Chronic corticosterone exposure disrupts bile acids homeostasis associated with circadian dysfunction in the liver of chickens

Abstract

Enterohepatic bile acid metabolism displays robust diurnal rhythmicity governed by the hepatic circadian clock, yet how chronic stress reshapes this rhythmic process in poultry remains poorly defined. We hypothesized that chronic corticosterone exposure disrupts bile acid homeostasis in association with hepatic circadian clock dysfunction. This study therefore investigated the diurnal alterations of hepatic bile acid metabolism and its linkage with core clock dysregulation under corticosterone-induced chronic stress. Seventy-two 63-day-old male yellow-feathered dwarf broilers were randomly assigned to vehicle (Con) or corticosterone-treated (Cort) groups. Cort group received daily subcutaneous injections of corticosterone (4 mg/kg) for 10 consecutive days. On day 11, tissues were sampled every 4 hours from 09:00 (ZT2) to 05:00 the following day (ZT22), with feed and water provided ad libitum. Separately, AML12 hepatocytes were cultured with vehicle, dexamethasone (DEX), or DEX plus RU486, and bile acid transport activity was assessed using a Transwell system. Results indicated that chronic corticosterone exposure disrupted the hepatic circadian clock in chickens, reducing BMAL1 and REV-ERBβ mesors to 0.75- and 0.65-fold of control levels, respectively (P < 0.05), suppressing REV-ERBβ amplitude to 0.27-fold (P < 0.0001), and inducing significant phase delays in multiple core clock genes. Corticosterone increased CYP8B1 mesor and amplitude by 3.04- and 6.42-fold, respectively (P < 0.05), and advanced the acrophases of FXR and SHP by 10.37 h and 9.49 h, respectively (P < 0.05). Hepatic (P < 0.01) and plasma (P < 0.05) total bile acid levels decreased at ZT18, whereas gallbladder bile acid levels increased at ZT22 (P < 0.01). Correlation analysis revealed corticosterone eliminated diurnal negative correlations between clock and bile acid genes, shifting to persistent positive associations. In vitro, dexamethasone upregulated MRP2 expression (1.39-fold, P < 0.05) and promoted its membrane translocation via the glucocorticoid receptor, enhancing apical bile acid efflux. These findings indicate that chronic stress disrupts bile acid homeostasis via hepatic clock dysregulation, while glucocorticoids enhance MRP2-mediated efflux, linking stress to enterohepatic bile acid remodeling.

Introduction

The circadian rhythm is an endogenous biological timing system that enables organisms to synchronize with photoperiodic cycles, thereby optimizing energy utilization and coordinating essential physiological processes - including metabolism and behavior - with the Earth's rotation (Patke et al., 2020). This rhythm system comprises a central pacemaker located in the hypothalamic suprachiasmatic nucleus (SCN) governing peripheral clocks present across diverse tissues.

At the molecular level, peripheral circadian clocks are sustained by interlocked transcriptional-translational feedback loops. In the core regulatory loop, brain and muscle ARNT-like 1 (BMAL1), the indispensable positive regulator of the circadian oscillator, forms heterodimers with circadian locomotor output cycles kaput (CLOCK) or its functional homolog neuronal PAS domain protein 2 (NPAS2). This heterodimer complex binds to E-box elements to activate transcription of negative regulators period circadian regulator (PER) and cryptochrome (CRY) family members. Accumulated PER and CRY proteins translocate into the nucleus to repress BMAL1-driven transcription, generating a self-sustaining ∼24-hour oscillation. An auxiliary loop comprising retinoic acid receptor-related orphan receptor (ROR) and reverse erythroblastosis virus beta (REV-ERB) family members fine-tunes clock stability by competitively modulating BMAL1 expression. This molecular network drives rhythmic expression of thousands of clock-controlled genes, coupling the circadian timer to core physiological processes including cellular metabolism and tissue homeostasis (Dibner, et al., 2010).

Accumulating evidence indicates that hepatic peripheral clocks play a pivotal role in maintaining metabolic homeostasis through the temporal regulation of glucose and lipid metabolism (Mukherji, et al., 2019). As a critical component of lipid digestion and metabolic signaling, bile acids, amphipathic steroid acids whose synthesis and diversity depend on both host and microbial metabolism, are deeply integrated into this temporal framework (Perino and Schoonjans, 2022). Following hepatic de novo synthesis, bile acids are conjugated with taurine or glycine and secreted into the canaliculus via the bile salt export pump (BSEP) and multidrug resistance-associated proteins (MRPs). Postprandially, bile acids are released into the intestinal lumen to emulsify dietary lipids and cholesterol and facilitate their absorption. Approximately 95% are then actively reabsorbed in the ileum and returned to the liver via the portal circulation, completing the enterohepatic cycle. Early studies demonstrated rhythmic bile secretion and enterohepatic circulation in rodents, and subsequent work confirmed diurnal fluctuations in circulating bile acids in humans (Balabaud, et al., 1975; Ho, 1976; Segers and Depoortere, 2021). Moreover, circadian timing interacts with dietary cues to coordinate bile acid homeostasis (Eggink, et al., 2017). Mechanistically, the hepatic circadian clock regulates bile acid homeostasis through CLOCK/BMAL1-controlled CYP7A1 expression, RORα-dependent CYP8B1 expression, and rhythmic regulation of the bile acid transporter NTCP (Ma, et al., 2009; Noshiro, et al., 2007; Pathak, et al., 2013a).

Classical circadian disruption models have further validated this causal link. In rodents, restricted feeding, chronic jet lag, and genetic ablation of core clock genes (e.g., BMAL1 or PER1/PER2 double knockout) all profoundly alter hepatic bile acid synthesis, transport kinetics, and enterohepatic homeostasis, leading to hepatic bile acid pool depletion and aberrant gallbladder storage (Cui, et al., 2022; Kettner, et al., 2016; Ma, et al., 2009). These findings collectively confirm that an intact hepatic clock is required for normal bile acid rhythmicity. Given the rapid growth and high metabolic demands of modern broilers, maintaining bile acid homeostasis and its temporal coordination is essential for efficient lipid utilization and metabolic health under intensive production systems. However, circadian research in chickens remains limited in scope: existing studies have predominantly focused on behavioral rhythms (Bessei, et al., 2023), reproduction using photoperiod manipulation (Geng, et al., 2022), and hepatic immune and metabolic rhythms by physiological time-course sampling (Li, et al., 2025). While our prior work demonstrated that chronic corticosterone exposure disrupts hepatic and intestinal bile acid metabolism in chickens (Wu, et al., 2023), whether chronic stress-induced glucocorticoids (GC) disrupt hepatic bile acid homeostasis through circadian clock dysfunction in chickens remains unclear.

In modern intensive farming, chickens are subjected to chronic stress from high temperatures, persistent humidity, sudden noise, and physical inactivity (Shields and Greger, 2013). These stressors activate the hypothalamic-pituitary-adrenal (HPA) axis, substantially increasing GC production (Miller and Auchus, 2011). Furthermore, GC serve as established hormonal cues that entrain peripheral oscillators (Shimba and Ikuta, 2020), and GC response elements participate in the transcriptional regulation of clock- and bile acid–related genes (Mitropoulos and Balasubramaniam, 1976; So, et al., 2009; Yamamoto, et al., 2005). In the present study, we established a corticosterone-induced chronic stress model in broiler chickens to test the hypothesis that chronic GC exposure disrupts hepatic circadian clock function, which in turn remodels the diurnal rhythm of bile acid metabolism and impairs enterohepatic homeostasis. We further validated the GC receptor-dependent regulatory mechanism of bile acid efflux transporters in vitro. Our findings provide novel insights into stress-induced metabolic disruption in poultry from a chronobiological perspective.

Materials and methods

Ethics statement

All procedures were carried out following the guidelines established by the Animal Ethics Committee of Nanjing Agricultural University (2012CB124703). Slaughter and sampling were conducted under the “Guidelines on Ethical Treatment of Experimental Animals” (2006) No.398 set by the Ministry of Science and Technology, China.

Animals and experimental design

Seventy-two 63-day-old male medium-growing yellow-feathered dwarf broilers were obtained from Jiangsu Lihua Animal Husbandry Co., Ltd. (Changzhou, China). Chickens were housed in standard wire cages (60 cm × 60 cm × 45 cm, L × W × H, housing 3 birds per cage) within an environmentally controlled room. The ambient temperature was maintained at 20–22°C, and relative humidity was kept at 50–60%. Throughout the experimental period, all birds had ad libitum access to fresh water through nipple drinkers and a commercial corn-soybean meal basal diet specifically formulated for medium-growing yellow-feathered broilers. The chickens were housed under a 12-h light:12-h dark cycle, with lights on from 07:00 to 19:00 daily. Following a 3-day acclimation period, chickens were randomly divided into the vehicle (Con) and corticosterone (Cort) group.

Corticosterone (Shanghai Aladdin Biochemical Technology Co., Ltd., China) was dissolved in absolute ethanol and diluted with sterile physiological saline immediately before administration. Chickens in the Cort group received subcutaneous injections into the neck at a total daily dose of 4 mg/kg body weight, administered twice daily (2 mg/kg per injection) at 09:00–10:00 and 18:00–19:00 for 10 consecutive days. The final corticosterone solution and vehicle both contained 15% ethanol. Chickens in the Con group received an equivalent volume of the vehicle via the same route and at the same time points. The dosage was chosen in accordance with earlier reports (Hu, et al., 2018; Shan, et al., 2024).

The experiment was conducted in Nanjing during winter, with sunrise at approximately 07:00 (designated as zeitgeber time 0, ZT0), under a 12:12-h light-dark cycle. After the 10-day injection treatment, on the following day (Day 11), six randomly selected birds per group were sampled at each of the time points ZT2, ZT6, ZT10, ZT14, ZT18, and ZT22. The blood was collected from the pterygoid venous plexus. The gallbladder was removed from liver and bile was collected. Subsequently, liver tissues were rapidly frozen in liquid nitrogen and stored at −80°C until further analysis.

Total bile acids analysis

Total bile acids (TBA) were extracted from liver tissues following a previously reported protocol (Hua, et al., 2018). Approximately 100 mg of liver tissue was homogenized with 95% ethyl alcohol (EtOH) and incubated at 60°C overnight. After centrifugation (8,000 rpm, 10 min), supernatants were collected. The pellets underwent two additional extractions with 80% ethanol, and pooled supernatants were retained for analysis. Gallbladder bile was diluted 4,000-fold. TBA concentrations in liver extracts, bile, and plasma were determined using an enzymatic cycling assay with a commercial kit (Ningbo MedicalSystem Biotechnology Co., Ltd., China) on a Hitachi 7020 automated biochemical analyzer (Hitachi 7020, Hitachi, Japan), following the manufacturer's instructions.

RNA extraction and cDNA synthesis

Total RNA was extracted from liver tissues and AML12 cells using TsingZol reagent (TSP401, TSINGKE, China) according to the manufacturer's instructions. Briefly, approximately 40 mg of liver tissue was homogenized in 1 mL of TsingZol reagent using a tissue homogenizer. For AML12 cells cultured in 6-well plates, 0.5 mL of TsingZol reagent was added to each well. The lysates were then transferred to microcentrifuge tubes, and phase separation was achieved by adding chloroform. Following centrifugation, the upper aqueous phase containing the RNA was collected, precipitated with isopropanol, washed with 75% ethanol, and finally dissolved in RNase-free water. RNA concentration and purity were assessed using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, USA), and values between 1.8 and 2.2 for the A260/A280 ratio and above 1.8 for the A260/A230 ratio were considered acceptable.

Subsequently, 1 μg of total RNA was reverse-transcribed into cDNA using the SynScript™ III cDNA Synthesis Mix (TSK322S, TSINGKE). Prior to reverse transcription, the RNA templates were treated with the provided gDNA Remover to eliminate residual genomic DNA. The first-strand cDNA was then synthesized by adding the SynScript III RT Enzyme Mix and RT Reaction Mix according to the manufacturer's protocol.

Primer design and quantitative real-time PCR

All specific primers (listed in Table 1) were designed using NCBI Primer-BLAST to span exon-exon junctions to prevent genomic DNA amplification and were synthesized by TSINGKE Co. (Nanjing, China). Quantitative real-time PCR was performed on an Mx3000P instrument (Stratagene, USA) using 2 × Universal SYBR Green Fast qPCR Mix (RK21203, ABclonal, China) with gene-specific primers and diluted cDNA templates. The amplification conditions were as follows: initial denaturation at 95°C for 3 min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. Melting curve analysis was performed to confirm the specificity of the PCR products.

Table 1.

The primers used for real-time quantitative PCR.

Gene	Primer sequence 5′ →3′	Product (bp)
CLOCK NM_001289834.2	F: TTTGATACCAGCCCCACACC	278
	R: GCTGAAAACTGGGACATGCTG
BMAL1 NM_001001463.3	F: CAGGAGGAACAAGAGCTGGG	164
	R: CTTCTTGCTTCCAGGGGAGG
NPAS2 NM_001030542.2	F: AAGAGGGCTTCGCGTAACAA	203
	R: GGCTTCCAGTCCTGCTGAAT
CRY1 XM_042889900.1	F: GTATGGCCTGGGGGAGAAAC	155
	R: AAACAGCCAAAGCGGAGGTA
CRY2 NM_204244.2	F: CAGCACCAGTACAGCAGTCAGA	106
	R: GGCTCGCTTGTACAGTTCTTC
PER2 NM_204262.2	F: CCGCTTCCTGTTTTTCCAGC	291
	R: AGGGGTAGTGAATGGAGGCT
PER3 XM_004947317.4	F: AGTCCGAACCCCTCTGAATGA	272
	R: GGTTGACATCTGCACACACC
REV-ERBα XM_040653346.1	F: AGGGCTGCAAGGGATTCTTC	232
	R: CCCACACTCATGGCGCTC
REV-ERBβ NM_205205.3	F: GTGGTGCATGTCTGCTTGTTA	148
	R: AGGATGACTGGAAACCGCTG
RORα XM_044980170.1	F: TAGAAACCGCTGCCAACACT	279
	R: GCTTTGCTACCTTCAGGGGT
CYP7A1 NM_001001753.2	F: GTAACGCCCTAGATGCCCTC	220
	R: GCTCTCTCTGTTTCCCGCTT
CYP27A1 XM_040676620.1	F: TATCCCCAAGATGCCGATGC	125
	R: TGGGGAAGAGGTAGTCTCCG
CYP8B1 NM_001005571.1	F: CGTTCTGGGAACCTTTCGGA	94
	R: CTGTCTGAAGTCCAGTGCGT
CYP7B1 XM_025147742.2	F: TAATGCCCTTTGGCTCTGGG	163
	R: AGGAGGATACCAAGGCCCAT
BAAT XM_040701449.1	F: CCTGGTGAAGCGTGATGTGA	262
	R: AGCCCGGAACTCAATGAGAC
BSEP XM_040676679.1	F: GCTGTGGTAAGAGCACCAGT	167
	R: AGCGATGCTGCAGTCGAATA
NTCP XM_040671950.1	F: TCAAGGGAGCCTCAAAGAGC	91
	R: CTGAGTGCAAATGGTGGTGC
OATP1B1 XM_040662375.2	F: GGGCTTCCTGCTTGGATCTT	106
	R: GGCACCAACCCATCGAGTAT
OSTα NM_001277697.1	F: CGTTCCATGATGGTGGTGGA	134
	R: ACCATGGGCACGTCCTTTAG
OSTβ XM_025153900.2	F: AGGAGCTGCTGTGGTTCTTC	90
	R: AACCCAAAATTGCGACCACG
MRP2 XM_015288821.3	F: CGGCTCCGTCTTTTGGAATG	138
	R: CATATGGGAAGGAGCTGCCA
PPIA NM_001166326.2	F: CTTCGAGCTCTTCGCTGACA	88
	R: GCCCTTGTAGCCAAATCCCT
β-actin NM_205518.2	F: GTGGATCAGCAAGCAGGAGT	182
	R: ATCCTGAGTCAAGCGCCAAA

Peptidylprolyl isomerase A (PPIA, for liver tissues) and β-actin (for AML12 cells) served as reference genes. Relative gene expression levels were quantified using the comparative Ct (2^-ΔΔCt) method, with values normalized to the control group at ZT2 (Con-ZT2), which was arbitrarily set to 1.0.

Cell culture and treatment

Cells were cultured at 37°C under 5% CO₂ in DMEM/F12 (Gibco) containing 10% fetal bovine serum (FBS, Ausgenex), 1% insulin-transferrin-selenium (ITS, Wako), and 1% penicillin-streptomycin (Gibco). Cytotoxicity screening established 500 nM DEX and 10 μM taurocholic acid as non-toxic over 36 h (Supplementary Fig. S1). Experimental groups comprised vehicle control, DEX (500 nM), RU486 (1 μM mifepristone; glucocorticoid receptor [GR] antagonist; MCE), and DEX+RU486 (500 nM DEX co-administered with 1 μM mifepristone).

Taurocholic acid (TCA) transport assay

Using the polarized architecture of AML12 cells (Decaens, et al., 2008; Wu, et al., 1994), directional transport of 10 μM TCA (B20918, Yuanye, China) was assessed in a Transwell system (0.4-μm inserts, JET BIOFIL; 500 μL apical / 1,500 μL basolateral medium), with the upper and lower chambers representing the apical (bile) and basolateral (blood) sides, respectively. Monolayer integrity was verified by transepithelial electrical resistance (TEER) (> 300 Ω·cm²) and permeability testing (Supplementary Fig. S2). At confluence, cells were exposed to vehicle, DEX (500 nM), or DEX+RU486 (1 μM) in the presence of 10 μM TCA for 36 h. TCA levels in the apical, basolateral, and intracellular compartments were subsequently determined.

Taurocholic acid (TCA) determination by high performance liquid chromatography (HPLC)

Following sample collection, culture media from the apical and basolateral chambers of Transwell inserts were collected separately. Cells were lysed by ultrasonication, and an aliquot was reserved for protein quantification using a bicinchoninic acid (BCA) assay. To extract TCA accurately, cell lysates and culture media were subjected to solid-phase extraction (SPE) using C18 reversed-phase silica cartridges (L100220, Dibai, China). The cartridges were sequentially preconditioned with methanol and distilled water. Samples were loaded onto the cartridges and allowed to pass through by gravity. After washing with distilled water, the retained TCA was eluted with methanol. The eluate was collected by centrifugation, filtered, and transferred to autosampler vials.

Chromatographic analysis was performed using an HPLC system (Acquity H-class, Waters, USA) equipped with a C18 reversed-phase column at ambient temperature. The mobile phase consisted of 70% acetonitrile and 30% phosphate buffer (15 mM KH₂PO₄/15 mM K₂HPO₄, pH 3.0 adjusted with phosphoric acid). The flow rate was maintained at 1.0 mL/min, and the injection volume was 2 μL. TCA was detected at 200 nm absorbance. Quantification was performed using a standard curve constructed with TCA standards, and the intracellular TCA content was normalized to the total protein concentration of the corresponding cell lysate.

Protein extraction

Total protein from liver tissues and AML12 cells was extracted using RIPA lysis buffer (P0013E, Beyotime, China), with 1 mL per 50 mg liver tissue or 150 μL per well of a six-well plate. After lysis and centrifugation, the supernatants were collected. Because MRP2-mediated bile acid transport requires proper membrane localization, we isolated membrane and cytoplasmic proteins from AML12 cells using a commercial kit (P0033, Beyotime) according to the manufacturer’s instructions. Briefly, the cell pellet (about 4 × 10⁷ cells) was resuspended in reagent A (1 mL), homogenized, centrifuged and the supernatant was further centrifuged to obtain the cytoplasmic fraction. The resulting pellet was resuspended in reagent B, vortexed, and centrifuged to collect the membrane fraction. Protein concentrations were determined via bicinchoninic acid assay kit (P0012, Beyotime). Protein samples were mixed with loading buffer before denaturation. Liver total-protein, whole-cell, and cytoplasmic samples were denatured at 95°C for 5 min, whereas membrane protein samples were denatured at 37°C for 30 min.

Western blotting

Equal amounts of total protein (40 μg for liver tissues and 20 μg for AML12 cells) were loaded per well and separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked with 5% non-fat milk in TBST for 2 h at room temperature and subsequently incubated overnight at 4°C with the following primary antibodies: PER1 (HA500097, HuaAn, China), PER2 (A13168, ABClonal, China), PER3 (ab177482, Abcam, USA), BSEP (A8467, ABClonal), and MRP2 (DF3873, Affinity, China). Following TBST washes, membranes were incubated with HRP-conjugated secondary antibodies. Protein bands were visualized using an enhanced chemiluminescence (ECL) kit and captured with a chemiluminescence imaging system (Tanon 3500, Tanon, China).

Due to corticosterone-induced alterations in conventional reference proteins, target protein signals were normalized to total protein. Gels were stained with BeyoBlue™ Plus Coomassie Solution (P0003S, Beyotime) and imaged using a Tanon 3500 system. Target protein intensity was normalized to the total lane density quantified by ImageJ. A prominent ∼45 kDa band is displayed as a visual loading indicator but was not used for quantification.

Immunofluorescence microscopy

Following 36-h treatment, cells were washed with PBS, fixed with 4% paraformaldehyde for 5 min, and permeabilized with 0.3% Triton X-100 in tris-buffered saline (TBST) for 30 min at room temperature. Following PBS washes, cells were blocked with 5% bovine serum albumin (BSA) for 10 min and incubated overnight at 4°C with primary antibodies against MRP2 (DF3873, Affinity, China) and E-cadherin (sc-8426, Santa Cruz, USA). After washing, cells were incubated with fluorescent-conjugated secondary antibodies for 30 min at 37°C in the dark. After repeated washes, nuclei were stained with DAPI for 5 min in the dark. Finally, samples were imaged via fluorescence microscopy (Axio Vert A1, Zeiss).

Statistical analysis

Data are presented as mean ± standard error of the mean (SEM) and analyzed using GraphPad Prism 8.0 and SPSS 20.0. Differences between two groups in animal experiments were analyzed using Student’s t-test, whereas in cell experiments, multiple-group comparisons were evaluated using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. Circadian rhythmicity of gene expression was assessed by Cosinor analysis, fitting data to a 24-h sinusoidal function: f(t) = M + A·cos(t·π/12 + B·π/12), where M = mesor, |A| = amplitude, and acrophase derived from parameter B (Liu, et al., 2019). To facilitate Cosinor fitting for graphical visualization only, data obtained at ZT2 were additionally designated as ZT26, thereby generating a continuous 24-h cycle with explicit start and end points. To statistically compare circadian parameters between the two groups, a Population-Mean Cosinor approach using a joint linearized regression model was applied. Differences in Mesor were evaluated via linear interaction terms, while the standard errors for Amplitude and Acrophase differences were derived using the Delta method, followed by Wald tests to determine statistical significance. The correlation analysis was performed by using the R package of circos plot and ggplot2.

Results

The effects of chronic corticosterone exposure on the expression of hepatic circadian clock gene expression

In the control group, 12 core hepatic clock genes (Fig. 1) exhibited clear 24- h rhythmicity. However, chronic corticosterone exposure markedly remodeled their temporal expression profiles. Compared with controls, the mesor of BMAL1, CRY1, REV-ERBβ, and RORα was significantly reduced to 0.75-, 0.69-, 0.65-, and 0.74-fold of control values, respectively (P < 0.05). In contrast, only CRY2 showed an upregulation, with its mesor increasing to 1.48-fold of controls (P < 0.001). At the amplitude level, REV-ERBβ exhibited the greatest reduction, falling to 0.27-fold of controls (P < 0.0001). With respect to circadian phase, corticosterone treatment induced phase delays in several core clock genes: the acrophases of BMAL1, NPAS2, CRY1, PER3, and REV-ERBβ were significantly delayed by 5.30 h, 4.72 h, 5.42 h, 5.28 h, and 5.04 h, respectively (P < 0.05).

Fig. 1.

The effects of chronic corticosterone exposure on circadian expression of clock genes in liver of chickens. (A) the circadian rhythm of CLOCK; (B) the circadian rhythm of BMAL1; (C) the circadian rhythm of NPAS2; (D) the circadian rhythm of CRY1; (E) the circadian rhythm of CRY2; (F) the circadian rhythm of PER2; (G) the circadian rhythm of PER3; (H) the circadian rhythm of REV-ERBα; (I) the circadian rhythm of REV-ERBβ; (J) the circadian rhythm of RORα; (K) the circadian rhythm of E4BP4; (L) the circadian rhythm of TEF. Relative mRNA expression levels are presented as fold changes relative to the Con group at ZT2 (defined as 1). Black dots and lines represent the vehicle control (Con) group, while red dots and lines represent the corticosterone (Cort) group. Rhythmicity was assessed using a 24-h Cosinor model. The solid curves represent the fitted Cosinor regression, and the horizontal dashed lines indicate the mesor (rhythm-adjusted mean). Data were presented as means ± SEM (n = 6), zeitgeber time (ZT), *P < 0.05, **P < 0.01. The embedded tables beneath each graph present the detailed circadian parameters (Mesor, Amplitude, Acrophase), goodness-of-fit (R²), and the statistical significance (P-values) comparing the two groups, as determined by Cosinor analysis.

Compared with the Con group, during the early light phase (ZT2), the Cort group was characterized by significantly decreased expression of PER2 (0.29-fold, P < 0.001), PER3 (0.24-fold, P < 0.01), CRY1 (0.51-fold, P < 0.01), REV-ERBβ (0.41-fold, P < 0.05), and RORα (0.51-fold, P < 0.01), alongside paradoxical elevations of NPAS2 (3.02-fold, P < 0.05) and CRY2 (1.38-fold, P < 0.01). As the diurnal cycle progressed to midday and late afternoon (ZT6 and ZT10), BMAL1 (0.37-fold at ZT6, P < 0.05; 0.27-fold at ZT10, P < 0.01) and CRY1 (0.44-fold at ZT6, P < 0.001; 0.49-fold at ZT10, P < 0.01) were significantly downregulated at both time points, while NPAS2 exhibited a sharp decline at ZT10 (0.28-fold, P < 0.001). Conversely, CRY2 remained significantly elevated (1.54-fold at ZT6, P < 0.05; 1.59-fold at ZT10, P < 0.01), PER3 (2.97-fold, P < 0.05), REV-ERBβ (2.03-fold, P < 0.01), and TEF (1.53-fold, P < 0.05) were significantly upregulated at ZT10. Early in the dark phase (ZT14), CRY2 remained elevated (1.93-fold, P < 0.05) while NPAS2 was notably suppressed (0.53-fold, P < 0.05). By mid-dark (ZT18), PER2 (0.50-fold, P < 0.05), PER3 (0.57-fold, P < 0.05), and REV-ERBβ (0.45-fold, P < 0.01) all exhibiting significant downregulation. Finally, in the late dark phase preceding dawn (ZT22), the system demonstrated a marked upregulation of BMAL1 (2.68-fold, P < 0.01) and NPAS2 (6.61-fold, P < 0.05), despite continued suppression of PER2, PER3, and REV-ERBβ (P < 0.01).

The effects of chronic corticosterone exposure on the circadian rhythm of the bile acid pool

Hepatic TBA levels in control chickens gradually increased from ZT6 to ZT18 and then declined at ZT22, whereas corticosterone treatment decreased hepatic TBA levels, with significant reductions observed at ZT14 (P < 0.05) and ZT18 (P < 0.01) (Fig. 2A). Gallbladder bile TBA showed an increasing trend from ZT14 to ZT22 after corticosterone treatment, and was significantly elevated at ZT22 compared with controls (P < 0.01, Fig. 2B). In control chickens, plasma TBA concentrations increased from ZT6 to ZT14, followed by a gradual decline toward ZT22. In contrast, corticosterone treatment resulted in a continuous reduction in plasma TBA levels from ZT2 to ZT18, with significantly lower levels observed at ZT18 compared with controls (P < 0.05, Fig. 2C). The liver/plasma and liver/gallbladder TBA ratios exhibited altered oscillatory patterns after corticosterone exposure, although no significant differences were detected (Fig. 2D and E).

Fig. 2.

The effects of chronic corticosterone exposure on the circadian rhythm of the bile acid pool in chickens. (A) The concentrations of total bile acids (TBA) in liver. (B) The concentration of TBA in gallbladder. (C) The concentrations of plasma TBA. (D) Liver/plasma TBA ratio (calculated based on the total pool size in µmol). (E) Liver/gallbladder TBA ratio (calculated based on the total pool size in µmol). Con: Control group (black); Cort: Corticosterone treatment group (red); X-axis shows the timescale in zeitgeber time (ZT) over the 24 h sampling period. Con = vehicle control; Cort = corticosterone; Data were presented as means ± SEM (n = 6), *P < 0.05, **P < 0.01.

The effects of chronic corticosterone exposure on circadian expression of genes governing bile acid synthesis, transport and regulation

In the control group, the expression of genes involved in hepatic bile acid synthesis and transport exhibited typical sinusoidal circadian rhythms, with peak and trough phases aligned with the light-dark cycle, while chronic corticosterone exposure significantly disrupted these rhythmic patterns. As is shown in Fig. 3, the mesor of BAAT expression decreased (0.88-fold of the control, P < 0.05), whereas the mesors of CYP8B1, NTCP, OSTβ, and OATP1B1 were significantly elevated by 3.04-, 1.75-, 1.64-, and 1.59-fold respectively (P < 0.05). Furthermore, Cort exposure amplified the diurnal amplitude of CYP8B1 and OATP1B1 by 6.42- and 1.93-fold (P < 0.05). Regarding the circadian phase, on one hand, CYP27A1, CYP7A1, FXR, and SHP underwent a phase advance, with their acrophases significantly shifted forward by 6.15 h, 9.07 h, 10.37 h, and 9.49 h respectively (P < 0.05). Conversely, CYP7B1 displayed a phase delay, with its acrophases significantly postponed by 5.50 h (P < 0.05).

Fig. 3.

The effects of chronic corticosterone exposure on circadian expression of bile acid synthesis and transport genes in liver of chickens. (A) the circadian rhythm of CYP7A1 gene; (B) the circadian rhythm of CYP27A1; (C) the circadian rhythm of CYP8B1; (D) the circadian rhythm of CYP7B1; (E) the circadian rhythm of BAAT. (F) the circadian rhythm of FXR; (G) the circadian rhythm of SHP. (H) the circadian rhythm of BSEP; (I) the circadian rhythm of MRP2; (J) the circadian rhythm of MRP3; (K) the circadian rhythm of MRP4; (L) the circadian rhythm of OSTα; (M) the circadian rhythm of OSTβ. (N) the circadian rhythm of NTCP; (O) the circadian rhythm of OATP1B1. Relative mRNA expression levels are presented as fold changes relative to the Con group at ZT2 (defined as 1). Black dots and lines represent the vehicle control (Con) group, while red dots and lines represent the corticosterone (Cort) group. Rhythmicity was assessed using a 24-h Cosinor model. The solid curves represent the fitted Cosinor regression, and the horizontal dashed lines indicate the mesor (rhythm-adjusted mean). Data were presented as means ± SEM (n = 6), zeitgeber time (ZT), *P < 0.05, **P < 0.01. The embedded tables beneath each graph present the detailed circadian parameters (Mesor, Amplitude, Acrophase), goodness-of-fit (R²), and the statistical significance (P-values) comparing the two groups, as determined by Cosinor analysis.

During the early light phase (ZT2), in Cort group, we observed significantly decreased expression of CYP7A1 (0.38-fold), BAAT (0.67-fold), and FXR (0.59-fold) (P < 0.05), along with elevated expression of CYP8B1 (2.67-fold, P < 0.05) and OSTβ (2.20-fold, P < 0.01). At midday and late afternoon (ZT6 and ZT10), the expression levels of NTCP (2.38-fold at ZT6, P < 0.01; 1.83-fold at ZT10, P < 0.05), OATP1B1 (3.47-fold at ZT6, P < 0.01; 6.80-fold at ZT10, P < 0.05), SHP (1.97-fold at ZT6, P < 0.05), and CYP8B1 (6.88-fold at ZT10, P < 0.05) were significantly upregulated. Early in the dark phase (ZT14), NTCP remained significantly elevated (1.69-fold, P < 0.05) while CYP7B1 expression was notably suppressed (0.78-fold, P < 0.05). By mid-dark (ZT18), MRP2 exhibited a significant downregulation (0.51-fold, P < 0.05). Finally, in the late dark phase preceding dawn (ZT22), CYP8B1 and NTCP were recurrently upregulated, by 4.20-fold (P < 0.05) and 2.39-fold (P < 0.01), respectively.

Correlations between bile acid homeostasis and hepatic metabolic and clock gene expression

In the Con group, correlation patterns exhibited pronounced circadian rhythmicity. Genes involved in bile acid synthesis and transport displayed negative associations with clock genes during the daytime (ZT2–ZT10), which transitioned to positive correlations during the nighttime (ZT14–ZT22). By contrast, corticosterone treatment attenuated this diurnal variation, resulting in marked positive correlations between bile acid synthetic and transport genes and clock genes throughout the circadian cycle (Fig. 4).

Fig. 4.

Temporal correlation networks between circadian clock genes and bile acid metabolism genes in both groups. Node size is proportional to expression level, while orange and gray lines indicate positive and negative correlations, with thickness denoting strength.

The effects of DEX treatment on the expression of genes involved in bile acid metabolism and circadian clock regulation in AML12 cells

In AML12 cells, DEX significantly upregulated PER1, BSEP, and OSTβ mRNA (1.21-, 5.17-, 1.48-fold, P < 0.05; Fig. 5A and B) and increased PER3 and MRP2 protein (3.32-, 1.39-fold, P < 0.05; Fig. 5C and D). To assess whether the DEX-induced increase in MRP2 mRNA and total protein translates into enhanced functional capacity, we next examined its subcellular distribution by isolating membrane and cytoplasmic fractions. Immunofluorescence and membrane fraction Western blotting confirmed that DEX promoted MRP2 translocation to the plasma membrane (Fig. 5E and F); these effects were fully reversed by co-administration of the glucocorticoid receptor antagonist RU486. However, no corresponding increase in BSEP protein abundance was observed.

Fig. 5.

The effects of DEX exposure on bile acid transport in AML12 cells. Cells were treated with vehicle (Con), 500 nM dexamethasone (DEX), 1 μM RU486, or a combination of both (RU486+DEX) for 36 h. (A) Relative mRNA expression level of clock genes; (B) Relative mRNA expression level of genes related to bile acid metabolism; (C) Relative protein expression level of clock genes; (D) Relative protein expression level of bile acid related genes; (E) Immunofluorescence microscopy analysis of MRP2 subcellular localization, with DAPI (blue), E-cadherin (red), and MRP2 (green); (F) Cytomembrane and cytoplasmic MRP2 protein expression; (G) Schematic illustration of the Transwell transport assay (left), and taurocholic acid (TCA) distribution profiles across the apical chamber, basolateral chamber, and intracellular compartments (right). For all Western blot analyses (C, D, and F), protein expression was normalized to total protein using Coomassie Brilliant Blue staining, with a prominent∼45 kDa band shown as a visual loading indicator. Values are means ± SEM (n = 6). Different lowercase letters indicate significant differences (P < 0.05).

The effects of DEX treatment on the capacity of bile acid efflux in AML12 cells

To determine whether elevated MRP2 expression translates into enhanced bile acid efflux, directional TCA transport was evaluated in AML12 cells using a Transwell system (Fig. 5G). Following basolateral loading with 10 μM TCA for 36 h, DEX treatment significantly increased the apical TCA fraction to 61.42% of the total, compared with 37.03% in controls. This effect was attenuated to 43.52% upon RU486 co-treatment. Concomitantly, the basolateral TCA fraction decreased to 30.88% in DEX-treated cells, relative to 55.19% in controls, and was restored to 49.23% in the presence of RU486. Intracellular TCA levels remained comparable across treatment groups.

Discussion

Stress responses and circadian rhythms are two fundamental regulatory systems that mediate environmental adaptation and synchronize internal physiology with daily cycles to maintain metabolic homeostasis, respectively (Nader, et al., 2010). Previous studies have established that environmental and psychological stressors can significantly perturb peripheral clock machinery. For example, social defeat stress delays the expression peak of the hepatic clock gene PER2 in mice (Ota, et al., 2020), and constant light exposure disrupts the circadian rhythm of BMAL1 in chick liver (Honda, et al., 2017; Zeman, et al., 2009). Consistent with these observations, our data demonstrated that chronic corticosterone exposure profoundly dampened the rhythmic amplitude and delayed the phase of multiple core clock genes in the broiler liver, indicating a desynchronization of the hepatic circadian clock. Furthermore, an intact circadian clock is essential for maintaining bile acid homeostasis, and its disruption directly impairs this metabolic balance. For instance, chronic environmental stressors have been documented to severely perturb peripheral clock machinery—particularly in the liver—and subsequent downstream metabolic rhythms in rodent models (Takahashi, et al., 2013). Taken together, these findings strongly suggest that chronic corticosterone exposure impairs hepatic peripheral clock organization, which is closely linked to the profound disruption of the enterohepatic bile acid pool observed in our study.

In diurnal animals such as chickens, peripheral clocks and bile acid secretion are tightly synchronized with environmental light-dark cycles and diurnal feeding-fasting transitions. Under normal physiological conditions, hepatic bile acid synthesis and transport are upregulated during the active daytime feeding phase and decline at night, opposite to the patterns reported in nocturnal animals such as mice (Yang and Zhang, 2020). Against this physiological rhythmic background, chronic corticosterone exposure significantly disrupted bile acid distribution, notably decreasing the liver-to-gallbladder bile acid ratio from ZT10 to ZT18. We hypothesize that chronic corticosterone may promote excessive hepatic bile acid efflux, potentially leading to increased gallbladder storage at late night (ZT22). Upon lights-on and the first feeding event, these excessive bile acids are presumably released into the intestine, reabsorbed into circulation, and subsequently suppress hepatic bile acid synthesis via feedback mechanisms.

Glucocorticoids have been shown to suppress hepatic CYP7A1 while inducing CYP8B1, NTCP, BSEP, and MRP2, indicating a direct regulatory role in bile acid metabolism (Al-Aqil, et al., 2018; Johnson and Klaassen, 2002; Mörk, et al., 2016; Rose, et al., 2011). Our findings demonstrate that chronic stress profoundly reorganizes the circadian landscape of the hepatobiliary system. Network analysis reveals that chronic corticosterone exposure reshapes the sparse, circadian-gated negative-correlation hepatic network into a dense, stress-synchronized positive-correlation network. This remodeling is strongly associated with the observed decrease in hepatic bile acids and abnormal gallbladder accumulation during ZT14–ZT18. Specifically, the correlation between clock genes and efflux transporters (e.g., MRP2) shifts from negative to positive at ZT14, which might favor excessive nocturnal biliary discharge. Coupled with the roughly 5.3-hour phase delay of BMAL1 and the rhythmic inversion of CYP7A1, these findings strongly indicate a stress-induced phase reprogramming mechanism. While further validation is warranted, we hypothesize that this disruption impairs the temporal coordination between gene expression and transport, shifting bile acid homeostasis from physiological oscillation to pathological desynchrony.

Mechanistically, the 5.3-hour phase delay of BMAL1 and the attenuated amplitude of PER2/PER3 collectively reprogram the temporal expression of downstream bile acid transporters. This aligns with evidence from mouse intestine that intestinal BMAL1 directly transactivates MRP2 through E-box elements in its promoter region, and that ablation of BMAL1 abolishes the circadian rhythm of MRP2 expression (Yu, et al., 2019). In control livers, the daytime negative correlation between clock genes and MRP2 (ZT2–ZT10) likely reflects physiological suppression of canalicular efflux during the resting phase, thereby conserving hepatic bile acid pools. This pattern is consistent with the established circadian regulation of bile acid transporters, where clock-driven repression during the fasting phase prevents excessive bile acid loss (Noshiro, et al., 2007; Pathak, et al., 2013b). The nocturnal positive correlation (ZT14–ZT22) normally permits moderate biliary secretion during the active phase. In the chronic corticosterone group, the loss of this circadian gating was accompanied by persistent positive correlations throughout the cycle, which may reflect an impairment of the daytime 'brake' on MRP2 and an amplified nocturnal efflux capacity. This phenotype resembles the bile acid disruption observed in circadian mutant models: PER1/PER2 double-knockout mice exhibit transiently increased hepatic bile acids despite suppressed CYP7A1, indicating that clock disruption uncouples synthesis from transport rhythms (Ma, et al., 2009). Critically, at ZT14—when the phase-delayed BMAL1 peak coincides with rising MRP2 expression—the normally transient positive correlation is prolonged and intensified, which may create a temporal window favoring increased biliary efflux that persists through ZT22. Such stress-induced phase misalignment between BMAL1 and its target transporters has been documented in sleep deprivation models, where circadian disruption similarly impairs the temporal coordination of bile acid synthesis and transport, resulting in hepatic depletion and compensatory gallbladder accumulation (Ferrell and Chiang, 2015). This temporal misalignment may partially explain why hepatic bile acids decline precipitously at ZT18 while gallbladder storage accumulates abnormally by ZT22.

In the liver, canalicular bile acid efflux is primarily mediated by transporters such as BSEP and MRP2. Although DEX increased BSEP mRNA expression, no corresponding increase in BSEP protein abundance was detected, suggesting that BSEP expression may be regulated beyond transcription. This interpretation is consistent with previous studies showing that BSEP abundance and canalicular localization are dynamically regulated through protein trafficking and endocytic recycling (Kubitz, et al., 2004; Soroka and Boyer, 2014). In contrast, DEX increased MRP2 protein expression and its canalicular membrane localization, accompanied by enhanced polarized TCA efflux. The complete abolition of these effects by the glucocorticoid receptor (GR) antagonist RU486 indicates that GC-induced MRP2 activation is GR dependent. Previous studies have demonstrated that MRP2 transcription can be regulated through glucocorticoid response elements (GREs) in its promoter (Kauffmann and Schrenk, 1998), while other evidence suggests that post-transcriptional mechanisms, including regulation of protein stability and membrane trafficking, also contribute to MRP2 expression and function (Cao, et al., 2001; Trauner, et al., 1997). Crucially, physiological GC secretion follows a robust diurnal rhythm, peaking during the active daytime phase and declining at night. Therefore, chronic high-dose GC exposure exerts its most disruptive impact during the nocturnal trough. During this nighttime fasting period, persistent GR activation forces abnormal, continuous MRP2-mediated bile acid efflux, abolishing the natural resting phase of the enterohepatic circulation. Given that previous research establishes the PER family as essential regulators of MRP2 rhythmicity (Ma, et al., 2009), the GC-driven alteration of PER3 observed in our study likely serves as a key mechanism breaking the natural daily timing of this liver transporter. Although birds and mammals differ in bile acid composition and metabolism (Hagey, et al., 2010; Thakare, et al., 2018), the core molecular architecture of the circadian clock, including the CLOCK/BMAL1–PER/CRY transcriptional–translational feedback loop, is highly conserved across vertebrates (Cassone, 2014), and GR-dependent regulation of bile acid transporters has been well documented in mammals (Prevoo, et al., 2011). Therefore, AML12 cells were used as a complementary in vitro system for mechanistic validation, while future studies in avian hepatocytes will further extend the physiological relevance of these findings.

Several limitations of the present study warrant consideration. Feeding is a critical determinant influencing the distribution of bile acids within the enterohepatic circulation. Because sampling was conducted under ad libitum conditions to preserve natural rhythms, individual differences in immediate feeding status may have introduced variance. Future investigations employing a larger sample size are necessary to mitigate this potential confounding effect and further validate our findings. In addition, our experiments only measured total bile acid concentrations. Future studies should combine targeted metabolomics with gut microbiota analysis to fully delineate the circadian remodeling of the bile acid pool and to clarify the interplay between gut microbiota and bile acids in chronic stress-induced metabolic disturbances.

Although our current findings are derived from broilers, the stress-induced diurnal alterations in hepatic bile acid metabolism may also contribute to severe metabolic disturbances in laying hens, such as fatty liver hemorrhagic syndrome (Zhang, et al., 2026). Given that laying hen production heavily relies on precise photoperiodic and rhythmic feeding strategies (Dan, et al., 2017; do Nascimento, et al., 2025; Liu, et al., 2024), temporally aligning dietary modulators with physiological peaks of bile acid efflux could offer a novel chronobiological strategy to alleviate stress-induced hepatic burden across broader poultry farming.

Conclusion

In summary, chronic corticosterone exposure disrupts hepatic circadian clock rhythmicity, uncouples clock genes from bile acid metabolic genes, and promotes excessive nocturnal bile acid efflux, which is likely coupled with a subsequent feedback inhibition of synthesis in broiler chickens. These findings provide novel insights into the crosstalk between stress signaling and the circadian regulation of bile acid homeostasis, and underscore the importance of chronobiological strategies for optimizing metabolic health and production performance in poultry.

Consent for publication

Not applicable

Author contributions

Aijia Zhang designed the study, developed the experimental methodology, and prepared the original manuscript. Chu Meng performed the experimental validation and prepared the original manuscript. Qi Liu contributed to the experimental procedures and organized the experimental data. Qian Shi contributed to the experimental procedures and performed the data processing and software-based analyses. Aizhi Cao performed the data analysis and interpretation. Qu Chen analyzed the data, participated in the experimental work, and critically revised the manuscript. Yimin Jia secured funding and resources for the study, supervised and coordinated the project, and critically revised the manuscript.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.