The liver cancer cell lines, HepG2 and Huh7, were obtained from American Type Culture Collection, and were cultured in DMEM and RPMI medium (Thermo Fisher Scientific, Inc.), respectively. Both cell lines were supplemented with 10% FBS (HyClone; GE Healthcare Life Sciences) and maintained in a humidified atmosphere containing 5% CO₂ at 37°C. Cell line authentication was performed by short tandem repeat profiling and interspecies contamination test (Applied Biological Materials Inc.). Cryopreserved normal primary human hepatocytes (PHH) derived from a pool of 5 donors were purchased from Sekisui XenoTech, LLC. and thawed with OptiThaw Hepatocyte Media (Sekisui XenoTech, LLC.) according to the manufacturer's protocol. Cells were seeded in 96-well BioCoat™ collagen I cellware (BD Biosciences) in OptiPlate hepatocyte media (Sekisui XenoTech, LLC.) for 4 h until sufficient confluency was reached. The media was then replaced with OptiCulture hepatocyte media containing Penicillin/Streptomycin (Sekisui XenoTech, LLC.) and the cells were further incubated for 48 h for the hepatotoxicity assays. The multi-kinase inhibitor, Sora, was ordered from LC Laboratories and dissolved in dimethyl sulfoxide (DMSO). Art was purchased from TGI Chemicals, Ltd. and dissolved in 100% ethanol.

Rabbit anti-VEGF receptor 2 (D5B1; cat. no. 9698; 1:1,000), rabbit anti-cleaved PARP (D64E10; cat. no. 5625, 1:1,000), rabbit anti-cleaved caspase 9 (D8I9E; cat. no. 9505; 1:1,000) and rabbit anti-GAPDH (14C10; cat. no. 2118; 1:2,000) were all purchased from Cell Signaling Technology Inc.. HRP conjugated goat anti-rabbit secondary antibody (cat. no. sc-2004; 1:1,000) was purchased from Santa Cruz Biotechnology Inc.

At 72 h prior to drug treatments, HepG2 and Huh7 cells (5×10³ cells/well) were seeded in 96-well plates. Cells were treated with Sora, starting at 40 and 20 µM for HepG2 and Huh7, respectively, and Art starting at 400 and 1,000 µM for HepG2 and Huh7, respectively, or the constant combination ratio of Sora to Art (1:10 in HepG2 and 1:50 in Huh7). To examine the hepatotoxicity of the combination of treatments, normal primary hepatocytes were treated in two constant combination ratios, 1:10 starting at 10 and 100 µM for Sora and Art, respectively and at ratio of 1:50 starting at 20 and 1,000 µM for Sora and Art, respectively. Cell viability was monitored using the CyQUANT Cell Proliferation Assay kit (Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. Cell morphology was observed under an inverted Nikon microscope and images were captured using a digital camera. Results from cells treated with individual drugs or the drug combination at a constant ratio (HepG2, Sora:Art, 1:10; Huh7, Sora:Art, 1:50) were processed, and drug combination and reduction indices and isobologram plots were calculated using the Calcusyn software v2.11 (Premier Biosoft International).

Protein extracts were prepared in 1% NP-40 lysis buffer (50 mM sodium fluoride, 1 mM orthovanadate, 10 mM iodoacetamide, 1 mM ethylenediaminetetraacetic acid, 0.25% sodium deoxycholate, 1 mM phenylmethylsulfonyl fluoride and protease inhibitor cocktail). Protein concentrations were determined by Pierce BCA protein assay kit (Thermo Fisher Scientific, Inc.). Cell protein lysates (50 µg) were separated by 10% SDS-PAGE gels and then transferred onto PVDF membranes. The membranes were blocked with Tris-buffered saline containing 3% BSA and 0.05% Tween-20 (TBST) for 30 min, followed by incubation with aforementioned primary antibodies overnight at 4°C and aforementioned secondary antibody for 1 h at room temperature. A total of three 5-min washes with TBST were performed after incubation with the primary or secondary antibodies. The protein expression was detected using RapidStep™ Enhanced Chemiluminescence (Merck KGaA). GAPDH was used as a reference protein.

HepG2 and Huh7 cells (20,000 cells/well) were seeded in triplicate in ImageLock 96-well plates (Essen Bioscience). Cells were treated with 1 mg/ml mytomycin C for 2 h in order to inhibit proliferation prior to wound scratching. Confluent cell layers were scratched using the Essen Bioscience Wound Maker in order to generate wounds that were 700–800 µm wide. Cells were washed twice with PBS and allowed to grow in 10% FBS growth medium with vehicle (DMSO or 100% ethanol), 2.5 µM Sora, 25 µM (HepG2)/125 µM (Huh7) Art or Sora + Art. Images were captured at 6 h intervals for 36 h using the IncuCyte ZOOM imaging system with time-lapse bright field microscopy (Essen Bioscience). Relative wound density was calculated based on the ratio of cell density in the wound/cell density outside the wound.

Liver cancer cell lines were plated into 6-well plates at a density of 300,000 cells/well and treated with vehicle, 2.5 µM Sora, 25 µM (HepG2)/125 µM (Huh7) Art or Sora + Art for 72 h. Cell cycle analysis with propidium iodide (PI) staining was performed according to a standard protocol. Briefly, drug-treated cells were harvested in cell suspensions in PBS and fixed in a final concentration of 70% ethanol on ice for 30 min, followed by incubation with the PI/Triton X-100 staining solution (50 mg/ml PI, 0.05% Triton X-100 and 100 mg/ml RNase A) at 37°C for 1 h. Data was acquired by collecting the area and width on a linear scale for the DNA channel in addition to forward scatter and side scatter. Relative DNA contents were analyzed using the FACSCanto II flow cytometer and BD FACSDiva software v5.0.3 (Becton, Dickinson and Company). Apoptotic cells were indicated by the percentage of subG₀/G₁ phase.

All values are presented as the means ± standard deviation of at least three independent experiments. Statistical analyses were performed to analyse differences in growth inhibition and apoptosis using one-way ANOVA followed by Tukey's post hoc analysis. The wound scratch assays were analysed using Friedman's test followed by the Dunn's post hoc test using GraphPad Prism v6.0.1 software (GraphPad Software, Inc.). P<0.05 (two-tailed) was considered to indicate a statistically significant difference.

To determine accurate IC₅₀ values of Sora and Art used in the present study, dose-response curves were initially generated for these drugs using the CyQUANT cell proliferation assay. This assay is a cell metabolic activity-independent and sensitive method, which is used to directly measure DNA content to quantify cells. The dose response curves of Sora in HepG2 and Huh7 cells were similar to those reported previously, with IC₅₀ values of 5.93–8.51 (mean 7.10) µM in HepG2 and 7.11–17.11 (mean 11.03) µM in Huh7 cells (Fig. 1; Table I) (20,21). It has been reported that the maximum drug concentration of Art in human plasma is 3,260 (1,020–164,000) ng/ml [8.48 (2.65–427.08) µM] and the terminal elimination half-life is 0.25 (0.1–1.8) h (22). To obtain clinically relevant concentrations, the inhibitory dose response of Art was tested for 72 h in the two cell lines at a starting concentration of 1,000 µM. This revealed that the IC₅₀ values of Art ranged between 63.28 and 99.85 (mean 79.49) µM in HepG2, and 344.70–1,099 (mean 615.40) µM in Huh7 cells (Fig. 1; Table I). The results indicated that HepG2 and Huh7 cells exhibited low sensitivity to Art treatment, compared with other types of cancer cell lines, including leukemia or colon cancer cell lines, which have been reported to be responsive in vitro to Art at 1.11±0.56 and 2.13±0.74 µM, respectively (4). However, the benefits of the drug combination include mutual enhancement of therapeutic effects, prevention of drug resistance and reduction of dose used per single treatment.

To leverage the synergistic inhibitory effects of Sora and Art, a constant combination ratio of Sora to Art (1:10 in HepG2 and 1:50 in Huh7) was used, based on IC₅₀ values of the individual drugs in liver cancer cell lines. The results demonstrated that the combination of Sora and Art significantly augmented cell growth inhibition compared with single drug treatments, the starting concentrations at which this effect became significant were 2.5 µM for Sora combined with Art at 25 µM for HepG2 and 2.5 µM Sora with 125 µM Art for Huh7 (Fig. 2). In addition, PHH were tested for the potential hepatotoxic effects associated with this drug combination. Due to the limitations of the utility of this model, the hepatotoxicity evaluation here is focused on cell viability. The cells were treated with the drug combination ratios of Sora to Art, 1:10 and 1:50, with Art starting at 100 µM and 1,000 µM, respectively. The former ration did not present hepatotoxicity in human normal primary hepatocytes, whereas with the latter ratio, Art treatment alone at higher concentrations (>250 µM), significantly reduced the cell viability by ~40% (the combination ratio at 1:50). Notably, no significant difference was observed between Art alone and the combination treatments, suggesting that the drug combination in PHH does not show augmented hepatotoxicity beyond the single drug treatments.

To further determine the types of drug interactions, the data were analyzed by median-drug effect analysis (Calcusyn) to determine antagonism [combination index (CI)>1], additivity (CI=1) and synergism (CI<1). The combinatorial effects of Sora and Art in HepG2 and Huh7 cells as identified by the CI, isobologram and dose reduction index (DRI) at three dose-effect levels of cell growth inhibition (IC₅₀, IC₇₅ or IC₉₀) are summarized in Table II. The combinations exhibited strong synergy in HepG2 cells (CI range, 0.11–0.22) or moderate to strong synergy in Huh7 cells (CI range, 0.75–0.43), which was also indicated by the isobolograms (Fig. 3; Table II). The isobolograms indicate the nature of the drug interaction at constant ratios in each liver cancer cell line. They present drug combination effects at the ED₅₀ (50% effective dose), ED₇₅ (75% effective dose), and ED₉₀ (90% effective dose) and the data points below the line, on the line or above the line suggest synergistic, additive or antagonistic effects, respectively (Fig. 3). Although there are differences between HepG2 and Huh7 cells in response to this drug combination which showed opposing trends in the CIs and DRIs (Table II), the CIs clearly indicated strong synergism in HepG2 and moderate synergism in Huh7 cells. Furthermore, the DRIs exhibited considerable dose reduction for Sora and Art as a result of synergism. When applied at the indicated constant ratio (Table II), the IC₅₀ of Sora was decreased 14.69-fold (HepG2) or 3.04-fold (Huh7). For Art, it was reduced 21.86-fold (HepG2) or 2.34-fold (Huh7). The dose reduction levels were specific to each cell line.

Additionally, a fixed constant ratio of drug combination (Sora:Art, 1:5) was tested in HepG2 and Huh7 cells for 48 h (Fig. S2). Art treatment at concentrations <100 µM demonstrated moderate effects on growth inhibition within 48 h (17–40%) for the two cell lines, while it exhibited synergistic effects with Sora on suppression of cell growth in HepG2 cells as indicated by CI (Fig. S2A). By contrast, concurrent treatment of Art and Sora in Huh7 cells for 48 h did not comply with the Calcusyn mathematical model since the median-effect curves in this cell line exhibited a negative slope due to its poor response to Art treatment (<100 µM; Fig. S2B). However, this was overcome by 24 h pre-treatment with Art, followed by the combination treatment (Fig. S2B), which suggested that sequential treatment of Art followed by Sora may improve the overall outcome of this drug combination in liver cancer treatment.

It has been previously reported that Art exerts anticancer properties in a variety of tumor types via the activation of mitochondrial apoptosis (17). Art also has been revealed to reduce tumor vessel formation via downregulation of VEGF and VEGFR protein expression levels (12,14). To further assess the biological effects of the combination of Sora and Art, concentrations (Sora:Art, 2.5:25 µM for HepG2; 2.5:125 µM for Huh7) at which the combination started exhibiting noticeable enhancement of growth inhibition at 72 h compared with the single drug treatments were selected (Fig. 2). This significantly reduced cell numbers, which was partially caused by cell apoptosis induction as indicated by the percentage of cells in the subG₀/G₁ phases of the cell cycle (Fig. 4A and B). Consistent with these results, western blot analysis revealed that the initiator and the final substrate of the caspase cascade, cleaved caspase-9 and cleaved poly ADP ribose polymerase expression levels were significantly increased following combination treatment. Sora and Art reduced VEGFR2 protein expression, and this effect was enhanced by drug combination in HepG2 and Huh7 cells (Fig. 4C).

Subsequently, the combined effect of the drugs on cell migration was determined using an in vitro wound healing assay (Fig. 5). Cell migration was kinetically monitored in Essen ImageLock plates using the IncuCyte Zoom imaging system. Cell growth was controlled by mitomycin C treatment. For the selected drug combination ratio (Sora:Art, 2.5:25 µM for HepG2 and 2.5:125 µM for Huh7), no significant apoptotic effects were observed at 24 h (data not shown). However, the results revealed a substantial inhibition of migration in the combination-treated cells compared with vehicle, Sora or Art treated cells (Fig. 5A and B).