Human Huh7 HCC cells were obtained from the Chinese Academy of Sciences Cell Bank (Shanghai, China). The cells were immediately expanded, and multiple aliquots were cryopreserved and used within 3 months of resuscitation. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Hyclone Labotatories; GE Healthcare Life Sciences, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Biowest, Nuaillé, France). The antibodies (Abs) against PERK, IRE1α, binding immunoglobulin protein (Bip), Calnexin, ER oxidoreductin-1-like (Ero1-L)α, protein disulfide isomerase (PDI), C/EBP homologous protein (Chop), ubiquitin specific peptidase 2 (USP2), itchy E3 ubiquitin protein ligase (ITCH), microtubule-associated protein 1 light chain 3 (LC3), P62, Autophagy related 5 (Atg5), cFLIP, caspase-8, cleaved caspase-8, caspase-9, caspase-3, cleaved caspase-3 and poly(ADP-ribose) polymerase (PARP) were purchased from Cell Signaling Technology, Inc. (Danvers, MA, USA). Ab against caspase-12 was from Abcam (Cambridge, MA, USA). The anti-β-actin Ab was from Sigma-Aldrich; EMD Millipore (Billerica, MA, USA). Sorafenib, 3-methyladenine (3-MA) and chloroquine phosphate (CQ) were from Selleck Chemicals (Houston, TX, USA). 4-phenylbutyric acid (4-PBA) was purchased from Sigma-Aldrich; EMD Millipore.

Sorafenib was dissolved in dimethyl sulfoxide to prepare a stock solution of 20 µmol/l. CQ was dissolved in phosphate-buffered saline (PBS) to prepare a stock solution of 50 mmol/l. 3-MA was dissolved in PBS at a stock concentration of 100 mmol/l by heating to 60–70°C immediately prior to use. The Annexin V-FITC/propidium iodide (PI) apoptosis detection kit was from Abcam. The MTT was purchased from Merck Millipore (Darmstadt, Germany) dissolved in sterile PBS at a stock solution of 5 mg/l. Lipofectamine 2000 was purchased from Thermo Fisher Scientific, Inc. (Waltham, MA, USA).

The cytotoxicity of sorafenib towards HCC cells was initially detected by treating the cells with gradually increasing concentrations of sorafenib in 96-well plates (10⁴ cells/well) at 37°C, with cell proliferation measured at 24, 48 and 72 h, respectively. Into the medium of cells, the indicated concentrations of sorafenib were added, which were just below their respective IC₅₀. The concentration of sorafenib was slowly increased by 0.25 µmol/l every week. After 25–30 weeks, the sorafenib-resistant cell line, termed Huh7-SR was obtained and was continuously maintained by culture in the presence of sorafenib.

The cytotoxicity of sorafenib towards the cells was determined using MTT, as previously described (16).

The cell death and apoptosis were evaluated by flow cytometry using the Annexin V/PI binding kit (Abcam). Briefly, following sorafenib treatment, the cells were trypsinized, stained with Annexin V/PI, and then analyzed with a flow cytometer.

A double-stranded siRNA targeting human cFLIP (5′-GCCUCAGAGCAUACCUGAATT-3′ and 5′-UCAUCUCGUACAUGACCACTT-3′) was produced by Shanghai GenePharma Co., Ltd. (Shanghai, China) The non-specific scrambled siRNA (5′-UUCUCCGAACGUGUCACGUTT-3′ and 5′-ACGUGACACGUUCGGAGAATT-3′) served as a control. The USP2cDNA clone plasmid (cat. no. RC200273) and empty vector plasmid (cat. no. PS100001) were purchased from OriGene Technologies, Inc. (Rockville, MD, USA). The transfection procedure was performed according to the previously described protocol (16).

The protein extracts were obtained by suspending the cells in RIPA lysis which containing protease inhibitor cocktail (Pierce; Thermo Fisher Scientific, Inc.). The protein concentrations were determined using a Bradford assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA) and 30 µg of cellular proteins were electroblotted onto a PVDF membrane following separation via 10% SDS polyacrylamide gel electrophoresis. The immunoblot was blocked for 1 h with 5% milk at room temperature, followed by incubation overnight at 4°C with 1:1,000 dilutions of primary antibodies against PARP, caspase proteins, ERS proteins, cFlip, p62, Atg5, LC3, ITCH, USP2 or β-actin. The blots were washed twice with Tween-20/Tris- buffered saline (TTBS) prior to the addition of a 1:1,000 dilution of HRP-conjugated secondary antibody (cat. no. 7074; Cell Signaling Technology Inc.) for 1 h at room temperature. The blots were washed again with TTBS, and developed by enhanced chemiluminescence using Supersignal West Femto chemiluminescent substrate (Pierce; Thermo Fisher Scientific, Inc.). The band intensities were quantified using UN-SCAN-IT gel analysis software (version 6; Silk Scientific, Inc., Orem, UT, USA). The OD values for the target proteins were calculated as a proportion of the OD value for β-actin. The western blot assays were repeated three times (16).

The cells were collected and fixed with 2.5% glutaraldehyde solution for 1 h, then with 1% osmic acid for 1 h, following which they were dehydrated with a graded series of ethanol and embedded. Ultrathin sections (60–80 nm) were cut using the LKB ultrotome with a diamond knife and double-stained with uranium acetate and lead citrate. The cells were observed under a transmission electronic microscope (HT7700; Hitachi, Ltd., Tokyo, Japan).

The Huh7-R cells were transfected with the USP2 plasmid for 48 h, followed by sorafenib (5 µM) treatment for 6 h, and the cells were washed three times with PBS. The pelleted cells (300 × g for 5 min at room temperature) were then prepared and analyzed by immunoprecipitation, as previously described (17).

Data are expressed as the mean ± standard deviation. Statistical significance was determined using SPSS 19.0 for Windows (IBM Corp., Armonk, NY, USA. Statistical comparison was performed using Student's t-test (unpaired). P<0.05 was considered to indicate a statistically significant difference.

Following incubation with 10 µM of sorafenib for 48 h, the cytotoxicity of Huh7-SR cells was 18.7%, respectively, which was significantly lower than that of the Huh7 parent cells (54.6%; Fig. 1A), as demonstrated by the MTT assay. When the concentration of sorafenib reached 20 µM, the cytotoxicity of the Huh7-SR cells was 45.9%, whereas almost complete Huh7 parent cell death was observed (Fig. 1A). The apoptotic rates of the Huh7 cells were ~4.2-, 3.4- and 2.3-fold higher than those of the Huh7-SR cells, following exposure to 5, 10 and 15 µM sorafenib, respectively (Fig. 1B). The apoptotic results were further supported by the expression of two key apoptotic proteins, caspase-3 and PARP (Fig. 1C), showing that PARP and caspase 3 were significantly activated in the Huh7 cells following 10 µM sorafenib treatment, compared with the Huh7-SR cells.

To investigate the role of ERS in sorafenib resistance, the Huh7 and Huh7-SR cells were incubated with 5, 10 and 20 µM sorafenib for 24 h, or with 10 µM sorafenib for 0, 3, 6, 12 or 24 h. The protein expression levels of PERK, IRE1α, PDI, Calnexin, Bip and Chop, the key molecules involved in ER stress, were detected. As shown in Fig 2A, the expression levels of IRE1α, Ero1-Lα, and Bip were higher in the Huh7-SR cells than those in the Huh7 cells. Following treatment with different concentrations of sorafenib, these proteins remained high in the Huh7-SR cells compared with the Huh7 cells. The apoptosis-related protein Chop was significantly lower in the Huh7-SR cells than in the Huh7 cells following sorafenib treatment. Similarly, the expression of IRE1α, Ero1-Lα, and Bip were higher following 10 µM sorafenib treatment for different times in the Huh7-SR cells than those in the Huh7 cells. Chop was increased in a time-dependent manner following sorafenib treatment in Huh7 cells, but not in Huh7-SR cells (Fig. 2A). Treatment with the ERS inhibitor 4-PBA + sorafenib inhibited the expression of IRE1α and Ero1-Lα, and increased the expression of Chop, compared with the sorafenib alone treatment group of Huh7-SR cells (Fig. 2B). In addition, 4-PBA increased sorafenib-induced Huh7 cell death (Fig. 2C). The results were supported by the expression of apoptotic proteins, PARP, caspase-3, −8 and −9, which were activated following 4-PBA + sorafenib treatment (Fig. 2D). However, ERS-induced apoptosis-related caspase-12 was inhibited following 4-PBA treatment.

P62, Atg 5 and LC3 are autophagy-related proteins. As shown in Fig. 3A, P62 was decreased, whereas Atg5 and the ratio of LC3-II/LC3-I were increased prior to and following treatment with sorafenib in Huh7-SR cells, compared with Huh7 cells. Co-incubation with the ERS inhibitor (4-PBA) inhibited the sorafenib-induced increase of LC3-II/LC3-I in the Huh7-SR cells (Fig. 3B). Electron microscopy showed the same results. Double membrane vacuolar structures with morphological features of autophagosomes and dilated ER lumens co-existed in the Huh7-SR cells and sorafenib-treated cells, but not in the 4-PBA- and 4-PBA + sorafenib-treated cells (Fig. 3C). In addition, co-incubation with lysosomal protease inhibitor (CQ) targeting the final steps of autophagic degradation, and 3-MA, a PI3K inhibitor, enhanced sorafenib-induced Huh7-SR cell apoptosis (Fig. 3D).

cFLIP is one of the specific inhibitors of caspase-8 (18). As shown in Fig. 4A, the level of cleaved caspase 8 was decreased in the Huh7-SR cells following sorafenib treatment, compared with that in the Huh7 cells. It was also found that cFLIP was continuously increased in the Huh7-SR cells following sorafenib treatment (Fig. 4B). Whether the knockdown of cFLIP by siRNAs can increase sorafenib-induced Huh7-SR and cytotoxicity and apoptosis was then examined. As shown in Fig. 4C, transfection of the Huh7-SR cells with cFLIP siRNA markedly reduced the expression of IRE1a prior to and following sorafenib treatment, and increased the expression of Chop following sorafenib treatment (Fig. 4C). Sorafenib-induced caspase-8 activation was also significantly increased in the Huh7-SR cells transfected with cFLIP siRNA (Fig. 4D). Sorafenib-induced Huh7-SR cell apoptosis was effectively increased in cells transfected with cFLIP siRNA (Fig. 4E), compared with control siRNA transfection.

A major post-transcriptional mechanism that controls protein levels in cells is ubiquitin-dependent proteasomal turnover (19). To determine whether the proteasome is involved in sorafenib-induced cFLIP decay, Huh7 cells were incubated with MG132, a specific proteasome inhibitor. As shown in Fig. 5A, pretreatment with MG132 significantly prevented sorafenib-induced cFLIP degradation in the Huh7 cells (Fig. 5A). Accordingly, it was found that the ubiquitin of cFLIP was higher in Huh7 cells than in Huh7-SR cells following sorafenib treatment (Fig. 5B). It has been reported that the knockdown of USP2, a de-ubiquitinating enzyme, can protect hepatocytes from tumor necrosis factor-α-induced apoptosis by decreasing the cellular levels of the ubiquitinligas ITCH, a negative regulator of cFLIP, and the subsequent turnover of cFLIP (20). Therefore, the present study determined the expression of USP2 and ITCH in Huh7 and Huh7-SR cells following sorafenib treatment. The expression of USP2 was decreased in the Huh7-SR cells following sorafenib treatment, compared with that in the Huh7 cells (Fig. 5C). The expression of ITCH in the Huh7-SR cells was also lower than that in the Huh7 cells (Fig. 5C). The overexpression of USP2 enhanced the expression of ITCH and decreased the expression of cFLIP in the Huh7-SR cells treated with sorafenib (Fig. 5D). These results suggested that the downregulation of USP2 following long term exposure to sorafenib may contribute to decreased proteasomal degradation of cFLIP.