There were 64 patients who underwent resection for HCC between 2011 and 2012 at the Department of Hepatobiliary Surgery, the Second Xiangya Hospital of Central South University. None of these patients received chemotherapy or radiotherapy before the operation. The group was composed of 52 men and 12 women at the time of operation. A summary of patient characteristics and pathological features is presented in Table I. Tumor specimens were either cut immediately after removal from the resected hepatic tissues, frozen in liquid nitrogen, and then stored at −80°C or collected in 10% formalin and then embedded in paraffin for histopathological analysis.

Phos-Akt (Ser473), Akt (C67E7) rabbit nuclear factor-κB (NF-κB) p65 (D14E12), phos-NF-κB p65 (Ser536), phos-GSK3 (Ser9), β-catenin (D10A8) and β-actin (8H10D10) used for immunoprecipitation were from Cell Signaling Technology, Inc. Antibodies to NEK2 (ab55550) for western blotting and immunocytochemistry were from Abcam (Cambridge, MA, USA). All antibodies were used according to the manufacturer's instructions. 5-Fluorouracil (5-FU) was purchased from Shanghai Xudong Haipu Pharmaceutical Co., Ltd. (Shanghai, China) and was diluted directly with cell culture medium to the desired concentration. Transient transfection of plasmid DNA was performed using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer's instructions. For RNA interference, siRNAs specific to NEK2, siRNA1 (5′-CAU UUG UUG GCA CAC CUU AUU-3′), siRNA2 (5′-GCU GAG AAA CAG AUG CUU GUU-3′) and siRNA3 (5′-UCU GUU GAA GAA CUA CAG CUU-3′) were purchased from GenePharma Co., Ltd., and transfected into the cells using Lipofectamine RNAi MAX (Invitrogen) according to the manufacturer's instructions. Non-specific control siRNA (5′-AAG TAG CCG AGCT TCG ATT GC-3′) was also used.

HepG2, BEL-7402, QGY-7703, SMMC7721 and Huh-7 were from the Institute of Biochemistry and Cell Biology (Shanghai Institutes for Biological Sciences, CAS). Cells were cultured in RPMI-1640 (Gibco, 31800022; 1.5 g/l NaHCO₃, 2.5g/l glucose, 0.11g/l sodium pyruvate) supplemented with 10% fetal calf serum (FCS; Invitrogen) in a humidified incubator under 5% CO₂ at 37°C.

Cell proliferation was determined using MTT assay of cell proliferation as described before (11). Cells were cultured into a 96-well plate at 1×10⁴ cells/well (n=4 for each time point) in a final volume of 100 µl. Cells were cultured for 36–48 h after transfection with NEK2-siRNA and ctrl-siRNA (non-specific control siRNA), respectively. Then, 20 µl MTT (5 mg/ml) was added to each well (0.8 mg/ml final concentration) for further incubation for 4 h. Then the medium was removed, and 100 µl DMSO was added to dissolve the solid formazan for 15 min. The absorbance of each well was read at 490 nm using a microplate reader (Thermo Fisher Scientific, Inc.). For drug resistance assay, cells were stimulated with different concentrations (0, 0.25, 0.5, 1, 20, 50 and 100 µg/ml) of 5-FU in culture medium for 48 h after transfection with siRNA. The IC₅₀ values were calculated by non-linear regression analysis using SPSS 17.0 software (SPSS, Chicago, IL, USA) as describe before (12).

Migration and invasion assay Transwell filters coated with collagen I or Matrigel (8-mm pore size; BD Biosciences) were used for migration or invasion assays, respectively. Cells (1.5×10⁴) were seeded into the upper chamber in RPMI-1640 with 0.5% fetal bovine serum (FBS). The bottom chamber contained RPMI-1640 with 10% FBS. Cells were allowed to migrate/invade at 37°C in 5% CO₂ for 24 h before they were fixed with methanol/methylene blue solution, stained with crystal violet (Beyotime, Shanghai, China), imaged, and counted with a microscope (Leica, UK). All experiments were performed in triplicate. Cell counts were performed on 10 fields per filter, and the mean was normalized to the migration/invasion cell count of control cells.

Total RNA was extracted with TRIzol reagent (Invitrogen) following the manufacturer's instructions. Real-time quantitative RT-PCR analysis was performed using the LightCycler (Roche Diagnostics) and SYBR RT-PCR kits (Takara Biotechnology Co., Ltd.). For mRNAs analysis, the primers were: human NEK2 forward, 5′-CCG CCC AAG TCA CAG CAG CAGA-3′ and reverse, 5′-GGG AGA AGA GCA GCA GCA GGA AGC-3′; human β-actin forward, 5′-CAT CCT GCG TCT GGA CCT GG-3′ and reverse, 5′-TAA TGT CAC GCA CGA TTT CC-3′; human ABCB1 forward, 5′-AGG CTC GCC AAT GAT GC-3′ and reverse, 5′-TCC TGT CCC AAG ATT TGC TAT-3′; human ABCG2 forward, 5′-TGA AAC CTG GTC TCA ACGC-3′ and reverse, 5′-AGA GTG CCC ATC ACA ACA TC-3′; human ABCC1 forward, 5′-CGC CTT CGC TGA GTT CCT GC-3′ and reverse, 5′-AGT TCT GCG GTG CTG TTG TGG-3′. The relative expression level of mRNAs was normalized to that of internal control GAPDH by using 2^−ΔΔCt cycle threshold method (13).

Whole cell lysates of QGY-7703 cells were prepared in radioimmunoprecipitation assay buffer [50 mmol Tris-HCl (pH 8.0), 150 mmol NaCl, 1% NP40, 0.5% sodium deoxycholate, and 0.1% SDS]. Briefly, cells were washed in ice-cold 1X PBS (pH 7.4) and lysed directly in radioimmunoprecipitation assay buffer on ice for 30 min, centrifuged at 4°C, 14,000 rpm, to remove insoluble material and total protein concentration of the resulting supernatant determined by bicinchoninic acid assay (Pierce Biotechnology, Inc., Rockford, IL, USA). Lysate, 50 µg, for each cell line was resolved by sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE), For immunoblot analysis, protein samples were subjected to SDS-PAGE and transferred onto a nitrocellulose membrane. The membrane was incubated with a primary antibody overnight at 4°C after blocking with 5% skim milk in 0.1% TBST (Tris-buffered saline with 0.1% Triton X-100) for 60 min. The second day, the membrane was incubated with a horseradish peroxidase-conjugated secondary antibody for 60 min after washing three times with 0.1% TBST, then, the membrane was incubated with the ECL solution after washing three times with 0.1% TBST, and exposed to X-ray film.

Immunohistochemical studies on NEK2 were performed on formalin-fixed, paraffin-embedded tissue sections obtained from the aforementioned patients with HCC. Tissue sections were deparaffinized and then boiled in 0.01 mol/l sodium citrate buffer (pH 6.0) in a 1,000 W microwave oven for 10 min to retrieve cell antigens. The primary antibody used was rabbit polyclonal NEK2 antibody (1:200 dilution; Bioss, China). All tissue sections were immunohistochemically stained using the avidin-biotin-peroxidase method and were counterstained with hematoxylin. The staining was scored by three independent investigators without knowledge of patient outcomes. The sections were evaluated at low magnification (×100 or ×200) to identify areas where NEK2 was evenly stained.

The data were subjected to statistical analysis using the SPSS software package (version 17.0). The clinicopathological parameters were tested by χ² test and bivariate analysis (14). All data are presented as the mean ± SD. The statistical differences were analyzed by Student's t-test and repeated measures of ANOVA. The differences were considered to be statistically significant at P<0.05.

NEK2 is considered an oncogene and is overexpressed in various tumors (8,15,16). In order to examine the roles of NEK2 in HCC development, we examined the expression of NEK2 in HCC tissues and cell lines. As shown in Fig. 1A, compared with negative control HFF (human fibroblast cell line), in HCC cell lines HepG2, BEL-7402, QGY-7703, SMMC7721 and Huh-7, we found that NEK2 was highly expressed. Furthermore, NEK2 expression was significantly increased in HCC tissues as compared to that in the matched non-tumor tissues. Hence, these data confirm that NEK2 expression is significantly increased in both HCC tissues and cell lines, which indicates an important role in HCC carcinogenesis and progression.

We further examined the correlation between NEK2 and the clinicopathological characteristics of patients with HCC. As summarized in Table I, NEK2 expression was not significantly correlated with age or histological grades. However, NEK2 expression was significantly correlated with tumor size, differentiation grading and lymph node metastasis (P=0.001, 0.045 and 0.029). Together, these data suggest that elevated NEK2 expression in HCC is correlated with the clinical progression of HCC patients.

As NEK2 expression is increased in HCC, we next investigated the roles of NEK2 increase in HCC development. RNA interference has emerged as natural and highly efficient mechanism for gene silencing (17–19). In order to test the functional role of NEK2 on cell growth, we have designed three siRNA candidates based on NEK2 gene sequence. Real-time PCR confirmed a remarkable downregulation of NEK2 expression in QGY-7703 cells after transfection of these NEK2-siRNAs (Fig. 2A). Additionally, relative amounts of NEK2 protein were decreased by siRNA1, 2 and 3 respectively, compared with those of control siRNA-treated cells (Fig. 2B). In contrast, transfection with control siRNA did not alter NEK2 expression significantly.

We then examined the effect of NEK2 knockdown on cell growth of QGY-7703 cells. As shown in Fig. 2C, the growth of QGY-7703 cells was substantially suppressed by treatment with NEK2 siRNA1 or 3 compared with control siRNA-treated cells. The results indicate that NEK2 can promote HCC cell proliferation.

Because NEK2 controls microtubule organization, overexpressed NEK2 might affect tumor invasion and migration. Therefore, we studied the effect of NEK2 siRNA administration on the invasion and migration of HCC cells. Compared with their control cells and control siRNA-treated cells, knockdown of NEK2 markedly reduced the invasion of QGY-7703 cells (Fig. 3A and B), and also showed significantly impaired migration of QGY-7703 cells (Fig. 3C and D), suggesting that knockdown of NEK2 may dampen the microtubule organization to inhibit their mobility potential. Together with the role of NEK2 in the promotion of HCC proliferation, we conclude that NEK2 promotes HCC progression.

We subsequently tested if knockdown of NEK2 expression could decrease drug resistance to fluorouracil injection (5-FU) using insensitive SMMC7721 cell lines. 5-FU inhibited cell proliferation of the SMMC7721 cell lines in a dose-dependent manner (Fig. 4A). Results of IC₅₀ from the representative SMMC7721 were 51.1±4.70, 48.69±2.57 and 15.61±1.85 µg/ml in SMMC7721 cells non-treated, treated with ctrl-siRNA or NEK2-siRNA, respectively (Fig. 4B). SMMC7721 cells with NEK2 knockdown showed a significant decrease of IC₅₀, indicating that highly elevated NEK2 could promote drug resistance to 5-FU.

NEK2 is known to regulate the mitotic centrosome separation through reversible phosphorylation of its substrates PP1 (20) and β-catenin (21) in yeast, which is also important for cancer cell drug resistance and proliferation (9). Another study demonstrated that the transcriptional level of NEK2 is increased in many aggressive types of cancer, and at least two pathways, PP1/Akt and Wnt signaling, are involved in NEK2-induced cancer cell progression and drug resistance. Consistently, knockdown of NEK2 by siRNA impaired the phosphorylation of Akt, glycogen synthase kinase-3 (GSK3), NF-κB and decreased nuclear β-catenin (Fig. 5A), which activates drug resistance and pro-survival gene members. Furthermore, we found that knockdown of NEK2 downregulated mitotic checkpoint protein ABC transporter family members, including ABCB1, the drug resistance protein ABCC1 (MRP1), and the breast cancer resistant protein ABCG2 (BCRP) (Fig. 5B). In addition, knockdown of NEK2 decreased the expression of pro-survival gene members of the BCL2 family (BCL2 and MCL1) in SMMC7721 cell lines, while promoted the pro-apoptotic gene members BAD and BAX, which are suppressed by Akt (Fig. 5C). These results strongly suggest that PP1 and β-catenin are the downstream targets of NEK2 in HCC cells, and increased NEK2 could contribute to HCC progression by amplifying the PP1/Akt and Wnt pathway.