EEC-derived Unpaired signaling regulates sleep.

a, upd3 expression levels in midguts expressing RNAi-mediated knockdown of upd3 in EECs using voilà-GAL4 in combination with Tubulin-GAL80ts (voilà>) (N=7). b, upd2 and upd3 expression levels in midguts in animals with EEC knockdown of upd2 or upd3 using voilà-GAL4 in combination with Tubulin-GAL80ts and R57C10-GAL80 (EEC>) (N=4-6). c, Confocal imaging of upd3-GAL4-driven UAS-GFP expression in EECs, co-stained with Prospero (Pros) as an EEC marker (Scale bar, 20 µm). d, Twenty-four-hour sleep profiles in animals with EEC-specific upd2 or upd3 knockdown (N=29-34). e, Total daytime sleep and f, nighttime sleep durations in animals with EEC-specific upd2- or upd3-knockdown flies (N=29-34). g, Twenty-four-hour sleep profiles for global upd2/3 deletion mutants (N=19-32). h, Daytime and i, nighttime sleep durations in global upd2/3 mutants (N=19-32). j, Sleep profiles following EEC-specific CRISPR-mediated upd2 or upd3 knockout (N=24-31). k, Daytime and l, night sleep durations in animals with EEC-specific CRISPR-mediated upd2 or upd3 knockout (N=25-31). m, Twenty-four-hour sleep profiles in animals with AstC-positive-EEC-specific knockdown of upd2 or upd3 using AstC-GAL4 combined with R57C10-GAL80 (AstCGut>) (N=29–30). n, Daytime sleep, and o, nighttime sleep durations in animals with AstCGut>-mediated knockdown of upd2 or upd3 (N=28–32). p, Twenty-four-hour sleep profiles in animals with Tachykinin-positive-EEC-specific knockdown of upd2 or upd3 using Tk-GAL4 combined with R57C10-GAL80 (TkGut>) (N=32). q, Daytime sleep, and r, nighttime sleep durations in animals with TkGut>-mediated knockdown of upd2 or upd3 (N=28–31). Statistical analyses performed using Mann-Whitney tests for panels a and b; ordinary one-way ANOVA with Dunnett’s multiple comparisons for panels e, j, h, i, n, o, q, and r; Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panels f, k, and l. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

EEC-derived Unpaired signaling regulates glial JAK-STAT activity that modulates sleep.

a, Day and night sleep measurements for flies with knockdown of IL-6 related Unpaired cytokine receptor domeless (dome), which activates JAK-STAT, in neurons (R57C10-GAL4, R57C10>) and glial cells (repo-GAL4, repo>) (N=25-39). b, Twenty-four-hour sleep profiles for controls and animals with glia-specific dome knockdown (N=25-32). c, Total day sleep duration and d, total night sleep duration for animals with glia-specific dome knockdown and control flies (N=25-32). e, Motion-bout length and f, motion-bout activity in animals with glia-specific dome knockdown and controls (N=25-32). g, Representative images of brains of from controls and animals with EEC knockdown upd2 or upd3 using voilà>, with 10xSTAT-GFP. GFP expression (green) reflects JAK/STAT activity, and Repo labeling (red) indicates glial cells (Scale bar, 50 µm). Insets show zoomed views of STAT-GFP+ and Repo+ glial cells (Scale bar, 15 µm). h, Quantitative analysis of GFP intensity in the layer of glial cells located at the surface of the brain in animals with EEC knockdown of upd2 or upd3 and controls (N=8). Statistical analyses were conducted using parametric t-tests for panel a; Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panels c, d, e, and f; and ordinary one-way ANOVA with Dunnett’s multiple comparisons for panel h. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

EEC-derived Unpaired signaling modulates sleep in response to intestinal oxidative stress.

a, Measurement of upd2 and upd3 expression in the midgut upon 20-hour treatment with 1% H2O2-laced food or with overexpression of upd3 (upd3-OE) using EEC> (N=4). b and c, Assessment of sleep duration over consecutive days during the daytime (ZT0-ZT12) and nighttime (ZT12-ZT24) in animals exposed to food containing (b) 0.1% H2O2 (N=26-31) or (c) 1% H2O2 (N=23-29). d, Daily sleep, and e, nightly sleep amounts measured over one day under standard food conditions followed by two consecutive days on 1% H2O2-containing food in animals with EEC-specific upd2 or upd3 knockdown and controls (N=23-30). Experiments measuring sleep levels in controls and animals lacking EEC-derived upd2 or upd3 were performed concurrently and share the “control” data, but results are presented in separate figures (b, c, d, and e) for clarity. In d, two-way ANOVA revealed significant genotype x diet interactions for upd2-i (p = 0.0076), upd3-iKK (p = 0.0003), upd3-iTRiP (p = 0.0204), and upd3-iGD (p = 0.0040), relative to the control, indicating that the sleep response to oxidative stress depends on EEC-derived Unpaired signaling. f, Sleep profiles and measurements of daytime (g) and nighttime (h) sleep across a two-day period, encompassing one day on standard diet followed by one day on 1% H2O2-laced food to induce oxidative stress, in flies with AstC-positive-EEC-specific knockdown of upd2 or upd3 using AstCGut> compared to controls (N=31–32). i, Survival rates under a 1% H2O2-induced oxidative stress diet in controls and animals with EEC-specific upd2 or upd3 knockdown (N=23-30). j, A 48-hour sleep profile comparison between global upd2/3 mutants and w1118 controls under one day of standard food conditions followed by one day of 1% H2O2-induced stress (N=18-63). k and l, Quantification of daytime and nighttime sleep durations in upd2/3 mutants versus w1118 controls under normal-food conditions and the following day exposed to food containing 1% H2O2 (N=18-63). In k, two-way ANOVA showed significant genotype × diet interaction (p < 0.0001), confirming a role for Unpaired cytokines in ROS-induced sleep modulation. m, Observation of sleep patterns, and n, measurements of daytime sleep across a three-day period, encompassing a day on standard diet, subsequent day on 1% H2O2-laced food to induce oxidative stress, and a final day back on standard diet to monitor recovery, in flies with EEC-specific overexpression of upd3 (upd3-OE) compared to controls (N=29-32). Statistical analyses were performed using Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panels a, b, c, d, e, i, and n; Mann-Whitney test for panels g, h, k, and l. Interaction effects were assessed using two-way ANOVA where indicated. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

Activation of glial JAK-STAT signaling by EEC-derived Unpaired cytokines in response to enteric oxidative stress.

a, Representative images of brains from flies expressing the STAT-::dGFP::2A::RFP reporter, where green (dGFP) reflects recent JAK-STAT activity due to its rapid degradation, and purple (RFP) indicates longer-term pathway activation due to its higher stability. Left panels show brains at lights-on (ZT0), middle panels show brains at lights off (ZT12), and right panels depict brains after 20 hours of oxidative stress induced by 1% H2O2-containing food, imaged at lights-on time ZT0. White dotted lines outline the brain perimeter. Scale bar, 50 µm. b, Ratio of dGFP to RFP fluorescence intensity at ZT0, at ZT12, and after oxidative stress (at ZT0), as depicted in panel a, to show dynamic changes in JAK-STAT activity (N=108-461, indicating the number of cells counted). c, Representative images of brains displaying 10xSTAT-GFP expression under homeostatic conditions and after oxidative stress in control flies (“Ctrl”) and flies with EEC-specific upd2 or upd3 knockdown. Scale bar, 50 µm. Inset panels provide magnified views of glia cells labeled by anti-Repo. Scale bar, 15 µm. d, Quantification of 10xSTAT-driven GFP intensity in glial cells under homeostatic and oxidative-stress conditions, demonstrating the impact of EEC-specific cytokine knockdown (N=8, indicating the number of brains). e, qPCR analysis of dome expression in the brains of flies with EEC-specific upd3 knockdown in comparison to voilà> controls (N=5). Statistical analyses were conducted using Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panel b; ordinary one-way ANOVA with Tukey’s multiple comparisons for panel d; and two-sided unpaired t-tests for panel e. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

EEC-derived Unpaired and glial Domeless signaling modulate sleep during intestinal oxidative stress.

a, Daytime sleep duration in flies with glia-specific dome knockdown under standard and oxidative-stress conditions induced by 1% H2O2 in food (N=25-32). Two-way ANOVA revealed significant genotype × diet interaction (p < 0.0001), indicating that glial Domeless is required for sleep regulation during oxidative stress. b, Daytime sleep duration in flies with repo-driven dome knockdown restricted to the adult stage using Tub-GAL80ts (repoTS>) under normal conditions and during exposure to 1% H2O2-containing food (N=29-32). Two-way ANOVA showed a significant genotype × diet interaction (p < 0.0001), further supporting a role for glial dome in regulating sleep in response to gut oxidative stress. c, Daytime sleep during a three-day period, encompassing a day on standard diet, subsequent day on 1% H2O2-laced food to induce oxidative stress, and a final day back on standard diet to monitor recovery, in controls and animals with glia-specific dome knockdown (N=24-32). Two-way ANOVA revealed significant genotype × diet interaction (p < 0.0001). d, Survival rates of controls and flies with glial-specific dome knockdown after exposure to oxidative stress by 1% H2O2-laced food (N=31). e, Sleep profiles, and f, daytime sleep duration for animals with EEC-specific upd3 knockdown compared to control flies across a 36-hour period encompassing 24 hours on standard diet followed by 12 hours on oxidative-stress conditions induced by 4% H2O2-containing food (N=15-30). Two-way ANOVA showed a significant genotype × diet interaction (p < 0.0001). g, Sleep profiles, and h, daytime sleep duration for animals with glia-specific dome knockdown compared to control flies across a 36-hour period encompassing 24 hours on standard diet followed by 12 hours under oxidative-stress conditions induced by 4% H2O2-containing food (N=20-32). Two-way ANOVA revealed significant genotype × diet interaction (p < 0.0001). i, Nighttime sleep durations for animals under 4% H2O2 oxidative-stress conditions in controls and animals expressing adult-restricted knockdown of dome in glia (N=31-32). Two-way ANOVA revealed significant genotype × diet interaction (p < 0.0001). Statistical tests: Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panels a, c, and d; Unpaired two-sided t-tests for panels b, f, and h; and Mann-Whitney test for panel i. Interaction effects were assessed using two-way ANOVA where indicated. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

BBB glia drive Domeless-mediated sleep responses to intestinal oxidative stress.

a, Representative images showing GFP expression driven by 10xSTAT-GFP in controls and animals with voilà-GAL4 (voilà>)-driven upd3 knockdown in EECs under intestinal ROS induced by 20 hours’ exposure to 1%-H2O2-containing food. The top panels depict overall brain and ventral nerve cord (VNC) structure with views of surface or deeper layers; the bottom panels provide zoomed-in views, highlighting the BBB glia at the interface between the brain and external environment. Dotted lines indicate brain and VNC perimeters. Scale bars, 50 µm (top) and 15 µm (bottom). b, Quantification of GFP intensity in the brain and VNC in BBB glia in controls and animals with EEC knockdown of upd3 under ROS stress, induced by exposure to 1% H2O2-laced food (N=7). c, Daytime, and d, nighttime sleep durations in flies with BBB-glia-specific knockdown of dome or overexpression of hopTumunder normal conditions, during exposure to 1% H2O2-containing food, and subsequent recovery on normal diet (N=23-32). In c, two-way ANOVA revealed significant genotype × diet interaction (p < 0.0001), indicating that BBB-glial Domeless is required for daytime sleep regulation under oxidative stress. e, Daytime, and f, nighttime sleep durations in flies with BBB-glia-specific knockdown of dome or overexpression of hopTum under normal conditions, during exposure to 4% H2O2-containing food, and subsequent recovery on normal diet (N=23-32). In e, two-way ANOVA revealed significant genotype × diet interaction (p < 0.0001), confirming the importance of BBB-glial Domeless signaling during higher levels of oxidative stress. Statistical tests used: Unpaired two-sided t-tests for panel b; Kruskal-Wallis ANOVA with Dunn’s multiple comparisons in panels c, d and f; ordinary one-way ANOVA with Dunnett’s multiple comparisons for panel e. Interaction effects were assessed using two-way ANOVA where indicated. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

Gut Unpaired cytokine signaling inhibits wake-promoting AstA signaling.

a, Images of brains from animals with AstA-R1-GAL4 and AstA-R2-GAL4 driving nuclear dsRed (nRFP, magenta) and co-stained with anti-Repo antibodies (green) show glial cells. The yellow dashed line indicates the interface between the brain and the external space, with the area below housing the (BBB glial cells (Scale bar, 10 µm). The yellow demarcation accentuates the separation between the cerebral interior and the external milieu, identifying the location of BBB glial cells underneath this partition (Scale bar is 10 µm). b, Relative expression of AstA-R1 and AstA-R2 in heads from animals with upd3 knockdown in EECs driven by voilà-GAL4 (voilà>) compared to the control group after 20 hours on 1% H2O2-laced food to induce oxidative stress (N=5-6). c, Relative expression of AstA-R1 and AstA-R2 in brains from animals with dome knockdown in glial cells driven by repo-GAL4 (repo>) compared to the control group after 20 hours on 1% H2O2-laced food to induce oxidative stress (N=6). d, Sleep patterns, and e, daytime sleep across a three-day period, encompassing a day on a standard diet, a subsequent day on 1% H2O2-laced food to induce oxidative stress, and a final day back on the standard diet to monitor recovery, in flies with BBB-glia-specific knockdown of AstA-R2 compared to control (N=28-32). In e, two-way ANOVA revealed significant genotype × diet interaction (p = 0.0114, supporting a role for glial AstA-R2 in ROS-induced sleep regulation. f, AstA transcript levels in brains and midguts from controls and animals with knockdown of AstA in AstA+ EECs using AstA-GAL4 (AstA>) in combination with R57C10-GAL80 to suppress neuronal GAL4 activity, referred to as AstAGut> (N=5-6). g, Sleep profiles, and h, daytime sleep on standard food of animals with AstA knockdown in AstA+ EECs using AstAGut> and controls (N=30-32). i, Sleep profiles, and j, daytime sleep on standard food of controls, animals with TrpA1-mediated activation of AstA+ EECs, and animals with TrpA1-mediated activation of AstA+ EECs with simultaneous knockdown of AstA (N=30-32). k-m, Quantification of AstA transcript levels in whole midguts (k) and AstA peptide levels in the R5 region of the posterior midgut (l) on standard diet, after 1 day on 1% H2O2-laced food to induce oxidative stress, and during recovery following H2O2 exposure (k: N = 5–6; l: N = 126–170). m, Representative images of R5 regions stained with anti-AstA antibody (scale bar: 25 µm). n, Diagram illustrating the role of enteroendocrine cells (EECs) in regulating wakefulness and sleep through Unpaired cytokine signaling under homeostatic and stress conditions. Left: Under homeostatic conditions, EECs release baseline levels of Unpaired, which interacts with the blood-brain barrier (BBB) to maintain normal JAK-STAT signaling and AstA transduction, promoting wakefulness. Right: In response to stress and disease, reactive oxygen species (ROS) increase in EECs, leading to elevated release of Unpaired. This surge in Unpaired upregulates JAK-STAT signaling in BBB glia, which inhibits wake-promoting AstA signaling by suppressing AstA receptor expression, thus resulting in increased sleep, a state termed “sickness sleep,” to promote recovery. The diagrams depict the gut lining with EECs highlighted, the interface with the BBB, and the resulting systemic effects on the organism’s sleep-wake states. EC; enterocyte. Statistical tests used: Unpaired two-sided t-tests for panel b, c and f; Ordinary ANOVA with Dunnett’s multiple comparisons for panel e, h, j, k, and l. Interaction effects were assessed using two-way ANOVA where indicated. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

Effects on sleep, feeding, and metabolic parameters of Unpaired cytokine manipulation in EECs.

a, Expression levels of upd2 and upd3 in brains of animals with EEC-specific knockdown using voilà-GAL4 in combination with Tubulin-GAL80ts and R57C10-GAL80 (together, “EEC>”), showing no significant changes in expression (N = 6). b, Representative fluorescent in situ hybridization images showing the co-expression of upd3 and the EEC marker prospero in the midgut (Scale bar, 10 µm). c, Twenty-four-hour sleep profiles for animals with EEC knockdown of upd2 or upd3 using voilà-GAL4 (voilà>). d, Day, and e, night sleep durations for EEC upd2 or upd3 knockdown and control groups (N=50-61). f, Sleep profiles for upd3 knockdown using R57C10-GAL80, Tub-GAL80ts, voilà> (EEC>) and UAS-upd3-RNAi controls (N=23-31). g, Day, and h, night sleep durations for UAS-RNAi controls and EEC-specific upd2 or upd3 knockdown (N=23-31). i, j, Motion-bout length and k, l, motion-bout activity for upd2 or upd3 knockdown using EEC> or voilà> and control flies (N=23-60). m, Motion-bout length and n, motion-bout activity for EEC upd2 or upd3 knockdown and UAS-RNAi control flies (N=21-32). o, Total feeding time for EEC-specific upd2 or -3 knockdown and control flies using FLIC (N=9-36). p, Food intake over 1 hour for animals with EEC-specific upd3 overexpression (upd3-OE) and controls (N=13-20). q, Relative triacylglyceride (TAG) levels in EEC-specific upd3 knockdowns and control flies (N=5-8). r, Expression levels of upd3 in midguts of animals with AstC-positive-EEC-specific knockdown of upd3 using AstC-GAL4 combined with R57C10-GAL80 (AstCGut>) (N=5). Statistical analyses were performed using two-sided unpaired t-tests for panel a, g, h, q, and r; Mann-Whitney tests for panels m, n, o, and p; ordinary one-way ANOVA with Dunnett’s multiple comparisons panel i; Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panels d, e, j, k, and l. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

Characterization of sleep patterns, activity, and metabolic impact of dome knockdown in glial cells.

a-c, Sleep profiles over a 24-hour period for animals with glia-specific dome knockdown and UAS-RNAi controls (N=25-32). d, Total daytime sleep, and e, nighttime sleep for UAS-RNAi controls and animals with glia-specific dome knockdown (N=25-32). f, Duration of motion bouts and g, motion-bout activity for the same groups as in a-d (N=25-32). h, Food intake measured over a 1-hour period for animals with dome knockdown in glia and UAS-RNAi control flies (N=23-31). i, Triacylglyceride (TAG) levels relative to controls in glia-specific dome knockdown flies. Statistical analyses were performed using Mann-Whitney tests for panels d, e, f, g, h, and i, except t-tests were performed for dome-iTRiPversus control in panel d and g and for dome-iGD versus control in panel d. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

Sleep duration influenced by EEC-specific Unpaired cytokine disruption and dietary changes.

a, Daytime and nighttime sleep duration in control flies fed standard food prepared and replaced daily without H2O2 to control for potential effects of the supplementation procedure (N=32). b-e, Daytime sleep duration in flies with upd2 or upd3 knockdown in EECs and UAS-RNAi controls measured over one day under standard food conditions followed by two consecutive days on 1% H2O2-containing food (N=21-32). f, g TUNEL staining of adult Drosophila brains following 24-hour feeding with 1% H2O2. TUNEL-positive cells per brain are shown in f (N=6), with representative images shown in g (scale bar, 50 µm). h, Sleep profiles over a 48-hour period for flies with EEC-specific upd3 knockdown under a transition from standard to 1% agar starvation diet compared to control flies (N=27-32). i and j, Daytime and nighttime sleep durations for flies with EEC-specific upd3 knockdown under standard diet and 1% agar starvation diet conditions (N=27-32). Statistical tests: t-tests for panel f; Kruskal-Wallis ANOVA with Dunn’s multiple comparisons was used for panels b, c, d, and e; two-way ANOVA with Sidak’s multiple comparisons was used for panels i and j. Data are presented as mean ± SEM. ns, non-significant (p>0.05).

Expression of upd2 and upd3.

Transcript levels of a, upd2, and b, upd3 in heads of animals with RNAi-mediated knockdown in EECs using voilà-GAL4 in combination with Tubulin-GAL80ts (voilà>) (N=6). Statistical analyses were performed using t-tests for panel a and b. Data are presented as mean ± SEM.

Impact of glia-specific dome knockdown on sleep patterns following oxidative stress and after sleep deprivation.

a-c, Daytime sleep measured over three days, starting with a normal diet, followed by a day with 1% H2O2-supplemented food to induce oxidative stress, and concluding with a return to a normal diet to assess recovery, in flies with UAS-RNAi constructs without dome knockdown (UAS-RNAi controls) and those with glia-specific dome knockdown (N=25-31). d, Sleep profiles during a night that included a 6-hour period of sleep deprivation, followed by a recovery phase, in control flies and those with glia-specific dome knockdown (N=17-24). e, Sleep quantity measured during the first 2 hours of the recovery period following sleep deprivation in control flies and those with glia-specific dome knockdown (N=17-24). Statistical tests: Kruskal-Wallis ANOVA with Dunn’s multiple comparisons for panels a, b, and e; ordinary one-way ANOVA with Dunnett’s multiple comparisons for panel c. Data are presented as mean ± SEM. ns, non-significant (p>0.05)

AstA signaling from EECs promotes wakefulness.

a, AstA-R2 expression in heads of flies with knockdown of AstA-R2 in BBB glia using moody-GAL4 (moody>) compared to the control group (N=5). b, Sleep profiles over two consecutive days, encompassing one day under standard diet conditions followed by one day on oxidative stress conditions, induced by 1% H2O2-containing food, in animals with BBB glia-specific knockdown of AstA-R1 or AstA-R2 and controls (N=30-32). c, Daytime sleep durations over two consecutive days, encompassing one day under standard diet conditions followed by one day on oxidative stress conditions, induced by 1% H2O2-containing food, in animals with BBB glia-specific knockdown of AstA-R1 or AstA-R2 and controls (N=30-32). d, Sleep profiles, and e, daytime sleep on standard food in animals with AstA knockdown in AstA+ EECs and UAS-RNAi controls (N=30-31). f, Single-cell RNA sequencing data from adult Drosophila midguts from the Fly Cell Atlas dataset52 were visualized with the SCope viewer. AstA (red), Tk (blue), and TrpA1 (green) expression is shown across the EEC clusters. AstA-positive EECs form a distinct cluster that does not overlap with TrpA1 expression. In contrast, Tk-positive EECs form a separate cluster, and a subset of these cells express TrpA1. Right: Zoom-in view of the Tk-positive EEC cluster highlights cells co-expressing Tk and TrpA1. These data indicate that TrpA1 is selectively expressed in Tk-positive EECs and not in AstA-positive EECs. Statistical tests used: Two-sided unpaired t-tests for panel a and e; two-way ANOVA with Sidak’s multiple comparisons was used for panel c. Data are presented as mean ± SEM.

Oligos used for cloning the upd2 and upd3 CRISPR constructs, with gRNA sequences indicated in bold and underlined.

List of qPCR primers