Background
Nasopharyngeal carcinoma (NPC) is a disease with remarkable geographic and racial distributions worldwide. It is one of the most common cancers in Southeastern Asia and is highly prevalent among populations originating from Southern China where the yearly incidence rate of NPC is 25-50 per 100,000 people [
1,
2]. In North America and other western countries, the yearly incidence is less than 1 per 100,000 [
1]. NPC is a particular type of squamous carcinoma of head and neck associated with EBV infection, environmental factors and genetic aberrance [
3]. Most NPCs are undifferentiated or poorly differentiated with the following characteristics: fast growth and a great tendency to invade adjacent regions as well as metastasize to regional lymph nodes and distant organs. Although NPCs are usually radiosensitive, local failure and metastasis still occur [
4,
5]. Nasopharyngeal carcinogenesis is a multi-step process with morphological progression involving multiple genetic and epigenetic events [
6]. Thus, identification of molecular and biological changes that occur during carcinogenesis and progression could facilitate investigation of the pathology of the disease and generate new prognostic markers to more accurately predict patients' clinical outcome, helping to individualize treatments for NPC patients.
CENP-F (or mitosin) is a member of the human centromeric proteins (CENPs) family, which is involved in centromere formation and kinetochore organization during mitosis [
7,
8]. Its expression and subcellular localization patterns are regulated in a cell cycle-dependent manner. No detectable expression of CENP-F has been reported in G0/G1, only low levels of expression have been detected in the nuclear matrix during S phase, and CENP-F proteins gradually accumulate in the nucleus in G2 then localizing to kinetochores in mitosis and reach the maximal expression in G2 and M cells [
9]. At the end of mitosis, CENP-F is rapidly proteolyzed by the proteasome [
10]. Accumulating evidence suggests that CENP-F is an important protein involved in chromosome alignment and kinetochore-microtubule interaction. Depletion of CENP-F results in chromosome misalignment and improper microtubule-kinetochore attachment [
11]. It interacts directly with many proteins including CENP-E, NudE, ATF4, and Rb, thereby modulating cell fate [
12]. The kinetochore-targeting domain is located near the C-terminus, a region that is sensitive to farnesyltransferase inhibitors (FTIs), which can prevent CENP-F farnesylation and cause mitotic chromosome alignment defects [
13]. A recent report showed that zoledronic acid (ZOL) can inhibit farnesylation of CENP-F and disrupt proper localization and functioning of the protein [
14].
Cell cycle-specific expression of CENP-F makes it a potential marker of proliferation. Indeed, CENP-F is correlated with tumor proliferation in a variety of human tumors, including lung cancer [
15], non-Hodgkin lymphoma [
16], salivary gland tumors [
17], and mantle cell lymphoma [
18]. CENP-F is also correlated with early recurrence in intracranial meningiomas [
19] and poor prognosis in breast cancer [
20]. The CENP-F gene is located on 1q32-q41, which is frequently amplified in NPCs as shown by comparative genomic hybridization analysis [
21]. Using a cDNA microarray, we analyzed the global gene expression profile of primary cultured NPC cells and found that CENP-F is significantly upregulated in NPC cells compared with normal nasopharyngeal epithelial cells [
22].
Our previous studies raised important questions regarding patterns of CENP-F expression in human NPC tissues, potential correlations with clinicopathologic grade and prognosis, and its potential role in chemotherapy. Here, we found that CENP-F was upregulated in NPC cell lines and tissues. Immunohistochemistry analysis revealed that CENP-F expression was positively correlated with clinicopathologic features and inversely correlated with overall survival. Cox regression analysis identified CENP-F as an independent factor for clinical prognosis. More importantly, we revealed that combining cisplatin with ZOL or FTI could have synergistic effects in NPC cell lines with high CENP-F expression. Taken together, our results suggest that CENP-F could be a potential prognostic biomarker for clinical outcome and a promising indicator for selective therapeutic treatment in NPC.
Discussion
In this study, we revealed that CENP-F is upregulated in NPC cell lines and NPC specimens at both the mRNA and protein levels in comparison with noncancerous nasopharyngeal epithelial cells and tissues. Overexpression of CENP-F was significantly associated with advanced clinical stage, higher T classification, skull-base invasion, and distant metastasis. Moreover, as an independent prognostic factor, overexpression of CENP-F was inversely correlated with the prognosis of NPC patients. Additionally, we found that ZOL or FTI-277 could significantly enhance the chemosensitivity to cisplatin of NPC cell lines with high expression of CENP-F, but not in cell lines with low expression of CENP-F, suggesting that CENP-F is a potential target of ZOL or FTI-277 and expression of CENP-F has potential therapeutic implications in NPC chemotherapy.
Our study suggests that CENP-F plays an important role in the progression of NPC. Upregulation of CENP-F was identified at both the transcriptional and translational levels in NPC cell lines in comparison with a primary NPEC2 and an immortalized NPEC2 Bmi-1 cells. In addition, high levels of CENP-F were detected in approximately half (48.5%) of NPC lesions. The importance of CENP-F in the progression of NPC was further highlighted by our finding that it is correlated with advanced stages and T classification, which was in general agreement with other tumor types [
17,
20,
25,
26]. Importantly, the current study was the first to identify an inverse correlation of CENP-F with skull-base invasion and distant metastasis, strongly suggesting that CENP-F could be used as a valuable factor to identify subsets of NPC patients with more aggressive tumors. Finally, our data show that the high expression of CENP-F correlates with poor prognosis and that the level of CENP-F is a potential independent prognostic factor for NPC, suggesting a function of CENP-F upregulation in the multistage pathogenesis of this disease. These findings are consistent with a previous study on breast cancer [
20].
A large amount of data collected from human tumors suggests that chromosome instability (CIN) plays a causative role in a substantial proportion of malignancies and correlates with tumor grade and prognosis [
27,
28]. CIN, which arises as a result of an abnormal mitosis, can occur because of alterations in mitotic timing, mitotic checkpoint control, or of microtubule or centrosome dynamics [
29]. CIN is commonly found in NPC and is thought to play a contributory role in tumor initiation and progression [
30]. Many kinetochore proteins are proved to be associated with CIN. CENP-E is required for efficient capture and attachment of spindle microtubules and responsible for mitotic checkpoint signal transduction [
31,
32]. Evidence has shown CENP-E silencing leads CIN [
33]. Recent research has revealed that CENP-H was upregulated in primary human colorectal cancers, and ectopic overexpression of CENP-H correlated with chromosome missegregation and aneuploidy [
34]. We have previously shown that CENP-H was also upregulated in most NPC tissues [
35]. The overexpression of CENP-F could affect other centromere-kinetochore components and disrupt normal kinetochore function, consequently causing mitotic delay and lagging chromosomes. Those might contribute to chromosome instability and induce the progression of NPC. A study on primary breast cancer showed that CENP-F expression was associated with CIN, including cyclin E overexpression, nuclear expression of survivin, c-
Myc amplification, aneuploid, and high telomerase activity and poor prognosis [
20]. Other studies in CENP-F depleted U2OS cells showed chromosome alignment defects, but the cells still proceeded through mitosis and became aneuploid [
7]. These results suggested that the relationship between kinetochore proteins may be crucial for appropriate localization and proper functioning of the kinetochore. Our study has also provided new insight into CIN in NPC. However, further studies are needed to clarify the possible link between the biological function of CENP-F and chromosome instability in NPC, which will provide important mechanistic understanding of the role of CENP-F in the development and progression of NPC.
In addition to serving as a potential prognostic biomarker, our
in vitro findings suggest that CENP-F may have a therapeutic implication. Numerous studies
in vitro have shown that ZOL exerts a direct cytotoxic effect on tumor cells via inhibition of cell growth and induction of cell apoptosis, in addition to its effect on osteoclasts [
36,
37]. A recent study on breast cancer identified CENP-F as a potential new molecular target for ZOL, which can cause the loss of CENP-F from the kinetochore by inhibiting farnesylation and be involved in the antitumor effect by impairing correct chromosome separation [
14]. Interestingly, our results did not show a direct antitumor effect of ZOL alone even at concentrations of 50 μM, which were not achievable
in vivo, either in immortalized NPEC2 Bmi-1 or in NPC cell lines. Those data indicate that ZOL alone cannot suppress the proliferation of NPC cells at clinically relevant concentrations. However, a synergistic effect was observed when cells were treated with cisplatin in combination with a clinically relevant concentration of ZOL in high CENP-F expression NPC cells. An additive effect was observed in medium CENP-F expression cells and only a marginal effect was observed in low CENP-F expressing immortalized cells. Likewise, the combination of FTI-277 with cisplatin has been shown to have similar synergistic effects against high CENP-F expression cells. These results suggest that the effects of combined treatments are correlated with high CENP-F expression. Moreover, we found ZOL or FTI-277 causes the reduction of CENP-F from the kinetochore, which is consistent with other reports in human breast cancer cells [
14] and head and neck tumor samples [
13]. In addition, these results indicate that the inhibition of CENP-F might be involved in the synergistic interaction. However, the molecular mechanism underlying the synergistic effects in high CENP-F expression cells is not known, and further work is required in order to confirm this effect. Both CENP-E and CENP-F are found on the kinetochores alongside microtubules and specifically localize to the outer kinetochore plate during M-phase. A recent report has identified that an allosteric inhibitor of CENP-E motor activity can decrease CENP-E function and induce tumor cell apoptosis and tumor regression [
38]. Thus, it will be important to identify novel chemicals, which target CENP-F more specifically, in order to understanding of the role of CENP-F in the development and progression of NPC as well as for the development of a novel targeted therapy.
Methods
Patients
Freshly frozen tissue samples of twelve nasopharyngeal carcinoma biopsies and eleven noncancerous nasopharyngeal biopsies were obtained under fiberoptic nasopharyngoscopy from the Department of Nasopharyngeal Carcinoma, Sun Yat-sen University Cancer Center. A total of 202 paraffin-embedded NPC samples, which were histologically and clinically diagnosed between 1996 and 1999 at the Cancer Center, Sun Yat-sen University, were also included in this study. Patients' consent and approval from the Institute Research Ethics Committee was obtained prior to the use of these clinical materials for research purposes. The clinical characteristics of the NPC patients are described in detail in Table
4. There were 151 men and 51 women, with a median age of 48 years (ranging from 13 to 74 years). The routine staging workup included a detailed clinical examination of the head and neck, fiberoptic nasopharyngoscopy, computed tomography (CT) imaging of the entire neck from the base of the skull, chest radiography, abdominal sonography, a complete blood count, and a biochemical profile. All patients' disease stages were classified, or reclassified, according to the 1992 NPC staging system of China as described previously [
2]. Clinical follow-up information was obtained from the patients' records.
Table 4
Clinicopathologic characteristics of 202 patients and expression of CENP-F
Gender | |
Male | 151 (74.8) |
Female | 51 (25.2) |
Age (y) | |
≤ 45 | 90 (44.6) |
> 45 | 112 (55.4) |
Histologic classification (WHO) | |
Type II | 10 (5.0) |
Type III | 192 (95.0) |
Clinical stage (92 stage) | |
I | 27 (13.4) |
II | 59 (29.2) |
III | 49(24.2) |
IV | 67 (33.2) |
T classification | |
T1 | 83 (41.1) |
T2 | 41 (20.3) |
T3 | 40 (19.8) |
T4 | 38(18.8) |
N classification | |
N0 | 118 (58.4) |
N1 | 33 (16.3) |
N2 | 29 (14.4) |
N3 | 22 (10.9) |
Distant metastases | |
No | 177 (87.6) |
Yes | 25 (12.4) |
Living status (at last follow-up) | |
Alive | 114 (56.4) |
Death from NPC | 88(43.6) |
Skull-base invasion | |
No | 159(78.7) |
Yes | 43 (21.3) |
Expression of CENP-F | |
Low | 104 (51.5) |
High | 98 (48.5) |
Chemicals
Zoledronic acid (ZOL) was supplied by Novartis Pharma AG (Stein, Switzerland). ZOL was dissolved in 0.9% NaCl solution as a 5 mM stock solution and stored at -20°C. FTI-277 was purchased from Sigma Chemical Co (St. Louis, MO, USA). The chemical was dissolved in DMSO at a concentration of 10 mM, and aliquots were stored at -20°C. The chemicals were diluted in fresh media before each experiment.
Cell culture
Primary NPEC2 cultures and immortalized NPEC2 induced with Bmi-1 were established as described previously [
39], and grown in keratinocyte/serum-free medium (Invitrogen, Carlsbad, CA). The NPC cell lines CNE1, CNE2, HNE1, HONE1, SUNE-1, 6-10B and C666 were maintained in RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS; Hyclone, Logan, UT), penicillin (100 units/ml), and streptomycin (100 units/ml) at 37°C in a humidified 5% CO
2 incubator.
Real-time RT-PCR
Total RNA from different cell lines and human tissues were extracted using Trizol reagent (Invitrogen, Carlsbad, CA). After reverse transcription of the total RNA, the first-strand cDNA was then used as a template for detecting of CENP-F expression. Real-time PCR and data collection were performed with an ABI PRISM 7900HT sequence detection system. The housekeeping gene GAPDH was used as an internal control to normalize the expression levels of CENP-F. The primer sequences are sense 5'- GTAGAGGACCAACACCTGCTACC-3', antisense 5'-GTCAGCAAACCCTTTCTTTACAACT-3' for CENP-F, and sense 5'-CTCCTCCTGTTCGACAGTCAGC-3', antisense 5'- CCCAATACGACCAAATCCGTT-3' for GAPDH. To ensure reproducibility of results, all genes were tested in triplicate.
Western blot analysis
Western blot analysis was performed as described previously [
39]. Briefly, cells were harvested and lysed in lysis buffer. The protein concentration was determined by the Bradford dye method (Bio-Rad Laboratories, Hercules, CA). Equal amounts of cell extract were subjected to electrophoresis in 4% SDS-PAGE and transferred to polyvinylidene difluoride membranes (Amersham Pharmacia Biotech, Piscataway, NJ). The membrane was probed with an anti-CENP-F rabbit polyclonal antibody (1:1000; Bethyl Laboratories, Montgomery, TX). Expression of CENP-F was determined with horseradish peroxidase-conjugated anti-rabbit immunoglobulin G (1:2000; Santa Cruz Biotechnology, Santa Cruz, CA) and enhanced chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ) according to the manufacturer's suggested protocols. An anti-α-tubulin mouse monoclonal antibody (1:1000; Santa Cruz Biotechnology) was used to confirm equal loading.
Immunohistochemical staining (IHC)
IHC staining was performed using the Dako Envision system (Dako, Carpinteria, CA) following the manufacturer's recommended protocols. All paraffin-embedded specimens were cut into 4 μm sections and baked for 1 h at 65°C. All sections were deparaffinized with xylenes and rehydrated with graded ethanol to distilled water. Sections were submerged in EDTA antigen retrieval buffer (pH 8.0) and microwaved for antigen retrieval. After being treated with 0.3% H2O2 for 15 min to block the endogenous peroxidase, the section were treated with normal goat serum for 30 min to reduce the nonspecific binding and then rabbit polyclonal anti-CENP-F antibody (1:200; Bethyl Laboratories) overnight at 4°C. After washing, the sections were incubated with biotinylated anti-rabbit secondary antibody (Zymed) followed by further incubation with streptavidin-horseradish peroxidase (Zymed) at 37°C for 30 min. For color reaction, diaminobenzidine (DAB) was used. For negative controls, the antibody was replaced by normal goat serum.
The immunohistochemically stained tissue sections were scored independently by two pathologists blinded to the clinical parameters. The final score for CENP-F was the average of the scores obtained by the two observers. Cases with major discrepancies in scoring (i.e., > 1) were reviewed by both observers on a multiheaded microscope. Based on previous studies [
40,
41], we used the intensity and extent of the staining to assess CENP-F. The entire tissue section was observed to assign scores. The staining intensity was scored as 0 (no staining), 1 (weak staining exhibited as light yellow), 2 (moderate staining exhibited as yellow brown), or 3 (strong staining exhibited as brown). Extent of staining was scored as 0 (0%), 1 (1 to 25%), 2 (26 to 50%), 3 (51 to 75%), or 4 (76 to 100%), according to the percentages of the positive staining areas in relation to the whole carcinoma area or entire section for the normal samples. The sum of the intensity and extent scores was used as the final staining score (0 to 7) for CENP-F. For the purpose of statistical evaluation, tumors having a final staining score of < 3 were grouped into low CENP-F expression and those with scores ≥3 were grouped into high CENP-F expression.
In Vitro Cytotoxicity Assays
Cytotoxicity tests were evaluated by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma, St. Louis, MO, USA) assay. Cells were grown in 96-well microtiter plates at appropriate densities and allowed to adhere for 24 h before addition of ZOL or FTI-277 alone or cisplatin and ZOL or FTI-277 together. To determine the cytotoxicity of ZOL and FTI-277, cells were exposed to increasing concentrations ranging from 0.1 μM to 50 μM for 72 h. The absorbance was determined at 570 nm in a multi-detection microplate reader (SpectraMax M5). To test the effect of ZOL or FTI-277 on the chemosensitivity of immortalized NPEC and NPC cells, ZOL (2 μM) was added to the medium with various concentrations of cisplatin in NPEC2-Bmi-1, SUNE1, CNE1, HONE-1 and 6-10B, and FTI-277 (1 μM) was added in CNE1 and HONE1. The concentrations required to inhibit growth by 50% (IC50) were calculated from survival curves using the Bliss method [
42]. The results from the assays were analyzed for the combination effect between ZOL and cisplatin according to the Zheng-Jun Jin method [
43]. This method provides a Q value, where Q < 0.85 indicates antagonism, 0.85 ≤ Q < 1.15 indicates additivity and Q≥ 1.15 indicates synergism. The formula is Q = E
a+b/(E
a+E
b-E
a×E
b), where E
a+b, E
a and E
b are the average effects of the combination treatment, ZOL only and cisplatin only, respectively. All treatments were performed in quadruplicate and experiments were repeated three times.
Immunofluorescence analysis
Immunofluorescence analysis was performed as described previously [
39]. Cell lines were plated on culture slides (Costar, Cambridge, MA), treated with ZOL or FTI-277 for 24 h, then fixed in ice-cold acetone for 5 min at -20°C. The cells were blocked for 30 min in 10% BSA (Sigma-Aldrich St. Louis, MO) in PBS and then incubated with rabbit polyclonal anti-CENP-F antibody (1:200; Bethyl Laboratories) for 2 hours at room temperature. After three washes in PBS, the slides were incubated for 1 h in the dark with secondary goat anti-rabbit antibodies (Invitrogen, Carlsbad, CA). After three further washes, the slides were stained with 4-,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich St. Louis, MO) for 5 min to visualize the nuclei, and examined using an Olympus confocal imaging system (Olympus FV100).
Statistical analysis
All statistical analyses were carried out using the SPSS 13.0 statistical software package. The significance of CENP-F mRNA levels and the MTT assays were determined by t-tests. Chi-square and Fisher's exact tests were used to analyze the relationship between CENP-F expression and clinicopathologic characteristics. Bivariate correlations between variables were calculated by Spearman's rank correlation coefficients. Survival curves were plotted by the Kaplan-Meier method and compared by log-rank test. Univariate and multivariate regression analyses were performed with the Cox proportional hazards regression model to analyze independent factors affecting prognosis. P < 0.05 was considered statistically significant.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JC and LL performed experiments and were responsible for data collection, analysis, interpretation of the results, and writing the manuscript. SC, XZ, YM were responsible for conducting the data analysis in cooperation with ZL and LF. ML and HZ were responsible for reviewing and scoring the immunostaining of sections. CQ and JS provided clinical samples for performance of experiments and validation of data. MZ and YX were responsible for experimental design, analysis and interpretation. All authors have read and approved the final manuscript.