Fibroblast Growth Factor 19 Alters Bile Acids to Induce Dysbiosis in Mice With Alcohol-Induced Liver Disease

Abstract

Background & aims: Excessive alcohol consumption can lead to alcohol-associated liver disease, a spectrum of conditions ranging from steatosis to fibrosis and cirrhosis. Bile acids regulate metabolic pathways by binding to cellular and nuclear receptors, and they also interact with the gut microbiome to control microbial overgrowth. Fibroblast growth factor 19 (FGF-19) is an ileum-derived hormone induced and released in response to bile acid activation of the nuclear receptor farnesoid X receptor. FGF-19 signaling is dysregulated with ethanol consumption and is increased in patients with alcoholic hepatitis. Here, we examined the effects of FGF-19 in a mouse model of chronic + binge ethanol feeding.

Methods: After injection of adeno-associated virus-green fluorescent protein or AAV-FGF-19, female C57BL/6J mice were pair-fed a Lieber DeCarli liquid diet (5% v/v) or control diet for 10 days and were given a bolus gavage of 5% ethanol or maltose control to represent a binge drinking episode. Tissues were collected for analysis 9 hours after the binge.

Results: Chronic + binge ethanol feeding induced steatosis regardless of FGF-19 expression. Interestingly, FGF-19 and ethanol resulted in significantly increased liver inflammation, as measured by Il6, Tgfβ, and Tnfα, compared with ethanol alone. Both ethanol and FGF-19 decreased bile acid synthesis, and FGF-19 significantly reduced secondary bile acids, leading to overgrowth of specific pathogenic bacteria including Enterococcus faecalis, Escherichia coli, and Clostridium perfringens.

Conclusions: Dysregulation of FGF-19 and consequent changes in bile acid synthesis and composition during alcohol consumption may be a contributing factor to alcohol-induced liver disease and dysbiosis.

Keywords: Ethanol; Inflammation; Microbiome.

Graphical abstract

Figure 1

FGF-19–AAV in ethanol-fed mice. (A) Serum FGF-19 quantified in control (GFP) and FGF-19–AAV–injected mice. (B) Ileal mRNA expression of human FGF-19. (C) Ileal mRNA expression of mouse Fgf-15. (D) Hepatic mRNA expression of Fgfr4. (E) Liver and adipose weight normalized to body weight. (F) mRNA gene expression in white adipose tissue. (G) Serum ALT and AST. n = 12 mice for all assays. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. ALT, alanine aminotransferase; AST, aspartate aminotransferase; Ctl, control; Exp., expression; Rel., relative; WAT, white adipose tissue; Wt, weight.

Figure 2

Hepatic lipid metabolism in FGF-19–AAV mice. (A) Hepatic triglycerides, serum triglycerides, hepatic cholesterol, and serum cholesterol. (B) Hepatic H&E staining, arrows indicate macrovesicular steatosis. (C) Hepatic Oil Red O (ORO) staining. (D) Hepatic BODIPY staining with nuclear detection in blue (DAPI). (E) Hepatic mRNA expression of genes involved in lipid homeostasis. (F) Hepatic protein expression. n = 12 mice for all assays, n = 6 mice for histology, and n = 3–4 mice for Western blot. White scale bar: 100 μmol/L; black scale bar: 200 μmol/L. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. Chol, cholesterol; Ctl, control; DAPI, 4′,6-diamidino-2-phenylindole; Exp., expression; Gapdh, glyceraldehyde-3-phosphate dehydrogenase; P-Acc, phospho-Acc; Rel., relative; T-Acc, total Acc; TG, triglycerides.

Figure 3

Bile acid homeostasis is altered by FGF-19–AAV and ethanol. (A) Serum bile acids. (B) Liver bile acids. (C) Gallbladder bile acids. (D) Intestine bile acids. (E) Bile acid pool. (F) Serum 7α-hydroxy-4-cholesten-3-one (C4). (G) Hepatic mRNA expression of bile acid synthesis genes. (H) Hepatic mRNA expression of genes that regulate bile acid homeostasis. (C) n = 12 mice for all assays. Data were analyzed by 2-way analysis of variance. ∗∗P < .01, and ∗∗∗P < .001. Ctl, control; BA, bile acid; Exp., expression; Rel., relative.

Figure 4

Serum bile acid composition is altered by FGF-19–AAV. (A) Unconjugated bile acids. (B) Taurine-conjugated bile acids. (C) Bile acid hydrophobicity. (D) Total primary and secondary bile acids. (E) Total unconjugated and conjugated bile acids. n = 11 mice for all assays. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. BA, bile acid; Ctl, control; DCA, deoxycholic acid; UDCA, ursodeoxycholic acid.

Figure 5

Hepatic inflammation is increased by FGF-19–AAV. (A) Hepatic mRNA expression. (B) Hepatic F4/80 staining. (C) Hepatic CD-68 staining. (D) Hepatic cytokeratin-19 staining. n = 12 mice for all assays; n = 6 mice for histology. White scale bar: 100 μmol/L; black scale bar: 200 μmol/L. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. Avg., average; Ctl, control; Exp., expression; Rel., relative.

Figure 6

Gut inflammation is increased by FGF-19–AAV. (A) Serum LPS. (B) Ileal mRNA expression. (C) Colon mRNA expression. (D) Colon protein expression of tight junction proteins. n = 12 mice for all assays; n = 3–4 mice for Western blot. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. Ctl, control; Exp., expression; Jam-A, junctional adhesion molecule A; Rel., relative.

Figure 7

FGF-19–AAV reduces microbial diversity and alters the cecal microbiome. (A) Phylogenetic diversity is reduced by FGF-19–AAV and ethanol. (B) Generalized UniFrac principal component analysis (PCA). (C) Phylum-level changes induced by ethanol and FGF-19–AAV. (D) Quantification of bacterial genera expressing bile salt hydrolase. (E) Quantification of bacterial families expressing 7α-dehydroxylase. n = 6 mice. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. Ctl, control.

Figure 8

FGF-19–AAV increases pathogenic bacterial abundance in cecum. (A) Percentage abundance of the pathogenic genera Enterococcus and Clostridium. (B) Percentage abundance of pathogenic microbial species. (C) Percentage abundance of beneficial microbial species. n = 6 mice. Data were analyzed by 2-way analysis of variance. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. Ctl, control.