Misregulation of Drosophila Myc Disrupts Circadian Behavior and Metabolism.

Drosophila Myc (dMyc) is highly conserved and functions as a transcription factor similar to mammalian Myc. We previously found that oncogenic Myc disrupts the molecular clock in cancer cells. Here, we demonstrate that misregulation of dMyc expression affects Drosophila circadian behavior. dMyc overexpression results in a high percentage of arrhythmic flies, concomitant with increases in the expression of clock genes cyc, tim, cry, and cwo. Conversely, flies with hypomorphic mutations in dMyc exhibit considerable arrhythmia, which can be rescued by loss of dMnt, a suppressor of dMyc activity. Metabolic profiling of fly heads revealed that loss of dMyc and its overexpression alter steady-state metabolite levels and have opposing effects on histidine, the histamine precursor, which is rescued in dMyc mutants by ablation of dMnt and could contribute to effects of dMyc on locomotor behavior. Our results demonstrate a role of dMyc in modulating Drosophila circadian clock, behavior, and metabolism.

INTRODUCTION

Circadian rhythms are 24-h cycles of sleep-wake, fasting-feeding, body temperature, and metabolism (Asher and Schibler, 2011; Bass and Takahashi, 2010; Sahar and Sassone-Corsi, 2012). Genetic regulation of circadian rhythms by the clock, which is found broadly from cyanobacteria to humans, was first discovered in Drosophila and can be entrained by external cues such as light and food (Dubowy and Sehgal, 2017). The molecular clock consists of transcription-translation feedback loops, in which clock genes autoregulate their own expression. In the major loop, the Drosophila period (per) and timeless (tim) genes encode proteins that rhythmically inhibit activity of their transcriptional activators, CLOCK (CLK) and CYCLE (CYC), to maintain cycles of gene expression (Dubowy and Sehgal, 2017). Light entrains this loop via degradation of TIM by Cryptochrome (CRY), a light-sensitive photoreceptor (Koh et al., 2006; Peschel et al., 2006). In a second feedback loop, CLK drives expression of its own activator and repressor, PDP1 and VRILLE (VRI), respectively. Transcriptional control also includes the Clockwork orange (CWO) protein, which contributes to repression by PER and TIM (Zhou et al., 2016).

The molecular clock in Drosophila is expressed in the large and small ventrolateral neurons (lLNvs and sLNvs, respectively), which serve as the central clock regulators similar to the suprachiasmatic nucleus (SCN) in mammals (Rieger et al., 2006). Of these, the sLNvs are crucial in maintaining circadian rhythms in constant darkness (Grima et al., 2004; Stoleru et al., 2005). Both lLNvs and sLNvs express neuropeptide pigment-dispersing factor (PDF), which synchronizes clock neurons in the fly brain. Loss of PDF desynchronizes clock neurons and alters circadian locomotor behavior (Hyun et al., 2005; Renn et al., 1999; Yoshii et al., 2009; Zhang et al., 2010).

CLK and CYC are basic helix-loop-helix (HLH) proteins that activate transcription of target genes by binding to specific DNA sequences, E boxes. Other HLH proteins include the MYC onco-protein, which transcriptionally orchestrates cell growth, cell cycle, and metabolism (Hsieh et al., 2015; Stine et al., 2015), and has a conserved role in Drosophila (Demontis and Perrimon, 2009; Gallant, 2013; Grewal et al., 2005). dMyc overexpression in Drosophila S2 cells stimulates glycolysis and suppresses oxidative phosphorylation (de la Cova et al., 2014). Ectopic expression of Drosophila Myc (dMyc) in flies results in larger body size (de la Cova et al., 2004) associated with increased cell size (Johnston et al., 1999) or apoptosis (de la Cova et al., 2004; Montero et al., 2008; Moreno and Basler, 2004). Conversely, loss of dMyc function underlies the diminutive phenotype with smaller cell and organismal body size (Pierce et al., 2004).

MYC overexpression in mammalian cancer cells can suppress oscillation of BMAL1 (homolog of Drosophila CYC) by inducing the BMAL1 repressor, REV-ERBα (Altman et al., 2015), and inhibiting MIZ-1 (Altman et al., 2017; Shostak et al., 2016). However, the role of dMyc in Drosophila circadian behavior, molecular clock, and metabolism is unknown. Here, we demonstrate that misregulation of dMyc expression perturbs circadian locomotor behavior, expression of clock genes, and metabolism in flies. Additional dMnt loss is able to reverse alterations of behavior, gene expression, and specific metabolite levels in hypomorphic dMyc flies. Our results demonstrate a role of dMyc in modulating Drosophila circadian locomotor behavior potentially through directly regulating components of the core molecular clock, metabolism, as well as an effect on the clock output PDF.

RESULTS

Overexpression of dMyc Disrupts Circadian Locomotor Activity

To study the effect of dMyc overexpression, we ectopically expressed dMyc with three different drivers (tim-Gal4, Pdf-Gal4, cry24Pdf-Gal4) using the GAL4/UAS (upstream activating sequence) system. To counter possible apoptotic effects of dMyc, p35, the Drosophila homolog of BCL2, was co-expressed (Greer et al., 2013). We examined transcript levels of clock genes in whole-head extracts of tim-Gal4 > UAS-Myc+p35 flies, compared to Gal4 control flies (tim-Gal4 > +) and flies overexpressing p35 (tim-Gal4 > UAS-p35). We found that dMyc mRNA is expressed rhythmically and peaked at zeitgeber time 14 (ZT14) in Gal4 control flies (tim-Gal4 > +) (Figure S1A, dMyc inset), suggesting that dMyc, similar to mammalian Myc, may be clock-regulated (Fu et al., 2002). Similar to mammalian Myc (Altman et al., 2015), overexpressed dMyc increased expression of several clock genes, including cyc, cwo, per, tim, and vri (Figure S1A). Immunoblots of whole-head lysates also showed a significant increase in PER protein levels (Figure 1A). But unlike mammalian Myc, dMyc did not suppress the BMAL1 (ARNTL) homolog Clk or cyc mRNA (Figure S1A). To evaluate whether dMyc directly binds clock genes, we analyzed dMyc chromatin immunoprecipitation-sequencing (ChIP-seq) data from Drosophila melanogaster Kc cells (Yang et al., 2013) and found endogenous dMyc binding at the promoters of cyc, cwo, pdp, and vri, but not of per and tim (Figure S1C).

Figure 1.

PER Immunoblot, Circadian Locomotor Rhythms, and PDF Expression of sLNvs in dMyc-Overexpressing Flies

(A) dMyc and PER protein levels determined by immunoblot in heads of tim-G > Myc+p35 flies compared to tim-G and + > Myc+p35 flies. Flies were entrained in light/dark 12:12-h cycle for 3–5 days. On the last day of entrainment, flies were snap frozen every 4–6 h and the heads were used to extract protein. α-Tubulin (α-Tub) is the loading control. Data are representative of three or more biological replicates. (B) Representative actogram of cry24Pdf-G, cry24Pdf-G > p35, cry24Pdf-G > Myc+p35 rhythmic (RR), cry24Pdf-G > Myc+p35 weakly rhythmic (WR), and cry24Pdf-G > Myc+p35 arrhythmic (AR). Flies were entrained to a light/dark cycle for 3 days before being monitored in constant darkness over 8 days.

(C and D) Decreased PDF expression in dorsal projections from sLNvs in dMyc-overexpressing flies with intact neuronal processes. Representative confocal image of brains from control flies (cry24Pdf-G > p35) and dMyc-overexpressing flies (cry24Pdf-G > Myc+p35) with membrane-labeled GFP (mCD8-GFP) (C) or without mCD8-GFP (D) subjected to immunofluorescence staining using antibodies directed against PDF (red in C; green in D) and PER (red in D) proteins at ZT1 and ZT13. A total of 12 brains from control flies and 34 brains from dMyc-overexpressing flies with mCD8-GFP were dissected, and the observations were consistent. Rectangles in (C) and arrows in (D): dorsal projections from sLNvs. Scale bars, 40 μm.

See also Figures S1, S2, and S5, and Table 1.

Given that dMyc increased expression of several clock genes, we then sought to determine the effects of dMyc overexpression on fly behavior. We assayed locomotor rhythms under constant darkness for flies with dMyc driven by different promoters: tim-Gal4 driving dMyc and p35 in all clock cells, Pdf-Gal4 in ventral lateral neurons, or cry24Pdf-Gal4 in ventral lateral neurons and cells expressing cryptochrome (Baik et al., 2017). dMyc overexpression in all clock cells (tim-Gal4 > Myc+p35) resulted in a lower percentage of rhythmic flies (30%) compared to the Gal4 control (56%) (Table 1; Figures S2C and S2D). Additional control with p35 overexpression was performed in tim-Gal4 flies with isogenic background. As expected, dMyc overexpression in tim-Gal4 (w/Y;tim-G/+;+/UAS-p35 versus w/Y;tim-G/+;+/UAS-Myc, UAS-p35) also showed a moderate effect on rhythmicity (91% versus 77% rhythmic, respectively; Table 1). Restricted dMyc expression in PDF-positive ventral lateral neurons (Pdf-Gal4 > Myc+p35) was similarly associated with weaker rhythms, with 87% of the flies rhythmic compared to 100% rhythmic in control flies. Rhythm strength was also significantly decreased in rhythmic flies (Table 1; Figures S2A and S2B). Notably, use of both cry24-Gal4 and Pdf-Gal4 to drive dMyc expression (cry24Pdf-G > Myc+p35) resulted in arrhythmia in about 70% of the flies and weaker rhythm in the rhythmic flies (Table 1; Figure 1B). These observations indicate that dMyc overexpression perturbs Drosophila circadian behavior. Perturbation of rhythmic behavior, however, is not always associated with the same changes in clock gene expression. Although head extracts of cry24Pdf-G > Myc+p35 flies showed overall higher levels of cry, cyc, cwo, and tim transcripts (Figure S1B), the oscillation and expression of other clock genes were largely unchanged.

Table 1.

Circadian Locomotor Rhythmicity of Flies with Misregulated dMyc Expression

Genotype[a]	n	No. Rhythmic	% Rhythmic	Period (h) ± SEM	FFT ± SEM[b]
Circadian Locomotor Rhythmicity of dMyc-Overexpressing Flies in Constant Darkness
w/Y;+;UAS-Myc, UAS-p35/+	32	32	100	23.80 ± 0.04	0.04 ± 0.004
w/Y;Pdf-G/+;Pdf-G/+	32	32	100	24.01 ± 0.04	0.06 ± 0.005
w/Y;Pdf-G/+;Pdf-G/UAS-Myc, UAS-p35	30	26	87	24.13 ± 0.10	0.02 ± 0.002[c]
cry24-G/Y;Pdf-G/+;+	31	31	100	24.49 ± 0.05	0.10 ± 0.008
cry24-G/Y;Pdf-G/+;+/UAS-p35	28	25	89	24.60 ± 0.05	0.04 ± 0.004
cry24-G/Y;Pdf-G/+;+/UAS-Myc, UAS-p35	26	8	31	24.46 ± 0.07	0.03 ± 0.004[d]
w/Y;TUG/+;+	32	32	100	24.49 ± 0.04	0.05 ± 0.005
w/Y;TUG/+;+/UAS-Myc, UAS-p35	32	32	100	23.97 ± 0.06[e]	0.07 ± 0.009[f]
yw/Y;tim-G/+;+	54	30	56	25.32 ± 0.11	0.01 ± 0.001
yw/Y;tim-G/+;+/UAS-Myc, UAS-p35	69	21	30	24.77 ± 0.07[g]	0.01 ± 0.001
w/Y;tim-G/+;+/UAS-p35	32	29	91	24.52 ± 0.08	0.04 ± 0.004
w/Y;tim-G/+;+/UAS-Myc, UAS-p35	64	49	77	24.06 ± 0.05[h]	0.04 ± 0.003
Circadian Locomotor Rhythmicity of dMyc Mutant Flies in Constant Darkness
w/Y; + ; +	28	24	86	24.01 ± 0.06	0.05 ± 0.004
dmP0/Y; + ; +	32	13	41	23.91 ± 0.03	0.03 ± 0.005[i]
dmP0/+; bw1/+ ; st1/+	31	29	94	24.17 ± 0.07	0.066 ± 0.008
dmP0/dm4; bw1/+ ; st1/+	30	9	30	24.03 ± 0.17	0.028 ± 0.004[j]
dmP0/dm4 dmnt1; bw1/+ ; st1/+	30	19	63	23.84 ± 0.07[k]	0.045 ± 0.006[l]

w: wISO31 isogenic white color eye mutant; yw: y1,w1118 yellow color body and white color eye mutant; tim-G: tim-Gal4; TUG: tim-(UAS)-Gal4; cry24-G: cry-Gal4 (line 24); Pdf-G: Pdf-Gal4; bw1: brown color eye mutant; st1: light orange eye mutant.

FFT, fast Fourier transform. Flies with FFT value > 0.01 are counted as rhythmic.

p < 0.005 compared to w/Y;Pdf-G/+;Pdf-G/+ control, by Student’s t test.

p < 0.05 compared to cry24-G/Y;Pdf-G/+;+ control, by Student’s t test.

p < 0.005 compared with w/Y;TUG/+;+ control, by Student’s t test.

p < 0.05 compared with w/Y;TUG/+;+ control, by Student’s t test.

p < 0.005 compared with yw/Y;tim-G/+;+, by Student’s t test.

p < 0.005 compared with w/Y;tim-G/+;+/UAS-p35, by Student’s t test.

p < 0.01 compared with w/Y; + ; + control, by Student’s t test.

p < 0.005 compared with dmP0/+; bw1/+ ; st1/+ control, by Student’s t test. p < 0.05 compared with dmP0/dm4 dmnt1; bw1/+ ; st1/+, by Student’s t test.

p < 0.005 compared with dmP0/+; bw1/+ ; st1/+ control, by Student’s t test.

p < 0.05 compared with dmP0/+; bw1/+ ; st1/+ control, by Student’s t test.

Our results from using three different drivers suggest that where, rather than how much dMyc is overexpressed, affects circadian behavior. With up to 500-fold increase in dMyc transcript and significant elevation of dMyc protein in tim-G > Myc+p35 flies, the percentage of rhythmic flies only dropped from 56% to 30% (Figure S1A; Table 1). However, cry24Pdf-G > Myc+p35 flies showed a severely arrhythmic phenotype, although dMyc transcripts were only 3-fold higher than controls (Figure S1B; Table 1). These results suggest that dMyc-mediated disruption of circadian behavior in flies may be dependent on where dMyc is expressed.

Previously, we documented that MYC suppresses BMAL1 through upregulation of REV-ERBα (Altman et al., 2015). Although the Drosophila REV-ERBα homolog Ecdysone-induced protein 75B (E75) suppresses Clk, the BMAL1 homolog (Kumar et al., 2014), we found that dMyc did not change the levels of four mRNA isoforms of E75 in cry24Pdf-G > Myc+p35 flies (Figure S2E).

dMyc-Overexpressing Flies Show Reduced PDF Expression but Preserved Axonal Morphology at the Dorsal Termini of Central Pacemaker Neurons (sLNvs)

Integrity of sLNvs and their secretion of the PDF neuropeptide are essential for rhythmic behavior in Drosophila, such that loss of Pdf neurons or expression result in arrhythmic flies (Renn et al., 1999). PDF expression in the dorsal axon of sLNvs is particularly relevant, with rhythmic axonal fasciculation observed at the terminus of this axon (Fernández et al., 2008). Intriguingly, overexpression of another E-box transcriptional factor Mef2 alters circadian sLNv dorsal axonal fasciculation and causes arrhythmic Drosophila behavior (Sivachenko et al., 2013). We thus asked whether ectopic dMyc alters sLNv morphology or PDF expression, rendering cry24Pdf-G > Myc+p35 flies arrhythmic.

We visualized PDF in the cell bodies and axons by immunohistochemistry (IHC) at two different time points—ZT1 (an hour after lights on) and ZT13 (an hour after lights off). While there was no significant change in PDF staining in the cell bodies of both lLNvs and sLNvs, PDF expression was reduced in the dorsal projection of sLNvs in cry24Pdf-G > Myc+p35 flies compared to controls (Figure 1D, arrowheads). To investigate whether reduced PDF expression is associated with dysmorphic dorsal projection, we expressed membrane-labeled GFP (mCD8-GFP) under the control of a Pdf- and cry-specific promoter in dMyc-overexpressing or control flies. The mouse lymphocyte transmembrane protein (mCD8) fused to GFP labels the cell surface, particularly axons and dendrites due to their high surface area (Lee and Luo, 1999). Although cry24Pdf-G > Myc+p35 flies exhibit less intense PDF staining at dorsal termini, the mCD8-labeled axons and their fasciculation are preserved (Figure 1C, rectangle), suggesting that dMyc overexpression does not affect axonal morphology but, instead, reduces PDF expression in the dorsal projection of sLNv cells. Because loss of PDF expression in the dorsal projection has been observed in Clk and cyc mutant flies (Park et al., 2000), we surmise that ectopic dMyc alters CLK-mediated PDF expression and contributes to dysrhythmia in these flies.

Hypomorphic dMyc Flies Exhibit Arrhythmic Behavior

Per2 is reported to regulate endogenous mouse Myc expression (Fu et al., 2002). However, whether endogenous Myc affects the circadian clock is not known. To investigate whether loss of dMyc affects Drosophila circadian locomotor behavior, we used hypomorphic dMyc alleles, dmP0 (P-element insertion 100 bp upstream of the transcription start site of dMyc) and dm4 (two exon deletion dMyc-null allele) (Figure 2A) (Pierce et al., 2004). Male dmP0/Y flies are viable but exhibit smaller body size (Figure 2B), and dm4 mutations are lethal at the early larval stage (Johnston et al., 1999). Assay of circadian locomotor behavior of dmP0/Y flies in constant darkness revealed robust rhythms in 86% of the sibling controls (w/Y), but only in 41% of dMyc mutants, with other mutant flies showing weak rhythms or arrhythmia (Table 1; Figures 2C–2E). These data demonstrate that endogenous dMyc function is necessary for normal circadian behavior.

Figure 2.

Hypomorphic dMyc Flies Exhibit Arrhythmic Behavior

(A) Schematic illustration (modified from Pierce et al., 2004) of the dmP0 and dm4 alleles of the dMyc gene.

(B) Picture of a dmP0/Y fly versus its sibling control w/Y.

C) Representative actogram of w/Y, rhythmic (RR) dmP0/Y, and arrhythmic (AR) dmP0/Y flies. Flies were entrained to a light/dark cycle for 3 days before being monitored in constant darkness over 7 days.

(D and E) Period length (D) and rhythm strength (E) of rhythmic dmP0/Y flies compared with control flies w/Y.

(F) Picture of dmP0/+, dmP0/dm4, and dmP0/dm4dMnt1 flies.

(G) Representative actogram of dmP0/+, rhythmic (RR) dmP0/dm4, arrhythmic (AR) dmP0/ dm4, and dmP0/ dm4dMnt1.

(H and I) Period length (H) and rhythm strength (I) of dmP0/+, rhythmic dmP0/dm4 (dmP0/dm4), and dmP0/ dm4dMnt1 (dmP0/dm4dmnt) flies. Data are represented as mean ± SEM. *p < 0.05 and **p < 0.001 by Student’s t test.

See also Figures S3 and S4, and Table 1.

dMnt Mutation Rescues Arrhythmicity in dMyc Mutant Flies

To determine whether dMyc transcriptional activity is involved in rhythmic behavior, we used flies mutant for MNT (dMnt; a homolog of mammalian MNT), a transcription repressor that antagonizes MYC function by dimerizing with the MYC-binding partner MAX (Hooker and Hurlin, 2006). Loss of dMnt in dMyc-null flies rescues growth arrest from the early larval to pupal stage (Loo et al., 2005; Pierce et al., 2008). Further knockdown of Mnt compensates for loss of Myc and accelerates cell proliferation in Myc-null cells, presumably by relieving transcriptional repression (Nilsson et al., 2004).

Therefore, we asked whether loss of dMnt is able to rescue the arrhythmic phenotype in dMyc hypomorphs. Because dm4/Y and dm4dmnt1/Y are inviable as adult males, we crossed male dMyc hypomorphic flies (dmP0/Y) with either wild-type females (wISO31), females heterozygous for a dMyc-null allele (dm4/FM7), or females carrying a dMyc-null allele along with a dMnt mutation (dm4dmnt1/FM7) to generate female heterozygotes dmP0/+, dmP0/dm4, and dmP0/dm4dmnt1, respectively. Consistent with the smaller body size observed in dmP0/Y flies, dmP0/dm4 flies also have smaller body size compared to dmP0/+ flies (Figure 2F). Strikingly, the smaller body size was reversed in dmP0/dm4dmnt1 flies that also carry the dMnt mutation (Figure 2F), suggesting loss of the dMyc suppressor dMnt is sufficient to de-repress dMyc-driven growth pathways.

Consistent with the arrhythmic behavior we observed in dMyc hypomorphic male flies (dmP0/Y), dmP0/dm4 flies exhibit substantial arrhythmicity compared to dmP0/+ flies (30% rhythmic versus 94%) (Table 1; Figures 2G–2I). Notably, among the 30% rhythmic dmP0/dm4 flies, the rhythm strength was also significantly reduced (Table 1; Figure 2I). Strikingly, loss of dMnt in the dm4 background rescued the rhythmicity from 30% to 63% (Table 1). Furthermore, reduced rhythm strength in rhythmic dmP0/dm4 flies was rescued in dmP0/dm4dmnt1 flies (Table 1; Figure 2I). Of note, there is no statistically significant difference in period length between w/Y and dmP0/Y flies or dmP0/+ and dmP0/dm4 flies (Table 1; Figures 2D and 2H), suggesting altered endogenous dMyc activity has minimal effect on period length.

Endogenous dMyc Activity Affects Molecular Oscillations of per and tim

We explored the mechanism of arrhythmic behavior in dmP0/dm4 flies by determining whether oscillation of per and tim is altered in brains of these flies (Zeng et al., 1996). As expected under constant darkness, dMyc mutants showed significantly lower dMyc transcript levels and dampened oscillations (Figure S3A). Loss of dMnt in hypomorphic dMyc background (dmP0/dm4dmnt1) did not affect the dMyc transcript level nor rescue dMyc oscillations (Figure S3A). Next, we examined whether decreased dMyc activity affects the cycling of per and tim mRNA levels. In dmP0/+ flies, per and tim mRNA have an oscillation pattern similar to that of wild-type controls with a peak at circadian time 10 (CT10) (CT reflects time in constant darkness conditions and roughly corresponds to the equivalent ZT times in 24-h light/dark cycles) (Jang et al., 2015). In dmP0/dm4 flies, instead of peaking at CT10, both per and tim mRNA levels display a trough at CT10 (Figures S3B and S3C). Notably, the peak of per and tim transcripts at CT10 re-emerged in dmP0/dm4dmnt1 flies.

We assayed the protein levels of dMyc and PER in the whole heads of dMyc hypomorphic flies dmP0/+, dmP0/dm4 at ZT2, ZT8, ZT14, and ZT20 during entrainment and maintained in a light/dark cycle (Figure S3D). dMyc was not detected, and PER protein expression was not significantly changed between dmP0/dm4 flies versus dmP0/+ flies (Figure S3D). We then assayed the expression of PER proteins in the whole brains of dmP0/+, dmP0/dm4, and dmP0/dm4dmnt1 flies, based on their rhythmicity, at CT2, CT6, CT10, CT14, CT18, and CT22 in constant darkness. Consistent with the mRNA expression profile, PER was lower in dmP0/ dm4 flies at times of peak expression in wild-type controls (dmP0/+) (Figures S3E and S3F). PER levels were rescued in dmP0/dm4dmnt1 flies, corroborating the observation with per transcripts (Figure S3B). Interestingly, among the dmP0/dm4 mutants, arrhythmic flies appear to have even lower PER expression than rhythmic flies (Figures S3E–S3G), suggesting that altered PER expression in hypomorphic dMyc flies may be linked to arrhythmicity.

Oscillation of PER Expression Is Altered in sLNvs of dMyc Hypomorphic Flies

As mentioned, ventral lateral neurons are central clock cells in Drosophila that are crucial for maintaining rhythmic behavior. To investigate whether clock protein oscillation is preserved in ventral lateral neurons, particularly in arrhythmic flies, we compared PER protein levels in dissected fly brains from rhythmic dmP0/+, dmP0/dm4, arrhythmic dmP0/dm4 and dmP0/dm4dmnt1 flies by immunohistochemistry. We focused on sLNvs, because they have robust PER oscillation, with nuclear PER expression reaching its peak at CT2 and trough at CT14. Although not statistically significant, nuclear PER accumulation in sLNvs was reduced in dmP0/dm4 flies relative to dmP0/+ flies at CT2 (Figures S4A and S4B). The reduced PER nuclear accumulation was rescued in dmP0/dm4dmnt1 flies (Figures S4A and S4B). In contrast, nuclear PER expression in sLNvs was increased at CT8 in dmP0/dm4 flies relative to dmP0/+ flies (Figures S4A and S4B). The elevated PER nuclear expression was similarly reversed in dmP0/dm4dmnt1 flies (Figures S4A and S4B). This observation echoes the oscillation pattern of per transcript in whole brain where at CT10, instead of reaching the peak as in dmP0/+ flies, the level is reduced in dmP0/dm4 flies (Figure S3B). Altogether, these data demonstrate that loss of endogenous dMyc in Drosophila perturbs per oscillation and disrupts circadian behavior, which can be rescued by further deletion of dMnt.

Overexpressed dMyc Alters Metabolism in Fly Heads

Mammalian Myc is known to orchestrate a number of metabolic pathways in proliferating cells (Hsieh et al., 2015; Stine et al., 2015). As alteration of cellular energetic status, such as the ratio of NAD+/NADH or ATP/AMP, provides feedback to the circadian clock (Lamia et al., 2009; Rutter et al., 2002), we hypothesized that overexpression of dMyc alters metabolism in flies, which in turn contributes to increased arrhythmic behavior. To test this idea, metabolites in cry24Pdf-G > Myc+p35 and control fly heads were extracted and analyzed by mass spectrometry. While the majority of amino acids did not change with dMyc overexpression, we noted that the relative enrichment of valine, L-cystathionine, and sarcosine was significantly reduced (Figure 3A). In addition, significant increase of histidine was observed in cry24Pdf-G > Myc+p35 flies compared to controls (Figure 3A). Furthermore, glutamine was slightly lowered with dMyc overexpression. These observations indicate that dMyc affects metabolism broadly in the fly head even when its expression is limited just to cry- and Pdf-expressing cells.

Figure 3.

Metabolic Alteration in dMyc-Overexpressing and dMyc Hypomorphic Flies

(A) Relative enrichment of metabolites detected by gas chromatography-mass spectrometry (GC-MS) compared with Nor-leucine extracted from heads of control flies (cry24Pdf-G > p35) or dMyc-overexpressing flies (cry24Pdf-G > Myc+p35). Each data point is averaged from four pooled head extracts. Data are represented as mean ± SEM. *p < 0.05 and **p < 0.001 by Student’s t test.

(B and C) Heatmap of metabolites (log value with normalized value between 0 and 1), detected by liquid chromatography-mass spectrometry (LC-MS) extracted from heads of dmP0/+ (+), dmP0/dm4 (dm4), and dmP0/dm4dMnt1 (dm4dMnt1) flies at ZT3 (B) and ZT15 (C).

(D) Model for the effect of dMyc and dMnt on the Drosophila molecular clock and locomotor activity. Schematic diagram of how dMyc and dMnt act on E-box-driven dMyc target genes (top). The bottom shows per/tim transcript oscillation (second column) and fly actogram (third column) in flies with dMyc overexpression (a), normal level of dMyc (b), low level of dMyc (c), and low level of dMyc with loss of dMnt (d).

See also Figure S5.

Loss of dMyc Alters Fly Head Metabolites in a Time-Dependent Manner that Is Partially Rescued by Further Loss of dMnt

Given the metabolic dysregulation observed in dMyc-overexpressing flies, we investigated whether there are comparable metabolic alterations in dMyc hypomorphic flies, particularly those metabolites or metabolic pathways that can be rescued by loss of dMnt. Flies (dmP0/+, dmP0/dm4, and dmP0/dm4dmnt1) were harvested at two times, ZT3 and ZT15, after light/dark entrainment. We calculated the ratio of the difference of metabolites between dmP0/dm4 flies versus dmP0/+ flies and dmP0/dm4 flies versus dmP0/dm4dmnt1 flies [(dmP0/+ – dmP0/dm4)/(dmP0/dm4dmnt1 – dmP0/dm4)] at ZT3. We searched for metabolites that were significantly reduced in the dMyc hypomorph flies but rescued by loss of dMnt by focusing on metabolites that exhibited a ratio between 0.5 and 1.5. We subsequently ranked the metabolites from the ratio closer to 0.5, representing the largest magnitude of rescue from loss of dMnt (Figures 3B and 3C). A heatmap was graphed based on the log value of each metabolite that was further normalized to yield a value between 0 and 1. At ZT3, selected amino acids (histidine, leucine/isoleucine, glutamate, tyrosine), neurotransmitters (glutamate, gamma-aminobutyric acid [GABA]), tricarboxylic acid cycle (TCA) cycle intermediates (citrate/isocitrate, fumarate, alpha-ketoglutarate), and oxidized energy transfer molecules (NADP+, cyclic AMP [cAMP]) were greatly reduced in dmP0/dm4 flies compared to dmP0/+ flies and were rescued in dmP0/dm4dmnt1flies (Figure 3B). At ZT15, the majority of amino acids (aspartate, serine, tyrosine, methionine, glutamine, threonine, lysine) were instead increased in dmP0/dm4 flies and were rescued in dmP0/dm4dmnt1 flies, demonstrating the abundance of certain type of metabolites at specific time points are different in dmP0/dm4 flies compared to either dmP0/+ flies or dmP0/dm4dmnt1 flies.

Putative Link between dMyc Regulation of Histidine Metabolism and Circadian Activity

Altered metabolites that could affect circadian activity include histamine, which is derived from histidine through histidine decarboxylase, Hdc, and glutamate that is produced from glutamine by glutaminase, GLS. Both histamine and glutamate have been shown to affect wakefulness (Abe et al., 2004; Zimmerman et al., 2017). Our data demonstrated that the ion count of glutamine is higher and glutamate is lower in the dmP0/dm4 flies, consistent with decreased conversion from glutamine to glutamate (Figures 3B, S5B, and S5C). These changes suggest a possible link to Drosophila GLS, which if diminished in low Myc states would decrease the conversion of glutamine to glutamate. Hence, we sought to determine whether Drosophila GLS is directly regulated by dMyc. We took advantage of the publicly available data of dMyc ChIP-seq and found that the endogenous dMyc binds to the promoter of GLS in Drosophila melanogaster Kc cells (Figure S5A) (Yang et al., 2013). However, a time-course study of GLS mRNA levels in tim-G > Myc+p35 with increased dMyc expression versus tim-G flies (Figures S5D and S5E) and of decreased dMyc activity in dmP0/+ versus dmP0/dm4 flies (Figures S5F and S5G) does not corroborate GLS as a dMyc target in Drosophila heads.

dMyc regulation of histidine metabolism, however, could affect circadian behavior. Histidine levels were significantly increased in dMyc-overexpressing flies (Figure 3A) and remarkably reduced in dMyc hypomorph flies (Figure 3B). Hence, we sought to determine whether dMyc regulates Hdc, which converts histidine to histamine, or the histidine transporters CG13743 and CG30394. dMyc target genes are expected to be bound by dMyc and have increased mRNA levels with dMyc gain of function and decreased levels with hypomorphic dMyc. Endogenous dMyc does not bind to the promoter site of Hdc from the dMyc ChIP-seq in Drosophila melanogaster Kc cells (Figure S5H) (Yang et al., 2013). By contrast, endogenous dMyc binds to the promoter regions of both CG13743 and CG30394 (Figure S5H) (Yang et al., 2013). Intriguingly, the mRNA levels of CG13743 and CG30394 differ in response to dMyc levels with CG30394 behaving as a dMyc target (Figures S5I–S5P). While both CG13743 and CG30394 mRNA levels seem to oscillate and peaked at ZT10 in tim-G and dmP0/+ flies (Figures S5I, S5K, S5M, and S5O), only CG30394 was increased in tim-G > Myc+p35 flies versus tim-G flies (Figures S5M and S5N) and decreased in dmP0/dm4 flies versus dmP0/+ flies (Figures S5O and S5P). Furthermore, it is notable that misregulation of dMyc expression resulted in disruption of CG30394 oscillation (Figures S5M and S5O). These observations demonstrate a significant difference of histidine levels and expression of the histidine transporter CG30394 at different times of the day with misregulation of dMyc expression, suggesting that dMyc regulation of histidine metabolism could be a potential link between dMyc-driven metabolism and circadian behavior.

DISCUSSION

In this report, we establish that misregulation of dMyc expression disrupts Drosophila circadian activity, adding to its pleiotropic roles in regulating cell growth, cell cycle, apoptosis, and metabolism (de la Cova et al., 2004, 2014; Gallant, 2013; Johnston et al., 1999; Montero et al., 2008; Moreno and Basler, 2004).

We report here that gain or loss of function of dMyc reduced rhythmicity and demonstrate a role of dMyc in Drosophila circadian behavior. We observed that overexpressing dMyc specifically in Pdf and cry-expressing cells results in loss of rhythm in more than 70% of flies. High levels of dMyc are associated with decreased expression of the neuropeptide PDF in dorsal projections of the sLNvs, which likely contributes to arrhythmicity. Moreover, we found that loss of endogenous dMyc significantly impairs rhythmicity that can be rescued by additional loss of the dMyc transcriptional antagonist dMnt. Through gene expression analysis, we found that low dMyc activity alters transcript oscillations of circadian clock genes, per and tim, and protein oscillation of PER in fly brain extracts as well as sLNvs. We also observe metabolic derangement resulting from loss of dMyc activity that could be rescued by concomitant loss of dMnt activity.

Unlike mammalian Myc (Altman et al., 2015), which suppresses BMAL1 expression, overexpression of dMyc in tim-positive cells increased expression of the Drosophila homolog of BMAL1, cyc. Furthermore, while mammalian REV-ERBα, repressor of BMAL1, was increased by Myc, its fly counterpart, E75, was unaffected by elevated dMyc in Pdf and cry-expressing cells. In addition to cyc, overexpressed dMyc induced mRNA levels of several E-box-containing, negative regulators of the clock—per, tim, vri, and cwo—and also increased PER protein expression in the fly heads. By examining publicly available data of dMyc ChIP-seq, we found that the endogenous dMyc directly binds to the promoter regions of cyc, pdp, vri, and cwo of Drosophila melanogaster Kc cells, suggesting that dMyc dysregulation of some clock genes contributes to dysrhythmia.

Although we found that dMyc overexpression affects circadian behavior, the arrhythmic behavior depended significante on background strain and the chosen promoter for ectopic dMyc expression. dMyc transcript and protein levels increased significante in tim-G > Myc+p35 flies; however, the percentage of rhythmic flies only marginally dropped. By contrast, cry24Pdf-G > Myc+p35 flies showed a severely arrhythmic phenotype, although dMyc transcripts were only moderately higher than controls. It is notable that ectopic dMyc expression also affects the expression of the neuropeptide PDF, central to fly circadian behavior. Because distinct subset of MYC target genes can differ as a function of MYC levels (Lorenzin et al., 2016), it is possible that the effect of ectopic dMyc expression on circadian behavior and clock genes is contextual. These observations suggest that dMyc may affect behavior through complex mechanisms in regulation of clock and metabolic gene expression.

Endogenous Drosophila Myc in Molecular Clock

Although dMyc has no prior known connection to the clock in flies, the increased arrhythmicity in dMyc mutant flies suggests that endogenous dMyc is involved in circadian homeostasis. In tim-Gal4 and cry24Pdf-Gal4 control flies, dMyc mRNA oscillates in a circadian fashion, suggesting dMyc is regulated by the Drosophila clock. Prior studies using microarray analysis to compare the mRNA expression profile of PDF neurons with that of ELAV (marker for the majority of neurons)-expressing neurons revealed that dMyc was highly enriched in PDF neurons, including both sLNvs (2.96 at ZT12 and 3.1 at ZT0 fold enrichment relative to ELAV neurons) and lLNvs (2.44 at ZT0 but not enriched at ZT12), and cycles within lLNvs (2.7-fold down at ZT12 compared to ZT0), suggesting a potential role of dMyc in circadian regulation (Kula-Eversole et al., 2010). Furthermore, given the observations of relatively stronger effects of dMyc on per/tim mRNA cycling in fly brains than on the PER protein oscillation in sLNvs, the action of dMyc in non-Pdf cells may also be crucial to maintain rhythmicity.

Drosophila Myc and Mnt

Our observation that loss of dMnt is able to rescue the arrhythmicity of dMyc mutant flies supports the notion that dMyc target genes are involved in circadian behavior. Mnt belongs to the extended Myc network, which includes Max, Mxd, Mlx, MondoA, and ChREBP (Conacci-Sorrell et al., 2014). Mnt represses Myc activities by interacting with Myc-binding partner Max and displacing Myc from binding Max (Conacci-Sorrell et al., 2014; Link et al., 2012). Although Myc was not linked to the molecular clock until recently (Altman et al., 2015), dMnt (isoform B) was found to be bound and regulated by CLK, and its mRNAs oscillate with peak expression at ZT14 (Abruzzi et al., 2011). Notably, dMnt expression was increased in sLNvs (2.2 at ZT12) and cycles within lLNvs (2.09-fold up at ZT12 compared to ZT0) (Kula-Eversole et al., 2010). The circadian expression of dMnt perhaps plays a role in maintaining the persistent oscillation of per and tim mRNAs in dMyc hypomorphic fly brains when dMyc level is low. The inverse abundance of dMyc and dMnt in PDF neurons at different times (ZT0 for dMyc and ZT12 for dMnt) may provide a potential explanation on the phase shift of per and tim transcripts, further supporting the hypothesis that both dMyc and dMnt are involved in the fly molecular clock.

Other members of the extended Myc network participate in metabolism and may affect circadian rhythm. MondoA was reported to be required for Myc-induced tumorigenesis both in vitro and in vivo (Carroll et al., 2015). MondoA, a glucose- 6-phosphate sensor, heterodimerizes with Mlx to regulate glucose-sensing, glutamine uptake, and glutamine-derived lipid biosynthesis (Carroll et al., 2015). Mondo and Mlx, the Drosophila orthologs of ChREBP/MondoA and Mlx, respectively, have also been shown to be crucial in sugar-sensing and metabolism (Havula et al., 2013; Mattila et al., 2015; McFerrin and Atchley, 2011). Interestingly, the direct target of Drosophila Mondo-Mlx, cabut (cbt), was found to repress gluconeogenesis enzyme pepck, and genes involved in lipolysis and glycerol metabolism with glucose stimulation (Bartok et al., 2015). Intriguingly, 85% of cbt-overexpressing flies exhibit arrhythmic or weakly rhythmic circadian behavior (Bartok et al., 2015).

Misregulation of Drosophila Myc Affects Metabolism

The metabolic alteration we observed in cry24Pdf-G > Myc+p35 flies and dMyc hypomorphic flies supports the role of dMyc in metabolism, as documented for mammalian MYC. Specifically, we surmise that alterations of glutamine, glutamate, and histidine levels by misregulated dMyc expression could affect fly circadian behavior. Previous studies demonstrated that Myc directly regulates mammalian glutamine metabolism to promote tumor growth (Gao et al., 2009; Xiang et al., 2015), but our data do not support a role for dMyc in regulating glutaminase, which converts glutamine to glutamate, a neurotransmitter, in fly head. However, among the metabolites that were reduced through loss of dMyc and rescued by dMnt, histidine also stood out, and its derivative histamine was previously shown to influence sleep-wake cycles (Oh et al., 2013; Parmentier et al., 2002). Histidine can be converted to histamine, which induces wakefulness in both mice and fruit flies (Oh et al., 2013; Parmentier et al., 2002; Thakkar, 2011). Interestingly, a recent study uncovered that the growth of dMyc-dependent neural dedifferentiation-derived clones in fruit flies is sensitive to histidine deprivation (Froldi et al., 2019). In that study, the mutant nerfin-1 clone (nerfin-1 is required for Drosophila neurons to maintain a differentiated state) that was generated in a Myc P0 (hypomorphic) heterozygous background failed to respond to dietary histidine supplementation, which can be rescued by dMyc overexpression. This result suggests that dMyc is critical for histidine import or utilization. Endogenous dMyc does not bind the Hdc promoter in Drosophila melanogaster Kc cells, suggesting that Hdc is not a direct dMyc target. We found, however, that dMyc bound histidine transporter genes, CG13743 and CG30394 (Yang et al., 2013), but only the expression of CG30394 behaved as expected for a dMyc target. The elevated histidine level in dMyc-overexpression flies and diminished histidine level in dMyc-hypomorphic flies support the hypothesis that dMyc upregulates histidine transporter to increase histidine import. As such, further studies are required to understand the putative connection between alteration of histidine metabolism with circadian behavior.

Collectively, our observations support the hypothesis that ectopic dMyc can affect the clock and that endogenous dMyc activity is essential for normal Drosophila circadian behavior. Furthermore, we surmise that similar to mammalian MYC, dMyc is able to induce negative regulators of the circadian clock and alter metabolism. These foundational observations linking dMyc to the Drosophila clock provide a path for additional studies, which will hopefully further link MYC with the circadian clock and metabolism in various organisms and settings.

STAR★METHODS

LEAD CONTACT AND MATERIALS AVAILABILITY

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Chi Van Dang (cdang@wistar.org). Drosophila lines generated in this study can be requested by contacting Dr. Amita Sehgal. This study did not generate new unique reagents.

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Fly stocks

Pdf-Gal4, tim-Gal4, cry24-Gal4, dmP0, UAS-p35 and UAS-mCD8::GFP lines were provided by the Bloomington Drosophila Stock Center. UAS-Myc UAS-p35 and dm4 were kindly shared by Dr. Julie Secombe (Albert Einstein College of Medicine, NY). dm4 dmnt1 lines were provided by Dr. Robert N. Eisenman’s lab. dm4 and dm4 dmnt1 lines were balanced with FM7i, Act-GFP. cry24-Gal4; Pdf-Gal4 lines were generated from crossing cry24-Gal4 to Pdf-Gal4 flies. dmP0/Y and its sibling control w/Y were generated by backcrossing dmP0 line from into isogenic w[1118] line (wlSO31). cry24-Gal4; Pdf-Gal4 > GFPCD8 lines were generated from crossing cry24-Gal4; Pdf-Gal4 to UAS-mCD8::GFP flies. Flies that used for the experiments were grown at 25°C.

METHOD DETAILS

Locomotor activity assay

Circadian locomotor activity assays were measured using the Drosophila Activity Monitoring System (Trikinetics). Approximately 5- to 7-day-old adult flies were entrained to light-dark (LD) 12 hour: 12 hour (12:12) cycles for three days, then transferred to constant darkness (DD) for 7–8 days. Locomotor activity data were analyzed using Clocklab software (ActiMetrics). Individual periods were calculated from 7–8 days activity data during DD using chi-square periodogram. Rhythm strength was determined by Fast Fourier Transform (FFT) analysis, then categorized as arrhythmic (< 0.01), weakly rhythmic (0.01–0.03) and rhythmic (> 0.05). Periods and rhythm strength were only calculated from rhythmic and weakly rhythmic flies.

Immunohistochemistry

Five-day-old adult flies were entrained to LD 12:12 cycle for 3 days and then transferred to DD. Brain dissection was done on day 4 or day 5 in DD. Adult fly brains were dissected and fixed for 30–60 min in 4% paraformaldehyde made in PBS with 0.3% Triton-X(PBST). Fixed brains were quickly rinsed twice in PBST, washed 3 × 10 min with PBST, followed by blocking with 5% normal donkey serum in PBST (NDST) for 1 hour. Brains were then incubated with primary antibody diluted in NDST at 4°C overnight. Brain samples were then washed 3 × 10 min with PBST, incubated with secondary antibody diluted in NDST at room temperature for 2 hours, washed with 3 × 10 min PBST again and mounted with Vectashield. Five to ten brains of each genotype at each time point were processed. A Leica TCS SP5 confocal microscope was used to acquire images for immunostained brain. Primary antibodies used include: rabbit anti-PDF (C7; Developmental Studies Hybridoma Bank) and Guinea Pig anti-Per (GP1140). Secondary antibodies used include: Alexa Fluor 488 goat anti-rabbit IgG (Life Technologies), Alexa Fluor 488 goat anti-Guinea Pig IgG (Life Technologies), Cy3 donkey anti-rabbit (Jackson ImmunoResearch Laboratories, West Grove, PA), Cy3 donkey anti-mouse (Jackson ImmunoResearch Laboratories). Quantitative analysis was performed by measuring integrated density on section of the largest area of sLNv with Fiji software. The data for each genotype/rhythmicity is an average value of approximately 40 sLNvs from 10 brain hemispheres.

Quantitative real-time PCR

Five- to seven- day old adult flies were entrained to LD 12:12 cycle for 3–5 days. Approximately 20 flies were collected on dry ice at each time points on the last day of LD for heads. Ten brains of each genotype and at each time point were dissected on the fourth day of DD after 3 days of LD. Brains were dissected in cold RNase-free PBS and spun down at 4C to remove PBS within 30 minutes. Adult fly heads or brains were homogenized in TRIzol® Reagent using VMR Pellet Mixers for 10 s. Total RNA was then extracted according to manufacturer’s instruction (TRIzol; Life Technologies) and cDNA was made by using TaqMan Reverse Transcription Reagents (Life Technologies). cDNA was used as template for quantitative real time PCR (RT-PCR) with specific Drosophila primers. RT-PCRs were performed using the ViiA 7 or QuantStudio™ 6 Flex Real-time PCR system (Applied Biosystems) with a SYBR Green kit (Life Technologies). Relative mRNA expression levels were normalized to actin and analyzed using absolute quantification method. Standard curve used for quantification was determined by serial dilution of mixed samples from all time points. Delta-Delta Ct method was also used in some of the experiments. Primers used in the experiments were dMyc forward (5′-GAGCAACAACAGGCCATCGATATAG-3′), dMyc reverse (5′-CCTTCAGACTGGATCGTTTGCG-3′), per forward (5′-CGTCAATCCATGGTCCCG-3′), per reverse (5′-CCTGAAAGACGCGATGGTG-3′), tim forward (5′-GCTGCGCCTTGTTTTCCTT-3′), tim reverse (5′-ATGATGTTCAGATCCTGCTGGA-3′), Pdp1 forward (5′-TTTGAACAGCTTGAAAGCGC-3′), Pdp1 reverse (5’-GAGATTTCCTGCCTGAGCTGG-3’), vri forward (5′-CGACTCTCTCGATGAACGGC-3′), vri reverse (5′-ACGGATGCAAGTTAGAAGCCTC-3′), Actin forward (5′-GCGCGGTTACTCTTTCACCA-3′), Actin reverse (5′-ATGTCACGGACGATTTCACG-3′), cry forward (5′-GCAGTACGTCCCGGAGTTGA-3′), cry reverse (5′-AGGGCTCGTGAACAAATTCCT-3′), Clk forward (5′-TTCTCGATGGTGTTCTCGGTG-3′), Clk reverse (5′-AGTTCGCAAAGCCAACGG-3′), cyc forward (5′-GTACGTTTCCGATTCGGTGT-3′), cyc reverse (5′-CTCTGTGGAACGTCGGTCTT-3′), cwo forward (5′-CAGGACTTTTGCCACC GACTATTG-3′), cwo reverse (5′-CTGCCGCCGCCTGCTGAC-3′), E75-RA forward (5′-CGGTAATCTGCACATTGTCGC-3′), E75-RA reverse (5′-TGCTTCAGCTGTTGGCTCTTG-3’), E75-RB forward (5′-TGCAACATCATCCGGAGGAT-3′), E75-RB reverse (5′-TCCTCCAGATGCAGCATCTCA-3′), E75-RC forward (5′-GACTTCTGTGATCTGCAGCACG-3′), E75-RC reverse (5′-CGACCTGTTACCCCCAAAATG-3′), E75-RD forward (5′-GCGAAGAACTCCCGATATTGAA-3′), E75-RD reverse (5′-TGAACTCACAGGTCTCGAGGTG-3′), GLS forward (5′-CGAGTGGCAGGAACTTCAA-3′), GLS reverse (5′-TGACAAGGGCGTTCATCAG-3′), CG13743 forward (5′-ATACCCTACGCCCTTCACA-3′), CG13743 reverse (5′-TGACCATCAGGATGAGTGAGTA-3′), CG30394 forward (5′-CTC CAATGCCTACCCATCTTC-3′), CG30394 reverse (5′-TGTCGAGACTCTGGTTGTTAATG-3′)

Western blot analysis

Five- to seven- day old adult flies were entrained to LD 12:12 cycle for 3–5 days. Approximately 8–10 flies were collected on dry ice at each time point on the last day of LD. Ten brains of each genotype and at each time point were dissected on the fourth day of DD after 3 days of LD. Brains were dissected in cold PBS and spun down at 4C to remove PBS within 30 minutes. Adult fly heads or brains were lysed and homogenized in 1X Passive Lysis Buffer (Promega) diluted in PBS supplemented with protease inhibitors and phosphatase inhibitor okadaic acid. Sample buffer (Thermo Cat:39000) was added to each sample. All samples were then boiled and spinned. Equal volume of supernatant from each sample was loaded into the Criterion pre-cast gradient gels (Bio-Rad, Hercules, CA, USA). Proteins were transferred onto nitrocellulose membranes using iBlot® Gel transfer system (Thermo). Primary antibodies used include: mouse anti-dMyc (from Eisenman lab), rabbit anti-α-Tubulin (Cell Signaling, Danvers, MA, USA) and Guinea Pig anti-PER (GP1140). Secondary antibodies used include: Peroxidase AffiniPure Goat Anti-Guinea Pig IgG (Jackson ImmunoResearch Laboratories), Goat anti-Mouse IgG1, HRP conjugate (Thermo). Immunoblots were then developed using ECL substrate (Thermo).

UCSC Genome Browser and Publicly Available Genomic Data

For Drosophila melanogaster Kc cell ChIP-Seq for endogenous dMyc peaks, publicly available data from Yang et al. was used [GEO accession GSE39521 (Yang et al., 2013) access date 2–13-2019]. dMyc peak tracks from GEO accession numbers GSM970847 were downloaded and analyzed on UCSC genome browser D. melanogaster Apr. 2006 (BDGP R5/dm3) assembly. The resulting images were exported using the Genome browser’s website and cropped for display.

Fly metabolites analysis by Gas Chromatography mass spectrometry (GC-MS)

Fly heads and bodies were separated before metabolite extraction. 100μL of cold methanol was added to the fly samples and homogenized using VMR homogenizer for 10 s. 100μL of cold methanol and chloroform was then added, followed by sonication on a power setting of 2 and a duty cycle of 20% for 1 minute. Nor-valine was also added to methanol-chloroform mix alone for comparison. The samples were centrifuged at 8,000 rpm at 4°C for 10 min. The metabolite-containing supernatant was evaporated under nitrogen at 40°C. 100 μL of N-tertbutyldimethylsilyl- N-methyltrifluoroacetamide (MTBSTFA, Regis, Morton Grove, IL, USA) and 100 μL of acetonitrile were added to the dried residue. The samples were heated in 4 mL sealed glass vials at 70°C for 90 minutes. The resulting silylated metabolites were transferred to 1.5ml Eppendorf tubes and centrifuged at 13,000 rpm to remove insoluble materials. One micro-liter of the supernatants was analyzed with an Agilent 7890A/5975A GC-MS system with a DB-5 column (Agilent, Santa Clara, CA, USA). Mass spectra were quantified with the MSD ChemStation software (Agilent).

Fly metabolites analysis by Liquid chromatography mass spectrometry (LC-MS)

Fly heads and bodies were separated before metabolite extraction. 600μL of cold 2:1 methanol:chloroform was added to the fly samples and homogenized in a bead-based tissue homogenizer at 25Hz for 4 minutes (TissuLyser II, QIAGEN, Hilden, Germany). 200μLof both water and chloroform was then added, followed by centrifugation at 18787×g for 7 minutes at 4°C. 350μL of the upper layer, comprising the aqueous layer, was separated and dried down overnight under vacuum. Samples were resuspended in 50μL of 50:50 water:acetonitrile for injection onto the mass spectrometer. Liquid chromatography conditions and mass spectrometer parameters for hydrophilic interaction chromatography (HILIC) analysis for small polar metabolites were executed as previously reported (Rhoades and Weljie, 2016). Then, 5 μL injections were performed for each sample in duplicate on a Waters Acquity H-Class UPLC coupled to a Waters TQ-S micro mass spectrometer (Milford, MA), using an XBridge BEH Amide column for chromatographic separation (2.1 × 150 mm, 2.5 μm). The LC solvents consisted of 95:5 water:acetonitrile with 20 mM ammonium acetate at pH 9 (mobile phase A) and acetonitrile for mobile phase B. The gradient was changed from 15% to 70% mobile phase A over 5 min at 0.15 mL/min, followed by an isocratic hold for 10 min. The column was washed in 98% mobile phase A and then re-equilibrated in starting conditions for 5 min before the next injection. The MS operated in ion-switching mode with a capillary voltage of 3 kV for electrospray positive mode (ESI+) and 2 kV for negative mode (ESI−). The desolvation gas flow was set to 900 L/h and desolvation temperature at 450°C, with the source temperature set at 150°C. Metabolites were detected using multiple reaction monitoring (MRM), with mass transitions and voltages optimized as previously described. LC-MS chromatograms were processed using TargetLynx under Mas-sLynx version 4.1. Ion counts were exported and processed in R (version 3.3), and normalized by the number of fly heads collected per genotype before calculation of ratios and generation of heatmaps.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical methods of experiments and the number of biological replicates are described in figure legends. Error bars represent standard error of the mean (SEM). All statistical analyses were performed using Microsoft Excel or GraphPad Prism and the statistical significance (p value) was determined using a two-tailed Student’s t test. Locomotor activity data were analyzed and quantified using Clocklab software (ActiMetrics). Circadian periods were calculated using chi-square periodogram. Rhythm strength was determined by Fast Fourier Transform analysis. Quantification of the western blot band intensities were measured using Image Studio Lite Version 4.0.21, according to the manufacturer’s instruction. Quantitative analysis of the confocal images was performed by measuring integrated density on a section of the largest area of the neuron of interest with Fiji software. Quantification of the mass spectrometry was described in details in method details.

DATA AND CODE AVAILABILITY

This study did not generate/analyze datasets/code.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse anti-dMyc	Eisenman Laboratory	Greer et al., 2013
Rabbit anti-α-Tubulin	Cell Signaling Technology	Cat#2144
Guinea Pig anti-PER (GP1140)	Cocalico Biologicals	Garbe et al., 2013
Rabbit anti-PDF (C7)	Developmental Studies Hybridoma Bank	AB_760350
Peroxidase AffiniPure Goat Anti-Guinea Pig IgG	Jackson ImmunoResearch Laboratories	AB_2337402
Goat anti-Mouse IgG1, HRP conjugate	Invitrogen	Cat#A10551
Chemicals, Peptides, and Recombinant Proteins
Pierce ECL Western Blotting Substrate	Thermo Scientific	Cat#32106
TRIzol reagent	Invitrogen	15596018
Reverse transcription kit	Invitrogen	N8080234
Passive Lysis 5X Buffer	Promega	Cat#E1941
Pierce Lane Marker Reducing Sample Buffer	Thermo Scientific	Cat#39000
N-tertbutyldimethylsilyl-N-methyltrifluoroacetamide	Regis Technologies	Cat#77377–52–7
Critical Commercial Assays
Power SYBR Green PCR Master Mix	Thermo Fisher	4368708
Deposited Data
Drosophila melanogaster Kc cell ChIP-Seq	Yang et al., 2013	GEO: GSE39521 (GSM970847)
Experimental Models: Organisms/Strains
D. melanogaster: Iso31	Sehgal Laboratory Stocks	N/A
D. melanogaster: w; Pdf-GAL4;Pdf-GAL4	Bloomington Drosophila Stock Center	BDSC_80939
D. melanogaster: yw; tim-GAL4	Sehgal Laboratory Stocks	N/A
D. melanogaster: w; tim-GAL4	Bloomington Drosophila Stock Center	BDSC_80941
D. melanogaster: cry24-GAL4	Bloomington Drosophila Stock Center	BDSC_24774
D. melanogaster: w; UAS-Myc UAS-p35	Gift of Julie Secombe	Greer et al., 2013
D. melanogaster: w; UAS-mCD8::GFP	Bloomington Drosophila Stock Center	BDSC_ 5137
D. melanogaster: w; tim (UAS)-GAL4	Sehgal Laboratory Stocks	N/A
D. melanogaster: w; UAS-p35	Bloomington Drosophila Stock Center	BDSC_5073
D. melanogaster: cry24-GAL4; Pdf-GAL4	This study	N/A
D. melanogaster: dmP0;bw1;st1	Bloomington Drosophila Stock Center	BDSC_11298
D. melanogaster: dmP0	This study	N/A
D. melanogaster: dm4/FM7i, Act-GFP	Gift of Julie Secombe	Pierce et al., 2008
D. melanogaster: dm4dmnt1/FM7i, Act-GFP	Eisenman Laboratory Stocks	Pierce et al., 2008
Oligonucleotides
For rtPCR primers, see method details sectionQuantitative real-time PCR	N/A	N/A
Software and Algorithms
Image Studio software, version 2.0	LI-COR	https://www.licor.com/bio/products/software/image_studio_lite/
Fiji	Fiji	https://fiji.sc/
Prism	GraphPad	N/A
Excel	Microsoft Office	N/A
Clocklab software	ActiMetrics	https://www.actimetrics.com/products/clocklab/
DAMSystem3 Data Collection Software	TRIKINETICS	https://trikinetics.com/
MSD ChemStation software	Agilent	https://www.agilent.com/en/products/software-informatics/massspec-workstations/gc-msd-chemstation-software
MassLynx version 4.1	Waters	https://www.waters.com/waters/en_US/MassLynx-MS-Software/nav.htm?locale=en_US&cid=513662