Hornerin promotes tumor progression and is associated with poor prognosis in hepatocellular carcinoma

Total of 271 HCC patients was involved in the study. The snap-frozen tumors and corresponding peri-cancerous tissues were collected during liver resection at the Department of Hepatobiliary Surgery, the First Affiliated Hospital of Sun Yat-sen University from January 2006 to December 2008. There was no gender discrimination in the treatment offered (surgery) to patients referred for HCC to our institution. The study was approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University. All patients signed informed consent. The tumor stages were assessed according to the tumor-node-metastasis (TNM) system of the 2010 International Union Against Cancer by the American Joint Committee. The histological grade of tumors was determined by the Edmondson Steiner grading system [10]. Postoperative patient follow-up was implemented as previously described [11, 12]. The durations of disease-free survival (DFS) and overall survival (OS) were defined as previously described [11, 12]. The last follow-up date was December 31, 2013.

The human HCC cell lines HepG2 (Catalogue Number: HB-8065™), Hep3B (Catalogue Number: HB-8064™) and PLC/PRF/5 (Catalogue Number: CRL-8024™) were purchased from the American Type Culture Collection (ATCC; Rockville, MD, USA). The human HCC cell line Huh7 (Catalogue Number: JCRB0403) was purchased from the Japanese Cancer Research Bank. The human HCC cell lines SMMC-7721 (Catalogue Number: TCHu 52), BEL-7402 (Catalogue Number: TCHu 10), QGY-7703 (Catalogue Number: TCHu 43) and normal liver cell line LO2 (Catalogue Number: GNHu 6) were obtained from Cell Bank (Shanghai, China).

The cells were cultured in low glucose Dulbecco’s modified Eagle media (DMEM), including 10% fetal bovine serum (FBS) supplemented with 100 U/ml penicillin and 0.1 mg/ml streptomycin, and incubated at 37 °C in a humidified atmosphere at 5% CO₂.

The three lentivirus plasmids containing human HRNR shRNAs, vector plasmid pLKO.1 puro, packaging plasmid pHR’8.2 deltaR dvpr and pCMV-VSV-G were purchased from Sigma (St. Louis, MO, USA). These plasmids were extracted according to the protocol (GeneJET Plasmid Maxiprep Kit, Thermo SCIENTIFIC). The lentiviral packaging cells, 293 T cells (CRL-3216™), were transfected with the three lentivirus plasmids containing human HRNR shRNAs or vector plasmid pLKO.1 puro and packaging plasmid pHR’8.2deltaR dvpr and pCMV-VSV-G at 70% confluence with the use of Lipofectamine 2000 (Invitrogen, Carlsbad, CA) to produce the lentivirus. Media containing the lentivirus were added to the target cells for 24 h. After 24 h, the original medium was replaced with fresh medium. The cells containing the shRNA constructs were selected in the medium containing puromycin and were cultured for approximately 2 weeks [13]. The stable cell lines were validated by western blotting.

Tissue microarray construction was done as described [14]. Two 1 mm diameter core biopsies were removed from the donor blocks; then, the samples were transferred to the recipient paraffin block. The immunohistochemical staining (IHC) was used to the avidin-biotin-peroxidase complex method. In brief, after rehydration and heating antigen retrieval, antibodies against human HRNR (1:200, NBP1–80807; Novus) were then used to the slides and incubated at 4 °C overnight. The secondary antibody incubation (Envision Polymer-HRP,anti-Rabbit/Mouse) was then performed at 37 °C for 30 min. The reaction products were visualized with diaminobenzidine staining and Meyer’s haematoxylin counterstaining. Two investigators who did not have any clinical or pathological information regarding the origin of the samples scored the IHC staining. The scores of IHC staining were determined as previously described [15, 16]. Based on the scoring system, HCC tissues were classified as follows: negative, weak, moderate, and strong. The expression levels of HRNR were divided into a HRNR-low group (negative/weak) and a HRNR-high group (moderate/strong). Each sample was scored in a blinded manner by two investigators who did not have any clinical or pathological information regarding the origin of the samples.

The cells were washed twice with ice-cold phosphatebuffered saline (PBS). Proteins were extracted from the cells using RIPA lysis buffer as previously described [17]. The protein concentration was decided with the Bradford reagent (Bio-Rad Laboratories, Hercules, CA, USA) using a bovine serum albumin standard. Equal amounts of total protein were separated on 10% SDS-PAGE gels and subsequently transferred onto PVDF membranes. The membranes were detected overnight at 4 °C with primary antibodies. Western blot bands were detected by electrochemical luminescence (ECL). Protein expression was confirmed by western blot using the following antibodies: hornerin (NBP1–80807; Novus), AKT and p-AKT (Ser473) (9272 and 9271, respectively, Cell Signaling Technology, Danvers, MA,USA), and GAPDH (sc-47,724, Santa Cruz).

The cells were placed into a 96-well plate (5000 cells/ well). At different points in time (1, 2, 3, 4, 5 and 6 days), 10 μl of MTT (5 mg/ml, Sigma, USA) was added to each well, and the plate was hatched for an additional 4 h. Then, the medium was exchanged by 150 μl of DMSO and shaken at room temperature for 10 min. The number of viable cells in each well was calculated by the absorbance value (λ = 490 nm).

For the colony formation assay, the cells were placed into a 6-well culture plate (1000 cells/well) and cultured for 2 weeks. The colonies were stained with 1% crystal violet and counted.

For the cell migration assay, transwells (24-well, 8-μm pore size; Millipore, Billerica, MA, USA) were used. A total of 3 × 10⁴ cells in 300 μl DMEM without FBS were seeded in the upper chamber and 800 μl of DMEM with 10% FBS was added to the lower chamber. The upper chamber cells were removed after 48 h incubation and those on the lower surface of the membrane were fixed with methanol, then, the cells were stained with crystal violet, counted (200× magnification), and photographed. The cell invasion assays were performed the same as the cell migration assays, except the transwells were precoated with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA). All above experiments were done in triplicate.

For the xenograft tumor model, 1 × 10⁶ cells were injected subcutaneously into the right upper flank of 5-week-old male BALB/C nude mice. Each group contained 6 mice. Tumor formation in nude mice was monitored over a 32-day period, and the length and width of the tumors were measured every 4 days and their volumes were calculated by the formula: V = 0.5 × length × width². The animal experiment was approved by and performed in accordance with the Ethic Committee on the Use of Live Animals in Teaching and Research at the First Affiliated Hospital of Sun Yat-sen University. The tumor-bearing mice were sacrificed by cervical dislocation.

Statistical analysis was performed with SPSS software (19.0; SPSS, Inc., Chicago, IL). Categorical data were analyzed by the chi-square or Fisher’s exact tests. Cumulative recurrence and survival rates were analyzed using Kaplan-Meier’s method and the log-rank test. Cox’s proportional hazards regression model was used to analyze independent prognostic factors. Variables analyzed by univariate analysis with P <  0.05 were involved in the multivariate Cox proportional hazards model. P <  0.05 was considered statistically significant.

To explore the role of HRNR in HCC, we analyzed the expression of HRNR in tumor samples from a cohort of 271 HCC patients. Our results showed that HRNR was expressed in 84.5% (229/271) of HCC tissues. High HRNR expression was found in 57.9% (157/271) of patient tissues. HRNR expression was localized mainly in the cytoplasm, with some expression identified on the cell membranes (Fig. 1).

Next, we evaluated whether there was any association of HRNR expression with the clinicopathologic factors of HCC patients. Based on the IHC results, the 271 HCC patients were distributed into two groups: the HRNR-high expression group (n = 157) and the HRNR-low expression group (n = 114). The results revealed that high HRNR expression in HCC positively correlated with vascular invasion (P = 0.002), poor tumor differentiation (P = 0.042) and advanced TNM stage (P <  0.001); however, the high expression of HRNR in HCCs had no significant correlation with age, gender, HCC family history, hepatitis B, liver function Child-Pugh stage, cirrhosis, tumor size, tumor number, encapsulation and alpha-fetoprotein (AFP) (all P > 0.05) (Table 1).

We further explored the prognostic value of HRNR expression. We found that the 1-, 3-, and 5-year DFS rates (29.9%, 16.6% and 13.4% VS 61.4%, 49.1% and 45.3%, P <  0.001) and OS rates (61.1%, 31.2% and 21.6% VS 78.9%, 65.8% and 63.1%, P <  0.001) of HCC patients in the high HRNR expression group were poorer than those in the low HRNR expression group (Fig. 2). Kaplan-Meier analysis indicated that Edmondson grading, tumor size, capsulation, tumor number, vascular invasion, HRNR expression and TNM stage were risk factors for DFS; gender, Edmondson grading, tumor size, capsulation, tumor number, vascular invasion, HRNR expression and TNM stage were risk factors for OS (Table 2). According to the multivariate Cox regression analysis, high HRNR expression was found to be an independent prognostic factor linked to both poor DFS (hazard risk [HR] = 2.209, 95% confidence internal [CI] = 1.627–2.998,P <  0.001) and OS (HR = 2.459,95% CI =1 .736–3.484, P <  0.001) (Table 3). These findings suggest that high HRNR expression was significantly associated with poor prognosis, indicating a potential role for HRNR in hepatic tumorigenesis.

To investigate the roles of HRNR in HCC progression, we first detected the expression levels of HRNR in different HCC cell lines. The result indicated that the expression levels of HRNR were different in HCC cell lines, with the highest expression detected in PLC/PRF/5 and QGY-7703 cell lines (Fig. 3a). Thus, we selected these two cell lines for further analysis. We determined whether reducing HRNR expression ameliorated tumor growth. We knocked down HRNR expression in PLC/PRF/5 and QGY-7703 cell lines using two independent shRNA constructs (Fig. 3b). The proliferation assay showed that when HRNR expression was knocked down by HRNR-shRNAs in PLC/PRF/5 cells, tumor cells proliferation was suppressed compared with the PLC/PRF/5 scramble control cells (P <  0.01). Similarly, the proliferation of QGY-7703-shRNA1-HRNR and QGY-7703- shRNA2-HRNR cells was also significantly decreased (P <  0.01) (Fig. 4).

We next determined the functional role of HRNR in aggressive growth properties of tumor cells by performing colony formation and migration assays. Our results showed that the silencing of HRNR with shRNA1 and shRNA2 inhibited colony formation in PLC/PRF/5 cells compared to control cells (P <  0.01). The same phenomena were also observed in QGY-7703 cells (P <  0.01) (Fig. 5). The transwell migration assay revealed an important suppression of cell migration in PLC/PRF/5-shRNA1-HRNR and PLC/PRF/5-shRNA2-HRNR cells compared with the PLC/PRF/5-scramble control cells. Similarly, when compared to the QGY-7703-scramble control cells, the migration was less in both QGY-7703-shRNA1-HRNR and QGY-7703-shRNA2-HRNR cells (Fig. 6a). Moreover, the invasion assays demonstrated that knocking down HRNR significantly impaired the invasiveness of both PLC/PRF/5 and QGY-7703 tumor cells (Fig. 6b).

To further explore the biological importance of HRNR in HCC, we examined the tumor growth in xenograft experiments. Human tumor cells were injected subcutaneously in nude mice and tumor growth was monitored. As represented in Fig. 7a, tumor growth in mice injected with PLC/PRF/5-shRNA1-HRNR and PLC/PRF/5-shRNA2-HRNR cells was significantly decreased compared with the control group. Furthermore, tumor weight was positively associated with the expression levels of HRNR. We also found that inhibiting HRNR reduced tumour growth in the xenograft model with QGY-7703 tumour cells (Fig. 7b). Collectively, these data suggest that HRNR plays a critical role in HCC tumor growth in vivo.

Finally, we explored the potential mechanism responsible for HRNR-mediated tumor growth. HRNR is a member of the S100 protein family. Emerging evidence has indicated that the functional role of S100 protein family members, such as S100A1, A100A4 and S100A16, is closely associated with AKT phosphorylation and activation [18‐20]. We proposed that the role of HRNR in HCC might also be through regulating AKT phosphorylation. To test this, we analyzed AKT expression and phosphorylation by western blot. We found that AKT phosphorylation was suppressed after knockdown of HRNA in PLC/PRF/5 and QGY-7703 cells, whereas the expression level of total AKT was not changed (Fig. 8). Together, our results imply that HRNR may promote HCC via AKT phosphorylation.