Effect of fucoidan on ethanol-induced liver injury and steatosis in mice and the underlying mechanism

Abstract

Background: Alcoholic liver disease is caused as a result of chronic alcohol consumption. In this study, we used an alcoholic liver injury mouse model to investigate the effect of fucoidan on ethanol-induced liver injury and steatosis and the underlying mechanisms.

Methods: All mice were randomly divided into four groups: 1) control group, 2) model group, 3) diammonium glycyrrhizinate treatment group (200 mg/kg body weight), and 4) fucoidan treatment group (300 mg/kg body weight). Administration of ethanol for 8 weeks induced liver injury and steatosis in mice.

Results: Fucoidan treatment decreased serum alanine aminotransferase activity, serum total cholesterol levels, and hepatic triglyceride levels, and improved the morphology of hepatic cells. Fucoidan treatment upregulated the expression of AMPKα1, SIRT1, and PGC-1α and inhibited the expression of ChREBP and HNF-1α. The levels of hepatic IL-6 and IL-18 were significantly decreased in the fucoidan group. Further, the levels of cytochrome P450-2E1 (CYP2E1), glucose-regulated protein (GRP) 78, and 3-nitrotyrosine (3-NT) in hepatic tissues were reduced in the fucoidan group as compared to the model group. Fucoidan significantly reversed the reduction of ileac Farnesoid X receptor (FXR) and fibroblast growth factor 15 (FGF15) levels induced by alcohol-feeding and reduced CYP7A1 (cholesterol 7α-hydroxylase) expression and total bile acid levels in the liver tissue. In addition, fucoidan regulated the structure of gut flora, with increased abundance of Prevotella and decreased abundance of Paraprevotella and Romboutsia as detected by 16S rDNA high-throughput sequencing.

Conclusion: Fucoidan inhibited alcohol-induced steatosis and disorders of bile acid metabolism in mice through the AMPKα1/SIRT1 pathway and the gut microbiota-bile acid-liver axis and protected against alcohol-induced liver injury in vivo.

Keywords: AMPKa1/SIRT1 pathway; ethanol-induced liver injury; fucoidan; gut microbiota–bile acid–liver axis.

Fig. 1

Weight gain and levels of serum aminotransferase and hepatic inflammatory factors. (a) Weight gain curve. There were no significant differences in the initial body weight of the mice. Over a period of 8 weeks, only control group mice exhibited weight gain. At the end of the experiment, ethanol-fed (model group) mice showed a significant decrease in the final body weight as compared to control mice. (b, c) Serum ALT (b) and AST (c) levels. Model group mice exhibited significantly higher serum AST and ALT levels as compared to control group mice. Diammonium glycyrrhizinate treatment decreased the levels of serum AST and ALT, and fucoidan treatment decreased the levels of serum ALT as compared to the model group (P < 0.05). (d–f) Levels of hepatic IL-1β (d), IL-6 (e), and IL-18 (f). Hepatic IL-1β, IL-6, and IL-18 levels were higher in the model group than in the control group. Compared to the model group, treatment with diammonium glycyrrhizinate decreased the levels of hepatic IL-1β and IL-6, and treatment with fucoidan significantly decreased the levels of hepatic IL-6 and IL-18. Note: Data are represented as mean ± SD. n = 8 in each group. *P ˂ 0.05 versus control; **P ˂ 0.01 versus control; ^# P ˂ 0.05 versus model.

Fig. 2

Pathological changes in the liver. (a) Histopathological analysis of the liver. Model group mice showed an irregular arrangement of hepatocytes, extensive fat droplets, Mallory bodies, and inflammatory infiltration in the liver tissues. Mice in the DG group and fucoidan group showed reduced lipid accumulation and less inflammatory infiltration in the liver as compared to model group mice. (b) Ultrastructural analysis of the liver using transmission electron microscopy. Model group hepatocytes exhibited irregular shape, increased lipid droplets, swollen and deformed mitochondria with fuzzy mitochondrial cristae, and degranulated endoplasmic reticulum. In the DG group and fucoidan group, the hepatic cell morphology was improved, with neatly arranged endoplasmic reticulum and clear mitochondrial cristae.

Fig. 3

Serum and hepatic lipid profiles. (a) Serum lipid profiles. (b) Hepatic lipid profiles. Compared to the control group, ethanol feeding significantly increased the levels of serum TG, CHOL, LDL-CH, and TBA. The levels of hepatic TG, LDL-CH, and TBA in the model group were also higher than that in the control group. However, fucoidan treatment decreased the levels of serum CHOL and hepatic TG and TBA and increased hepatic HDL-CH levels as compared to the model group. Note: Data are represented as mean ± SD. n = 8 in each group. *P ˂ 0.05 versus control; **P ˂ 0.01 versus control; ^#P ˂ 0.05 versus model.

Fig. 4

Protein expression levels of CYP2E1, Grp78, and 3-NT in the liver. (a) Western blot analysis of the expression of CYP2E1, Grp78, and 3-NT in liver tissues. Model group mice showed significantly higher levels of CYP2E1, Grp78, and 3-NT compared to control group mice. Compared to the model group, the levels of hepatic CYP2E1, Grp78, and 3-NT were markedly downregulated in the fucoidan group. Note: *P ˂ 0.05 versus control; **P ˂ 0.01 versus control; ^#P ˂ 0.05 versus model. (b) Immunofluorescence analysis of CYP2E1. Liver tissues of fucoidan group mice displayed significantly weaker CYP2E1 staining compared to that of model group mice.

Fig. 5

Effect of fucoidan on the AMPK/SIRT1 signaling pathway. Western blot analysis showed that the expression of p-AMPKα, SIRT1, and PGC-1α was suppressed by ethanol feeding. However, treatment with fucoidan significantly reversed this effect of ethanol on hepatic p-AMPKα, SIRT1, and PGC-1α levels. In addition, ethanol-fed mice had significantly higher ChREBP and HNF-1α levels compared to control mice but treatment with fucoidan reduced these levels. Note: *P ˂ 0.05 versus control; **P ˂ 0.01 versus control; ***P ˂ 0.001 versus control; ^#P ˂ 0.05 versus model.

Fig. 6

Effect of fucoidan on fecal bile acid profiles and the bile acid–FXR–FGF15 axis. (a) Fecal bile acid analysis using LC-MS. Ethanol-fed mice showed significantly higher fecal CA and MCA levels but lower DCA, TDCA, and GDCA levels as compared to control mice. However, fucoidan treatment reduced CA and MCA levels but increased TDCA levels as compared to the model group. The bile acid profiles of the fucoidan intervention group were similar to those of the control group. (b) Western blot analysis of the expression of ileac FXR and FGF15 and hepatic CYP7A1. Ethanol-fed mice exhibited reduced levels of ileac FXR and FGF15 and increased expression of hepatic CYP7A1. However, treatment with fucoidan significantly reversed this effect of ethanol on ileac FXR and FGF15 and hepatic CYP7A1 levels. Note: *P ˂ 0.05 versus control; **P ˂ 0.01 versus control; ^#P ˂ 0.05 versus model.

Fig. 7

Diversity analysis of the gut flora of mice.(a) PCA analysis; (b) Weighted Unifrac PCOA analysis (Adonis test). The diversity analysis based on OTU abundance showed that there was a significant difference in the structure of gut flora among the three groups. C: control group; M: model group; F: fucoidan group.

Fig. 8

Composition of the gut flora of mice. (a) Analysis of the composition of gut flora at the phylum level. Alcohol feeding increased the abundances of Proteobacteria and Euryarchaeota as compared to the control group. After treatment with fucoidan, the abundance of Firmicutes was lower compared to that in the model group, and the abundance ratio of Firmicutes/Bacteroidetes was also lower compared to that in the control and model groups. (b) Analysis of the composition of gut flora at the genus level. (c) Heat map analysis of the composition of gut flora. Among the top 20 genera, ethanol feeding decreased the abundances of Prevotella, Lactobacillus, and Ruminococcus, and increased the abundances of Alloprevotella, Escherichia, Paraprevotella, Methanobrevibacter, and Romboutsia as compared to the control group. The genus distribution of fucoidan group mice was similar to that of control group mice. Treatment with fucoidan increased the abundance of Prevotella, and decreased the abundances of Paraprevotella, Romboutsia, and Clostridium sensu stricto. C: control group; M: model group; F: fucoidan group.

Fig. 9

Spearman correlation analysis between genus species and serological indicators. The abundance of Prevotella was negatively correlated with serum ALT. The abundance of Methanobrevibacter was positively correlated with AST and TG. The abundance of Clostridium sensu stricto was positively correlated with ALT, TG, and LDL-CH. In addition, the abundances of Romboutsia, Fusobacterium, and Turicibacter were positively correlated with TG. X-axis, serological indicators; Y-axis, genus species. The depth of color visually shows the correlation between genus species and serological indicators. ⁺P < 0.05; *P < 0.01.