The microbiome stabilizes circadian rhythms in the gut

Significance We demonstrate how the microbiome and rhythms of feeding impact circadian rhythms in the gut. Using a Drosophila model, we show that restricting feeding to specific times of the day strengthens circadian rhythms of gene expression. However, as compared with ad libitum feeding, time-restricted feeding increases sensitivity of the animals to stressors. This challenges the idea that time-restricted feeding, previously associated with metabolic benefits, promotes fitness. The microbiome, conversely, reduces the strength of circadian cycling in the gut. It also tempers the response of the gut clock to changes in the day:night cycle, thereby allowing circadian rhythms in the gut to remain synchronized with rhythms in the brain. The findings indicate that the microbiome promotes circadian synchrony in the animal.


Generation and maintenance of fly lines
Sterile flies were generated with the Wohlbachia-minus iso31 fly line. Newborn fly embryos within 12 hours were rinsed in 100% ethanol, dechorionated in 10% bleach for 2 minutes, and immediately rinsed three times in sterile PBS. The sterile embryos were fed autoclaved standard sterile molasses-cornmeal-yeast medium containing 1 mM kanamycin, 650 μM ampicillin (61-238-RH, MediaTech), and 650 μM doxycycline (D9891, Sigma-Aldrich). Germ-free fly lines were maintained on sterile medium with three antibiotics for four generations and then maintained on the same medium without antibiotics for the next four generations. Bacterial contamination of flies was monitored by homogenizing larvae and testing for bacterial growth on DeMan, Rogosa and Sharpe (MRS)-agar plates. For timed feeding, adult flies were provided standard sterile medium in groups of 40 for 3 days. 5-7 day old flies were separated by gender and subjected to the feeding paradigm in the presence of light:dark cycles starting on day 5. Ad libitum (AF) and flies subjected to timed feeding (TF) were shifted to sterile medium vials at zeitgeber time 0 (ZT0), (which corresponded to actual time of 9:00 am) and then were transferred to either a new sterile medium or a 1.1% agar vial at ZT10 (7pm), ZT12 (9pm) or ZT14 (11pm) for 14-h, 12-h or 10-h fast respectively.

Sample collection and 16S sequencing
Feces from the walls of the sterile tube were collected by using sterile cotton swabs at time points ZT0, ZT4, ZT8, ZT12, ZT16 and ZT20. For each strain, iso31 or per 01 at least 4 replicates were included at all time points for female and male flies respectively. The spin swab with feces was transferred into a reaction tube and the stick was carefully cut with sterile scissors. DNA was extracted by using the Qiagen DNeasy PowerSoil Pro kit and extracted DNA was quantified with the Quant-iT PicoGreen Assay Kit. For 16S rRNA marker gene sequencing, PCR amplification was performed using primers that targeted the V1-V2 region of the 16S rRNA gene, followed by sequencing using 250 bp paired-end sequencing on an Illumina MiSeq instrument. Sequence data were analyzed using the QIIME2 bioinformatics pipeline [PMID 31341288]. Taxonomic abundances were compared using linear mixed effects models after log transformation. Cyclic patterns in alpha diversity and taxon abundance were detected using the JTK_Cycle method from the MetaCycle R package.

Immunity and anti-stress test
For infection, LB containing bacteria (S. marcescens) was diluted in PBS (to OD600=0.02) together with 1% food coloring (Brilliant Blue FCF). The injection control solution contains equivalent amounts of LB medium and 1% blue food coloring in PBS. Bacterial and control solutions were stored on ice until the time of injection, at ZT2 or ZT12. Survival of infected flies was measured daily by counting the number of live flies in each test tube. For the heat shock experiment, flies were shifted to 37°C after 4 days and 21 days continuously fast and then surveyed every 12-h for survival rate. For the starvation test, AF and TF flies were transferred into 1.1% agar vials at 25°C under 12:12 LD conditions and surveyed every 12-h for survival rate. Each experiment was repeated at least 4 time, with 15-16 flies per repetition.

RNA-seq and data analysis
The quality of RNA isolated with the SV Total RNA Isolation kit (Z3105, Promega) was assessed by Fragment Analyzer High Sensitivity RNA Assay (DNF-471-0500, Agilent Technologies) and quantified by Qubit 2.0 RNA HS assay (Q32855, ThermoFisher). Paramagnetic beads coupled with oligo d(T)25 were combined with total RNA to isolate poly(A)+ transcripts based on the NEBNext® Poly(A) mRNA Magnetic Isolation Module manual (E7490L, NEB). Prior to first strand synthesis, samples were randomly primed (5´ d (N6) 3´ [N=A,C,G,T]) and fragmented based on manufacturer's recommendations. The first strand was synthesized with the Protoscript II Reverse Transcriptase with a longer extension period, approximately 40 minutes at 42 ⁰C. All remaining steps for library construction were according to the NEBNext® Ultra™ II Non-Directional RNA Library Prep Kit for Illumina® (E7760L, NEB). Final libraries were quantitated by Qubit 2.0 (Q32855, ThermoFisher) and quality was assessed by TapeStation D1000 ScreenTape (5067-5582, Agilent). Illumina® 8-nt dual-indices were used. Equimolar pooling of libraries was performed based on QC values and sequencing was conducted by Admera on an Illumina® NovaSeq S4 (Illumina, USA) with a read length configuration of 150 PE for 40 M PE reads per sample (20 M in each direction). For RNA-sequence analysis, FASTQ data were mapped to the fly genome (Number) by using Hisat2 (http://daehwankimlab.github.io/hisat2/) and alignment results were assayed by the LiBiNorm tool (https://warwick.ac.uk/fac/sci/lifesci/research/libinorm/) based on the reference genome GENCODE. Raw data count was analyzed by IDEP v0.95 (http://bioinformatics.sdstate.edu/idep/) for differentially expressed genes (DEGs) between different conditions. The threshold value CPM>0.5 and n libraries >3 and regularized log transformation were respectively used to filter out low-expressed genes and transform raw count data for further clustering and PCA analysis, then the TPM data were used for rhythmicity analysis through Nitecap (https://nitecap.org/) and MetaCycle (https://academic.oup.com/bioinformatics/article/32/21/3351/2415176).
Cycling transcription factors were identified through (https://flybase.org/reports/FBgg0000745.html). Each condition assayed in the RNAseq experiment was represented by three biological replicates.

Quantitative real-time PCR and Western blot analysis
Total RNA was extracted by using an RNA kit (Z3105, Promega) and cDNAs were synthesized by using a high-capacity cDNA Archive kit (4374966, Applied Biosystems). Assays were run on a ViiA7 Real-Time PCR system (prism 7100, Applied Biosystems) by using a SYBR Green kit (4364346, Applied Biosystems).
Relative gene expression was calculated using the 2 -∆∆CT method normalizing to RP49. Primer sequences are listed in Dataset S1M. For Western blot analysis, female bodies (flies injected with HDAC inhibitors) or male guts from flies subjected to ad libitum or 14-h timed feeding regimes at indicated time points were collected for protein extraction. All the protein samples were isolated by using the lysis buffer (87788, Thermo Scientific) with protease inhibitor (11873580001, Roche). Primary antibodies Rabbit anti-Acetylated Histone H4 (AHP418, Bio-Rad) and Rat anti-GAPDH (10494-1-AP, Proteintech) were used at 1:1000 and 1:1000 dilution, respectively. FIJI ImageJ software was used for analysis.

Fly feeding rhythm, activity and weight assays
The CAFE feeding assay was performed as described previously with small modifications (1). In brief, plastic vials containing agar on the bottom were placed in a custom made glass chamber filled with water that allowed air flow while maintaining high humidity. Glass micropipettes (5ul, catalog no. 53432-706; VWR, West Chester, PA) were filled with 1 cm of mineral oil and liquid medium containing 5% sucrose and 0.05% Brilliant Blue. The experiment was conducted under an 12:12 LD cycle in an incubator at 25℃ and >60% humidity in the chamber. Four male flies (5-7 day old) per vial were trained for 12-h. During the assay, capillaries were measured and replaced every 3-h across one day (ZT1, ZT4, ZT7, ZT10, ZT13, ZT16, ZT19, ZT23). At every time point a vial without flies was used to control for evaporation and the change in volume was subtracted from the experimental readings. Fly activity was recorded by using an updated multi-beam (MB) version of the Drosophila Activity Monitor (DAM) system, in which 15 beams of infrared detectors were used to detect the movement of single fly in a glass locomotor tube. We used the total number of movements (MT) data collected by the MB-DAM system, as it reflects the breaking of two beams, and thereby active locomotion of the fly. The total and mean MT per minute over 10-h after fly loading were calculated by R and used to indicate the fly's physical activity. An electronic balance (Sartorius CP124S, precision 0.1 mg) was used for weighing fly bodies in groups of 100.     Figure S4. Feeding rhythms do not differ between AS and AM flies A: Food consumption across 24h in AS and AM male flies. nAS=20, nAM=20, twoway ANOVA with Tukey's multiple comparisons test (*** p adj < 0.001; **** p adj < 0.0001), only relevant comparisons shown. B: Total day (average of four measurements from ZT1 to ZT13) and night intake (average of four measurements from ZT13 to ZT1) by AS and AM flies. nAS=20, nAM=20, p<0.05, one-way ANOVA, only relevant comparisons shown. Figure S5. Transcript cycling in timed fed germ-free flies is stronger than in microbiome-containing flies A: Polar histogram plots of the peak phase for transcripts that cycle in TM and TS conditions. B: Amplitude comparison for transcripts that overlap in cycling in TM and TS. Data are mean ± SEM, ****p < 0.0001 shown by Student's t test. C: Phasesorted heatmaps for transcripts whose oscillations are lost, shared or gained in TS flies versus TM flies. GO biological process enrichment analysis of oscillating transcripts is shown at the bottom. D: Heatmap showing the amplitudes of cycling transcripts shared across AM, TM, AS and TS with p < 0.01, JTK_cycle in 12:12 LD. GO biological process enrichment analysis of the shared oscillating transcripts is shown at the bottom. E: Comparison of amplitudes of cycling transcripts shared across AM, TM, AS and TS with p < 0.01, JTK_cycle in 12:12 LD. Data are mean ± SEM, **p < 0.01 and ***p < 0.001 determined by one-way ANOVA and Tukey's multiple comparison test.   Figure S7. The cycling of metabolic transcripts is modulated by timed feeding and by the microbiome Transcript abundance and JTK_amplitude analysis of sixteen metabolic genes at different timepoints in a 12:12 L:D cycle under AM, TM, AS and TS conditions (by RNAseq). Data are mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001 determined by one-way ANOVA and Tukey's multiple comparison test. Based upon the timing of peak expression under TM and TS, as determined by JTK_cycle, transcripts are shown in (A-F) peak expression ZT6-9, (G-L) peak expression ZT0-3 and (M-P) peak expression ZT12-21.  . Plots on the right show a comparison of amplitudes for cycling transcripts shared by each set of two conditions. p < 0.05 by JTK_cycle was used to assess cycling. Data are mean ± SEM, *p < 0.05 shown by Student's t test). G: Polar histogram plots of the peak phase for oscillating transcription factors under AM, TM, AS and TS conditions, using a JTK_cycle value of p < 0.05. The distribution of phases changes with timed feeding, but does not very much in sterile and microbiome-carrying flies. H: Polar histogram plots of the peak phase for oscillating transcripts under AM, TM, AS and TS conditions, using a JTK_cycle value of p < 0.05. The distribution of phases changes with timed feeding, but does not very much in sterile and microbiome-carrying flies. I-N: Transcript abundance and JTK_amplitude analysis of metabolism-related transcription factors at different timepoints in a 12:12 h L:D cycle under AM, TM, AS and TS conditions. Data are mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001 determined by one-way ANOVA and Tukey's multiple comparison test. CT-Control. Data are mean ± SEM, *p < 0.05, **p < 0.0, ***p < 0.01 and ****p < 0.01 indicated by Student's t test. N=4 biological replicates. B-C: Expression of two representative transcription factors (CrebA and Foxo) and three clock genes (per, tim and Clock) 24-h after injection of VA. Data are mean ± SEM, *p < 0.05, **p < 0.0, ***p < 0.01 and ****p < 0.01 indicated by Student's t test. N=4 biological replicates. D: Four metabolic genes show different responses to PBA and VA. Data are mean ± SEM, **p < 0.0, ***p < 0.01 and ****p < 0.01 determined by Student's t test. N=4 biological replicates for VA and 7 for PBA.

Fig.S11
Figure S11. Flies maintained on timed feeding paradigm are more sensitive to starvation A: Schematic showing experimental protocol for ad lib and timed feeding treatments. Survival curves with 95% confidence interval were generated in response to indicated stressors, N=4-5 biological replicates. B-C: Starvation of female flies subjected to ad lib (A) and timed feeding (T) treatments. The timed feeding consisted of a daily overnight 10-h or 14-h fast for 4 days (B) or 21 days (C). For these and all other assays, survival curves were compared via the logrank (Mantel-cox)-test. Each curve represents four biological replicates of n = 15-16 flies each per condition, Neither the 10-h or 14-h fast for 4 days (B) had a significant effect on response to starvation (p>0.10,). The 21 day TF significantly affected survival for the 10-h (p<0.0006) and 14-h fast (p< 0.0024).

Fig.S12
Figure S12. Effect of timed feeding on body weight and locomotor activity (A and B): Body weight of flies fed ad lib or subjected to timed feeding. Timed feeding consisted of a 10-h, 12-h or 14-h fast each day for 4 (A) or 21 days (B). (C and D): Locomotor activity was measured as beam crossings over a 10 hour period in Drosophila using the multibeam Activity Monitoring system (DAMS), n=48 flies in 3 independent replicates. Flies were either ad lib fed or maintained on 10-h, 12h or 14-h fast each day for 4 (C) or 10 days (D). Data are mean ± SEM, **p < 0.05 and *p < 0.05 determined by Student's t test. Body weight was determined for groups of 100 females. and AS (sterile) heads on the ninth day of LD changes, mRNA levels were measured by qPCR and normalized to the RP49. Data are mean ± SEM (n = 3) *p<0.05, and **p<0.01 determined by Student's t test. Phase is based on meta2d analysis. B: Expression pattern of per and tim in AM and AS fly guts on the third day of LD changes. mRNA levels measured by qPCR are normalized to RP49, data are mean ± SEM (n=4), (Control AM--black line; light:dark changes AM-dashed black line; Control AS--blue line; light:dark changes AS--dashed blue line).