What this is
- This research investigates the role of the neuropeptide pigment-dispersing factor () in regulating the Drosophila circadian clock.
- New fluorescent reporters were developed to monitor transcriptional and post-transcriptional rhythms in individual neurons.
- Findings reveal that enhances clock gene transcription in a time-of-day-dependent manner, influencing the synchrony and stability of molecular clocks.
Essence
- amplifies molecular clock rhythms in Drosophila by upregulating clock gene transcription at night, independent of neuronal activity. This modulation occurs through both cAMP-dependent and cAMP-independent pathways.
Key takeaways
- application at ZT14 significantly upregulates CLK/CYC-mediated transcription and PER levels in specific neurons. This effect is not observed at ZT2, indicating a nighttime-specific response.
- The upregulation of PER-TDT by is independent of neuronal excitability, while CLK/CYC-mediated transcription requires neuronal activity. This bifurcation suggests distinct regulatory pathways for clock gene expression.
- The study confirms that signaling influences the synchrony of pacemaker neurons, enhancing molecular clock oscillations and potentially contributing to phase synchrony in .
Caveats
- The study's findings are limited to specific neuronal subtypes and may not generalize to all clock neurons. Further research is needed to explore 's role across different cell types.
- The experimental conditions, such as the culture environment and genetic background, may influence the observed rhythmicity and period of the fluorescent reporters.
Definitions
- circadian rhythms: Biological processes that display an endogenous oscillation of about 24 hours, influenced by external cues like light and temperature.
- PDF (pigment-dispersing factor): A neuropeptide involved in regulating circadian rhythms by modulating neuronal activity and synchronizing molecular clocks.
Simplified
Results
Novel circadian reporters for live fluorescence imaging
Transcriptional activation from E-box-containing promoters and the feedback transcriptional inhibition by PER comprise the essential core loop of the molecular clock. Post-transcriptional regulatory features that control the stability, activity and subcellular localization of PER are also critical for the generation of circa 24-hr period rhythms. Therefore, to monitor the functioning of the transcriptional and post-transcriptional machinery of the molecular clock at single-cell resolution, we generated two fluorescent reporters: one to monitor CLK/CYC transcriptional activity rhythms and the other to follow PER protein levels and localization (). 23 Fig. 1a
The CLK/CYC transcriptional activity reporter×expresses short-lived yellow fluorescent protein fused to a nuclear localization signal (VENUS-NLS-PEST, VNP)under the control of three tandem repeats of the clock regulatory sequence (CRS), which is the E-box-containing 69-bp enhancer of. The CRS is necessary and sufficient for-like spatial and temporal transcriptional expression, and the CRS trimer has been shown to drive CLK/CYC-dependent, high-intensity and high-amplitude circadian reporter expression. The PER protein reporterwas modified from the previously establishedreporterand was designed to express the N-terminal two-thirds of PER fused with tandem TOMATO (TdT) red fluorescent protein, under the control ofregulatory sequences and the 3′UTR. The N-terminal PER moiety harbours most of the known functional domains, including two PER-ARNT-SIM (PAS) domains and an NLS sequence, but it lacks the CLK/CYC inhibition domain and thus does not replace endogenous PER function. 3 69-VNP per per in vivo per-TdT BG-luc per 24 25 26 27 27
As expected, both molecular clock reporters were expressed rhythmically in adult and larval brain clock neurons (, and). A detailed analysis of the simple circadian circuit of third-instar larvae revealed that 3 × 69-VNP peaked between ZT14 (2 hr after lights off under a 12 hr:12 hr light-dark (LD) cycle) and ZT18, which corresponds approximately to the peak phases ofmRNA accumulation, CLK/CYC binding toE-boxes, and genome-wide CLK/CYC binding. PER-TDT peaked approximately 8 hr after the 3 × 69-VNP peak in the LNvs and DN1s, recapitulating the characteristic lag betweenmRNA and protein accumulation. PER-TDT rhythms in the DN2s were anti-phasic to the LNvs and DN1s, which is consistent with endogenous PER cycling(and S1a). Furthermore, the cycling profile of PER-TDT nuclear-cytoplasmic localization was identical to that of PER (). The cycling expression of both reporters and the rhythms in PER-TDT nuclear/cytoplasmic ratio were largely maintained in DD (). These results establish that the×andreporters provide reliable readouts of the molecular clockwork at a single-cell resolution; 3 × 69-VNP reports CLK/CYC-mediated transcriptional rhythms and also mimicsmRNA rhythms, whereas PER-TDT mimics PER protein oscillations. Fig. 1b,c S1a–d 1 28 29 29 30 Fig. 1c Fig. 1d 31 Fig. S1c and d per per per 3 69-VNP per-TdT per
Next, we tested whether our new reporters can be also used for monitoring rhythms in live preparations. Cultured whole brains of adult flies have been shown to generate rhythms in luciferase reporter activity in a tissue-autonomous manner. We chose to use cultured third-instar larval brains for live imaging, as larval circadian circuits are numerically simpler yet composed of fully differentiated and functional clock neurons. As expected, clock neurons in cultured larval brains displayed circadian rhythms in 3 × 69-VNP and PER-TDT reporter expression, although with a longer average period (~29 hr) (,and). Each neuron expressed varying levels of fluorescence; thus, not all the neurons were detected. Among the detected neurons, 57% of the LNvs, 50% of DN1s and 76% of DN2s showed circadian rhythms in 3 × 69-VNP levels, and 75% of the LNvs showed rhythmic PER-TDT expression. Since the long periods were likely due to the combination of the longer endogenous rhythms of the reporter flies (free-running locomotor rhythms of>::×flies: 25.1 ± 0.13 hr;>::flies: 25.3 ± 0.08 hr), the 3 hr temporal resolution and culture conditions, we limited the use of brain explants to relatively short (<8 hr) fluorescent live-imaging at 1 hr intervals in subsequent experiments. 32 33 34 35 Fig. 2a–c Movies S1 S2 1982Clk mCD8 RFP, 3 69-VNP gal1118 mCD8 Venus, per-TdT
PDF enhances CLK/CYC-mediated transcription at night
The neuropeptide PDF mediates normal circadian locomotor rhythms by affecting the synchrony, amplitude and pace of the molecular clocksand by coordinating the phase of pacemaker neuron activity. PDF activates adenylate cyclase and increases cAMP levels in PDF receptor (PDFR)-expressing neurons. The rise in cAMP levels activates protein kinase A (PKA) and stabilizes TIM and PER, which contributes to the phase resetting and speed control of the molecular clocks. PDF also enhances neuronal activity in a cell-autonomous manner. Importantly, the stabilization of TIM via PDF signalling is activity-independent; therefore, PDF/PDFR signalling at least bifurcates into the pathway controlling neuronal excitability and the one controlling PER/TIM stability. However, the possible effects of PDF signalling on clock gene transcription have not been directly examined. 15 16 18 19 20 36 37 38 14 17 36 39 40 17
To better dissect the effect of PDF signalling on the clockwork, we monitored 3 × 69-VNP and PER-TDT expression in larval brain explants following a bath application of PDF. LD-entrained larvae were dissected at ZT1 or ZT13 and immediately subjected to live imaging in DD. PDF was applied at ZT2 or ZT14 (i.e., 1 hr after the start of live imaging). PDF application at ZT2 did not alter 3 × 69-VNP or PER-TDT expression in any larval clock neuron subtype. In contrast, PDF application at ZT14 upregulated 3 × 69-VNP and PER-TDT levels in the LNvs and DN1s. PDF application did not detectably affect reporter expression in the DN2s (and b). Reporter expression levels continued to rise in the LNvs and DN1s during the course of imaging following the application of PDF at ZT14. This is consistent with the previous finding by Klose. that signalling downstream of PDFR continues to be active as long as PDF is present in the medium inpreparations. As the addition of PDF had no effect on the expression profiles of 3 × 69-VNP and PER-TDT in cultures prepared from thehypomorphic mutant larvae, even at ZT14, the effect of PDF is specific and occurs via PDFR (). Furthermore, PDF upregulated 3 × 69-VNP and PER-TDT levels at ZT14 but not at ZT2 also inmutants (). These results indicate that PDF upregulates CLK/CYC-mediated transcription and PER levels in the LNvs and DN1s at night but not during the day, and this nighttime-specific response of the molecular clock is controlled independently of the timing of PDF release. Fig. 3a 41 Fig. S2a Fig. S3 et al ex vivo pdfr pdf01
To test whether neuronal excitability is involved in the reporter upregulation induced by PDF at ZT14, we co-applied the voltage-gated sodium channel blocker tetrodotoxin (TTX) and PDF to the brain explants. PER-TDT upregulation by PDF was insensitive to TTX (). This finding parallels the observation that PDF stabilizes TIM through a mechanism independent of neuronal activity. In contrast, TTX blocked the PDF-mediated upregulation of 3 × 69-VNP (), indicating that PDF affects CLK/CYC-mediated transcription via time-of-day- and activity-dependent mechanisms. Fig. 3b 17 Fig. 3a
To further test if the upregulation of PER-TDT by PDF is mediated by PKA activation, we applied the cAMP analogue Sp-adenosine-3′, 5′- cyclic monophosphorothioate triethyamine (Sp-cAMPS), which specifically activates PKA, to the larval brain explants at ZT14. Sp-cAMPS did not affect 3 × 69-VNP levels but significantly increased PER-TDT levels (), consistent with the role of cAMP-PKA signalling in PER stabilization. Taken together, these results demonstrate that PDF/PDFR-signalling modulates the molecular clockwork specifically at night via two pathways: one pathway involving an increase in CLK/CYC-mediated transcription in an activity-dependent mechanism and the other involving enhancement of PER stability through cAMP-PKA activation, independent of neuronal excitability. 14 42 Fig. 4 14
Cell-autonomous, activity-dependent upregulation of clock gene transcription by PDF
TTX blocks spontaneous firing generated by cell-intrinsic mechanisms as well as the action potentials triggered by synaptic inputs. To distinguish between these two possibilities, we monitored 3 × 69-VNP expression in dissociated cultured neurons in real time. Cultures were prepared from LD-entrained larvae expressing 3 × 69-VNP and the clock neuron marker (::) at approximately ZT11, then incubated for two days in DD and then subjected to time-lapse imaging starting at CT18. The cultured neurons were morphologically intact, and their neurites continued to grow and elaborate as previously shown, even during the time-lapse imaging (). 43 44 45 Fig. S2b 1982clk-gal4, UAS-mCD8 RFP
All of the>::clock neuron marker-positive cells co-expressed 3 × 69-VNP; however, 3 × 69-VNP levels showed steady increases over time without detectable rhythms. Remarkably, the addition of PDF caused a significant further increase in 3 × 69-VNP levels (,). The addition of PDF had no significant effect on 3 × 69-VNP expression in the culture prepared from themutant larvae (and e), confirming that the effect of PDFis specific and via PDFR, as in brain culture. Furthermore, TTX or the inhibitory neurotransmitter GABA completely inhibited the upregulation of 3 × 69-VNP levels in response to acute PDF treatment (and d). These results indicate that PDF upregulates CLK/CYC-mediated transcription by a cell-autonomous, activity-dependent mechanism. 1982clk mCD8 RFP pdfr in vitro Movie S3 Fig. 5a and d Fig. 5b Fig. 5c
Discussion
Bioluminescence-based circadian reporters allow non-invasive, long-term recording with a high temporal resolution and have been successfully used to monitor rhythmicity in real time. In contrast, fluorescent reporters can achieve the higher spatial resolution necessary to distinguish individual cells in a cluster and even to characterize intracellular protein dynamics. Using two new fluorescent reporters of themolecular clock, we were able to separately interrogate two important aspects of the clockwork: CLK/CYC-mediated transcription from E-boxes and PER intracellular dynamics, both at a single-cell resolution and in real time. 32 33 34 Drosophila
Our study confirms and extends the mechanistic links between network interactions and molecular clockworks, particularly those mediated by PDF signalling. PDF has been shown to trigger multiple downstream effects even within the same neuron. We showed that PDF bath application upregulates PER-TDT in whole-brain cultures, consistent with previous findings indicating that PDF stabilizes PER and TIM via PKA activation independently of neuronal activity. In addition, we demonstrated for the first time that PDF also enhances CLK/CYC-mediated transcription through a cell-autonomous, activity-dependent but cAMP-PKA signalling-independent mechanism. Furthermore, the effect of PDF on PER stability and clock gene transcription occurs at night but not during daytime (,and). 14 17 Figs 3 4 6
PDF signalling occurs via both cAMP-dependent and cAMP-independent pathways in cockroach circadian pacemaker neurons to increase the intracellular Caconcentration. Similarly, we demonstrated here the presence of cAMP-independent pathway inPDF signalling, which leads to the activation of clock gene transcription (). Activity-dependent gene expression triggered by increases in Caconcentration is a widespread phenomenon in the nervous system. Although it is beyond the scope of this study to identify the mechanisms by which PDF leads to the activity-dependent (TTX-sensitive) upregulation of CLK/CYC-transcription, we speculate that Cainflux downstream of PDF/PDFR signalling leads to an upregulation of clock gene transcription. As implied by Agrawal., PDFR-dependent increase in Caconcentration may be mediated by IP/Casignalling. 2+ 2+ 2+ 2+ 2+ 40 Fig. 4a 46 47 Drosophila et al 3
How does PDF confer its time-of-day-dependent effects on the molecular clock? PDF rhythmically accumulates at the s-LNv dorsal termini with a peak around ZT2, and is predicted to be released at a higher rate at ZT2 than at ZT14. However, our observation of ZT14-specific reporter upregulation even in the absence of endogenous PDF () indicates that intracellular timing mechanisms (but not the rhythms in PDF availability) gate the molecular clock transcriptional and post-transcriptional responses to PDF. Rhythms in PDFR expression might be a possible mechanism of time-of-day-dependent PDF response, although the precise daily patterns of PDFR expression are yet to be characterized. Klose. showed that the sensitivity of the s-LNvs to PDF, measured based on the cAMP concentration, exhibits daily rhythms and peaks at dawn. This suggests that cAMP signalling is permissive for the PER stabilization effect of PDF even at its nadir and the mechanisms downstream of cAMP production gate the PER stabilization effect at night. 48 49 Fig. S3 41 et al
The electrical activity of clock neurons is rhythmic and highest from the late night to early morning. As intracellular Calevels in the s-LNvs and DN1s peak at approximately the same time, rhythms in neuronal excitability and Camay gate the transcriptional response to PDF. We showed that TTX inhibits the nighttime-specific upregulation of clock gene transcription via PDF in brain explant cultures (). These results suggest the possibility that synaptic inputs to the LNvs and DN1s control the neuronal activity relevant for gating the transcriptional response to PDF. On the other hand, PDF addition to dissociated neurons uniformly triggered TTX-sensitive transcriptional upregulation despite the lack of rhythms (). This result suggests that the transcriptional upregulation induced by PDF may be the default state, i.e., that the clock may negatively regulate clock gene transcription unresponsive to PDF in the morning. As TTX can also inhibit spontaneous firing occurring independently of synaptic inputs, these results also imply that cell-autonomous control of neuronal firing rhythms by the molecular clock may be involved in the mechanisms gating the response to PDF. Fluorescent circadian reporter imaging combined with neuronal silencing or excitation by genetic tools will facilitate further studies to investigate whether the neuronal activity relevant for the response to PDF is mediated by circuit properties or by cell-autonomous mechanisms. 49 50 51 20 Fig. 3a Fig. 5a,c and d 2+ 2+
The nighttime-specific effect of PDF on clock gene transcription may be related to the mechanism of phase shift. Light increases the firing rate of pacemaker neuronsand likely induces PDF release from the LNvs. It has also been shown that acute induction of firing mimics the effect of light in phase shift. Therefore, an increase in PDF release in the early night by a light pulse when PDF levels should normally be low may trigger phase delay via enhanced clock gene transcription. Our finding that PDF-mediated activation of clock gene transcription is activity-dependent (and) also supports this possibility. 52 12 Figs 3a 5
The time-of-day-dependent modulation of the molecular clockwork by PDF may also enhance the synchronization of clock neurons. CLK/CYC-transcription is active during the late day-early night and reaches a maximum around ZT14. PER maximally accumulates somewhat later, in the middle or near the end of the night. As reported previously, the effect of PDF on PER stabilization likely occurs in the late night to slow down the pace of intrinsically fast-paced clock neurons, such as the s-LNvs and DN1a (same as larval LNvs and DN1s). Furthermore, by stimulating CLK/CYC-mediated transcription in the early night, PDF signalling can amplify molecular clock oscillations, which contributes to phase synchrony (). Interestingly, a computational study predicts that VIP, the functional ortholog of PDF, can mediate synchrony across SCN neurons by inducingexpression only if VIP signalling occurs in-phase withtranscription. Our results strikingly parallel this model and suggest that neuropeptide signalling in-phase with core clock gene transcription is a conserved principle for achieving pacemaker neuron synchrony. 23 14 19 Fig. 6 53 per per
It is noteworthy that reporter upregulation upon PDF application to brain explants was observed in the LNvs and DN1s but not in the DN2s (and). Adult DN2s express PDFR and can be activated by PDF. Although larval DN2s also express PDFR, the signalling downstream of PDFR in these cells may differ from that in the LNvs and DN1s. This finding underscores that PDF has multiple cell-type-specific roles. Figs 3 4 38 54 30
One of our most intriguing observations is the lack of discernible circadian rhythms in 3 × 69-VNP expression in dissociated neuron cultures (). This finding suggests a possibility that robust transcriptional rhythms may require intact network communication and parallels the notion that circuit properties strongly influence the robustness of the mammalian SCN clocks. Fig. 5 6 55
A recent study by Mezan. from Kadener’s group used another fluorescent circadian transcriptional reporters and showed that PDF signalling negatively regulates clock gene expression. The differences in experimental setup and timescale of observations likely contributed to the differences in the conclusions between our and their studies. Whereas we focused on analysing acute response to PDF, Mezan. investigated the effect ofmutation ontranscription at steady-state or following the induction of CLK-GR transgene nuclear localization to increase CLK/CYC-mediated transcription. Before induction, CLK-GR acts dominant negatively to inhibit clock gene transcription, and it takes at least 24 hr to observe the clock genetranscriptional upregulation after CLK-GR induction. It will be interesting to compare short- and long-term effect of PDF on clock gene transcription using both experimental systems in future studies. et al et al pdfr tim tim 56
Cell type-specific differences in PDF signalling, downstream molecular components and relationships with neuronal activity are important issues to be addressed in future studies. There are likely other important neuropeptides and neurotransmitters that link neuronal communication with the molecular clockwork. Our new fluorescent reporters can be combined with optogenetic tools and genetically encoded calcium reporters to manipulate and analyse neuronal activity; however, the period estimated from the fluorescence time-lapse imaging could be imprecise due to its inherent lower temporal resolution compared with bioluminescence recordings. As is often the case in anytransgenic line, we also observed that genetic background affects the period length of the reporter lines independently of the copy number of the fluorescent reporters. Our results also suggest that culture conditions of brain explants impact the rhythmicity and period of the reporters. Given these advantages and caveats, the application of our fluorescent reporters to imaging in intact fly brainsmay offer powerful tools to further decipher the intricate interactions between neuronal signalling and the molecular clockwork. Drosophila 20 57
Methods
Fly strains
were reared at 25 °C on a corn-meal medium under 12 hr:12 hr light-dark (LD) cycles. Theline was provided by N.R. Glossop. The GAL4 enhancer trap line, and,(Bloomington stock center, nb 26654) have been previously characterized. Drosophila 1982clk-gal4 gal1118 pdfr pdf 58 59 60 5304 1
::and::constructs were generated by exchanging the GFP coding sequence of theconstruct provided by L. Luo61tocDNA orcDNA. The resulting constructs were introduced into the genome by P-element-mediated germline transformation. UAS-mCD8 VENUS UAS-mCD8 RFP pUAST/mCD8-GFP Venus mRFP 61
To generate×flies, thebasal promoter sequence from(GenBank accession number), thecoding sequenceand 1 kb of the3′ UTR were PCR-amplified and ligated into the vector containing three tandem copies of the 69-bp CRS of(pCaSpeR-×), thus replacing the luciferase coding sequence and the SV40 3′ UTR. The resulting×vector was used for P-element-mediated germline transformation. 3 69-VNP hs43 pCaSpeR-hs43-lacz Venus-NLS-PEST1 per per per69 3-luc pCaSpeR-3 69-Venus-UTR X81643 24 26
To generate thetransgenic line, thevector was modified by removing the coding sequences of 5xUAS and EGFP-NLS and adding thesequence between the P-element 3′ end and ainsulator. A 13.2 kbgenomic fragment (from thesite at −4.2 kb to thesite located ca. 2 kb downstream of exon 8) was cloned into a modifiedvector. The tdTomato (TdT) coding sequence, amplified from theplasmid (generated by R. Tsien and provided by L. Luo), and a 1 kb3′UTR sequence were cloned into theplasmid. A-fragment from this plasmid containing the′was then ligated to the-fragment of, which resulted in the(exon 1–5)--′fusion construct (vector). Thestrains were generated by integrating thevector into the attP16 (2R) and attP40 (2L) landing sites by PhiC13-mediated site-specific integration. Injections were performed by BestGene Inc. The) and) lines were crossed to recombine both transgenes on the 2chromosome. per-TdT pStinger attB gypsy per BamHI EcoRI pStinger pRSET-B/tdTomato per pBS BamHI PsiI TdT-3 UTR BamHI PsiI pStinger-per13.2 per TdT 3 UTR per-TdT per-TdT per-TdT per-TdT (attP16 per-TdT (attP40 nd
Larval brain culture
The brains of LD-entrained non-wandering L3 larvae were dissected in ice-cold saline solutionat ZT11. The imaginal discs were left intact to prevent any tearing of the tissue. Dissected brains were kept on ice in modified Schneider’s medium (SM)supplemented with 5 mM Bis-Tris (Sigma). Forbrain culture, the brain explants were mounted on a glass-bottom dish (35 mm MatTek petri dish, 20 mm microwell with 0.16/0.19 mm coverglass) in a fibrinogen clot prepared by adding thrombin (bovine thrombin, Sigma) to fibrinogen (bovine fibrinogen, Calbiochem) as previously described. The glass-bottom well was filled with the SMmedium and covered by a Teflon membrane permeable to oxygen. The cultured brains were kept at 25 °C in 80% relative humidity and in the dark for a few hours prior to time-lapse imaging to allow the brains to set. Time-lapse imaging was performed in the same culture conditions, with images acquired every 3 hr. 62 63 64 active active ex vivo
For pharmacological experiments on brain explants, brains were dissected either at ZT1 or at ZT13 and cultured in dissecting saline solution (equivalent to hemolymph-like saline) without a Teflon membrane. The brain explants were immediately imaged to establish a baseline. Approximately 1 hr after the start of the time-lapse imaging, the neuropeptide PDF (custom-made by Chi Scientific, H-NSELINSLLSLPKNMNDA-OH; 2 μM) was added alone or in combination with 100 nM TTX (Cayman Chemical). For the control condition, the vehicle (DMSO) was added. Sp-cAMPS (sc-201571 from Santa Cruz Biotechnology, 150 μM) was added alone and the vehicle (HO) was used as a control. Time-lapse imaging was performed in the same conditions as described above with images acquired every hour for 8 hr. 2
Primary neuron culture
Dissociated neuron culture was performed as previously describedwith the following modifications. In brief, the dissected brains were enzymatically treated with 50 units/mL papain (Worthington) and mechanically dissociated. The cell suspension was then plated on glass-bottom dishes (35 mm MatTek petri dish, 10 mm microwell with 0.16/0.19 mm coverglass) coated with concanavalin A (Sigma). Once the cells were attached to the glass bottom, the dish was flooded with SM. The dissociated neuron culture was incubated at 25 °C in 80% relative humidity and constant darkness for 2 days prior to time-lapse imaging. For pharmacological experiments, 2 or 20 μM PDF, 100 nM TTX, 10 μM GABA (Sigma), or the vehicle (DMSO or ddHO) was added to the cell culture medium just before the start of the time-lapse imaging. Time-lapse experiments were conducted in the same culture conditions, and the images were captured every 3 hr. 45 64 active 2
Microscopy and image analysis
A Leica TCS SP5 tandem scanner confocal microscope was used for fluorescence imaging. The same parameter settings were used to image all samples of the same type (dissected brains, cultured brains or cultured neurons). Freshly dissected larval and adult brains were scanned using a 40x water-immersion objective with the galvo scanner at 400 Hz. For time-lapse imaging, 20x, 40x or 63x objectives were used, and the images were acquired using the resonant scanner at 8,000 Hz with high-sensitivity HyD detectors. A bi-directional scan was used together with an 8x line average. Z-section steps of 1.7 μm × 8 and 2 μm × ~40 were used for imaging the dissociated cultured neurons andbrain culture, respectively. A 514-nm laser was used to excite the VENUS fluorophore (0.68 μW/cmto image VNP, 0.40 μW/cmfor mCD8::VENUS), and a 561-nm laser was used for the TdTomato and mRFP fluorophores (14.8 μW/cmfor PER-TDT, 5.71 μW/cmfor mCD8::RFP). The laser intensity was measured at the level of the sample with a microscope slide power meter (Thorlabs, S170C). ex vivo 2 2 2 2
The raw data of the larval and adult brain images taken at different time points and the time-lapse movies of thebrain cultures were analysed with FIJI software. Briefly, a SUM-stack containing the clock neuron cluster of interest was constructed, and the area of each neuron was manually determined. The mean fluorescence intensity of the defined area in the SUM-stack was measured. Three nearby areas were also measured to analyse the background fluorescence level. The corrected total relative intensity of each cell was calculated as follows: ex vivo 65
The fluorescence intensities in the time-lapse movies of the cultured neurons were measured using Imaris software (Bitplane). A 3D mask (region of interest in 3D) was built for each cell by thresholding the 3 × 69-VNP or PER-TDT fluorescence levels after background subtraction. The intensity SUM in each 3D mask was then extracted from the statistical data that were automatically generated by the program. For presentation purposes only, some of the images and time-lapse movies were processed with a 10x iterative deconvolution using AutoQuant (MediaCybernetics) and Imaris. When necessary, time-lapse images were also treated for 3D correction drift using the ImageJ plugin (NIH).
Time series data analysis
Fluorescence-intensity time-series data were normalized to the value at t = 0. Heatmaps representing the fluorescence-intensity time course were generated with an in-house R script. Each row represents the data of a single cell. Rows were ordered by the highest intensity over 24 hr.
For rhythm analysis, a combination of manual inspection and Maximum Entropy Spectral Analysis (MESA) was used to detect rhythmicity and estimate the period. MESA was chosen because, unlike other methods, it is adapted for the detection of rhythms in short or noisy time series. First, time course plots of the intensity were generated using Excel or Prism (GraphPad Prism version 6.0c for Mac, GraphPad Software, San Diego, California, USA,). Because we found that 6-order polynomial regression models best fit the data, we superimposed the 6-order polynomial trend lines on the graphs to facilitate the detection of rhythmicity. The time-series data with circadian rhythms, defined by a peak-to-peak interval from 18 to 36 hr, were then identified by manual inspection. However, the polynomial trend lines were not used to determine the period. In parallel, the intensity time-series data without normalization were analysed with MESA without filtering. The time-series data were scored as circadian only when both the manual and MESA analyses detected the rhythms with a period between 18 and 36 hr. 66 www.graphpad.com↗ th th
Statistical analysis
Statistical analyses were performed with GraphPad Prism software. To compare the effects of the drug treatments in the pharmacological experiments (–), a two-way ANOVA with a Sidak test to correct for multiple comparisons was used. Figs 3 5
Additional Information
: Sabado, V.. Fluorescence circadian imaging reveals a PDF-dependent transcriptional regulation of themolecular clock., 41560; doi: 10.1038/srep41560 (2017). How to cite this article 7 et al Drosophila Sci. Rep.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher's note: