Abstract Background Acute liver failure (ALF) and acute-on-chronic liver failure (ACLF) are the two most common subtypes of liver failure. They are both life-threatening clinical problems with high short-term mortality. Although liver transplantation is an effective therapeutic, its application is limited due to the shortage of donor organs. Given that both ACLF and ALF are driven by excessive inflammation in the initial stage, molecules targeting inflammation may benefit the two conditions. MicroRNAs (miRNAs) are a group of small endogenous noncoding interfering RNA molecules. Regulation of miRNAs related to inflammation may serve as promising interventions for the treatment of liver failure. Aims To explore the role and mechanism of miR-125b-5p in the development of liver failure. Methods Six human liver tissues were categorized into HBV-non-ACLF and HBV-ACLF groups. Differentially expressed miRNAs (DE-miRNAs) were screened and identified through high-throughput sequencing analysis. Among these DE-miRNAs, miR-125b-5p was selected for further study of its role and mechanism in lipopolysaccharide (LPS)/D-galactosamine (D-GalN) -challenged Huh7 cells and mice in vitro and in vivo. Results A total of 75 DE-miRNAs were obtained. Of these DE-miRNAs, miR-125b-5p was the focus of further investigation based on our previous findings and preliminary results. We preliminarily observed that the levels of miR-125b-5p were lower in the HBV-ACLF group than in the HBV-non-ACLF group. Meanwhile, LPS/D-GalN-challenged mice and Huh7 cells both showed decreased miR-125b-5p levels when compared to their untreated control group, suggesting that miR-125b-5p may have a protective role against liver injury, regardless of ACLF or ALF. Subsequent results revealed that miR-125b-5p not only inhibited Huh7 cell apoptosis in vitro but also relieved mouse ALF in vivo with evidence of improved liver histology, decreased alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels, and reduced tumor necrosis factor-α (TNF-α) and IL-1β levels. Based on the results of a biological prediction website, microRNA.org, Kelch-like ECH-associated protein 1 (Keap1) was predicted to be one of the target genes of miR-125b-5p, which was verified by a dual-luciferase reporter gene assay. Western blot results in vitro and in vivo showed that miR-125b-5p could decrease the expression of Keap1 and cleaved caspase-3 while upregulating the expression of nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and heme oxygenase-1(HO-1). Conclusion Upregulation of miR-125b-5p can alleviate acute liver failure by regulating the Keap1/Nrf2/HO-1 pathway, and regulation of miR-125b-5p may serve as an alternative intervention for liver failure. Keywords: acute liver failure, acute-on-chronic liver failure, microRNA-125b-5p, Kelch-like ECH-associated protein 1, high-throughput sequencing, inflammation Introduction Liver failure is a life-threatening clinical problem with high short-term mortality. It can present as acute liver failure (ALF, without pre-existing chronic liver disease), acute-on-chronic liver failure (ACLF, an acute deterioration of underlying chronic liver disease) or an acute decompensation of an end-stage liver disease ([35]1). Liver transplantation is an effective therapeutic option, irrespective of the etiology of liver failure. However, the application of liver transplantation is limited due to the shortage of donor organs ([36]2). Thus, liver failure is still a severe clinical challenge, and other interventions assisting alleviating liver injury are being explored. ALF and ACLF are the most widely discussed owing to their high incidence worldwide. Better knowledge of the pathophysiology of these diseases can provide insights into novel therapies. It has been reported that the development of both ALF and ACLF is driven by immune dysfunction and inflammatory imbalance, although the conditions are distinct clinical entities ([37]3–[38]5). The immune and inflammatory status of the diseases is dynamic, progressing from intensive inflammation to the development of immunoparalysis ([39]3, [40]4, [41]6). In the initial phase of ALF, immune cells participating in the innate response are activated to produce proinflammatory mediators, which can stimulate a systemic inflammatory response. Patients with ACLF display an excessive innate immune response, which is characterized by leukocytosis, neutrophilia and lymphopenia, together with high levels of inflammatory mediators ([42]7, [43]8). Initial systemic inflammatory response syndrome (SIRS) due to acute insult and/or subsequent secondary infection due to immunoparalysis can lead to extrahepatic organ failure ([44]4, [45]9, [46]10). Thus, strategies modulating immune and proinflammatory mediators can be potential targets to alleviate liver failure. MicroRNAs (miRNAs) are a class of small endogenous noncoding interfering RNA molecules and can induce gene silencing and translational repression by binding specific sequences in target mRNAs, thereby playing key roles in biological processes and in the development of various diseases ([47]11). Accumulating studies have demonstrated that miRNAs are involved in the modulation of immunity and inflammation, and regulation of these miRNAs may be potential therapeutics for clinical problems ([48]12–[49]14). For instance, inhibition of miR-34b-5p could attenuate inflammation and apoptosis in acute lung injury, and thus miR‐34b‐5p and its target progranulin might be a potential intervention pathway for the treatment of acute lung injury ([50]15). Thus, the regulation of miRNAs and their target genes may improve the outcome of various diseases by regulating immunity and inflammation. In the present study, we identified miRNAs that were differentially expressed in human HBV-ACLF tissues compared to HBV-non-ACLF tissues. Then, we further explored the role of a certain miRNA in the development of liver failure, aiming to determine whether the miRNA may improve liver failure by regulating intensive inflammation, hoping to pave the way for miRNA-targeting therapies for liver failure. Materials and methods Study population Six patients with chronic HBV infection were enrolled. They all received antiviral therapy with nucleos(t)ide analogs. They were divided into HBV-non-ACLF (n=3) and HBV-ACLF (n=3) groups based on their liver histopathology and liver function. The histological assessments were performed using the METAVIR scoring system. Briefly, the degree of inflammation in the liver biopsies was assessed with the standard METAVIR histology activity index scoring system, defined as A0, no inflammation; A1, mild inflammation; A2, moderate inflammation, and A3, severe inflammation. The histological appearance of fibrosis was classified as F0 to F4, ranging from no fibrosis to cirrhosis ([51]16). HBV-non-ACLF referred to patients with chronic hepatitis B (CHB) who had abnormal liver function and mild-to-moderate inflammatory activity (≤A2) in the liver tissue due to HBV infection, but did not meet the diagnostic criteria of ACLF. ACLF is diagnosed according to the recommendations of the Asian Pacific Association for the Study of the Liver (APASL) ([52]1). Herein, patients who were previously diagnosed with CHB or cirrhosis belonged to the HBV-ACLF group if they manifested with jaundice (bilirubin>5 mg/dL), coagulopathy [prolonged international normalized ratio (INR)>1.5], encephalopathy, and ascites within 4 weeks. The histopathological findings of the HBV-ACLF group showed evident infiltration of inflammatory cells (A3), accompanied by the formation of pseudolobuli (F4). All HBV-ACLF patients later received liver transplantation. Patients were excluded if they had any of the following conditions: (1) infections with other hepatitis viruses (including A, C, D, and E) or human immunodeficiency virus (HIV); (2) evidence of drug-induced liver injury, alcoholic liver disease, autoimmune liver diseases, severe systemic illnesses; (3) malignancies, such as hepatocellular carcinoma. This study was carried out in accordance with the Declaration of Helsinki. All patients provided verbal informed consent. Detailed information about the study cohort is described in [53]Table 1 . Table 1. Clinical characteristics of the two groups of patients. Groups Patients Gender Age(year) ALT(IU/L) AST(IU/L) TBil(μmoL/mL) INR HBV-DNA (log10 IU/mL) Inflammationactivity index Fibrosis score HBV-non-ACLF Patient 1 male 39 50 44 9.9 1.24 4.049 A1 F0 Patient 2 male 40 59 56 14.9 1.10 3.328 A2 F1 Patient 3 male 46 114 93 14.0 1.05 3.755 A2 F1 HBV- ACLF Patient 4 male 47 559 353 326.6 2.07 2.538 A3 F4 Patient 5 male 52 442 293 396.3 1.95 2.661 A3 F4 Patient 6 male 46 417 356 402.6 2.23 3.326 A3 F4 [54]Open in a new tab Upper limit of normal (ULN) of ALT: 50 IU/mL for male; ULN of TB: 28 μmol/L. HBV DNA<100 IU/mL is defined as undetectable serum HBV DNA. ALT, Alanine aminotransferase; TBil, total bilirubin. INR, International Normalized Ratio. High-throughput sequencing Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. The quantity and purity of total RNA were analyzed with an Agilent 2100 Bioanalyzer (Agilent, USA) with RIN> 6.5. Small RNAs of different length were separated using denaturing polyacrylamide gel electrophoresis (PAGE). Fragments between 18 and 30 nt in length were gel-purified and ligated to adaptors at both the 3′- and 5′-ends. The ligation products were subsequently reverse- transcribed into cDNA, and PCR amplification was performed using an Illumina sequencing kit (Illumina, USA) to generate a cDNA library according to previous studies ([55]17, [56]18). High-throughput sequencing was performed by Chengdu Life Baseline Technology Co., Ltd. As shown in [57]Supplement 1 , the raw read sequences were filtered to remove low-quality reads, 5′ adaptor contaminant reads, reads without 3’ adaptor sequences, reads containing poly (A) and adapter sequences, sequences <18 nt and sequences >32 nt. The clean reads were obtained and their length distributions were calculated using Fastx-Toolkit ([58]19). Then, the clean reads were mapped and aligned to the human reference genome group and other small RNA databases using Bowtie2 software ([59]20). The known miRNAs and novel miRNAs were identified and predicted using miRbase and miRDeep2, respectively ([60]21, [61]22). The miRNA expression levels between the two groups were compared to identify differentially expressed (DE)-miRNAs. The expression of miRNA was normalized using transcripts per million (TPM) as the following formula: TPM = (mapped read count/total clean read count) ×10^6 ([62]23). DE-miRNAs were defined as |log2 (fold change)| >1 between two groups with a false discovery rate (FDR) of <0.05. Bioinformatics analysis for DE-miRNAs To understand the functions of these DE-miRNAs, their potential target genes were predicted by RNAhybrid ([63]https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid/) and miRanda ([64]http://www.microrna.org/microrna/home.do) ([65]24, [66]25). Only the target genes predicted by both methods were considered reliable targets for further analysis. Gene ontology (GO) functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed on these target genes ([67]26, [68]27). GO analysis was conducted to provide functional annotation for predicted target genes of miRNAs by analyzing the classifications of Biological Process, Cellular Component and Molecular Function. Pathway analysis was based on KEGG, which is a database resource for understanding the high-level functions and utilities of biological systems. p<0.05 was regarded as the cutoff to select significantly enriched terms. Cell culture and transfection The human HCC cell line Huh7 was preserved in our laboratory. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, D6429, Sigma, USA) containing 10% fetal bovine serum (12103C, Sigma, USA) and 1% penicillin/streptomycin under standard culture conditions (a humidified 5% carbon dioxide incubator at 37°C). The miR-125b-5p overexpression vector and negative control (NC) vector were designed and provided by Heyuan Biotechnology (OBIO, Shanghai, China). After growth to 60%-70% confluence in six-well plates, Huh7 cells were transfected with miR-125b-5p vector or NC vector using Lipofectamine™ 2000 (11668019, Invitrogen, USA). On the following day, the cells were incubated with normal medium. Flow cytometric analysis Apoptotic cells were assessed using an Annexin V- Fluorescein isothiocyanate (FITC) apoptosis detection kit (APOAF, Sigma, USA) according to the manufacturer’s protocol. Huh7 cells were collected and washed twice with PBS. Then, the cells were resuspended in 500 μL of 1× binding buffer and stained with 5 μL of Annexin V-FITC conjugate and 10 μL of PI solution. After incubation for 15 min in the dark at room temperature, stained cells were analyzed by flow cytometry (BD Accuri™ C6 flow cytometer). Real−time quantitative PCR analysis Total RNA was extracted from Huh7 cells and liver tissues using TRIzol Reagent (15596026, Invitrogen, USA) according to the manufacturer’s instructions. Reverse transcription was performed using a First Strand cDNA Synthesis Kit (B300537, Sangon Biotech, China). The expression of miR-125b-5p was determined by quantitative real-time PCR using a miRNA qPCR detection kit (B532461, Sangon Biotech, China), as previously reported ([69]28). The forward primer sequence for miR-125b-5p was CGTCCCTGAGA- CCCTAACTTGTGA. The reverse primers for miR-125b-5p were universal adaptor primers designed and provided by Sangon Biotech Company (Shanghai, China). The level of miR-125b-5p was calculated using the relative quantification 2^-ΔΔCT method and normalized to the U6 transcript (Bio-Rad CFX Manager software), as previously described ([70]29, [71]30) Animals Male C57BL/6J mice (6-8 weeks old, weighing 20-25 g) were purchased from Huaxi Laboratory Animal Center of Sichuan University (Chengdu, China). All mice were maintained under controlled conditions (24°C, 55% humidity and 12-h day/night rhythm) and given free access to water and food. The mice received humane care under guidance from the Institutional Review Board in accordance with the Animal Protection Art of Sichuan University. After 1 week of acclimation, the mice were prepared for further study. Mouse model of acute liver failure The mouse model of ALF was established using lipopolysaccharide (LPS) and D-galactosamine (D-GalN) as previously described ([72]31). In brief, C57BL/6J mice were given 700 mg/kg D-GalN (G0500, Sigma, USA) and 10 μg/kg LPS (Escherichia coli, 0111:B4, L2630, Sigma, USA) by intraperitoneal injection. Mice in the present study were randomly divided into four groups (n = 8/group): a normal control group, an LPS/D-GalN group, an LPS/D-GalN+ negative control (NC) group and an LPS/D-GalN+ miR-125b-5p group. The miR-125b-5p overexpression vector or negative control vector was administered to mice via tail vein prior to the establishment of ALF as previously reported ([73]32, [74]33). Seven hours after ALF model establishment, all mice were sacrificed, and serum and liver samples were harvested and stored for further analysis. H&E staining Liver tissue was obtained and fixed with 4% paraformaldehyde at 4°C for 48 h and then embedded in paraffin. After immobilization, samples were cut into sections and then stained with hematoxylin-eosin (HE) using a standard protocol. The sections were visualized under a light microscope, and representative images are presented. Biochemical detection of aminotransferase Serum samples were collected from mice to detect alanine aminotransferase (ALT) and aspartate aminotransferase (AST) by an automatic biochemical analyzer. Enzyme-linked immunosorbent assay (ELISA) The levels of serum tumor necrosis factor -α (TNF-α, EMC102a, NeoBioscience, China) and IL-1β (EMC001b, NeoBioscience, China) were detected using ELISA kits according to the manufacturer’s instructions. Western blotting Protein was extracted from Huh7 cells and liver tissues. The isolated proteins were separated by polyacrylamide gel electrophoresis and then transferred onto polyvinylidene fluoride (PVDF) membranes. Subsequently, the blots were exposed to primary antibodies against Kelch-like ECH-associated protein 1 (Keap1, #8047, CST, USA), nuclear factor (erythroid-derived 2)-like 2 (Nrf2, sc-365949, Santa Cruz, USA), heme oxygenase-1 (HO-1, #43966, CST, USA), and cleaved caspase-3 (#9661, CST, USA). Anti-GAPDH (TA-08), anti-β-tubulin (TA-10) and anti-β-actin (TA-09) were used as internal references, and were