Conjugated secondary 12α-hydroxylated bile acids promote liver fibrogenesis

Abstract

Background: Significantly elevated serum and hepatic bile acid (BA) concentrations have been known to occur in patients with liver fibrosis. However, the roles of different BA species in liver fibrogenesis are not fully understood.

Methods: We quantitatively measured blood BA concentrations in nonalcoholic steatohepatitis (NASH) patients with liver fibrosis and healthy controls. We characterized BA composition in three mouse models induced by carbon tetrachloride (CCl₄), streptozotocin-high fat diet (STZ-HFD), and long term HFD, respectively. The molecular mechanisms underlying the fibrosis-promoting effects of BAs were investigated in cell line models, a 3D co-culture system, and a Tgr5 (HSC-specific) KO mouse model.

Findings: We found that a group of conjugated 12α-hydroxylated (12α-OH) BAs, such as taurodeoxycholate (TDCA) and glycodeoxycholate (GDCA), significantly increased in NASH patients and liver fibrosis mouse models. 12α-OH BAs significantly increased HSC proliferation and protein expression of fibrosis-related markers. Administration of TDCA and GDCA directly activated HSCs and promoted liver fibrogenesis in mouse models. Blockade of BA binding to TGR5 or inhibition of ERK1/2 and p38 MAPK signaling both significantly attenuated the BA-induced fibrogenesis. Liver fibrosis was attenuated in mice with Tgr5 depletion.

Interpretation: Increased hepatic concentrations of conjugated 12α-OH BAs significantly contributed to liver fibrosis via TGR5 mediated p38MAPK and ERK1/2 signaling. Strategies to antagonize TGR5 or inhibit ERK1/2 and p38 MAPK signaling may effectively prevent or reverse liver fibrosis.

Fundings: This study was supported by the National Institutes of Health/National Cancer Institute Grant 1U01CA188387-01A1, the National Key Research and Development Program of China (2017YFC0906800); the State Key Program of National Natural Science Foundation (81430062); the National Natural Science Foundation of China (81974073, 81774196), China Postdoctoral Science Foundation funded project, China (2016T90381), and E-institutes of Shanghai Municipal Education Commission, China (E03008).

Keywords: 12α-hydroxylated bile acids; ERK1/2; G protein-coupled bile acid receptor; Hepatic stellate cell; Liver fibrosis; TGR5; p38MAPK.

Fig. 1

The BA concentrations were significantly higher in the sera of NASH patients with liver fibrosis (n = 99) than in the controls (n = 99). (A) Heatmap of serum BA concentrations (scaled to 0-1) in NASH patients and controls. Red and white represent high and low BA concentration, respectively (see color scale). (B) Concentration of 12α-OH BAs, non-12α-OH BAs, and total BAs that were higher in NASH patients with fibrosis than in the controls. (C) Changes of 12α-OH BAs, non-12α-OH BAs, and total BAs in NASH patients with fibrosis vs. controls. (D) Immunofluorescence staining showed that α-SMA levels were significantly higher in the liver of fibrosis patients who had higher serum 12α-OH BAs (> 5 µg/mL, n = 15) than in those with lower serum 12α-OH BAs (< 1 µg/mL, n = 15). *p<0.05, compared to controls. (E) Masson trichrome staining showed significant higher level of fibrosis score in patients with high 12α-OH BAs than in those with low 12α-OH BAs in serum.

Fig. 2

Mice with liver fibrosis had higher concentrations of BAs in plasma and liver than control mice. (A) H&E, Sirius red staining and immunochemistry staining for α-SMA in stained liver sections from control and long term HFD-treated mice at week 82. Original magnifications X 200. (B,C) BAs levels in both plasma and liver were markedly increased in long term HFD-treated mice than in control mice. (D) H&E, Sirius red staining and immunochemistry staining for α-SMA in stained liver sections from control and STZ-HFD-treated mice at week 12. Original magnifications X 200. (E,F) BAs levels in both plasma and liver were markedly increased in STZ-HFD-treated mice than in control mice. (G) H&E, Sirius red staining and immunochemistry staining for α-SMA in stained liver sections from control and CCl₄-treated mice. Original magnifications X 200. (H,I) BAs levels in both plasma and liver were markedly increased in CCl₄-treated mice than in control mice.

Fig. 3

(A) Function of secondary BA biosynthesis and bile salt hydrolase were significantly increased in mice with liver fibrosis. (B) Abundance of BA metabolism related microbiota at the genus level in ileum of fibrosis mice. (C) Abundance of BA metabolism related microbiota at the species level in ileum of fibrosis mice. (D) Spearman correlation between BA concentration and OTUs of BA metabolism-related microbes in mice with liver fibrosis. The color of each spot in the heatmap corresponds to the R value of the spearman correlation analysis between microbial abundance and BAs concentration, and the spot with * in red color spot refers to the significant positive correlation with R>0.3 and P<0.05 while the spot with * in blue color spot refers to the significant negative correlation with R<-0.3 and P<0.05.

Fig. 4

Effects of conjugated primary 12α-OH BAs (12α-OH BA1) and conjugated secondary 12α-OH BAs (12α-OH BA2) on human hepatic stellate cells (HSC) LX-2. (A) Heatmap of gene expression level of α-SMA, TGF-β and COL I in LX-2 cells treated with different BAs (50 µM) for 24-72 hours. Relative gene expression was measured using quantitative real-time PCR (qRT-PCR), with GAPDH as housekeep gene. (B) Immunofluorescence staining for α-SMA and TGF-β in LX-2 cells treated with 50 µM 12α-OH BA1 or 12α-OH BA2 for 48 hours. The Mean Florescence Intensities (MFI) of five random microscopic views of each condition were quantified using Image J software. (C) Western-blot analysis of the protein expression level of α-SMA, TGF-β and COL I in LX-2 cells treated with 50 µM 12α-OH BA1 or 12α-OH BA2 for 48 hours. (D) Growth curves of LX-2 cells treated with vehicle control, 12α-OH BA1 or 12α-OH BA2 analyzed with a WST-8-based Cell Counting Kit. (E) Immunofluorescence staining of 3D cultured micro-tissues containing L02 and LX-2 (L02: LX-2 = 100:1) treated with 50 μM 12α-OH BA1, or 12α-OH BA2. (F) The Mean Florescence Intensities (MFI) of five random microscopic views of each condition were quantified using Image J software and presented for α-SMA, COL I, TGF-β, and TGR5 expression in α-SMA positive cells. (G) Phosphorylation of ERK1/2, p38MAPK, and JNK in LX-2 cells. Mean ± SD. * p<0.05 and ** p<0.01.

Fig. 5

TGR5 antagonist and inhibitors for ERK1/2 and p38MAPK abolished 12α-OH BA2-induced LX-2 activation. (A) Phosphorylation of ERK1/2 and p38MAPK detected by western blot in LX-2 cells treated with 12α-OH BA1 or 12α-OH BA2 (50 µM) for 5 min to 72 hours. (B) TGR5 antagonist 5β-cholanic acid (50 μM) inhibited 12α-OH BA2-induced overexpression of α-SMA and COL-1 proteins in LX-2 cells, as detected by western blot. (C) 5β-cholanic acid inhibited 12α-OH BA2-induced phosphorylation of ERK1/2 and p38/MAPK in LX-2 cells, as detected by western blot. (D,E) The pro-proliferative effect of 12α-OH BA2 in LX-2 cells was abolished by 5β-cholanic acid (D), ERK1/2 inhibitor SCH772984 (100 µM), and p38 MAPK inhibitor SB239063 (50 µM) but not by JNK inhibitor SP600125 (50 µM) (E). Cell numbers were quantified using a CCK-8 kit. (F) SCH772984 and SB239063, but not SP600125 inhibited 12α-OH BA2-induced overexpression of α-SMA, COL-I, and TGR5 in LX-2 cells, as detected using western blot. (G) SCH772984 and SB239063, but not SP600125 inhibited 12α-OH BA2-induced overexpression of α-SMA and TGF-β in the L02 and LX-2 composed 3D micro-tissues. The protein expression was detected using immunofluorescence staining. The data are presented as Mean ± SD. * p<0.05 and ** p<0.01. Treatments lasted for 48 hours if not stated otherwise.

Fig. 6

Expression of TGR5, p-p38 MAPK and p-ERK1/2 was significantly increased in STZ-HFD-treated mice and liver fibrosis patients. (A) Liver fibrosis detected by Masson Trichrome staining, and protein expression of α-SMA, TGF-β, COL-1, and PDGF in the mouse liver detected by immunohistochemistry staining. (B) Protein expression of TGR5, TGF-β, and phosphorylation of p-ERK1/2 and p-p38 MAPK, as detected by immunofluorescence staining. The MFI of 15 random microscopic views of each condition were quantified using the Image J software. (C) Immunofluorescence staining for ERK1/2 and p38MAPK phosphorylation and TGR5 expression in α-SMA-positive cells in the liver of patients with liver fibrosis and controls. The MFI of 15 random visions of each sample were quantified using Image J software. Mean±SD. ^a p<0.05 compared to normal, ^b p<0.05 compared to STZ-HFD. * p<0.05 and ** p<0.01 vs. normal. Mouse study, n=8 per group; human study, n=15 per group.

Fig. 7

TDCA and GDCA significantly promote liver fibrosis in CCl₄-induced liver fibrosis C57BL/6J model mice. (A) Bar plots of liver to body weight ratio, serum levels of ALT and AST. (B) Tgr5 expression, activation of ERK1/2 and p38MAPK in α-SMA positive cells in mice liver as determined by IHC staining (× 200). (C) Protein expression of α-SMA, TGF-β, and COL-1 in the liver detected by immunohistochemistry staining. The collagen level in mice liver as determined by Masson Trichrome staining (× 100).

Fig. 8

Depletion of Tgr5 in HSCs significantly decreased liver fibrosis in a CCl₄-induced liver fibrosis model mouse. (A) Proliferation of primary HSCs cells as evaluated using a CCK-8 kit. (B) The protein expression of α-SMA, Collagen I and Tgf-β in HSCs^{CAS9 KI} and HSCs^{CAS9 KI+Tgr5sgRNA} treated with 50 µM 12α-OH BA2 and vehicle control, a, p<0.01, two-way ANOVA. (C) The collagen level and α-SMA in mice liver as determined by Masson Trichrome and IF staining (× 100). (D) Tgr5 expression, activation of p-Erk1/2 and p-p38Mapk in α-SMA positive cells in mice liver as determined by IF staining (× 300). *, compared to CAS9 KI Corn Oil, by Student's t-test; #, compared to CAS9 KI+ CCl₄, by Student's t-test.