Activation of gut FXR improves the metabolism of bile acids, intestinal barrier, and microbiota under cholestatic condition caused by GCDCA in mice

C57BL/6J male mice at 7 weeks of age were used for the intervention in accordance with our previous study (14). The mice were divided into three groups: the control group (10% dimethyl sulfoxide (DMSO) + 40% PEG300 + 5% Tween-80 + 45% saline; n = 5), the GCDCA gavage group (GCDCA powder dissolved in 10% DMSO, 40% PEG300, 5% Tween-80, and 45% saline; a dose of 1,000 mg/kg body weight of GCDCA was administered via a specific gavage needle; n = 4), and the GCDCA + GW4064 gavage group (GW4064 [30 mg/kg body weight] + GCDCA [1,000 mg/kg body weight], the dose of GW4064 was the same as that of GCDCA; n = 5). All the mice were tested three times a week for 12 weeks. All the mice were housed under a 12 h light/dark cycle at 20°C–22°C and 45 ± 5% humidity in a specific pathogen-free environment.

The liver tissues and terminal ileum contents were collected for identification of the BA profiles using ultra-performance liquid chromatography‒tandem mass spectrometry (UPLC‒MS/MS) by LC-Bio Technology Co., Ltd. (Hangzhou, China). Then, 50 mg of the sample was placed into a 1.5 mL Eppendorf (EP) tube (on ice), and 0.5 mL of 80% methanol in water (precooled at −20°C) was added for protein precipitation. The sample was subsequently ground in a freeze-grinder for 3 minutes (50 Hz, 4°C), shaken, extracted in a multifunctional mixer for 20 minutes, and centrifuged (4°C, 20,000 rcf) for 15 minutes. The supernatant was transferred to a new 1.5 mL EP tube and centrifuged (4°C, 20,000 rcf) for 15 minutes. Finally, 200 µL of solution was transferred into an injection bottle for LC‒MS/MS analysis.

The standard substances were dissolved or diluted and mixed to obtain a 2 µg/mL mixed standard stock solution. The mixed standard stock solution was diluted stepwise to prepare standard solutions of different concentrations for standard curve generation.

A Vanquish liquid-phase system (Thermo) equipped with an Agilent Eclipse Plus C18 RRHD column (3.0 * 150 mm, 1.8 µm) was used for separation. Mobile phase A was a 10 mmol/L ammonium acetate aqueous solution, and mobile phase B was a methanol-acetonitrile solution (volume ratio = 1:1). The column temperature was set to 40°C, and the automatic sampler temperature was set to 4°C. The injection volume was 5 µL. The flow rate was 0.45 µL/min. The mobile phase elution gradient was as follows: 0–6 minutes, 40%–50% B; 6–8 minutes, 50% B; 8–11.5 minutes, 50%–60% B; 11.5–12 minutes, 60% B; 12–15.5 minutes, 60%–90% B; 15.5–17.5 minutes, 90%–98% B; 17.5–18 minutes, 98%; 18–18.1 minutes, 98–40% B; and 18.1–21 minutes, 40% B.

A TSQ Altis triple quadrupole mass spectrometer (Thermo) with an electrospray ionization (ESI) interface was used. The typical ion source parameters were as follows: negative ion voltage = 3,000 V, sheath gas = 20 Arb, collision gas = 10 Arb, auxiliary gas = 0 Arb, evaporation temperature = 350°C, and ion transfer tube temperature = 350°C. Flow injection analysis was conducted by injecting the standard solution of a single analyte into the ESI source of the mass spectrometer to optimize the multiple reaction monitoring (MRM) parameters of each target analyte. During the optimization process for each compound, the Q1/Q3 ratio with the highest response was selected as the quantitative ion, and the others were selected as the qualitative ions. Thermo Scientific Xcalibur (version 4.4) software was used for MRM data collection and processing (). Supplementary S1

The 341 forward primer (5′-CCTACGGGNGGCWGCAG-3) and the 805 reverse primer (5′-GACTACHVGGGTATCTAATCC-3) were used for prokaryotic 16S rRNA gene amplification of the V3-V4 region, and the 5´ ends of the primers were tagged with specific barcodes. The total volume of the reaction mixture was 25 µL, which consisted of 50 ng of template DNA, 12.5 µL of PCR Premix, 2.5 µL of each primer, and PCR-grade water. The PCR procedure involved the following steps: (i) initial denaturation at 98°C for 30 seconds; (ii) 32 cycles of denaturation at 98°C for 10 seconds, annealing at 54°C for 30 seconds, and extension at 72°C for 45 seconds; and (iii) a final extension at 72°C for 10 minutes.

The PCR products were confirmed with 2% agarose gel electrophoresis. Throughout the DNA extraction process, ultrapure water, instead of sample mixture, was used to exclude the possibility of false-positive PCR results as a negative control. The PCR products were purified with AMPure XT beads (Beckman Coulter Genomics, Danvers, MA, USA) and quantified with a Qubit fluorometer (Invitrogen, USA). The amplicon pools were prepared for sequencing, and the size and quantity of the amplicon library were assessed on an Agilent 2100 Bioanalyzer (Agilent, USA) and with a Library Quantification Kit for Illumina (Kapa Biosciences, Woburn, MA, USA), respectively. The libraries were sequenced on a NovaSeq PE250 platform by LC-Bio Technology Co., Ltd. (Hangzhou, China).

The liver and terminal ileum tissue sections were first deparaffinized and rehydrated. Antigen retrieval was then performed, followed by blocking of nonspecific binding sites with 5% bovine serum albumin (BSA) for 20 minutes. The sections were incubated overnight at 4°C with primary antibodies against FXR (1:200, 25055-1-AP, Proteintech), FGF15 (1:200, PK15481, ABmart), mucin2 (1:1,000, GB11344, Servicebio), claudin-1 (1:1,000, GB15032, Servicebio), occludin (1:500, GB111401↗, Servicebio), and ZO-1 (1:1,000, GB115686↗, Servicebio). Afterward, the samples were incubated in secondary antibody at room temperature for 60 minutes. Finally, the sections were examined under a microscope after diaminobenzidine (DAB) and hematoxylin staining.

Cutadapt (v.1.9) was used to remove the sequencing primer from the demultiplexed raw sequences. Paired-end reads were merged using FLASH (v.1.2.8). The low-quality reads (quality scores <20), short reads (<100 bp), and reads containing more than 5% “N” records were trimmed using the sliding window algorithm to obtain high-quality clean tags in fqtrim (v.0.94). Chimeric sequences were filtered using Vsearch software (v.2.3.4). After denoising for amplicon sequence variations (ASVs) using DADA2, the ASV feature table and feature sequence were obtained. BLAST was used for sequence alignment, and the feature sequences were annotated with the SILVA and NT-16S databases for each representative sequence. Alpha diversity was used to analyze the complexity of species diversity in a sample on the basis of three indices, namely the Chao1, Shannon, and Simpson indices. All the indices for the samples were calculated with QIIME2. The beta diversity was calculated with QIIME2, and the graphs were drawn with the R package. On the basis of the ASV abundance table, principal component analysis (PCA) was conducted. Principal coordinate analysis (PCoA) and nonmetric multidimensional scaling (NMDS) based on the Bray-Curtis distance were performed. Finally, the abundance of species at each level in each sample was calculated on the basis of the ASV abundance table. The linear discriminant analysis (LDA) effect size (LEfSe, LDA ≥3.0, P-value <0.05) was calculated using nsegata-Lefer. The R package (v.3.5.2) was used for image production.

The correlations between the abundances of gut microbes and BAs were determined and visualized using OmicStudio Cloud (https://www.omicstudio.cn/↗). A P-value <0.05 was considered to indicate statistical significance.

The BA profile of the liver was detected (Fig. 1). Compared with those in the control group, the levels of beta-muricholic acid (beta-MCA), murideoxycholic acid (MDCA), tauroursodeoxycholic acid (TUDCA), taurohyodeoxycholic acid (THDCA), ursodeoxycholic acid (UDCA), glycocholic acid (GCA), taurocholic acid (TCA), hyodeoxycholic acid (HDCA), cholic acid (CA), allocholic acid (ACA), taurochenodeoxycholic acid (TCDCA) and chenodeoxycholic acid (CDCA) were greater in the GCDCA group, but the differences were not significant. In contrast, treatment with GW4064, an agonist of FXR, markedly reduced the concentrations of T-beta-MCA, beta-MCA, TUDCA, THDCA, UDCA, GCA, TCA, CA, TCDCA, and CDCA, which were strongly induced by GCDCA (Fig. 1; Supplementary S2).

The BA levels in the terminal ileum contents were also determined. The levels of T-alpha-MCA, T-beta-MCA, TUDCA, THDCA, and TCA were greater in the GCDCA group than in the control group. The levels of hydroguinonecarboxylicacid (DBH), UDCA, and lithocholic acid (LCA) were lower in the GCDCA group than in the control group. The differences in the BA levels between the GCDCA group and the GCDCA + GW4064 group were not significant (Fig. 2; Supplementary S3).

Compared with the control group, the GCDCA group presented higher total BA levels in the liver (187468.62 ± 96522.81 ng/g vs 97401.77 ± 48151.23 ng/g, P = 0.1082), and GW4064 markedly reduced the total BA levels in the liver compared with the GCDCA group (35316.63 ± 21224.21 ng/g vs 187468.62 ± 96522.81 ng/g, P = 0.048) (Fig. 3A). In addition, the GCDCA group presented a trend toward increased levels of primary BAs in the liver compared with the control group (182730.65 ± 94457.22 ng/g vs 93971.6 ± 46420.51 ng/g, P = 0.1051), and the GCDCA + GW4064 group presented lower liver primary BA levels than did the GCDCA group (33000.58 ± 20062.75 ng/g vs 182730.65 ± 94457.22 ng/g, P = 0.048) (Fig. 3B). The levels of total secondary bile acids in the liver did not differ among the three groups. In the terminal ileum, FXR activation promotes the expression of fibroblast growth factor 15 (mouse), which reaches the liver through the portal vein and inhibits the expression of Cyp7a1, attenuating BA production. The immunohistochemistry results indicated that GCDCA administration inhibited gut FXR and FGF15 expression, whereas GW4064 restored their expression (Fig. 3C). These results indicated that the administration of GCDCA induced cholestasis, whereas GW4064 alleviated cholestasis in the liver.

Compared with the control group, the GCDCA group presented greater total bile acid levels in the gut (8886052.41 ± 6776545.86 ng/g vs 2783280.64 ± 2937875.84 ng/g, P = 0.17), and there was no marked reduction in total BA levels in the gut in the GCDCA + GW4064 group compared with the GCDCA group (10304124.67 ± 8805289.13 ng/g vs 8886052.41 ± 6776545.86 ng/g, P = 0.793) (Fig. 3D). In addition, the GCDCA group presented increasing trends in the gut primary BA levels compared with those of the control group (8696497.00 ± 6573145.83 ng/g vs 2329070.26 ± 2610175.75 ng/g, P = 0.147), and no reduction in the levels of primary BAs was detected in the GCDCA + GW4064 group compared with those of the GCDCA group (9622555.02 ± 9213349.57 ng/g vs 8696497.00 ± 6573145.83 ng/g, P = 0.871) (Fig. 3E). The gut primary BA levels in the control group, GCDCA group, and GCDCA + GW4064 group were 2329070.26 ± 2610175.75 ng/g, 8696497.00 ± 6573145.83 ng/g, and 9622555.02 ± 9213349.57 ng/g, respectively. Compared with those in the GCDCA group, the gut primary BA levels in the GCDCA + GW4064 group tended to increase. These findings indicated that GCDCA administration increased the bile content in the gut, whereas GW4064 administration potentially induces bile discharge into the gut.

First, no marked difference in the alpha diversity of the gut microbiome was found among the control, GCDCA, and GCDCA + GW4064 groups on the basis of the Chao1, Shannon, and Simpson indices (Fig. 4A; Supplementary S4). With respect to the beta diversity of the microbial community structure, a PCoA and an NMDS analysis based on Bray-Curtis distances revealed that the GCDCA group was markedly different from the control group and that the GCDCA + GW4064 and control groups had similar microbial communities (Fig. 4B).

At the phylum level, the gut bacteria in the control group, GCDCA group, and GCDCA + GW4064 group consisted of Firmicutes (53.28%, 67.99%, and 62.27%, respectively), Desulfobacterota (25.16%, 1.5%, and 13.28%, respectively), Proteobacteria (1.90%, 16.61%, and 1.17%, respectively), Patescibacteria (1.28%, 0.07%, and 0.22%, respectively), Campylobacterota (0.52%, 0.01%, and 0.0028%, respectively), Bacteroidota (13.08%, 2.55%, and 15.50%, respectively), Actinobacteriota (3.86%, 1.22%, and 6.72%, respectively), and Cyanobacteria (0.27%, 9.52%, and 0.23%, respectively) (Fig. 5A; Supplementary S5). Firmicutes was the main microbiota constituent in the three communities. The abundances of Desulfobacterota, Bacteroidota, and Actinobacteriota were lower in the GCDCA group than in the control group; however, the abundances of Proteobacteria, Cyanobacteria, and Patescibacteria were significantly greater in the GCDCA group than in the control group. Among the three groups, the microbiota structure of the GCDCA group significantly differed from that of the control group at the phylum level, and the microbiota structure of the GCDCA + GW4064 group was similar to that of the control group, as revealed by the Bray-Curtis distances (Fig. 5B and C). Taken together, these findings indicated that GW4064 administration improved the microbiota structure at the phylum level.

At the genus level, the gut bacteria in the control, GCDCA, and GCDCA + GW4064 groups included Desulfovibrio (25.16%, 1.48%, and 13.27%, respectively), Lactobacillus (24.39%, 13.43%, and 36.25%, respectively), Ligilactobacillus (5.63%, 25.93%, and 4.40%, respectively), Enterorhabdus (2.58%, 0.25%, and 3.26%, respectively), Clostridia_UCG-014_unclassified (1.47%, 0.14%, and 0.48%, respectively), Candidatus_Saccharimonas (1.21%, 0.02%, and 0.21%, respectively), Streptococcus (0.23%, 4.12%, and 0.32%, respectively), Paramuribaculum (0.70%, 0.06%, and 0.35%, respectively), Mitochondria_unclassified (0.25%, 10.55%, and 0.15%, respectively), Helicobacter (0.52%, 0.01%, and 0.0028%, respectively), Faecalibaculum (0.02%, 0.04%, and 4.36%, respectively), Ralstonia (0.16%, 1.12%, and 0.13%, respectively), Turicibacter (0.01%, 0.04%, and 0.93%, respectively), Methyloversatilis (0.04%, 0.19%, and 0.04%, respectively), Staphylococcus (0.04%, 1.49%, and 0.05%, respectively), and Pseudoalteromonas (0.06%, 0.56%, and 0.04%, respectively) (Fig. 5D; Supplementary S6). The GCDCA group exhibited markedly decreased abundances of Desulfovibrio, Lactobacillus, Enterorhabdus, Clostridia_UCG-014_unclassified, Candidatus_Saccharimonas, Paramuribaculum, and Helicobacter as well as markedly increased abundances of Ligilactobacillus, Mitochondria_unclassified, Pseudoalteromonas, Streptococcus, Staphylococcus, and Ralstonia compared with the control group. Moreover, GW4064 restored the abundance of these microbiota at the genus level. Among the three groups, the microbiota structure of the GCDCA group was markedly different from that of the control group, and the microbiota structure of the GCDCA + GW4064 group was similar to that of the control group (Fig. 5E and F). GW4064 administration mitigated the imbalance of the microbiota at the genus level.

LEfSe analysis results revealed changes in the gut microbiota structure. Ninety-two taxa (containing six grading levels) were shared between the control group and the GCDCA group (Fig. 6A), 59 taxa (containing six grading levels) were shared between the GCDCA group and the GCDCA + GW4064 group (Fig. 6B), and 24 taxa (containing six grading levels) were shared between the control group and the GCDCA + GW4064 group (Fig. 6C). On the basis of the differences in taxa at the six grading levels, the GCDCA group presented markedly altered gut microbiota structure compared with the control group, and the administration of GW4064 alleviated the gut microbiota imbalance caused by BA disorders in the gut.

To further investigate the effects of GCDCA on the intestinal mucosa, the expressions of mucin2, claudin-1, occludin, and ZO-1 were investigated. Immunohistochemistry analysis revealed that mucin2, claudin-1, occludin, and ZO-1 were downregulated in the GCDCA group, whereas GW4064 upregulated their expression (Fig. 7). Figure 3 also indicates that GW4064 activated gut FXR. These results suggested that gut FXR activation alleviated intestinal injury caused by GCDCA.

Spearman correlation analysis was conducted to determine the correlation between gut microbiota constituents that showed significant differences at the phylum and genus levels and the levels of BAs that showed significant differences in the liver and gut based on the following criteria: |r| ≥ 0.5 and P < 0.05.

At the phylum level, the abundances of Desulfobacterota and Patescibacteria were positively related to UDCA levels in the gut (r = 0.587, P = 0.027; r = 0.574, P = 0.031, respectively) (Fig. 8A). At the genus level, the abundance of Helicobacter was positively related to gut DBH levels (r = 0.673, P = 0.008) and negatively related to GCDCA (r = −0.563, P = 0.035). The abundance of Clostridia_UCG_014_unclassified was positively correlated with the gut UDCA (r = 0.642, P = 0.013) and DBH (r = 0.571, P = 0.032) levels. The abundance of Faecalibaculum was positively correlated with the gut GCDCA concentration (r = 0.638, P = 0.013). The abundance of Candidatus_Saccharimonas was positively correlated with the gut UDCA concentration (r = 0.599, P = 0.023), and the abundance of Desulfovibrio was positively correlated with the gut UDCA concentration (r = 0.587, P = 0.027). The abundance of Ligilactobacillus was negatively related to gut LCA levels (r = −0.562, P = 0.036) (Fig. 8B). These results suggested that some important gut microbes may be involved in the dysregulation of gut BA metabolism caused by GCDCA.

At the phylum level, the abundance of Proteobacteria was positively related to the levels of liver TCA (r = 0.73, P = 0.003), TCDCA (r = 0.709, P = 0.005), CA (r = 0.679, P = 0.007), TUDCA (r = 0.679, P = 0.009), beta-MCA (r = 0.67, P = 0.01), THDCA (r = 0.668, P = 0.008), CDCA (r = 0.653, P = 0.011), T-beta-MCA (r = 0.652, P = 0.013), GCA (r = 0.627, P = 0.016), and UDCA (r = 0.618, P = 0.018). The abundance of Actinobacteria was negatively correlated with the level of liver TCA (r = −0.717, P = 0.003). The abundance of Desulfobacterota was negatively correlated with the levels of liver CDCA (r = −0.552, P = 0.04) and TCA (r = −0.543, P = 0.044) (Fig. 8C).

At the genus level, the abundance of Faecalibaculum was negatively related to the levels of beta-MCA (r = −0.814, P = 0.0003), T-beta-MCA (r = −0.677, P = 0.007), TUDCA (r = −0.677, P = 0.007), THDCA (r = −0.676, P = 0.007), TCA (r = −0.658, P = 0.01), and CDCA (r = −0.598, P = 0.023) in the liver. The abundance of Enterorhabdus was negatively correlated with the liver levels of TCA (r = −0.668, P = 0.008) and GCA (r = −0.54, P = 0.046). The abundance of Ligilactobacillus was positively correlated with the levels of CA (r = 0.664, P = 0.009), TCA (r = 0.543, P = 0.044) and TCDCA (r = 0.538, P = 0.049) in the liver. The abundance of Desulfovibrio was negatively correlated with the liver levels of CDCA (r = −0.552, P = 0.04) and TCA (r = −0.543, P = 0.044) (Fig. 8D). These results suggested that some important gut microbes may be involved in the dysregulated BA metabolism in the liver caused by GCDCA.

GCDCA induced cholestasis and disturbed BA metabolism in the gut and liver, as well as the intestinal barrier and structure of the gut microbiota. Activation of gut FXR improved intestinal barrier injury and alleviated BA metabolism dysfunction and dysbacteriosis caused by GCDCA under cholestatic conditions.