Tauroursodeoxycholic acid inhibits intestinal inflammation and barrier disruption in mice with non-alcoholic fatty liver disease

Abstract

Background and purpose: The gut-liver axis is associated with the progression of non-alcoholic fatty liver disease (NAFLD). Targeting the gut-liver axis and bile acid-based pharmaceuticals are potential therapies for NAFLD. The effect of tauroursodeoxycholic acid (TUDCA), a candidate drug for NAFLD, on intestinal barrier function, intestinal inflammation, gut lipid transport and microbiota composition was analysed in a murine model of NAFLD.

Experimental approach: The NAFLD mouse model was established by feeding mice a high-fat diet (HFD) for 16 weeks. TUDCA was administered p.o. during the last 4 weeks. The expression levels of intestinal tight junction genes, lipid metabolic and inflammatory genes were determined by quantitative PCR. Tissue inflammation was evaluated by haematoxylin and eosin staining. The gut microbiota was analysed by 16S rRNA gene sequencing.

Key results: TUDCA administration attenuated HFD-induced hepatic steatosis, inflammatory responses, obesity and insulin resistance in mice. Moreover, TUDCA attenuated gut inflammatory responses as manifested by decreased intestinal histopathology scores and inflammatory cytokine levels. In addition, TUDCA improved intestinal barrier function by increasing levels of tight junction molecules and the solid chemical barrier. The components involved in ileum lipid transport were also reduced by TUDCA administration in HFD-fed mice. Finally, the TUDCA-treated mice showed a different gut microbiota composition compared with that in HFD-fed mice but similar to that in normal chow diet-fed mice.

Conclusions and implications: TUDCA attenuates the progression of HFD-induced NAFLD in mice by ameliorating gut inflammation, improving intestinal barrier function, decreasing intestinal fat transport and modulating intestinal microbiota composition.

Figure 1

TUDCA attenuates HFD‐induced hepatic steatosis and inflammatory responses in NAFLD mice. (A, left) Representative macroscopic images of the livers of NCD, HFD and TUDCA mice, (A, middle) Liver weight and (A, right) the ratio of the liver weight and body weight of the three groups. (B, left) Representative H&E‐treated and Oil Red O‐stained sections in liver samples and (B, right) NAS of three groups. (C) The serum and hepatic TG and TC contents of mice in the indicated groups were measured using the elisa. (D) The mRNA expression levels of genes associated with fatty acid synthesis, transport and β‐oxidation, (E) inflammatory cytokines and (F) innate immunity components in the liver were detected by qPCR. The data are presented as the means ± SEM. One‐way ANOVA followed by Newman–Keuls post hoc test for multiple comparison. NCD group, n = 8; HFD group, n = 9; and HFD + TUDCA, n = 6. *P < 0.05. Hmgcr, 3‐hydroxy‐3‐methylglutaryl CoA reductase. ND, not detected.

Figure 2

TUDCA ameliorates HFD‐induced obesity and insulin resistance in NAFLD mice. (A, left) Representative macroscopic pictures and (A, right) body weight of the NCD, HFD and HFD + TUDCA groups. (B) Fasting glucose, (C) fasting insulin and (D) HOMA‐IR in mice treated with NCD, HFD or HFD + TUDCA. Fasting glucose and fasting insulin levels were measured at the endpoint of this experiment. HOMA‐IR was calculated as HOMA‐IR = (FBG (mM) × FINS (ng·mL⁻¹))/22.5. (E) The IPGTT and (F) IPITT assays were performed to evaluate the insulin sensitivity of mice in the indicated groups treated with NCD, HFD or HFD + TUDCA. The data are presented as the mean ± SEM for (A–D) and mean ± SD for (E–F). One‐way ANOVA followed by Newman–Keuls post hoc test for multiple comparison. NCD group, n = 8; HFD group, n = 9; and HFD + TUDCA, n = 6. *P < 0.05 versus the NCD group; ^# P < 0.05 versus the HFD group.

Figure 3

TUDCA attenuates gut inflammation. (A) Representative ileum H&E staining sections (left panel) and corresponding histopathological score (right panel) in the NCD, HFD and HFD + TUDCA groups; (B, C) the mRNA expression levels of inflammatory cytokines and components of innate immune signalling were measured in the indicated groups; (D) the mRNA expression levels of inflammatory cytokines and components of innate immune signalling were measured in the indicated groups in Caco‐2 cells. CON, control group treated with vehicle; PA + LPS group, Caco‐2 cells co‐stimulated with 600 μM palmitate and 10 μg·mL⁻¹ LPS for 6 h; PA + LPS + TUDCA group, Caco‐2 cells treated with the combination of palmitate (600 mΜ), LPS (10 μg·mL⁻¹) and TUDCA (500 mM) for 6 h. The data are presented as the means ± SEM. One‐way ANOVA followed by the Newman–Keuls post hoc test for multiple comparisons. NCD group, n = 8; HFD group, n = 9; and HFD + TUDCA, n = 6. In vitro experiment, n = 5. *P < 0.05.

Figure 4

TUDCA improves intestinal barrier function. (A) The mRNA levels of ileum tight junction molecules were measured by real‐time PCR. (B, C) The mRNA levels of gut mucin lysozyme, Iap, Muc2 and C3gnt in the mice of different groups were detected by qPCR. (D) The TEERs of the CON, PA + LPS and PA + LPS + TUDCA groups were detected in Caco‐2 cells. CON, control group treated with vehicle; PA + LPS group, Caco‐2 cells co‐stimulated with 600 μM palmitate and 10 μg·mL⁻¹ LPS for the indicated times; PA + LPS + TUDCA group, Caco‐2 cells treated with the combination of palmitate (600 mΜ), LPS (10 μg·mL⁻¹) and TUDCA (500 mM) for the indicated times. The data are presented as the means ± SEM. One‐way ANOVA followed by the Newman–Keuls post hoc test for multiple comparisons. NCD group, n = 8; HFD group, n = 9; and HFD + TUDCA, n = 6. *P < 0.05; in vitro experiment, n = 5. *P < 0.05 versus the CON group; ^# P < 0.05 versus the PA + LPS group.

Figure 5

TUDCA suppresses gut lipid transport genes expression. (A) The mRNA levels of ileum lipid transport‐related genes (Cd36, Fabp, Fatp4 and Ffar3) in the NCD, HFD and HFD + TUDCA groups were measured. NCD group, n = 8; HFD group, n = 9; and HFD + TUDCA, n = 6. *P < 0.05. (B) The expression levels of ileum lipid transport‐related genes were detected in Caco‐2 cells (n = 5). CON, control group treated with vehicle; PA + LPS group, Caco‐2 cells co‐stimulated with 600 μM palmitate and 10 μg·mL⁻¹ LPS for 6 h; PA + LPS + TUDCA group, Caco‐2 cells treated with the combination of palmitate (600 mΜ), LPS (10 μg·mL⁻¹) and TUDCA (500 mM) for 6 h. The data are presented as the means ± SEM. One‐way ANOVA followed by the Newman–Keuls post hoc test for multiple comparisons. *P < 0.05.

Figure 6

TUDCA treated mice show different gut microbiota composition compared with HFD‐fed mice. (A) PCoA score plot; PCoA score plot based on unweighted (left panel) and weighted (right panel) UniFrac metrics; (B) average phylum distribution of gut microbiomes in the NCD, HFD and HFD + TUDCA groups; and (C, D) comparison of the taxonomic abundance among the indicated groups. The data are presented as the means ± SEM. The statistical significance in bacterial composition among the different samples was assessed by the ANOSIM test. NCD group, n = 8; HFD group, n = 9; and HFD + TUDCA, n = 6. *P < 0.05.