DAYWAKE implicates novel roles for circulating lipid-binding proteins as extracerebral regulators of daytime wake–sleep behavior

Abstract

Sleep during the midday, commonly referred to as siesta, is a common trait of animals that mainly sleep during the night. Work using Drosophila led to the identification of the daywake (dyw) gene, found to have anti-siesta activity. Herein, we show that the DYW protein undergoes signal peptide-dependent secretion, is present in the circulatory system, and accumulates in multiple organs, but, surprisingly, it is not detected in the brain where wake–sleep centers are located. The abundance of DYW in adult flies is regulated by age, sex, temperature, and the splicing efficiency of a nearby thermosensitive intron. We suggest that DYW regulates daytime wake–sleep balance in an indirect, extracerebral manner, via a multi-organ network that interfaces with the circulatory system.

Graphical Abstract

Drosophila daywake (dyw) is the first anti-siesta gene discovered. As in humans, key wake–sleep centers in Drosophila are found in the brain. Surprisingly, we show that the DYW protein, which belongs to the tubular lipid-binding protein (TULIP) superfamily, is not detected in the brain but accumulates in multiple organs and is secreted into the circulatory system, suggesting a novel signaling pathway for regulating wake–sleep balance.

Introduction

A hallmark feature of animal life is the daily changes in wake and sleep states. The timing of when animals are active during a daily cycle varies and can be broadly categorized as diurnal (day-time) or nocturnal (night-time), although other activity patterns are prevalent [1]. The main partition of human sleep occurs during the night, which is also similar to that observed in the model research organism Drosophila, or common fruit fly [2,3]. Indeed, studies using D. melanogaster have been highly informative in understanding genetic, molecular and neurological mechanisms underlying nighttime sleep [4,5]. Although diurnal animals are mainly active during the day, they can exhibit a period of increased quiescence during the mid-day to mid-afternoon, which is commonly referred to as a siesta. Analysis of several pre-industrial societies indicates that daytime napping is more prevalent in the summer compared to winter [6]. Daytime inactivity or siesta during hot days minimizes risks from heat, solar radiation and water loss, threats that are especially deleterious to small animals such as insects that cannot thermoregulate [7–9].

Studies in humans suggest health benefits from a short daytime nap or siesta [10]. These include improved cognitive performance, lower blood pressure and increased cardiovascular function. However, excessive daytime sleep is linked to worse prognosis in numerous diseases such as diabetes and neurological disorders [10]. Large-scale genomic analysis has shown that genetic factors contribute to variant daytime sleep behavior in humans and that they differ from those associated with nighttime sleep (e.g., [11–13]. We have been using D. melanogaster as a model to elucidate mechanisms governing daytime wake-sleep balance and discovered a prominent role for a gene we termed daywake (dyw) [14].

Earlier work showed that the splicing efficiency of a small thermosensitive intron in the 3’ untranslated region of the key circadian clock gene period (per) regulated siesta in D. melanogaster [15–17]. Splicing of this intron termed dmpi8 (Drosophila melanogaster period intron 8) is progressively increased as daily temperatures decrease [15]. Higher rates of dmpi8 removal are causally linked to decreases in daytime siesta and more activity, a shift in balance appropriate for the more favorable cooler conditions. Surprisingly, the effects of dmpi8 splicing on daytime wake-sleep levels are not mediated by per function but by the slightly overlapping reverse-oriented dyw gene [14]. Increasing the splicing efficiency of dmpi8 somehow leads to higher dyw mRNA levels, consistent with an ‘anti-siesta’ role for dyw. While genetic manipulations of dyw affect wake-sleep levels during daylight, little to no effect was observed on nighttime sleep [14]. To the best of our knowledge dyw is the first described gene that specifically regulates siesta.

The function of dyw in daytime wake-sleep balance is not clear. The DYW protein is predicted to be a secreted lipid carrier of the Takeout-Juvenile Hormone Binding family (JHBP). Although JHBPs are only found in arthropods, they are part of the larger TULIP (Tubular Lipid Binding Protein) superfamily present in bacteria to humans [18]. Spatio-temporal and regulatory aspects of dyw mRNA levels in flies have been reported [14,19], most notably from annotated databases based on genome-wide RNA-seq data. Nonetheless, since DYW is predicted to be secreted, the relationship with its reported mRNA profiles is not clear. Herein we biochemically characterize DYW protein and show that a portion is secreted into the hemolymph (blood equivalent) but the majority of steady-state levels accumulate in tissues where its mRNA is highly expressed. Intriguingly, all the DYW detected in flies appears to be cleaved of its signal peptide. Our results suggest that DYW is part of an intricate inter-organ network comprising surface tissues (e.g., eyes), internal organs (digestive system) and systemic presence in regulating daytime wake-sleep balance.

Materials and Methods

Drosophila Maintenance and Strains.

Drosophila stocks were routinely maintained at 25°C in vials or bottles containing Bloomington Standard Medium (1.6% yeast, 0.5% soy flour, 7% yellow cornmeal, 1% agar, 7.5% light corn syrup, 0.5% propionic acid, 0.2% tegosept), unless otherwise noted. In addition, flies used in this report were maintained in daily 12 h light/12 h dark cycles [LD; where Zeitgeber 0 (ZT0) = lights-on] at the indicated temperature, prior to removal at desired times. Flies used in this study include, wildtype Canton-S (BDSC #64349), w¹¹¹⁸ and natural populations from Australia [20].

Generation of anti-DAYWAKE Antibodies.

The dyw coding region encompassing aa 26-260 was amplified using the primers NdeI_0.9(26-260)_F and BamHI_0.9R (see Table S1) in the presence of plasmid RH43234 (RRID:DGRC_10923), which contains the entire dyw open reading frame. The resulting DNA fragment was subcloned into the pET-15b vector (Novagen^®; SigmaAldrich cat.# 69662) using restriction cites NdeI and BamHI such that the coding region for the plasmid derived 6xHis tag was situated in frame at the N terminus of the start of dyw sequences, generating the final construct, termed 6His-DYW(Δ1-25). The plasmid was transformed into Tuner^™ (DE3)pLysS competent cells [Novagen^® F⁻ ompT hsdS_B (r_B⁻ m_B⁻) gal dcm lacY1(DE3) pLysS (Cam^R); SigmaAldrich cat.# 69451; a kind gift from Elliot Campbell and Steven Anderson (Rutgers University). Expression was induced by the addition of isopropyl-β-D-thiogalactoside (IPTG) to a final concentration of 0.5mM when the media reached an OD₆₀₀ = 0.6. After 2 hours of induction, cells were harvested by centrifugation at 3,000xg for 15 minutes. Subsequently, cells were resuspended in 5ml of Lysis Buffer (20mM Tris·HCl [pH 8.0], 150mM NaCl, 6M Urea), sonicated and incubated on ice for 30 minutes. Insoluble debris was removed by centrifugation at ~27,000xg for 20 minutes at 4°C, and the supernatant was used for downstream affinity purification.

His-TALON^® Superflow Affinity Resin (MiliporeSigma, cat.# GE28-9574-99) was washed in cold 1xTBS (150 mM NaCl, 50 mM Tris [pH 7.6]) and equilibrated in lysis buffer for 10 minutes. Supernatant was loaded onto the resin and incubated at 4°C with end over end rotation for 30 minutes. Resin was collected via centrifugation and washed 3 times in cold 1x Wash Buffer (20mM Tris [pH 8.0], 150mM NaCl, 25mM imidazole, 6M urea). The bound fraction was eluted by incubating resin in 900 μl of Elution Buffer (20mM Tris [pH 8.0], 150mM NaCl, 200mM imidazole, 6M urea) at room temperature for 10 minutes with occasional mixing. Eluates were analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) followed by Coomassie staining to determine DYW(Δ1-25) protein purity and yield. Briefly, protein samples were resolved by 12% SDS-polyacrylamide gel electrophoresis (PAGE; 12% resolving gel; 29.6%-acrylamide:0.4%N,N’-methylene-bis-acrylamide, 0.4M Tris·HCl [pH 8.8], 0.1% SDS, 0.05% ammonium persulfate, 0.02% temed) and then fixed in a 50% methanol, 10% glacial acetic acid solution for 1 hour at room temperature. After fixation, the gel was stained using a 0.1% Coomassie Brilliant Blue R-250, 50% methanol, 10% glacial acetic acid solution for 20 minutes at room temperature. Gels were then de-stained in a 40% methanol, 10% glacial acetic acid solution at room temperature, replacing the de-staining solution as necessary until background staining disappeared.

Aliquots of purified DYW(Δ1-25) protein were sent to Cocalico Biologicals, Inc.^™ (Stevens, PA, USA) where the immunogen was mixed with Freund’s Adjuvant, followed by injection of one guinea pig and two rat hosts. In this study, we used the guinea pig anti-DYW antibody (GP49) since it has the highest signal to noise ratio (data not shown).

Generation of daywake constructs and expression in Drosophila S2 cells.

Dyw containing inserts were generated by amplifying the dyw coding sequence using forward primers NdeI_0.9FL_F or NdeI_0.9(26-260)_F with reverse primer BamHI_0.9R for generating constructs pMT-DYW and pMT-DYW(Δ1-25), respectively (primers listed in Table S1). The resulting amplified fragments were cloned into pMT-V5/His (ThermoFisher cat# V412020) using NdeI and BamHI. Constructs were transformed into JM109 E. coli competent cells (endA1, recA1, gyrA96, thi, hsdR17 (r_k⁻, m_k⁺), relA1, supE44, Δ(lac-proAB), [F′ traD36, proAB, laqI^qZΔM15]) (Promega cat# L2005). After antibiotic selection on LB agar plates, plasmid DNA was isolated from 5 transformants and the correct sequences verified by Sanger sequencing using the MT Forward primer (Table S1).

The DYW(N182) mutant was generated from the pMT-DYW construct using the QuikChange II XL site directed mutagenesis kit (Agilent cat# 200521) using primers 09_N182A_F and 09_N182A_R (Table S1). Constructs were transformed into XL-10 Gold Ultracompetent Cells (Agilent Technologies cat.# 200315; Tet^rΔ (mcrA)183 Δ(mcrCB-hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac Hte[F’ proAB lacI^qZΔM15 Tn10 (Tet^r) Amy Cam^r]); transformants were verified as described above. Recombinant plasmid DNA was prepared using Nucleospin Plasmid recovery kits according to manufacturer’s protocol (Macherey Nagel cat# 740588.250).

Constructs were transfected into cultured Drosophila S2 cells (Invitrogen^™, cat# R69007) using the Effectene transfection system (Qiagen cat# 301427). Briefly 1x10⁵ cells were seeded into 6-well tissue culture plates (Falcon cat# 353046) containing 1.6ml Schneider’s Media (Sigma-Aldrich, cat# S0146). Cells were allowed to grow for 24 hours after which 400ng of recombinant plasmid DNA was added to the media. After incubation for 3 days, cells were washed with 1xPhosphate Buffered Saline (PBS; 137mM NaCl, 2.7mM KCl, 10mM Na₂HPO₄, KH₂HPO₄). After washing, 1.6ml Schneider’s media containing CuSO₄ at a concentration of 0.5mM was added to induce expression of metallothionein promoter-based constructs (pMT), and cells were cultured for another 24 hours.

Protein Extract Preparation from Drosophila S2 Cells and media.

Cells were collected in 15ml conical tubes by centrifuging at 2000 RPM (Beckman model TJ-6). Culture media was removed and stored for further analysis at 4°C (see below). Following centrifugation, pelleted cells were resuspended in 1ml 1xPBS and transferred to 1.5 microfuge tubes, washed 3x by resuspension in 1xPBS followed by centrifugation at ~17,000xg. Cell-free extract preparation was done essentially as described [21]. Briefly, after washing, cell pellets were homogenized in 50μl modified RIPA buffer (50mM Tris·HCl [pH 7.5], 150mM NaCl, 1% Nonidet P-40, 0.25% Sodium Deoxycholate, 1x cOmplete EDTA-free Protease Inhibitor Cocktail [Roche cat# 11873580001]) with a motorized pellet pestle (Kimble cat# 749521-1590; Kontes cat# 749540-0000), followed by 30 min incubation on ice. Homogenized samples were subsequently centrifuged at ~17,000xg and the supernatant protein was assayed for total protein concentration using the Pierce^™ Bradford Plus Protein Assay Kit (cat# 23236) and following manufacturer’s instructions. For analysis by immunoblotting, each lane contained either 20μg of total protein extract generated from spun cells (representing ~1x10⁴ cells or ~3% of total cells) or 5μl of culture media (representing ~0.3-0.5% of the available media).

Endoglycosidase H treatment of DAYWAKE produced in S2 cells.

We used Endo H (Promega, cat# V4871) and followed manufacturer’s protocol. Essentially, 20μg of total protein extract generated from spun cells (representing ~1x10⁴ cells or ~3% of total cells) and 5μl of culture media (representing ~0.3-0.5% of the available media) were added to ddH₂O and 1μl 10xDenaturing Buffer (5% SDS, 400mM DTT) to a final volume of 10μl. This mixture was incubated for 5 minutes at 95°C and then allowed to cool to room temperature. Subsequently, 2μl of 10xEndoH Reaction buffer, 2500U EndoH and ddH₂O were added to a final volume of 20μl (final concentrations; 1μg/μl protein, 0.25% SDS, 20mM DTT, 50mM sodium citrate [pH 5.5]). The reaction was allowed to continue at 37°C for 4 h whereupon equal volumes of the reaction mixtures for cell extracts and media were resolved via 12% SDS-PAGE gel and probed for DYW by immunoblotting as described below.

Protein extracts from adult flies.

Flies were collected at the indicated times. For the analysis of extracts from isolated heads, isolated bodies or whole flies, the collected flies were immediately frozen on dry ice. Heads and bodies were isolated by mechanical agitation and sorted using nested sieves (Newark Wire Cloth Company, Sieve nos. 40 and 25, 425μm and 710μm, openings respectively). For timed collections post-eclosion, numerous vials were cleared of all emerged flies in the morning following lights-on (ZT0), a time in the day when eclosion rate is at its peak. Newly emerged flies within two hours post-clearance were transferred to fresh vials and placed in 18°C or 25°C LD incubators to age for specified times before briefly anesthetizing with CO₂, sorting by sex and quick-freezing replicate sets of 8 individuals for each group. Independent experiments for age-matched replicates were collected at similar ZT times, which represent flies within a 2 hr range in age. For more accurate temporal resolution, the 0-30 min replicates were placed in separate incubators 1.5 hr in advance of clearing, to equilibrate the vials to the test temperature (i.e., 18° or 25°C). In cases where we dissected tissues, the collected flies were quickly anesthetized using CO₂, placed in Ringer’s solution and desired tissues such as retinas and guts dissected, followed by snap freezing.

Cell-free extracts for immunoblotting were prepared as previously reported [21]. Briefly, fly samples were homogenized in modified RIPA buffer (50mM Tris·HCl [pH 7.5], 150mM NaCl, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1x cOmplete EDTA-free Protease Inhibitor Cocktail [Roche cat# 11873580001]) using a motorized pellet pestle (Kimble cat# 749521-1590; Kontes cat# 749540-0000). Homogenized samples were then centrifuged at ~17,000xg for 20 minutes at 4°C to remove lipid and insoluble protein fractions. The concentration of soluble protein was determined using the Pierce^™ Bradford Plus Protein Assay Kit (cat# 23236) and following manufacturer’s instruction. Samples were then stored at −80°C or immediately analyzed by immunoblotting (see below).

Immunoblotting.

Samples prepared from adult flies, S2 cell extracts or media were mixed with 4x SDS-PAGE sample buffer (final concentration 62.5mM Tris·HCl [pH 6.8], 2.5% SDS, 0.002% Bromophenol Blue, 10% glycerol) and boiled at 95°C for 5 minutes. Where indicated, we compared samples with equal total protein, equal fly body parts or a selected volume of S2 media. For extracts prepared from flies or S2 cells, equal amounts of total protein were analyzed for each experiment, typically 20-60 μg. In all cases, samples were resolved using BioRad electrophoresis systems. These included hand cast mini 12% SDS-polyacrylamide gels (12%; 29.6%-acrylamide:0.4%-N,N’-methylene-bis-acrylamide, 0.4M Tris·HCl [pH 8.8], 0.1% SDS, 0.05% ammonium persulfate, 0.02% temed), or 12% Criterion^™ TGX^™ Precast Midi Protein Gels (cat# 5671044). Resolved proteins were transferred onto a 0.45μm pore size nitrocellulose membrane (GE Amersham cat# 10600002 or ThermoFisher, cat# 88018) using the BioRad Trans-Blot SD Semi-Dry Transfer Cell (cat# 1703940) or TransBlot Turbo Transfer System (cat# 1704150).

Membranes were routinely stained with reversible dye [e.g., Totalstain Q (NC), cat#AC2227; Azure Biosystems] to verify protein transfer prior to immunoblotting, subsequently washed in 1x Tris Buffered Saline with Tween (1xTBST: 20mM Tris [pH 7.5], 150mM NaCl, 0.05% Tween-20) for 10 minutes and blocked in blotting-grade 1% nonfat dry milk (BioRad cat# 1706404) diluted in 1xTBST or EveryBlot Blocking Buffer (BioRad, cat# 12010020) for 30 minutes. After blocking, membranes were washed with 1xTBST and incubated with shaking for 1 hour at 4°C with anti-DYW antibody GP-49 diluted 1:4000 in 0.1% Blocker (diluted in 1xTBST). In some cases, membranes were also incubated overnight at 4oC in the presence of GP-49 (1:1000) diluted in 1% Blocker-TBST. Membranes were then washed with 1xTBST and incubated for 1 hour in anti-guinea pig HRP conjugated antibody diluted 1:10,000 in 0.1% Blocker (diluted in 1xTBST). Signal was detected using Amersham ECL Prime (GE cat# RPN2232) and imaged either on an Azure Biosystems c600 imager (AZI600-01) or autoradiography using film (XAR ALF 1318; LabScientific).

Results and Discussion

DYW is N-glycosylated and secreted in a signal peptide-dependent manner.

DYW is predicted to be a secreted JHBP of 260 amino acids. To better analyze DYW regulation and function we generated anti-DYW polyclonal antibodies in guinea pigs and rats (Fig. 1). The antigen we used was the full-length protein missing the first 25 amino acids that comprise the signal peptide (Figs. 1A and B). SDS-polyacrylamide gel electrophoresis (PAGE) followed by immunoblotting in the presence of anti-DYW antibodies revealed two to three closely migrating DYW-specific electrophoretic bands, suggesting different DYW isoforms (Fig. 1C; further discussed below). These bands migrated with an estimated molecular weight of approximately 22-25kDa (data not shown), consistent with its predicted polypeptide mass. Several experiments using flies and transfected Drosophila tissue culture cells validated the specificity of our anti-DYW antibodies (e.g., Fig. 1C, compare lanes 3 and 4 to 1 and 2).

Fig. 1. DYW produced in S2 cells is secreted in a signal-peptide dependent manner and stabilized by glycosylation at amino acid N182.

(A) Residues 1-25 of DYW are predicted to be a signal peptide. (B) Bacterial cells were transformed with a construct expressing 6His-dyw(Δ1-25). Shown is a Coomassie stained gel of the input cell-free extract material (lane 1) and the resultant affinity purified 6His-DYW(Δ1-25) (lane 2) used as immunogen. (C) Example of a validation experiment for the anti-DYW antibody (GP49) used in this study. Extracts prepared from either E. coli expressing 6His-dyw(Δ1-25) or adult fly heads were immunoblotted in the presence of GP49 (lanes 1 and 2) or the pre-immune sera (lanes 3 and 4). Note the presence of two closely migrating DYW bands in fly extracts, whereas only one mobility isoform is observed for DYW(Δ1-25) (B, lane 2; C, lane 1). (D-F) Drosophila S2 cells were transfected with the indicated constructs expressing either full length dyw (FL) or altered versions missing the first 25 aa (Δ1-25) or N182A mutation. Aliquots from cell extracts (E and F, top panels) and media (E and F, bottom panels) were either mock-treated (E, lanes 1 and 2), or treated with Endoglycosidase H (EndoH) (E, lanes 3 and 4). For analysis by immunoblotting, each lane contained either 20 μg of total protein extract generated from spun cells (representing ~1x10⁴ cells or ~3% of total cells) or 5 μl of culture media (representing ~0.3-0.5% of the available media). Samples were then immunoblotted in the presence of anti-DYW antibodies. In S2 cells, DYW exhibits 2-3 mobility isoforms that appear to be influenced by N-glycosylation.

To examine the possibility that DYW is exported out of cells, we first measured the ability of DYW expressed in Drosophila tissue culture cells to be secreted into the media (Fig. 1D). Indeed, side-by-side comparison of DYW levels reveals that approximately equal amounts of DYW are present in cells and secreted extracellularly (Fig 1E, lane 1, compare top and bottom panels). Although we did not determine if the stability of DYW differs when present in cells or the extracellular milieu, our results clearly indicate that DYW can be secreted. That the presence of DYW observed in the media represents functional secretion was verified by analyzing a version of DYW missing the signal peptide (DYWΔ1-25). In sharp contrast to wildtype DYW, DYWΔ1-25 accumulates in cells but not the media (Fig 1E, lane 2, compare top and bottom panels). Moreover, the fastest moving isoform(s) of DYW co-migrates with (DYWΔ1-25) (e.g., Fig. 1C, compare lanes 1 and 2; and Fig. 1E, compare lanes 1 and 2), strongly suggesting that DYW protein is subject to signal peptide cleavage.

Many proteins that enter the secretory pathway through the endoplasmic reticulum and Golgi apparatus on their route to extracellular export undergo N-glycosylation, which can enhance maturation, stability and function [22,23]. Indeed, several JHBPs from higher insects have been shown to undergo N-glycosylation [24–26]. Endoglycosidase H (Endo H) cleaves the chitobiose core of high mannose from N-linked glycoproteins [27]. Treatment of S2 cell extracts expressing wildtype DYW with EndoH collapses the multiple electrophoretic isoforms into one detectable band that comigrates with DYWΔ1-25 (Fig. 1E, compare lanes 1 and 3, top panel). The same treatment of DYW in the media was not accompanied by any visible changes in electrophoretic mobility (Fig. 1E, compare lanes 1 and 3, bottom panel). Since EndoH does not cleave complex glycans it is possible that glycosylation is modified in DYW secreted outside the cell and/or a structural change occurs that makes enzymatic cleavage refractory once DYW enters the extracellular milieu. We also identified a classic N-glycosylation site (Asn-X-Ser/Thr, where X = any aa except Pro; [28]) on DYW at position 182 (Fig. 1F, top). Indeed, mutating the asparagine to alanine prevented DYW from accumulating (Fig. 1F, compare lanes 1 and 2), consistent with a key role for N-glycosylation in the stability and possibly function of DYW.

DYW protein levels peak early in adulthood and increased in cold temperatures and females.

Genome-wide RNA profiling of different Drosophila developmental stages indicate that dyw mRNA levels are low or below detection limits during embryonic, larval and early pupal stages, peak during the last day of eclosion, drop during the first day of adult emergence and remain detectable but constitutively low throughout adulthood (RNA-seq data annotated at flybase.org). The relatively high levels of dyw transcripts during eclosion are consistent with an early report examining expression from the dyw locus, which at the time was called the 0.9 gene [19]. In addition, we previously showed that dyw mRNA levels persist in adults and remain relatively constant throughout a daily cycle [14].

To determine temporal aspects of DYW protein levels, we analyzed extracts prepared from whole flies that were collected at several hour intervals during the first 24 hr post-eclosion followed by more limited sampling for several more days. To ensure we obtained sufficient numbers of properly aged flies for analysis, we simultaneously seeded many fly cultures, cleared them of adults and continuously transferred only flies that emerged within the first two hours (0-2 hr post-eclosion) to fresh vials (Fig. 2A). These freshly transferred flies were then aged for specific durations, followed by freezing and pooling those from the same time window. Males and females were analyzed separately. Moreover, since cooler temperatures increase dyw mRNA levels [14], flies were maintained at either 18°C or 25°C beginning with the 0-2 hr post-eclosion pool of flies. The results for DYW protein levels are consistent with expectation based on extant dyw mRNA profiles (Fig. 2B–E). Most notably, DYW protein levels peak around 10 hrs post-eclosion followed by a steady decline during the first day of emergence (Fig. 2B, C and E). In addition, in both males and females the steady-state abundance of DYW is higher at 18°C compared to 25°C (Fig. 2B, C, E). Finally, when standardized against total fly protein levels, the abundance of DYW is higher in females compared to males (Fig. 2D and E). Females manifest less midday siesta compared to males [29] but a causal connection to sexual dimorphism in DYW protein levels has not been established. We did not observe any consistent difference in the electrophoretic migration patterns of the various DYW bands as a function of time post-eclosion, sex or temperature.

Fig. 2. DYW levels are regulated by age, sex and temperature.

(A-E) Male and female adult flies were collected at the indicated temperatures (18° and 25°C) and ages post-eclosion (hr). Whole fly extracts were prepared and DYW protein visualized by immunoblotting in the presence of anti-DYW antibodies. For each experiment and/or gel image shown, equal amounts of total protein were loaded in each lane, typically 20-60 μg. (A) Schematic of experimental outline. Cultures with newly eclosing flies were maintained at 25°C under 12:12 hr light-dark conditions where ZT0 is lights-on. At ZT23 of the experimental day, half of the vials containing flies were kept at 25°C and the other half moved to 18°C. One hour later (ZT0) all adult flies were quickly removed, vials returned to the same temperature for an additional 2 hrs, followed by transfer of the 0-2 hr old flies to fresh vials for collection at the indicated times. (B, C) Representative examples highlighting the effects of temperature on DYW levels. Note, two exposures are shown for each. (D) Representative example highlighting the sexual dimorphism in DYW levels. (E) Quantification of DYW levels as a function of age, temperature and sex. Values for each time point represent at least three biological replicates and error bars indicate standard error of the mean (SEM).

In our initial characterization of dyw, older flies were used to analyze its expression. We showed that dyw mRNA levels in head extracts are higher at 18°C compared to 25°C, show little to no daily changes and is upregulated in flies with higher dmpi8 splicing efficiency [14]. To determine if DYW protein levels are similarly regulated, we generated head extracts from 4-5 day old flies kept at either 18° or 25°C and exposed to several days of 12 hr light:12 hr dark conditions [LD; where Zeitgeber time 0 (ZT0) is defined as lights-on). The levels of DYW protein are higher at 18°C compared to 25°C in male and female flies (Fig. 3A, compare lanes 1 and 2 to 3 and 4). We did not observe large variations in DYW protein levels at different times of day (Fig. 3A, compare lane 1 to 2; and lane 3 to 4; also, Fig. 3B), although we cannot rule out minor but physiologically relevant temporal changes.

Fig. 3. DYW protein levels are relatively constant throughout a daily cycle and elevated in flies with higher dmpi8 splicing efficiency.

Young adult flies (3-5 days old) were kept at 25°C and exposed for several days in 12:12LD cycles and collected at the indicated times (ZT, where ZT0 = lights-on). Extracts from isolated heads were probed for DYW staining by immunoblotting. For each experiment and/or gel image shown, equal amounts of total protein were loaded in each lane, typically 20-60 μg. (A) Males and females were separated prior to start of LD exposure and kept at the indicated temperature. (B) Male flies were used and kept at 25°C. Note that flies carrying the SNP3G genotype have enhanced dmpi8 splicing efficiency compared to SNP3A. Representative examples of at least two independent experiments are shown.

To determine if differences in dmpi8 splicing efficiency modulate DYW protein levels we analyzed LD-entrained natural populations of flies previously shown to have different dmpi8 splicing efficiencies that affect midday siesta levels [20]. Specifically, flies from Australia that carry the SNP3G allele in the 3’ UTR of per have higher dmpi8 splicing efficiency and lower midday siesta compared to Australian counterparts with the SNP3A allele [20,30]. Indeed, strains of flies with SNP3G have higher DYW protein levels compared to those with SNP3A (Fig. 3B, compare lanes 1-4 to 5-6). Overall, our results indicate that steady-state levels of DYW protein in the head and body are regulated by temperature and dmpi8 splicing efficiency in a manner consistent with regulatory effects on dyw mRNA levels.

DYW protein accumulates in multiple organs and the circulatory system.

Dyw transcripts have been reported to be highest in several body parts of adult flies, most notably the head, eyes and gut; in addition to spatially undefined carcass or cuticle preparations, essentially representing the exoskeleton (RNA-seq data annotated at flybase.org). However, because DYW can be secreted it is not clear if DYW accumulates in these tissues. Moreover, we sought to determine if DYW protein is detected in the hemolymph.

To determine relative levels of DYW in different body parts we first separated males and females, and then pooled dissected body parts. Extracts were prepared from body parts representing equal numbers of flies followed by immunoblotting (Fig. 4A). DYW protein is present in numerous isolated body parts, including abdomen, thorax, heads and guts. However, the distribution of DYW in the different body parts shows significant differences between males and females (Fig. 4A, compare top and bottom panels). For example, the total abundance of DYW in males is relatively higher in the abdomen and thorax compared to the head, which is opposite in females (Fig. 4A, compare lanes 2 and 3 to 6). The signal intensity of DYW protein in the different body parts is not simply related to the total amount of protein analyzed. This is clearly observed with legs and antennae in females whereby more total DYW is present in 6 legs or 2 antennae compared to the much larger abdomen (Fig. 4A, compare lanes 5 and 9 to 2, bottom panel). We also checked for the presence of DYW in adult brains, co-staining with antibodies against synapsin (Syn) since it is predominately expressed in the brain. Surprisingly, we do not observe DYW in the brain compared to the larger head even when loading equal amounts of total protein (Fig. 4B). This suggests that although DYW protein is enriched in the head, it does not regulate wake-sleep balance via a direct presence in the brain.

Fig. 4. DYW protein levels vary in different body parts and is detected in the hemolymph but not brain.

Different parts of young male or female flies (approx. 10 - 12 hr post-eclosion) were isolated, extracts prepared and DYW visualized by immunoblotting. (A) Total amount of material loaded in each lane represents equal number of flies; e.g., 1 abdomen, 6 legs; but not total protein. (B) Male and female flies (0-5 hr post-eclosion) were collected and immediately used for isolating whole heads and brains, followed by extract preparation. Equal total protein (60 μg; thus, relatively more brains were analyzed compared to whole heads) was loaded in each lane and immunoblotted in the presence of either anti-Synapsin (top) or anti-DYW (bottom) antibodies. Synapsin is associated with synaptic vesicles and predominately expressed in the brain. Note that the image shown is a “long” exposure to better ensure the absence of DYW in the brain, but as a result the relative DYW levels between males and females are beyond the linear range required for direct comparison. (C) Hemolymph (hemo) was extracted from punctured flies and equal total protein of hemolymph (lane 1), remaining fly material (lane 2) and mock-treated flies (lane 3) was loaded. Note that the amount of hemolymph loaded represents approximately 10-20x as many flies as the total fly preparations. Thus, although DYW is present in the hemolymph, the far majority appears to be localized in solid tissues.

Based on early studies whereby JHBPs were isolated from insect hemolymph it was generally thought that they mostly function as carriers that bind and deliver JH to target tissues [31,32]. However, many hemolymph JHBPs (hJHBPs) shown to bind JH were studied in larger insects [33–37] and share little homology with DYW found in Drosophila [18,38,39]. Our analysis of extracted hemolymph indicates that DYW is present in the circulatory system, consistent with other JHBPs (Fig. 4C), including Drosophila TAKEOUT [40]. Nonetheless, several side-by-side comparisons indicate that on a total fly basis the overall levels of DYW in the hemolymph are much less compared to that in the combined solid tissues (Fig. 4C, compare lane 1 to 2 and 3). While it is difficult to determine if we extracted all the hemolymph, our findings indicate that DYW has a systemic presence but that the majority of DYW accumulates locally at major sites where its mRNA is expressed.

Summary.

Earlier work showed that dyw expression regulates daytime wake-sleep balance in a manner consistent with an ‘anti-siesta’ function [14]. Based on our own work and annotated databases of genome wide transcriptomics in Drosophila, dyw expression peaks during the last pupal stage and in adults is preferentially observed in the head, eye and hindgut/rectal pad (annotated RNA-seq data at flybase.org; data not shown). In addition, we previously showed that dyw transcript levels are regulated by the splicing efficiency of dmpi8 [14]. However, DYW is predicted to be secreted extracellularly, potentially skewing its spatial accumulation from sites-of-production. In this report we determined key spatio-temporal attributes of DYW protein that strongly suggest it has both local and systemic functions in daytime wake-sleep control.

Post-translational modifications of ectopic DYW in cultured cells and native DYW in flies appear to result in the same 2-3 major electrophoretic isoforms, most likely due to differential glyco-moieties and/or -branching (Fig. 1). This strongly suggests that irrespective of site-of-production, most of the steady-state DYW produced in flies is cleaved of its signal peptide and secreted. Yet only a small fraction of this total DYW present in flies appears to be secreted into the hemolymph (Fig. 4C). While we cannot rule out enhanced degradation of DYW in the hemolymph, it is possible that only one or a few tissues where DYW protein is synthesized has competence to secrete it into the circulatory system. In summary, our findings suggest that DYW participates in an inter-organ network that integrates with the circulatory system, possibly with sex-specific functionality. Adding to the complexity of this putative inter-organ/systemic network for DYW action is that to our surprise we have not observed DYW in the brain. Clearly, it is possible that very low but physiologically relevant levels of DYW are present in the brain. Yet our findings raise the intriguing possibility that DYW works ‘at a distance’ to modulate brain centers that ultimately control siesta. Ongoing studies are aimed at better understanding the relationships between the many locations of DYW protein throughout the body and hemolymph and its ability to modulate daytime wake-sleep balance.

List of Abbreviations

dyw

daywake

per

period

dmpi8

Drosophila melanogaster period intron 8

JH

juvenile hormone

JHBP

juvenile hormone-binding protein

Endo H

Endoglycosidase H

LD

12 hr light:12 hr dark cycles

ZT

Zeitgeber time (where ZT0 = lights-on)

SNP

single nucleotide polymorphism

Data Availability.

All data shown in this manuscript are images of chemiluminescent signals from immunoblots that were captured either by autoradiography and converted to digital form or via direct imaging using azure biosystems. Original digital images shown in the manuscript can be requested. In cases where digital images were used for quantification, Excel Tables containing the values recorded and corresponding statistical analysis are available upon request.