvs) also express pdf; these emerge during metamorphosis and send projections into the optic lobe and across the brain midline. Larvae and adults also contain pdf mRNA in the posterior extremity of the CNS (Park, 2000).

Northern blots reveal no daily rhythm of pdf mRNA abundance, but they could have failed to detect pdf mRNA cycling in a subset of the cells. Thus temporal in situ hybridizations were performed; neither category of pdf-expressing neurons exhibit systematic fluctuations in signal intensities. Therefore, there is no pdf mRNA rhythm for clock mutations to affect (Park, 2000).

Anti-Drosophila PDF antibodies give cell labeling identical to that obtained by in situ hybridization. Neither method leads to marking of cells in the dorsal brain of adults that are stained by anti-crab-PDH. This indicates that the dorsally located antigen is cross-reacting material and does not have to be considered in terms of effects of clock mutations on pdf expression (Park, 2000).

Expression of pdf in the arrhythmic ClkJrk mutant has been found to be strikingly abnormal. In ClkJrk brains, neither pdf mRNA nor PDF is detectable in larval LN cells and in the s-LNvs of adults. The same defects were observed in mutant animals heterozygous for ClkJrk and a deletion of the locus. These results suggest that Clk is required for pdf transcription, although only in certain cells: the larval LNs and the s-LNvs into which they develop. Dorsally projecting axonal processes arising from the s-LNv cells terminate near the calyx of the dorsal-brain mushroom body. In accord with the absence of perikaryal s-LNv immunoreactivity, these projections are absent from ClkJrk brains. In contrast, expression in the l-LNvs and abdominal-ganglionic cells of adults is apparently unaffected by ClkJrk and D); this includes normal staining of centrifugal and interhemispheric projections within the fly's head. However, certain features of projections from l-LNv cells are aberrant in ClkJrk. Approximately 50% of the mutant brains showed abnormal projections; in others, one or two axons from this region project further and irregularly toward a dorsal or median region of the brain. None of these projections is similar to the more dorsal-reaching projections in the brains of wild-type adults (Park, 2000).

Because the Cyc protein cooperates with Clk in their transcriptional-activation roles, pdf expression was examined in cycle mutants. The effects were similar to but less severe than those caused by ClkJrk. Most of the larval LNs homozygous for either of two cyc0 mutations show much weaker expression of both mRNA and peptide, as compared with wild type, but the mutant expression levels are variable even within a single brain hemisphere: some cells contained signal, whereas others are extremely difficult to detect. The numbers of antibody-stained s-LNvs in cyc0 adults are well above zero, compared with the elimination of such signals in ClkJrk flies. Numbers of l-LNv cells in the brains of cyc-mutated adults are normal, similar to the results obtained in the ClkJrk background. About 25% of the adult cyc0 brains exhibit an abnormal dorsal projection. In approximately 30% of these mutant specimens, the projections are asymmetric within a single individual: one hemisphere can contain a bundle of dorsally projected axons; but in the contralateral hemisphere, only one or two axons project into the dorsal brain. In other cyc0 adults, axons project irregularly into a median brain region (Park, 2000).

The major conclusions from examining pdf expression in the Clock and cycle mutants are that (1) both genes appear to be positive regulators of pdf RNA levels but only in the s-LNvs and their larval precursors; (2) the effects of ClkJrk are stronger than those of the cyc0 mutations; and (3) there are developmental defects, because PDF-containing processes in the adult CNS are aberrant in both types of mutants (Park, 2000).

Do Clk and Cyc activate pdf transcription directly? If that is the case, there could be an E-box in this gene's regulatory region. Indeed, within a 2.4-kb segment 5' to the pdf ORF a CACGTG sequence ~1.4 kb upstream of the transcription-start site has been found. The 2.4-kb DNA fragment was fused to the (yeast) GAL4 gene; transgenic strains were generated and crossed to flies carrying UAS-lacZ. The doubly transgenic progeny show faithful beta-galactosidase-reported expression of pdf. To determine whether the E-box is important for the pdf's transcriptional activation, further transgenics were generated. Deletions missing either half or all of the E-box are sufficient to drive brain expression indistinguishable from that observed in wild type. Interestingly, the smallest 5'-flanking region examined mediates the normal brain pattern but does not lead to abdominal-ganglionic expression in the larval CNS. That the influences on pdf expression of Clock and cycle do not operate through a circadian E-box, and thus seem to be indirect, is consistent with the lack of pdf mRNA cycling and Clock/cycle-independent expression in the l-LNv cells (Park, 2000).

Protein Interactions: Clk interaction with Per and Tim

Drosophila Clock protein (dClock) is a transcription factor that is required for the expression of the circadian clock genes period (per) and timeless (tim). dClock undergoes circadian fluctuations in abundance, is phosphorylated throughout a daily cycle, and interacts with Per, Tim, and/or the Per-Tim complex during the night but not during most of the day. Both Per and Tim copurify with dClock in a time-of-day-specific manner: Per and Tim are first detected at ZT12 (beginning of the dark period), followed by increases in amounts that reach peak values at ZT23.9 (just before the lights go on). Between ZT16 (a third of the way through lights off) and ZT23.9, the amounts of all three proteins in immune complexes increase, even though the total levels of Tim and Per in head extracts peak at ZT16 and ZT20, respectively. This suggests that during the night dClock is present in limiting amounts compared to Per and Tim. Despite the higher levels of immunoprecipitated dClock between ZT4 and ZT8 compared to values obtained between ZT12 and ZT16, very little, if any, Per and Tim are detected. A likely explanation for this is that between ZT4 and ZT8 the total levels of Per and especially those of Tim are at, or close to, trough values. Thus, the interaction of Per and Tim with dClock is mainly restricted to nighttime hours (Lee, 1998).

Analysis of immune complexes derived from a period mutant clearly indicate that in the absence of Per, Tim can still interact with dClock. Because Tim is apparently located exclusively in the cytoplasm in the absence of Per, this result could suggest that the nuclear localization of dClock also requires Per or a functional oscillator. Alternatively, low levels of Tim might be able to enter the nucleus in the absence of Per. In contrast, several attempts to visualize a specific interaction between Per and dClock in the absence of Tim were unsuccessful. There are at least two nonmutually exclusive reasons that might account for thr inability to detect Per in dClock-containing immune complexes prepared from tim mutant flies: (1) the levels of Per are very low in tim mutant flies and as such the amounts of Per that copurify with dClock are below the detection limit, and (2) the interaction of Per with dClock requires Tim, possibly via formation of the Per-Tim complex and/or a dependence for nuclear localization (Lee, 1998 and references).

Attempts were made to measure the relative amounts of dClock that interact with Per and Tim as a function of time in an LD cycle. Head extracts were incubated with antibodies against either Per or Tim, and immune complexes probed for dClock, Per, and Tim. At ZT20 almost identical levels of dClock copurify with antibodies directed against either Per or Tim. Equivalent amounts of Per were also present in both immune pellets, but 1.6-fold more Tim is immunoprecipitated with antibodies to Tim, as compared to those directed against Per. These results are almost identical with a previous study showing that (1) in head extracts prepared from flies collected at ZT20, 80% of the total amount of Per is bound to Tim in a 1:1 stoichiometric relationship, and (2) there is 1.5-1.8 times more Tim, as compared to Per. Thus, the current results suggest that at ZT20 the majority of the Per and Tim proteins that interact with dClock are in the form of a heterodimeric Per-Tim complex. During the early day, only low levels of dClock are detected in immune complexes obtained using either antibodies to Per or Tim, in agreement with results using anti-dClock antibodies. Furthermore, it is mainly versions of Per and Tim that are essentially free of one another that interact with dClock during the early day (Lee, 1998).

How might a trimeric complex containing Per, Tim, and dClock be assembled? Presumably the HLH domain of dClock does not participate in mediating protein-protein interactions in this putative trimeric complex, because neither Per nor Tim seems to have a similar dimerization region. The only other regions that have been shown to mediate protein-protein interactions are the PAS domain found in Per and dClock and a not so well characterized region in Tim that spans 400 amino acids and interacts with the PAS domain of Per. It is tempting to speculate that one or both of these domains has the capacity to engage in at least trimeric formation. Although these studies do not address the nature of the trimeric interaction, they indicate that PAS-containing proteins are not limited to binary interactions (Lee, 1998).

These results suggest that Per and Tim participate in transcriptional autoinhibition by physically interacting with dClock or a dClock-containing complex. Nevertheless, in the absence of Per or Tim, the levels of dClock are constitutively low, indicating that Per and Tim also act as positive elements in the feedback loop by stimulating the production of dClock. Although Per and Tim inhibit dClock activity, Per and Tim are required for the high-level production of dClock protein and mRNA. Thus, Per and Tim appear to be the main "motor" of the Drosophila circadian oscillator, driving both positive and negative elements of the transcriptional-translational feedback loop. These observations suggest an explanation for the previously unexplained finding that the levels of Per mRNA in per mutant flies are approximately half as high as those obtained at peak times in wild-type flies. In contrast, mutations that abolish Neurospora FRQ activity result in high levels of frq RNA, suggesting that the frq-based circadian oscillator in Neurospora is based on a more simple negative transcriptional feedback loop. How Per and Tim stimulate dClock expression is not clear. They may interact with other transcription factors and act as coactivators. Alternatively, they may block the function of negative factors leading to the stimulation of gene expression. In addition to regulating the transcriptional activity of the dClock-CYC complex, Per and Tim might also interact with other transcription factors that are not involved in the circadian oscillator and as such molecularly couple the timekeeping mechanism to downstream effector pathways (Lee, 1998).

A novel C-terminal domain of Drosophila PERIOD inhibits dCLOCK:CYCLE-mediated transcription

The essence of the Drosophila circadian clock involves an autoregulatory feedback loop in which Period (Per) and Timeless (Tim) inhibit their own transcription by association with the transcriptional activators Clock (Clk) and Cycle (Cyc). Because Per, Clk, and Cyc each contain a PAS domain, it has been assumed that these interaction domains are important for negative feedback. However, a critical role for PAS-PAS interactions in Drosophila clock function has not been shown. Nuclear transport of Per is also believed to be an essential regulatory step for negative feedback, but this has not been directly tested, and the relevant nuclear localization sequence (NLS) has not been functionally mapped. These critical aspects of Per-mediated transcriptional inhibition have been evaluated in Drosophila Schneider 2 (S2) cells. The dCLK:CYC inhibition domain (CCID) of Per has been mapped; it lies in the C terminus, downstream of the PAS domain. Using deletion mutants and site-directed mutagenesis, a novel NLS has been identified in the CCID of Per that is a potent regulator of Per's nuclear transport in S2 cells. Nuclear transport, primarily through this novel NLS, is essential for the inhibitory activity of Per. The data indicate that nuclear Per inhibits Clk:Cyc-mediated transcription through a novel domain that additionally contains a potent NLS (Chang, 2003).

Thus, a key step in the Drosophila circadian negative feedback loop, Per inhibition of Clk:Cyc transcription, is not mediated by the PAS domain of Per. Instead, the previously uncharted C terminus of Per contains a novel domain (CCID; aa 764-1034) responsible for transcriptional inhibitory activity. The functional importance of this C-terminal region of Per is corroborated by an earlier in vivo experiment, in which a per transgene extending only up to amino acid 876 failed to rescue behavioral rhythms in per null mutant (per01) flies (Zehring, 1984). This truncated Per would still possess binding sites for Tim, Double-time, and Cryptochrome, but, as the experiments reveal, its CCID would be disrupted. The monopartite NLSs previously predicted by sequence analysis are relatively weak in regulating Per localization. Instead, a novel, bipartite NLS in the CCID is the dominant NLS in S2 cells. However, there may be several NLSs that contribute to Per nuclear transport in vivo since transgenic fly experiments suggest that there is a competent NLS in the first 95 amino acids of Per, and N-terminal fragments of Per do show some nuclear localization in S2 cells. The nuclear transport of Per is essential for its inhibition of Clk:Cyc-mediated transcription. These results advance understanding of Per function and thus understanding of the Drosophila circadian clock mechanism (Chang, 2003).

Clock genes: Stress response genes protect against lethal effects of sleep deprivation in Drosophila

Sleep is controlled by two processes: a homeostatic drive that increases during waking and dissipates during sleep, and a circadian pacemaker that controls its timing. Although these two systems can operate independently, recent studies indicate a more intimate relationship. To study the interaction between homeostatic and circadian processes in Drosophila, homeostasis was examined in the canonical loss-of-function clock mutants period (per01), timeless (tim01), clock (Clkjrk) and cycle (cyc01). cyc01 mutants show a disproportionately large sleep rebound and die after 10 hours of sleep deprivation, although they are more resistant than other clock mutants to various stressors. Unlike other clock mutants, cyc01 flies show a reduced expression of heat-shock genes after sleep loss. However, activating heat-shock genes before sleep deprivation rescues cyc01 flies from its lethal effects. Consistent with the protective effect of heat-shock genes, is the observation that flies carrying a mutation for the heat-shock protein Hsp83 (Hsp8308445) show exaggerated homeostatic response and die after sleep deprivation. These data represent the first step in identifying the molecular mechanisms that constitute the sleep homeostat (Shaw, 2002).

A sleep-like state has been described in Drosophila on the basis of its similarities to mammalian sleep. This state is characterized by increased arousal thresholds and is regulated homeostatically. Like mammalian sleep, it is abundant in young flies, decreases in older animals and is modulated by stimulants and hypnotics. Perhaps the most important similarity between mammals and flies is homeostatic regulation: when flies are kept awake, they show a large compensatory increase in sleep the next day (Shaw, 2002).

In mammals, the circadian pacemaker alternately promotes and maintains both wakefulness and sleep. Although the circadian pacemaker and the sleep homeostat can interact, little is known about the underlying mechanisms. To evaluate this relationship, homeostasis was evaluated in clock mutants maintained in constant darkness (DD) and deprived of sleep for 3, 6, 9 and 12 h. Under these conditions, sleep is evenly distributed across the day. Upon release from sleep deprivation, wild-type Canton-S flies recover ~30%-40% of the sleep that they lost within 12 h. per01 and Clkjrk show a more prominent sleep rebound, reclaiming ~100% of lost sleep within 12 h. tim01 flies did not show a homeostatic response after 3–6 h of sleep deprivation but displayed a sleep rebound similar to that of per01 and Clkjrk flies after 7, 9 and 12 h of sleep deprivation (Shaw, 2002).

Surprisingly, cyc01 mutants showed an exaggerated response to 3 h of sleep deprivation, reclaiming ~3 min of sleep over baseline for each minute of sleep lost. Further increasing sleep debt produces a change in the regulation of sleep not seen in other clock mutants: cyc01 flies showed large increases in sleep that persisted for as long as the flies were recorded (up to 16 days). These periods of quiescence are associated with increased arousal thresholds, indicating that the deprivation has produced an increase in the amount of sleep and does not result in an injured fly. If sleep deprivation produces an increased need for sleep in cyc01 flies, they should show higher amounts of sleep when sleep deprived for a second time. Indeed, deprivations of an additional 6 h results in further increases in sleep (Shaw, 2002).

The extreme sensitivity of cyc01 flies is revealed when sleep deprivation extends past 10 h: the flies begin to die. This effect was not observed in wild-type flies, in mutant lines representing 45 other genetic loci, or in other clock mutants, indicating that the mutations do not in themselves increase vulnerability to sleep deprivation. Note that the most sensitive of the clock mutants (cyc01) is the only one that does not cycle (Shaw, 2002).

To determine whether death in the cyc01 flies was due to the stimuli used for sleep deprivation, flies were deprived of sleep for 30 min each hour for 24 h, ensuring that the flies received the same number of stimuli that accrued during 12 h of sleep deprivation but without producing 12 h of continuous wakefulness. No deaths were observed after this protocol, indicating that the deprivation stimulus was not responsible for the deaths. Further supporting this conclusion, stress-sensitive B (sesB1) flies, which are extremely sensitive to mechanical shock, survived 12 h of sleep deprivation and showed activity patterns during the deprivation that were similar to those of wild-type flies. cyc01 flies were also deprived of sleep by gentle handling. The proportion of flies that succumbed to sleep deprivation and the size of the homeostatic response in surviving flies were indistinguishable from the automated deprivation method. Similar results were obtained with a rotating deprivation apparatus (Shaw, 2002).

To determine whether death in cyc01 flies is due specifically to sleep deprivation or to hypersensitivity to any environmental challenge, per01, tim01, Clkjrk, cyc01 and Canton-S flies were subjected to several stressors including heat stress, oxidative stress, starvation, desiccation and physical stress. cyc01 flies were as sensitive, but no more so than other genotypes to desiccation and vortex-mixing and survived longer than per01, tim01 and Clkjrk flies when challenged with heat, oxidative stress and starvation. Canton-S flies, which have an intact clock, were more resistant to starvation and desiccation than tim01, Clkjrk and cyc01 flies. These data indicate that cyc01 mutants are vulnerable to prolonged wakefulness in itself and are not merely hypersensitive to non-related stressors (Shaw, 2002).

To confirm that this phenotype maps to the cyc locus, cyc01 homozygotes were crossed with flies carrying the appropriate deficiency Df(3L)kto2/TM6B, Tb1. The resulting cyc01/Df transheterozygote flies showed an exaggerated homeostatic response and deaths after 12 h of sleep deprivation. Furthermore, cyc01 heterozygotes with and without a functioning clock (cyc01/+ and per0w;cyc01/+) also showed exaggerated homeostasis (Shaw, 2002).

Prolonged sleep deprivation (2–4 weeks) is invariably fatal in normal rats. Is the rapid demise after a few hours of sleep deprivation the result of an anomalous reaction in cyc01 mutants, or is it an increased susceptibility to the lethal consequences of sleep loss? Individual Canton-S flies were kept awake for 70 h by tapping on their tubes when they stopped moving to ensure that lethality was not due to excessive handling; 2 of 12 Canton-S flies died after ~60 h of continuous wakefulness, whereas 2 more died by ~67–70 h. The behavior of the flies during the last hours of the sleep deprivation protocol resembled that seen in cyc01 flies, indicating that the deaths were due to sleep loss and not to the deprivation stimulus itself. To test this, an additional group of Canton-S flies was kept awake by using a different deprivation method and again it was found that flies began to die between 60 and 70 h. These data indicate that sleep does indeed serve a vital biological role in the fly and that specific mutations that increase susceptibility to death might help to clarify such a role (Shaw, 2002).

Given that cyc01 flies are equally well or better equipped than other clock mutants to tolerate chronic heat and other stressors, why do they die in response to sleep deprivation? There is much evidence that stress response genes can protect an organism during challenging conditions. The ability of heat and sleep deprivation to activate stress response genes was examined, by using real-time quantitative polymerase chain reaction (QPCR). All clock mutants respond to 3 h of heat with an induction of genes such as hsp70, Hsp83, droj1 and hsc70-3 coding for chaperone proteins. After 3 h of sleep deprivation, the levels of these genes were near baseline in Clkjrk flies (with the exception of hsp70) and were unchanged or mildly increased in per01, tim01 and Canton-S flies. Interestingly, levels of chaperone proteins are also elevated after sleep deprivation in rodents. However, sleep deprivation produces a decrease in the expression of these genes in cyc01 flies. Genes activated by qualitatively different stressors, including metabolic stress (SNF4a, Hif1), chemical stress (mpk2) and humoral stress (turandot), were reduced in all lines, indicating that sleep deprivation is not inherently stressful (Shaw, 2002).

To evaluate the relationship between heat-shock genes and sleep deprivation in cyc01 flies, heat-shock genes were induced before sleep deprivation. When 12 h of sleep deprivation was preceded by 3 h of heat exposure at 36°C, the mortality rate was reduced compared to unheated cyc01 flies (note that one would have predicted increased mortality because preheating results in a further 3 h of wakefulness). Moreover, heat exposure reduces homeostatic drive in cyc01 and Clkjrk flies). When cyc01 flies were pre-exposed to 37°C, homeostasis was reduced further (Shaw, 2002).

The importance of heat-shock genes in sleep deprivation was also shown by examining flies mutant for Hsp83 (Hsp8308445). After 12 h of sleep deprivation, Hsp8308445 mutants exhibit a mortality rate similar to that of cyc01 and show a homeostatic response corresponding to fivefold that of wild-type flies. The sensitivity to sleep deprivation in Hsp8308445 mutants is present even in heterozygous flies, which have only a modest reduction in gene product. Heterozygous Hsp8308445 flies display a sleep rebound that is not statistically different from either homozygous Hsp8308445 or heterozygous Hsp83e6A flies. However, both Hsp83 heterozygotes exhibit a sleep rebound that is significantly different from that of Hsp60RA75 heterozygotes, indicating that a limited set of chaperone proteins are involved in homeostasis. Finally, whereas preheating cyc01 flies prevents the lethal effects of sleep deprivation, this does not occur in Hsp8308445 flies. It should be noted that sleep homeostasis has been evaluated in mutant lines representing 45 other genetic loci; cyc01 and Hsp8308445 flies are the only mutants that show both an exaggerated homeostatic response and death after sleep deprivation. Although it is unlikely that these are the only two genes involved in the sleep homeostat, it is worth noting that their mammalian homologs have been shown to interact physically. Nevertheless, it is possible that the increased homeostatic response and lethality that is observed in cyc01 mutants is due to factors other than Hsp83 (Shaw, 2002).

Although it is believed that sleep is an essential biological process, its function remains a mystery. So far, death after chronic total sleep deprivation in the rat provides the best evidence in support of a vital role for sleep. The data reinforce these findings and indicate that the vital role of sleep extends beyond mammals; the data also indicate a connection between vulnerability to sleep loss and increased homeostatic drive. Most importantly, the observation that the induction of certain chaperone proteins protects against the lethal effects of sleep loss provides a first hint about the functional targets of sleep and its molecular mechanisms (Shaw, 2002).

Gender dimorphism in the role of cycle (BMAL1) in rest, rest regulation, and longevity in Drosophila melanogaster

The central clock is generally thought to provide timing information for rest/activity but not to otherwise participate in regulation of these states. To test the hypothesis that genes that are components of the molecular clock also regulate rest, the authors quantified the duration and intensity of consolidated rest and activity for the four viable Drosophila mutations of the central clock that led to arrhythmic locomotor behavior and for the pdf mutant that lacked pigment-dispersing factor, an output neuropeptide. Only the cycle (cyc01) and Clock (ClkJrk) mutants had abnormalities that mapped to the mutant locus, namely, decreased consolidated rest and grossly extended periods of activity. All mutants with the exception of the cyc01 fly exhibited a qualitatively normal compensatory rebound after rest deprivation. This abnormal response in cyc01 was sexually dimorphic, being reduced or absent in males and exaggerated in females. Finally, the cyc01 mutation shortened the life span of male flies. These data indicate that cycle regulates rest and life span in male Drosophila (Hendricks, 2003).

PER-dependent rhythms in CLK phosphorylation and E-box binding regulate circadian transcription

Transcriptional activation by Clock-Cycle (Clk-Cyc) heterodimers and repression by Period-Timeless (Per-Tim) heterodimers are essential for circadian oscillator function in Drosophila. Per-Tim has been found to interact with Clk-Cyc to repress transcription, and this interaction is shown to inhibit binding of Clk-Cyc to E-box regulatory elements in vivo. Coincident with the interaction between Per-Tim and Clk-Cyc is the hyperphosphorylation of Clk. This hyperphosphorylation occurs in parallel with the Per-dependent entry of Double-time (Dbt) kinase into a complex with Clk-Cyc, where Dbt destabilizes both Clk and Per. Once Per and Clk are degraded, a novel hypophosphorylated form of Clk accumulates in parallel with E-box binding and transcriptional activation. These studies suggest that Per-dependent rhythms in Clk phosphorylation control rhythms in E-box-dependent transcription and Clk stability, thus linking Per and Clk function during the circadian cycle and distinguishing the transcriptional feedback mechanism in flies from that in mammals (Yu, 2006).

ChIP studies demonstrate that Clk-Cyc is only bound to E-boxes when target genes are being actively transcribed. Since Per-Dbt/Per-Tim-Dbt complexes interact with Clk-Cyc to inhibit transcription (Darlington, 1998; Lee, 1998), these data imply that these Per-containing complexes inhibit transcription by removing Clk-Cyc from E-boxes. It is also possible that binding of these Per-containing complexes to Clk-Cyc effectively blocks Clk and Cyc antibody access, in which case Per complexes would inhibit transcription while Clk-Cyc is bound to E-boxes. Given that the polyclonal Clk and Cyc antibodies used in this study were raised against full-length proteins and have been used to immunoprecipitate Per-containing complexes (Lee, 1998), it is highly unlikely that all Clk and Cyc epitopes are fully blocked by Per complex binding. Thus, it is concluded that Clk-Cyc rhythmically binds E-boxes in concert with target gene activation (Yu, 2006).

Per complex binding could remove Clk-Cyc from E-boxes by directly altering their conformation or by promoting Clk phosphorylation. The region of Per that inhibits Clk-Cyc transcription, called the Clk-Cyc inhibitory domain or CCID, is near the C terminus. The CCID can act independently of the N terminus of Per, where the Dbt-binding domain resides. This observation argues that Per does not inhibit Clk-Cyc binding to E-boxes by promoting Dbt-dependent Clk phosphorylation. Dbt- and CK2-dependent phosphorylation nevertheless enhances transcriptional repression in S2 cells by potentiating Per inhibition or by inhibiting Clk activity directly. Unfortunately, these disparate results from S2 cells do not allow distinguishing between the different effects of Per complex binding to inhibit transcription outlined above (Yu, 2006).

In mammals, mCry complexes bind to CLOCK-BMAL1 and repress transcription without removing CLOCK-BMAL1 from E-boxes. This contrasts with the situation in flies, where Per complexes inhibit transcription by inhibiting Clk-Cyc E-box binding, and suggests that these Per and mCry complexes repress transcription via different mechanisms. Although mCry complexes do not remove CLOCK-BMAL1 from E-boxes, they repress transcription by inhibiting the CLOCK-BMAL1-induced acetylation of histones by blocking p300 histone acetyl transferase function or introducing a histone deacetylase. Even though Per complexes repress transcription by inhibiting Clk-Cyc binding to E-boxes, this does not exclude the possibility that rhythms in histone acetylation are also involved in regulating rhythmic transcription in flies. Since chromatin remodeling is generally accepted as a prerequisite for transcription initiation, it would be surprising if rhythms in transcription were not accompanied by rhythms in histone acetylation or some other form of chromatin remodeling (Yu, 2006).

A rhythm in Clk phosphorylation has been defined in which hyperphosphorylated Clk predominates during times of transcriptional repression and hypophosphorylated Clk predominates during times of transcriptional activation. This rhythm occurs in parallel to the rhythm in Per phosphorylation; hyperphosphorylated Per and Clk accumulate in nuclei during the late night and early morning, then these forms are degraded and hypophosphorylated forms of Per and Clk accumulate in the cytoplasm and nucleus, respectively, during the late day and early evening. The rhythm in Clk and Per phosphorylation are not merely coincidental; the accumulation of hyperphosphorylated Clk is Per dependent. Although Per is not itself a kinase, it is bound by Dbt kinase. Per brings Dbt into the nucleus, where Per-Dbt or Per-Tim-Dbt complexes bind Clk-Cyc to inhibit transcription (Yu, 2006 and references therein).

Since Dbt enters a complex containing Clk-Cyc at times when Clk becomes hyperphosphorylated, Dbt may also act to phosphorylate Clk. However, an in vitro assay for Dbt phosphorylation is not available, thus it iw not known whether Dbt directly phosphorylates Clk. Dbt acts to reduce Clk levels in S2 cells even though Per levels are very low. It is therefore possible that Dbt can act to destabilize Clk in a Per-independent manner, although it is believed this is unlikely to be the case since Clk hyperphosphorylation and complex formation with Dbt are both Per dependent (Yu, 2006).

Clk is phosphorylated to some extent in the absence of Per and is hyperphosphorylated in the absence of functional Dbt, indicating that other kinases act to phosphorylate Clk. The accumulation of hyperphosphorylated Clk in dbtAR/dbtP flies suggests that Dbt triggers Clk degradation subsequent to Clk hyperphosphorylation. A similar situation is seen for Per, where hyperphosphorylated Per accumulates in the absence of functional Dbt, and phosphorylation by CK2 precedes Dbt-dependent phosphorylation and Per destabilization. In addition, rhythmically expressed phosphatases may also contribute to Clk phosphorylation rhythms (Yu, 2006 and references therein).

Rhythms in Clk phosphorylation may function to modulate Clk stability, subcellular localization, and/or activity. Clk levels do not change appreciably throughout the daily cycle despite approximately fivefold higher levels of Clk mRNA at dawn than at dusk. If a less stable hyperphosphorylated form of Clk accumulates when Clk mRNA is high and a more stable hypophosphorylated form of Clk accumulates when Clk mRNA is low, they would tend to equalize total Clk levels over the daily cycle. This possibility is supported by results in ARK flies, which express Clk mRNA in the opposite circadian phase (i.e., Clk mRNA peak at dusk rather than dawn). The overall level of Clk cycles in ARK flies with a peak in (hypophosphorylated) Clk around dusk, consistent with hypophosphorylated Clk being more stable than hyperphosphorylated Clk. This possibility is also supported by Dbt-dependent destabilization of Clk in S2 cells since Dbt associates with Clk as hyperphosphorylated Clk accumulates in wild-type flies. If hypophosphorylated Clk is relatively stable, higher levels of Clk might be expected to accumulate in per01 flies. However, constant low levels of Clk mRNA likely limit Clk accumulation in per01 flies. Clock phosphorylation is coupled to its degradation in cultured mammalian cells, yet degradation of phosphorylated Clock does not lead to a rhythm in Clock abundance even though Clock mRNA levels are constant (Yu, 2006).

Studies in cultured mammalian cells also demonstrate that Clock phosphorylation promotes Clock-BMAL1 nuclear localization, although the significance of this nuclear localization is not clear given that Clock-BMAL1 binding to E-boxes is either constant or more robust during transcriptional repression in vivo. In contrast, Clk is nuclear throughout the daily cycle in flies (Yu, 2006).

The coincidence between Clock phosphorylation and transcriptional repression in mice supports the possibility that phosphorylation inhibits Clock-BMAL1 activity, perhaps by promoting HDAC binding or inhibiting HAT binding. Likewise, hypophosphorylated and hyperphosphorylated Clk accumulate in parallel with target gene activation and repression, respectively, in flies. This relationship suggests that the state of Clk phosphorylation may alter its ability to activate target genes. Given that target gene activation occurs when Clk-Cyc is bound to E-boxes and that E-box binding coincides with the accumulation of hypophosphorylated Clk, it is possible that Clk hyperphosphorylation compromises Clk-Cyc binding to E-boxes and, consequently, target gene transcription is repressed. Precedent for such a regulatory mechanism is seen in the Neurospora clock, where limiting levels of FREQUENCY (FRQ) promote phosphorylation of WHITE COLLAR 1 (WC1) and WHITE COLLAR 2 (WC2), thereby inhibiting WC1-WC2 binding to C-box regulatory elements and repressing transcription. In contrast to FRQ in Neurospora, Per is considerably more abundant than Clk in Drosophila and forms stable complexes with Clk-Cyc. In addition, Per/Per-Tim can release Clk-Cyc from E-boxes in vitro, thus demonstrating that Per/Per-Tim binding is sufficient to release Clk-Cyc from E-boxes independent of Clk phosphorylation. Taken together with the in vitro E-box binding results, the high levels of Per relative to Clk and the formation of stable Per-Tim-Clk-Cyc complexes in flies argue that Per/Per-Tim binding may also be sufficient to inhibit E-box binding by Clk-Cyc in vivo, although they do not rule out a role for Clk hyperphosphorylation in inhibiting Clk-Cyc E-box binding. For instance, Per/Per-Tim binding could function to initially remove Clk-Cyc from E-boxes, and subsequent Clk phosphorylation could maintain Clk-Cyc in a form that is incapable of binding E-boxes (Yu, 2006).

The constant levels and rhythmic phosphorylation of Clk defined in this study are similar to those previously characterized for mammalian Clock. This similarity extends beyond metazoans to fungi, where positive elements of the Neurospora circadian feedback loop; i.e., WC1 and WC2, are also rhythmically phosphorylated. In each of these organisms, phosphorylation of positive factors increases when they interact with their respective negative feedback regulators, and decreases when they activate target gene transcription in the absence of these feedback inhibitors. This remarkable similarity suggests that phosphorylation controls one or more critical aspects of positive element function, and consequently, the rhythm in the positive element phosphorylation has become a conserved feature of circadian feedback loops in eukaryotes (Yu, 2006 and references therein).

Per-dependent regulation of Clk-Cyc binding to E-boxes, Per-dependent formation of Per-Dbt and/or Per-Dbt-Tim complexes with Clk-Cyc, and Per-dependent rhythms in Clk phosphorylation suggest a model for the regulation of rhythmic transcription. During the late day and early evening, hypophosphorylated Clk-Cyc binds E-boxes to activate transcription of per, tim, and other genes within and downstream of the transcriptional feedback loop. Accumulating levels of per mRNA peak during the early evening, but Per accumulation is delayed due to Dbt-dependent (and possibly CK2-dependent) phosphorylation, which destabilizes Per. Per is subsequently stabilized via Tim binding, which inhibits further phosphorylation of Per by Dbt. Phosphorylation of Tim by SGG then promotes the translocation of Tim-Per-Dbt complexes into the nucleus, where they bind (hypophosphorylated) Clk-Cyc and repress transcription by inhibiting E-box binding and promoting Clk hyperphosphorylation and degradation. These transcriptional repression mechanisms are not mutually exclusive; Clk hyperphosphorylation may inhibit E-box binding as well as promote Clk degradation. Dbt is able to enter the nucleus in per01 flies, but does not associate with Clk in the absence of Per. This suggests that Per is required to either bring Dbt into a complex with Clk-Cyc, enable phosphorylation of Clk after Dbt enters the complex, or both. Once the Tim-Per-Dbt-Clk-Cyc complex has formed, hyperphosphorylated Per and Clk levels decline in a coordinated fashion by mid-day. Tim is eliminated prior to hyperphosphorylated Per and Clk via separate light-dependent and light-independent mechanisms. As hyperphosphorylated Clk and Clk mRNA decline during the day, hypophosphorylated Clk accumulates. This hypophosphorylated Clk forms complexes with Cyc and binds E-boxes in the absence of nuclear Tim-Per-Dbt complexes, thus initiating the next cycle of transcription (Yu, 2006 and references therein).


Clock: Biological Overview | Evolutionary Homologs | Developmental Biology | Effects of Mutation | References

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.