Bile acids attenuate hepatic inflammation during ischemia/reperfusion injury

Abstract

Background & aims: Persistent cholestasis has been associated with poor prognosis after orthotopic liver transplantation. In this study, we aimed to investigate how the accumulation of tauro-beta-muricholic acid (TβMCA), resulting from the reprogramming of bile acid (BA) metabolism during liver ischemia/reperfusion (IR) stress, attenuates liver inflammation.

Methods: Ingenuity Pathway Analysis was performed using transcriptome data from a murine hepatic IR model. Three different models of hepatic IR (liver warm IR, bile duct separation-IR, common bile duct ligation-IR) were employed. We generated adeno-associated virus-transfected mice and CD11b-DTR mice to assess the role of BAs in regulating the myeloid S1PR2-GSDMD axis. Hepatic BA levels were analyzed using targeted metabolomics. Finally, the correlation between the reprogramming of BA metabolism and hepatic S1PR2 levels was validated through RNA-seq of human liver transplant biopsies.

Results: We found that BA metabolism underwent reprogramming in murine hepatocytes under IR stress, leading to increased synthesis of TβMCA, catalyzed by the enzyme CYP2C70. The levels of hepatic TβMCA were negatively correlated with the severity of hepatic inflammation, as indicated by the serum IL-1β levels. Inhibition of hepatic CYP2C70 resulted in reduced TβMCA production, which subsequently increased serum IL-1β levels and exacerbated IR injury. Moreover, our findings suggested that TβMCA could inhibit canonical inflammasome activation in macrophages and attenuate inflammatory responses in a myeloid-specific S1PR2-GSDMD-dependent manner. Additionally, Gly-βMCA, a derivative of TβMCA, could effectively attenuate inflammatory injury in vivo and inhibit human macrophage pyroptosis in vitro.

Conclusions: IR stress orchestrates hepatic BA metabolism to generate TβMCA, which attenuates hepatic inflammatory injury by inhibiting the myeloid S1PR2-GSDMD axis. Bile acids have immunomodulatory functions in liver reperfusion injury that may guide therapeutic strategies.

Impact and implications: Our research reveals that liver ischemia-reperfusion stress triggers reprogramming of bile acid metabolism. This functions as an adaptive mechanism to mitigate inflammatory injury by regulating the S1PR2-GSDMD axis, thereby controlling the release of IL-1β from macrophages. Our results highlight the crucial role of bile acids in regulating hepatocyte-immune cell crosstalk, which demonstrates an immunomodulatory function in liver reperfusion injury that may guide therapeutic strategies targeting bile acids and their receptors.

Keywords: CYP2C70; GSDMD; Liver transplantation; Macrophages; Pyroptosis; S1PR2.

Graphical abstract

Fig. 1

IR orchestrates BA metabolism and participates in inflammatory injury. Sham and IR-stressed (ischemia for 1 h, reperfusion for 6 h) mouse liver lobes were subjected to transcriptome sequencing, followed by IPA (n = 3/group, DEG cut-offs: |log2FoldChange| >0.8, p-adj <0.1). (A) DEGs were used to perform canonical pathway analysis (the top 15 pathways of -log(p-value) are shown). (B) Correlation analysis between different molecules and enrichment pathways. (C) Upstream regulator prediction (the top 20 regulators of the activation z-score are shown). p values were calculated using Fisher’s exact test. (D) Expression bubble map of genes related to BA secretion and BA biosynthesis in sham and IR-stressed livers, value is the normalized FPKM data, S: sham; I: IR. (E) qRT-PCR analysis of BA synthesis-related gene expression in sham and IR-stressed livers (I1R0: ischemia for 1 h; I1R1: ischemia for 1 h, reperfusion for 1 h; I1R3: ischemia for 1 h, reperfusion for 3 h; I1R6: ischemia for 1 h, reperfusion for 6 h, n = 6/group). (F) qRT-PCR analysis of BA transporter expression in sham and IR-stressed livers (n = 6/group). (G) Western blot-assisted detection of CYP7A1, CYP8B1, CYP7B1, CYP27A1, and OST-β in sham and IR-stressed livers. (H-I) Bile ducts were intubated and draining bile was collected. Representative images of biliary drainage. The volume of bile was measured at different time points and the amount of bile was calculated (n = 4/group). (J) C57BL/6 mice were pretreated with cholestyramine (2% or 5%, w/w) or a normal diet before IR surgery, and liver/serum was collected after IR stress (n = 5/group). (K) BA levels in the liver. (L) Serum ALT/AST levels. (M) Representative H&E staining of liver sections. Percentages of necrotic areas were quantified (scale bar, 100 μm). (N) qRT-PCR analysis of pro-inflammatory gene expression in livers. Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by Student’s t test or one-way ANOVA analysis. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BA, bile acid; Bsep (Abcb11), bile salt export pump; CHO, cholestyramine; Cyp7a1, cholesterol 7α-hydroxylase; Cyp8b1, sterol 12α-hydroxylase; Cyp7b1, oxysterol 7α-hydroxylase; Cyp27a1, sterol 27-hydroxylase; Cyp2b10, cytochrome P450, family 2, subfamily b, polypeptide 10; Cyp2c70, cytochrome P450, family 2, subfamily c, polypeptide 70; IRI, ischemia-reperfusion injury; Mrp2-4, multidrug resistance associated protein 2-4; N, necrotic area; Ntcp (Slc10a1), sodium/taurocholate cotransporting polypeptide, solute carrier family 10, member 1; Oatp1b1, solute carrier organic anion transporter family, member 1B1; Ost-α/β, organic solute transporter alpha/beta; qRT-PCR, quantitative reverse-transcription PCR.

Fig. 2

Hepatic BAs attenuate murine IRI and inhibit IL-1β release. (A) C57BL/6 mice were subjected to IR or CBDL-IR stress, followed by liver/serum sampling after 6 h of reperfusion. (B) Serum ALT/AST levels (n = 6-8/group). (C) Liver tissue appearance and representative H&E staining of IR-stressed or CBDL-IR-stressed livers (n = 6/group). Percentages of necrotic areas in H&E staining were quantified (scale bar, 100 μm). (D) C57BL/6 mice were subjected to IR or BDS-IR stress, followed by liver/serum sampling after 6 h of reperfusion. Microscopic views before and after 1 h of ischemia. (E) Serum BA levels (n = 3-6/group). (F) Liver BA levels (n = 7-10/group). (G) Serum ALT/AST levels (n = 7/group). (H) Representative H&E and immunohistochemical staining of Ly6C⁺ or CLEC4F⁺ cells in liver sections (n = 7/group). Percentages of necrotic areas and the number of Ly6C⁺ or CLEC4F⁺ cells were quantified (scale bar, 100 μm, and 50 μm). (I) qRT-PCR analysis of pro-inflammatory gene expression in the livers (n = 7/group). (J) Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT. (K) Serum IL-1β levels (n = 10/group). (L) C57BL/6 mice were pretreated with 2% cholestyramine or a normal diet, followed by BDS-IR stress. Serum ALT/AST levels were measured (n = 4/group). (M) Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT in the livers. (N) Serum IL-1β levels (n = 4/group). (O) BA-targeted metabolomic analysis in sham, BDS-IR-stressed, and IR-stressed livers (n = 4-10/group). Total liver BA levels were measured. (P) Correlations between total liver BAs and serum ALT/AST levels were analyzed by non-parametric Spearman’s method (n = 16). Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by Student’s t test or one-way ANOVA analysis. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BDS-IR, bile duct separation with IR; CBDL, common bile duct ligation; CHO, cholestyramine; IRI, ischemia-reperfusion injury; MoMFs, monocyte-derived macrophages; N, necrotic area; qRT-PCR, quantitative reverse-transcription PCR; TBA, total bile acid.

Fig. 3

IR induces metabolic pathway shifts to promote TβMCA production. C57BL/6 mice were subjected to IR or BDS-IR stress, and liver samples were obtained at 6 h of reperfusion for BA-targeted metabolomic analysis (n = 4-10/group). (A) Concentrations of FBAs and CBAs. (B) Concentrations of various CBAs. (C) Concentrations of MCAs. (D) Proportions and concentrations of different MCAs. (E) BA synthesis pathway in mice. (F) Combined transcriptomic and metabolomic analyses of the sham and IR-stressed livers. Significantly altered gene expression and bile acid levels after IR stress are shown. (G) Hepatic CDCA/CA ratio in sham, BDS-IR-stressed and IR-stressed livers. (H) C57BL/6 mice were subjected to BDS-IR stress, and liver samples were obtained at 0-6 h of reperfusion. Western blot-assisted detection and relative density ratio of CYP7A1, CYP8B1, CYP7B1, CYP27A1, and OST-β in the sham and BDS-IR-stressed livers. (I) qRT-PCR analysis of BA synthesis-related gene expression in sham, BDS-IR-stressed, and IR-stressed livers (n = 5/group). (J) HPCs and NPCs were isolated from BDS-IR-stressed and IR-stressed livers, qRT-PCR analysis of Cyp2c70 levels (n = 6/group). Western blot-assisted detection and relative density ratio of CYP2C70. (K, L) Correlations between hepatic TβMCA levels and serum ALT/AST levels, as well as Il-1β, Cxcl1, Cxcl2, Ccl2 and Tnf-α levels were analyzed by non-parametric Spearman’s method (n = 16). Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by one-way ANOVA analysis. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BAs, bile acids; BDS-IR, bile duct separation with IR; CBAs, conjugated bile acids; FBAs, free bile acids; HPCs, hepatic primary cells; IR, ischemia-reperfusion; LC-MS, liquid chromatography-mass spectrometry; MCAs, muricholic acids; NPCs, non-parenchymal cells; qRT-PCR, quantitative reverse-transcription PCR.

Fig. 4

Inhibition of hepatic Cyp2c70 exacerbates IR inflammatory injury. (A) AAV8-Cyp2c70-shRNA was used to generate CYP2C70 knockdown mice (AAV-shCyp2c70). AAV8-Cyp2c70-shRNA was injected into 6-week-old mice for 4 weeks, followed by liver/serum sampling. (B) Western blot-assisted detection and relative density ratio of CYP2C70 in sham-operated AAV8-control and AAV8-shCyp2c70 mice. (C) Body weight measurements (n = 6/group). (D) Ratio of the liver-to-body weight (n = 6/group). (E) Serum ALT/AST levels (n = 4/group). (F) Representative H&E staining in liver sections. (G) qRT-PCR analysis of Il-1β, Tnf-α, Cxcl1, Cxcl2, and Ccl2 levels in the livers (n = 4/group). (H) AAV8-control and AAV8-shCyp2c70 mice were subjected to IR stress, followed by liver/serum sampling after 6 h of reperfusion (n = 6/group). Western blot-assisted detection and relative density ratio of CYP2C70 in IR-stressed AAV8-control and AAV8-shCyp2c70 mice. (I) Serum ALT/AST levels. (J) Representative H&E staining of liver sections (n = 6/group). Percentages of necrotic areas were quantified (scale bar, 100 μm). (K) qRT-PCR analysis of Il-1β, Tnf-α, Cxcl1, Cxcl2, and Ccl2 levels in the livers (n = 6/group). (L) AAV-control and AAV-shCyp2c70 mice were subjected to IR stress, followed by liver/serum sampling after 6 h of reperfusion. Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT in the livers. (M) Serum IL-1β levels (n = 4-6/group). (N) BA-targeted metabolomic analysis in sham or IR-stressed livers of AAV-control/AAV-shCyp2c70 mice (n = 4/group). Global sample distribution profiles were analyzed by the PCA. (O) Abundance of TβMCA in each group. (P-Q) Abundance of CDCA and its derivatives in sham and IR groups. (R) IR and BDS-IR models were established in AAV-control/AAV-shCyp2c70 mice, followed by liver/serum sampling after 6 h of reperfusion (n = 6/group). Serum ALT/AST levels. (S) Serum IL-1β levels. Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by Student’s t test. AAV, adeno-associated virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BA, bile acid; BDS-IR, bile duct separation with IR; IR, ischemia-reperfusion; LC-MS, liquid chromatography-mass spectrometry; N, necrotic area; PCA, principal component analysis; qRT-PCR, quantitative reverse-transcription PCR.

Fig. 5

TβMCA inhibits BA receptor S1PR2 in Ly6C⁺ macrophages. C57BL/6 mice were subjected to IR or BDS-IR stress, followed by liver/serum sampling after 6 h of reperfusion. (A) Expression of different BA receptors including S1pr2, Vdr, Car, Pxr, Rorγt, Chrm2, Lxrα and Lxrβ were detected by qRT-PCR (n = 5/group). (B) Western blot-assisted detection and relative density ratio of S1PR2 in the sham, BDS-IR-stressed, and IR-stressed livers. (C) Western blot-assisted detection and relative density ratio of S1PR2 in sham or IR-stressed livers of AAV-control/AAV-shCyp2c70 mice. (D) C57BL/6 mice were treated with JTE-013 (0.1 mg/kg, or 1 mg/kg) or vehicle 1 h before ischemia (pre-ischemia) or at the beginning of reperfusion (pre-reperfusion), followed by liver/serum sampling after 6 h of reperfusion. (E-F) Serum ALT/AST and IL-1β levels (n = 3/group). (G) Representative H&E staining of liver sections (n = 3/group). Percentages of necrotic areas were quantified (scale bar, 100 μm). Vehicle: injection of vehicle at the beginning of reperfusion. (H) qRT-PCR analysis of S1pr2 levels in HPCs and NPCs isolated from sham, BDS-IR-stressed, and IR-stressed livers (n = 5/group). (I) Representative immunofluorescence staining of CLEC4F (green), S1PR2 (red), merged, and enlarged images of IR-stressed or BDS-IR-stressed livers (scale bar, 50 and 20 μm). (J) Representative immunofluorescence staining of Ly6C (green), S1PR2 (red), merged, and enlarged images of IR-stressed or BDS-IR-stressed livers (scale bar, 50 and 20 μm). (K) The percentages of S1PR2⁺ CLEC4F⁺ cells were quantified (n = 4/group). (L) The percentages of S1PR2⁺ Ly6C⁺ cells were quantified (n =4/group). Scale bars, 50 and 20 μm. (M) Mouse BMDMs were treated with CA, CDCA, TCA, TCDCA, TDCA, UDCA, αMCA, βMCA, TβMCA, ω-MCA (50 μM), or DMSO for 3 h. qRT-PCR analysis of S1pr2 levels in cell lysates (n =3/group). (N) BMDMs were treated with LPS (0.1 μg/ml) and TβMCA (0, 50, 100, 150 or 200 μM) for 3 h, followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of S1PR2, AKT1, p-AKT1, ERK1/2, and p-ERK1/2 in cell lysates. Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by Student’s t test or one-way ANOVA analysis. AAV, adeno-associated virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BAs, bile acids; BARs, bile acid receptors; BDS-IR, bile duct separation with IR; BMDMs, bone marrow-derived macrophages; HPCs, hepatic primary cells; IR, ischemia-reperfusion; KCs, Kupffer cells; LPS, lipopolysaccharide; LSECs, liver sinusoidal endothelial cells; MoMFs, monocyte-derived macrophages. N, necrotic area; NPCs, non-parenchymal cells; qRT-PCR, quantitative reverse-transcription PCR.

Fig. 6

TβMCA inhibits myeloid macrophage-specific S1PR2-GSDMD-mediated IL-1β release. (A) BMDMs were treated with LPS (0.1 μg/ml) and TβMCA (0, 50, 100, 150 or 200 μM) for 3 h, followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT, Pro-IL-1β and caspase1/p-20-caspase1 in cell lysates. (B) IL-1β (n = 6/group) and LDH levels (n = 4/group) in the supernatants. (C) BMDMs were treated with LPS (0.1 μg/ml) + JTE-013 (5 mM) or LPS (0.1 μg/ml) +CYM-5520 (5 mM) for 3 h, followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT, Pro-IL-1β and caspase1/p-20-caspase1 in cell lysates. (D) IL-1β and LDH levels in the supernatants (n = 3/group). (E) Molecular docking of TβMCA and JTE-013 with S1PR2 was performed using CB-Dock2. The Vina score, location of the binding center, docking size, and interaction residues were evaluated. (F) siControl- or siS1pr2-transfected BMDMs were treated with LPS and TβMCA (100 μM and 150 μM) for 3 h, followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT and S1PR2 in cell lysates. (G) IL-1β and LDH levels in supernatants (n = 3/group). (H) BMDMs were treated with LPS (0.1 μg/ml) + JTE-013 (5 mM), LPS (0.1 μg/ml) +CYM-5520 (5 mM), or LPS (0.1 μg/ml) +DMSO for 3 h, followed by ATP treatment (5 mM, 0.5 h). The cell lysates were collected for transcriptomic sequencing and IPA analysis (n = 4/group). (I) DT-pretreated CD11b-DTR mice infused with siControl or siS1pr2 BMDMs were subjected to IR stress, followed by liver/serum sampling after 6 h of reperfusion (n = 5/group). (J) Serum ALT/AST levels. (K) qRT-PCR-assisted detection of hepatic S1pr2. (L) Representative H&E staining of liver sections. Percentages of necrotic areas were quantified (scale bar, 100 μm, 40 μm). (M) qRT-PCR analysis of Il-1β, Tnf-α, Cxcl1, Cxcl2, and Ccl2 levels in the livers. (N) Western blot-assisted detection and relative density ratio of hepatic GSDMD/GSDMD-NT and S1PR2. (O) Serum IL-1β levels. Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by Student’s t test or one-way ANOVA analysis. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BAs, bile acids; BMDMs, bone marrow-derived macrophages; DT, diphtheria toxin; IR, ischemia-reperfusion; LDH, lactate dehydrogenase; LPS, lipopolysaccharide; MoMFs, monocyte-derived macrophages; NPCs, non-parenchymal cells; N, necrotic area; qRT-PCR, quantitative reverse-transcription PCR.

Fig. 7

Gly-βMCA attenuates IR inflammatory injury and inhibits human macrophage pyroptosis. (A) 2D molecular structure of TβMCA and Gly-βMCA and molecular docking results of Gly-βMCA with S1PR2. (B) BMDMs were treated with LPS (0.1 μg/ml) and Gly-βMCA (0, 50, 100, 150 or 200 μM) for 3 h, followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT, Pro-IL-1β, caspase-1/p-20-caspase-1, S1PR2, AKT1, p-AKT1, ERK1/2, and p-ERK1/2 in cell lysates. (C) IL-1β and LDH levels in the supernatants (n = 4/group). (D) THP-1 cells were activated with 50 nM PMA for 48 h and then treated with LPS (1 μg/ml) and Gly-βMCA for 4 h, followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT, pro-IL-1β, caspase-1/p-20-caspase-1, and S1PR2 in cell lysates. (E) IL-1β and LDH levels in supernatants (n = 3/group). (F) siControl- or siS1pr2-transfected BMDMs were treated with LPS and Gly-βMCA (100 and 150 μM), followed by ATP treatment (5 mM, 0.5 h). Western blot-assisted detection and relative density ratio of GSDMD/GSDMD-NT and S1PR2 in cell lysates. (G) IL-1β and LDH levels in supernatants (n = 3/group). (H) WT mice were treated with vehicle or different doses of Gly-βMCA (2.5 mg/kg, 5 mg/kg or 10 mg/kg i.v.) 1 h before IR, followed by liver/serum sampling after 6 h of reperfusion (n = 4-5/group). (I) Serum ALT/AST levels. (J) qRT-PCR analysis of Il-1β, Tnf-α, Cxcl1, Cxcl2, and Ccl2 levels in the livers. (K) Representative H&E staining (scale bar, 100 μm) and immunofluorescence staining of Ly6C (green), S1PR2 (red), and merged images in liver sections (original scale bar, 400 μm; magnification scale bar, 50 μm). Percentages of necrotic areas and the intensity of S1PR2 staining were quantified. (L) Western blot-assisted detection and relative density ratio of hepatic GSDMD/GSDMD-NT and S1PR2. (M) Serum IL-1β levels. Data are shown as mean±SEM. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 by Student’s t test or one-way ANOVA analysis. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BAs, bile acids; BMDMs, bone marrow-derived macrophages; Gly-βMCA, Glycine-beta-muricholic; IPA, Ingenuity Pathway Analysis; IR, ischemia-reperfusion; LDH, lactate dehydrogenase; LPS, lipopolysaccharide; MASH, metabolic dysfunction-associated steatohepatitis; MoMFs, monocyte-derived macrophages; N, necrotic area; OLT, orthotopic liver transplantation; PMA, phorbol 12-myristate 13-acetate; qRT-PCR, quantitative reverse-transcription PCR.