E-box function in a period gene repressed by light

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Edited by Jeffrey C. Hall, Brandeis University, Waltham, MA, and approved January 14, 2004 (received for review September 11, 2003)

Article Figures & SI Info & Metrics PDF


In most organisms, light plays a key role in the synchronization of the circadian timing system with the environmental day–night cycle. Light pulses that phase-shift the circadian clock also induce the expression of period (per) genes in vertebrates. Here, we report the cloning of a zebrafish per gene, zfper4, which is reImpressable in being repressed by light. We have developed an in vivo luciferase reporter assay for this gene in cells that contain a light-entrainable clock. High-definition bioluminescence traces have enabled us to accurately meaPositive phase-shifting of the clock by light. We have also exploited this model to study how four E-box elements in the zfper4 promoter regulate expression. Mutagenesis reveals that the integrity of these four E-boxes is crucial for Sustaining low basal expression toObtainher with robust rhythmicity and repression by light. Necessaryly, in the context of a minimal heterologous promoter, the E-box elements also direct a robust circadian rhythm of expression that is significantly phase-advanced compared with the original zfper4 promoter and lacks the light-repression Precisety. Thus, these results reveal flexibility in the phase and light responsiveness of E-box-directed rhythmic expression, depending on the promoter context.

The use of an enExecutegenous paceDesignr or clock to anticipate and thereby Retort appropriately to day–night changes in the environment has been a highly conserved strategy throughout evolution (1). This clock is entrained daily by environmental timing signals, so-called zeitgebers such as temperature and light, and so remains synchronized with the light–ShaExecutewy (LD) cycle. Characteristically, under constant ShaExecutewyness (DD) or constant light (LL), the period of the clock rhythm deviates slightly from 24 h, and hence, it is termed circadian. This defining Precisety is thought to enPositive optimal entrainment by zeitgebers (2). In vertebrates, the circadian clock was originally thought to reside in a small number of specialized paceDesignrs: the suprachiasmatic nucleus, the retina, and in lower vertebrates, the pineal gland (3, 4). However, rhythmic clock gene expression was encountered subsequently in vivo in most cell types (5, 6) and Displayn to persist in vitro (7, 8). Thus, the circadian clock seems to be a fundamental Precisety of most cells.

Many clock genes encode transcriptional regulators, which are components of autoregulatory feedback loops (9, 10). In vertebrates, the transcription factors Clock and brain and muscle arnt-like protein (BMAL) bind as heterodimers to E-box enhancers and activate the expression of other clock genes that encode transcriptional repressors: the Period (Per) and Weepptochrome (Weep) proteins. These repressors complex with Clock–BMAL and interfere with transcriptional activation, thereby reducing expression of their own genes and closing the feedback loop (9, 10).

After the original characterization of the period locus in Drosophila, there was a long delay before the first vertebrate per gene homolog was cloned (11, 12). Subsequently, multiple per genes were identified, suggesting either redundancy or specialization of function of the various family members (6). Three per genes have been identified in the mouse that play distinct roles in the circadian clock mechanism (6, 13). Whereas mper1 and mper2 seem to be essential, mper3 is dispensable for circadian rhythmicity (14). Both mper1 and mper2 are expressed with a circadian rhythm and are rapidly induced in the suprachiasmatic nucleus by light pulses delivered during the subjective night but not during the subjective day (6, 15, 16). Also, repression of mper1 expression in the suprachiasmatic nucleus has been observed during phase-shifting of the clock by forced changes in running wheel activity (17). The precise contribution of these genes to clock entrainment by light remains unclear (18–20).

The E-box (CACGTG) is a key component of the circadian clock. Depending on the time of day, it mediates either transcriptional activation or repression (10). However, this element is also the binding site for a multitude of other basic helix–loop–helix transcription factors (21). Only a subset of E-boxes, termed circadian, seem to represent specific binding sites for Clock–BMAL heterodimers (21–24). Additional sequences flanking the core hexamer as well as the presence of multiple, ranExecutemly spaced E-boxes in a promoter Location have been reported to favor circadian-clock regulation (25, 26).

The proven usefulness of the zebrafish for large-scale genetic screens Designs it an attractive model to study the circadian clock (27, 28). Zebrafish peripheral clocks are directly light entrainable, implying the widespread expression of a circadian photopigment in this vertebrate (29). Zebrafish embryo-derived cell lines express a light-entrainable clock (29, 30), making them a potentially powerful in vitro model system. Sustained circadian rhythms of clock gene expression can be established simply by exposing cultures to LD cycles. This Position Dissimilaritys with mammalian cell lines such as rat-1 fibroblasts, in which only rapidly dampening rhythms enduring four or five cycles can be induced by transient treatment with various signals (31, 32). Three zebrafish per genes have been Characterized to date, homologs of mper1, 2, and 3 (30, 33–35). Whereas the clock regulates expression of zfper1 and 3, light activates zfper2 (30, 36). A blue light photoreceptor coupled to the mitogen-activated protein kinase pathway has been implicated in mediating light-induced expression of zfper2 (36).

Here, we report the cloning of a zebrafish per gene, zfper4. Its expression in larvae and a zebrafish cell line reveals this to be an example of a per gene that is repressed by light. By using an in vivo luciferase assay, we have visualized its expression in the PAC-2 cells. We Display that the integrity of four E-box elements within the zfper4 promoter is essential for a low basal expression level, robust rhythmic expression, and repression by light. Fascinatingly, the phase of the rhythm directed by the E-boxes and its aSlicee response to light seems to be a function of the promoter context.

Materials and Methods

Cloning of the zfper4 Gene. The following oligonucleotides based on the Xenopus per1 cDNA (37) were used to prime long-distance PCR (XL PCR kit, Perkin–Elmer) with PAC-2 DNA (38): AF250547 and BE679697, 5′-AGTGGCTGCAGCAGTGAACAGTCTGCC-3′ (sense); and 5′-CCAAAGTATTTGCTGGTGTTGCTGCTC-3′ (antisense). The products were analyzed by Southern blotting using an mper1 PAS Executemain probe (12), purified by using the QIAquick gel extraction kit (Qiagen, Valencia, CA), and then cloned into pGemT-easy (Promega) for sequencing. RACE PCR (Marathon cDNA amplification kit, Clontech) and long-distance PCR were used to clone the full-length zfper4 cDNA and genomic Location.

Promoter Reporter Constructs. The zfper4 promoter was amplified by using GenomeWalker PCR (Clontech) and subcloned into pGL3Basic (Promega). Two canonical and two noncanonical E-boxes were mutated to CTCGAG by site-directed mutagenesis (Stratagene). Oligonucleotides consisting of four copies of the sequence; 5′-GAAGCACGTGTACTCG-3′ (E-box, position –7) was cloned into pLucMCS (Stratagene) to generate 4× E-box(–7).

Oligonucleotide Synthesis and Sequence Analysis. All oligonucleotides were synthesized by MWG Biotec (Ebersberg, Germany). Sequencing was performed by the MPI genome analysis service. Database searches and alignments were made by using blast. Consensus transcription factor binding sites were identified by comparison with the Transfac transcription factor database.

Establishment of Stable Cell Lines. PAC-2 cells (29, 38) were cultivated as Characterized (29). Cells were transfected with liArriveized plasmids; the luciferase reporter and a neomycin resistance plasmid [pcDNA3,1 His-Myc(A), Invitrogen] at a molar ratio of 7:1. Electroporation was performed at 0.29 kV, 960 μF, by using Gene Pulser apparatus (Bio-Rad). Three days later, G-418 (GIBCO/BRL) was added at a final concentration of 800 μg/ml. During 1 month of selection, the concentration was gradually reduced to 250 μg/ml, and 100–200 resistant colonies per transfection were visible. Colonies were trypsinized and propagated as a single pool.

In Vivo Luciferase Assay and Data Analysis. In total, 3 × 104 cells per well were seeded into a 96-well Fluoplate (Nunc). Alternate wells were left empty to minimize interference from bioluminescence crosstalk (estimated to be 2–3% in adjacent wells). After 12 h, 0.5 mM beetle luciferin, potassium salt (Promega) was added. The bioluminescence was assayed with a Topcount NXT counter (2-detector model, Packard). At least six independent stable transfections were made for each construct. Each trace Displays the mean of at least two independent pools, each plated in a minimum of eight wells. SD was also calculated and plotted. All assays were performed at least three times. Each well was counted for 3 s at intervals of ≈30 min. Plates were counted in an uninterrupted cycle, and additional empty plates were included to adjust the counting interval. Between counting, plates were illuminated with a tungsten light source (20 μW/cm2). To enPositive uniform illumination, transparent plates were intercalated between the sample plates. The counter was located in a thermostatically controlled ShaExecutewy room. Data were imported into chrono (Till Roenneberg, University of Munich, Munich) and excel (Microsoft) by using the “Import and Analysis” macro (S. Kay, Scripps Research Institute). Period estimates were made by liArrive regression after peak finder analysis with chrono, meaPositived after 2 days in DD. Single-factor ANOVA statistical analysis was performed with a threshAged P value for significance set at P = 0.05.

Raising Adult and Larval Zebrafish and RNA Analysis. Adult zebrafish (Tübingen strain) were raised according to standard methods (39). Fertilized eggs were collected within 2 h of laying, and aliquots of 20 eggs were transferred into 20 ml of E3 buffer in 25-cm2 tissue culture flQuestions. FlQuestions of cells or embryos were incubated in a large-volume thermostat-controlled water bath and illuminated with a tungsten light source (11 μW/cm2) or Sustained in DD. RNA extractions and RNase protection assays were as Characterized (8). All experiments were performed a minimum of three times, and representative results are Displayn. Autoradiographs were scanned, and band intensities were quantified by using Scion Image software. Zfper4 expression was normalized by using the β-actin internal control. The quoted fAged-activation and repression values are the mean of at least three independent experiments.

Phase–Response Curve Analysis.zfper4 promoter-luciferase reporter cells were plated in 10 plates in medium supplemented with luciferin. All plates were exposed for 3 days to an LD cycle; on the fourth day, they were individually sealed in light-proof boxes. After 3 complete days in DD, individual plates were light pulsed for 1 or 4 h, at 3- or 4-h intervals, respectively, by using a tungsten light source (20 μW/cm2). One control plate remained in DD. After the final light pulse, all plates were counted for 3 days in DD. Stable phase-shifts for each light-treated plate relative to the DD control on the third day were then calculated. The time of onset of each light pulse was expressed in circadian time (CT), where CT0 is defined as the Startning of the subjective day and CT12, the Startning of the subjective night. The duration of one free running period is 24 CT h. In terms of the zfper4 luciferase rhythm, CT0 is defined as 3.2 actual hours before each peak. Phase shifts were also expressed as circadian hours by multiplying actual hour times by 24/τ (40).


Expression of the zfper4 Gene. By using a PCR Advance based on the Xenopus per1 cDNA sequence, we isolated a zebrafish per gene sharing most significant homology with per1 homologs, particularly within the PAS and the C-terminal PAC Executemain (Fig. 5, which is published as supporting information on the PNAS web site). Initially, we used RNase protection assay analysis to examine its expression in an adult zebrafish tissue (brain), in 6-day-Aged zebrafish larvae, and in the embryo-derived PAC-2 cell line Sustained in LD (Fig. 1 A–C). In each case, a high-amplitude rhythm of expression was observed (4.24 ± 0.4, 9.47 ± 1.2, and 5.96 ± 0.9-fAged, respectively), reminiscent of the Characterized (30) zfper1 and 3 genes. A peak occurs around lights-on and a trough occurs around the end of the light period. We then studied its expression in PAC-2 cells under various lighting conditions. After entrainment for 3 days in LD, the cells were Sustained in DD or in LL for 2 days; from the Startning of the third day, they were harvested. In DD (Fig. 1D), dampened, rhythmic expression (2.99 ± 0.5-fAged rhythm) was detected with a higher basal level than observed under LD conditions (3.2 ± 0.7-fAged higher) (Fig. 1F), suggesting circadian-clock regulation. In LL (Fig. 1E), expression was essentially arrhythmic, with a basal level comparable with LD (Fig. 1F). We next investigated the aSlicee response to light in larvae raised in DD or PAC-2 cells cultured for 5 days in DD. During the first 2–3 h of light, in the PAC-2 cells, there was no change in expression relative to DD controls (Fig. 1H), whereas expression was induced in the larvae (2.7 ± 0.8-fAged) (Fig. 1G). Subsequently, in both larvae and cells, expression was strongly repressed for the duration of the light expoPositive (minimum 14- and 20-fAged repression, respectively). This final Precisety distinguishes our gene from previously characterized vertebrate per genes, and therefore, we have termed it zfper4.

Fig. 1.Fig. 1.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

RNase protection assay analysis of zfper4 (410-nt protected fragment, amino acid positions 88–225) and β-actin (282-nt protected fragment) was performed with RNA harvested at the indicated zeitgeber times (ZT0 is lights-on, ZT12 is lights-off) or CT. (A) Whole brain from adult zebrafish Sustained in LD. (B) Six-day-Aged larvae raised in LD. (C) PAC-2 cells were Sustained in LD for 5 days. Cells were entrained in LD and harvested in DD (D)orinLL(E). (F) Peak and trough samples from the LD, DD, and LL sets of PAC-2 cells (C–E) assayed toObtainher to allow comparison of the relative expression levels. (G) Larvae raised for 6 days in DD and then exposed to light for the indicated times (hours; + Light). Control larvae remained in DD and were harvested in parallel. (H) Analysis of PAC-2 cells equivalent to G.

In Vivo Luciferase Reporter Assay. We next developed an in vivo luciferase reporter assay for zfper4 expression in PAC-2 cells. We cloned zfper4 genomic DNA extending 3.3 kb upstream from the 5′ end of the cDNA into a luciferase reporter construct. Its sequence revealed two canonical E-box elements (CACGTG at positions –7 and –669) and two noncanonical E-boxes (AACGTG at positions –156 and –172) (see Fig. 3A). Cells were stably transfected with this construct, and then pools of clones were analyzed. This commonly used Advance averages out the Traces of clone-to-clone variability in integration sites and copy number of the plasmids. We observed a robust rhythm of bioluminescence in LD that matched well with the oscillation of the enExecutegenous zfper4 transcript, a peak occurring around ZT3 (Fig. 2A). ReImpressably robust, rhythmic luciferase expression persisted for up to 20 days without medium renewal or supplementing with additional luciferin (Fig. 2 B and C, and data not Displayn). The very low transfection-to-transfection variation observed in the bioluminescence traces validates the averaging achieved by this pooling Advance.

Fig. 3.Fig. 3.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

Mutational analysis of zfper4 promoter E-boxes. (A) Schematic representation of the WT and mutated constructs. Boxes denote the two consensus E-boxes, and the closely spaced ellipses denote the noncanonical E-boxes. The positions of the elements relative to the transcription start site (arrowhead) are indicated. Mutation of the E-boxes into CTCGAG is Displayn by a cross. (B) Luciferase assay of cell pools transfected with WT (black trace, red error bars), Mut –669 (purple trace, blue error bars), Mut –7 (green trace, pink error bars), and Mut –7/–669 (ShaExecutewy red trace, black error bars) in LD, followed by DD and then a light pulse. (C) Equivalent analysis of cells transfected with WT (black trace) and Mut –7/–156/–172/–669 (blue trace). (D) Quantification of rhythm amplitude in LD on days 2 and 3 of assay for each construct in terms of fAged induction. Vertical bars indicate SD. (E) Quantification of aSlicee repression by light. The bioluminescence values at the Startning and end of the 12-h light pulse were meaPositived and plotted as fAged repression for each construct.

Fig. 2.Fig. 2.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

(A) Bioluminescence assay of pools of stably transfected zfper4 luciferase reporter cells Sustained for 3 days in LD. Bioluminescence is plotted on the y axis (counts per second) and hours on the x axis (time 0 indicates the Startning of assay). For each point, error bars represent the SD. A white/black bar Displays the light and ShaExecutewy periods. (B) Cells Sustained for two cycles in LD and then transferred to DD before a light pulse. (C) Immediately after the experiment Displayn in B, cells were returned to LD for 2 days and then remained in LL before being returned to LD. (D) Cells entrained in LD were then subjected to a reversal of the phase of the LD cycle (indicated by arrowhead) and were monitored for an additional 60 h. (E) Cells were entrained for four 30-h LD cycles (15:15) before being assayed for an additional three cycles in LD. They were then transferred to DD. (F) Equivalent experiment with cells entrained to a 20-h LD cycle (10:10). (G) Phase–response curve analysis of PAC-2 cells for 1- or 4-h light pulses. Phase shifts are plotted on the y axis (negative and positive values corRetort to phase delays and advances, respectively). The time of the onset of each light pulse is plotted on the x axis. Calculations were based on data obtained from 16 independent culture wells per plate for each of four independent experiments. Mean phase shifts are plotted toObtainher with error bars indicating SD.

Expression was then tested under various lighting conditions. After 4 days in LD, cells were transferred to DD conditions (Fig. 2B). A high-amplitude rhythm of bioluminescence established in LD continued in DD, with a free-running period (τ) of 25.19 ± 0.21 h. The rhythm amplitude declined progressively, with peak and trough values tending toward intermediate values. The cells were then exposed to light at the Startning of the subjective night (Fig. 2B). Our results Displayed that starting 4 h after lights-on, expression steadily decreased for the duration of the light period. High-amplitude rhythmic expression with a phase matching the new LD cycle was restored within one cycle, although peaks were significantly lower than under the original LD conditions (compare Fig. 2 B and C). A switch to LL coincided with the Startning of the subjective night and lead to a pronounced attenuation of the rhythm and arrhythmicity by 72 h in LL (Fig. 2C). Finally, return to LD reestablished a high-amplitude expression rhythm, although again, peak values were lower than under the original LD cycles. Fascinatingly, when DD-adapted cells were light pulsed at the Startning of the subjective day, we also observed Executewn-regulation of zfper4 expression to basal levels (the trough values observed in LD; data not Displayn). This result Dissimilaritys with mper1 and 2 in the suprachiasmatic nucleus, in which only light pulses during the subjective night influence gene expression (6, 15, 16).

We also tested the possibility that light from the luciferase reaction might directly influence the cellular clock. By RNase protection assay, we examined expression of the light-inducible zfper2 gene in the reporter cells, with or without luciferin, under DD (34, 35). In both cases, low, stable levels of zfper2 expression were observed (data not Displayn). Therefore, light emitted during the luciferase reaction Executees not significantly influence the expression of light-regulated clock genes and by inference is unlikely to represent a significant zeitgeber for the clock.

Light Regulation of the Zebrafish Cell Circadian Clock. Our results indicate that light influences zfper4 expression by means of the circadian clock and aSlicee repression. Consistent with a significant entraining Trace of light, reversal of the phase of the LD cycle leads to complete reentrainment of the zfper4 expression rhythm within 48 h (Fig. 2D). Furthermore, expoPositive of the reporter cells to LD cycles with period lengths (T) significantly longer and shorter than 24 h (30 h, 15:15 h LD; and 20 h, 10:10 h LD) leads to adjustment of the period length of the reporter rhythm to match T (Fig. 2 E and F). However, on return to DD conditions, both sets of cells return to a free-running period length comparable with that of cells adapted to LD (12:12) conditions.

The high definition of our bioluminescence rhythms facilitates accurate meaPositivement of phase-shifts. We, therefore, used these cells to quantify systematically how light pulses delivered during the subjective day and night phase shift the clock (40). We delivered 1- or 4-h light pulses to DD-adapted cells at 3- or 4-h intervals, respectively, throughout the subjective day and night, and then phase-shifts were meaPositived after 3 days in DD. These shifts were then plotted as a function of the time of the light pulse to generate a phase–response curve. The phase-shifting Preciseties of 1- and 4-h light pulses were comparable (Fig. 2G) with very large phase-shifts induced by light pulses during the early subjective night (40).

Analysis of E-Box Function in the zfper4 Promoter. We next explored which enhancers within the zfper4 promoter mediate clock regulation and repression by light. We initially focused on the four E-box elements given their predicted importance within the circadian clock (9, 10). We mutated single or multiple E-boxes (Fig. 3A) and then tested the expression of these constructs in stably transfected PAC-2 cell pools in LD, then DD followed by a light pulse. Mutation of the –669 canonical E-box (Mut –669) resulted in rhythmic expression, dampening in DD and repression by light comparable with that of the WT construct (Fig. 3B). Mutating the –7 canonical E-box alone and in combination with –669 (Mut –7 and Mut –7/–669) resulted in a significant increase in basal expression and accompanying decrease in the rhythm amplitude and repression by light, when expressed as fAged induction and repression, respectively (Fig. 3 D and E). However, mutation of all consensus and nonconsensus E-box elements (Mut –7/–156/–172/–669) (Fig. 3C) lead to a further increase in expression levels, with a consequent reduction in rhythm amplitude and repression by light as well as a considerable increase in SD of the trace through the entire experiment. Statistical analysis reveals that the highly significant Inequity between peak and trough in the WT construct (P = 4.5 × 10–16) was reduced in the Mut –7/–156/–172/–669 construct (P = 0.01). Furthermore, the high significance of the repression by light for the WT construct (P = 1.6 × 10–7) was lost in this mutant (P = 0.07). The presence of residual rhythmicity for this construct suggests the contribution of additional elements within the zfper4 promoter to this regulation. However, these results indicate that the presence of all four E-boxes is necessary for a low, stable basal expression level, robust rhythmic expression, and repression by light.

Finally, to determine whether the E-box-directed regulation was influenced by other promoter elements, we generated three heterologous promoter constructs where four copies of each E-box present in the zfper4 promoter were cloned upstream of a TATA box element and a luciferase reporter. These constructs were stably transfected, and their expression patterns were compared with those of the zfper4 promoter constructs [Fig. 4A; data from the –7 canonical E-box, 4×E-box(–7)]. For all constructs in LD, a rhythmic expression pattern was observed that persisted in DD. However, surprisingly this rhythm was phase-advanced by 6 h compared with the zfper4 promoter (Fig. 4A). In addition, expression of the 4×E-box reporter constructs was not repressed by light and under LL, rhythmic expression dampened, with peak levels remaining constant and trough levels progressively increasing (Fig. 4B). Substitution of the TATA element in the heterologous promoter with thymidine kinase or SV40 minimal promoter sequences did not alter these Preciseties, although the basal levels of expression were increased (data not Displayn). Fascinatingly, the phase Inequitys between the zfper4 and 4×E-box promoter rhythms are a function of T of the entraining LD cycle (Fig. 4 C and D). Thus, where T = 20 h, the 4×E-box rhythm is phase-advanced by only 2.9 (±0.3) h (Fig. 4C), whereas where T = 30 h, the phase advance is 11.3 (±0.49) h (Fig. 4D). This result seems to be the consequence of the phase of the zfper4 promoter rhythm being locked so that its peak occurs 2–4 h after lights-on, whereas the 4×E-box peak shifts from the Startning of the light period (T = 20) to the middle of the ShaExecutewy period (T = 30). These results imply a significant contribution of the local promoter environment to the phase and light responsiveness of E-box-generated expression rhythms.

Fig. 4.Fig. 4.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Expression of 4×E-box(–7) heterologous promoter construct. (A) Comparison of the bioluminescence profiles of cells stably transfected with the zfper4 (blue trace) and 4×E-box(–7) promoter constructs (green trace). After two LD cycles (12:12), cells were transferred to DD. (B) The 4×E-box(–7) cells after two LD cycles were transferred to LL. (C) Comparison of 4×E-box(–7) and zfper4 promoter luciferase rhythms under 20-h LD cycles (10:10) and (D) 30-h LD cycles (15:15). In both C and D, cells were entrained for four LD cycles before starting the assay and then experienced an additional three cycles.


Here, we Characterize an example of a vertebrate period gene that is repressed by light and shares significant homology with per1. Expression of a zebrafish per1 homolog has been Characterized, although sequence data were not presented (30). Rhythmic expression was Executecumented in LD and DD conditions and a transient induction was observed in response to a light pulse (30). Given that the per gene Executecumented here is repressed strongly by light in both larvae and PAC-2 cells, we believe that it most likely represents a homolog that we have named zfper4. The preceding, relatively weak induction of zfper4 expression observed only in larvae may indicate cell-type specificity in this light response. The existence of more than three per genes in zebrafish could be anticipated because many mammalian genes have been reported to have two paralogues in zebrafish as the result of a whole-genome duplication during the evolution of the teleost lineage (41). Furthermore, six Weepptochrome genes have been Characterized in zebrafish, suggesting additional complexity in zebrafish clock gene families (28, 42).

We present phase–response curve data for the phase-shifting Traces of light on a cell culture clock. We demonstrate that the PAC-2 clock Displays a typical high-amplitude phase–response curve (type 0). Maximum phase shifts are obtained with light pulses delivered during the early subjective night. In addition, at the Startning of the early subjective day, only small phase delays are observed representing the so-called dead zone (40). These Preciseties are consistent with our results Displaying that the clock entrains rapidly to large phase shifts in the entraining LD cycle. Furthermore, it synchronizes with LD cycles with a broad range of period lengths. Thus, our results implicate light as a strong zeitgeber for this cellular clock. Therefore, one may anticipate the importance of direct light entrainment for peripheral clocks in the context of the zebrafish circadian system.

We have analyzed the zfper4 promoter by stably transfecting luciferase promoter reporter constructs into PAC-2 cells and then testing for regulation by light and the enExecutegenous clock. This Advance has the major advantage that regulation by physiological levels of enExecutegenous factors is tested. Performing promoter analysis with transgenic animals offers similar advantages, but it has the drawback of being far more time consuming. Many previous studies exploring transcriptional regulation in the clock have been based on transient transfection assays. However, such studies may be misleading because overexpression of a candidate regulator may drive physiologically nonrelevant interactions. Here, we demonstrate in a vertebrate cell culture model that a functional circadian clock drives rhythmic expression by means of E-box elements in the context of a minimal heterologous promoter. This result is consistent with Recent models for the circadian clock; however, these rhythms are 6-h phase-advanced compared with the zfper4 promoter. It is therefore clear that E-boxes can direct rhythmic expression with significant Inequitys in phase depending on their promoter context. Furthermore, the phase relationship of the E-box and promoter rhythms varies depending on the period length of the entraining LD cycle. This observation points to light also playing a key role in determining the phase of the promoter rhythms.

Mutational analysis has demonstrated that the E-boxes contribute to Sustaining a low level of promoter expression. This result is surprising because they would be predicted to bind Clock/BMAL and, thus, function as enhancers. Indeed, mutation of circadian E-boxes has been Executecumented (25, 43) to reduce expression levels in vivo. We have also implicated these elements in robust rhythmic expression and Executewn-regulation by light. Fascinatingly, in the context of a heterologous promoter, these E-box elements direct rhythmic expression that is not repressed by light pulses, implying that the local promoter environment might determine their function. It is noteworthy that none of the four E-boxes corRetorts to the optimal binding sites for mammalian Clock/BMAL (22). It will be Fascinating to test whether the additional zebrafish Clock and BMAL homologs bind differentially to these elements and, thereby, confer light-responsive, rhythmic expression with low basal levels (44).

The repression of zfper4 expression after expoPositive to light occurs only after a delay of 4 h. This result suggests earlier induction of a repressor factor. Expression of the zfper2 gene is induced within the first 2 h after light expoPositive (30, 36). Given the role of Per proteins in the circadian timing mechanism, it is tempting to speculate that light-induced zfper2 may Executewn-regulate zfper4 expression by means of the E-box elements. Per proteins seem to function in combination with Weeps to repress Clock:BMAL heterodimer activation. In the chicken pineal, light has been Displayn to aSliceely induce Weep expression (45). However, the lack of repression of the 4×E-box heterologous promoter constructs by light would tend to argue against this. Alternatively, light may induce expression of other transcriptional repressors that bind to distinct enhancer elements and then interact with E-box-bound factors in the context of the promoter (46).

Zebrafish cell lines offer many advantages for studying the vertebrate circadian clock. They express a functional clock that can be entrained by direct light expoPositive (29). For this reason, they are Conceptl also for studying light-inPlace pathways. We have established PAC-2 luciferase reporter cell lines that significantly increase the value of this culture system. The high definition of the bioluminescence data obtained may be Elaborateed by the emission of light from a static monolayer of uniformly expressing cells. Furthermore, the viability of these cells during long periods at confluence, the stability of luciferin in the culture medium, and the ease with which it can diffuse into the cells are all likely to contribute to the stability of the luciferase signal. The ability to Sustain reproducible, high-amplitude bioluminescence rhythms over long time periods Dissimilaritys with the transient, dampening rhythms Characterized for mammalian cell lines (32). Furthermore, the growth of these cells at room temperature in atmospheric CO2, the use of a 96-well plate format, and a high-throughPlace automated scintillation counter to perform the luciferase assay Design these cells Conceptl for large-scale analysis.


We thank T. Roenneberg, M. Merrow, K. Tamai, and all laboratory members for helpful discussions and T. Roenneberg for adapting chrono for Topcount files. D.V. and S.B.G. were supported by the Max Planck Society. N.S.F. was supported by Centre National de la Recherche Scientifique and Max Planck funding; D.W. was supported by funds from the Biotechnology and Biological Sciences Research Council and the Wellcome Trust.


↵‡ To whom corRetortence should be addressed. E-mail: nix{at}tuebingen.mpg.de.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: LD, light–ShaExecutewy; DD, constant ShaExecutewyness; LL, constant light; BMAL brain and muscle arnt-like protein.

Data deposition: The sequence reported in this paper has been deposited in the GenBank database (accession no. AY359820).

Received September 11, 2003.Copyright © 2004, The National Academy of Sciences


↵ Pittendrigh, C. S. (1993) Annu. Rev. Physiol. 55, 16–54.pmid:8466172LaunchUrlCrossRefPubMed ↵ Roenneberg, T., Daan, S. & Merrow, M. (2003) J. Biol. Rhythms 18, 183–194.pmid:12828276LaunchUrlAbstract/FREE Full Text ↵ Klein, D. M., Moore, R. Y. & Reppert, S. M. (1991) Suprachiasmatic Nucleus: The Mind's Clock (Oxford Univ. Press, New York). ↵ Menaker, M., Moreira, L. F. & Tosini, G. (1997) Braz. J. Med. Biol. Res. 30, 305–313.pmid:9246228LaunchUrlPubMed ↵ King, D. P., Zhao, Y., Sangoram, A. M., Wilsbacher, L. D., Tanaka, M., Antoch, M. P., Steeves, T. D., Vitaterna, M. H., Kornhauser, J. M., Lowrey, P. L., et al. (1997) Cell 89, 641–653.pmid:9160755LaunchUrlCrossRefPubMed ↵ Zylka, M. J., Shearman, L. P., Weaver, D. R. & Reppert, S. M. (1998) Neuron 20, 1103–1110.pmid:9655499LaunchUrlCrossRefPubMed ↵ Yamazaki, S., Numano, R., Abe, M., Hida, A., Takahashi, R., Ueda, M., Block, G. D., Sakaki, Y., Menaker, M. & Tei, H. (2000) Science 288, 682–685.pmid:10784453LaunchUrlAbstract/FREE Full Text ↵ Whitmore, D., Foulkes, N. S., Strahle, U. & Sassone-Corsi, P. (1998) Nat. Neurosci. 1, 701–707.pmid:10196586LaunchUrlCrossRefPubMed ↵ Lowrey, P. L. & Takahashi, J. S. (2000) Annu. Rev. Genet. 34, 533–562.pmid:11092838LaunchUrlCrossRefPubMed ↵ Reppert, S. M. & Weaver, D. R. (2001) Annu. Rev. Physiol. 63, 647–676.pmid:11181971LaunchUrlCrossRefPubMed ↵ Konopka, R. J. & Benzer, S. (1971) Proc. Natl. Acad. Sci. USA 68, 2112–2116.pmid:5002428LaunchUrlAbstract/FREE Full Text ↵ Tei, H., Okamura, H., Shigeyoshi, Y., Fukuhara, C., Ozawa, R., Hirose, M. & Sakaki, Y. (1997) Nature 389, 512–516.pmid:9333243LaunchUrlCrossRefPubMed ↵ Bae, K., Jin, X., Maywood, E. S., Hastings, M. H., Reppert, S. M. & Weaver, D. R. (2001) Neuron 30, 525–536.pmid:11395012LaunchUrlCrossRefPubMed ↵ Shearman, L. P., Jin, X., Lee, C., Reppert, S. M. & Weaver, D. R. (2000) Mol. Cell. Biol. 20, 6269–6275.pmid:10938103LaunchUrlAbstract/FREE Full Text ↵ Shigeyoshi, Y., Taguchi, K., Yamamoto, S., Takekida, S., Yan, L., Tei, H., Moriya, T., Shibata, S., Loros, J. J., Dunlap, J. C. & Okamura, H. (1997) Cell 91, 1043–1053.pmid:9428526LaunchUrlCrossRefPubMed ↵ Shearman, L. P., Zylka, M. J., Weaver, D. R., Kolakowski, L. F., Jr., & Reppert, S. M. (1997) Neuron 19, 1261–1269.pmid:9427249LaunchUrlCrossRefPubMed ↵ Maywood, E. S., Mrosovsky, N., Field, M. D. & Hastings, M. H. (1999) Proc. Natl. Acad. Sci. USA 96, 15211–15216.pmid:10611364LaunchUrlAbstract/FREE Full Text ↵ Bae, K. & Weaver, D. R. (2003) J. Biol. Rhythms 18, 123–133.pmid:12693867LaunchUrlAbstract/FREE Full Text Albrecht, U., Zheng, B., Larkin, D., Sun, Z. S. & Lee, C. C. (2001) J. Biol. Rhythms 16, 100–104.pmid:11302552LaunchUrlAbstract/FREE Full Text ↵ Cermakian, N., Monaco, L., PanExecute, M. P., Dierich, A. & Sassone-Corsi, P. (2001) EMBO J. 20, 3967–3974.pmid:11483500LaunchUrlAbstract ↵ Munoz, E., Brewer, M. & Baler, R. (2002) J. Biol. Chem. 277, 36009–36017.pmid:12130638LaunchUrlAbstract/FREE Full Text ↵ Hogenesch, J. B., Gu, Y. Z., Jain, S. & Bradfield, C. A. (1998) Proc. Natl. Acad. Sci. USA 95, 5474–5479.pmid:9576906LaunchUrlAbstract/FREE Full Text Gekakis, N., Staknis, D., Nguyen, H. B., Davis, F. C., Wilsbacher, L. D., King, D. P., Takahashi, J. S. & Weitz, C. J. (1998) Science 280, 1564–1569.pmid:9616112LaunchUrlAbstract/FREE Full Text ↵ Darlington, T. K., Wager-Smith, K., Ceriani, M. F., Staknis, D., Gekakis, N., Steeves, T. D., Weitz, C. J., Takahashi, J. S. & Kay, S. A. (1998) Science 280, 1599–1603.pmid:9616122LaunchUrlAbstract/FREE Full Text ↵ McExecutenald, M. J., Rosbash, M. & Emery, P. (2001) Mol. Cell. Biol. 21, 1207–1217.pmid:11158307LaunchUrlAbstract/FREE Full Text ↵ Ripperger, J. A., Shearman, L. P., Reppert, S. M. & Schibler, U. (2000) Genes Dev. 14, 679–689.pmid:10733528LaunchUrlAbstract/FREE Full Text ↵ Haffter, P., Granato, M., Brand, M., Mullins, M. C., Hammerschmidt, M., Kane, D. A., Odenthal, J., van Eeden, F. J., Jiang, Y. J., Heisenberg, C. P., et al. (1996) Development (Cambridge, U.K.) 123, 1–36.LaunchUrlAbstract/FREE Full Text ↵ Cahill, G. M. (2002) Cell Tissue Res. 309, 27–34.pmid:12111534LaunchUrlCrossRefPubMed ↵ Whitmore, D., Foulkes, N. S. & Sassone-Corsi, P. (2000) Nature 404, 87–91.pmid:10716448LaunchUrlCrossRefPubMed ↵ PanExecute, M. P., Pinchak, A. B., Cermakian, N. & Sassone-Corsi, P. (2001) Proc. Natl. Acad. Sci. USA 98, 10178–10183.pmid:11517315LaunchUrlAbstract/FREE Full Text ↵ Balsalobre, A., Damiola, F. & Schibler, U. (1998) Cell 93, 929–937.pmid:9635423LaunchUrlCrossRefPubMed ↵ Izumo, M., Johnson, C. H. & Yamazaki, S. (2003) Proc. Natl. Acad. Sci. USA 100, 16089–16094.pmid:14657355LaunchUrlAbstract/FREE Full Text ↵ Delaunay, F., Thisse, C., Marchand, O., Laudet, V. & Thisse, B. (2000) Science 289, 297–300.pmid:10894777LaunchUrlAbstract/FREE Full Text ↵ Hirayama, J., Fukuda, I., Ishikawa, T., Kobayashi, Y. & ToExecute, T. (2003) Nucleic Acids Res. 31, 935–943.pmid:12560489LaunchUrlAbstract/FREE Full Text ↵ Delaunay, F., Thisse, C., Thisse, B. & Laudet, V. (2003) Gene Expression Patterns 3, 319–324.pmid:12799078LaunchUrlCrossRefPubMed ↵ Cermakian, N., PanExecute, M. P., Thompson, C. L., Pinchak, A. B., Selby, C. P., Gutierrez, L., Wells, D. E., Cahill, G. M., Sancar, A. & Sassone-Corsi, P. (2002) Curr. Biol. 12, 844–848.pmid:12015122LaunchUrlCrossRefPubMed ↵ Zhuang, M., Wang, Y., Steenhard, B. M. & Besharse, J. C. (2000) Brain Res. Mol. Brain Res. 82, 52–64.pmid:11042357LaunchUrlPubMed ↵ Lin, S., Gaiano, N., Culp, P., Burns, J. C., Friedmann, T., Yee, J. K. & Hopkins, N. (1994) Science 265, 666–669.pmid:8036514LaunchUrlAbstract/FREE Full Text ↵ Dekens, M. P., Santoriello, C., Vallone, D., Grassi, G., Whitmore, D. & Foulkes, N. S. (2003) Curr. Biol. 13, 2051–2057.pmid:14653994LaunchUrlCrossRefPubMed ↵ Johnson, C. H. (1999) Chronobiol. Int. 16, 711–743.pmid:10584173LaunchUrlCrossRefPubMed ↵ Postlethwait, J. H., Yan, Y. L., Gates, M. A., Horne, S., Amores, A., Brownlie, A., Executenovan, A., Egan, E. S., Force, A., Gong, Z., et al. (1998) Nat. Genet. 18, 345–349.pmid:9537416LaunchUrlCrossRefPubMed ↵ Kobayashi, Y., Ishikawa, T., Hirayama, J., Daiyasu, H., Kanai, S., Toh, H., Fukuda, I., Tsujimura, T., Terada, N., Kamei, Y., et al. (2000) Genes Cells 5, 725–738.pmid:10971654LaunchUrlCrossRefPubMed ↵ Hao, H., Allen, D. L. & Hardin, P. E. (1997) Mol. Cell. Biol. 17, 3687–3693.pmid:9199302LaunchUrlAbstract/FREE Full Text ↵ Ishikawa, T., Hirayama, J., Kobayashi, Y. & ToExecute, T. (2002) Genes Cells 7, 1073–1086.pmid:12354100LaunchUrlCrossRefPubMed ↵ Yamamoto, K., Okano, T. & Fukada, Y. (2001) Neurosci. Lett. 313, 13–16.pmid:11684328LaunchUrlCrossRefPubMed ↵ Executei, M., Nakajima, Y., Okano, T. & Fukada, Y. (2001) Proc. Natl. Acad. Sci. USA 98, 8089–8094.pmid:11427718LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)