EHHM was isolated and identified as described in a previous study (9). EHHM was dissolved in dimethyl sulfoxide (DMSO) at a stock solution of 50 mM, stored at −20°C and diluted with culture medium prior to each experiment. The human HCC cell line HepG2, which was obtained from the American Type Culture Collection (Manassas, VA, USA), was routinely maintained in Dulbecco's modified Eagle medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) with 10% fetal calf serum (Gibco; Thermo Fisher Scientific, Inc.) and incubated in a humidified atmosphere with 5% CO₂ at 37°C. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was purchased from Roche Diagnostics (Indianapolis, IN, USA). Trypan blue was purchased from Sangon Biotech Co., Ltd. (Shanghai, China). The pEGFPc1-LC3 plasmid was a gift from Professor Marja Jäättelä (Institute of Cancer Biology, Danish Cancer Society, Copenhagen, Denmark) (25). Benzyloxycarbonyl-Val-Ala-Asp (OMe) fluoromethylketone (Z-VAD-FMK) was purchased from Sigma-Aldrich (Merck Millipore, Darmstadt, Germany). Lipofectamine 2000 and FITC Annexin V/Dead Cell Apoptosis kit were purchased from Invitrogen (Thermo Fisher Scientific, Inc.). JC-1 Mitochondrial Membrane Potential Assay kit was purchased from Nanjing KeyGen Biotech Co., Ltd. (Nanjing, China). 4′,6-Diamidino-2-phenylindole and Hoechst 33258 were obtained from Sangon Biotech Co., Ltd. Bafilomycin A1 (Baf A1) and 3-methyl adenine (3-MA) were purchased from Sigma-Aldrich (Merck Millipore). ECL Western Blotting Substrate kit was purchased from Abcam (Cambridge, MA, USA). Antibodies were purchased from the following suppliers: Anti-β-actin antibody (cat. no. AF0003) from Beyotime Institute of Biotechnology (Haimen, China); anti-caspase 3 antibodies (cat. no. ab13585) from Abcam; and anti-light chain 3 (LC3; cat. no. 3868), anti-autophagy protein 5 (Atg5; cat. no. 12994) and anti-Beclin 1 (cat. no. 3495) antibodies from Cell Signaling Technology, Inc. (Danvers, MA, USA).

Cell proliferation was determined by MTT assay as described previously (26). Briefly, 90 µl cells (10⁵ cells) were seeded into a 96-well plate and precultured for 24 h, prior to be treated with EHHM for 24 or 48 h. Next, 10 µl of MTT (5 mg/ml) was added to each well and incubated for additional 4 h at 37°C. The medium was removed carefully, and 150 µl of DMSO was added into the plate, and then the 96-well plate was agitated for 10 min. The absorbance (A) was measured at 570 nm by an automated microplate reader (BioTek Instruments, Inc., Winooski, VT, USA). The cell death rate was calculated as follows: Cell death (%) = (mean A₅₇₀ of the control group - mean A₅₇₀ of the experimental group)/(mean A₅₇₀ of the control group - mean A₅₇₀ of the blank group) × 100%. Cell viability was estimated by trypan blue dye exclusion (27).

Colony formation in soft agar was tested by suspending 1×10³ cells in 1 ml of DMEM with 0.3% low melting point agarose (Merck Millipore) in the presence of different concentrations of EHHM, and then plating them on a bottom layer containing 0.6% agarose and 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.) in a 6-well plate in triplicate. After 15 days, the plates were stained with 0.1% gentian violet, and the colonies were counted under a light microscope (28,29).

Cell morphological changes in the nucleus were determined by Hoechst 33258 staining as described previously (30). Briefly, cells treated with different concentrations of EHHM were fixed with 4% polyoxymethylene for 10 min and then stained with Hoechst 33258 (1 mg/ml) at room temperature in the dark for 5 min. Morphological changes of nuclei were observed and imaged using a fluorescence microscope with an excitation wavelength of 330–380 nm. Cells with round nuclear morphology were considered normal, while those with condensed/fragmented and bright nuclei were regarded as apoptotic.

Cells were treated with EHHM, 3-MA or Baf A1 for the indicated concentrations and time points. Apoptotic cells were analyzed using annexin V-fluorescein isothiocyanate (FITC) /propidium iodide (PI) double staining methods as described previously (31). Briefly, HepG2 cells were treated with the indicated concentrations of EHHM in the presence or absence of 3-MA (5 mM) or Baf A1 (10 nM) for 24 h. Cells were digested into a single-cell suspension with ehylenediaminetetraacetic acid-free trypsin and stained according to the manufucturer's protocol of the FITC Annexin V/Dead Cell Apoptosis kit. The stained cells were immediately analyzed by flow cytometry. Cells undergoing early-stage apoptosis are stained with annexin V-FITC only (green fluorescence), while cells at late-stage apoptosis and necrotic cells are stained with both annexin V-FITC and PI (both green and red fluorescence) (32). In addition, to determine whether EHHM-induced apoptosis is caspase-dependent, HepG2 cells were pretreated with 20 mM Z-VAD-FMK for 1 h followed by treatment with 15 µM EHHM for 24 h, and the inhibition rate was determined by MTT assay, as described previously.

siRNA targeting human Atg5 (5′-CCUUUGGCCUAAGAAGAAATTdTdT-3′) was synthesized by Shanghai GenePharma Co., Ltd. (Shanghai, China), and a nonspecific siRNA served as a negative control. Cells seeded in 60-mm dishes were transfected with siRNAs for 36 h prior to treatment with EHHM.

Cell mitochondrial membrane potential changes were analyzed by JC-1 staining as described previously (33). Cells were incubated with or without EHHM for 24 h, followed by incubation with JC-1 working solution, and then the cells were incubated in 5% CO₂ for 20 min. Cells was washed twice with JC-1 incubation buffer, and then the cells labeled with JC-1 were detected with a fluorescence microscope. JC-1 can selectively enter the intact mitochondria with high membrane potential and form polymers to emit red fluorescence, whereas in mitochondria with low membrane potential, it remains in the monomeric form in the cytosol and emits a green fluorescence (16).

Cells were treated with different concentrations of EHHM in the presence or absence of 3-MA (5 mM) or Baf A1 (10 nM) for 24 h, and then lysed in lysis buffer [50 mM Tris-HCl (pH 6.8), 2% w/v sodium dodecyl sulfate (SDS), 10% glycerol, 10 mM dithiothreitol, 1 mM phenylmethane sulfonyl fluoride and 1X protease inhibitor cocktail (Roche, Basel, Switzerland)]. Equal amounts of protein samples were quantified using the RC DC ™ Protein Assay Kit I (Bio-Rad Laboratories, Hercules, CA, USA) and electrophoresed by 8–12% SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene fluoride membrane. The membranes were blocked with a 5% non-fat milk powder solution in Tris-buffered saline with Tween 20 (TBST) buffer [50 mM Tris-HCl (pH 7.6), 150 mM NaCl and 0.05% Tween 20] for 1 h at room temperature and then incubated with the corresponding primary antibodies (1:1,000) overnight at 4°C. Upon washing with TBST, membranes were incubated with horseradish peroxidase-conjugated anti-mouse (cat. no. 31430; 1:5,000) and anti-rabbit (cat. no. 31460; 1:5,000) IgG secondary antibodies (Thermo Fisher Scientific, Inc.) for 2 h at room temperature and detected with ECL Western Blotting Substrate kit (Thermo Fisher Scientific, Inc.) (34).

Cells plated in 6-well plates were transfected with the pEGFPc1-LC3 plasmid using Lipofectamine 2000. After 24 h of transfection, cells were treated with different concentrations of EHHM in the presence or absence of autophagy inhibitors, 3-MA (5 mM) or Baf A1 (10 nM), for 12 h. The patterns of GFP-labeled LC3 in transfected cells were examined and imaged by fluorescence microscopy.

A total of 24 female 6–7 week-old nude immunodeficient mice (nu/nu) (15–20 g) were purchased from the Guangdong Province Medical Animal Center (Guangzhou, China) and bred in the laboratory animal facility of Tsinghua University (Shenzhen, China). The mice were housed at 25°C and exposed to 12 h light/dark cycles with free access to food and water. All animal studies were approved by the Institutional Animal Care and Use Committee of Tsinghua University. The mice were subcutaneously injected with HepG2 cells (2.0×10⁶ cells) in 100 µl of DMEM without serum. The mice were randomly divided into four groups (6 mice/group) when the tumors reached a palpable size and then treated with vehicle control, 3-MA (24 mg/kg) and EHHM (50 mg/kg), alone or in combination with 3-MA, every 2 days. Vernier caliper measurements of the longest perpendicular tumor diameters were performed every 2 days to estimate the tumor volume with the following formula: 4π/3 × (width/2)² × (length/2), which represents the three-dimensional volume of an ellipsoidal tumor tissue (35).

Data from three independent experiments were expressed as the mean ± standard deviation, unless otherwise noted. The Student's t test was used to evaluate the variables in the present study. P<0.05 was considered to indicate a statistically significant difference.

EHHM is a novel plant metabolite isolated from the fruit of L. chinensis (Fig. 1A). In the present study, the cytotoxicity of EHHM towards human HepG2 cells was examined. By MTT assay, it was demonstrated that EHHM had a moderate cytotoxicity towards HepG2 cells, with a half maximal inhibitory concentration of 41.62±4.57 µM (Fig. 1B). Meanwhile, using trypan blue dye exclusion analysis, it was demonstrated that EHHM inhibited the proliferation and growth of HepG2 cells in a time- and dose-dependent manner (Fig. 1C). Next, the effect of EHHM on colony formation was tested using soft agar colony formation assays. The results revealed that the clonogenic activity of HepG2 cells was significantly suppressed (Fig. 1D and E). These results suggested that EHHM can significantly inhibit the proliferation and growth of HCC cells in a time- and dose-dependent manner.

It was tested whether EHHM can induce the apoptosis of HCC cells. By Hoechst 33258 staining, HepG2 cells treated with the indicated concentrations of EHHM exhibited chromatin condensation and fragmentation, which are typical apoptotic nuclear morphological changes (Fig. 2A). Meanwhile, flow cytometry assay using annexin V/PI staining also confirmed that treatment with EHHM induced apoptosis in HepG2 cells in a time- and dose-dependent manner (Fig. 2B). Caspase 3, a critical executor of apoptosis, is either partially or totally responsible for the proteolytic cleavage of numerous key proteins such as the nuclear enzyme poly (ADP-ribose) polymerase (PARP) (36). The inactive full-length caspase 3 is processed during apoptosis, generating an activated 17-kDa fragment termed cleaved caspase 3 (37). In the present study, both cleaved caspase 3 and cleaved PARP were detected subsequent to EHHM treatment at different concentrations (15, 45, 90 and 150 µM), and a reduction in pro-caspase 3 (through cleavage) was observed (Fig. 2C). To further confirm that EHHM-induced cell apoptosis is caspase-dependent, HepG2 cells were pretreated with Z-VAD-FMK (20 mM), a pan-caspase inhibitor, for 1 h and then treated with 15 µM EHHM for additional 24 h. As shown in Fig. 2D, EHHM-induced apoptosis was significantly suppressed. Since the activation of caspase 3/caspase 9 involves mitochondria permeability changes and the release of cytochrome c, the present study next examined changes in the mitochondrial transmembrane potential following EHHM treatment. As shown in Fig. 2E, in EHHM-treated HepG2 cells, JC-1 remained in the cytoplasm in its monomeric form and emitted green fluorescence, whereas in the cells without EHHM treatment, JC-1 accumulates and aggregates in the mitochondria, emitting red fluorescence. This result suggested that EHHM treatment damaged the mitochondrial transmembrane of HepG2 cells. Meanwhile, cytochrome c release from mitochondria was also investigated. As shown in Fig. 2C, a significant increase in released cytochrome c was detected at 24 h after 15-µM EHHM treatment.

Overall, these results suggested that EHHM treatment significantly suppressed HepG2 proliferation and induced apoptosis in a time- and dose-dependent manner, and that EHHM-induced apoptosis in HepG2 cells is mediated via mitochondria-dependent caspase cascade activation.

Notably, several vacuoles were observed in EHHM-treated HepG2 cells, but not in untreated cells, by light microscopy, indicating that EHHM treatment may be associated with cell autophagy.

Autophagy, a cellular pathway involved in protein and organelle degradation, plays complicated roles in cancer development (38). During the autophagy process, LC3-I is converted into LC3-II via a ubiquitin-like system involving Atg7 and Atg3, which promotes LC3-II relocalization to autophagic vesicles (39). The conversion of LC3-I into LC3-II and the presence of LC3-II in autophagosomes have been used as indicators of autophagy activation (40). To test whether EHHM induces autophagy formation, HepG2 cells transfected with GFP-LC3 were treated with EHHM, and autophagic vacuoles were detected. As shown in Fig. 3A, the number of autophagic vacuoles with a punctate staining pattern of GFP-LC3-II was significantly increased in EHHM-treated cells compared with untreated cells. Accordingly, the protein levels of LC3-II were also significantly increased in HepG2 cells treated with EHHM in a dose-dependent manner (Fig. 3B). In addition, several studies have demonstrated that Beclin 1 participates in the initiation of the autophagosome formation, and that Atg5, an E3 ubiquitin ligase, is important in autophagosome elongation (41,42). Therefore, the expression levels of these two proteins were also investigated in EHHM-treated cells. As shown in Fig. 3B, the expression levels of Atg5 and Beclin 1 proteins in cells exposed to EHHM were also increased in a dose-dependent manner.

The phosphatidylinositol 3-phosphate kinase (PI3K)/Akt/p70 ribosomal protein S6 kinase (p70S6K) signaling pathway is known to regulate autophagy, growth, proliferation and survival in response to a variety of stresses in mammalian cells (43). To investigate whether the Akt/mTOR/p70S6K signaling pathway was involved in EHHM-induced apoptosis and autophagy, western blotting was performed to detect the expression levels of the associated proteins. As shown in Fig. 3C and D, EHHM inhibited the phosphorylation of Akt, mTOR and p70S6K in a time- and dose-dependent manner.

Taken together, these data indicated that EHHM was able to induce autophagy in HepG2 cells, which involves the Akt/mTOR/p70S6K signaling pathway.

As demonstrated above, EHHM treatment induces cell apoptosis and autophagy in HepG2 cells. As previously reported, autophagy may serve different roles in response to different cellular stresses, protecting or killing the cells (44). To identify which type of autophagy is induced by EHHM, HepG2 cells were transfected with Atg5 siRNA (Fig. 4A), and the proliferation effect of EHHM on HepG2 cells was evaluated using an MTT assay. As illustrated in Fig. 4B, knockdown of Atg5 expression significantly enhanced the inhibition effect of EHHM on HepG2 cells compared with control siRNA. Simultaneously, knockdown of Atg5 resulted in the decrease of LC3-I/II conversion and the increase in cleaved caspase 3 expression (Fig. 4C). A flow cytometry assay further confirmed the above results. As shown in Fig. 4D, knockdown of Atg5 significantly enhanced EHHM-induced apoptosis of HepG2 cells.

Furthermore, the effects of autophagy specific inhibitors (3-MA and Baf A1) on EHHM-induced growth inhibition and apoptosis were investigated. 3-MA inhibits autophagy via blocking the autophagosome formation in the early stage of autophagosome formation (45), whereas Baf A1, a macrolide antibiotic derived from Streptomyces griseus, inhibits autophagy by blocking the fusion of autophagosomes with lysosome in the late stage of autophagy (46). As shown in Fig. 5A, the EHHM-induced upregulation of Atg5 was inhibited by 3-MA and Baf A1 treatment. As expected, the accumulation of GFP-LC3-II on autophagic vacuoles induced by EHHM was also decreased following 3-MA treatment (Fig. 5B and C). Accordingly, 3-MA pretreatment significantly promoted EHHM-induced growth inhibition and apoptotic cell death in HepG2 cells (Fig. 5D and E). Baf A1 treatment markedly promoted the accumulation of GFP-LC3-II autophagic vacuoles in EHHM-treated cells, indicating that Baf A1 inhibited the progression of autophagy by blocking the fusion of autophagosomes with lysosome (Fig. 5B). Similar to 3-MA, Baf A1 pretreatment also significantly promoted EHHM-induced growth inhibition of HepG2 cells (Fig. 5D). These results suggested that EHHM-induced autophagy is a protective response to cell death, and that inhibition of this type of autophagy can enhance the anti-HCC effect of EHHM.

To further evaluate the anti-tumor effect of EHHM in vivo, 2.0×10⁶ HepG2 cells suspended in 100 µl of serum-free DMEM were subcutaneously inoculated into nude mice to generate xenografted murine models. When the tumors could be measured, the tumor-bearing mice were randomly divided into four groups (n=6 mice/group) and administered vehicle control, 3-MA (24 mg/kg) or EHHM (50 mg/kg), alone or in combination with 3-MA, every 2 days. As shown in Fig. 6A, the tumor growth was markedly inhibited by EHHM compared with the vehicle control and 3-MA groups, and 3-MA treatment significantly promoted the inhibitory activity of EHHM. By contrast, the above treatments did not significantly reduce the mice body weight, which suggested that EHHM plus 3-MA had no apparent side effects (Fig. 6B). These results preliminarily suggested that EHHM has a potent antitumor activity both in vitro and in vivo, and that the combination of EHHM and autophagy inhibitors provides synergistic anti-proliferative and pro-apoptotic effects in human HCC cells.