Human foetal liver cell line (LO2) and human hepatoma cell line (HepG2), widely used for establishing injured liver cell models [13, 14], were obtained from the Senior Department of Infectious Diseases, the Fifth Medical Center of PLA (People’s Liberation Army) General Hospital. Human umbilical vein endothelial cells (HUVECs) were obtained from Otwo Biotech (Guangzhou, China). LO2 and HepG2 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen, Shanghai, China), containing 10% foetal bovine serum (FBS; Gibco Life Technologies, Grand Island, NY, USA), under 5% CO2 at 37 °C, while the HUVECs were cultured in endothelial cell medium (ECM; Sciencell, Carlsbad, CA, USA), containing 5% FBS and 1% endothelial cell growth factor (ECGS; Sciencell) in a 5% CO2 atmosphere at 37 °C.
Establishment of the injured liver cell model
LO2 and HepG2 cells (1.5 × 105 cells/well) were seeded in 96-well plates. After 24 h, D-galactosamine (D-GalN; ST1213, Beyotime Biotechnology, Shanghai, China) was added at final concentrations of 6, 8, 10, and 12 mg/mL and CCK8 assays (K1018, APExBIO Technology LLC, Houston, TX, USA) were performed after 24 h to confirm cell injury. Thereafter, the suitable D-GalN concentration and treatment duration were adopted for subsequent experiments.
Collection of conditioned medium (CM)
LO2 and HepG2 cells (2 × 105 cells/well) were seeded in 6-well plates and cultured in the presence of D-GalN (final concentration, 10 mg/mL) for 24 h. The experimental group was treated with G-CSF at a final concentration of 10,000 ng/mL; the total culture time (72 h) included 24 h before, 24 h during, and 24 h after D-GalN treatment. Thereafter, the medium was replaced with ECM and the cells were further cultured for 48 h. Finally, the supernatant was collected as the CM for subsequent experiments.
The CCK8 assay (K1018, APExBIO Technology LLC) was performed following the manufacturer’s protocol. For injured liver cells and their corresponding counterparts treated with varying concentrations of G-CSF (500, 1,000, and 10,000 ng/mL), CCK8 assays were performed at 0, 24, 48, 72, and 96 h.
HUVECs were plated for 24 h, and the culture medium was replaced with CM from injured liver cells and G-CSF-treated injured liver cells (10,000 ng/mL). The CCK8 assay was performed 48 h later.
Quantitative real time-polymerase chain reaction (qRT-PCR)
RNA was extracted from the cells using TRIzol reagent (15596018, Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s instructions. Thereafter, cDNA was synthesized using the PrimeScript RT-PCR Kit (FSQ-301, Toyobo, Osaka, Japan). Next, qRT-PCR was performed using the cDNA as a template and the Universal PCR Master Mix (CW0957M, Beijing Cowin Biotech Co., Ltd., Beijing, China) on a CFX96 Touch Deep WellTM Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). The amplification results were then quantitated using the 2(-ΔΔCt) method. Sequences of the primers used are shown in Supplement Table 1.
Total protein was separated and transferred onto polyvinylidene difluoride membranes (Millipore Corp., Billerica, MA, USA). Thereafter, the membranes were incubated overnight with primary antibodies (Supplement Table 2) at 4 °C and in peroxidase-linked goat anti-rabbit-IgG (1:5,000; ab6721, Abcam, Cambridge, UK) and goat anti-mouse-IgG (1:5,000; ab6789, Abcam) at room temperature (approximately 25 °C) for 1 h. The blots obtained were developed using enhanced chemiluminescence reagents (32106, Pierce, Rockford, IL, USA), and analysed quantitative using the ImageJ software (National Institute of Health, Bethesda, MD, USA).
Tubule formation assay
HUVECs (3 × 104 cells/well) were plated in Matrigel (356234, Corning Inc., Corning, NY, USA)-precoated 48-well plates. After culturing in CM for 12 h, the tubular structures were quantified via manual counting at three random ×100 fields. Thereafter, the number of structures was averaged.
Establishment of the injured liver mouse model
A total of 18 male C57BL/6 mice (Pengyue Laboratory Animal Breeding Co., Ltd., Jinan, China) were assigned to the control (n = 4), injured liver (n = 7), and treatment (n = 7) groups. Mice in the control group were not administered any treatment. Those in the injured liver group received a peritoneal injection of D-GalN (1,000 mg/kg), while those in the treatment group received a subcutaneous injection of G-CSF (250 µg/kg) for five days before the peritoneal injection of D-GalN (1,000 mg/kg). The mice were decapitated 24 h after D-GalN injection Blood and liver samples were collected, and the latter was soaked in formalin. Haematoxylin–eosin (HE) staining of the liver tissue was performed and alanine aminotransferase (ALT) levels in the serum were detected to confirm liver injury. The animal experiments were performed in accordance with the principles of the 1975 Declaration of Helsinki and were approved by the Ethics Committee of the Fifth Medical Center of PLA General Hospital (Approval ID: IACUC-2013-022).
Blood samples from the mice was centrifuged at 1,000 rpm for 10 min and sera samples were collected for ALT assay, which was performed using the ALT assay kit (C009-2-1, Nanjing Jiancheng Bioengineering Institute, Nanjing, China) in accordance with the manufacturer’s instructions.
Mice liver samples were embedded in paraffin, cut into 3-μm-thick slices, and incubated overnight with primary antibodies against Ki67 (1:100, AF0198, Affinity Bioscience, Jiangsu, China) or CD31 (1:1500, ab182981, Abcam) at 4 °C. Normal rabbit IgG was used as the negative control. The slices were then incubated with HRP goat anti-rabbit IgG polymer (KIT-5005, Maixin, Biotechnology Development Co., Ltd., Fuzhou, China) and stained with 3,3′-diaminobenzidine, while the cell nuclei were stained with haematoxylin. The staining scores were simultaneously evaluated by two independent pathologists at ×200 magnification. The proportion score was recorded in at least four random fields and presented as the fraction of stained cells (0 < 10%; 1 = 10–25%; 2 = 26–75%; 3 > 75%). Further, the intensity score represented the average staining intensity (0 = none; 1 = weak; 2 = intermediate; 3 = strong). The expression of Ki67 was determined as the product of the proportion and intensity scores. Furthermore, micro vessel density (MVD) was determined using CD31 immunoreactivity and quantified via manual counting at five random ×200 fields to obtain the average. Images were then scanned using iViewer (Suzhou Youna Medical Equipment Co., Ltd., Jiangsu, China) and CaseViewer systems (3DHISTECH Ltd., Budapest, Hungary).
All statistical analyses were performed using SPSS software version 19.0 (IBM, Chicago, IL, USA) and data are presented as the mean ± SD. Specifically, Student’s t-test was performed to analyse differences between groups, while one-way ANOVA was performed to analyse differences among groups. p < 0.05 was considered statistically significant. GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA) was used to generate histograms. The experiments were performed in triplicate.