The present study was approved by the Institutional Ethics Committee of the Eastern Hepatobiliary Surgery Hospital in Shanghai, China, and was performed in compliance with the Declaration of Helsinki (1975) and its amendments. Patients whose ICC tumor tissues were used in the present study, provided written informed consent. A total of 80 cases of ICC patients diagnosed by surgery at the Eastern Hepatobiliary Surgery Hospital between January 2012 and January 2013 were enrolled. Their tumor tissues and clinical information were collected for data analysis.

The UC-MSCs were purchased from Cyagen Biotechnology Co., Ltd. (cat. no. HUXUC-03011-440). MSCs were cultured in Dulbecco's modified Eagle's medium (DMEM) nutrient mix F12 (cat. no. 10565018; Thermo Fisher Scientific, Inc.) with 10% fetal bovine serum (FBS; cat. no. 16140071; Gibco™; Thermo Fisher Scientific, Inc.). ICC cell line RBE (cat. no. TCHu179; Cell Bank of Chinese Academy of Sciences) was cultured in RPMI-1640 medium (item no. 31800; Solarbio Life Sciences) containing 10% FBS and ICC cell line QBC939 (Tongpai Biotechnology Co., Ltd.) was cultured in DMEM containing 10% FBS. All cells were cultured at 37°C in a humidified atmosphere containing 5% CO₂.

Herpesvirus entry mediator (HVEM)-overexpressing adenovirus was synthesized by Obio Technology Corporation (http://www.obiosh.cn/). Enhanced Green Fluorescent Protein (EGFP)-overexpressing adenovirus (OBiO Technology Corp., Ltd.) was used as the control. IL-6 cytokine (cat. no. 206-IL-010/CF) was purchased from R&D Systems, Inc. Chloroquine (CQ) was purchased from Sigma-Aldrich; Merck KGaA. In order to determine whether autophagy was involved in IL-6-induced chemoresistance in cholangiocarcinoma cells, the inhibitor of autophagy CQ was added to the medium of cholangiocarcinoma cells at a concentration of 10 µM. Compound C was purchased from Selleck Chemicals. To determine the role of the AMPK pathway in IL-6-induced autophagy, the AMPK inhibitor, compound C, was used to suppress AMPK signaling at a concentration of 5 µM.

ICC tissues were minced and digested at 37°C with 2.5 mg/ml collagenase IV (Sigma Aldrich; Merck KGaA) and 0.1 mg/ml DNase (Sigma Aldrich; Merck KGaA). Digestion was performed for 30 min-1 h and stopped once no pieces of tissue were left. Then, the suspension was filtered through a 100-µm nylon cell strainer and spun 5 min at 300–400 × g, at 4°C to obtain single-cell suspension. The suspension was then incubated with FITC-conjugated SSEA-4 (cat. no. 330409; BioLegend, lnc.) for 30 min at 4°C. Cells were washed 3 times with PBS, and then FITC-positive cells were assessed via flow cytometry. Flow cytometry was carried out with MoFlo™ XDP Cell Sorter (Beckman Coulter, Inc.) and the data analysis was performed using FlowJo software v.7.6.5 (BD Biosciences).

RBE human cholangiocarcinoma cells (1×10⁶) and QBC939 human cholangiocarcinoma cells (1×10⁶) were plated into 6-well plates, 50% conditioned medium of control MSCs and HVEM-overexpressing MSCs were added. In addition, chemotherapeutic drugs, 5-FU (25 µg/ml) (cat. no. HY-90006; MedChemExpress) and cisplatin (5 µg/ml) (cat. no. HY-17394; MedChemExpress) were added to the culture system for 48 h. Cells were stained by Annexin V Alexa Fluor 647 (cat. no. R37175; Invitrogen; Thermo Fisher Scientific, Inc.) and propidium iodide (PI; cat. no. P3566; Invitrogen; Thermo Fisher Scientific, Inc.) at 4°C for 30 min, and then the cells were washed with PBS. Annexin V- and PI-positive cells were analyzed by flow cytometry. Flow cytometry was carried out with MoFlo™ XDP Cell Sorter and the data analysis was performed using FlowJo software v.7.6.5.

RBE human cholangiocarcinoma cells (1×10⁶) were plated into 6-well plates, and treated with various IL-6 concentrations (0.1, 1, 5, 10, 20 and 50 ng/ml) for 6–8 h. Then, chemotherapeutic drugs, 5-FU (25 µg/ml) and cisplatin (5 µg/ml) were added to the culture system for 48 h. Western blotting was used to detect the level of autophagy, and CCK-8 and apoptosis assays were employed to examine cell viability and cell death. The lowest concentration of IL-6 (10 ng/ml) which could effectively promote autophagy and chemoresistance was used for subsequent experiments.

Cholangiocarcinoma cells (3,000 cells/well) were plated into 96-well plates, and 50% conditioned medium of control MSCs and HVEM-overexpressing MSCs were added. In addition, chemotherapeutic drugs, 5-FU and cisplatin were added to the culture system at 25 µg/ml and 5 µg/ml, respectively. After 48 h, a CCK-8 assay (cat. no. ab228554; Abcam) was used to assess the cell viability. CCK-8 reagent was added into each well at a ratio of 1:10 with medium. After incubation for 1 h at 37°C, the absorbance was detected using a microplate reader at a wavelength of 450 nm.

A recombinant adenovirus of HVEM overexpression was purchased from Genechem Biotech, Inc. This adenovirus was amplified directly from the original adenovirus strain. The original adenovirus strain was constructed by co-transfecting plasmid and packaging plasmid to adenovirus. The plasmid vector was pADV-mCMV-HA-P2A and pADV-mCMV-HA-P2A-EGFP was used as a control. Plasmid (1 µg/µl) was transfected with Lipofectamine 3000 transfection reagent (cat. no. L3000015; Invitrogen; Thermo Fisher Scientific, Inc.) into the adenovirus at 37°C for 24 h. 293 cells (1×10⁶; cat. no. GNHu18; Cell Bank of Chinese Academy of Sciences) were transfected to amplify the adenovirus. Adenovirus particles were collected and the concentration was detected. MSCs were transfected by this adenovirus at MOI 20 for 12 h with serum-free medium. Then, 48 h later, HVEM expression was verified by western blotting.

Total protein was extracted from cells using RIPA buffer (Beyotime Institute of Biotechnology), and the protein concentration was detected using a BCA assay. A total of 20 µg protein per lane was separated via 12% SDS-PAGE (Zhao Rui Biotech Co., Ltd.). The proteins were then transferred onto nitrocellulose membranes. Next, 5% non-fat milk was used to block non-specific sites for 2 h at room temperature. The membranes were incubated with primary antibodies against: HVEM (product code ab47677; Abcam), LC-3 (product no. 12741), p62 (product no. 88588) and GAPDH (product no. 5174) (all 1:1,000; from Cell Signaling Technology, Inc) at 4°C overnight. The membranes were then washed with TBST [TBS + 0.1% (v/v) Tween-20) 3 times, and then incubated with corresponding secondary antibodies goat anti-rabbit or goat anti-mouse IgG H&L (HRP) (cat. nos. ab205718 or ab205719, respectively; 1:5,000; Abcam) at room temperature for 2 h and washed with TBST another 3 times. ECL (Cell Signaling Technology, Inc.) was used for visualization.

Total RNA was extracted from cells using TRIzol^® (cat. no. 15596-026; Invitrogen; Thermo Fisher Scientific, Inc.) and reverse-transcribed into cDNA using a Bestar qPCR RT kit (cat. no. DBI-2220; Shanghai Xinghan Sci&Tech Co., Ltd.) according to the manufacturer's protocol. The PCR cycle was performed according to manufacturer's instructions with initial denaturation at 37°C for 15 min, 95°C for 2 min followed by 11 cycles at 95°C for 30 sec, 65°C for 30 sec and 72°C for 1 min and a final extension at 72°C for 5 min. mRNA expression of HVEM was detected by RT-PCR using a Bestar real-time SYBR Green PCR master mix (DBI-2043; Shanghai Xinghan Sci&Tech Co., Ltd.) with an ABI PRISM 7300 system. GAPDH was used as the internal control and data analysis was performed using the 2^−ΔΔCq method (23). The primers used were as follows: HVEM forward, 5′-TCATCGTCATTGTTTGCTCCA-3′ and reverse, 5′-ACCTTGACTACATCACCCCTT-3′; GAPDH forward, 5′-GGAGCGAGATCCCTCCAAAAT-3′ and reverse, 5′-GGCTGTTGTCATACTTCTCATGG-3′.

Conditioned medium of control MSCs and HVEM-MSCs were collected following culture with serum-free medium for 24 h. Cytokines in the medium were detected by a Bio-Plex Pro™ Human Cytokine 27-plex assay (cat. no. M500KCAF0Y). Briefly, serially diluted standards and undiluted conditioned media (50 µl) were added to a microfilter plate containing antibody-coupled beads for each of the 27 analytes and incubated for 60 min with continuous shaking at room temperature between 20°C and 30°C. After washing, the biotinylated detection antibodies were added for 30 min with shaking. The microfilter plate was washed again, and Streptavidin-PE (50 µl) was added and incubation continued at room temperature with shaking (90 × g for 1 min followed by 10 × g for 15 min). Assay buffer (125 µl) was added to each well of the microfilter plate before being read on a Bio-Plex 200 machine (Bio-Rad Laboratories, Inc.).

Patient tissues were fixed with 10% formalin at room temperature for 24 h, embedded in paraffin, and sliced into 5-µm thick sections for the immunohistochemistry assay. Sections were de-paraffinised with xylene and rehydrated with three successive changes in ethanol, and 1% bovine serum albumin (Gibco; Thermo Fisher Scientific, Inc.) was used to block non-specific sites at 37°C for 30 min. The sections were then incubated with primary antibodies against: HVEM (1:100), LC-3 (1:200) and p62 (1:200), at 4°C overnight. PBS was used to wash the sections 3 times, and then sections were incubated with corresponding secondary antibodies goat anti-rabbit and goat anti-mouse IgG H&L (HRP) at 37°C for 30 min. After washing with PBS 3 times, DAB (1 mg/ml; cat. no. D8417; Sigma-Aldrich; Merck KGaA) was used at room temperature between 20°C and 30°C for 3 min for color development. Hematoxylin was used to stain nucleic acid at room temperature. A light field microscope (magnification, ×200) was used to observe and quantify positive cells per mm² (Leica Microsystems, Inc.). We used immunohistochemical scoring standards to define high or low IL-6 levels. The intensity was scored as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. The frequency of positive cells was defined as follows: 0, <5%; 1, 5–25%; 2, 26–50%; 3, 51–75%; and 4, >75%. The staining index (values, 0–12) was determined by multiplying the score for staining intensity with the score for positive area. For statistical analysis, scores of 0 to 7 were considered low expression and scores of 8 to 12 were considered high expression) (24).

For immunofluorescence staining, the sections were boiled with citrate (pH 6.0), permeabilized in PBS supplemented with 0.2% Triton X-100 (PBST) and blocked with 2% normal donkey serum (cat. no. ab7475; Abcam) for 1 h at room temperature. Primary antibodies: HVEM (1:200) and SSEA-4 (1:500; cat. no. ab16287; Abcam, Inc.) were then incubated at 4°C overnight. Subsequently, after washing with PBS, the sections were incubated with corresponding secondary antibodies goat anti-rabbit or goat anti-mouse IgG H&L (Alexa Fluor^® 488) (cat. no. ab150077 or ab150113, respectively; 1:1,000; Abcam), stained with DAPI (5 µg/ml; cat. no. D8417; Sigma-Aldrich; Merck KGaA) for 15 min at room temperature and then embedded using Vectashield^® (Vector Laboratories, Inc.). Images were captured on SP8 confocal microscope (magnification, ×200) (Leica Microsystems, Inc.).

Statistical analyses were performed using the SPSS 20.0 (IBM Corp.). Categorical variables were compared using the χ² test. Continuous variables were compared using the Mann-Whitney U test as appropriate. Pearson's correlation coefficient was used to determine correlations between continuous normally distributed variables. The overall survival curve was drawn by the Kaplan-Meier method. Log-rank test was used to compare the survival time of patients between each group. Data sets were analyzed by analysis of variance (ANOVA) with a posteriori contrast by least significant difference (for comparisons among multiple groups) or by paired Student's t-test (for comparison between two groups). Immunohistochemical images were analyzed by Image-Pro^® Plus 6.0 (IPP; Media Cybernetics, Inc.). Each experiment was performed in triplicate. P<0.05 was considered to indicate a statistically significant difference.

At first, immunofluorescence staining of SSEA4 and HVEM was performed in the liver specimens from several ICC patients. As revealed in Fig. 1A, the expression of the SSEA4 (red) and the HVEM (green) exhibited high overlap. Then, using the fresh liver specimens collected from the patients with ICC, MSCs were isolated via fluorescence-activated cell sorting (FACS). SSEA4 antibody was used to identify MSCs (25,26). As presented in Fig. 1B, there were ~7.5% SSEA-positive cells detected. Then, SSEA4-positive cells were collected and HVEM was detected in ICC-MSCs via western blotting and RT-qPCR. Cholangiocarcinoma originates from abnormal cell differentiation caused by chronic inflammation. Thus, there is chronic inflammation in the para-carcinoma tissue. Since the function of MSCs is easily affected by inflammatory factors, MSCs in para-carcinoma tissue cannot be used as a normal control (27,28). Umbilical cord-derived MSCs are derived from normal tissues and are currently the most commonly used MSCs in research and applications (29). Therefore, UC-MSCs were used as a normal control in this study. As presented in Fig. 1D, it was revealed that HVEM expression in ICC-MSCs was significantly higher than that in UC-MSCs. The results of the RT-qPCR demonstrated the same trend (Fig. 1C).

In addition, the level of HVEM between the primary MSCs sorted by FACS and HVEM-overexpressed MSCs by adenovirus were detected by western blotting. As revealed in Fig. S1, there was no significant difference between the expression of HVEM between the primary MSCs sorted by FACS and the HVEM-overexpressed MSCs by adenovirus.

Since HVEM was expressed at high levels in MSCs in cholangiocarcinoma, in order to detect the role of HVEM-overexpressed MSCs in the chemoresistance of cholangiocarcinoma, the present study constructed an HVEM-overexpressing adenovirus, transfected with normal MSCs, which were isolated from healthy individuals. As presented in Fig. 2A, the adenovirus effectively induced HVEM expression in MSCs. Then, the supernatant liquid of HVEM-overexpressing MSCs was added to the RBE and QBC939 cell culture medium. The chemotherapeutic drugs, 5-FU and cisplatin (25 and 5 µg/ml) were added to the culture system for 48 h. A CCK-8 assay was used to examine the cell viability. Compared with the control group, HVEM-overexpressing MSCs demonstrated a significantly higher cell viability (Fig. 2B-E). Consistent with this, the apoptosis assay also demonstrated that HVEM-overexpressing MSCs could significantly enhance the resistance to chemotherapeutic drugs (Fig. 2F-K). The aforementioned data indicated that HVEM-overexpressing MSCs promoted the capacity of chemoresistance in cholangiocarcinoma cells.

To determine the underlying mechanism by which HVEM-overexpressing MSCs promoted chemoresistance in cholangiocarcinoma cells, a bioplex assay was performed to detect the cytokines in HVEM-overexpressing MSCs. As presented in Fig. 3A, compared with the control group, the HVEM-overexpressing group exhibited a significantly higher level of IL-6 in MSCs. In order to confirm the role of IL-6 in the chemoresistance of cholangiocarcinoma cells, IL-6 was used to treat cholangiocarcinoma cells directly at a concentration of 10 ng/ml. The cell viability and death were detected to evaluate the chemoresistance in cholangiocarcinoma cells. Following treatment with 5-FU and cisplatin, IL-6 increased the cell viability in RBE and QBC939 cell lines (Fig. 3B-E). A decreased amount of 5-FU or cisplatin-induced cell death was observed in IL-6-pretreated cholangiocarcinoma cells (Fig. 3F-J). These results indicated that HVEM-overexpressing MSCs promoted chemoresistance in cholangiocarcinoma cells through IL-6.

The present study then examined the mechanism underlying IL-6-induced chemoresistance in cholangiocarcinoma cells. Autophagy has been reported to play an important role in tumor cell chemosensitivity (30,31). Therefore, the level of autophagy was detected in IL-6-treated RBE cells via western blotting. The results indicated that LC-3 expression was upregulated, whereas p62 expression was downregulated, which are two of the most important markers of autophagy (32) (Fig. 4A). In order to determine whether autophagy was involved in IL-6-induced chemoresistance in cholangiocarcinoma cells, the inhibitor of autophagy CQ was added to the medium of cholangiocarcinoma cells at a concentration of 10 µM. As revealed in Fig. 4B, CQ could effectively suppress LC3 expression and increase p62 expression in RBE cells. The CCK-8 and apoptosis assays were also employed to examine cell viability and cell death in RBE cells treated with 5-FU and cisplatin. As anticipated, the enhancement of IL-6-induced survival in RBE cells was diminished when autophagy was inhibited by CQ (Fig. 4C-F). Consistent with this, a significantly higher rate of cell apoptosis was detected in RBE cells in the CQ treatment group compared with that in the control group (Fig. 4G-K), thus indicating that apoptosis can be inhibited by autophagy. Therefore, it was established that IL-6 promoted chemoresistance via activating autophagy in cholangiocarcinoma cells.

The present study then investigated the mechanism underlying IL-6-mediated autophagy activation. Previous studies have reported that the AMPK/mTOR signaling pathway mediates cell autophagy activation (33,34), and that activated AMPK negatively regulated mTOR and thereby enhanced autophagy flux. Therefore, phosphorylated (p)-AMPK, AMPK, p-mTOR and mTOR levels were detected in cholangiocarcinoma cells following IL-6 treatment using western blotting in the present study. As presented in Fig. 5A, compared with the control group, the level of p-AMPK was increased, while the p-mTOR protein level was decreased in IL-6-disposed RBE cells. Furthermore, to determine the role of the AMPK pathway in IL-6-induced autophagy, the AMPK inhibitor, compound C, was used to suppress AMPK signaling. The western blotting results revealed that compound C effectively inhibited p-AMPK expression (Fig. 5B). Then, LC3 and p62 expression levels were also detected in RBE cells treated with both IL-6 and compound C. Compared with IL-6-treated alone group, autophagic activity was significantly decreased in both the IL-6 and compound C-treated group (Fig. 5C). These findings indicated that the AMPK/mTOR pathway was involved in IL-6-induced autophagy in cholangiocarcinoma cells.

In order to evaluate the prognostic value of IL-6 in ICC, 80 patients with ICC were recruited and tumor tissues were collected. According to the immunohistochemical scoring standard (scores of 0–7 were considered low expression and scores of 8–12 were considered high expression) (24); the patients were divided into two groups: The high-IL-6 level group (n=44) and the low-IL-6 level (n=36) group; It was observed that the expression of LC3 in the high-IL-6 group was increased compared with the low-IL-6 group and the expression of p62 in the high-IL-6 group was decreased compared with the low-IL-6 group (Fig. 6A). The baseline features between these patients in the two groups were well-balanced (Table I). Furthermore, the integrated optical density (IOD) of LC3 and p62 expression was calculated by IPP analysis following immunohistochemistry in the IL-6-high and IL-6-low expression groups. Then, the association between the expression levels of IL-6 and autophagy-associated markers was identified. As presented in Fig. 6B, the IL-6 level was significantly associated with LC3 and p62 in ICC tissues. Furthermore, the survival rate of ICC patients was evaluated using the Kaplan-Meier survival analysis. Patients with high IL-6 levels were likely to have significantly poorer survival rate than those with low IL-6 levels (Fig. 6C). The aforementioned results demonstrated that high IL-6 levels could be a risk factor in the prognosis of patients with HCC.