Introduction

Hepatocellular carcinoma (HCC) is a kind of lethal malignant primary liver cancer and ranks fourth on the cancer-related death list and sixth in the terms of incident cases worldwide, which has a low 5-year survival rate1,2. Liver diseases, such as chronic hepatitis caused by hepatitis B or C virus (HBV or HCV), alcoholic liver disease and non-alcoholic fatty liver disease, increase the risk of hepatocellular carcinoma development3. Generally, HCC carcinogenesis is caused by the repeated cycle of hepatic injury, induced liver inflammation, and compensatory hepatocyte proliferation4. The liver injury plays a critical role in this vicious cycle, and death of hepatocytes is able to release damage-associated molecular patterns (DAMPs), which activates immune cells in the liver, recruits circulatory inflammatory cells, and initiates liver inflammation. The induced liver inflammation stimulates compensatory hepatocyte proliferation, while hepatic inflammatory responses also worsen hepatic damage and aggravate liver inflammation. Sustained hepatic injury induces the chronic liver inflammation and repeated compensatory hepatocyte proliferation, which eventually lead to hepatocarcinogenesis5,6. The underlying regulatory mechanisms of hepatocarcinogenesis have attracted much attention but remain largely unknown, which needs further investigation.

Cell death plays pivotal roles in the initiation of hepatocarcinogenesis. Apoptosis, necroptosis, pyroptosis, and ferroptosis are the intensively explored types of cell death during tissue damage, and different types of hepatocyte death eventually lead to different types of liver cancer7. For instance, the necroptosis of hepatocytes incubates an environment, which determines intrahepatic cholangiocarcinoma (ICC) outgrowth. However, if hepatocytes are in the environment created by apoptotic hepatocytes, they are inclined to become HCC8. Hepatocyte-specific deletion of the IκB kinase (IKK) subunit NEMO/IKKγ sensitizes hepatocyte apoptosis by NF-κB inhibition, which spontaneously forms mouse HCC in 12 months9. Similarly, mice lacking the anti-apoptotic myeloid cell leukemia-1 (Mcl-1) in hepatocytes have severe liver damage caused by spontaneous apoptosis, and tumor formation is observed in over 50% of mice in 8 months10. Besides, mice with combined knockout of receptor-interacting serine/threonine-protein kinase 1 (RIPK1) and Rel-like domain-containing protein A (RelA) in hepatocytes show increased hepatocyte apoptosis and develop spontaneous HCC11. Furthermore, other proteins regulating hepatocyte apoptosis, such as TGF-β-activated kinase 1(TAK1), TNF receptor-associated factor 2 (TRAF2), and IκB kinase subunit beta (IKKβ), also participate in hepatocarcinogenesis12,13. Therefore, hepatocyte apoptosis is critical in the initiation of hepatocarcinogenesis, and the molecular regulatory mechanisms of hepatocyte apoptosis and their roles in hepatocarcinogenesis have raised widespread concerns.

MicroRNAs (miRNAs) are a class of single-strand RNA with approximately 22 nucleotides in length. The functions of miRNAs have been widely investigated, and they are determined to play critical roles in cancer progression, especially in HCC. A set of miRNAs, such as miR-330-5p, miR-520a, and miR-483-3p, have been suggested to participate in cancer progression, by modulating proliferation, apoptosis, migration, and invasion41. All animal experiments were undertaken in accordance with National Institutes of Health’s Guide for the Care and Use of Laboratory Animals with the approval of the institutional research ethics committee of Second Military Medical University, Shanghai, China. HL-7702 and BNL CL.2 cell lines were obtained from cell bank of Chinese Academy of Sciences (Shanghai, China). All cell lines have been authenticated using STR profiling and tested for mycoplasma contamination by Genechem (Shanghai, China). The knockout cell lines were constructed using CRISPR/Cas9 method (Supplementary Table S2)42. HL-7702 was cultured in RPMI 1640 with 10% Fetal Bovine Serum (FBS), and BNL CL.2 was cultured in dulbecco’s modified eagle medium (DMEM) with 10% FBS and 1% non-essential animo acid (NAA).

Reagents

Antibodies specific to caspase-3 (9662), cleaved caspase-3 (9661), caspase-8 (4790), human cleaved caspase-8 (9496), mouse cleaved caspase-8 (9429), BAK (12105), BAX (2772), BCL-2 (3498), BCL-XL (2764), PDCD4 (9535), and horseradish peroxidase-coupled secondary antibodies (7074 and 7076) were from Cell Signaling Technology (Danvers, MA). Antibody specific to β-actin (A5441), DEN (N0258), DHE (D7008), CHX (239764), LPS (L3024) and MEM NAA (M7145) were from Sigma-Aldrich (St. Louis, MO). Antibody specific to cytochrome C (sc-13156) was from Santa Cruz Biotechnology (CA, USA). Recombinant human TNF-α (300-01A) and murine TNF-α (315-01 A) were from Pepro Tech (Rocky Hill, NJ). TUNEL assay kit (11684817910) was from Roche (Shanghai, China). CCl4 (CAS: 56-23-5, C805329) was from MAKLIN (Shanghai, China). Olive oil (CAS: 8001-25-0, A502795-0100), APAP (CAS: 103-90-2, A506808), and D-hanks buffer (B548148-0500) were from Sangon Biotech (Shanghai, China). Cell mitochondria isolation kit (C3601) and tissue mitochondria isolation kit (C3606) were from Beyotime (Shanghai, China). Percoll solution (17-0891-01) was from GE Healthcare Life Science (Little Chalfont, UK). Type IV collagenase (LS004140) was from Worthington Biochemical Corporation (Lakewood, NJ). Protease inhibitor cocktail (539134-1SML) and apoptosis assay kit (PF032) were from Calbiochem (Darmstadt, Germany).

Establishment of hepatocarcinogenesis model

To construct DEN-induced hepatocarcinogenesis model, the 2-week-old male mice were injected with 25 mg/kg body weight DEN intraperitoneally43. Mice were sacrificed 8 months later and tumors were analyzed. As for DEN plus CCl4 model, the 2-week-old male mice were injected with 25 mg/kg body weight DEN intraperitoneally, and 4 weeks later, mice were injected with CCl4 (0.5 ml/kg body weight, dissolved in olive oil at a ratio of 1:3) weekly lasting for 15 weeks19. Mice were sacrificed 8 weeks after last CCl4 injection.

Establishment of acute hepatic injury model

To constructed DEN-induced acute hepatic injury model, 8-week-old male mice were injected with DEN (100 mg/kg) intraperitoneally and mice were sacrificed according to the time points. As for APAP-induced acute hepatic injury model, eight-week-old male mice were fasted overnight (16 h) and then injected with APAP (400 mg/kg) intraperitoneally. Mice were fed immediately after injection and sacrificed as indicated. Serum ALT and AST were measured by automatic biochemical analyzer FDC-7000i (Shanghai, China).

In vivo AAV8 administration

The rAAV (serotype 8) vector expressing PDCD4 under promoter CAG was constructed as previously described18. For AAV8 administration, 1 × 1012 vg AAV8 per mouse in 200 μl saline buffer was injected into the mice through tail vein. Four weeks later, the protein expression of PDCD4 was measured by Western blot.

Isolation of primary hepatocytes

Primary hepatocytes were isolated from mouse liver using two-step collagenase perfusion procedure44. Mice were anesthetized and first perfused with 45 ml D-hanks buffer and then with type IV collagenase digestion (1 mg/ml) for 5 min at a rate of 2 ml/min through the portal vein. After digestion, the liver was excised, minced, filtered through 70-micron membrane, spin for 5 min at 50 × g, and purified using 50% Percoll solution. The obtained hepatocytes were resuspended in DMEM, and trypan blue exclusion assays indicated that cell viability was >95%. Primary hepatocytes were cultured in DMEM with 10% FBS and 1% Penicillin-Streptomycin.

RNA isolation and quantitative PCR analysis

Total RNA was extracted from mouse liver tissues or cell lines using RNAiso Plus reagent (Takara, Dalian, China) according the recommended protocol. Real-time quantitative RT-PCR (qRT-PCR) analysis was performed using LightCycler 2.0 (Roche, Switzerland) and SYBR RT-PCR kit (Takara) as previously described18. The relative expression level of the individual genes was normalized to that of internal control by using 2-ΔΔCt cycle threshold method in each sample (Supplementary Table S3)45.

Western blot

Cells or tissues were lysed on ice with cell lysis buffer (Cell Signaling Technology) supplemented with protease inhibitor cocktail (Calbiochem) at a ratio of 1:200. Protein concentrations of the extracts were measured with bicinchoninic acid (BCA) assay (Pierce). Equal amount of the extracts was loaded to sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred onto nitrocellulose membrane for immunoblot analysis as described previously20.

Histology

For HE, IHC and TUNEL, liver tissues were fixed in 4% paraformaldehyde for 48 h, embedded by paraffin and sliced up into 5 μm thick sections, which were processed to HE, IHC, and TUNEL as kit protocols described. For reactive oxygen species (ROS) measurement, fresh liver tissues were embedded by Tissue-Tek® O.C.T compound and immediately sliced up into liver sections. The slices were incubated in DHE (1 μM) for 30 min. After washing with PBS for three times, the red fluorescence was measured by fluorescence microscope.

Dual-luciferase reporter assay

The PDCD4 luciferase reporter was made by amplifying the mouse pdcd4 mRNA by PCR and cloned into the 3ʹUTR region of pMIR-promoter-Firefly plasmids. The luciferase reporter plasmids, RL-TK-Renilla plasmids, and indicated miRNAs (final concentration 20 nM) were co-transfected into HEK 293T cells. After 24 h, the activities of Firefly and Renilla luciferases were measured using the Dual-Luciferase Reporter Assay System (Promega) as previously described46. Data was normalized for transfection efficiency by dividing Firefly luciferase activity with that of Renilla luciferase.

Flow cytometry

The control and knockout cells were treated with APAP (10 mM) for 24 h. Cells were collected and labeled by apoptosis assay kit and subjected to flow cytometry analysis on LSRII. Data were analyzed using FACSDiva software (Becton Dickinson).

Statistical analysis

As indicated in the figure legends, sample size of the experiments depended on the assay type. There were no blind experiments for the investigators both in cells and mice experiments. The mice were randomly distributed to the experimental group. For all groups that are statistically compared, the variance within each group was similar. Data are shown as mean ± SD from one representative of three independent experiments. Statistical comparisons between experimental groups were analyzed by unpaired Student’s t-test or chi-square test in GraphPad Prism 8.0, and a two-tailed P < 0.05 was taken to indicate statistical significance.