All experiments were conducted in biological replication, with each experiment repeated three times.

Cell culture

Hepa1-6 and Huh7 cell lines were procured from the Shanghai Cell Bank at the Chinese Academy of Sciences and the American Typical Culture Center, respectively. The Hepa1-6 cell line was specifically isolated from BW7756 liver cancer induced in C57/L mice. For the Huh7 cell culture, we used Dulbecco’s Modified Eagle’s medium (Thermo Fisher Scientific, C11995500BT) supplemented with 10% fetal bovine serum (GibcoLife Technologies, Carlsbad, CA) and 1% penicillin/streptomycin (P1400, Solarbio, China). The cells were maintained in an incubator at 37 °C with 5% CO2. Meanwhile, the Hepa1-6 cells were cultured in Roswell Park Memorial Institute 1640 medium (Thermo Fisher Scientific, C11875500BT) with the same supplements and incubator conditions as the Huh7 cells.

Reagents and antibodies

Donafenib and Dimethyl sulfoxide (DMSO) was purchased from MedChemExpress. Donafenib was dissolved in DMSO. Bcl-2 monoclonal antibody (#68,103–1-Ig), p53 polyclonal antibody (#10,442–1-AP), Caspase8 monoclonal antibody (#66,093–1-Ig), Beta Actin monoclonal antibody (#66,009–1-Ig), HIF-1 alpha monoclonal antibody (#66,730–1-Ig), GPX4 monoclonal antibody (#67,763–1-Ig), Caspase3 monoclonal antibody (#66,470–1-Ig), CD71 monoclonal antibody (#66,180–1-Ig), PD-L1 monoclonal antibody (#66,248–1-Ig), Bax polyclonal antibody (#50,599–2-Ig), SLC7A11 polyclonal antibody (#26,864–1-AP), HRP conjugated Effinipure goat antirabbit IgG (H + L) (#SA00001-2), and HRP conjugated Effinipure goat antimouse IgG (H + L) (#SA00001-1) were analyzed using a cell apoptosis analysis kit (#C1052) purchased from Protein Group and Beyotime Biotechnology.

Cell viability assay

A Cell Counting Kit-8 (Biosharp) was used to detect the degree of cytotoxicity of donafenib on Hepa1-6 and Huh7 cells, and each group was repeated three times. The cell density was adjusted to 5 × 104 cells/mL, 100 μL was added to a 96-well plate, cultured at 37 °C in a 5% CO2 incubator overnight for 24 h, and then treated with different concentrations of donafenib for intervention. 10 µL CCK8 was added at different detection time points and incubated for 2 h. The optical density (OD) was read with an absorbance value of 450 nm on the microplate reader. The semicirculated inhibitory concentration (IC50) of acetylides at lodges on each cell was calculated by Prism 9.0 software program, and the proliferation curve was made. Statistical analysis using Prsim 9.0.

Transwell migration/invasion assay

The matrix gel was diluted in serum-free medium at a ratio of 1:9 and placed on the upper surface of the Transwell chamber bottom membrane. It was incubated at 37 °C for 2 h. After 2 h, suck out the excess matrix gel. Cells were starved by serum for 12 h. When measuring cell migration, no matrix gel was added. When measuring cell invasion, matrix gel was added. Add 300 µL serum-free medium and 8 × 104 Huh7 or Hepa1-6 cells to each Transwell chamber. Add 800 µL of culture medium containing 10% FBS into the lower chamber of a 24 well plate. The experimental group was divided into a control group (add 1 µL DMSO to every 1 mL of culture medium), donafenib L group (5 µM/L), donafenib M group (10 µM/L), and donafenib H group (20 µM/L). The DMSO content in each group’s culture medium should not exceed one thousandth. After culturing with donafenib for 24 h, the cells were fixed and stained with 0.1% crystal violet for 60 min. The cells that crossed the membrane were observed and counted under a microscope. Quantitative analysis was performed using ImageJ software. Statistical analysis using Prsim 9.0.

Wound-healing invasion assay

Hepa1-6 and Huh7 cells were digested and seeded into 6-well plates. According to the calculated IC50, they were divided into a control group (add 1 µL DMSO to every 1 mL of culture medium), donafenib L group (5 µM/L), donafenib M group (10 µM/L), and donafenib H group (20 µM/L). When the cell density reaches 80%, use a 100 µL plastic pipette to draw a perpendicular line perpendicular to the well plate and the marked line, so that the scratch intersects with the marked line, forming several intersection points as fixed detection points and washed three times with phosphate buffered saline (PBS). After being washed, the cells were added to a 2% FBSDMEM medium containing different concentrations of donafenib, and pictures were taken at 0 and 24 h under an inverted microscope (Olympus). A transwell chamber (8 µm pore size, BD) was coated with 100 µL of Matrix (285 µg/mL, Corning) and placed in an incubator at 37 °C for 1 h to gelatinize.

We then added 600 µL of culture medium (containing 10% FBS) to each well of a 24-well plate. We placed a transwell culture chamber over the wells and added 200 µL of culture medium (excluding 10% FBS) containing 5 × 104 cells and the corresponding donafenib concentrations to each chamber. After a 48-h cultivation period, we rinsed the upper chamber with PBS, discarded the cells from the upper chamber, affixed the cells to the chamber’s base using methanol, and colored them with 0.1% crystal violet for 30 min at room temperature. Following washing and drying, the stained cells were deemed to have penetrated into the lower chamber. Photos were taken under an inverted microscope. Statistical analysis using Prsim 9.0.

Cell apoptosis analysis

Hepa1-6 and Huh7 cells were digested, and 2 mL of culture medium (2 × 105 cells/ml) were added to each well on a 6-well plate. The experimental group was divided into a control group (add 1 µL DMSO to every 1 mL of culture medium), donafenib L group (5 µM/L), donafenib M group (10 µM/L), and donafenib H group (20 µM/L). Following normal cultivation for 24 h, the different groups of cells were treated with the corresponding concentrations of donafenib for 24 h. The Beyotime cell apoptosis analysis kit was employed to detect apoptosis in the cells, while flow cytometry was used to assess the distribution pattern of apoptotic cells. Stain with Servicebio TUNEL fluorescence kit, take photos under fluorescence microscope, and calculate and analyze fluorescence intensity using ImageJ. Statistical analysis using Prsim 9.0.

RNA sequencing

The Huh7 cells were digested in a suspension, maintaining a cell concentration of 2 × 105 cells per milliliter. These cells were dispensed into a 6-well plate, with 2 ml added to each well. We next set up a blank control group and an experimental group (20 µM/L), repeating the process three times. After 24 h of normal culture, we collected the cells and extracted their RNA using the Trizol method. This step requires measuring the concentration and purity of the extracted RNA.

We enriched the mRNA with poly(a) tails using oligo (DT) microbeads and randomly fragmented it in fragment buffer. We employed a designated reverse transcriptase system to generate the initial cDNA strand, using mRNA fragments as templates and arbitrary oligonucleotides as initiators. We degraded the RNA strand with RNaseH and harnessed the double-stranded cDNA for the ligation of sequencing connectors. We filtered cDNA fragments ranging from 370 to 420 base pairs using Ampure XP beads for the purpose of PCR amplification. Subsequently, the purified PCR products were again refined with Ampure XP beads. Our constructed library was subjected to transcriptome sequencing via the illuminahiseq4000 platform. We performed a differential analysis of the obtained data, selecting genes with significant differences and identifying genes with P values of less than 0.05 as significant differential genes. We drew a volcano plot and performed an enrichment analysis using the Kyoto Encyclopedia of Genes and Genomes and Gene Ontology.

Intracellular ROS determination

The Hepa1-6 and Huh7 cells were digested, and we distributed 1 mL of culture medium, which comprised 1 × 106 cells, into each well of a 6-well plate. The plate was maintained at 37 °C in 5% CO2 for 24 h. Subsequently, we introduced complete cell culture media containing varying concentrations of donafenib into the cell plate, dividing it into distinct groups, for an additional 24 h. The process was completed according to the instructions in the Reactive Oxygen Specifications Assay Kit (Beyotime). We observed the results, took photos under an inverted fluorescence microscope and analyzed them on a flow cytometer. We used ImageJ software to calculate and analyze the fluorescence intensity. Statistical analysis using Prsim 9.0.

Western blotting analysis

The Hepa1-6 and Huh7 cells, which were processed under different conditions, were treated separately, and protein lysis was performed on ice. Once the cultivation process was finalized, it was necessary to cool the centrifuge to a temperature of 4 °C. Subsequently, we discarded the culture medium and rinsed the cells thoroughly three times using PBS. The cells were exposed to RIPA lysis solution for 15 min. The treated sample was transferred to a cooled centrifuge and centrifuged at 4 °C for 10 min. The supernatant obtained following centrifugation contained the desired protein.

We collected the supernatant, measured the protein concentration using a BCA reagent kit, and added a loading buffer (about a quarter of the volume of the protein in the new EP tube). Electrophoresis was performed at 90 V for 10 min and then at 130 V for 70 min. The membrane transfer time was adjusted according to the target protein. After the membrane transfer was completed, 5% skim milk was sealed. After 30 min at 37 °C, the membrane was washed at room temperature for 30 min using TBST. After the sealing, the membrane was washed 3 times for 5 min each time. The primary antibodies (SLC7A11, BAX, Bcl-2, GPX4, CD71, Caspase3, Caspase8, HIF-1α, PD-L1, p53, and β-actin) were incubated overnight at 4 °C or at room temperature for 1 h. The membrane was then washed three times: the first time for 10 min, the second for 7 min, and the third for 5 min. The secondary antibodies were incubated at room temperature for 1 h. The film was washed 3 times for 5 min each. Following exposure, ImageJ software is used to quantitatively analyze the grayscale values of target protein bands and calculate the relative expression levels of target proteins. Statistical analysis using Prsim 9.0.

Xenotransplantation experiments and immunohistochemistry

Five-week-old BALB/c nude mice were acquired from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All experimental surgical procedures and protocols undergone by these mice underwent stringent evaluation and received clearance from the Animal Care and Use Committee of Guangxi Medical University. The mice were housed in a pathogen-free environment with a controlled light–dark cycle of 12 h each. After the mice successfully acclimated to their new surroundings, 5 million Huh7 cells were administered via subcutaneous injection into the right side of each mouse, with 4 mice in each group, to establish liver cancer xenograft models.

Tumor sizes were measured and recorded once every 3 days, with 1/2 × (length × width 2) used as the formula to calculate tumor volumes. Taking as a starting point the average tumor volume, which was around 100 mm3, the mice were divided into four groups according to the different treatment methods [control group, donafenib L group (5 mg/kg), donafenib M group (10 mg/kg), and donafenib H group (20 mg/kg)], and the relevant drug, i.e., donafenib (100 μL) or an equal amount of physiological saline (100 μL), was intraperitoneally injected every 2 days. This procedure continued for 2–3 weeks. After the treatment, the mice were euthanized, and the tumors were removed for histological analysis. We performed IHC staining according to the manufacturer’s instructions (Zhongshan Jinqiao, Beijing, China). Quantify staining intensity and positive cell count using ImageJ analysis, and perform statistical analysis using Prsim 9.0.

HE staining

The tissue was soaked in 4% paraformaldehyde and fixed at 4 °C for more than 24 h. We used a paraffin-embedding machine to embed the tissue and cut it into 4-μm slices using a tissue-slicing machine. We let the specimen dry for 24 h at room temperature. After the wax was removed using xylene and ethanol, the paraffin-embedded tumor and normal tissue sections underwent rehydration. Subsequently, they were stained with eosin and hematoxylin. Finally, we examined and captured images of the slices under a microscope.

Statistical analysis

A statistical analysis was performed using the Prism 9.0 software program. The outcomes are expressed as mean values along with the standard deviation as derived from three or more observations. The statistical significance of the findings was determined by applying either Student’s t-test or a one-way analysis of variance. Statistical significance was set at a p-value < 0.05. All bioinformatic analyses were conducted using R software (4.1.2).