CRTC Potentiates Light-independent timeless Transcription to Sustain Circadian Rhythms in Drosophila

Abstract

Light is one of the strongest environmental time cues for entraining endogenous circadian rhythms. Emerging evidence indicates that CREB-regulated transcription co-activator 1 (CRTC1) is a key player in this pathway, stimulating light-induced Period1 (Per1) transcription in mammalian clocks. Here, we demonstrate a light-independent role of Drosophila CRTC in sustaining circadian behaviors. Genomic deletion of the crtc locus causes long but poor locomotor rhythms in constant darkness. Overexpression or RNA interference-mediated depletion of CRTC in circadian pacemaker neurons similarly impairs the free-running behavioral rhythms, implying that Drosophila clocks are sensitive to the dosage of CRTC. The crtc null mutation delays the overall phase of circadian gene expression yet it remarkably dampens light-independent oscillations of TIMELESS (TIM) proteins in the clock neurons. In fact, CRTC overexpression enhances CLOCK/CYCLE (CLK/CYC)-activated transcription from tim but not per promoter in clock-less S2 cells whereas CRTC depletion suppresses it. Consistently, TIM overexpression partially but significantly rescues the behavioral rhythms in crtc mutants. Taken together, our data suggest that CRTC is a novel co-activator for the CLK/CYC-activated tim transcription to coordinate molecular rhythms with circadian behaviors over a 24-hour time-scale. We thus propose that CRTC-dependent clock mechanisms have co-evolved with selective clock genes among different species.

Figure 1. Drosophila CRTC is necessary for robust circadian behaviors.

(a) A schematic diagram of crtc mutant alleles. Dark-gray boxes, the translated regions of exons; light-gray boxes, untranslated regions. (b) Male flies homozygous or trans-heterozygous for crtc mutant alleles show long but poor rhythms in circadian behaviors. Normalized activity profiles in LD cycles (top) or on the first day of DD cycles (middle) were averaged from individual flies. Averaged actograms throughout the behavioral analyses were double-plotted (bottom). Anticipatory increase in locomotor activities prior to lights-on (i.e., morning anticipation) was quantified by calculating morning index in individual flies as described in Materials and Methods. Averaged morning index values +/− SEM were shown in the LD activity profiles. White arrow heads, morning anticipation in LD cycles; black arrow heads, morning anticipation in the first DD cycle; white/black bars, LD cycles; gray/black bars, DD cycles; ZT, zeitgeber time; CT, circadian time. Error bars indicate SEM. (c) Circadian periods in DD locomotor rhythms were averaged from rhythmic flies (P-S > 10; see below). ***P < 0.001 to wild-type (w¹¹¹⁸) and all heterozygous controls as determined by one-way ANOVA, Tukey post hoc test. Error bars indicate SEM. (d) Rhythmicity in free-running locomotor behaviors was determined by measuring power (P) - significance (S) values from the chi-squared periodograms of individual flies and averaged per each genotype. ***P < 0.001 to wild-type (w¹¹¹⁸) and all heterozygous controls as determined by one-way ANOVA, Tukey post hoc test. Error bars indicate SEM.

Figure 2. Drosophila clocks are sensitive to the dosage of CRTC in PDF Neurons.

(a) Circadian pacemaker neurons in adult fly brain. tim-Gal4 is expressed in all clock neurons covering lateral neurons (LNs), dorsal neurons (DNs), and lateral posterior neurons (LPNs). Pdf-Gal4 is expressed specifically in PDF-positive large and small ventral LNs (l-LNv and s-LNv, respectively). LNd, dorsal LN; 5^th s-LNv, a single s-LNv not expressing a PDF neuropeptide. (b) CRTC depletion in circadian pacemaker neurons leads to arrhythmic circadian behaviors. crtc RNAi transgene was co-expressed along with the RNAi-enhancing DCR2 by tim-Gal4 or Pdf-Gal4 driver. Rhythmicity in free-running locomotor behaviors in DD cycles was measured similarly as in Fig. 1d. **P < 0.01, ***P < 0.001 to controls heterozygous for Gal4 > DCR2 or UAS-crtc RNAi as determined by one-way ANOVA, Tukey post hoc test. Error bars indicate SEM. (c) SIK2 overexpression in circadian pacemaker neurons by tim-Gal4 or Pdf-Gal4 driver leads to arrhythmic circadian behaviors. ***P < 0.001 to controls heterozygous for Gal4 or UAS-SIK as determined by one-way ANOVA, Tukey post hoc test. Error bars indicate SEM. (d) CRTC overexpression in PDF neurons rescues 24-hour behavioral rhythms in crtc mutants. Averaged actograms in DD cycles were double-plotted with their genotypes on top. (e,f) Circadian periods and rhythmicity in DD locomotor behaviors were measured as above. *P < 0.05, **P < 0.01 and ***P < 0.001 to controls heterozygous for Pdf-Gal4 or UAS-CRTC in wild-type (light-gray bars) or crtc mutants (dark-gray bars) as determined by one-way ANOVA, Tukey post hoc test. Error bars indicate SEM.

Figure 3. crtc mutation phase-delays circadian gene expression in adult fly heads.

(a) Circadian expression of clock mRNAs in adult fly heads of wild-type (gray lines) and crtc mutants (orange lines). Flies were collected at six different time-points in LD cycles (top) or during the transition from the first to the second DD cycle following LD entrainment (bottom) and total RNAs from heads were purified. Relative levels of Clk, per, tim, and Pdp1 mRNAs were quantified by real-time RT-PCR. X-axis indicates zeitgeber time (ZT) in LD (top) or circadian time (CT) in DD (bottom). Y-axis indicates relative expression levels (%) at each time-point, calculated by normalizing to the peak value (set as 100). White/black bars, LD cycles; gray/black bars, DD cycles. Data represent the average of three independent experiments. *P < 0.05, **P < 0.01 and ***P < 0.001 as determined by Student’s t-test. Error bars indicate SEM. (b) Circadian expression of PER and TIM proteins in adult fly heads of wild-type (gray lines) and crtc mutants (orange lines). Head extracts were prepared from flies harvested during LD (top) or the first DD (bottom) cycle and immunoblotted with anti-PER, anti-TIM and anti-TUBULIN (TUB, loading control) antibodies under the same experimental conditions. Images were cropped from full-length blots shown in Supplementary Fig. 12. Protein band intensities in each lane were quantified using ImageJ software and normalized to that of TUB protein. Y-axis indicates the relative expression levels (%) of PER and TIM proteins, calculated by normalizing to the peak value (set as 100). White/black bars, LD cycles; gray/black bars, DD cycles. Data represent the average of three independent experiments. *P < 0.05, **P < 0.01 and ***P < 0.001 as determined by Student’s t-test. Error bars indicate SEM.

Figure 4. crtc mutation impacts on PER and TIM oscillations in circadian pacemaker neurons.

(a) Circadian expression of PER and TIM proteins in circadian pacemaker neurons of wild-type (gray lines) and crtc mutants (orange lines). Adult fly brains were dissected at different time-points in LD cycles (top) or during the transition from the first to the second DD cycle following LD entrainment (bottom). Whole-mount immunostaining was performed using anti-PER, anti-TIM, and anti-PDF antibodies. Confocal brain images were obtained from 12–13 hemispheres at each time-point (X-axis). The fluorescence intensity of anti-PER and anti-TIM antibody staining was quantified from individual neurons using ImageJ software and averaged for each group of circadian pacemaker neurons. Y-axis indicates the relative expression levels (%) of PER and TIM proteins, calculated by normalizing to the peak value in each graph (set as 100). White/black bars, LD cycles; gray/black bars, DD cycles. s-LNv, PDF-expressing small ventral lateral neurons; LNd, dorsal LN. *P < 0.05, **P < 0.01 and ***P < 0.001 as determined by Student’s t-test. Error bars indicate SEM. (b) Representative confocal images of small LNv in wild-type (w¹¹¹⁸) and crtc mutants (crtc^25-3). 1CT/2CT, circadian time in the first/second DD cycle.

Figure 5. timeless is a primary clock target of Drosophila CRTC.

(a) Drosophila S2 cells in 12-well plates were co-transfected with reporter plasmids (50 ng of per-luc or tim-luc; 50 ng of renilla luciferase) and expression vectors for V5-tagged CLK (0.2 ng) and HA-tagged CRTC (0, 50 or 250 ng). Dual luciferase reporter assays were performed 40 hours after transfection. Firefly luciferase activity was first normalized to that of renilla luciferase. Relative fold-activation was then calculated relative to baseline luciferase activity in the absence of any effectors. Data represent the average from three independent experiments. Error bars indicate SEM. n.s., not significant; **P < 0.01, ***P < 0.001 as determined by one-way ANOVA, Tukey post hoc test. (b) Reporter plasmids (50 ng of cre-luc or tim-luc; 50 ng of renilla luciferase) and expression vectors for V5-tagged PKA (5 ng) and HA-tagged CRTC (0, 50 or 250 ng) were cotransfected into S2 cells in 12-well plates. Data represent average +/− SEM (n = 3). ***P < 0.001 as determined by one-way ANOVA, Tukey post hoc test. (c) S2 cells in 12-well plates were pre-incubated with 10 μg of double-stranded RNAs against CRTC (dsCRTC) or EGFP (dsEGFP, control). Forty-eight hours after dsRNA treatment, cells were co-transfected with reporter plasmids (50 ng of per-luc or tim-luc; 50 ng of renilla luciferase) and expression vector for V5-tagged CLK (0.2 and 0.5 ng). Data represent average +/− SEM (n = 3). n.s., not significant, *P < 0.05, **P < 0.01 as determined by Student’s t-test. (d) dsRNA-treated S2 cells were transfected with 10 ng of V5-tagged CLK expression vector. Cell extracts were prepared 2 days after transfection, resolved by SDS-PAGE and immunoblotted for specific proteins under the same experimental conditions. Images were cropped from full-length blots shown in Supplementary Fig. 13. (e) TIM overexpression in PDF neurons rescues circadian behaviors in crtc mutants. Each clock gene was overexpressed in PDF neurons of wild-type or crtc mutants flies. Rhythmicity in DD locomotor behaviors was measured as in Fig. 1d. Data represent average +/− SEM (n = 16–57). n.s., not significant, **P < 0.01 to Pdf-Gal4 controls in crtc mutants as determined by one-way ANOVA, Tukey post hoc test.

Figure 6. A model for the evolution of CRTC-dependent clocks.

In ancestral organisms, environmental time cues such as light or the availability of nutrients might have been directly accessible to circadian clock cells. Timing information could have converged on the regulation of CRTC through various intracellular signaling pathways to modulate CREB-dependent transcription in the earliest transcription-translation feedback loop (TTFL), thereby entraining and sustaining molecular clocks. In poikilothermic Drosophila, environmental changes in light could be cell-autonomously sensed by circadian pacemaker neurons, whereas metabolic cues are instead provided systemically. Accordingly, post-translational regulation of TIM stability has evolved as the primary strategy for clock entrainment by light since the blue-light photoreceptor CRY and the F box protein JETLAG (JET) trigger light-dependent TIM degradation. CREB-dependent transcriptional regulation of the tim promoter, on the other hand, has been conserved from the original TTFL. Given that Drosophila tim is essential for sustaining molecular rhythms, a transcriptional co-activator function of CRTC in CLK/CYC-induced tim transcription contributes to free-running behaviors in DD. In homeothermic mammals, light cues are indirectly transmitted to master pacemaker neurons in the suprachiasmatic nucleus (SCN) by the synaptic input from the retinohypothalamic tract (RHT). Consequently, the mammalian tim homolog became dispensable for clock function whereas the light-sensing activity of CRY homologs has been replaced by their transcriptional repressor function. Instead, mammalian Per took over a role in light entrainment by retaining the CREB/CRTC-dependent transcriptional regulation from the primitive TTFL.