Cell treatment and transfection

Human hepatocellular carcinoma cell lines Huh-7 (CL-0120) and HCCLM3 (CL-0278), and 293 T cells procured from Procell Life Science & Technology Co., Ltd. (Wuhan, China) were cultured in DMEM media added with 10% fetal bovine serum (FBS) and 100 U/mL penicillin/streptomycin in a humidified incubator (37 ℃, 5% CO2).

For β-Sitosterol treatment or GSK3B inhibitor (CHIR-98014), β-Sitosterol (Sigma-Aldrich, St. Louis, USA) was first dissolved in DMSO and then diluted with complete culture medium. Huh-7 and HCCLM3 cells were incubated with 5 μg/mL, 10 μg/mL, or 20 μg/mL of β-Sitosterol for 24, 48 or 72 h. CHIR-98014 (MedChem Express, Shanghai, China) was treated to Huh-7 and HCCLM3 cells at 50 nM concentration for 48 h [21].

The GSK3B overexpression plasmid (pLVX-Puro-GSK3B plasmid) and blank pLVX-Puro vector (named vector-NC) were provided by Oribio (Changsha, China). Polymerase chain reaction (PCR) was performed using DNA containing the GSK3B sequence as the template. The amplification products were isolated and purified using 1% agarose gel electrophoresis. The target fragments were retrieved and double digested with XhoI and EcoRI. After further purification, the GSK3B CDS fragment was obtained, and then was sub-cloned into pLVX vector by homologous recombination method using Seamless Cloning Kit (Beyotime, Shanghai, China). Then, the GSK3B pLVX vector and blank pLVX vector were transfected into Huh-7 and HCCLM3 cell lines to a final concentration of 1 μg/ml using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, USA) according to the manufacturer’s protocols. The primer sequences of the GSK3B pLVX vector construction are shown as follows: Forward: 5′-ATGATGACGATAAAGGATCCATGTCAGGGCGGCCCAGA-3′. Reverse: 5′-CCGGTAGAATTATCTAGATCAGGTGGAGTTGGAAGCTG-3′. The GSK3B pLVX vector reports were provided in the Supplementary Material_1. After transfection for 48 h, the cells were treated with 10 μg/mL β-Sitosterol for another 48 h.

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay

Huh-7, HCCLM3 and 293 T cell lines were plated onto 96-well plates (5 × 103 cells/well) and cultured in 100 μL of media. All these cell lines were treated with β-Sitosterol at gradient concentrations (5 μg/mL, 10 μg/mL, and 20 μg/mL) for 24, 48, and 72 h, respectively. Next, each well was introduced to 20 μL of MTT (5 mg/mL). After incubation at 37 ℃ for 4 h, each well was added with 100 μL DMSO, followed by a 10 min gentle shaking. A multifunctional microplate reader (PerkinElmer, Shanghai, China) was employed to measure the optical density (OD) at 490 nm.

Cell counting kit-8 (CCK-8) assay

Huh-7 and HCCLM3 cell lines were seeded onto 96-well plates (5 × 103 cells per well). After GSK3B overexpression plasmid transfection and β-Sitosterol or CHIR-98014 treatment or for 24, 48 and 72 h, 10 μL of CCK-8 solution was added to each well. After 2 h incubation, a multifunctional microplate reader (PerkinElmer) was employed to measure the OD value at 450 nm.

Colony formation assay

Huh-7 and HCCLM3 cell lines were plated onto six-well plates (1 × 103 cells per well) and incubated overnight. The cells were then treated with β-Sitosterol or CHIR-98014 or 48 h or transfected for 48 h, followed by an incubation for 1–2 weeks with weekly medium refreshed. After the experiment, the culture media was removed. The cells were rinsed in phosphate-buffered saline (PBS), fixed in 4% paraformaldehyde at room temperature (RT) for 20 min, and stained with 0.1% crystal violet for 1 min. Finally, colonies were counted using an optical microscope (Nikon, Tokyo, Japan).

EdU assay

The cell proliferation detection kit (Guangzhou RiboBio Co., Ltd, Guangzhou, China) was used to detect cell proliferation. In brief, Huh-7 and HCCLM3 cell lines were seeded onto 96-well plates (1 × 104 cells per well) and treated with β-Sitosterol or transfected with plasmids for 48 h. After that, 100 μL EdU media (50 μM) was added to each well for 2 h incubation. After PBS washing, each well was added with 50 μL PBS solution containing 4% paraformaldehyde for 30-min incubation at RT. The fixed solution was then removed. Next, 50 μL glycine solution (2 mg/mL) and 100 μL PBS solution containing 0.5% Triton X-100 were added to each well for 10-min incubation. Afterward, cells were subjected to 30 min staining at RT with 100 μL 1 × Apollo staining reaction solution in the dark. The nucleus was counter-stained using a Hoechst33342 reaction solution. An inverted fluorescence microscope (Olympus, Tokyo, Japan) was employed to observe EdU-positive cells (red) [22].

Flow cytometry

A PI staining kit (Beyotime) was employed to detect the cell cycle based on previous research method [23]. In short, Huh-7 and HCCLM3 cell lines were subjected to treatment with transfection or β-Sitosterol with plasmids for 48 h, centrifuged, and re-suspended in pre-cooled PBS. Cell density was set to 2 × 105 cells/mL and then fixed (4 ℃, overnight) in 1 mL 70% ethanol. Cells were centrifugated, and then re-suspended within pre-cooled PBS and then centrifuged again. The supernatant was removed. Next, cells were added with 500 µL PI staining solution (containing RNase), and slowly and thoroughly re-suspended. Cells were incubated (37 ℃, 1 h) in a warm water bath in a dark environment. After that, cells were placed on ice and detected using a flow cytometer (BD Biosciences, San Jose, USA) within 24 h.

The Annexin V-PI double staining kit (Nanjing Keygen Biotech CO., Ltd, Nanjing, China) was used to detect cell apoptosis. In brief, Huh-7 and HCCLM3 cell lines were seeded onto six-well plates at a density of 1 × 105 cells per well and transfected with plasmids and treated with β-Sitosterol or CHIR-98014 for 48 h. Next, cells were collected and re-suspended in 500 μL of buffer. After that, 5 μL of Annexin V-FITC was added and mixed, followed by adding 5 μL of PI and mixing for reaction (room temperature, 15 min) away from light. A flow cytometer (BD Biosciences) was employed to analyze cell apoptosis immediately.

Scratch assay

The scratch assay were applied to determine migration ability of Huh-7 and HCCLM3 cells [24]. Huh-7 and HCCLM3 cell lines were transfected for 48 h. Then, the cells were seeded into six-well plates an cultivated until they reached 90% confluence. A 200 μL pipette tip was used to scratch the cells. Followed ,by 3 times PBS washes, the cells were pretreated with mitomycin C for 1 h and receiving β-Sitosterol treatment in a humidified incubator (37 ℃, 5% CO2). After incubation, the cells were imaged at 0 and 48 h.

Transwell assay

The Matrigel-coated Transwell chambers were used to measure invasion ability of Huh-7 and HCCLM3 cells according to previously described method [25]. Cells transfected with β-Sitosterol or CHIR-98014 treatments were placed in the top chamber, while the bottom chamber was filled with DMEM containing 10% FBS. The bottom surface of the upper chamber was stained with crystal violet dye for 20 min. Finally, a microscope was used to observe the number of invasive cells.

Western blot

RIPA lysis buffer (Beyotime) was used to extract total protein from Huh-7 and HCCLM3 cells. Protein content was detected using a bicinchoninic acid (BCA) reagent kit (Beyotime). After separation by polyacrylamide gel electrophoresis, the protein was electroblotted onto nitrocellulose filter (NC) membranes (Merck Millipore, Billerica, USA). The membranes were then block with TBST solution containing 5% non-fat milk at room temperature for 2 h, followed by overnight incubation at 4 ℃ with primary antibodies including anti-BAX, BCL2, anti-cleaved caspase3, anti-total caspase3, anti-E-cadherin, anti-Snail, Anti-Vimentin and anti-GAPDH. Antibody dilution and resources are listed in Table S1. Membranes were washed three times with TBST solution before being incubated for 2 h at room temperature with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:2000; ab6721; Abcam, Cambridge, USA), followed by three times washes with TBST solution. GAPDH served as the internal reference. Color development was carried out using enhanced chemiluminescence (ECL) (New Cell & Molecular Biotech Co., Ltd., Suzhou, China) [26].

Bioinformatics analysis

The drug target genes of β-Sitosterol were collected from PubChem (https://pubchem.ncbi.nlm.nih.gov/) [27]. The GSE112790 microarray containing 183 hepatocellular carcinoma specimens and 15 normal control specimens, was downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The GSK3B gene expression analysis was performed on this dataset using the R language limma package with t test analysis. The raw data of GSK3B gene expression in GSE112790 dateset are provided in the Supplementary Material_2. Using the Cancer Genome Atlas (TCGA) combined with Genotype Tissue Expression (GTEx) cohort, GSK3B expression was analyzed between tumors and corresponding normal tissues in 20 tumors, including TCGA-Liver Hepatocellular Carcinoma (TCGA-LIHC), via SangerBox online tool (http://sangerbox.com/). The Kaplan–Meier plot analysis based on SangerBox tool was applied to evaluate the prognostic significance of GSK3B in hepatocellular carcinoma patients. Using the auto select best cutoff algorithm on the website, hepatocellular carcinoma patients were divided into two cohorts (high vs. low expression). Then, the hazard ratio (HR) and p value were calculated. In addition, the receiver operating characteristic (ROC) curve based on SangerBox tool was constructed to evaluate the feasibility of using the GSK3B expression level to distinguish hepatocellular carcinoma patients from normal subjects.

Statistical analysis

The cell experiments were performed for three biologic replicates. All data were presented as mean ± SD. The One-way analysis of variance (ANOVA) followed Tukey’s post hoc test was performed using GraphPad Prism 8.0 software. The significance level was set at p < 0.05.