Impact of peripheral circadian misalignment and alcohol on the resiliency of intestinal barrier and microbiota

Average weekly food and liquid intake were significantly affected by treatment and sex effects. Males consumed more food, as expected, but consumed less liquid compared to female mice. Body weight was assessed and was significantly affected by alcohol treatment and sex effects. Male mice were more affected by alcohol treatment, resulting in decreased average body weight. Average weekly alcohol intake was measured and there was a significant impact of the food timing paradigm on alcohol consumption (Supplemental Figure S1).

The blood alcohol levels were assessed with significantly elevated alcohol effects noted in the right-time eating alcohol-fed group (RTE EtOH), consistent with blood collected following 12 h of food intake and 20% EtOH (v/v) availability. It is significantly increased compared to the H₂O groups. However, there was not a significantly elevated blood alcohol in the wrong-time alcohol-fed (WTE EtOH) group due to a study constraint. The time of sacrifice occurred in the morning although we anticipate this trend to have continued with significantly increased levels if this group was sacrificed at night when the mice are coming off 12 h of food intake and 20% EtOH (v/v) availability (Supplemental Figure S2).

Previous studies of central circadian misalignment resulted in colonic hyperpermeability.^20,21 The apical junctional complex (AJC) of intestinal epithelial cell proteins contains the tight junction protein zonula occludens protein 1 (ZO-1), occludin, and the adherens junction protein E-cadherin that are key regulators of gut leakiness. We sought to determine whether our altered food timing paradigm and alcohol treatment resulted in changes in the expression of these key AJC proteins that regulate permeability in colon tissue.

Colon tissue from mice was assessed for any histological evidence of inflammation or tissue damage (Supplemental Table S1). Male mice from the RTE EtOH group experienced increased scores of histological evidence of inflammation and tissue damage. Overall, there is a sex and alcohol treatment effect (p = 0.0195).

However, we also evaluated the impact of alcohol treatment effect on stool microbial communities (i.e., RTE H₂O + WTE H₂O vs. RTE EtOH + WTE EtOH). While there were no significant differences in alpha diversity observed (i.e., variation within each sample), the analysis of beta diversity (i.e., variation between samples/ groups) revealed a significant impact of alcohol on the microbiota.

PERMANOVA analysis showed no differences between male and female groups. Subsequent microbiota analysis was run with combined male and female data. Alcohol-fed groups (RTE EtOH and WTE EtOH) had significantly different effects. Alcohol treatment drove the global changes to the microbiota profile seen in our group comparisons (Table 2).

PERMANOVA and PERMDISP results are based on Aitchison matrix distances for the multi-amplicon (V4) sequence data at the feature level. Significant values are based on 9,999 permutations and corrected for multiple testing using the Benjamini–Hochberg method (bolded, q < 0.05). Total mice sample sizes: RTE H₂O (n = 18); RTE EtOH (n = 17); WTE H₂O (n = 18); WTE EtOH (n = 14).

We analyzed the distance measure of the samples using Non-metric multidimensional scaling (NMDS) (Supplemental Figure S3). The distance between each group is indicative of each group’s dissimilarity and similarity to each other. The RTE H₂O and WTE H₂O groups are more like each other compared to both the RTE EtOH and WTE EtOH groups.

Phylum and genus taxonomic level differential abundance and compositional analysis of the microbial profiles between mice groups were analyzed with both DESeq2 and centered log-ratio Kruskal–Wallis (CLR-KW). These were compared to Boruta to identify important taxa. CLR-KW identified the same and more taxa compared to DESeq2. Of interest, significant taxa that were identified between H₂O and EtOH groups that were associated with alcohol consumption included: Clostridium sensu stricto 1, Muribaculaeceae, Bifidobacterium, Turicibacter, and Atopobiaceae (all q < 0.05) (Supplemental Excel sheet S1).

Taken together, the richness (alpha diversity) of the intestinal microbiota was not impacted by either food timing or alcohol treatment. However, the differences between the microbial communities (beta diversity) were driven mainly by alcohol treatment. The identified taxa of significance are associated with alcohol consumption, supporting how alcohol alters the microbial communities in our mouse groups.

We also assessed the biological pathways present between the RTE H₂O and WTE EtOH cecal content supernatant data using Reactome to predict equivalent human pathways.³³ A total of 55 pathways were identified (all p < 0.05) and a few pathways of interest to note are adherens junctions interactions, G-protein-coupled receptors (GPCR) signaling and binding signaling by GPCR, and γ-aminobutyric acid (GABA) signaling (supplemental Excel sheet S2). These pathways are decreased in the WTE EtOH samples which are implicated in alcohol consumption.³⁴

Significant correlations were noted between sucralose:lactulose ratio and Turicibacter (p-value = 0.034269, R-value = 0.394355), Occludin relative density and Clostridium sensu stricto 1 (p-value = 0.016971, R-value = −0.50322), ZO-1 relative density and Clostridium sensu stricto 1 (p-value = 0.035408, R-value = −0.42979). These results indicate that Turicibacter and Clostridium sensu stricto 1 may be important mediators of microbiota and epithelial barrier integrity in the context of alcohol consumption and serve as potential targets in future studies (supplemental excel sheet S3).

Food timing and alcohol treatment can induce intestinal inflammation that can subsequently lead to peripheral inflammation in organs throughout the body. We assessed a few inflammatory markers in the serum and stool. First, the pro-inflammatory cytokine IL-6 was examined in the serum for systemic inflammation. While there was no significant difference between groups, there was a slight increase in IL-6 levels seen in the alcohol-treated groups. Second, lipopolysaccharide (LPS) is a component in the outer membrane of gram-negative bacteria, and LPS-binding protein (LBP) is a type 1 acute-phase protein that binds to LPS to facilitate an immune response and is a well-accepted marker of intestinal barrier integrity and endotoxemia. There was a relative increase of serum LBP between food timing in both WTE groups. Third, we assessed calprotectin which is present within neutrophils and throughout the body. Increased stool calprotectin levels are an indication of neutrophil migration into gastrointestinal tissue due to inflammatory processes. Although there was not a significant difference with food timing or alcohol treatment, the levels of calprotectin increased with food timing effects (Supplemental Figure S5).

All experiments were approved by the Rush University Institutional Animal Care and Use Committee (IACUC). Mice with a Per2 (mPer2Luciferase [mPer2Luc]) knock-in mutation, obtained from Jackson Labs (Bar Harbor, ME), were bred in-house.³⁹ A total of 89 mice (51 male and 38 female) were used for experiments, with 12–13 males and 9–10 females. All mice were at least 8-weeks old when the experiments began. The animals were maintained on a 12-h light:12-h dark (12∶12 LD) cycle with lights on at 7 AM and lights off at 7 PM Central Standard Time.²² Animals were individually housed in ventilated, light-tight cabinets with 40W GE “Cool White” fluorescent bulbs individually controlled by an electronic timer.⁴³

Mice were placed on a food-restricted paradigm. As mice are nocturnal and consume their food in the dark, food timing in this experiment is defined with either day- (rest-phase= “wrong-time”, 7 AM) or nighttime (active-phase= “right-time”, 7 PM). Mice received access to food for 12 h for 5 consecutive weekdays followed by a 2-day ad-libitum access to food to mimic human conditions on the weekend.⁹ Mice were fed Envigo 2018 standard rodent chow (Teklad, Madison, WI) and given either normal drinking water or alcohol in drinking water (described below). This paradigm lasted a total duration of 10 weeks (Supplemental Figure S7). The general condition and health of the mice were monitored by daily observation, weekly body weight, food intake, and liquid intake measurements.⁴⁷

The control groups received drinking water (H₂O). The mice in alcohol-fed groups were given 20% alcohol v/v treatment in drinking water.⁴⁸ The 20% alcohol was only provided during the mice’s food intake periods, so it was available ad-libitum during the weekends when the mice had unrestricted access to both food and liquid.⁴⁹ Rodents reportedly metabolize alcohol much more rapidly than humans, at around 5.5 times the human rate.⁵⁰ To achieve blood alcohol levels (BAL) in mice that are pharmacologically relevant and comparable to BAL in alcohol-consuming human (~1.0 g/L), a 20% alcohol dose in drinking water was required.⁵¹ To acclimate the mice to alcohol consumption, a 14-day ramp-up was used, starting at 3% on days 1–3, increasing to 5% on days 4–7, 10% on days 8–10, 15% on days 11–14, and then reaching the target treatment of 20% on day 15. The volume was replaced as needed, with a maximum fill of 30 mL, and daily intake of both water and alcohol was measured. After 10 weeks, spontaneously voided stool samples were collected with sterile forceps to assess the intestinal microbial communities, followed by a urine test for intestinal barrier integrity. Finally, at the end of the study, the mice were euthanized via guillotine. Then, trunk blood and tissue samples were collected (Supplemental Figure S8).

The mice were fasted overnight prior to the urine permeability test, which is performed the next morning.^22,30 A 200 µl solution containing lactulose (3.2 mg), sucrose (0.45 mg), sucralose (0.45 mg), and mannitol (0.9 mg) is administered via oral gavage, followed by a subcutaneous injection of 2 mL of 0.9% saline to promote urine production. The mice were then placed into a metabolic chamber with water available ad-libitum for 5 h, after which the urine produced was collected and the total volume was recorded. Urine samples were stored at −80°C until further experimental analysis. This 5-h timeframe allows for intestinal transit, enabling assessment of barrier integrity across the entire intestine, including the colon. Intestinal barrier integrity is determined by measuring urinary concentrations of the administered sugars using gas chromatography, which enables calculation of the percentage of the oral dose that was exerted in the urine over 5 h. Then, intestinal permeability is assessed at the end of the 10-week study period.³⁰

PEAR (paired-end read merger) (v0.9.11) software package (Dalhousie University, Halifax, Nova Scotia, Canada) was used to analyze merged raw sequences.⁵⁴ A DADA2 algorithm from the QIIME2 (v 2020.8.0) workflow was utilized to remove sequences shorter than 240 bases, with denoising, were then removed before the merged were processed.^55,56 The resulting amplicon sequence variants (ASVs) were used for all subsequent analyses. Each ASV was allotted a taxonomy using an unbiased Bayes classifier and the SILVA 138 99% database reference for OTUs.^57,58 In total, 4,229,677 sequencing clusters were generated, with an average depth of 44,059 sequences per sample (median = 41,906; min = 1; max = 113,563). Three reagent contaminant ASVs (Rhodococcus, Allorhizobium-Neorhizobium-Pararhizobium-Rhizobium, and Pseudomonas) were identified and removed using the decontam algorithm based on their prevalence in the reagent negative blank controls (n = 4).⁵⁹ Unassigned and host-associated taxa, such as mitochondrial ASVs, chloroplast, and eukaryote were also removed from the datasets prior to statistical analyses.⁶⁰ After analysis, the raw sequence data were deposited in the NCBI Sequence Read Archive under BioProject PRJNA881333.

At the end of the study, the mice were euthanized via guillotine. Blood was collected by trunk collection, then centrifuged at 2,000 RPM for 35 min at 22°C. The serum was collected and stored at −80°C until experimental analysis. Murine tissue samples were collected: taking sections of the liver and the large intestine. The large intestine was flushed clean with cold phosphate-buffered saline (PBS) at least three times to remove excess debris. The tissue segments were snap-frozen and stored at −80°C, or formalin-fixed and paraffin-embedded (FFPE), as previously described.^42,61

Rhythmic expression of the clock gene PER2 (PER2 luciferase, PER2:LUC) persists in these colonic tissues and organoids, allowing constant monitoring of the robustness of these oscillations.⁴⁰ Freshly collected mouse colonic tissue was immediately transferred onto Matrigel-coated 35 mm dishes to minimize tissue movement in the media-filled dishes. Floating colonic organoids were also assessed in these dishes separately. The Kronos Dio apparatus (AB-2500, Atto Co., Tokyo, Japan) quantified bioluminescence levels of the PER-2 luciferase protein (E1602, Promega) in luciferin-containing media. Measurements were taken for 1 min every 15 min, at a controlled temperature of 37°C and 5% CO2, for a minimum of 5 days.

The mouse intestinal epithelial organoid cultures from the colon were prepared using a previously described method.^42,62–64 Briefly, the colon was washed with ice-cold PBS lacking Ca++ and Mg++ (Thermo Fisher Scientific, #100-10-023) to remove any loose debris. The samples were cut longitudinally, and excess fat was gently scraped off. After two PBS washes, the samples were cut into small pieces and transferred into a 50 mL conical tube. The tissue was then washed five times with PBS, transferred to a separate 50 mL conical with 2.5 mm EDTA chelating buffer, and incubated on a shaker at 4°C for 1 h. The supernatant was removed, and the tissue pieces were washed in PBS, discarding the first fraction. Three mL of PBS was added, and the crypts were gently resuspended and filtered through a 70 μm cell strainer into a fresh conical tube. This washing and filtering was repeated three times. The crypt suspension was centrifuged at 300 g for 5 min after adding 10% FBS. The supernatant was discarded, and the cell pellet was resuspended in 15 mL of organoid media minus growth factors. Base organoid media consisted of: Advanced DMEM/F-12 (Thermo Fisher Scientific), Glutamax (Thermo Fisher Scientific), Penicillin/Streptomycin (Sigma Aldrich), HEPES (Sigma Aldrich), B27-supplement (Thermo Fisher Scientific), N2-supplement (Thermo Fisher Scientific), and N-Acetylcysteine (Sigma Aldrich).

The cell suspension was then centrifuged at 150 g for 3 min. Then, this washing/centrifugation step was repeated twice. After the final centrifugation, the crypts were counted, resuspended in cold Matrigel (60 μL per well), and plated on a warm 24-well culture plate. Then, the plate was incubated at 37°C for 30 min to allow the Matrigel to polymerize before adding 500 μL of organoid medium with the growth factors to each well. Growth factors for colonic organoids included: recombinant mouse EGF (Sigma Aldrich), recombinant mouse Noggin (Sigma Aldrich), recombinant human R spondin 1 (PeproTech), recombinant mouse WNT-3A (PeproTech), A-83-01 (Tocris Bioscience), SB202190 (Sigma Aldrich), Y-27632 (Sigma Aldrich), and nicotinamide (Sigma Aldrich). The organoids were incubated at 37°C/5% CO2. Organoids were observed daily to ensure proper organoid health and robustness. Then, the organoids were passaged at least three times (roughly 3 weeks from the initial preparation) before any experimentation was performed.

We generated apical out organoids from these initial organoids using a previously described method.⁶⁵ Organoid growth was assessed under the microscope daily. To ensure proper health and robustness, the organoids were passaged at least three times (approximately 3 weeks from the initial preparation) before any experimentation was performed.

The cecal content supernatant sample preparations were adapted from previously published methods.⁶⁶ The cecal contents were pooled together (n = 4) for each group, weighed on ice, and 500 μL of cold PBS was added 100 mg cecal content. The sample tubes were then vortexed for 5 min and centrifuged at 13,000 g for 15 min at 4°C. The resulting supernatant was filtered through a 0.22 μM membrane with a 1 ml syringe before the sample was stored at −80°C until experimental use.

To assess epithelial barrier integrity, apical-out organoids were treated for 3 h with either control media, 0.2% EtOH, or a 1:100 dilution of cecal content supernatant (RTE H₂O or WTE EtOH) with control media. The organoids were then pelleted, and washed in PBS (4°C, 5 min, 5,000 RPM) before being incubated for 30 min in a solution (2 mg/ml) of 4 kDa fluorescein isothiocyanate (FITC)-dextran (Sigma, FD4-1 G) in a 37° incubator.⁶⁵ Organoids are pelleted and washed (4°C, 5 min, 5,000 RPM) in a live cell imaging solution (Thermo Fisher, #A14291DJ). Then, the organoids were placed in 8-well chamber slides (Nunc, Rochester, NY) and subsequently imaged using a Zeiss LSM 700 confocal microscope. All images were taken at × 20 magnification with 20–25 enteroids per group. Then, all experiments were performed in triplicate. Data are reported as mean ± standard error of the mean (SEM).

Tissue samples were immediately frozen in liquid nitrogen and stored at −80°C until used for protein extraction and concentration prior to western blot analysis. To ensure organoid growth and robustness for experimentation, the organoid protein was collected at least 21 days (approximately 3 weeks) after colon isolation and organoid generation. The commercially available kit, NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific 78,833), was used to obtain cellular material from both colon tissue and organoid samples.

2-mercaptoethanol (Bio-Rad) in a laemmli sample buffer with was used to prep the samples. For 2 h each gel was electrophoresed at 100 V with 20 μg of protein per lane. A 4%/7.5% stacking acrylamide Tris gel was used. As previously described, the membrane transference, blocking, and antibody incubation of both primary and secondary were performed.^20,21 Autoradiography film (HyBlot CL, Denville Scientific, Metuchen, NJ) with chemiluminescent substrate (ECL, GE Healthcare) was then placed on the membrane, and protein imaging was attained. Image J software was used to examine optical density by densitometric analysis. The specific antibodies are detailed in Supplemental Table S3 (below).

To assess histologic changes within the colonic tissue, 4- to 5-µm thick sections of fixed colon tissue samples were stained with hematoxylin and eosin, as previously described.^61,69 A gastrointestinal pathologist blinded to the sample identities, scored the tissues across seven categories to measure colon inflammation, tissue injury, and repair: inflammatory cells (0–4), goblet cells (0–4), mucosa thickening (0–4), submucosa cell infiltration (0–4), destruction of architecture (0–4), ulcers (0–1), and crypt abscess (0–1).⁶¹

Lipopolysaccharide (LPS) is a component found in the outer membrane of Gram-negative bacteria. LPS-binding protein (LBP) is a type 1 acute-phase protein that binds to LPS, facilitating an immune response. LBP is a well-accepted marker of intestinal barrier integrity and endotoxemia. Serum collected at the time of euthanasia was stored at −80°C until experimental analysis to measure LBP levels using an LBP ELISA kit (Hycult Biotech, HK205), following the manufacturer’s instructions.

Serum levels of the cytokine interleukin-6 (IL-6) were measured using a Meso Scale V-PLEX Pro-inflammatory Mouse Kit (Cat. #K152QXG–1, Meso Scale Diagnostics, Rockville, MD). All samples were analyzed in duplicate on a QuickPlex SQ120 instrument (Meso Scale Diagnostics), following the manufacturer’s protocol.

At the end of the study, mouse fecal samples were collected and stored at −80°C until experimental analysis. The fecal samples were subjected to only a single freeze-thaw cycle to maintain stool integrity. The calprotectin levels were measured using an ELISA kit (Cat. #KR6936, Eagle Biosciences, Amherst, NH).

Blood samples (25 µl each) were removed from storage at −80°C and thawed at 4°C for 15 min. To each sample, a standard (0.1%, or 0.804 mg/ml of n-propanol [nPrOH]; 5 μL) was then added to each sample. Each sample was then transferred and mixed to a 2 mL GC vial by pipette, which was capped right away. The sealed samples were placed on the GC autosampler and allowed to equilibrate at room temperature (22–23°C) for a minimum of 15 min prior to injection. The 25 μl headspace vapor from each sample was injected into a Trace 1310 GC coupled to a Thermo ISQ-LT MS, with a split ratio of 10:1 using a gastight 100 μl syringe. The GC inlet was held at 250°C. A 30 m DB-WAX UI column (0.25 mm ID, and 0.25 μm film thickness) was used to achieve optimal separation. The oven temperature was kept at 40°C for 4 min, and then ramped at 40°C/min to 120°C and held for 1 min. Helium carrier gas flow was maintained at 1.2 mL/min. 250°C was used for both transfer line and ion source. The SIM mode was used to scan ions, at m/z 31 and 45 (EtOH), m/z 31 and 42 (nPrOH), and m/z 59 and 43 (iPrOH [IPA]).