Introduction
Head and neck cancer (HNC) is the sixth most common cancer worldwide, the incidence of HNC is estimated at 560,000 new cases and 300,000 deaths annually [
1]. Approximately 90% of HNC arises in the mucosa of the oral cavity, oropharynx, larynx and hypopharynx. According to a histopathological perspective, more than 90% of HNC are squamous cell carcinomas [
2]. Both environmental factors and genetic inheritance can give rise to the development of HNC. Tobacco, alcohol consumption and areca nut are major risk factors for the development of this disease. Patients with early stage can be cured by therapy; however, two-third of HNC patients with advanced disease at time of diagnosis, due to HNC has a high potential for local recurrent, invasion and lymph node metastasis [
3]. The most common treatment modalities for HNC include surgery, radiation, and chemotherapy, often in combination. Despite new treatment options for HNC patients, the 5-year mortality rate has not improved over the past 3 decades. Therefore, investigations specifically aimed at further understanding the molecular basis involving in HNC carcinogenesis can facilitate the integration of diagnosis and therapy for HNC in the future.
Aurora-A, also designated as STK15/STK6, is a serine/threonine kinase that plays a crucial role in mitosis and spindle assembly during various stage of mitosis. Aberrant Aurora-A amplification and/ or overexpression has been reported in human malignancies, such as colon, neuroblastoma, breast, oral, NPC, pancreas, ovary, lung, esophagus cancers [
4-
13]. Increased Aurora-A expression may cause genomic instability, tumorigenesis, metastasis and chemoresistance, correlating with its pro-survival function in cancer cells [
14]. Accumulated reports indicate that evaluated Aurora-A expression is not only correlated with advanced stage of tumors, but also associated with a poorer outcome of patients. Recent studies have shown that oncogenic signaling pathways, such as GSK-3β, c-Myc, AKT, β-catenin, p53 and NF-κb involved in Aurora-A function in cancers [
15]. Based on these results, Aurora-A plays a key converging point of a complex network of oncogenic signaling pathway. However, the role of Aurora-A in the signaling transduction pathway involved in tumorigenesis of HNC has not been fully clarified.
FLJ10540, a protein of 464 amino acids, has been mapped to the 10q23 chromosomal region. This protein contains three coiled-coil domains, a peroxisomal targeting signal 2, a nuclear export signal, and structure maintenance of chromosome (SMC) domain. Frequent overexpression of FLJ10540 has been reported in HCC, lung, OSCC, and NPC [
16-
20]. FLJ10540 is a mitotic phosphoprotein and has been reported to be an essential regulator of cytokinesis controlled by Cdk1 and ERK2 [
21]. Deregulated expression of FLJ10540 may cause cytokinesis defects, genetic instability and oncogenic transformation [
22]. Highly FLJ10540 expression promotes progrowth signaling pathways resulting in cancer cell proliferation, metastasis and poor patient prognosis in human cancers [
16] [
17,
20]. Therefore, the oncogenic potential and essential roles in cell cycle of FLJ10540 make it an intriguing target for anticancer therapeutic intervention.
In the present study, we demonstrate for the first time that Aurora-A induced FLJ10540 expression not only influences FLJ10540/PI3K complex, but also regulates MMP-7 and MMP-10 activations, thereby leading to the proliferation, metastasis of HNC cells and their resistance to cisplatin treatment in HNC.
Materials and methods
Reagents
MLN8237 and GM6001 were purchased from Selleck Chemicals. Cisplatin was purchased from Sigma-Aldrich. All chemicals were dissolved in dimethyl sulfoxide (DMSO) for in vitro studies.
Human HNC tissue samples and IHC
Commercially purchased tissue microarrays (TMAs) included 80 samples of 11 cases in early stage, 59 cases in advanced stage and 10 normal tissue (US Biomax, Inc., Rockville, MD, USA; catalog number HN802). This study was approved by the Medical Ethics and Human Clinical Trial Committee at Chang Gung Memorial Hospital. Tissues were fixed with 10% buffered formalin embedded in paraffin and decalcified in 10% EDTA solution. Representative blocks of the formalin-fixed, paraffin-embedded tissues were cut to 4 mm and deparaffinized with xylene and rehydrated in a series of ethanol washes (100, 90, 80, and 70%). Slides were washed with phosphate-buffered saline (PBS) and treated with 3% H
2O
2 for 30 minutes to block endogenous peroxidase activity. Next, the sections were microwaved in 10 mM citrate buffer, pH 6.0, to unmask the epitopes. After antigen retrieval, the sections were incubated with diluted anti-Aurora-A, anti-FLJ10540, anti-MMP-7 and anti-MMP-10 antibodies for 1 h followed by washing with PBS. Horseradish peroxidase/Fab polymer conjugate (PicTure™-Plus kit; Zymed, South San Francisco, CA, USA) was then applied to the sections for 30 min followed by washing with PBS. Finally, the sections were incubated with diaminobenzidine for 5 min to develop the signals. A negative control was run simultaneously by omitting the primary antibody. The reactivity level of the immunostained tissues was evaluated independently by two pathologists who were blind to the subjects’ clinical information. Between 15 and 20 high-power fields were viewed. Criteria were developed for quantitating the immunoreactivities of the Aurora-A, FLJ10540, MMP-7 and MMP-10 stainings in both the normal and tumor sections using a score range of 0 to +3, where 0 indicated no positive cell staining, +1 less than 10% positive cell staining, +2 10-30% positive cell staining, and +3 more than 30% positive cell staining. Similarly, the stain intensity was graded as +0, +1, +2, or +3 as previously described [
23].
Cell culture, transient transfection, the establishment of stable clones, and luciferase assay
FaDu and SAS cell lines were obtained from the American Type Culture Collection. All cell culture-related reagents were purchased from Gibco-BRL (Grand Island, NY, USA). FaDu and SCC4 cells were grown in DMEM containing 10% FBS and 100 U/ml penicillin and streptomycin (Gibco-BRL) Flag-vector (pcDNA3.1), Flag-Aurora-A and Flag-FLJ10540 were transiently transfected into cancer cells using Lipofectamine (Invitrogen) according to the manufacturer’s instructions. FaDu cells mixed-stably expressing Aurora-A or FLJ10540 were selected with 400 μg/ml G418 (Calbiochem Novabiochem, San Diego, CA, USA) for two weeks. The cell were then harvested and analyzed for exogenous Aurora-A and FLJ10540 expressions by Western blotting. 5′-upstream fragments of FLJ10540 gene (−1 ~ −2000) was amplified from human genomic DNA and verified by sequencing. The PCR fragments were cloned into firefly luciferase reporter vector pGL3-Basic (Promega) NheI and HindIII sites which were designed into the forward and the reverse primers, respectively. For co-transfection experiments, FaDu cells were co-transfected with 100 ng firefly luciferase reporter plasmids (pGL3-Basic or pGL3-FLJ10540), and 10 ng of pRL-TK Renilla luciferase internal control plasmid. After 24 h, the luciferase activity was measured using Dual Glo™ Luciferase Assay System (Promega). Two double-stranded synthetic RNA oligomers (5′-GCAGAGAACUGCUACUUAUtt-3′ deduced from human Aurora-A; and 5′-GGACTTTTAGCAAAGATCTtt-3′ deduced from human FLJ10540; Ambion; Taipei, Taiwan) deduced from human Aurora-A, and one negative control siRNA (#4611G; Ambion) were used in the siRNA experiments.
Immunoblot analysis
For tissue protein extraction, frozen samples were homogenized in RIPA lysis buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% Na-deoxycholate, and 0.1% SDS). The protein concentration in each sample was estimated by Bio-Rad Protein Assay (Bio-Rad, Hercules, CA, USA). Immunoblotting was performed according to standard procedures. Antibodies used in this study include Aurora-A (monoclonal; Epitomics, Burlingame, CA, USA), MMP-7 (monoclonal; Millipore), MMP-10 (monoclonal; Millipore), FLJ10540 (generated by us) and β-actin (monoclonal; Santa Cruz Biotechnology, Santa Cruz, CA, USA). The first antibodies were detected by incubation with secondary antibodies conjugated to HRP (Bio/Can Scientific, Mississauga, ON, Canada) and developed using Western Lighting Reagent. The proteins were explored by X-ray films. The protein expression levels were quantified by ImageJ Software and represented as the densitometric ratio of the targeted protein to GAPDH or β-actin.
Indirect immunofluorescence analysis and microscopy
The indirect immunofluorescence staining on the HNC cells treated with MLN8237 or Aurora-A/siFLJ10540 was performed with anti-FLJ10540, at RT for 2 h. The sections were then washed three times with PBST and incubated with DAPI and goat-anti-rabbit-FITC (Jackson, ImmunoResearch) at RT for 1 h. After washing with PBST, the sections were mounted with GEL/Mount (biomeda corp, Foster, CA). The fluorescence images on the slips were examined using a confocal microscope (Olympus FV10i).
Samples were frozen in liquid nitrogen and stored at −80°C prior to RNA extraction. The cells were homogenized using a Mixer Mill Homogenizer (Qiagen, Crawley, West Sussex, UK). Total RNA was prepared from the frozen tissue samples using an RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. The RNA (2 μg) was then reverse transcribed into cDNA using SuperScript II Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA). For Q-RT-PCR, Aurora-A, FLJ10540 and MMPs Taq-Man probe (ABI) were used to perform the study. Data were represented as mean ± s.d. To analyze the distribution of control and experimental groups, we performed the Wilcoxon signed rank test between two groups for statistical analysis. A P-value of less than 0.05 was significant. GAPDH (ABI) was used as an internal control for comparison and normalization the data. Assays were performed in triplicate using Applied Biosystems Model 7700 instruments.
Migration, and invasion assays
Migration and invasion assays were conducted with FaDu-vehicle, FaDu/Aurora-A/negative, FaDu/Aurora-A/siFLJ10540, FaDu/Aurora-A/DMSO, and FaDu/Aurora-A/GM6001 cells using 24-well Transwell chambers (8-μm pore size polycarbonate membrane; Costar, Corning, NY). For the migration (5 x 103) and invasion (1 x 104) assays, cells were suspended in 400 μl of DMEM containing 10% FBS, then seeded into the upper chamber; 600 μl of DMEM containing 10% FBS were added to the outside of the chamber. After being cultured at 37°C under 5% CO2/95% air for 24 h, the cells on the upper surface of the membrane were removed with a cotton-tipped applicator and the migratory cells on the lower membrane surface were fixed with methanol and stained with Giemsa (Sigma, USA). Cell migration was evaluated by counting the number of FaDu-vehicle, FaDu/Aurora-A/negative, FaDu/Aurora-A/siFLJ10540, FaDu/Aurora-A/DMSO, and FaDu/Aurora-A/GM6001 cells that had migrated by 200× phase-contrast microscopy on three independent membranes, then normalized against the vehicle cells to determine the relative ratio. For the invasion assays, 80 μg/ml of Matrigel (BD Biosciences) were added to the upper surface of the membrane and allowed to gel at 37°C overnight. A total cells (1×105) in 400 μl of DMEM containing 10% FBS were seeded into the upper chamber, while 600 μl of DMEM containing 10% FBS were added to the outside of the chamber. The rest of the protocol was the same as that for the migration assays.
Chromatin immunoprecipitation (ChIP)
ChIP assays were performed according to the protocol from Millipore (EZ–Magna ChIP G Chromatin Immunoprecipitation Kit, Millipore). FaDu cells transfected with vehicle control and HA fused Aurora-A according to the manufacturer’s instruction that was describe above. Chromatin was precipitated using anti-HA antibody and protein A agarose at 4°C overnight and immune complexes were collected by centrifugation. Normal human IgG was used as a control. Cross-links were then reversed at 65°C overnight. The purified DNA was amplified by PCR using FLJ10540 promoter primers pre-denaturation for 3 min at 94°C, denaturation at 94°C for 20 sec., annealing at 47°Cfor 30 sec., and extension at 72°C for 30 sec. for a total of 30 cycles).
Measurement of MMP-7 and MMP-10 protein
The amount of MMP-7 and MMP-10 proteins in the conditioned media was determined using the human MMP-7 and MMP-10 Quantikine ELISA kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. The amounts of MMP-7 and MMP-10 were calculated from a standard curve.
Cell viability assay and colony formation assay
Viability of sub-confluent cells was analyzed by 3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction assay. FaDu/vehicle + negative control, FaDu/vehicle + siFLJ10540, FaDu/Aurora-A + negative control, FaDu/Aurora-A + siFLJ10540, SAS/negative control, SAS/vehicle + negative control, SAS/FLJ10540 + negative control, SAS/siAurora-A, SAS/vehicle + siAurora-A, and SAS/FLJ10540 + siAurora-A cells were seeded at 5 × 103 cells/well in 96-well plates. Next day, cells were treated with MLN8237 (0.025 nM) or cisplatin (15 μM) or in combination for 48 h. After that, MTT solution was added to each well. The plates were stored at 37°C for 4 hour, and then 100 μL DMSO buffer was added and incubated in the dark for 10 min. Absorbance was measured on a microplate reader at 540 nm. The OD values were normalized with the value of control group. For colony formation assay, cells were seeded in 60-mm dishes at a density of 5 × 103 cells. Next day, cells were treated with MLN8237 (0.025 nM) and cisplatin (15 μM) or GM6001 (3 μM) for 48 h. After washing with PBS, the cells were incubated in drug-free complete medium for 15 days. Subsequently, cell colonies were counted after staining with 0.01% crystal violet.
Animal experiments and immunohistochemistry
Parental SAS, SAS/negative control and SAS/siFLJ10540 cells were harvested, washed in PBS, and suspended in a mixture of PBS and Matrigel (BD Biosciences, San Jose, CA, USA). 1 × 106 cells were injected into the flanks of female nude mice. All animal experiments were carried out in accordance with protocols approved by the Animal Use and Management Committee of Kaohsiung Chang-Gung Memorial Hospital. After tumor growth reached 100 mm3, mice were assigned to receive MLN8237 or DMSO by oral gavage for 14 days. Mice were monitored daily and tumor volumes and body weights were measured twice weekly. At the completion of the study, tumors were excised, formalin-fixed and paraffin-embedded for immunohistochemical analysis.
Statistical analysis
All in vitro experiments were performed in triplicates. ANOVA analysis was used to evaluate statistical difference between groups. The correlation between two parameters was determined by the Spearman correlation. The P value of 0.05 or less was considered statistically significant.
Discussion
The surgery, radiotherapy and chemotherapy used alone or in combination, are commonly employed for the treatment of HNC patients. Recently, molecular targeted therapies are in development with the goal of selective approaches to prevent the growth of HNC cells. Aurora-A and FLJ10540 overexpression has been explored in a variety of human cancers. However, before this study, the relationship between these two molecules in HNC has not been investigated.
In this study, we first demonstrated that both Aurora-A and FLJ10540 were not only commonly co-amplified, but also had a similar expression pattern in HNC tumor tissues and cell lines by using public accessory microarray database, Oncomine database, Q-RT-PCR, Western blotting and immunohistochemistry approaches. Gain and loss of function assays on Aurora-A indicated that Aurora-A regulates FLJ10540 expression via binding to the FLJ10540 promoter in HNC cells. In addition, inhibition of Aurora-A kinase activity by MLN8237 was not only significant decrease tumor growth and the FLJ10540 expression in vitro and in vivo, but also prevented the formation of FLJ10540-PI3K complex in HNC cells. Aurora-A overexpression in HNC cells elicited the characteristics of aggressive malignancy, such as proliferation, migration, invasion and centrosome abnormality were dramatic decreased, while endogenous FLJ10540 was suppressed. However, FLJ10540 overexpression significantly reversed the sensitivity of Aurora-A-depleted cells to cisplatin. Next, increased MMP-7 and MMP-10 activities by Aurora-A or FLJ10540 were required for increasing cell motility. In addition, Aurora-A elicited cell motility was dramatic suppressed, while FLJ10540 and MMP-7/MMP-10 were inhibited simultaneously comparing to Aurora-A alone. Finally, the expression of Aurora-A in HNC tissue microarray was correlated with elevated FLJ10540, MMP-7, and MMP-10 expressions. Taken together, we demonstrate a new molecular mechanism that aberrant Aurora-A expression in HNC plays a crucial role as positive activator of MMP-7 and MMP-10 via the function of FLJ10540 to regulate HNC progression.
It is known that overexpressed Aurora-A in cancer cells lead to carcinogenesis in multiple types of human cancers. Previous study has shown that Aurora-A participates in PI3K pathway for cancer cell survival [
34]. Our previous data indicated that FLJ10540 could form a complex with PI3K for promoting tumor growth in HCC [
16]. Recently, we had demonstrated that Aurora-A was activated in advanced stage of squamous cell carcinoma of head and neck cancer [
12]. In addition, FLJ10540 expression was correlated with aggressiveness of oral cavity squamous cell carcinoma [
19]. Thus, these results suggest that Aurora-A and FLJ10540 may have functionally linked in head and neck cancer patients. In this study, we first demonstrated that Aurora-A regulated FLJ10540 expression in transcriptional and post-transcriptional levels in HNC cells, suggesting that Aurora-A might regulate PI3K activation through the FLJ10540 function. As expect, Aurora-A kinase activity inhibition in HNC cells led to significant decrease the FLJ10540-PI3K complex formation as well as FLJ10540 protein levels by using Aurora-A inhibitor of MLN8237. The decreased FLJ10540-PI3K association was also observed while Aurora-A was depleted by Aurora-A-mediated siRNA (data not shown). This is the first time to unravel how Aurora-A participates in PI3K signaling pathway in human cancer. These results not only demonstrated that Aurora-A was an upstream regulator of FLJ10540, but also revealed cross-talk between Aurora-A and FLJ10540 in modulating PI3K pathway.
It is noteworthy that down-regulation of Aurora-A or its targets improves chemosensitivity of human cancers [
35]. However, whether FLJ10540 involved in cisplatin chemoresistance in Aurora-A depleted cells has not been reported. In the present study, compared with the corresponding control cells, the viability of Aurora-A depleted HNC cells was reduced under cisplatin treatment in a dose-dependent. Moreover, FLJ10540 overexpressed in Aurora-A-depleted HNC cells or FLJ10540 transfectants treated with MLN8237 could overcome the chemosensitive characteristic to cisplatin compared to Aurora-A-depleted cells or vehicle control. Thus, these results suggest that overexpressed Aurora-A-raised chemoresistant to cisplatin are at least partly, through regulating FLJ10540 expression.
According to current concept, MMP expression could maintain tumor microenvironment such as stroma cells, so inhibited MMP expression would be more effective in a cancer treatment [
33,
36]. Several studies reported that Aurora-A enhances cancer cell metastasis through MMP-2 expression in human cancers [
37,
38]. By using Q-RT-PCR screening approach, we identified that MMP-7 and MMP-10 expressions, but not MMP-2 were dramatic increased by both Aurora-A and FLJ10540 regulations in HNC cells. In addition, MMP-7- and MMP-10-elicited HNC cell motility were regulated by Aurora-A-FLJ10540 expression. This is the first study to unravel that one of the important upstream regulators of FLJ10540 and MMP-7/MMP-10 may be regulated by Aurora-A in HNC cells. Finally, the immunohistochemical analysis showed a significant correlation that overexpression of Aurora-A was not only elevated expression of FLJ10540, but also increased MMP-7 and MMP-10 expression in HNC tissues. However, the molecule mechanisms of FLJ10540 regulated MMP-7 and MMP-10 expression via Aurora-A modulation need to be further elucidated.
In summary, this study highlights the importance of Aurora-A signaling pathway in the biology of HNC at clinical tissue, cell line and murine tissue levels. The impact of suppression of Aurora-A pathway is imposed across all translational related characteristics of HNC cells such as viability, invasiveness, chemoresistance and in vivo tumorigenesis.
Acknowledgements
We also thank the Center for Translational Research in Biomedical Sciences, Kaohsiung Chang Gung Memorial Hospital, to provide the instruments for this study (CLRPG871342-43). We also would like to Chang Gung Medical Foundation Kaohsiung Chang Gung Memorial Hospital Tissue Bank (CLRPG8B0033) for providing the study materials.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
C.H.C and C.Y.C: clinical tissue collection, experimental design and drafted manuscript. A.Y.W.C and S.H.L.: clinical tissues collection, the design and interpretation data of animal mode. H.T.T, L.Y.S, L.J.S, W.L.W, and S.D.L: carried out microarray analysis, cellular and molecular studies. T.J.C, and T.L.H: clinical tissues collection. All authors read and approved the final manuscript.