a-c. VCO₂, VO₂, and respiratory quotient (RQ) profiles across the 24-hour cycle in different genotypes. VCO₂, VO₂ and RQ (VCO₂/VO₂) was continuously recorded at one-second intervals from Zeitgeber Time (ZT) 0 to 24 and subsequently averaged into 5-minute bins for analysis. Traces represent mean VCO₂, VO₂, and RQ values across the 24-hour recording period for wild-type flies under light-dark conditions (WT-LD), short-sleep mutants fumin (fmn) and sleepless (sss), the circadian clock mutant period01 (per01), and wild-type flies maintained in constant darkness (WT-DD). Source data are the continuous respirometry recordings. d-f. Genotype-specific differences in VCO₂, VO₂, and RQ measured over a full 24-hour recording period. Boxplots show average values of (left to right) respiratory quotient (RQ), carbon dioxide production (VCO₂), and oxygen consumption (VO₂) across genotypes and lighting conditions. Measurements were taken continuously over a 24-hour period using a flow-through MAVEn system. Genotypes include wild-type under light-dark (WT-LD) and constant darkness (WT-DD), short-sleep mutants fumin (fmn) and sleepless (sss), and circadian mutant per01. Group differences were assessed using the Kruskal-Wallis test, followed by Dunn’s multiple comparisons post hoc test. Significance is denoted as: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***); ns = not significant. g-i. Body Weight and Normalized VCO₂ and VO₂ Across Different Genotypes. Body weight (mg) and respiratory parameters (VCO₂ and VO₂) normalized to body weight (mg) were measured in wild type (WT), fmn, sss, and per01 mutant flies. Statistical significance was assessed using one-way ANOVA followed by Dunnett’s multiple comparisons test against WT. p < 0.05 (*), p < 0.01 (**), p < 0.001 (***); ns = not significant. For all panels with error bars or shaded error bands, values are presented as mean ± SEM. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers.

Twenty-four-hour time course of respiratory parameters across genotypes.

Normalized temporal profiles of carbon dioxide production (VCO₂) (top), oxygen consumption (VO₂) (middle), and respiratory quotient (RQ) (bottom) plotted over a 24-hour cycle (ZT0–24) for genotypes: wild-type in light-dark (WT-LD) and constant darkness (WT-DD), short-sleep mutants fumin (fmn) and sleepless (sss), and circadian mutant per01. Curves reveal genotype-specific rhythmicity and metabolic dynamics across the 24-hour recording period. Significant genotype × time interactions were observed for all variables (VCO₂, VO₂, RQ), indicating genotype-dependent alterations in respiratory rhythmicity. p < 0.05 considered significant. For all panels with error bars or shaded error bands, values are presented as mean ± SEM. Source data are the same continuous respirometry recordings shown in Figure 1; Figure 2 presents rhythmicity analysis of VCO₂, VO₂, and RQ across the 24-hour cycle. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers.

Phase distribution of respiratory rhythms across genotypes.

Polar plots depicting the phase (peak timing) of rhythmic expression for carbon dioxide production (VCO₂), oxygen consumption (VO₂), and respiratory quotient (RQ) over a 24-hour cycle (ZT0-24) for each genotype. Genotypes include wild-type in light-dark (WT-LD) and constant darkness (WT-DD), short-sleep mutants fumin (fmn) and sleepless (sss), and circadian mutant per01. Statistical significance of rhythmicity was determined using the RAIN algorithm: darker-colored bars indicate significant rhythms (p < 0.05), and lighter-colored bars indicate non-significant rhythms (p > 0.05). Where applicable, plotted values are presented as mean ± SEM. Source data are the same continuous respirometry recordings shown in Figure 1; Figure 3 presents phase/peak-timing analysis derived from the rhythmicity analysis

Rhythmicity metrics for respiratory parameters across genotypes

Temporal Profiling of Respiratory Quotient in Wild-Type Flies Under Light-Dark Conditions.

Respiratory quotient (RQ) was extracted every 2 h across a 24-hour light-dark (LD) cycle in WT flies. Lag analyses were performed by advancing or delaying the RQ data by −120, −60, −30, −15, −5, +5, +15, +30, +60-, and +120-minutes relative to Zeitgeber Time (ZT). Each panel represents the RQ profile corresponding to a specific time shift, illustrating phase-dependent respiratory dynamics across the 24-hour LD cycle. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers. RQ rhythmicity itself was assessed from the continuous respirometry time series (Table 1); Figure 4 shows the day-night RQ pattern used as the reference for the lag-based metabolite correlations, to which the metabolomics data contribute.

a. Rank-based correlation between Respiratory Quotient (RQ) and selected metabolites. Scatter plots depict the rank-order relationships between RQ and (A) Hydroxyhexadecenoylcarnitine and (B) Quinolinate. Each point represents a timepoint after aligning data with respective time shifts. Spearman’s rank correlation coefficient (ρ) and corresponding p-values are indicated on each panel. A positive lag (+120 minutes for Hydroxyhexadecenoylcarnitine) or negative lag (−120 minutes for Quinolinate) denotes the temporal shift in RQ relative to the metabolite dataset. Statistical significance was defined as p < 0.05. b. Temporal alignment of metabolite with respiratory quotient (RQ) in WT-LD flies. Z-scored time series of RQ, Hydroxyhexadecenoylcarnitine, and Quinolinate across a 24-hour light-dark cycle. Top Panel: Hydroxyhexadecenoylcarnitine shows a positive lag of +120 minutes and strong positive correlation with RQ (ρ=+0.78), suggesting its rise after changes in RQ. Bottom Panel: Quinolinate shows a negative lag of –120 minutes and strong negative correlation with RQ (ρ=-0.77), indicating it precedes changes in RQ.

Metabolite-Respiratory Quotient Correlations and Pathway Enrichment in WT-LD Flies.

(A) Heatmap depicting Spearman correlations (|ñ| > 0.7, p < 0.05) between respiratory quotient (RQ) and metabolite across the 24-hour LD cycle in wild-type flies maintained under light-dark (LD) conditions. Metabolites were clustered based on correlation patterns, highlighting groups with similar temporal associations with respiratory activity. (B) Pathway enrichment analysis of significantly correlated metabolites, illustrating metabolic pathways most closely linked to respiratory dynamics. Pathways with a p-value less than 0.05 were considered statistically significant.

Correlation of Metabolites with Respiratory Quotient and Pathway Enrichment in Short-Sleep Mutants.

(A, C) Heatmaps showing Spearman correlations (|ρ| > 0.7, p < 0.05) between respiratory quotient (RQ) and metabolites in the short-sleep mutants fmn (A) and sss (C). (B, D) Pathway enrichment analyses of metabolites significantly correlated with RQ in fmn (B) and sss (D). Pathways with p-values less than 0.05 were considered statistically significant.

Correlation of Metabolites with Respiratory Quotient and Pathway Enrichment in Circadian Mutant and Wild-Type Flies in Constant Darkness.

(A, C) Heatmaps displaying Spearman correlations (|ρ| > 0.7, p < 0.05) between respiratory quotient (RQ) and metabolites in the circadian mutant per01 (A) and wild-type flies maintained in constant darkness (WT-DD) (C). (B, D) Pathway enrichment analyses of metabolites significantly correlated with RQ in per01 (B) and WT-DD (D). Pathways with p-values less than 0.05 were considered statistically significant.

Baseline oxygen consumption in isolated gut tissue from fmn and per01 mutants.

Baseline oxygen consumption rate (OCR, fmol/mm²/s) measured by tissue respirometry in individual dissected guts from the short-sleep mutant fmn (A) and the circadian-clock mutant per01 (B) relative to iso31 genetic-background controls. For each gut, OCR was averaged over a 24-hour window beginning ∼12 h after loading, once tissue respiration had stabilized (12–36 h). To combine independent runs, OCR was normalized to the median of the iso31 controls within each run and log2(x+10)-transformed. One iso31 gut in the third run of the per01 experiment was excluded as an outlier; no other values were removed. Baseline OCR differed significantly between iso31 and each mutant (Mann-Whitney test, p < 0.05; pooled across 3 independent runs; fmn vs iso31, n = 12 vs 12; per01 vs iso31, n = 12 vs 11 after the iso31 outlier exclusion).

a-e. VCO₂ profiles across the 24-hour recording period in different genotypes. VCO₂ was continuously measured at one-second intervals from Zeitgeber Time (ZT) 0 to 24 and averaged into 5-minute bins for analysis. The traces display mean VCO₂ values across the 24-hour recording period for the following genotypes: wild-type flies under light-dark conditions (WT-LD), short-sleep mutants fumin (fmn) and sleepless (sss), the circadian clock mutant period01 (per01), and wild-type flies in constant darkness (WT-DD). For all panels, traces are presented as mean ± SEM.. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers.

a-e. VO₂ profiles across the 24-hour recording period in different genotypes. VO₂ was continuously measured at one-second intervals from Zeitgeber Time (ZT) 0 to 24 and averaged into 5-minute bins for analysis. The traces display mean VO₂ values across the 24-hour recording period for the following genotypes: wild-type flies under light-dark conditions (WT-LD), short-sleep mutants fumin (fmn) and sleepless (sss), the circadian clock mutant period01 (per01), and wild-type flies in constant darkness (WT-DD). For all panels, traces are presented as mean ± SEM. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers.

a-e. Respiratory quotient (RQ) profiles across the 24-hour recording period in different genotypes. RQ (VCO₂/VO₂) was continuously recorded at one-second intervals from Zeitgeber Time (ZT) 0 to 24 and subsequently averaged into 5-minute bins for analysis. Traces represent mean RQ values across the 24-hour recording period for wild-type flies under light-dark conditions (WT-LD), short-sleep mutants fumin (fmn) and sleepless (sss), the circadian clock mutant period01 (per01), and wild-type flies maintained in constant darkness (WT-DD). For all panels, traces are presented as mean ± SEM. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers.

Day-night differences in genotype-specific metabolic rates.

Boxplots depict day (ZT0-12) and night (ZT12-24) averages plotted side-by-side for each genotype to facilitate direct day-night comparison. Panels show genotype-specific differences in (A) carbon dioxide production (VCO₂), (B) oxygen consumption (VO₂), and (C) respiratory quotient (RQ). Genotypes include wild-type flies under light-dark conditions (WT-LD), wild-type flies maintained in constant darkness (WT-DD), short-sleep mutants fumin (fmn) and sleepless (sss), and the circadian mutant period01 (per01). Measurements were acquired continuously using a flow-through MAVEn system and binned into day and night intervals. Group differences were assessed using the Kruskal-Wallis test followed by Dunn’s multiple comparisons post hoc test. Significance is denoted as: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***); ns = not significant. Source data are the same continuous respirometry recordings shown in Figure 1. Data represent approximately 300 flies per genotype across three experimental days (25 flies/chamber, four chambers/experiment). The chamber was used as the experimental unit; n denotes the number of chambers.

Metabolites showing strong lag-dependent correlations with respiratory quotient (RQ) across genotypes and lighting conditions.

Note: Metabolites were included if they showed a strong metabolite–RQ association at one or more temporal lags, defined as |ρ| ≥ 0.7 with nominal p < 0.05. Nominal p-values are reported at the strongest lag for each metabolite. Benjamini–Hochberg false discovery rate (BH-FDR) adjusted p-values were calculated from the full metabolite × lag nominal p-value set within each genotype/condition, including all metabolites and all lag time points tested, and are reported at the strongest lag. BH-FDR adjusted p < 0.05 was considered significant after multiple-testing correction.

Metabolite Set Enrichment Analysis (MSEA ORA)

Experimental unit, sample size, and error representation for respirometry-based figures.

This table summarizes the experimental unit, chamber-level sample size, and error representation for all respirometry-based analyses shown in Figures 1-3 and Figures S1-S4. For all respirometry analyses, the chamber was used as the experimental unit. Each chamber contained 25 flies, and each genotype/condition was measured across four chambers per experiment over three experimental days, corresponding to n = 12 chambers per genotype/condition unless otherwise indicated.