Society: AASLD

Background: Chronic cholestatic liver diseases, such as primary sclerosing cholangitis (PSC), often lead to end-stage liver disease requiring liver transplantation. PSC also significantly increases the risk of cholangiocarcinoma (CCA). There is currently no effective therapeutic agent for PSC due to a limited understanding of its pathogenesis. Exosomes carry a variety of bioactive molecules, including miRNAs, lipids and proteins. The dysregulation of miRNAs is closely related to various pathological aspects of PSC. In this study, we aimed to investigate the expression profiles and function of exosomal miRNAs under cholestatic conditions.
Methods: One-year-old Mdr2^-/- mice (FVB background) and age-/gender- matched wild-type (WT) mice were used in this study. The serum exosomes were isolated using qEV isolation system (from IZON). Total RNA was isolated from the serum exosomes and liver tissues using Trizol. Total RNA transcriptome was determined by RNAseq. The miRNA profiles in the exosomes and livers were analyzed using the NanoString nCounter® miRNA Expression panel. TargetScan and mirnet2.0 were used to identify the potential target genes. Protein-protein interaction (PPI) was constructed based on the STRING database. Hub genes under cholestatic conditions were visualized by Cytoscape software. The human CCA patient data were downloaded from the Cancer Genome Atlas (TCGA) database.
Results: Bioinformatic analysis identified 74 differentially expressed miRNAs (DE-miRs) and 137 DE-miRs in the male and female Mdr2^-/- mice compared to corresponding WT controls, respectively. The identified DE-miRs were used to predict the target genes. GO and KEGG analysis further showed that these DE-miRs were linked to multiple signaling pathways related to cholestatic liver injury. We also identified 431 up-, and 1106 down- DE genes (DEGs) in the Mdr2^-/- mice compared to WT mice. From the PPI network analysis of the DEGs, the top 100 hug genes were determined. We identified 11 upregulated DE-miRs in Mdr2^-/- mice, which overlapped with the predicted miRs based on the the top 100 hug genes. Interestingly, among the overlapping DE-miRs, two miRs, miR-16 and let-7e, were also upregulated in the serum exosomes of Mdr2^-/- mice. Among the 100 hug genes, we identified nine downregulated genes that were potential targets of miR-16 and let-7e. Further analysis of TCGA CCA data showed upregulation of miR-16 and let-7e, and 7 out of the 9 identified targeted genes were also significantly downregulated in CCA compared to healthy controls.
Conclusion: This study identified potential DE-miRs in the liver and serum exosomes which may contribute to cholestatic liver injury. This study supports the possible use of exosomal miRs as diagnostic and prognostic markers and therapeutic targets for cholestatic liver diseases.

<b>Fig 1. Liver DE-miRs and DEGs identification in Mdr2<sup>-/-</sup> mice compared to corresponding WT controls, target prediction and analysis of these DE-miRs and DEGs, and the overlapping miRNAs in cholestatic liver injury.</b><br /> (A)<b> </b>Liver DE-miRs expression profiles between male Mdr2<sup>-/-</sup> and WT controls. (B) Liver DE-miRs expression profiles between female Mdr2<sup>-/- </sup>and WT controls. (C)<b> </b>Liver DEGs expression profiles between Mdr2<sup>-/-</sup> and WT controls. (D) The top 50 hug genes of the PPI network of upregulated DEGs and the top 50 hug genes of the PPI network of downregulated DEGs. (E) Predicted network of miRNAs – 100 identified hug genes of DEGs between Mdr2<sup>-/- </sup>and WT controls. (F) Identification of 11 overlapping upregulated miRNAs in cholestatic liver injury. FC, fold change; miR, micro RNA. DE, differentially expressed, Mdr2<sup>-/-</sup> , multidrug resistance 2 gene knockout; WT, wild type.

Fig 1. Liver DE-miRs and DEGs identification in Mdr2^-/- mice compared to corresponding WT controls, target prediction and analysis of these DE-miRs and DEGs, and the overlapping miRNAs in cholestatic liver injury.
(A) Liver DE-miRs expression profiles between male Mdr2^-/- and WT controls. (B) Liver DE-miRs expression profiles between female Mdr2^-/-and WT controls. (C) Liver DEGs expression profiles between Mdr2^-/- and WT controls. (D) The top 50 hug genes of the PPI network of upregulated DEGs and the top 50 hug genes of the PPI network of downregulated DEGs. (E) Predicted network of miRNAs – 100 identified hug genes of DEGs between Mdr2^-/-and WT controls. (F) Identification of 11 overlapping upregulated miRNAs in cholestatic liver injury. FC, fold change; miR, micro RNA. DE, differentially expressed, Mdr2^-/- , multidrug resistance 2 gene knockout; WT, wild type.

<b>Fig 2. Exosome DE-miRs identification in Mdr2<sup>-/- </sup>mice compared to corresponding WT controls, the overlapping DE-miRs in the liver and serum exosomes, and comparison of the expression levels of these DE-miRs and target genes between normal and CHOL samples in TCGA CCA database.</b><br /> (A) Exosome DE-miRs expression profiles between female Mdr2<sup>-/-</sup> and WT controls. (B) Identification of 2 overlapping upregulated DE-miRs in the serum exosomes which may contribute to cholestatic liver injury. (C) The expression levels of these 2 overlapping upregulated DE-miRs between normal and CHOL samples. (D) Heat map, the expression levels of these target hug genes of 2 overlapping upregulated DE-miRs between normal and CHOL samples.

Fig 2. Exosome DE-miRs identification in Mdr2^-/-mice compared to corresponding WT controls, the overlapping DE-miRs in the liver and serum exosomes, and comparison of the expression levels of these DE-miRs and target genes between normal and CHOL samples in TCGA CCA database.
(A) Exosome DE-miRs expression profiles between female Mdr2^-/- and WT controls. (B) Identification of 2 overlapping upregulated DE-miRs in the serum exosomes which may contribute to cholestatic liver injury. (C) The expression levels of these 2 overlapping upregulated DE-miRs between normal and CHOL samples. (D) Heat map, the expression levels of these target hug genes of 2 overlapping upregulated DE-miRs between normal and CHOL samples.

Background: Following biliary injury/senescence in rodent models of cholestatic liver injury and in patients with PSC, mast cells (MCs) infiltrate the liver and migrate toward cholangiocytes. Proteases such as chymase and tryptase are indicative of MC origination and function, while activated MCs release histamine (HA) that promotes liver inflammation. In human PSC, chymase gene expression is significantly higher than tryptase in these patients. MC chymase induces detrimental effects in cardiac dysfunction, asthma, and lung fibrosis. In vivo, inhibition of chymase reverses acute liver failure and CCl₄-induced damage. Similarly, in a PSC mouse model, Mdr2^-/-, has increased chymase- and tryptase-positive MC presence at baseline, but after treatment with cromolyn sodium (to block MC activation), these mice display reduced biliary damage/senescence, hepatic fibrosis, and inflammation. Aim: To determine the effects of INVA8001, a specific chymase inhibitor, on PSC phenotypes in Mdr2^-/- mice. Methods: 10 wk male Mdr2^-/- mice were treated with INVA8001 (20 mg/kg BW/day) for 14 days by daily IP injection. Serum, liver, cholangiocytes, and cholangiocyte supernatants were collected. FVB/NJ (wild-type, WT) and untreated Mdr2^-/- mice were utilized for comparison. Liver damage was assessed by H&E staining in liver and total bile acids (TBAs) were measured in serum by EIA. Changes in intrahepatic bile duct mass (i.e., ductular reaction, DR) were determined by semiquantitative CK-19 immunohistochemistry (IHC) and biliary senescence by immunofluorescence (IF) for p16 (costained with CK-19) and qPCR for p16, p21, and p18 in total liver mRNA. Hepatic fibrosis and collagen deposition were evaluated by quantitative Fast Green/Sirius Red staining and qPCR for collagen type 1a and TIMP1 in total liver samples. MC presence was determined by staining for mouse MC protease-1 (marks chymase-positive MCs) and tryptase beta 2 (marks tryptase-positive MCs) and activation by HA serum content and qPCR for mMCP-1 and FCeR1 in total liver. Changes in senescence-associated secretory phenotypes (SASPs) were determined by serum TGF-β1 content using EIA and total liver expression by qPCR. Results: Inhibition with INVA8001 ameliorated DR, hepatic damage, senescence/SASP, and fibrosis along with chymase-positive MC infiltration and activation in Mdr2^-/- mice compared to WT and untreated Mdr2^-/- mice. Conclusion: Specific inhibition of chymase targets PSC phenotypes ameliorating pathogenesis by decreased biliary senescence and SASP factors. Our studies demonstrate the damaging effects of chymase-positive MCs in PSC and the therapeutic benefit of INVA8001 treatment with potential applicability to other liver diseases wherein MCs promote pathogenesis.

Fibrosis development is associated with poor prognosis in the patient with chronic liver disease. The formation of bridging fibrosis (BF) is a critical step for developing advanced fibrosis and further cirrhosis. Preventing BF formation could be a strategy for inhibiting further advancement of fibrosis to cirrhosis. For that, underpinning the mechanisms of BF formation is to crucial. We previously showed Wilms tumor 1 (WT1) is a transcriptional regulator of profibrogenic HAS2 in liver fibrosis. Here, we further identified a novel profibrogenic function of WT1 in hepatic stellate cell (HSC) activation and invasion that is associated with BF formation in liver fibrosis.
First, we identified WT1 expression is upregulated in HSCs of BF area in advanced fibrosis in NAFLD patients, which is determined by our new experiment by combining scRNA-seq and spatial transcriptome analysis. Because HSC invasion to liver parenchyma and collagen production is critical for BF formation, we investigated the molecular mechanisms of HSC invasion using a collagen hydrogel invasion model in vitro. We performed RNA-seq for non-invaded HSCs and HSCs during and after invasion by PDGF treatment in vitro. During the invasion, HSCs quickly acquired the capacity to express WT1 and HSCs showed much higher proliferative and fibrogenic phenotype after invasion compared to HSCs without invasion. This invasive, proliferative, and fibrogenic HSC phenotype was inhibited by WT1 deletion. Expectedly, HSC-specific WT1 knockout mice showed reduced HSC invasion and BF development in murine liver fibrosis models induced by bile duct ligation (BDL) and choline-deficient high fat diet (CD-HFD) feeding, concluding WT1 is a potent pro-fibrogenic factor. Then, we investigated the regulatory mechanism of WT1 induction. We found WT1 promoter contains hypoxia inducible factor (HIF) binding site and hypoxic challenge is a strong inducer of WT1 and HSC invasion. Furthermore, our bioinformatics analysis revealed that WT1 is the most enriched transcription factor for hypoxia-induced invasive phenotype in HSCs. As the downstream effectors of WT1, our RNA-seq analysis by comparing wild-type and HSC-specific WT1 knockout mice found PDGFRB and Cathepsin K (CTSK). CTSK is also upregulated by WT1 overexpression in HSCs. Pharmacological inhibition of CTSK inhibited HSC activation in vitro and prevented BF formation in the BDL model. In conclusion, the WT1-CTSK-mediated profibrogenic actions in HSCs are crucial for HSC invasion and BF formation in liver fibrosis. Thus, targeting the WT1-CTSK axis to suppress invasive HSCs could be a novel strategy to prevent BF formation in patient with chronic liver disease.

Background: Cholestatic liver disease can lead to endothelial dysfunction and dedifferentiation of liver sinusoidal endothelial cells (LSECs) with loss of fenestrations, deposition of a basement membrane, and surface expression of CD31 and CD34, a process that has been termed sinusoidal capillarization and that precedes liver fibrosis. We aimed to define the microRNA regulated vascular remodeling, intussusceptive angiogenesis, and liver fibrosis during cholestatic liver injury.
Methods: To demonstrate the overall ultrastructure of the vascular damage during cholestatic liver injury, we injected mice via the portal vein with resin to create vascular corrosion casts of the liver with BDL in miR-34a knockout and WT control mice for scanning electron microscopy analysis. CD34⁺ cells were isolated from mouse liver using laser capture microdissection (LCM). A capture probe covalently bound to an oligonucleotide containing biotin and a color-coded reporter probe were designed for 84 endothelial function-related genes and analyzed with the nCounter Single Cell Gene Expression Assay.
Results: Using BDL mouse model of cholestatic liver injury to evaluate vascular injury and deranged angioarchitecture of the liver by 3-dimensional morphology of the hepatic microcirculation, results showed that BDL induced extensive remodeling of the sinusoids with a plexus-like appearance. Casts revealed pores and endovascular pillars, the corresponding intraluminal structures to the pores of intussusception, the hallmark of intussusceptive angiogenesis after vascular injuries. The process of vascular injury associated angiogenesis, vascular remodeling and rearrangement after BDL led to an increased variation of the sinusoidal diameter and increased branching within the microcirculation. Lack of miR-34a in vivo reversed the serum ALT level, and restored the levels of Sirt1 coupled with decreased NOS3 expression as well as the reduced levels of TNFα, CCl2, IL-1β, IFNβ and IL-7 in LCM isolated CD34⁺ cells analyzed by nCounter single cell gene expression assay. Depletion of miR-34a in vivo also induced a significant down-regulation of profibrogenic genes and MMPs in total liver tissues and LCM isolated CD34⁺ cells by single cell gene assay from BDL mice liver, along with the reduced vascular remodeling and intussusceptive angiogenesis.
Conclusion: By 3-dimensional morphology of the hepatic microcirculation and single cell analysis, our discovery that microRNA-34a as an important signaling pathway in hepatic vascular endothelium that governs vascular structure and function in the liver, regulates intussusceptive angiogenesis, and contributes to liver fibrosis during cholestatic liver injury implicates an exciting field in which the epigenomic microRNAs of endothelial dysfunction may be manipulated with potential therapeutic benefits.

Background: Growth hormone (GH) and GH receptor (GHR) signaling regulates cell growth and promotes liver regeneration. GH activates downstream insulin-like growth factor 1 (IGF1) and its receptor (IGF1R) signaling, and GH and IGF1 promote cholangiocyte proliferation. We have previously demonstrated that downregulation of microRNA let-7 leads to cholangiocyte proliferation and ductular reaction. We hypothesize that upregulation of GH/GHR signaling in liver diseases induces robust cholangiocyte proliferation leading to ductular reaction, which is associated with liver fibrosis via downregulated let-7.
Methods: Human liver sections of healthy donors and patients with primary sclerosing cholangitis (PSC), primary biliary cholangitis (PBC), biliary atresia, NAFLD, non-alcoholic steatohepatitis (NASH), or alcoholic steatohepatitis (ASH) were analyzed by CK-19 immunohistochemistry (IHC) to detect ductular reaction, Sirius Red staining to determine liver fibrosis, and GHR IHC. Patients’ liver sections were also used for double immunofluorescence staining to identify GHR or IGF1R expression in CK-19⁺ cholangiocytes. Mdr2^-/- mice (PSC model, female, 12-wk old) were administered with vehicle (1×PBS), human recombinant GH (hrGH, 0.6 µg/g body weight/day) or an FDA-approved GHR antagonist (Somavert, 18 µg/g body weight/day) for 2 wks via implanted osmotic minipumps. Liver tissues were characterized by H&E, CK-19 IHC, and Sirius Red staining. RT-PCR was performed for human PSC cholangiocytes or primary murine cholangiocytes isolated from Mdr2^-/- mice after treatments to determine expression levels of let-7.
Results: Patients with PSC, PBC, biliary atresia, NASH, or ASH showed extensive ductular reaction and liver fibrosis, as well as elevated immunoreactivity against GHR in the portal area compared to healthy individuals. The NAFLD patient showed only mild ductular reaction and liver fibrosis without elevated GHR immunoreactivity. Elevated expression of GHR and IGF1R was observed in CK-19⁺ cholangiocytes of PSC, PBC, and NASH patients. Administration of hrGH exacerbated, but Somavert attenuated liver damage, portal infiltration, ductular reaction, and liver fibrosis in Mdr2^-/- mice compared to vehicle control. Expression levels of let-7 were downregulated in PSC cholangiocytes compared to control cholangiocytes. Target prediction showed that let-7 families might inhibit expression of GHR, IGF1, and IGF1R. Treatments of hrGH decreases let-7 expression in cholangiocytes in vivo, showing the association of the GH/GHR axis with liver fibrosis via downregulation of let-7 and upregulation of cholangiocyte proliferation via IGF1 signaling.
Conclusion: GH/IGF1 signaling is associated with liver fibrosis in liver diseases with extensive ductular reaction. Blocking GH/GHR signaling using Somavert could be a promising therapeutic approach for liver fibrosis.

Keisaku Sato

Appearances