Journal of Oncology Research and Therapy

Volume 2017; Issue 01
5 Aug 2017

Aurora-A Overexpression Plays a Role in Developing Aneuploidy in a Subset of Oral Squamous Cell Carcinoma

Research Article

Al-Hazmi NA1, Alhazzazi TY1, Albahiti Haji M2, Williams GH3, Stoeber K4, Speight PM5, Al-Dabbagh RA6*

1Department of Oral Biology, King Abdulaziz University, Faculty of Dentistry, Jeddah, Saudi Arabia.
2Specialist Endodontist, Department of Endodontics, King Abdulaziz University Faculty of Dentistry, Jeddah, Saudi Arabia.
3Medical Director at Oncologica® London, United Kingdom
4Department of Pathology, UCL Cancer Institute, University College London, London, UK
5School of Clinical Dentistry, University of Sheffield, Claremont Crescent, Sheffield, United Kingdom
6Department of Oral and Maxillofacial Prosthodontics, King Abdulaziz University, Jeddah,Saudi Arabia

*Corresponding Author: Nadia Abed Al-Hazmi, Department of Oral Biology, King Abdulaziz, Saudi Arabia, Tel: +966 6402000; Email:

Received Date: 1 December, 2016; Accepted Date: 23 December, 2016; Published Date: 2 January, 2017







Suggested Citation





Globally, oral cancer is the eighth most prevalent cancer in men and of these 94% are squamous cell carcinomas [1]. The survival rate of oral cancer is poor and averages 50%. Despite improvements in treatment, the survival rates have only marginally improved during the past three decades [2]. This is due to inconsistency in traditional clinical (e.g. TNM stage), radiologic, and histopathologic (e.g. the histological grade of tumour differentiation) parameters used to determine the stage and subsequent treatment strategies required [3]. As a result several attempts have been made to search for molecular and genetic biomarkers thatpredict tumour behaviour and aggressiveness as well as enabling a guided personalised treatment plan [4-6].


Aurora-A is a member of the mitotic serine/threonine kinases family (Aurora-A, -B & -C), and has essential functions in controlling centrosome maturation and separation, chromosome condensation, segregation and orientation on the metaphase plate, as well as during cytokinesis [7]. In addition, Aurora-A is expressed in most cell types, but differs in localization and time of activation [8].


The expression levels of the Aurora kinases mRNAs and proteins are elevated in certain cancers, but they are not always correlated with gene amplification [9-11]. The human Aurora-A gene is located at chromosome 20q13.2-q13.3, and itis amplified in several cancer types including colorectal, pancreatic and head and neck cancers [12-14].


Aurora-A is an oncogene due to its transformation and induction of NIH3T3 and rat1 cells when overexpressed ectopically. Furthermore, these transformed cells induced tumours when introduced to nude mice [12,15] However, this overexpression did not induce transformation in primary Mouse Embryonic Fibroblasts (MEF) cells [16]. Because NIH3T3 and rat1 cells contain several genetic abnormalities, one can assume that Aurora-A is a weak oncogene and other defects cooperate with it to induce oncogenesis [17].


Nevertheless, the precise link between Aurora-A and mitotic aberrations, aneuploidy and carcinogenesis in OSCC is still unknown. Therefore, the aim of this study was to perform a detailed immunohistochemical analysis of Aurora-A expression in primary OSCC and fibro-epithelial polyps and then correlate the expression data with aneuploidy, histopathological parameters and clinical outcome.



Material and Methods


Patients and Archival Tissue Samples


Ethical approval was obtained from the Joint Research and Ethics Committee in the Eastman Dental Institute (EDI) and Hospitals for patients diagnosed with primary OSCC who had incisional or excisional biopsies or surgeries. Histopathological records of the EDI and Department of Pathology of the University College London, UK were used to identify subjects to be included in the study. All samples had originally been fixed in a 4% dilution of 10% formalin saline concentrate, processed and embedded in paraffin wax. Exclusion criteria included patients who underwent pre-operative radiotherapy or chemotherapy. Patients’ notes were used to retrieve the following information: date of birth, gender, and site of primary tumour, date of diagnosis, differentiation, date and type of surgery, TNM staging and follow-up periods.


Three and five-year survival data were retrieved from the Thames Cancer Registry Office. Histologic diagnoses were subjectively classified according to the Broder’s criteria using a modification of his classification system [18]. The TNM system was used to classify tumours clinically as follows: Stage I, T1N0M0; Stage II, T2N0M0; Stage III, T3N0M0, or any T with N1 M0; and Stage IV, any T with N2M0, N3M0, or any N with M1.


In addition, grading of the mode of invasion was objectively carried out according to Odell et al. [19] as follows: Grade 1; well delineated borderline; Grade 2; infiltrating cords, bands and strands; Grade 3; small groups or cords of less than 15 cells; and Grade 4; marked extensive cellular dissociation in small groups or single cells. For reliability of histopathological classification, 10% of randomly selected cases were re-assessed by two different pathologists (Bill Barrett and Garth Thomas). As a control group, 18 fibroepithelial polyps were identified from the records. Patients’ notes were used to gather the following information:  date of birth, gender and date of diagnosis.





The anti Aurora-A mouse anti-human antibody (clone JLM28) was obtained from Novacastra Laboratories (Newcastle, UK). The anti Aurora-A was raised against a 131 amino acid region of the N-terminus of human Aurora-A kinase, which was cytoplasmic and formed a 46-kDa polypeptide in Western blots (Novocastra Data Sheet 2004).


Sodium Dodecyl Sulphate – Polyacrylamide Gel Electrophoresis (SDS-PAGE) and Western Blots


The protein was examined in synchronized HeLa53 cell lysates. Lysates were denatured by boiling in 4xlamelli sample buffer and separated according to the size by SDS-PAGE. Subsequently, the proteins were transferred to a nitrocellulose membrane and incubated with the primary and secondary antibodies. Finally, immune complexes were visualised using an Enhanced Chemi Luminescence (ECL).


In detail, the experiment was carried out as follows: a comb and a white strip from a pre-packaged 10% gel was removed and placed in a gel tank. The central tank was filled with 1x SDS buffer until it covered the wells and the outer tank (1/3 total volume). The cell lysates and the controls were boiled in 4x laemlli buffers for 5 minutes. The samples were spun down at 130,000rpm for 5 minutes and loaded into the wells. The gel was run at 80mV for 10 minutes and at 125mV for 90 minutes. Just before transferring the gel with the samples, a fresh transfer buffer (1.2M Tris, 0.04M CAPS, 0.08% SDS and deionized water) was prepared and 10 pieces of 3MM (7cm x 9cm) and 1 piece of Hybond C-extra nitrocellulose membrane were soaked with the buffer. Then, 5 pieces of the 3MM were built on a transfer apparatus, and the membrane was laid on top. The gel was peeled off into the transfer buffer and laid on top of the membrane. Finally, the remaining five pieces of 3MM were placed on top. The sample was transferred from the gel to the nitrocellulose membrane at 15mV for 30 minutes. Next, proper transfer was confirmed by placing the membrane into Ponceau S and shaking it for a few minutes. The Ponceau S was then poured off and washed with UHQ water. After transfer, the membrane was blocked overnight in milk/Tween-20 (5% milk/0.1% Tween made in Phosphate Buffered Saline (PBS). The next day, the membrane was washed 3 times for 5 minutes (using PBS/Tween20 for this and subsequent washes), and the primary antibody was incubated for 2 hours at Room Temperature (RT) (Aurora-A 1:100). Again, the membrane was washed 3 times for 10 minutes.  Then, the secondary antibody (1:2500) was incubated for 1 hour and washed 3 times for 5 minutes. To visualise the reaction, 1ml of each of the ECL-developing reagents were mixed per blot, added to the blot on clingfilm and left for 1-2 minutes followed by exposure to an auto-rad film.


Antibody Optimisation


Four-μm-thick sections were cut using a sledge microtome (Leica SM2400) and mounted on organosiliane-coated slides that were dried and baked overnight in a 60ºC oven. Parallel batches of sections from the chosen blocks were put into two different antigen retrieval pH solutions. One batch was pressure-cooked in citrate buffer at pH 6.0 for 2 minutes, and the other batch was microwave cooked in TE pH9.0 for 25 minutes. Labelling used four different detection systems: ChemMate (LSAB) (Dako), Envision (DakoChemMate K5007), Bond x (Vision Biosystem) and Bond enhancer (Vision Bio-system). Different primary antibody concentrations were incubated in parallel with each detection system (Aurora-A 1: 50 and 1:100 and 1: 200). The subsequent steps for the automated staining were as described in the next section (immunohistochemistry).


For manual staining, the primary antibody was washed twice for 2 minutes, drained, and excess buffer was removed. The sites of primary antibody binding were identified by incubation with the secondary antibodies for 60 minutes at RT, washing twice for 2 minutes, and draining to remove excess buffer.  Immunostaining was developed with 3,3′- -diaminobenzidinechromagen for 7 minutes, rinsing with TBS/Tween20 and washing in tap water. The sections were then counterstained with Mayer’s haematoxylin for 2 minutes, differentiated in 1% acid alcohol, dehydrated through graded alcohols, cleared in xylene, and mounted in a Leica CV mount (Leica CV5000) with cover slips.




Cover slips were autoclaved and placed into culture dishes. The 25 ml of Dulbecco’s Modified Eagle’s Medium (DMEM) including 10% Foetal Cell Serum (FCS) and streptomycin were added to the culture dishes. Then, 2ml Hela cells in DMEM were added to the dishes. After a few days, the cells were at least 60% confluent, and they were washed in PBS, fixed in 4% paraformaldehyde for 5 minutes and washed twice in PBS for 5 minutes. They were then permeabilized in PBS/0.1% Triton/ 0.02% SDS for 5 minutes, washed in PBS 5 minutes and blocked in PBS/1% Bovine Serum Albumin (BSA) for 10 minutes. Next, 50ml of the primary antibody (Aurora-A 1:100) in PBS/ 2% BSA were incubated per well in a humidified chamber for 1 hour at 37oC and subsequently washed twice in PBS/0.01% triton/0.02% SDS and twice PBS/2% BSA for 5 minutes.


Next, a fluorochrome-coupled secondary antibody (50 ml per well; anti-mouse and anti-rabbit) (IgG/FITC)(1/200), the counter stain propidium iodide (50ng/ml), and RNAse A (50ng/ml) in PBS/2% BSA were added. The secondary antibody was incubated in a dark humidified chamber for 1 hour at 37oC, followed by 3 washes in PBS/0.01% Triton/0.02% SDS for 5 minutes. Finally, the coverslips were mounted on slides using 3.5ml of mounting media (Vectorshield) per slide, sealed with nail polish, labelled and imaged under a confocal microscope (Leica TCS DMRE).




Paraffin blocks were available for all patients; blocks contained tissue peripheral to the tumour. The fixation methods at both institutes were similar. The fixative was a 4% dilution of 10% formalin saline concentrate.  Sections from the chosen blocks were serially cut, prepared, deparaffinized, rehydrated and antigen retrieval was as mentioned in the section Antibody Optimisation. Next, immunostaining was automated using a Bond immune stainer (Vision Biosystem). Non-specific activity of endogenous peroxidase was quenched with peroxidase blocking solution for 5 minutes. The slides were then washed (using a wash solution for this and subsequent washes) 3 times for 2 minutes and drained.  Subsequently, the primary antibody was incubated for 30 minutes, washed 3 times for 2 minutes and drained. The best working antibody dilution in this series was Aurora-A 1/40 (A: TBS). Then, a post primary antibody was incubated for 20 minutes and washed 3 times for 2 minutes. After that, a polymer-based secondary antibody and peroxidase molecules were incubated for 20 minutes, washed 3 times for 2 minutes and drained.


Immunostaining used DAB for 10 minutes with subsequent washing in deionised water 3 times for 1 minute with draining. Next, an enhancer was applied for 5 minutes, washed with deionised water 3 times for 1 minute and drained.  This enhancer is a copper sulphate (CuSO4) compound, which reacts with the peroxidase molecules and precipitates them. The precipitate is dark brown or black in colour, i.e. it enhances the chromogen’s colour. Finally, the slides were counterstained with Mayer’s haematoxylin for five minutes, washed with deionised water, drained, dehydrated, cleared, mounted and analysed as mentioned above. For quality control, appropriate tissue sections from the colon were used as positive controls; these underwent identical sample preparation.


Protein Expression Profile Analysis


The Labelling Indices (LI) of each tumour were calculated to analyse the Aurora-A protein expression. Initially, a pathologist used the Olympus BX51 microscope to scan all slides under low powerand identify a representative area. Next, images from the selected area were captured with a CCD camera and Analysis software (SIS, Munster, Germany) from the invasive front towards the periphery (Kodani et al. 2001) at 10x magnification for orientation purposes. A tumour axis was determined and another 2 to 5 images were taken following this axis at 20x magnification. These 20x magnification images were enlarged to 30x magnification for counting. Then, images were printed for quantitative analysis. Analysis was done without knowledge of the clinico-pathological records. Tumour cells were counted at the chosen fields ignoring any stromal or inflammatory cells. The presence of any nuclear staining was considered positive, and the LI was calculated by dividing the number of positive cells by the total number of counted cells.


In addition, the intensity was assessed semi-quantitatively as follows: ++, high expression was detectable within the lesion; +, moderate expression was detectable within the lesions; +/-, expression was weakly detectable in part of the lesions; and -, expression was not detectable within the lesions. This semiquantitative scoring was adapted from Tanadaet al., however we modified it for statistical significance [20].


Image Cytometric Analysis for Ploidy


The DNA content was analysed using the method of Sudboet al. [21]. The most representative area of the lesion was outlined in each slide. The selected areas were identified in the blocks and marked. These areas were micro-dissected, and 1-6 sections that were 50-micron thick were cut and placed in Falcon tubes. The sections were cleared by adding 10ml of xylene for 10 minutes with removal of supernatant (repeat twice). They were then hydrated by adding 10 ml of 100% and 95% ethanol for 5 minutes each with 1ml of deionised water for 5 minutes (the last step was repeated twice). Next, 10 ml of cold PBS was added for 5 minutes, and thiswas repeated twice.


Enzymatic digestion was achieved by adding 2 ml of protease XXIV for 2.5 hours in a shaker water bath at 37ºC. After digestion, 3 ml of chilled PBS was added, filtered through a nylon mesh into 15ml Falcon tubes and centrifuged for 5 minutes at 1500 rpm. The pellets were re-suspended in 3 ml of fresh PBS. Then 100-200µl of the re-suspended pellets wascytospun in a cytocentrifuge at 1500 rpm for 5 minutes and fixed in 4% formalin overnight. The next day, the slides were rinsed in tap water, placed in 5 M HCL for 1 hour for hydrolysis, rinsed in distilled water and stained with Schiff’s reagent for 2 hours (the last step is done in the dark).  Subsequently, the slides were washed in running tap water for 10 minutes and dehydrated in increasing concentrations of ethanol (70%, 95% and 100%) for 10 seconds each. Finally, the slides were mounted and scanned using an Axioplan 2 imaging microscope. The nuclei were analysed with Fairfield imaging software, Kent, UK. The lymphocytes were used as internal controls to guide histogram scaling.


Histograms were interpreted based on Sudbo’s classification of aneuploidy [21]. Diploid lesions were defined by the presence of only one 2c peak or if the number of nuclei in 4c did not exceed 10% of the total number of epithelial nuclei or if the number of nuclei with more than 5c DNA content was less than 1% of the total number of nuclei. The lesions were classified as tetraploid when their G0/G1 (4c) peak was present together with its G2 peak (8c)or when the fraction of the nuclei in the tetraploid region exceeded 10% of the total number of nuclei.  Lesions were classified as aneuploid by the presence of non-euploid peaks or if the number of nuclei with DNA content was greater than 5c or 9c exceeded 1% [21].


Statistical Analysis


Independent sample tests to evaluate equality of means were carried out for reliability of analysis, and these showed no significant differences. The expression profiles of Aurora-A were then compared to both OSCC and FEP using student’s t test and the chi-square test. Association between Aurora-A expression and other factors were assessed using the Mann-Whitney U test, the Kruskal-Wallis test, the Jonckheere-Terpstra test, and Spearman’s rank correlation coefficient test. All statistical results were two-sided. The expression levels of Aurora-A in relation to overall survival periods and disease free survival periods were tested with Kaplan-Meier analysis using the log-rank test. This statistical analysis used SPSS 12.0.1 for Windows (SPSS, Inc., Chicago, IL).




Demographic Data


Of 172 HNSCC specimens collected, 125 OSCC cases were suitable for immunohistochemical analysis. Here, 9, 12, and 26 cases were excluded because ofexposure to radiotherapy or photodynamic therapy, extra oral origin (8 from the oropharynx, 2 from the maxillary antrum, 1 from the scalp, and 1 from the parotid gland), and insufficient tissue or proper orientation for analysis, respectively. The immunohistochemical samples included 35 biopsies and 90 resections from 80 males and 45 females with a male to female ratio of 1.8:1. The age range was 27-96 yearswith a median age of 60 years. The primary site of the OSCC included 54 cases from the tongue (22 from the lateral border and the rest were not specified), 39 from the alveolar mucosa, 21 from the floor of the mouth, 6 from the buccal mucosa (1 from the commissures), 4 from the lower lip, and 1 from the hard palate (Table 1).


The 18 fibroepithelial polyps were all suitable for analysis. Five were males, and the remaining 14 cases were females. The age ranged from 18 to 78 with a median age of 44.5 years.


Histologicpathologic, Clinical Staging and Clinical Outcome Data


From the 125 OSCC, 113 cases were conventional OSCC, 9 were verrucous, 2 were adenosquamous and one was basaloid. Twenty-six cases were well differentiated, 71 were moderately differentiated, and 28 were poorly differentiated.  In terms of tumour invasiveness, 9 were classified as having pushing fronts, 53 had bands at the front, 38 had cords at the front and 25 had a diffuse front.There were 82 cases with TNM staging information including 41 with lymph node metastasis. Information on follow up periods was available over a three-year survival period in 79 cases (63.2%): 31 survived and 48 did not. There was 5-year survival data from 65 cases (52%): 14 survived and 51 did not (Table 1).


Antibody characterization and optimisation


IF and WB confirmed the specificity of the Aurora-A antibody. This antibody is specific based on the protein’s expected size and cellular localisation. Figure 1 shows positive Aurora-A immunofluorescence staining in the cytoplasm and localized in the centrosomes. There is an expected band size of 46 kDa in Western blots.  After utilizing different detection methods and antibody concentrations, the optimum Aurora-A expression was achieved at a 1/40 concentration with the bond enhancer detection method.


Aurora-A Expression in FibroEpithelial Polyps and OSCC


A total of94.4% (17/18) of the FEP showed expression of Aurora-A. This expression was demonstrated in the cytoplasm of a small number of basal cells of the FEP (Median LI; 3.7%) (Figure 2). The limited expression of Aurora-A was further complimented by the difference of its LI versus other proliferative markers such as Mcm2, Ki67 and Geminin (not published results) [Median: Mcm2, 41.6%; Ki67, 21.15%; Geminin, 5.5%; Aurora-A, 3.4% (paired t test, P < 0.001 each)] (Figure 3).  The FEP specimens were predominantly (55.6%) undetectable (-, 1 FEP from a total of 18 FEP specimens) or weak (+, 9/18) expression of Aurora-A in the cytoplasm. We noted that 44.4% (8/18) of the cases had medium to high expression. Furthermore, all analysed FEP had normal diploid DNA content.


However, Aurora-A’s expression was noted in 88.8% (111/125) of OSCC with a diffusely distributed expression in the cytoplasm including at the peripheral portions of the OSCC nests. They were sometimes found scattered throughout some tumours (Figure 4). The diffuse cytoplasmic distribution of Aurora-A might imply that this protein is ectopically overexpressed in OSCC rather than its normal localisation in the centrosomes and mitotic spindle. In addition, Aurora-A was expressed in proliferating cells in OSCC as seen by its intensity association with Ki67 expression (Jonckheere-Terpstra test, P= 0.02) (Figure 5). However, only a subset of cycling cells expressed Aurora-A as indicated by its low LI versus other proliferative markers such as Mcm2, Ki67 and Geminin (data not shown) [Median: Aurora-A, 4.51%; Mcm2, 71.9%; Ki67, 55.2%; and Geminin, 18.8% (paired t test, P < 0.001 each)] (Figure 6).


The OSCC specimens had frequently (68.8%, 86/ 125) high (++, 51/ 125) or medium (+, 35/125) expression in the cytoplasm of tumour cells. This elevated intensity might represent ectopic overexpression of Aurora-A in cell-cycle phases other than the normal G2/M phases. Thismight implicate Aurora-A in OSCC carcinogenesis. Moreover, the higher Aurora-A expression in OSCC relative to FEP (Chi-square test, P= 0.025) might further implicate it in OSCC carcinogenesis. Approximately, 31% of cases had weak expression (+/-, 25/ 125) or no detectable expression (-, 14/ 125).


Relationship between Aurora-A expression in OSCC, aneuploidy, histopathologic, and outcome parameters


Next, we used image cytometry of DNA content to investigate the link between Aurora-A expression and DNA content. A nuclear monolayer was prepared for 125 OSCC subjects. Of these, 103 OSCC were suitable for analysis and histogram interpretation (Figure 7 shows histograms and their interpretation of representative OSCC). The number of nuclei analysed per case ranged from 354-1053 nuclei (median 809) with a median of 10 lymphocytes as internal controls. The 19 tumours had diploid DNA content, and 85 cases had aneuploid DNA content. Of the 85 aneuploid OSCC, 64 (75.3%) had medium to high Aurora-A expression versus the 31.6% of the diploid tumours with moderate to high expression (Chi-square test, P= 0.01) (Figure 8). Thus, ectopic overexpression of Aurora-A might play an important role in inducing aneuploidy and tumour formation in OSCC. However, 31.6% (6/19) of the diploid tumours and 42% (8/19) of the diploid fibroepithelial polyps had ectopic overexpression of Aurora-A (medium to high expression). This might imply that Aurora-A ectopic overexpression does not independently induce aneuploidy and tumourgenesis in OSCC (Table 2).


Aurora-A expression measured by LI did not correlate with any histopathologic parameters, clinical outcome parameters or DNA content. Accordingly, there was no association between Aurora-A expression and differentiation including early and stage (Jonckheere-Terpstra test, P= 0.315, P=0.59 and P=0.85 respectively). There was no association between Aurora-A expression and gender, age or DNA content (Mann-Whitney U test, P=0.302, P=0.94, P=0.295 and P=0.752 respectively). There was no correlation between Aurora-A expression and lymph node metastasis (Spearman’s correlation coefficient, P= 0.688). Additionally, Aurora-A expression was not a good predictor of five-year survival rates (Log-Rank test, P= 0.17) (Table 2 and 3).




Oral cancer is the eighth most common tumour worldwide, and approximately half of all patients die within five years [22,23] This poor survival is attributed to late-stage survival when lymph node metastasis and/or invasion of surrounding tissues have occurred [24]. Therefore, there is a need to understand the molecular mechanisms of oral cancer,particularly OSCCto enable new approaches for diagnosis and treatment. Cancer development is partly due to alteration in the expression and/or mutations of cell-cycle regulators. Abrogation of mitosis and its checkpoints have been reported to result in abnormalities in nuclei and chromosomal segregation with subsequent aneuploidy development [25].


In our study, we showed that 88.8% (111/125) of the OSCC expressed Aurora-A, and the majority (68.8%, 86/125) were found to overexpress it diffusely throughoutthe cytoplasm. This diffuse overexpression suggests that Aurora-A is ectopically overexpressed in OSCC throughout the cytoplasm rather than its normal localisation in the centrosomes and mitotic spindle. Additionally, the high Aurora-A intensity seen in most cells expressing this protein suggests that it is expressed independently of the cell-cycle phases rather than the typical increase seen in G2/M phases (cell-cycle phase dependent). This spatial and temporal ectopic overexpression of Aurora-A promotes oral tumourgenesis via excessive phosphorylation of normal substrates and abnormalphosphorylation of cytoplasmic proteins (involved in oncogenesis) or proteins expressed in other phases of the cell cycle [7]. Aurora-A was overexpressed in 55.6% of FEP, which possibly suggests that Aurora-A is a weak oncogene and that other defects cooperate with it to induce oral carcinogenesis.


To the best of our knowledge, this was the first immunohistochemical report to show that primary OSCC overexpress Aurora-A (67.7%). Tatsukaet al. reported Aurora-A gene amplification (36.36%, 4/11) and mRNA overexpression (100%, 11/11) in tongue and gingival carcinoma, but did not analyse the protein expression [14].  Furthermore, Aurora-A was expressed in only a subset of cycling cells, which might suggest that Aurora-A is only an indicator of cancer cell proliferation relative to other existing proliferative markers.


There was no association between Aurora-A overexpression and clinic pathological details or outcome parameters. These finding weresupported by other reports of breast tumours [26]. One reason might be that the Aurora-A expression analysis by IHC is not the most effective method of analysis [27]. Other methods such as gene amplification and mRNA might be better atcorrelating with clinical and histopathologic parameters. This might be because when the sample is subdivided according to different histopathologic parameters, the subgroups become small, and the statistically significant differences become less obvious. It is also possible that Aurora-A overexpression might be found in patients with poor prognosis, but adjuvant treatments altered the natural history of the diseaseleading to no survival differences. Thus, a correlation might still exist, but this study was too small to detect it. Moreover, recent reports support the involvement of Aurora-A in early events in tumour carcinogenesis; thus, it is possible that Aurora-A overexpression could be an initiation marker rather than a prognostic marker.


Although a high level of Aurora-A expression appeared to be common in OSSC, 24.7% (21/85) of the aneuploid tumours showed weak or no expression of Aurora-A. This suggests that in a fraction of OSCC, other genes are associated with the development of anueploidy.  Thus, Aurora-A overexpression most probably acts together with other gene products involved in chromosomal segregation to induce aneuploidy in OSCC. One good candidate is inactive p53 tumour suppressor protein. This has been shown to correlate with Aurora-A overexpression in hepatocellular carcinoma when mutated (TP53). This observation provides clinical evidence that Aurora-A overexpression and TP53 cooperate in tumour formation [28].


This study suggests that Aurora-A protein overexpression in OSCC cooperates with other oncogenes to disrupt the signalling cascade that regulates equal segregation of chromosomes leading to pronounced aneuploidy. Future studies in Aurora-A regulators could help explain the mechanisms underlying these processes. The results might enable us to develop new approaches to early diagnosis and treatment.



  1. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JW, et al. (2013) Cancer Incidence and Mortality Patterns in Europe: Estimates for 40 countries in 2012. Eur J Cancer 49: 1374-1403.
  2. SEER Stat Fact Sheets: Oral Cavity and Pharynx Cancer Accessed in 1-8-2016.
  3. Sciubba JJ (2001) Oral cancer: The importance of early diagnosis and treatment. Am J ClinDermatol 2: 239-251.
  4. Bankfalvi A andPiffko J (2000) Prognostic and predictive factors in oral cancer: the role of the invasive tumour front. J Oral Pathol Med 29: 291-298.
  5. Bettendorf O, Piffkò J, Bànkfalvi A (2004) Prognostic and predictive factors in oral squamous cell cancer: important tools for planning individual therapy?. Oral Oncol 40 110-119.
  6. Santosh AB, Jones T, Harvey J (2016) A review on oral cancer biomarkers: Understanding the past and learning from the present. J Cancer Res Ther 12:486-492.
  7. Marumoto T, Zhang D, Saya H (2005) Aurora-A – a guardian of poles. Nat Rev Cancer 5: 42-50.
  8. Hu HM, Chuang CK, Lee MJ, Tseng TC, Tang TK (2000) Genomic organization, expression, and chromosome localization of a third aurora-related kinase gene, Aie1. DNA Cell Biol 19: 679-688.
  9. Dutertre S, Cazales M, Quaranta M, Froment C, Trabut V, et al. (2004) Phosphorylation of CDC25B by Aurora-A at the centrosome contributes to the G2-M transition. J Cell Sci 117: 2523-2531.
  10. Sakakura C, Hagiwara A, Yasuoka R, Fujita Y, Nakanishi M, et al. (2001) Tumour-amplified kinase BTAK is amplified and overexpressed in gastric cancers with possible involvement in aneuploid formation Br J Cancer 84: 824-831.
  11. Tatsuka M, Sato S, Kitajima S, Suto S, Kawai H, et al. (2005) Overexpression of Aurora-A potentiates HRAS-mediated oncogenic transformation and is implicated in oral carcinogenesis. Oncogene 24: 1122-1127.
  12. Bischoff J R, Anderson L, Zhu Y, Mossie K, Ng L, et al. (1998) A homologue of Drosophila aurora kinase is oncogenic and amplified in human colorectal cancers. EMBO J 17: 3052-3065.
  13. Rojanala S, Han H, Munoz RM, Browne W, et al. (2004) The mitotic serine threonine kinase, Aurora-2, is a potential target for drug development in human pancreatic cancer Mol Cancer Ther 3: 451-457.
  14. Tatsuka M, Sato S, Kitajima S, Suto S, Kawai H, et al. (2005) Overexpression of Aurora-A potentiates HRAS-mediated oncogenic transformation and is implicated in oral carcinogenesis. Oncogene 24: 1122-1127.
  15. Zhou H, Kuang J, Zhong L, Kuo WL, Gray JW, et al. (1998) Tumour amplified kinase STK15/BTAK induces centrosome amplification, aneuploidy and transformation. Nat Genet 20: 189-193.
  16. Anand S, Penrhyn-Lowe S, Venkitaraman AR (2003) AURORA-A amplification overrides the mitotic spindle assembly checkpoint, inducing resistance to Taxol. Cancer Cell 3: 51-62.
  17. Giet R, Petretti C, Prigent C (2005) Aurora kinases, aneuploidy and cancer, a coincidence or a real link? Trends Cell Biol 15: 241-250.
  18. Bryne M (1991) Prognostic value of various molecular and cellular features in oral squamous cell carcinomas: a review. J Oral Pathol Med 20: 413-420.
  19. OdellEW, Jani P, Sherriff M, Ahluwalia SM, Hibbert J, et al. (1994) The prognostic value of individual histologic grading parameters in small lingual squamous cell carcinomas. The importance of the pattern of invasion. Cancer 74: 789-794.
  20. Tanaka T, Kimura M, Matsunaga K, Fukada D, Mori H, et al. (1999) Centrosomal kinase AIK1 is overexpressed in invasive ductal carcinoma of the breast. Cancer Res 59: 2041-2044.
  21. Sudbo J, Kildal W, Risberg B, Koppang HS, Danielsen HE, et al. (2001) DNA content as a prognostic marker in patients with oral leukoplakia. N Engl J Med 344: 1270-1278.
  22. Ferlay J, Parkin DM, Bray F, Pisani P (2005) GLOBOCAN 2002 Descriptive Epidemiology Group.
  23. Neville BD, Damm DD, Allen CM, Bouquot JE (2002) Epithelial Pathology in Oral and Maxillofacial Pathology. Second edition, Forest E, Rudolph P eds., W.B Saunders Company, Philadelphia, Pennsylvania. 315-387.
  24. Reis LAG, Harkins D, Krapcho M, Mariotto A, Miller BA, et al. (2006) SEER cancer statistics review, 1975-2003, National Cancer Institute. based on 2005 SEER data submission.
  25. Sherr CJ (1996) Cancer cell cycles. Science 274: 1672-1677.
  26. Royce ME, Xia W, Sahin AA, Katayama H, Johnston DA, et al. (2004) STK15/Aurora-A expression in primary breast tumors is correlated with nuclear grade but not with prognosis. Cancer 100: 12-19.
  27. Shin SO, Lee KH, Kim JH, Baek SH, Park JW, et al. (2000) Alternative splicing in 5′-untranslational region of STK-15 gene, encoding centrosome associated kinase, in breast cancer cell lines. ExpMol Med 32: 193-196.
  28. Jeng YM, Peng SY, Lin CY, Hsu HC (2004) Overexpression and amplification of Aurora-A in hepatocellular carcinoma. Clin Cancer Res 10: 2065-2071.



Figure1: Antibody characterization.

(A) Positive Aurora-A immunofluorescence staining is cytoplasmic and localized at the centresomes (green/yellow). Nuclei of cells not expressing these proteins are counterstainedwith propidium iodide (red).

(B) Western blotshows the expected band size of Aurora-A(46 kDa).


Figure 2: Immunohistochemical staining of fibroepethelial polyps for Aurora-A (positive cells stained brown).  Aurora-A expression was predominantly cytoplasmic and at the basal layers of these specimens. Original magnification x 400


Figure3: The median (solid red line), interquartile range (boxed), and range (enclosed by lines) of Mcm2, Ki67, Geminin and Aurora-A expression are shown in oral fibroepethelial polyps (outlying cases are shown by isolated points designated by the case number). The median of Aurora-A wassignificantly less than the rest of the markers (paired t test, P< 0.001 each).

Figure 4:  Immunohistochemical staining of OSCC for Aurora-A (positive cells stain brown).  Aurora-A staining is diffusely distributed in the cytoplasm, at the peripheral portions of the cancer nests (A) and sometimes found scattered throughout some tumours (B). Original magnification was x 100 and x 400 (the enlarged images).



Figure 5:Median, interquartile range, and range of Kki67 according to intensity of Aurora-A expression is shown in OSCC. The median of Ki67 is significantly higher in cases with high Aurora-A intensity (Jonckheere-Terpstra test, P= 0.02)



Figure 6: Median, interquartile range, and range of Mcm2, Kki67, Geminin, and Aurora-A expression are shown in OSCC. The median and interquartile range of Aurora-A are significantly less than those of the rest of the markers (paired t test, P < 0.001 each).


Figure 7: DNA histograms generated by measurements of the nuclear DNA content of Fuelgen-stained OSCC cells (grey) of diploid and aneuploid tumours.  The histograms were scaled on the basis of DNA content of a diploid standard (lymphocytes, red). Histogram interpretation for DNA content was based on Sudbo’s classification for aneuploidy.




Figure 8: Aurora-A ectopic overexpression and its relation with aneuploidy in OSCC. Among the 85 aneuploid OSCC, 64 cases had moderate to high Aurora-A expression; while only 6 cases from the 19 diploid OSCC had moderate to high expression (Chi-square test, P= 0.01). Thus, it is possible that Aurora-A ectopic overexpression induces aneuploidy in OSCC.



Clinical Features Total Clinical Features Total
All cases 125 Stage
Gender I 12
 Male 80 II 9
Female 45 III 3
Unknown 0 IV 58
Age Unknown 43
≤ 60 years 62 Tumour size
≥ 60 years 62 T1+T2 36
      Unknown 1 T3+T4 50
Primary tumour site Unknown 39
      Tongue 54 Lymph node metastasis
  Alveolar mucosa 39 No 64
 Floor of the mouth 21 Yes 41
Buccal mucosa 6 Unknown 20
Others 5 Three-years survival  
Unknown 0 Survived 31
Phenotype Did not survive 48
Conventional 113  Unknown 46
Verrucous/papillary 9  Five-years survival
Adenosquamous 2 Survived 14
Basaloid 1 Did not survive 51
Unknown 0 Unknown 60
Histopathologic differentiation DNA content
Well 26  Diploid 19
Moderate 71 Aneuploid 85
Poor 28 Tetraploid 0
Unknown 0  Unknown 21
Invasive front AA intensity expression
Pushing 9  No expression 14
Bands 53 Weak expression 25
Cords 38 Medium expression 35
Diffuse 25  High expression 51
Unknown 0 Unknown 0


Table 1: Clinicopathological and outcome parameters in OSCC.


Clinical Features Total Aurora-A intensity Aurora-A LI
+/- ++/+ P value LI P value
All cases 125 14 25 86
Male 80 5 17 58 0.10 6.025* 0.302
Female 45 9 8 28 5.968*
60 years 62 9 8 45 0.10 6.111* 0.940
 ≥ 60 years 62 5 17 40   5.918*  
Conventional 113 14 22 77 1 6.048* 0.644
Verrucous/papillary 9 0 3 6   5.880*  
Basaloid 1 0 0 1   9.200*  
Adenosquamous 2 0 0 2   2.485*  
Histopathologic differentiation              
Well 26 3 7 16 1 4.554* 0.315
Moderate 71 9 12 50   6.462*  
 Poor 28 2 6 20   6.192*  
Invasive front              
Pushing 9 1 3 5 1 5.547* 0.594
 Bands 53 8 7 38   5.653*  
Cords 38 4 9 25   6.124*  
Diffuse 25 1 6 18   6.731*  
I 12 5 0 7 0.10 5.41* 0.845§
II 9 0 2 7   6.700*  
III 3 0 1 2   4.943*  
IV 58 6 11 41   5.909*  
Tumour size              
T1+T2 36 5 4 27 1 6.447* 0.295
T3+T4 50 6 11 33   5.541*  
 Lymph node metastasis              
 No 64 11 9 44 0.20 5.648* 0.688
Yes 41 3          
Abbreviations:*Mean; expressed as percentage, Chi-square test, Mann-Whitney U test, §Jockheere-Terpstra test, Kruskal-Wallis H test, and Spearman’s Correlation coefficient.


Table 2: Relationship between Aurora-A expression and Histopathologic and clinical parameters in OSCC.


Clinical Features Total Aurora-A intensity Aurora-A LI
+/- ++/+ P value LI P value
 Five-years survival              
Survived 14 1 2 11 0.17° 4.878* 0.113°
Did not survive 51 5 15 31   5.935*  
DNA content              
Diploid 19 4 9 6 0.01 5.348* 0.752
Aneuploid 85 8 13 64   6.097*  
Abbreviations:*Mean; expressed as percentage, Chi-square test, Mann-Whitney U test and °Log-Rank survival test.


Table 3: Relationship between Aurora-A expression and survival rates and DNA content in OSCC.

Suggested Citation


Citation: Al-Hazmi NA, Alhazzazi TY, Albahiti Haji M, Williams GH, Stoeber K et al. (2017) Aurora-A Overexpression Plays a Role in Developing Aneuploidy in a Subset of Oral Squamous Cell Carcinoma. JOncol Res Ther2017: J112.


Leave a Reply