CD147 promotes Bip expression in response to ERS in human HCC cells

To follow the kinetics of CD147 and Bip expression in response to ERS, 7721 and HepG2 human HCC cells were exposed to the ER stressors (i.e., reagents that induce ERS), tunicamycin (Tm, an inhibitor of N-glycosylation) and thapsigargin (Tg, an inhibitor of ER Ca²⁺-ATPase). A relatively low level of CD147 and Bip expression was detected before Tm or Tg treatment. However, upon ER stressor treatment, there was a substantial increase in CD147 and Bip protein levels in a time-dependent manner (Figure 1a).

Open in a separate window

Figure 1

CD147 promotes Bip expression in response to ERS in human HCC cells. (a) 7721 and HepG2 cells were treated with 1μM Tm or 100nM Tg for 0, 6, 12, 18, 24h. Western blot was used to examine CD147 and Bip protein levels. (b) Immunofluorescence was used to assess the localization of CD147 in 7721 cells after treatment with Tm or Tg (Bar=10μm). (c) The 7721 cells were treated with CD147 siRNA, incubated for 24h and then stimulated with Tm or Tg for 18h. Western blot was performed to examine the CD147 and Bip protein levels. Control siRNA transfected 7721 cells were used as a negative control. (d) Real-time PCR was performed to examine the CD147 and Bip mRNA levels. (e) The 7721 cells were treated with CD147 siRNA, incubated for 24h and then stimulated with Tg for 0, 6, 12 and 18h. Western blot was performed to examine the decrease of CD147 and Bip by CD147 siRNA for various time periods. Bars represent each sample performed in triplicate, and the error bars represent the standard deviations. *P<0.05, **P<0.01, by Student's t-test

As the ER is a key organelle for protein synthesis, ERS may induce the accumulation of misfolded proteins in the ER.²⁵ To examine whether the increased CD147 after ER stressors treatment is misfolded protein accumulated in the ER or is properly folded protein mostly expressed at the cell surface, we tested the localization of CD147 in 7721 cells after ER stressors treatment. Immunofluorescence assay demonstrated that the increased CD147 was mostly localized on the HCC cell membrane (Figure 1b). This indicated that the increase of CD147 upon ER stressor treatment is not due to the accumulation of misfolded CD147 in the ER.

To directly detect whether Bip expression was correlated with CD147 expression, we used small interfering RNA (siRNA) to knockdown CD147 in 7721 cells. CD147 siRNA led to a marked reduction of CD147 levels in 7721 cells (Figure 1c). When the cells were stimulated with Tm or Tg for 18h, CD147 siRNA markedly diminished Bip mRNA and protein expression levels in 7721 cells (Figures 1c and d). As CD147 is a principal marker of different types of cancers, we further detected the induction of Bip by CD147 in MDA-MB-231 human breast cancer cells (MM231) and A549 human lung adenocarcinoma cells after Tm or Tg treatment. As shown in Supplementary Figure S2, Bip expression significantly diminished in CD147 siRNA transfected cells compared with corresponding control siRNA transfected cells. Together, these results suggest that CD147 may play roles in Bip induction in response to ERS.

Next, to more directly assess the requirement of CD147 in Bip induction, we monitored the decrease of CD147 and Bip by CD147 siRNA for various time periods. After the exposure to Tg for 6h, Bip protein levels in CD147 siRNA transfected 7721 cells were induced to similar levels, as in control siRNA transfected 7721 cells. However, at 12h and later, the level of Bip protein decreased by almost 50% in CD147 siRNA transfected 7721 cells compared with the control siRNA transfected 7721 cells at each time point (Figure 1e). This suggested that CD147 is important for the induction of Bip expression during ERS, particularly during the late phase of UPR signaling. This may be related to the transcriptional induction of CD147 mRNA after ERS.

CD147 induces nuclear localization of TFII-I and the binding of TFII-I to the Bip promoter to increase the transcription of Bip in response to ERS in human HCC cells

IRE1, PERK and ATF6 are three key factors that trigger UPR signaling and subsequently promote the expression of Bip.^{12, 13, 14} Therefore, we examined the expression and phosphorylation levels of these genes in CD147 siRNA transfected 7721 cells. Unexpectedly, CD147 siRNA did not significantly change these proteins expression levels with or without Tg treatment (Figure 2a).

Open in a separate window

Figure 2

CD147 induces nuclear localization of TFII-I and the binding of TFII-I to the Bip promoter to increase the transcription of Bip in response to ERS in human HCC cells. (a) The 7721 cells were treated with CD147 siRNA, incubated for 24h and then were either left untreated or treated with Tg for 18h. Western blot was used to examine the expression levels of IRE1, p-IRE1, PERK, p-PERK and cleaved ATF6. (b) The expression levels of YY1, TFII-I and p-TFII-I were examined in CD147 siRNA transfected 7721 cells either untreated or treated with Tg for 18h. (c) The TFII-I phosphorylation levels were tested in the nuclear and cytoplasmic fractions of CD147 siRNA transfected 7721 cells treated with Tg for 18h. Protein expression levels in the nuclear and cytosolic fractions were normalized to Lamin B1 and β-actin, respectively. (d) Immunofluorescence staining was performed to observe the location of phosphorylated TFII-I in CD147 siRNA transfected 7721 cells treated with Tg for 18h. (Bar=10μm). (e) The 7721 cells were transfected with CD147 expression plasmid (CD147-pcDNA3.1), incubated for 48h and then were either left untreated or treated with Tg for 18h. Western blot was used to examine the expression levels of CD147, TFII-I and p-TFII-I. (f) The TFII-I phosphorylation levels were tested in the nuclear and cytoplasmic fractions of CD147 overexpressing 7721 cells treated with Tg for 18h. (g) The activity of the Bip promoter was tested by Luciferase reporter assay in HEK-293 cells. Bip promoter linked to the Luc promoter gene was co-transfected with either CD147-pcDNA3.1, or TFII-I (WT)-pcDNA3.1, or Tyr248 mutated TFII-I (MT)-pcDNA3.1 expression plasmids, or pcDNA3.1 empty vector, alone and in combination. Five hours after transfection, cells were treated with Tg for 18h, and the luciferase activity was measured. (h) ChIP assay using antibody against TFII-I was performed in Tg-treated CD147 siRNA transfected 7721 cells. The no antibody (no Ab) was used as a blank control, the normal mouse IgG was used as a negative control and input indicates 5% input DNA, a positive amplification control. The 160-bp fragment was amplified by the Bip promoter-specific primers contains the TFII-I binding sequence, whereas the control PCR reactions generated a 160-bp fragment does not contain any TFII-I binding sequence. (i) The TFII-I expression plasmid was co-transfected with a series of 5′-deletion mutants of the human Bip promoter or (j) with a series of three ERSE mutations of the human Bip promoter linked to the Luc promoter gene into HEK-293 cells. Five hours after transfection, cells were treated with Tg for 18h, and the luciferase activity in each sample was measured. Bars represent each sample performed in triplicate, and the error bars represent the standard deviations. *P<0.05, by the ANOVA analysis or Student's t-test

We then tested the expression levels of two major transcription factors, TFII-I and YY1, which can directly bind to the ERS elements (ERSEs) in the Bip promoter in HeLa cells.²⁶ We observed that CD147 siRNA did not decrease the YY1 expression levels in 7721 cells (Figure 2b). However, CD147 siRNA markedly diminished the p-TFII-I levels after Tg treatment (Figure 2b). These results indicate that CD147 may induce tyrosine phosphorylation of TFII-I in response to ERS. Transcription factors might play roles via their translocation from the endochylema to the nucleus. Thus, cytoplasmic and nuclear proteins were extracted and analyzed. The results showed that TFII-I phosphorylation significantly decreased in the nucleus of CD147 siRNA transfected 7721 cells (Figure 2c). Immunofluorescent staining for phosphorylated TFII-I further confirmed that in CD147 siRNA transfected 7721 cells, the phosphorylated TFII-I fluorescence was weak and predominantly localized to the cytoplasm (Figure 2d). To further confirm the above results, we transfected CD147 expression plasmid into 7721 cells to obtain CD147 overexpressing cells (Figure 2e). As shown in Figure 2e, after Tg treatment, the p-TFII-I levels were significantly increased in CD147 overexpressing 7721 cells. Nuclear and cytoplasmic protein extraction results showed that TFII-I phosphorylation markedly increased in the nucleus of CD147 overexpressing 7721 cells after Tg treatment (Figure 2f). Collectively, these results indicate that CD147 induces TFII-I tyrosine phosphorylation and nuclear localization in response to ERS in human HCC cells.

To detect whether CD147-promoted Bip induction was mediated through TFII-I, a Bip (−149/+7)/Luc promoter construct was co-transfected with CD147 or TFII-I expression plasmids, alone or in combination. We observed that either CD147 or TFII-I alone stimulated Bip (−149/+7)/Luc reporter gene activity after Tg treatment. However, the combination of both resulted in no further enhancement (Figure 2g, lanes 2–4). This result may indicate that CD147 and TFII-I perform functional overlapping roles on enhancing the Bip promoter activity in ERS. The functional overlap suggests a molecular link between CD147 and TFII-I in Bip induction. The tyrosine phosphorylation motif at residues 248 of TFII-I is critical for transcriptional activity of TFII-I.²⁷ To further address whether TFII-I is required for CD147-promoted Bip induction in ERS, we co-transfected the Bip (−149/+7)/Luc promoter construct and the CD147 expression plasmid together with either TFII-I (WT) or Tyr248 mutated TFII-I (MT) expression plasmids. As shown in Figure 2g in lanes 4 and 5, compared with CD147 co-transfected with TFII-I (WT), CD147 co-transfected with TFII-I (MT) suppressed Bip promoter activation. The inhibitive effect of TFII-I (MT) on Bip promoter activation indicate that CD147-promoted Bip induction is induced by TFII-I.

To directly elucidate the effects of CD147 on the binding of TFII-I to the Bip promoter, chromatin immunoprecipitation (ChIP) assays were performed in Tg-treated CD147 siRNA transfected 7721 cells. We found that CD147 siRNA markedly decreased binding activity of TFII-I to the Bip promoter (Figure 2h). These findings suggested that CD147 may increase binding activity of TFII-I to the Bip promoter.

Bip promoter contains multiple ERSE copies, with the consensus sequence CCAAT(N9)CCACG located upstream of the TATA element.²⁶ To map the TFII-I target sites within the Bip promoter, the full-length TFII-I expression plasmid was co-transfected into HEK-293 cells with a series of 5′-deletion mutants of the human Bip promoter, which was linked to the Luc promoter gene. Compared with the reporter activity of a minimal promoter containing essentially only the TATA element of the Bip promoter, we observed an almost 10-fold reporter activity of the −149/+7 Luc promoter, which contains three ERSE copies cloned upstream of the TATA element, an eightfold reporter activity of the −115/+7 Luc promoter, which contains two ERSE copies and a fivefold reporter activity of the −82/+7 Luc promoter, which contains one ERSE copy (Figure 2i). Next, we mutated the three ERSEs in the Bip promoter. Mutation of ERSE3 alone reduced the TFII-I-induced reporter activity by ~22% compared with the wild-type promoter, mutation of ERSE2 alone reduced the TFII-I-induced reporter activity by ~20%, and mutation of ERSE1 alone decreased the TFII-I-induced reporter activity to <70% (Figure 2j). This more severe effect suggests that the TFII-I targets all three ERSEs sites of Bip promoter but preferentially acts on the ERSE1 site.

FAK and Src mediate CD147-promoted Bip transcription in response to ERS in human HCC cells

TFII-I is phosphorylated at Tyr248 in a FAK–Src-dependent fashion.¹⁸ To identify whether FAK and Src is involved in CD147-induced TFII-I tyrosine phosphorylation, the expression and phosphorylation levels of FAK and Src were tested. CD147 siRNA markedly decreased p-FAK and p-Src levels in 7721 cells, in both the absence and presence of Tg (Figures 3a and b). These data indicate that CD147 may activate the FAK–Src signaling pathway with or without Tg treatment.

Open in a separate window

Figure 3

FAK and Src mediate CD147-induced Bip induction in response to ERS in human HCC cells. (a) The 7721 cells were treated with CD147 siRNA, incubated for 24h and then were either left untreated or treated with Tg for 18h. Western blot was used to test the FAK expression and phosphorylation levels. (b) The Src expression and phosphorylation levels were tested in CD147 siRNA transfected 7721 cells either untreated or treated with Tg for 18h. (c) After FAK inhibitor PF573288 or Src inhibitor S2075 treatment, the protein levels of TFII-I, p-TFII-I and Bip were examined. (d) Bip protein levels were tested after integrin β1 antibody treatment

We next used the FAK inhibitor PF573288 and the Src inhibitor S2075 to examine whether CD147-induced Bip induction is dependent on the FAK–Src signaling pathway after Tg treatment. The results showed that CD147 siRNA did not further reduce p-TFII-I and Bip levels following the FAK or the Src inhibitor treatment (Figure 3c). Our previous results demonstrate that CD147 facilitates FAK phosphorylation at Tyr397 by interacting with α3β1 integrin.²² We confirmed these results by the observation that CD147 siRNA did not further reduce Bip levels after blocking α3β1 integrin with the addition of 10μM integrin β1 antibody (Figure 3d). These results suggest that α3β1 integrin, FAK and Src are involved in CD147-induced Bip expression.

CD147 inhibits ERS-induced HCC cell apoptosis and decreases chemosensitivity by inducing Bip expression

Due to its anti-apoptotic property, Bip functions as a cytoprotective protein in stressed cells by activating the UPR.²⁵ Because CD147 increases Bip during ERS, CD147 may protect cells from ERS-induced cell apoptosis. To test this hypothesis, cell death was examined by microscope (Supplementary Figure S3) and TUNEL staining (Figure 4a), and then quantified by Annexin V/PI (propidium iodide) double staining (Figure 4b). The results demonstrated that CD147 siRNA caused less resistance to ERS-induced cell apoptosis. To further confirm that Bip is involved in this process, Bip siRNA was used (Supplementary Figure S4). Annexin V/PI double staining demonstrated that CD147 siRNA did not further increase the apoptotic cells following Bip depletion (Figure 4c). To further confirm these results, the activation of caspase-3 was detected. The results showed that CD147 siRNA increased caspase-3 activation during ERS and did not further increase caspase-3 activation following Bip depletion (Figure 4d). From the above results, we conclude that CD147 may protect HCC cells from ERS-induced apoptosis, and Bip is required in this process.

Open in a separate window

Figure 4

CD147 inhibits ERS-induced HCC cell apoptosis and decreases chemosensitivity by inducing Bip expression. (a) The 7721 cells were treated with CD147 siRNA, incubated for 24h, and then were either left untreated or treated with Tg for 18h. Cell death was tested by TUNEL staining (Bar=100μm) and (b) quantified by Annexin V/PI double staining. (c) The 7721 cells were treated with CD147 or Bip siRNA, alone and in combination, and then were stimulated with Tg for 18h. Annexin V/PI double staining was performed to detect the apoptotic cells. (d) Caspase-3 assay kit was used to detect the caspase-3 activity. (e) A range of concentrations of ADR and VP-16 were used to evaluate the drug sensitivity of the CD147 siRNA transfected 7721 cells by colony formation assays. (f) The 7721 cells were treated with CD147 or Bip siRNA, alone and in combination, and then were stimulated with Tg for 18h. ADR (0.2μg/ml) and VP-16 (2μg/ml) were used to evaluate drug sensitivity by colony formation assays. Bars represent each sample performed in triplicate, and the error bars represent the standard deviations. *P<0.05, by Student's t-test

ADR (Adriamycin) and VP-16, two widely used anti-cancer drugs, were used to directly test whether CD147 protects HCC cells resistant to drug by activating UPR. Colony formation assays showed that ADR and VP-16 induced cell death in a dose-dependent manner in Tg-treated 7721 cells. Further, ADR and VP-16 caused significantly more cell death in CD147 siRNA transfected 7721 cells than control siRNA transfected 7721 cells at every concentration measured (Figure 4e). CD147 siRNA, however, did not further increase cell death following Bip depletion after ADR or VP-16 treatment (Figure 4f). These results show that CD147 causes a statistically significant decrease in drug sensitivity by inducing Bip during ERS.

Gene silencing

The sense sequences for CD147 siRNAs were 5′-GUUCUUCGUGAGUUCCUCTT-3′ (Ambion, Austin, TX, USA). Bip siRNAs were 5′-GGAGCGCAUUGAUACUAGA-3′ (Ambion). The cells were transfected with siRNA using the LipofectAMINE 2000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Silencer negative control siRNA (control siRNA) (Ambion) was used as a negative control under similar conditions.

Immunohistochemistry

A total of 98 cases of human HCC tissues from patients who underwent hepatectomies and 42 normal liver tissues from healthy liver transplant donors (supplied by the Department of Pathology of Xijing Hospital at the Fourth Military Medical University and the Cancer Hospital of Tianjin Medical University) were fixed with 10% formalin and paraffin embedded. Sections were deparaffinized and incubated with primary antibodies, including mouse CD147 mAb (prepared in our laboratory) and goat anti-Bip (BD Biosciences, San Jose, CA, USA) antibodies, followed by visualization with a Histostain-Plus Kit (Invitrogen). Immunopositivity was independently evaluated by two pathologists who were blinded to the clinical data. Immunostaining was scored as positive and negative, as described previously.¹⁵

Western blot analysis

Cells were lysed in 1% OG buffer (20mM Tris–HCl, pH 8.0, 150mM NaCl, 1% OG, 1mM EDTA, 10μg/ml leupeptin, 2μg/ml aprotinin and 1mM PMSF). A BCA Protein Assay Kit (Pierce Biotechnology, Rockford, IL, USA) was then used to determine the total protein density, and equal amounts of protein were separated by 10% SDS-PAGE and transferred to a polyvinylidene fluoride microporous membrane (Millipore, Billerica, MA, USA). After being blocked with 5% non-fat milk, the membrane was incubated for 2h at room temperature with the designated antibody. A Western-Light chemiluminescent detection system (Applied Biosystems, Foster City, CA, USA) was used for immunodetection.

Real-time quantitative PCR and data analysis

Total RNA was extracted with TRIzol reagent (Invitrogen) and reverse transcribed into cDNA with a ReverTra Ace-a kit (Toyobo, Shanghai, China). All primers and probes were as follows: CD147 forward primer, 5′-TCGCGCTGCTGGGCACC-3′, and reverse primer, 5′-TGGCGCTGTCATTCAAGGA-3′ Bip forward primer, 5′-CACAGACGGGTCATTCC-3′, and reverse primer, 5′-CCTATGTCGCCTTCACT-3′ and β-actin forward primer, 5′-CCCAGCCATGTACGTTGCTA-3′, and reverse primer, 5′-TCACCGGAGT-CCATCACGAT-3′. β-Actin mRNA was used as an internal control. Real-time quantitative PCR was performed with SYBR green PCR reagent (Takara, Shiga, Japan) and a Stratagene M × 3005P Multiplex Quantitative PCR System (Agilent Technologies, Frankfurt am Main, Germany).

Immunofluorescence

After 100nM Tg (Calbiochem, La Jolla, CA, USA) treatment for 18h, the cells were allowed to attach to glass coverslips for 6h. They were then fixed in 3.7% formaldehyde in PBS, permeabilized with 0.5% Triton X-100 and blocked with 1% BSA in PBS for 1h. Coverslips were then incubated with the indicated antibodies (BD Biosciences) for 20min. Cells were washed in PBS, incubated with their corresponding FITC secondary antibodies (Pierce) for 1h and stained with DAPI (Vector Labs, Burlingame, CA, USA). Finally, the cells were mounted using glycerol and observed with an FV1000 laser scanning confocal microscope (Olympus, Tokyo, Japan).

Nuclear and cytoplasmic protein extraction

Nuclear and cytoplasmic protein extraction was performed using a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, Shanghai, China). Briefly, 2 × 10⁶ cells were resuspended in 200μl of cytoplasmic protein isolation solution A and homogenized on ice. Next, 10μl cytoplasmic protein isolation solution B was added, and the cells were homogenized on ice. The homogenate was centrifuged at 10000 × g for 5min at 4°C. The resulting supernatant was the cytoplasmic protein fraction. The pellet was resuspended in 50μl nuclear protein isolation solution, homogenized on ice and centrifuged at 10000 × g for 10min. The resulting supernatant was the nuclear protein fraction.

Chromatin immunoprecipitation

ChIP assay was performed using the ChIP-IT kit (Millipore) according to the manufacturer's protocol. Briefly, the cells were exposed to 100nM Tg for 18h. Cell lysates were incubated with no antibody, immunoglobulin G (IgG), or anti-TFII-I antibody (Abcam, Cambridge, UK). The immunoprecipitated DNA was amplified by the Bip (NC_000009.11) promoter-specific primers. The experimental PCR reactions generated a 160-bp product from the regulatory region of the Bip promoter containing three ERSEs (−174/−15), whereas the control PCR reactions generated a 160-bp product from a distal region without any ERSEs (+8/+167). Bip ChIP primer: forward 5′-GCGAAGCTTAGCCGCCGCCGGTCGA-3′ reverse 5′-CGAGTAGGCGACGGTGAGGT-3′. Bip control primer: forward 5′-TCTGTCTAACTGGATAACCCCACAA-3′ reverse 5′-TGCTCCTGTGCTACGGC-3′.

Plasmid construction

Full-length TFII-I (WT) was PCR-amplified, and the primers were designed as follows: forward primer, 5′-TTGGATCCCGCCCCCCCGCTTCCCCGCACGCGC-3′ (BamHI) and reverse primer, 5′-GCTCTAGAAAACTATGCCGTCACAG-3′ (XbaI). The Tyr248 mutant TFII-I (MT) was PCR-amplified using a QuickChange mutagenesis kit (Stratagene, La Jolla, CA, USA) according to the manufacturer's instructions. The tyrosine codon at amino acid 248 in TFII-I cDNA was changed to a phenylalanine codon. The mutation primers used were (mutations are shown in bold): forward primer, 5′-TCAGAAGATCCTGATTTTTATCAATATAACAT-3′ and reverse primer, 5′-ATGTTATATTGATAAAAATCAGGATCTTCTGA-3′. The products of full-length TFII-I (WT) and the Tyr248 mutant TFII-I (MT) were confirmed by sequencing (Shanghai Sangon, Shanghai, China) and then cloned into the pcDNA3.1 vector (Promega, Madison, WI, USA), respectively.

The Bip promoter subregions (−149/+7, which contains all three ERSEs; −115/+7, which contains two ERSEs; −82/+7, which contains just one ERSE; −49/+7, which contains only the minimal TATA promoter; and −82/−50, which contains the first ESRE) were synthesized by Wolsen Biological Technology Co and then cloned into the pGL3-Basic vector (Promega). The three BiP promoter ERSEs were disrupted by mutating their respective sequences using a QuickChange mutagenesis kit (Stratagene) as recommended by the manufacturer. The Bip mutation primers used were (mutations are shown in bold): for the first ESRE mutation, forward primer, 5′-AATCGGCGGCGGAATTATTGGTGGC-3′ and reverse primer 5′-TTAGCCGCCGCCTTAATAACCACCG-3′ for the second ESRE mutation, forward primer, 5′-AATGAACGGCCTAATTATAGCAGG-3′ and reverse primer 5′-TTACTTGCCGGATTAATATCGTCC-3′ and for the third ESRE mutation, forward primer, 5′-TCGGCGGCCTAATTAACGGGGCTGG-3′ and reverse primer 5′-AGCCGCCGGATTAATTGCCCCGACC-3′. The products were cloned into the pGL3-Basic vector respectively.

Transfection experiment

The CD147 expression vector (CD147-pcDNA3.1) was developed in our laboratory.³⁶ After cells were grown to 60–80% confluency, transfection was performed using the LipofectAMINE 2000 reagent (Invitrogen) according to the manufacturer's instructions. The pcDNA3.1 empty vector was used as a negative control under similar conditions.

Reporter gene assays

For the luciferase assay, cells were grown to 60–80% confluence. Bip promoter-reporter gene constructs were then co-transfected with the following vectors using Lipofectamine 2000 (Invitrogen) as recommended by the manufacturer: CD147-pcDNA3.1, TFII-I (WT)-pcDNA3.1 and TFII-I (MT)-pcDNA3.1. The amount of total DNA in each transfection was normalized by the addition of an empty vector to the reaction mixture. For stress induction, the cells were treated with 100nM Tg for 18h after transfection. The cells were then prepared for the luciferase assays using the Dual-Luciferase Reporter Assay System (Promega). The relative activity was defined as the ratio of firefly luciferase activity to renilla luciferase activity.

Annexin V/PI double staining

The number of dead cells was determined by Annexin V/ PI double staining. Cells were exposed to 100nM Tg for 18h and then incubated with FITC-conjugated Annexin V in binding buffer (0.01M HEPES, 0.14M NaCl and 2.5mM CaCl₂, pH 7.4) for 20min at 37°C in the dark. After incubation, the cells were washed and resuspended in 200μl PBS with 1% FCS and additionally incubated with 10μl of 1mg/ml PI solution. The Annexin V-positive cells were detected using a FACSCalibur flow cytometer (BD Biosciences), and the results were analyzed using CellQuest software (BD Biosciences). Annexin V-FITC conjugates were detected with the FL1 channel of the FACSCalibur machine. PI was read on the FL2 channel.

Detection of caspase-3 activity

Caspase-3 activity was measured spectrophotometrically via the detection of pNA cleavage from caspase-3-specific substrates. These experiments were completed using a caspase-3 assay kit (Beyotime). After cell lysates were incubated with Ac-LEVD-pNA for 2h at 37°C, the samples were read at 405nm.

Colony formation assays

Twelve hours after incubation with 100nM Tg, the cells were seeded in six-well plates. The next day, the medium was changed to include 0.02–2μg/ml ADR (Hisun Pharma. Co., Zhejiang, China) or 0.2–20μg/ml Etoposide (VP-16, Qilu, Jinan, China). After a change to fresh medium 48h later, the cells were allowed to form colonies for 10 days in the absence of drugs. The colonies were then stained with 0.2% crystal violet (in 75% ethanol) for 10min and counted.

Hepatoma stably expressing GFP

The cells were infected with a reconstructed pWPT–GFP–puro lentiviral vector, which contains a puromycin selection marker as previously reported.³⁷ For selection, 2μg/ml puromycin was added to the medium, and after 14 days, the medium was changed to medium without puromycin. In this manner, the cells stably expressing GFP were established.

Establishment of an orthotopic HCC model in nude mice

For each mouse, 2 × 10⁶ cells were resuspended in 40μl serum-free DMEM/matrigel. Then, each nude mouse (six in each group, BALB/c-nu/nu) was orthotopically inoculated under anesthesia in the left hepatic lobe with a microsyringe through an 8-mm transverse incision in the upper abdomen. After 25 days, the mice were killed, and their livers were dissected, fixed with phosphate-buffered neutral formalin and prepared for standard histological examination. Animal studies were approved by the local regulatory agency (Laboratory Animal Research Centre of the Fourth Military Medical University).

TUNEL staining

Apoptosis in human tumor xenografts from nude mice sections was detected using the DeadEnd Colorimetric Apoptosis TUNEL System (Promega), which end-labels the DNA from the fragmented apoptotic cells using a modified TUNEL staining according to the manufacturer's instructions.

This work was supported by grants from the National Basic Research Program of China (2009CB521704), the National Natural Science Foundation of China (31101005, 81530058, 81071691) and the National S&T Major Project (2011ZX09102-001-21, 2012ZX10002017-002).