From January 2009 to December 2010, after obtaining approval from the Human Subjects Committee of the First Affiliated Hospital of Xi’an Jiaotong University and informed consent of the patients, HCC tissues (including adequately sized tumor tissue samples and the corresponding adjacent non-cancerous tissue samples obtained 5–10 cm from the tumor) were obtained from patients undergoing liver resection for hepatic cancer at the Department of Hepatobiliary Surgery, First Affiliated Hospital of Xi’an Jiaotong University. All tissues were obtained within 30 min following excision from the patient’s body. A part of the sample was cut into small pieces, snap-frozen in liquid nitrogen immediately and stored at −80°C to prepare for analysis of CENP-H gene expression, the other part was fixed with 4% formalin and paraffin-embedded for histological studies. Of the 60 patients, there were 49 men and 11 women. Median age at the time of surgery was 52 years (range, 25–74 years). The histological type in all 60 patients was hepatocellular carcinoma, diagnosed according to the World Health Organization histological classification of tumors of the liver and intrahepatic bile ducts (2000)(26). The disease stages of all the patients were classified or reclassified according to the classification of the International Union Against Cancer (27) and the Chinese Anti-Cancer Association, as outlined in Table I. Clinical information of the samples is described in detail in Table II. Patients with preoperative anticancer treatment or with evidence of other malignancies were excluded from the study. All patients were followed up, and the median duration of follow-up was 20 months (range, 2–39 months).

Total RNA was extracted from HCC tumors and the corresponding non-tumor tissues with TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. First-strand cDNA was synthesized from total RNA with PrimeScript^® RT Master Mix [Takara Biotechnology (Dalian) Co., Ltd., Dalian, China].

The PCR primers sets were as follows: i) CENP-H, forward 5′-CAGTCTAGTGTGCTCATGGAT-3′ and reverse 5′-TCCA TCTGTAGGTTTTGTCG-3′; ii) glyceraldehyde-3-phosphate dehydrogenase (GAPDH), forward 5′-CAAGCTCATTTCC TGGTATGAC-3′ and reverse 5′-CAGTGAGGGTCTCTCTC TTCCT-3′. The RT-PCR conditions included an initial denaturation step for 5 min at 94°C followed by 30 cycles of amplification: 94°C for 30 sec, 55°C for 30 sec and 72°C for 30 sec. After the last cycle, a final extension was performed at 72°C for 10 min, and the RT-PCR products were separated by electrophoresis on 1.5% agarose gels.

Real-time quantitative qPCR of CENP-H cDNA was carried out (Bio-Rad Laboratories, Hercules, CA, USA) with SYBR^® Priemix Ex Taq™ II (Takara Biotechnology). The cycling conditions were as follows: initial denaturation at 95°C for 30 sec, and then 40 cycles of denaturation at 95°C for 5 sec, annealing and elongation at 60°C for 30 sec. Bio-Rad CFX Manager 2.1 software was used for analysis of qPCR. The housekeeping gene GAPDH was used as an internal control for both RT-PCR and real-time qPCR. The optimization and synthesis of primers was carried out by Sangon Biotech Co., Inc. (Shanghai, China). Serial dilutions of the template cDNA were made for reactions to optimize the PCR products within the linear range.

Frozen tissue samples were first solubilized 1 h in lysis buffer (Beyotime Institute of Biotechnology, Shanghai, China) in ice using homogenizers, then centrifugated (14,000 × g; Eppendorf, Germany) 30 min at 4°C. After denaturation, equal amounts of supernatant proteins were separated electrophoretically on 10% SDS/polyacrylamide gels (SDS-PAGE) and transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA) in a tank transfer apparatus (Bio-Rad). The membranes were blocked with 5% skim milk in Tris-buffered saline with Tween (TBS-T) for 2 h, then incubated with mouse anti-CENP-H antibody diluted 1:200 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) overnight at 4°C. The next day, horseradish peroxidase-conjugated anti-mouse immunoglobulin G (HRP; Santa Cruz Biotechnology) diluted 1:5,000 in phosphate-buffered saline (PBS pH 7.4) was used as the secondary antibody. Antigens on the membrane were detected with enhanced chemiluminescence horseradish peroxidase (HRP) substrate (Millipore) according to the manufacturer’s instructions. The intensity of each band was measured using Image Lab 4.0 (Bio-Rad). To confirm equal loading, mouse anti-β-actin antibody (Santa Cruz Biotechnology) diluted 1:1,000 was used as the primary antibodies.

CENP-H expression was detected by immunohistochemistry through a standardized streptavidin-peroxidase (SP) method. For immunohistochemistry, 4-μm paraffin sections were adhered to glass slides, which then were deparaffinized in fresh xylene and rehydrated by passage through an ethanol series. Following antigen retrieval in citrate buffer (0.01 M, pH 6.0) in a pressure cooker for 1–2 min after air jetting, slides were first cooled to room temperate and washed by distilled water and PBS twice successively, then incubated for 10 min with 0.3% (v/v) hydrogen peroxide to block activity of endogenous peroxidase. After blockage with 10% normal goat serum for 10 min, the slides were incubated with mouse monoclonal anti-CENP-H (1:40 dilution; Santa Cruz Biotechnology) overnight at 4°C. The next day, the slides were washed in PBS three times and incubated with biotinylated goat anti-mouse IgG (1:100 dilution; Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China) as the secondary antibodies for 20 min. After being washed with PBS and incubated with peroxidase-conjugated streptavidin (1:100 dilution; Beijing Zhongshan Golden Bridge Biotechnology) for 10 min, each slide was colored with 100 μl 0.02% 3,3′-diaminobenzidine (DAB) (Sigma, St. Louis, MO, USA). Finally, all the paraffin sections were rinsed by running water to terminate coloration, counterstained with hematoxylin, differentiated through 0.1% hydrochloric acid (HCl), washed with PBS, dehydrated in graded ethanol and coverslipped with neutral gum. All the incubations were carried out in a humidified chamber, and non-immune mouse serum replaced the primary antibody as the negative control. The slides were read and scored by two independent experiments under a microscope (Olympus Optical Co., Ltd., Tokyo, Japan). According to Guo et al(21), the extent of staining and the proportion of stained cells were used as criteria of evaluation. First, according to the intensity of staining, the cells were scored on a scale of 1, no staining; 2, weak staining (light yellow); 3, moderate staining (yellowish brown); 4, strong staining (brown). Second, according to the percentage of positively stained tumor cells, the proportion of CENP-H-positive cells varied from 0 to 100%: <5% of the cells, 1; 6–35% of the cells, 2; 36–70% of the cells, 3; >71% of the cells, 4. If the final score, calculated by multiplying the above two scores, was ≥4, the tumor was considered to have high expression; otherwise, the tumor was considered to have low expression.

The liver cancer cell line Hep3B obtained from the Transform Medical Center of Xi’an Jiaotong University was cultured in Dulbecco’s modified Eagle’s medium (DMEM)(Invitrogen) supplemented with 10% fetal bovine serum (HyClone, Logan, UT, USA) and 1% penicillin/streptomycin, and incubated at 37°C in a humidified 5% CO₂ atmosphere.

To observe the expression and localization of CENP-H in Hep3B cells, 1.6×10⁵ cells were fixed by acetone, and incubated with mouse monoclonal anti-CENP-H primary antibody (1:40 dilution; Santa Cruz Biotechnology) overnight after antigen retrieval and blockage, followed by fluorescein isothiocyanate (FITC), conjugated rabbit anti-mouse IgG (1:100 dilution; Beijing Zhongshan Golden Bridge Biotechnology). 4,6-Diamidino-2-phenylindole (DAPI) (Sigma) was used to stain the nucleus. Fluorescent imaging was observed with a fluorescence microscope (Leica QFISH; Leica Microsystems, Tokyo, Japan).

All statistical analyses were carried out using SPSS 19.0 statistical software (SPSS, Inc., Chicago, IL, USA). CENP-H protein and mRNA levels were determined by the t-test. Chi-square and Fisher’s exact tests were used to analyze the relationship between CENP-H protein expression and the clinicopathological characteristics. Survival rate was calculated by the Kaplan-Meier method, and differences were examined by the log-rank test. Factors found to be associated with survival rate were then selected for a stepwise Cox’s multivariate proportional hazard model to determine their prognostic values. P<0.05 was considered to indicate a statistically significant result.

To investigate whether CENP-H is upregulated in the HCC tissue, western blot analysis was performed in 10 matched pairs of HCC tissues and corresponding non-cancerous tissues from the same patient. The result revealed that CENP-H was highly upregulated in all 10 pairs of HCCs (Fig. 1A). The relative expression of CENP-H protein in tumor samples compared with adjacent non-cancerous samples varied from 1.75 to 13.13.

To determine whether upregulation of CENP-H is the result of increased transcription, CENP-H mRNA levels in HCC tissues and corresponding non-cancerous tissues were examined using RT-PCR and real-time qPCR. All HCC tissues showed higher expression of CENP-H mRNA when compared with that in non-cancerous tissues (Fig. 1C). Therefore, relative mRNA levels in tumor samples correlated well with the relative protein levels (Fig. 1B and C). These results revealed that CENP-H was upregulated at both the mRNA and protein levels in clinical HCC tissues and that the overexpression of CENP-H may occur at the transcription level.

To further study the expression and subcellular location of CENP-H protein, immunohistochemical analysis was performed in 60 pairs of paraffin-embedded HCC samples and corresponding adjacent non-cancerous samples. We observed that CENP-H expression was significantly higher in HCC tissues (38/60, 63.3%; Fig. 2B–D) than that in the corresponding adjacent non-cancerous tissues (21/60, 35%; Fig. 2A) (P=0.003). Moreover, the expression of CENP-H appeared to increase with histological grade of HCC; well-differentiated HCC samples had the lowest expression of CENP-H (Fig. 2B), and moderately differentiated samples had higher expression (Fig. 2C), while the expression of poorly differentiated samples was the highest (Fig. 2D). CENP-H was predominantly localized in the nucleus.

After we performed immunohistochemistry, we investigated the correlation between the expression of CENP-H protein and the clinicopathological features in 60 HCC cases. As shown in Table II, statistical analysis revealed no statistical correlation between the CENP-H protein level and age, gender, lymph node metastasis, distant metastasis and venous invasion. However, the expression of CENP-H protein was closely related to histological grade, with higher histological grade associated with a higher frequency of CENP-H overexpression in HCC patients (P=0.001). In addition, statistics showed a significant difference in CENP-H expression in patients categorized according to tumor size (P=0.032). Therefore, expression of CENP-H protein was statistically correlated to both TNM stage (P=0.002) and Chinese (P=0.008) clinical stage.

To study the expression and localization of CENP-H protein in HCC cells, Hep3B cells were stained with anti-human CENP-H monoclonal antibody using an immunofluorescence assay. Consistent with the results of the immunohistochemical analysis CENP-H was present in the nucleus of Hep3B cells (Fig. 3).

To determine the effect of classic clinicopathological characteristics and CENP-H expression on survival in HCC, we analyzed the cumulative survival of patients through Kaplan-Meier analysis (Fig. 4). A log-rank test showed that the low CENP-H expression group had longer survival time, whereas the high CENP-H expression group was associated with shorter survival (P<0.05).

Multivariate analysis in our study revealed that venous invasion (P=0.037), advanced TNM stage (P=0.026) and Chinese stage (P=0.020) were independent prognostic factors for overall survival time in patients with HCC. Overexpression of CENP-H protein was also an independent prognostic factor for overall survival (P=0.046). Tumor size and other clinical parameters were not independent prognostic factors.