Corticosterone disrupts the temporal coordination of the gut–liver bile acid axis in broiler chickens

Abstract

Elevated corticosterone levels resulting from chronic stress impair metabolic balance in broilers, frequently culminating in lipid dysregulation. While bile acids (BAs) are essential for managing lipid metabolism, how chronic stress alters BA dynamics and drives metabolic failure across the gut-liver axis is not fully understood. To address this, we developed a chronic corticosterone-exposure model in broilers (divided into control and stressed groups) to mimic prolonged stress. We evaluated temporal changes across two zeitgeber time (ZT) points (ZT2 and ZT14) using transcriptomics for both the liver and ileum, alongside ZT2 hepatic metabolomics and ZT14 ileal metabolomics paired with ileal microbiota analysis. The stress model exhibited retarded growth, hepatic fat accumulation, and hyperlipidemia. Crucially, corticosterone disrupted the diurnal variation in BAs in a highly time- and tissue-dependent manner. During the day (ZT2), hepatic dysfunction was most pronounced, marked by the suppression of BA synthesis (cytochrome P450 family 7 subfamily A member 1, CYP7A1) and conjugation (bile acid-CoA:amino acid N-acyltransferase, BAAT), reduced fatty acid oxidation, and a lower conjugated-to-unconjugated BA ratio. In contrast, nocturnal (ZT14) impairments were localized to the ileum, where decreased expression of intestinal barrier components and BA transporters (organic solute transporter beta, OSTβ) coincided with the accumulation of total and unconjugated BAs. Our findings reveal the spatiotemporal specificity of gut–liver bile acid axis dysregulation under chronic stress, and establish a robust theoretical foundation for precision temporal bile acid interventions to alleviate stress-induced metabolic disorders in yellow-feathered broiler production.

Introduction

Chronic environmental and management-related stressors can activate the hypothalamic–pituitary–adrenal (HPA) axis and elevate circulating corticosterone levels in broiler chickens. This hormonal surge triggers a cascade of detrimental physiological and behavioral responses, including oxidative damage, immunosuppression, compromised intestinal integrity, and ultimately, diminished growth and production performance (Ncho et al., 2025). In chickens, prolonged exposure to corticosterone is known to induce hepatic steatosis and severely perturb bile acid (BA) metabolism across the gut-liver axis (Wu et al., 2023). Similarly, chronic environmental challenges, such as heat stress, also exacerbate liver fat accumulation and disrupt BA equilibrium in broilers (Zhang et al., 2023).

Compared to mammals, broilers exhibit a distinctly different metabolic profile characterized by exceptionally high metabolic rates and a strong biological drive for energy efficiency (Korver, 2023). The liver—rather than adipose tissue—serves as the central hub for lipid synthesis and processing in avian species (Zaefarian et al., 2019). Consequently, BAs are indispensable for regulating hepatic and systemic lipid balance. Beyond their classical role as amphipathic molecules that emulsify dietary fats and aid in the absorption of fat-soluble vitamins, BAs function as potent signaling molecules. By activating receptors such as the farnesoid X receptor (FXR) and G protein-coupled bile acid receptor 1 (TGR5), they modulate lipid, glucose, and overall energy homeostasis, thereby buffering against metabolic dysfunction. The availability of BAs is tightly controlled through enterohepatic circulation: synthesized in the liver and stored in the gallbladder, BAs are released into the duodenum upon feeding. Up to 95 % of these BAs are actively reclaimed in the ileum and shuttled back to the liver via the portal vein for resecretion (Fleishman and Kumar, 2024). This highly efficient recycling system relies on a network of synthesis and transport genes operating between the liver and intestine. Notably, supplementing poultry diets with BAs has proven effective in mitigating stress, repairing intestinal barriers, and correcting lipid disorders (Chen et al., 2024a; Hu et al., 2024a; Hu et al., 2024b).

A critical aspect of BA metabolism is its diurnal rhythmicity, which enables the host to synchronize digestive and metabolic activities with feeding cycles and environmental cues. This temporal coordination is vital for sustaining metabolic health. While research in humans (Bello et al., 2024; Galman et al., 2005) and rodents (Eggink et al., 2017; Zhang et al., 2011) clearly demonstrates significant daily fluctuations in BA pool size, composition, and the rhythmic expression of BA-related genes, such circadian dynamics remain poorly defined in avian models. Specifically, despite known associations between glucocorticoid elevation and BA alterations in poultry, the diurnal dysregulation of BAs under chronic stress has not been thoroughly investigated.

We hypothesized that chronic corticosterone exposure disrupts the temporal coordination of the gut–liver BA axis in broilers, leading to alterations in BA metabolism that are associated with both impaired lipid homeostasis and intestinal injury. Therefore, the aim of this study was to characterize these temporal responses by comparing the hepatic and ileal effects of chronic corticosterone exposure across different times of the day. By deciphering these phase-specific mechanisms of metabolic failure, our findings provide a robust theoretical basis for the precise timing and application of dietary BAs in poultry production. Ultimately, understanding these pathways is of great significance for formulating practical strategies to enhance stress resilience and improve overall efficiency in yellow-feathered broiler farming.

Materials and methods

Establishment of a chronic corticosterone-induced stress model in broilers

Given the sex-dependent differences in stress and metabolic responses in broilers (Marin et al., 2002), only males were selected to minimize biological variation. A total of 24 male dwarf yellow-feathered broilers (63 days old), sourced from Jiangsu Lihua Animal Husbandry Co., Ltd. (Changzhou, Jiangsu, China), were housed in battery cages (eight cages in total, with three birds per cage and four cages per experimental group) within an environmentally controlled room at the animal research facility of Nanjing Agricultural University. Each cage measured 60 × 60 × 45 cm (length × width × height). The ambient temperature was maintained at 20–22°C, and relative humidity was kept at 50–60 %. The facility operated under a strict 12-h light:12-h dark photoperiod (lights on from 07:00 to 19:00), with an illumination intensity of approximately 10 lux. Before the initiation of the experiment, all cages were randomly distributed throughout the animal room, and cages assigned to the normal control (Norm) and corticosterone-treated (Cort) groups were interspersed to minimize potential cage-location effects associated with environmental position. To ensure baseline uniformity, the birds were weighed and randomly allocated into two experimental groups—a normal control (Norm) and a corticosterone-treated group (Cort)—using a random number generator in Microsoft Excel. No significant difference in initial body weight was observed between the two groups (P > 0.05). Each group consisted of 12 birds housed in four cages (three birds per cage). Throughout the experiment, all birds were provided feed and water ad libitum. Feed intake was recorded daily at the cage level, whereas the body weight of each individual bird was measured separately every other day. The basal diet was formulated based on the National Agricultural Standard for Nutritional Requirements for Yellow-feather Broiler Chickens (NY/T 3645-2020). The ingredient composition and calculated nutrient levels of the experimental diet are presented in Supplementary Table S1.

Corticosterone powder (C104537, Aladdin, Shanghai, China) was initially dissolved in absolute ethanol to create a 1 % stock solution, which was shielded from light and stored at 4°C. Prior to administration, this stock was diluted with physiological saline to achieve a final vehicle concentration of 15 % ethanol. Following a 3-day adaptation period, birds in the Cort group were administered daily subcutaneous injections in the neck region (4 mg/kg body weight) for 10 successive days. The 10-day treatment duration and corticosterone dosage were selected based on established chronic corticosterone exposure protocols in poultry (Chen et al., 2025; Hu et al., 2018). In these studies, repeated corticosterone administration for 7 days or longer was used to induce sustained physiological and metabolic responses, while the selected dose was shown to effectively elevate serum corticosterone concentrations from baseline levels to stress-associated levels in broilers (Yang et al., 2022). Sampling was strategically conducted on day 11 (one day after the final injection) to eliminate the acute pharmacological peak effects of the final dose. The Norm group received an equivalent volume of the 15 % ethanol vehicle.

Sample collection

Sampling occurred one day after the final injection. Six birds from each group were randomly selected for sampling at two specific zeitgeber times (ZT): daytime (ZT2, 09:00, two hours post-illumination) and nighttime (ZT14, 21:00, two hours post-darkness). We chose these specific intervals to capture critical metabolic transitions: ZT2 reflects the early active feeding phase and the corresponding surge in post-prandial BA secretion, whereas ZT14 represents the entry into the resting, post-absorptive state.

At each time point, final body weights were recorded, and blood samples were drawn from the brachial wing vein into 5 mL heparinized tubes. The birds were subsequently euthanized via decapitation. Liver and abdominal fat pads were rapidly excised and weighed. To ensure the validity of subsequent cross-tissue analyses, the liver, ileum, and corresponding digesta samples collected at a specific time point were rigorously paired to the exact same individual bird. A representative liver fraction was preserved in 4 % paraformaldehyde at 4°C for histological evaluation. All other collected specimens—including the remaining liver tissue, ileum, ileal digesta, and colonic contents—were flash-frozen in liquid nitrogen and transferred to a −80°C freezer until further processing. Specifically, ileal digesta were selected for targeted metabolomics and 16S rRNA sequencing as the ileum represents the principal anatomical site for active enterohepatic bile acid reabsorption, whereas colonic contents were utilized to capture the end-point excretion and cumulative retention profiles of total bile acids.

Biochemical analysis

Heparinized blood was centrifuged at 3,000 rpm for 5 min to harvest the plasma. An automated clinical chemistry analyzer (Hitachi 7020, Tokyo, Japan) was then utilized to quantify the plasma concentrations of glucose (GLU), triglycerides (TG), total cholesterol (TC), high- and low-density lipoprotein cholesterol (HDL-C and LDL-C), non-esterified fatty acids (NEFA), and total BAs (TBA). Plasma corticosterone concentrations were quantified using an Enzyme Immunoassay kit (ADI-900-097, Enzo, Farmingdale, NY) in accordance with the manufacturer’s instructions.

Bile acid analysis of colonic contents

We extracted colonic TBA following a modified procedure described previously (Hua et al., 2018). First, colonic digesta were oven-dried at 60°C to a constant weight and pulverized into a fine powder. A ∼200 mg aliquot of this powder was homogenized in 1 mL of 95 % ethanol and subjected to an overnight water bath incubation at 60°C. The mixture was centrifuged (8,000 rpm, 10 min), and the resulting supernatant was transferred to a fresh 5 mL collection tube. To maximize extraction, the residual pellet was resuspended in 1 mL of 80 % ethanol, vigorously vortexed, and subjected to a second overnight incubation at 60°C. Following another centrifugation step under identical conditions, this second supernatant was pooled with the first. Finally, the remaining pellet was suspended in 1 mL of a chloroform-methanol solution (2:1, v/v) and allowed to stand at room temperature for approximately 12 hours. After a final centrifugation (8,000 rpm, 10 min), all three supernatant fractions were combined and thoroughly mixed. The TBA concentration of this pooled final extract was subsequently measured using the Hitachi 7020 automated analyzer, adhering strictly to the manufacturer's guidelines.

Determination of hepatic triglyceride and total cholesterol levels

To assess lipid accumulation in the liver, we measured triglyceride (TG) and total cholesterol (TC) levels utilizing specific colorimetric kits (E1025-105 for TG; E1026-105 for TC; Applygen Technologies, Beijing, China). Briefly, ∼50 mg of liver tissue was lysed in 1 mL of the provided buffer. After a 10-minute incubation at 70°C, the lysates were centrifuged to harvest the supernatant. A fraction of this supernatant was reserved for total protein quantification using a bicinchoninic acid (BCA) Protein Assay Kit (P0012, Beyotime, Shanghai, China). For lipid measurements, 10 μL of each sample was reacted with 190 μL of the working solution at 37°C for 20 min. Optical densities were recorded at 550 nm on a BioTek ELx800 microplate reader (BioTek, Winooski, VT). Final TG and TC values were derived against standard curves and adjusted for protein concentration.

Histological evaluation (H&E) and Oil Red O staining

Liver specimens preserved in 4 % paraformaldehyde underwent routine histological processing. For morphological assessment, samples were embedded in paraffin, sectioned, and stained with Haematoxylin-Eosin. To visualize neutral lipids, separate tissue fragments were cryopreserved, sectioned, and treated with an Oil Red O staining kit (G1261, Solarbio, Shanghai, China) following the manufacturer's guidelines. Microscopic imaging of all slides was captured using a Zeiss Axio Imager Z2 light microscope (Carl Zeiss, Göttingen, Germany).

Hepatic and ileal transcriptome sequencing

Total RNA was isolated from liver and ileum samples utilizing TsingZol reagent (TSP401, TSINGKE, Nanjing, Jiangsu, China), with yield and purity evaluated via a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE). We captured polyadenylated mRNA using Oligo(dT) magnetic beads (Cat. No. 25-61005, Thermo Fisher Scientific, Waltham, MA) and subjected it to ion-catalyzed fragmentation, yielding ∼300 bp fragments. These fragments served as templates for first-strand cDNA synthesis primed with random hexamers, followed by second-strand synthesis. The resulting double-stranded cDNA was amplified via polymerase chain reaction (PCR) and size-selected to generate ∼450 bp library constructs.

After verifying library integrity using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA), libraries were pooled according to their specific index barcodes and target sequencing depth. The pool was diluted to 2 nM, subjected to alkaline denaturation, and sequenced on an Illumina HiSeq platform (Illumina, San Diego, CA) in paired-end mode. Raw reads were rigorously filtered to eliminate low-quality sequences. The resulting clean reads were mapped against the Gallus gallus reference genome (v5.0, Database version 94.5). Genes exhibiting an absolute fold change > 2 and an adjusted P-value < 0.05 were defined as differentially expressed genes (DEGs) for subsequent clustering and functional annotation.

Transcriptomic interaction analysis and data visualization

Data wrangling and transformation were performed using the dplyr (version 1.2.0) and tidyr (version 1.3.2) packages. To identify genes whose expression patterns were significantly disrupted by corticosterone treatment and time, we performed an interaction analysis using the limma R package (version 3.66.0). A two-factor experimental design matrix (Time × Treatment) was constructed, and the interaction effect was evaluated using the specific contrast formula. P-values were moderated using the Empirical Bayes (eBayes) method. The top 100 significant interaction genes were visualized using the pheatmap package (version 1.0.13) with row-wise Z-score standardization. To preserve the biological temporal logic while ensuring smooth visual transitions, global column clustering was disabled, and an intra-group hierarchical clustering approach based on Euclidean distances was applied to reorder biological replicates strictly within each specific subgroup.

Transcriptomic data preprocessing and statistical analyses were performed using R software. To account for technical variations arising from separate sequencing batches for ZT2 and ZT14 samples, the removeBatchEffect function from the limma package was rigorously applied to the expression matrices prior to downstream temporal comparisons. Differential gene expression analysis was conducted using the linear modeling features and empirical Bayes statistics (eBayes) of the limma package. Specifically, cross-sectional comparisons between the Corticosterone (Cort) and Normal (Norm) groups were performed independently within each circadian time point (ZT2 and ZT14). Statistical significance was defined as P < 0.05 and denoted by asterisks (* P < 0.05, **P < 0.01, *** P < 0.001). For the temporal dynamic analysis of predefined core pathways, absolute transcript abundance (BaseMean) was extracted, and the relative expression levels of genes were row-scaled to Z-scores. Within each targeted module, the top 10 genes exhibiting the highest fluctuation amplitude (calculated as the difference between the maximum and minimum mean Z-scores across all groups) were extracted. To accurately reflect macroscopic biological trends while mitigating the statistical bias of low-abundance transcripts, a weighted average Z-score was computed for each time point (ZT2 and ZT14), using the gene's absolute abundance (BaseMean) across all samples as the weight. Time-dependent alteration plots were generated using ggplot2 (version 4.0.2) and ggrepel (version 0.9.8). In these visualizations, the dynamics of the top fluctuating genes within each pathway are represented by thin lines, with point sizes logarithmically scaled to their absolute abundance. Up to 10 genes with the highest fluctuation amplitude were displayed for each pathway; for pathways containing fewer than 10 enriched genes, all available genes were included. The macroscopic weighted average trends of the entire pathway for both the Cort and Norm groups are highlighted by thick solid lines, demonstrating the temporal shifts between the two treatments.

Bile acid-targeted metabolomics

Sample extraction: Approximately 50 mg of either liver tissue or ileal digesta was combined with 1 mL of methanol and bead-beaten (three 90-s cycles at 60 Hz). Mixtures were sonicated for 30 min at ambient temperature and cleared by centrifugation (12,000 rpm, 10 min, 4°C). Next, 100 μL of the cleared extract was further diluted with 900 μL of methanol, and a 300 μL portion was filtered (0.22 μm) directly into autosampler vials.

Liquid chromatography-tandem mass spectrometry (LC-MS/MS) conditions: Extracts (5 μL injections) were resolved on a Waters ACQUITY ultra-performance liquid chromatography (UPLC) platform fitted with a BEH C18 column (2.1 × 100 mm, 1.7 μm; Waters, Milford, MA) held at 40°C. The binary mobile phase comprised 0.01 % aqueous formic acid (Buffer A) and pure acetonitrile (Buffer B) at a flow rate of 0.25 mL/min. The gradient program for B was: 25 % (0–4 min), 25–30 % (4–9 min), 30–36 % (9–14 min), 36–38 % (14–18 min), 38–50 % (18–24 min), 50–75 % (24–32 min), 75–100 % (32–35 min), and returning to 25 % (35–38 min).

Detection was performed in negative electrospray ionization (ESI) mode utilizing an AB Sciex 5500 triple quadrupole mass spectrometer (AB Sciex, Framingham, MA). Optimal MS settings were: source temperature 500°C, ion spray voltage −4500 V, collision gas 6 psi, curtain gas 30 psi, and nebulizer/auxiliary gases at 50 psi. Scanning occurred in multiple reaction monitoring (MRM) mode. Analytical validation confirmed robust performance: all standard curves yielded R² > 0.99, recovery efficiencies spanned 85–110 %, and intra-day precision (RSD) remained under 10 %.

16S ribosomal RNA (16S rRNA) gene sequencing

To profile the gut microbiome, total microbial DNA was extracted from the cryopreserved ileal digesta and verified for quality. Specific primers were utilized to amplify hypervariable regions of the 16S rRNA gene. The resulting amplicons were purified, quantified, and formulated into sequencing libraries using the Illumina TruSeq Nano DNA LT Kit. The microbial community composition was evaluated via paired-end sequencing on an Illumina MiSeq platform (Illumina, San Diego, CA).

Statistical analysis

Statistical analyses were performed using SPSS Statistics (version 25.0, IBM Corp., Armonk, NY), while data visualization and graphing were generated using GraphPad Prism 9 (GraphPad Software, Boston, MA). Results are expressed as the mean ± standard error of the mean (SEM). Prior to comparative analysis, the normality of data distribution and homogeneity of variances were rigorously assessed using the Shapiro-Wilk test and Levene’s test, respectively. For comparisons between two groups at the same time point, an unpaired Student’s t-test was utilized for normally distributed data; in instances where data failed the normality assumption, the non-parametric Mann-Whitney U test was applied.

For transcriptomic data, differential expression analysis was performed using the DESeq2 package in R (version 4.0.3), with genes identified as differentially expressed at an adjusted P-value < 0.05 and an absolute fold change > 2. Gene set enrichment analysis (GSEA) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were conducted using the clusterProfiler package. Correlation analysis between bile acid profiles and lipid metabolism indicators was performed using Spearman's correlation coefficient. For correlation analyses, data matrices were constructed by strictly matching the hepatic and ileal data points generated from the same biological replicate. In all analyses, a P-value < 0.05 was considered statistically significant.

Results

Corticosterone exposure suppresses growth and provokes hepatic steatosis

During the 10-day corticosterone treatment period, chronic corticosterone administration severely blunted body weight gain in broilers (Table 1). Average daily feed intake was largely comparable between groups throughout the experiment, with the exception of a transient increase in the Cort group on day 4 (P < 0.05; Table 2). Anatomical assessments revealed distinct hepatomegaly in stressed birds, as evidenced by significantly higher absolute liver weights and liver-to-body weight ratios (Fig. 1A and B). Additionally, the proportion of abdominal fat was notably elevated (Fig. 1D). Histological evaluations via H&E and Oil Red O staining corroborated these gross observations, displaying massive lipid droplet deposition within hepatocytes (Fig. 1E). This morphological evidence of hepatic steatosis perfectly aligned with the biochemical detection of sharply increased triglyceride concentrations in the liver (Fig. 1F).

Table 1.

Effects of corticosterone on body weight in broilers.

Time (day)	Norm group (kg)	Cort group (kg)
1	1.79 ± 0.036	1.78 ± 0.033
3	1.85 ± 0.033	1.77 ± 0.032
5	1.97 ± 0.031	1.85 ± 0.034 *
7	2.09 ± 0.033	1.91 ± 0.033 ⁎⁎
9	2.21 ± 0.035	1.91 ± 0.034 ⁎⁎
11	2.33 ± 0.035	1.98 ± 0.038 ⁎⁎

Note: Data are presented as mean ± SEM. For growth performance measurements, all birds in each group were included (n = 12 per group). Norm, control group; Cort, corticosterone-treated group.

^⁎P < 0.05.

^⁎⁎P < 0.01 compared with the Norm group at the same time point.

Table 2.

Effects of corticosterone on average daily feed intake in broilers.

Time (day)	Norm group (g/bird/day)	Cort group (g/bird/day)
1	84.81 ± 10.05	95.93 ± 15.41
2	120.00 ± 10.12	134.58 ± 19.71
3	156.67 ± 6.24	183.33 ± 11.00
4	150.91 ± 8.54	178.89 ± 6.81 *
5	173.19 ± 4.44	187.22 ± 6.55
6	157.78 ± 5.16	163.89 ± 7.77
7	159.72 ± 7.14	182.78 ± 8.53
8	155.28 ± 10.52	141.11 ± 9.20
9	155.00 ± 4.63	152.22 ± 6.44
10	163.06 ± 4.62	170.56 ± 13.97

Note: ADFI = average daily feed intake. Data are expressed as mean ± SEM (n = 4 replicate cages per group, 3 birds per cage). Norm, control group; Cort, corticosterone-treated group.

^⁎P < 0.05 compared with the Norm group.

Fig. 1.

Impact of corticosterone exposure on physical lipid deposition and hepatic morphology. Absolute weights and calculated relative indices for the liver (A, B) and abdominal fat pad (C, D). (E) Representative histological micrographs of liver sections stained with Oil Red O and H&E, illustrating hepatic lipid accumulation. Biochemical quantification of hepatic (F) triglycerides and (G) total cholesterol. Values are presented as the mean ± SEM. For growth performance measurements, all birds in each group were included (n = 12 per group). Norm: normal control group; Cort: corticosterone-treated group. Data from ZT2 and ZT14 were pooled for general phenotypic and morphological assessments due to the stable nature of organ weights and total lipid accumulation across the diurnal cycle. Asterisks denote statistical significance: *P < 0.05, **P < 0.01.

Spatiotemporal shifts in systemic lipids and bile acid distribution

Blood biochemistry analyses at ZT2 and ZT14 highlighted a profound disruption of systemic lipid homeostasis under chronic stress. Corticosterone robustly elevated circulating TC and LDL-C irrespective of the diurnal phase. Time-specific lipid alterations were also evident, with TG being significantly elevated at ZT14 and HDL-C rising significantly at ZT2 (Fig. 2A, 2B, 2D, 2E). While plasma GLU and NEFA remained relatively stable across treatments (Fig. 2C, 2F), corticosterone induced a striking phase-dependent redistribution of bile acids across the gut-liver axis. During the active phase (ZT2), plasma TBA spiked dramatically, mirroring a concurrent depletion of colonic TBA (Fig. 2G). Conversely, during the resting phase (ZT14), this dynamic inverted, resulting in a substantial buildup of TBA within the colonic contents of the Cort group compared to controls (Fig. 2H). To verify the efficacy of the chronic corticosterone exposure model, plasma corticosterone concentrations were measured at both ZT2 and ZT14. Compared with the Norm group, the Cort group showed significantly higher plasma corticosterone concentrations at ZT2 (34.1 ± 4.7 vs. 21.3 ± 1.7 ng/mL, P < 0.05) and ZT14 (36.1 ± 5.6 vs. 16.6 ± 1.4 ng/mL, P < 0.01) (Table 3).

Fig. 2.

Alterations in plasma biochemistry and bile acid pools induced by corticosterone. Shown are the plasma levels of (A) TC, (B) TG, (C) GLU, (D) HDL-C, (E) LDL-C, and (F) NEFA. Additionally, TBA concentrations were quantified in both plasma (G) and colonic contents (H). Data are reported as mean ± SEM (n = 6 birds per group at each zeitgeber time point). ZT: zeitgeber time; Norm: control group; Cort: corticosterone group. Significant differences are indicated by *P < 0.05 and **P < 0.01.

Table 3.

Plasma corticosterone levels in broilers at different Zeitgeber times.

Zeitgeber times (ZT)	Norm group (ng/mL)	Cort group (ng/mL)
ZT 2	21.3 ± 1.7	34.1 ± 4.7 *
ZT 14	16.6 ± 1.4	36.1 ± 5.6 ⁎⁎

Note: Data are expressed as mean ± SEM (n = 6 birds per group at each zeitgeber time point). Norm, control group; Cort, corticosterone-treated group; ZT, Zeitgeber time.

^⁎P < 0.05.

^⁎⁎P < 0.01 indicate significant differences compared with the Norm group at the same time point.

Effects of corticosterone exposure on the transcriptome of broiler liver and ileum

In the liver transcriptomic analysis, there were more DEGs during the day (541 up, 477 down) than at night (202 up, 91 down), and principal coordinate analysis (PCoA) revealed a clearer separation between the Cort and Norm groups at ZT2 than at ZT14 (Fig. 3A and B). We next performed limma interaction analysis to capture the combined effect of corticosterone treatment and time (ZT2/ZT14) and selected the top 100 genes with the most significant interaction. Heatmap clustering showed that the expression patterns of these interaction-significant genes were markedly time-dependent: at ZT2 (daytime), Cort and Norm samples formed clearly separated clusters, whereas at ZT14 (nighttime), the two groups intermingled and could not be distinguished (Fig. 3C). We analyzed the dynamic trajectories of core gene modules at ZT2 and ZT14 (Fig. 3D). The macroscopic weighted average trends (thick lines) revealed a profound temporal dysregulation, characterized by prominent trajectory cross-overs in multiple metabolic modules. Additionally, statistical significance was evaluated separately at ZT2 and ZT14 for individual genes (indicated by asterisks). In BA synthesis, the Norm group showed a physiological downward trend, whereas Cort exposure inverted this into a dramatic surge at ZT14. Specifically, synthesis genes like cytochrome P450 family 39 subfamily A member 1 (CYP39A1) and cytochrome P450 family 27 subfamily A member 1 (CYP27A1) were significantly upregulated in the Cort group at ZT14, while CYP7A1 and CYP39A1 were significantly downregulated at ZT2. A subtle intersection was also observed in BA transport, where the normal diurnal upregulation was blunted under Cort treatment. Furthermore, overall BA regulation pathways were constitutively suppressed at both time points in the Cort group, with key regulators like FXR and peroxisome proliferator-activated receptor alpha (PPARα) showing significant downregulation in Cort group specifically at ZT2. Similarly, a trajectory cross-over occurred in the Lipid Degradation module. In the Norm group, lipid degradation genes started at a high expression level at ZT2 and declined towards ZT14. In contrast, the Cort group exhibited a lower baseline of degradation at ZT2 (characterized by the significant downregulation of carnitine palmitoyltransferase 1A [CPT1A] and acyl-CoA oxidase 1 [ACOX1]) and a slower rate of change, leading to an intersection of the two trends. Moreover, while no intersection was seen in Lipid Synthesis, the Cort group failed to show the physiological decline observed in the Norm group, maintaining an aberrantly high synthetic state, which was accompanied by a significant upregulation of stearoyl-CoA desaturase (SCD) and elongation of very long chain fatty acids protein 6 (ELOVL6) in the Cort group at ZT14. Lipid transport displayed an abnormal elevation at both ZT2 and ZT14 in Cort group, driven by the significant upregulation of lipid droplet and binding proteins (perilipin 2 [PLIN2], PLIN4, and the fatty acid-binding protein [FABP] family) at ZT2, despite a significant downregulation of the cluster of differentiation 36 (CD36) transporter.

Fig. 3.

Corticosterone-induced shifts in the diurnal hepatic transcriptome. (A) Volcano plots highlighting differentially expressed genes (DEGs). (B) Principal coordinate analysis (PCoA) of the transcriptomic profiles. (C) Heatmap depicting the top 100 genes with the most significant Time × Treatment interaction effects. Data are row-scaled via Z-score standardization. Side color bars designate the zeitgeber time (yellow: ZT2; dark blue: ZT14) and experimental groups (red: Cort; light blue: Norm). (D) Temporal trajectory plots illustrating the dynamic diurnal changes in core metabolic modules (bile acid and lipid metabolism). Thin lines trace the relative expression (mean Z-scores) of up to 10 genes exhibiting the greatest fluctuation amplitude within each pathway, with marker sizes scaled to their absolute transcript abundance (Log10 BaseMean). The thick solid lines represent the abundance-weighted macroscopic averages for the Cort (red) and Norm (blue) groups, visualizing the overall pathway-level temporal deviations caused by chronic stress. Asterisks indicate significant differences between the Cort and Norm groups at the corresponding time point (*P < 0.05, **P < 0.01). n = 6 birds per group at each zeitgeber time point.

Transcriptome analysis of broiler ileum showed that, compared to ZT2 (daytime), corticosterone treatment had a greater impact at nighttime, with the Cort group showing markedly more DEGs (Fig. 4A) and a significantly clearer separation from the Norm group in PCoA (Fig. 4B). Heatmap clustering revealed that the expression patterns of the top 100 genes with significant interaction exhibited clear time-dependent characteristics. Unlike in the liver, at ZT14 (nighttime) compared to ZT2 (daytime), the Cort and Norm group samples formed more distinctly separated clusters (Fig. 4C). We analyzed the temporal dynamic trajectories of core gene modules at ZT2 and ZT14 (Fig. 4D). The macroscopic weighted average trend lines (thick lines) revealed that, in terms of bile acid transport, intestinal barrier & tight junctions, and stem cells & epithelial proliferation, the Cort group exhibited similar diurnal differences to the Norm group, but their overall gene expression levels were persistently suppressed. Statistical analysis of key individual genes (indicated by asterisks) further detailed these changes. Specifically, in the bile acid transport module, organic solute transporter alpha (OSTα) was significantly downregulated at ZT2. In the gut barrier & junctions module, zonula occludens-2 (ZO-2) was significantly downregulated at both time points and Occludin was significantly downregulated at ZT14, while ZO-1 was significantly upregulated at ZT14. In the stem/epithelial growth module, proliferating cell nuclear antigen (PCNA) was significantly upregulated at ZT14. Notably, the divergence between the two groups across these modules was more pronounced during the nighttime (ZT14). Furthermore, corticosterone treatment induced a continuous downregulation in FXR-FGF19 (fibroblast growth factor 19) signaling and goblet cell function & differentiation, highlighted by a significant downregulation of Krüppel-like factor 4 (KLF4) at ZT2 in the goblet cell module. Strikingly, in contrast to the Norm group, the pro-inflammatory mediators pathway was persistently activated by chronic stress at both ZT2 and ZT14, which was validated by the significant upregulation of interleukin-1 beta (IL-1β) at both time points and cyclooxygenase-2 (COX-2) at ZT14.

Fig. 4.

Corticosterone-induced shifts in the diurnal ileum transcriptome. (A) Volcano plots displaying differentially expressed genes (DEGs). (B) Principal coordinate analysis (PCoA) of the sample profiles. (C) Heatmap mapping the top 100 genes that exhibit the most pronounced Time × Treatment interaction. Data are standardized using row-wise Z-scores. Side color annotations specify the zeitgeber time (yellow: ZT2; dark blue: ZT14) and experimental groups (red: Cort; light blue: Norm). (D) Temporal trajectory charts illustrating the diurnal dynamics of targeted core functional modules between ZT2 and ZT14. For each pathway, thin lines map the relative expression (mean Z-scores) of the 10 most highly fluctuating genes, with marker sizes indicating their absolute transcript abundance (Log10 BaseMean). The thick solid lines denote the abundance-weighted macroscopic averages for both the Cort (red) and Norm (blue) groups, capturing the broader pathway-level temporal deviations driven by corticosterone. Asterisks indicate significant differences between the Cort and Norm groups at the corresponding time point (*P < 0.05, **P < 0.01). n = 6 birds per group at each zeitgeber time point.

Corticosterone-induced alterations in the hepatic transcriptome at ZT2

Given that our initial interaction models (Fig. 3) identified the active daytime phase (ZT2) as the window of maximum corticosterone-induced transcriptomic divergence in the liver, we zeroed in on the specific hepatic alterations occurring at this time point.

Gene Set Enrichment Analysis (GSEA) demonstrated that chronic stress systematically dismantled major metabolic cascades, notably repressing primary bile acid (BA) biosynthesis, fatty acid metabolism, unsaturated fatty acid synthesis, and PPAR signaling (Fig. 5A). This broad suppression was corroborated by Kyoto Encyclopedia of Genes and Genomes (KEGG) mapping, which showed massive enrichment in disrupted lipid and BA regulatory networks, particularly steroid hormone biosynthesis, xenobiotic metabolism via cytochrome P450, and the PPAR signaling pathway (Fig. 5B).

Fig. 5.

Hepatic transcriptomic profiling during the daytime active phase (ZT2). (A) Gene Set Enrichment Analysis (GSEA) demonstrating the alterations in major metabolic cascades between treatments, specifically highlighting primary bile acid biosynthesis, unsaturated fatty acid biosynthesis, fatty acid metabolism, and the PPAR signaling pathway. (B) Bubble chart detailing the top enriched KEGG pathways. (C) Box plots depicting the relative expression levels of crucial genes governing bile acid metabolism (SCP2, CYP7A1, CYP39A1, CYP8B1, SLC27A5, BAAT, and FXR) across the normal and stressed groups. (D) Heatmap representation of differentially expressed gene (DEG) signatures. (E) Spearman rank correlation matrix mapping the functional associations among specific metabolic markers. Values represent the mean ± SEM (n = 6 birds per group at ZT2). Norm: control group; Cort: corticosterone-exposed group. Statistical significance is denoted by *P < 0.05 and **P < 0.01.

At the individual transcript level, corticosterone drastically skewed the expression of core BA metabolic components. Crucial drivers of BA synthesis, transport, and receptor signaling—specifically sterol carrier protein 2 (SCP2), CYP7A1, solute carrier family 27 member 5 (SLC27A5), bile acid-CoA:amino acid N-acyltransferase (BAAT), and FXR—were markedly downregulated in stressed birds. In contrast, cytochrome P450 family 8 subfamily B member 1 (CYP8B1) experienced a significant compensatory upregulation (Fig. 5C).

This stress-induced metabolic shift was equally evident in the hepatic lipid networks. Heatmap visualizations revealed a distinct transcriptional divergence favoring fat accumulation: genes facilitating lipogenesis and storage (e.g., SCD, acyl-CoA cholesterol acyltransferase 1 [ACAT1], apolipoprotein C3 [APOC3]) were sharply upregulated, whereas those governing lipid oxidation and breakdown (e.g., CD36, CPT1A, ACOX1) were severely repressed (Fig. 5D). Correlation mapping further exposed a tight functional coupling between these two disrupted systems. Broadly, the transcriptional abundance of BA synthetic genes shared a robust positive correlation with markers of lipid catabolism, while exhibiting a strong negative correlation with lipogenic networks (Fig. 5E).

Corticosterone-induced alterations in the ileal transcriptome and gut microbiota at ZT14

Building on our prior interaction models (Fig. 4), which highlighted the nighttime phase (ZT14) as the period of maximum ileal sensitivity to corticosterone, we conducted a targeted transcriptomic analysis of the ileum at this specific time point.

Pathway enrichment analysis indicated that chronic stress significantly impacted primary bile acid (BA) biosynthesis, alongside key structural networks including tight junctions, focal adhesions, and cell adhesion molecules (Fig. 6A). Correspondingly, stressed birds exhibited a marked downregulation in the transcription of critical BA transporters (such as OSTβ) and essential regulatory elements (small heterodimer partner [SHP] and cholesterol 25-hydroxylase [CH25H], Fig. 6B). Furthermore, heatmap clustering visualized a widespread suppression of differentially expressed genes (DEGs) required for maintaining mucosal barrier integrity, epithelial homeostasis, and local immune responses (Fig. 6C). To further substantiate these transcriptomic findings, Western blot analysis confirmed a significant reduction in the protein expression of key tight junction components, ZO-1 and Occludin, in the ileal mucosa of stressed birds (Supplementary Fig. S1). Together, these transcriptional declines strongly suggest that corticosterone exposure inflicts substantial intestinal injury.

Fig. 6.

Ileal transcriptomic and microbial profiling during the nighttime resting phase (ZT14). (A) Gene Set Enrichment Analysis (GSEA) highlighting the core pathways altered between the two groups, including primary bile acid biosynthesis, tight junctions, focal adhesions, and cell adhesion molecules (CAMs). (B) Box plots illustrating the relative transcript abundance of critical bile acid transporters and regulatory elements (OSTα, OSTβ, ASBT, SHP, and CH25H). (C) Heatmap visualization comparing the differentially expressed gene (DEG) expression patterns between the normal and stressed birds. (D) Evaluation of gut microbial alpha diversity indices. (E) Principal coordinate analysis (PCoA) plots representing beta diversity across four distinct distance metrics. (F) PCoA evaluation of the predicted functional potential of the microbiome, based on annotations from the Enzyme Commission and KEGG Orthology databases. Data are presented as the mean ± SEM (n = 6 birds per group at ZT14 for transcriptomic analyses in A–C; n = 5 biological replicates per group for 16S rRNA sequencing in D–F). Norm: control group; Cort: corticosterone-exposed group. Asterisks indicate statistical significance: *P < 0.05, **P < 0.01.

Interestingly, profiling the microbial composition of the ileal contents revealed that corticosterone exposure had no significant impact on the gut microbiome. Alpha diversity metrics—including the Chao1, Simpson, Shannon, and observed species indices—confirmed that microbial richness and evenness remained highly stable between the treatment and control groups (Fig. 6D). Additionally, PCoA failed to show any distinct sample clustering across multiple distance metrics (Bray-Curtis, Jaccard, and UniFrac), reflecting virtually identical bacterial community structures (Fig. 6E). Finally, predictive modeling of the microbiota's functional potential, encompassing enzymatic and metabolic pathways, also revealed no stress-induced shifts (Fig. 6F). Collectively, these findings imply that corticosterone disrupts BA and systemic metabolic homeostasis primarily through direct actions on host tissue physiology; however, the elevation of bacterially-derived oxidized BAs (6-oxolithocholic acid and 7-oxolithocholic acid), classic products of bacterial hydroxysteroid dehydrogenase activity, indicates that altered microbial function within a compositionally stable community may also contribute.

Corticosterone-induced shifts in the hepatic bile acid profile at ZT2

Prolonged corticosterone treatment led to distinct alterations in the hepatic BA composition (Fig. 7A). While the absolute quantities of total, primary, secondary, conjugated, and unconjugated BAs remained statistically similar between treatments (Fig. 7B-G), the ratio of conjugated to unconjugated BAs dropped significantly in stressed birds (Fig. 7H). Because conjugated BAs are critical for lipid emulsification, a depressed ratio frequently precipitates lipid metabolic dysfunction. Conversely, an elevated proportion of unconjugated BAs—which possess strong cytotoxicity and poor aqueous solubility—can trigger hepatocellular injury and impair broader liver functions.

Fig. 7.

Corticosterone-induced alterations in the hepatic bile acid profile. (A) Orthogonal partial least squares discriminant analysis (OPLS-DA) of the liver metabolome. Absolute quantification of (B) total, (C) primary, and (D) secondary bile acids. (E) The calculated ratio of secondary to primary bile acids. Hepatic concentrations of (F) conjugated and (G) unconjugated bile acids, alongside (H) the conjugated-to-unconjugated ratio. (I) Detailed compositional profile of individual bile acid metabolites. Values represent the mean ± SEM (n = 6 birds per group at ZT2). Norm: control group; Cort: corticosterone-exposed group. Statistical significance and trends are indicated as follows: #P < 0.1, *P < 0.05, **P < 0.01, ***P < 0.001.

Among the 19 specific BAs quantified, taurocholic acid (TCA) and taurochenodeoxycholic acid (TCDCA) dominated the hepatic pool. Compared to normal broilers, the Cort group displayed a marked depletion in several specific metabolites, including chenodeoxycholic acid (CDCA), lithocholic acid (LCA), TCDCA, tauro-alpha-muricholic acid (T-alpha-MCA), tauro-beta-muricholic acid (T-beta-MCA), and the combination of taurohyodeoxycholic acid and tauroursodeoxycholic acid (THDCA+TUDCA). Concurrently, stressed birds exhibited a significant elevation in 6-oxolithocholic acid (6-keto-LCA) (Fig. 7I).

Corticosterone-induced shifts in the ileal bile acid profile at ZT14

In the ileal digesta, chronic stress severely disrupted the nocturnal BA composition (Fig. 8A). Although the levels of primary, secondary, and conjugated BAs were unaffected (Fig. 8C-E), corticosterone exposure triggered a massive accumulation of both total and unconjugated BAs (Fig. 8B, F). This accumulation was accompanied by a noticeable downward trend in the conjugated-to-unconjugated BA ratio (Fig. 8H). Profiling of individual metabolites revealed a significant spike in the unconjugated bile acid cholic acid (CA) within the Cort group. Additionally, several other metabolites—including 3beta-CA, alpha-muricholic acid (alpha-MCA), lithocholic acid 3-sulfate (LCA-3S), and 7-oxolithocholic acid (7-keto-LCA)—displayed a strong upward trend under stress conditions (Fig. 8I).

Fig. 8.

Shifts in the ileal bile acid pool induced by corticosterone exposure. (A) Orthogonal partial least squares discriminant analysis (OPLS-DA) of the ileal metabolome. Measured levels of (B) total, (C) primary, and (D) secondary bile acids. (E) The calculated ratio of secondary to primary bile acids. Ileal contents of (F) conjugated and (G) unconjugated bile acids, paired with (H) the conjugated-to-unconjugated ratio. (I) Distribution map of specific bile acid metabolites within the ileum. Data are presented as the mean ± SEM (n = 6 birds per group at ZT14). Norm: control group; Cort: corticosterone-treated group. Statistical differences and trends are denoted as #P < 0.1, *P < 0.05, and **P < 0.01.

Correlation between bile acid dynamics and lipid metabolism

To untangle the complex interactions among serum biochemistry, BA metabolites, and gene expression, we conducted a redundancy analysis (RDA). This multidimensional mapping clearly separated the stressed and control groups. Within the liver, variables representing BA synthesis genes and the majority of BAs (predominantly conjugated forms) plotted directly opposite to lipid metabolic indicators, revealing a strong inverse distribution (Fig. 9A). A similar opposing pattern emerged in the ileum, where transcripts governing BA reabsorption and markers of intestinal injury were inversely distributed against lipid indicators and ileal BAs (which were primarily unconjugated) (Fig. 9B).

Fig. 9.

Multivariate integration and correlation networks linking bile acid and lipid metabolism. (A) Redundancy analysis (RDA) illustrating the functional interplay between hepatic bile acid profiles and lipid indicators during the daytime phase (ZT2). (B) RDA mapping the relationships between the ileal bile acid pool and lipid metabolism during the nighttime phase (ZT14). (C) Comprehensive correlation matrix evaluating the cross-tissue associations between liver/ileum bile acid dynamics and overall lipid homeostasis.

Deeper Spearman correlation mapping solidified these findings: hepatic lipid deposition was robustly and negatively correlated with hepatic BA synthesis genes, the conjugated BA pool, and specifically CDCA. Furthermore, lipid accumulation exhibited a significant negative correlation with ileal BA reabsorption genes, T-alpha-MCA, and CA (Fig. 9C). Finally, the analysis uncovered a highly significant cross-tissue linkage, demonstrating that the expression of ileal BA reabsorption genes is tightly correlated with hepatic BA metabolic function.

Discussion

Consistent with prior research (Liu et al., 2024; Zaytsoff et al., 2019; Zhang et al., 2025), our chronic corticosterone-exposure model successfully induced a profound metabolic syndrome in broilers, highlighted by stunted growth, dyslipidemia, and significant fatty liver. Because feed consumption was largely stable across groups throughout most of the experimental period, except for a transient increase in the Cort group on Day 4, these metabolic deficits were clearly not driven by sustained differences in feed intake. Instead, as documented in earlier literature, elevated glucocorticoids force a systemic reallocation of energy in poultry—diverting nutritional substrates away from skeletal muscle development and toward intense lipogenesis and fat storage, which often culminates in muscle wasting (Permpoon et al., 2025; Swarbrick et al., 2021). Consequently, chronic stress may compromise feed efficiency and production performance in broiler chickens.

We specifically targeted ZT2 and ZT14 to capture the dynamic metabolic transitions of the gut-liver axis. ZT2 marks the onset of the active feeding phase and the corresponding post-prandial surge in BA circulation, whereas ZT14 reflects the transition into the dark, resting, and post-absorptive state. By mapping these two critical windows, our trajectory analyses unveiled a striking spatiotemporal divergence in stress responsiveness. During the daytime (ZT2), the liver bore the brunt of the metabolic damage: corticosterone severely blunted the physiological decline in BA synthesis and lipid catabolism, while locking lipogenic and transport pathways in an abnormally elevated state. This created an environment of unchecked lipid accumulation. Conversely, ileal dysfunction peaked at night (ZT14). Corticosterone induced a persistent suppression of pathways linked to barrier integrity and BA transport, while unnaturally sustaining an elevated pro-inflammatory state across the diurnal cycle. These patterns indicate a flawed, hyperactive compensatory response in the gut paired with broken BA feedback loops. Ultimately, this stress-induced desynchronization of the BA circadian rhythm directly sabotages the dynamic balance of lipid digestion, absorption, and breakdown, inevitably triggering systemic lipid disorders (Fiorucci et al., 2021).

Because the liver exhibited its most drastic transcriptional shifts during the day (ZT2), we zeroed in on this phase to untangle the underlying hepatic mechanisms. Pathway analyses confirmed that stress severely crippled primary BA biosynthesis. The marked suppression of key synthetic drivers—namely CYP7A1, CYP39A1, and SCP2—demonstrated a blockade in the conversion of cholesterol into primary BAs, a bottleneck further evidenced by elevated circulating LDL-C. Interestingly, corticosterone upregulated CYP8B1, which promotes CA synthesis and determines the CA/CDCA balance (Wang et al., 2024b). In our Cort group, increased CYP8B1, together with decreased hepatic CDCA at ZT2 and increased ileal CA at ZT14, suggested a shift toward a more hydrophilic BA profile. Because CDCA is a potent endogenous FXR ligand (Modica et al., 2009), its depletion may have contributed to the observed reduction in hepatic FXR signaling. Reduced FXR activity may, in turn, be associated with impaired lipid oxidation, consistent with the downregulation of CPT1A and ACOX1.

Hepatic FXR-SHP and intestinal FXR-FGF19 signaling normally repress CYP7A1 as part of the negative feedback control of BA synthesis (Chiang and Ferrell, 2020). In our research, the reduction in CYP7A1 may involve glucocorticoid-mediated effects through additional regulatory pathways, such as FGF21-mediated inhibition of CYP7A1 transcription and modulation of other transcriptional regulators involved in BA metabolism (Al-Aqil et al., 2018; Rosales et al., 2013; Wang et al., 1996). Meanwhile, reduced FXR signaling may represent a parallel response affecting lipid metabolism. These findings are consistent with mammalian evidence that CYP8B1-mediated alterations in the CA/CDCA ratio influence lipid metabolism (Bertaggia et al., 2017) and that glucocorticoids can increase CYP8B1 expression (Mork et al., 2016). The discrepancy with a previous chicken study reporting opposite CYP8B1 and FXR responses to stress (Wu et al., 2023) may reflect differences in stress duration, genetic background, or feedback dynamics.

Furthermore, corticosterone heavily repressed SLC27A5 and BAAT. SLC27A5 is vital for channeling free BAs into the endoplasmic reticulum, while BAAT catalyzes their subsequent conjugation with amino acids (Pellicoro et al., 2007). This conjugation renders BAs less toxic and highly efficient at emulsifying dietary fats. The inhibition of these enzymes, corroborated by a depressed conjugated-to-unconjugated BA ratio in the hepatic metabolome, suggests that chronic stress aggravates liver injury and steatosis through the cytotoxic accumulation of free, unconjugated BAs (Wu et al., 2023).

At night (ZT14), transcripts maintaining mucosal barrier integrity, immune homeostasis, and BA reabsorption were broadly suppressed in the ileum. Metabolically, this coincided with a massive pooling of total and unconjugated BAs in the ileal contents. Normally, BAs are actively reclaimed via the apical sodium-dependent bile acid transporter (ASBT), ferried across the enterocyte by the ileal bile acid-binding protein (IBABP), and extruded into the portal circulation by the OSTα/OSTβ heterodimer. Silencing intestinal SHP downregulates OSTα/β, crippling BA reabsorption (Nguyen et al., 2021), and OSTβ deficiency blocks ileal BA efflux, leading to intracellular BA accumulation and massive fecal loss (Ferrebee et al., 2018). Crucially, the ileum absorbs conjugated BAs highly efficiently (>95 %) via active ASBT transport, whereas unconjugated BAs rely on slow, passive diffusion (Ticho et al., 2019). Therefore, the combination of reduced hepatic conjugated BA synthesis and the suppression of ileal transporters (OSTβ) inevitably triggered the severe BA stasis observed in our Cort group. The resulting buildup of cytotoxic unconjugated BAs predictably exacerbated the observed intestinal barrier damage.

Interestingly, while previous poultry studies link stress-induced BA imbalances to gut dysbiosis (Zhang et al., 2023), our 16S rRNA sequencing revealed that corticosterone did not alter the alpha diversity or structural makeup of the ileal microbiome. This stability suggests that the metabolic disturbances in our model are likely primarily host-driven—stemming from direct hormonal disruption of endogenous signaling pathways (the FXR axis and CYP7A1/OSTβ transcription)—rather than from major microbial shifts. However, the presence of oxidized BA metabolites hints at potential alterations in microbial enzymatic functions, underscoring the need for future studies to assess the functional capacity of the gut microbiota. A limitation of this study is that our microbiota profiling was restricted to ileal digesta. The cecum and colon are the major sites for microbial-mediated secondary bile acid biotransformation; therefore, we cannot exclude potential microbial contributions to BA metabolism and metabolic phenotypes in the hindgut. Future work incorporating hindgut microbiota and metabolome profiling will be needed to fully delineate the microbial roles in stress-induced gut-liver axis disruption.

Exogenous BAs are already widely utilized as feed additives in the poultry industry. Supplementation can alleviate stress-induced fatty liver by suppressing lipogenesis (Liu et al., 2024), and it improves feed conversion in high-fat diets via the FXR-IGF2 (insulin-like growth factor 2) axis (Chen et al., 2024b). Their efficacy is deeply influenced by dietary fat sources (Geng et al., 2022) and interacts with specific bacterial taxa to downregulate fatty acid synthase (FAS) (Wang et al., 2024a). However, blind supplementation carries risks; for instance, isolated TCDCA administration can actually worsen adiposity by overstimulating BA receptors in fat tissues (Sun et al., 2025).

In our yellow-feathered broiler model, we observed distinct BA alterations at ZT2 in the liver and ZT14 in the ileum, raising the possibility that chrononutrition could offer a new dimension to BA supplementation in this production system. Human clinical chronotherapy has proven that timing drug delivery to match the diurnal peaks and troughs of hepatic detoxification drastically improves efficacy (Ohdo et al., 2019). Similar chronobiological strategies are gaining traction in animal husbandry, such as evening iron supplementation for better absorption in piglets (Dong et al., 2021), split-feeding regimens (morning high-protein/evening high-calcium) to optimize eggshell quality in laying hens (Akter et al., 2025), and mapping the diurnal balance of muscle anabolism (Kelu et al., 2020). Because our data explicitly show that stress induces daytime-dominant hepatic lipid accumulation and nighttime-dominant ileal BA stasis, the current industry standard of continuous, untargeted BA supplementation may be sub-optimal. Future strategies must move toward precision application—timing BA interventions to synchronize with the bird's natural circadian rhythms. This approach will maximize economic returns, optimize lipid metabolism, and avoid the adverse side effects of BA toxicity in yellow-feathered broiler farming. Our findings are derived from a medium-growing yellow-feathered broiler model under controlled conditions using male birds only, and their applicability to fast-growing commercial strains, diverse farming systems, or female birds requires further validation.

Conclusion

Chronic corticosterone exposure drives severe metabolic dysfunction in broilers and is associated with alterations in the gut–liver bile acid (BA) axis. Transcriptomic profiling revealed time-dependent patterns, with hepatic transcriptomic changes most pronounced at ZT2 and ileal transcriptomic changes more evident at ZT14. Targeted BA metabolomics at the corresponding tissue–time points was consistent with these transcriptomic trends, showing suppressed BA synthesis and lipid catabolism in the liver and impaired BA transport with accumulation of unconjugated BAs in the ileum. These insights establish a robust theoretical foundation for precision chronobiological BA interventions in broiler production.

Ethics approval and consent to participate

The ethical protocol for this animal study was approved by the Institutional Animal Care and Use Committee at Nanjing Agricultural University. This approval was granted in compliance with the national animal welfare guidelines (Guidelines on Ethical Treatment of Experimental Animals, Ministry of Science and Technology, China, 2006, No. 398). The project number is NJAU.No20201221083.

Consent for publication

Not applicable.

Funding

This work was supported by the Key Technologies Research and Development Program of China (2023YFD1300802) and National Natural Science Foundation of China (No.32604412). We gratefully acknowledge the research funding provided by Shandong Longchang Animal Health Product Co., Ltd.

Author contribution

Aijia Zhang: study conception and design, experiments, data analysis, and manuscript drafting. Xifei Xiao: experimental work, data interpretation, and manuscript writing and critical review. Yunan Wu: data acquisition and analysis, and writing and critical review of the corresponding sections. Zixuan Zhou: experimental methodology and data analysis, and writing and critical review of the corresponding sections. Qu Chen: data analysis and interpretation, and critical review and revision of the corresponding sections. Yimin Jia: study conception, supervision, data interpretation, and critical review and revision of the manuscript. All authors contributed to the manuscript, approved the final version for publication, and agreed to be accountable for all aspects of the work, including the accuracy and integrity of the research.

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.