The PF extract used was from the same batch as that in a previous study (36). Briefly, the dry immature fruits of P. trifoliata were purchased from an Oriental medical store in Geuman, Korea. The fruits were processed with methanol at room temperature for 24 h. The PF methanol extract was concentrated under reduced pressure using a rotary evaporator. The residues were resuspended in distilled water and further extracted with hexane and ethyl acetate. The ethyl acetate-soluble fraction was subjected to silica gel chromatography using the phase solvent method and a chloroform-methanol solvent. The active fractions were combined, concentrated, and subjected to chromatography on a Sephadex LH-20 column eluted with methanol. Again, the active fraction was subjected to C18 reverse-phase column chromatography eluted with 80% aqueous methanol followed by high-performance liquid chromatography (HPLC) eluted with 85% aqueous methanol to obtain the desired compound.

Hep3B and Huh7 cells were obtained from the American Type Culture Collection (ATCC). Cell line authentication was systematically conducted using a panel of ATCC short tandem repeats (STR) and routinely monitored for mycoplasma contamination and tested negative for mycoplasma contamination. The cell lines were cultured in Dulbecco's Modified Eagle's Medium (DMEM; HyClone^®, Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS; HyClone^®, Thermo Fisher Scientific, Inc.) and 1% antibiotics (HyClone^®, Thermo Fisher Scientific, Inc.) at 37°C in humidified conditions with 5% CO₂. The cells were treated for 24 h with PF dissolved in phosphate-buffered saline (PBS) (0, 20, 30, or 40 µM).

Hep3B and Huh7 cells were seeded into 96-well plates at a cell density of 5×10³ per well and cultured for 24 h. The cells were treated with the indicated drug concentrations (0, 20, 30, 40 µM). The medium was removed, and the cells were incubated for 2 h in solution with 50 µl of 5 mg/ml 2,5-diphenyl tetrazolium bromide (MTT; Sigma-Aldrich; Merck KGaA). The transformed purple formazan crystals were solubilized in dimethyl sulfoxide (DMSO) and the optical density was measured at a wavelength of 575 nm.

Hep3B and Huh7 cells were seeded in 6-well plates (1,000 cells per well) for 24 h. The medium was replaced with fresh medium containing PF (20, 30, or 40 µM) for 24 h. After 24 h, the medium was removed, and fresh medium was added to the cells for 10–14 days. Afterward, the cells were fixed for 15 min with ice-cold methanol, washed with PBS, stained with 1% crystal violet solution (bioWORLD) for 1 h at room temperature, and the colonies with >10 cells were counted by using densitometric software Clono-Counter as described previously (37).

Apoptosis of the Hep3B and Huh7 cells was detected by flow cytometry. Briefly, Hep3B and Huh7 cells were seeded at a cell density of 5×10⁴ in a 60-mm culture dish overnight and treated with medium containing PF (0, 20, 30, and 40 µM) for 24 h. The cells were trypsinized and the supernatants containing the cells were collected. The cells were centrifuged at 1,500 rpm for 3 min at 15°C and frequently washed with PBS. Then, an Annexin V-FITC apoptosis kit was used to determine apoptosis by flow cytometry according to the manufacturer's instructions.

The mitochondrial membrane potential was determined using a tetramethylrhodamine ethyl ester perchlorate (TMRE) mitochondrial membrane potential assay kit (Abcam) according to the manufacturer's instructions. Hep3B and Huh7 cells were seeded at a density of 10,000/well in 96-well plates. The cells were cultured with PF (0, 20, 30, and 40 µM) for 24 h. Then, the cells were stained with TMRE (400 nmol/l) for 20 min and washed twice with PBS. The TMRE dye intensity was measured at excitation and emission wavelengths of 549/575 nm, respectively, using a fluorescence plate reader.

The levels of intracellular ROS in the Hep3B and Huh7 cells were measured using a fluorescent dihydroethidium (DHE) probe. Intracellular DHE is oxidized to ethidium, which binds to DNA and stains the nuclei a bright red fluorescent color. Hep3B and Huh7 cells were cultured in glass-bottom confocal dishes and pretreated with PF as previously described. The cells were washed twice with PBS and then fixed at room temperature with ice-cold methanol for 10–15 min. Then, DHE was diluted in PBS to a final concentration of 5 µM and applied to the cells for 35 min at 37°C in the dark. The cells were labeled with 4′,6-diamidino-2-phenylindole, dihydrochloride (DAPI) for 5 min after washing twice with PBS, and examined under a fluorescence microscope using a ×63 oil immersion objective lens (Axioskop 2 Plus, Carls Zeiss, Gottingen, Germany).

The migration and invasion potential of the Hep3B and Huh7 cells was evaluated using a Transwell chamber with 8.0-µm pore polyester membrane inserts and 8.0-µm pore polyester membrane inserts with Matrigel-coated chambers (Corning, Inc.). For the migration assay, Hep3B and Huh7 cells treated with PF at the indicated concentrations were seeded in the upper chamber at a density of 5×10⁴ cells per well in 200 µl of serum-free medium, and 500 µl of 10% FBS-containing medium was added to the lower chamber. The suspension was discarded after 24 h and the cells on the top layer of the membrane were removed with cotton swabs. The cells that migrated to the lower membrane surface were stained with the Diff-Quick kit solution. The migrated cells were counted under light microscopy in three randomly chosen fields. The Matrigel-coated chamber was rehydrated with serum-free medium in a humidified tissue culture incubator to perform the invasion assay, followed by the migration assay. After 24 h, the suspension medium was discarded and the cells on the upper layer of the membrane were removed using cotton swabs. The cells that invaded the lower surface of the membrane were fixed with ice-cold methanol and stained with the Diff-Quick kit solution. The invasive cells were detected, photographed, and counted in three randomly chosen areas under light microscopy.

The cells were washed with PBS and lysed with lysis buffer (iNtRON Biotechnology) containing a cocktail of phosphatase inhibitor-1. Western blotting was performed as previously described (38). The membranes were incubated overnight with primary antibodies specific for anti-rabbit cleaved caspase-3 (cat. no. 9661) and −9, anti-mouse caspase-9 (cat. no. 9508), cleaved PARP (cat. no. 5625), E-cadherin (cat. no. 3195), vimentin (cat. no. 5741), Snail (cat. no. 3879), MMP-9 (cat. no. 2270) (dilution 1:1,000; Cell Signaling Technology, Inc.), anti-mouse actin, α-SMA (cat. no. A2547, dilution 1:1,000; Sigma-Aldrich; Merck KGaA), anti-rabbit Bcl-2 (cat. no. BS1511), Bax (cat. no. BS6420), MMP-2 (cat. no. BS1236) (dilution 1:1,000; Bioworld Tech. Inc.), anti-rabbit caspase-3 (cat. no. sc-7148), anti-mouse Pin1 (cat. no. sc-46660), anti-rabbit cyclin D1 (cat. no. sc-717) (dilution 1:1,000, Santa Cruz Biotechnology, Inc.), anti-mouse β-catenin (cat. no. 610154), and N-cadherin (cat. no. 610920) (dilution 1:1,000, BD Biosciences). The membranes were washed and incubated with horseradish peroxidase (HRP)-conjugated anti-mouse (cat. no. ADI-SAB-100-J) or anti-rabbit secondary antibodies (ADI-SAB-300-J) (dilution 1:3,000, Enzo Life Sciences). Protein expression was detected using an enhanced chemiluminescence detection kit (Millipore Corp.).

Gelatin zymography was performed to determine the levels of MMPs, such as MMP-2 and MMP-9. The HCC cell supernatants were collected by centrifugation at 4,000 × g at 4°C and the protein content was determined using a BCA protein assay kit. Equal amounts of protein mixed with non-reducing 6X loading buffer were separated by SDS-PAGE containing 0.1% gelatin. After electrophoresis, the gels were rinsed thrice with 0.25% Triton X-100 for 15 min at room temperature and then incubated at 37°C in developing buffer (50 mM Tris-HCL pH 7.4, 5mM CaCl₂, 200 mM NaCl) for 42 h. After incubation, the MMP-2 and MMP-9 activities were measured by Coomassie blue staining and the gel was imaged using an LAS 3000 (Fuji) (39).

All the experimental results were obtained by repeating each experiment at least three times and data are expressed as means ± standard deviation. Statistical significance was analyzed by one-way analysis of variance (ANOVA) using the Dunnett's test or Turkey's post hoc test in Prism 7 software (GraphPad Software, Inc.). A P-value less than 0.05 (P<0.05) was considered statistically significant.

MTT and colony-forming assays were performed to evaluate the anti-proliferative effects of PF in Hep3B and Huh7 cells. As shown in Fig. 1A, PF significantly reduced the cell proliferation of Hep3B and Huh7 cells in a dose-dependent manner. The inhibitory effect of PF on cell proliferation was also measured using a clonogenic assay. As shown in Fig. 1B and C, PF significantly inhibited the colony formation of Hep3B and Huh7 cells in a dose-dependent manner. These results indicate that PF had a significant inhibitory impact on the viability and proliferation of HCC cell lines.

To explore whether PF induced cancer cell apoptosis, flow cytometry was conducted. As shown in Fig. 2A and B, treatment with PF significantly induced apoptosis in the Hep3B and Huh7 cells in a dose-dependent manner compared to the control.

The present study demonstrated the mechanism of PF-induced apoptosis on HCC cells (Hep3B and Huh7) by analyzing the apoptosis-related proteins, and quantifying the mitochondrial membrane potential (ΔΨm) and intracellular ROS levels. As shown in Fig. 3A, western blot analysis revealed that PF significantly decreased the expression of procaspase-3, procaspase-9, PARP and anti-apoptotic protein Bcl2, whereas it increased the expression of apoptotic proteins Bax, cleaved caspase-3, cleaved caspase-9 and cleaved PARP in a dose-dependent manner. The loss of mitochondrial membrane potential (ΔΨm) results in mitochondrial membrane permeabilization, which is regarded as an important hallmark of early apoptosis and plays a key role in the intrinsic apoptotic pathway (40–42). Therefore, we further analyzed the disruption of the mitochondrial membrane by PF. As shown in Fig. 3B, PF induced a significant, concentration-dependent depletion of mitochondrial membrane potential. Mitochondria are considered the main source of ROS, which is considered a trigger of apoptosis (43). Therefore, we determined the intracellular ROS generation in Hep3B and Huh7 cells by assessing the DHE fluorescence intensity after PF treatment. As shown in Fig. 3C, PF treatment induced stronger fluorescence intensity in Hep3B and Huh7 cells compared to the control, which indicated increased levels of ROS. These results suggest that PF induced apoptosis of the Hep3B and Huh7 cells by effects on the mitochondrial apoptosis pathway.

The migratory and invasive abilities of tumor cells are considered as significant hallmarks of HCC and play a key roles in the metastasis of various tumors (44). To investigate whether PF exhibits anti-migratory and anti-invasion effects, Transwell assays were performed. As shown in Fig. 4A and B, the migratory ability of Hep3B and Huh7 cells was inhibited by PF treatment in a dose-dependent manner. Consistently, the Transwell invasion assay showed that PF treatment inhibited the number of invasive cells (Fig. 4C and D).

EMT is vital in tumor cell growth, causing the invasion and metastasis of malignant tumors (45). Therefore, the protein levels of Pin 1, cyclin D1, and EMT markers such as β-catenin, E-cadherin, vimentin, N-cadherin, Snail, and α-SMA hepatic cancer cells were evaluated. As shown in Fig. 5A, the treatment of Hep3B and Huh7 cells with PF markedly increased the expression of E-cadherin, whereas it decreased the levels of Pin1, cyclin D1, β-catenin, N-cadherin, vimentin, Snail, and α-SMA in a dose-dependent manner. These results suggested that PF inhibited cancer cell progression by decreasing the EMT process. A previous study demonstrated that matrix metalloproteinases (MMPs) such as MMP-2 and MMP-9 play crucial roles in basement membrane degradation during tumor cell invasion (46). Therefore, we determined the expression of MMP-2 and MMP-9 in Hep3B and Huh7 cells. As shown in Fig. 5B, PF treatment markedly downregulated the protein levels of MMP-2 and MMP-9. To confirm whether the activities of extracellular MMPs were inhibited by PF, a gelatin zymography assay was performed. As shown in Fig. 5C, PF decreased the activities of MMP-2 and MMP-9 in the Huh7 culture medium. These results suggest that PF inhibits cancer cell migration and invasion by attenuating the activities of the metalloproteinases.