An assessment for unique biomarker profiles in oral squamous cell carcinoma

© 2018 Dr. Simone Helene BelobrovPurpose: Biomarkers can be used to predict the incidence or outcome of disease. Currently there are no established biomarkers for oral squamous cell carcinoma (OSCC). Identification of these markers and molecular profiles of these cancers could not only optimise treatment, but also have the potential to improve clinical outcomes with early detection of primary OSCC and relapse. If found, these biomarkers could be used to establish screening techniques for precursor changes of the mucosa that precede the development of invasive tumours and so an accurate prediction of malignant transformation risk could be developed. For this to occur the molecular mechanisms of the carcinogenetic process of OSCC need to be elucidated, proven to be more efficient than currently available, ultimately resulting in improved prognosis. Tumour biomarkers can be physical, chemical or biological molecules that include proteins, DNA and RNA. These can be expressed singularly or in combination representing a specific molecular profile. In an ideal setting they would be easy and reliable to analyse, cost-effective, and add additional information to current clinical diagnostic techniques. If clinicopathological parameters are only taken into account, OSCC outcome prediction is difficult. Hence, biomarkers have the potential to aid in the identification of tumour aggressiveness and prognosis prediction. At present there are no clinically approved biomarkers for OSCC tumour classification and diagnosis. Even though substantial progress has been made in recent years to understand the molecular basis of OSCC, much more needs to be understood. Hypothesis and Aim: The overall hypothesis was that the increase expression of cancer associated protein biomarkers for OSCC correlate to clinicopathological features of oral cancer patients and their prognosis. The overall aim was to investigate the relationship between cancer associated protein biomarkers in human OSCC tissues and clinical outcomes. Materials and Methods: Formalin fixed paraffin embedded (FFPE) tumour blocks from 129 consecutive patients (71 males, 58 females) treated for OSCC at the Royal Melbourne Hospital with well documented clinical and histopathological features were chosen as the patient cohort for analysis. The FFPE blocks were sectioned and a retrospective immunohistochemical (IHC) staining was performed to identify the expression of a chosen panel of protein markers which included p53, p16, Cyclin D1, Epidermal Growth Factor Receptor (EGFR) and Survivin. Subsequently forty-six tumour blocks from this cohort underwent Laser capture microdissection to isolate invasive tumour tissue. DNA was extracted and tested for high-risk human papillomavirus (HPV) types using a PCR-ELISA method based on the L1 SPF10 consensus primers, and a real-time PCR method targeting HPV-16 and HPV-18 E6 region. Genotyping of HPV-positive cases were performed using a reverse line blot hybridization assay (Inno-LiPA). RNAScope (a chromogenic RNA in situ hybridization assay) was utilized to detect E6/E7 mRNA of known high-risk HPV types for detection of transcriptionally active virus. Four oral malignant keratinocyte cell lines (BICR16, BICR56, H357 and H400) were cultured and employed as an in vitro model for OSCC carcinogenesis. Firstly, DNA was extracted from all cell lines and underwent Sanger sequencing to determine the presence or absence of CDKN2A mutations. Protein expression profiling was performed to compare the difference of expression between a mutated CDNKN2A cell line (H357) and wild-type (non-mutated) CDKN2A cell line (H400). Cell proliferation (growth curve) and migration (Transwell) assays were employed to determine if the two cell lines differed in phenotype and if this could be attributed to p16 expression. Protein was extracted from both cell lines and analysed using the Explorer Antibody Array (Full Moon BioSystems, CA, USA) which is an ELISA-based antibody array containing a panel of 656 highly specified and well-characterized antibodies involved in more than 10 signalling/biochemical pathways. Proteins of interest were subjected to further analysis by Western Blotting to validate the antibody array findings. To induce expression of p16 the cell lines were subjected to treatment with 200μM, 400μm and 600μM H202 at 2h, 4h and 24h time points. Protein was extracted and subjected to Western Blot analysis to determine p16 expression. The relevance of epidermal growth factor receptor expression in oral cancer cells was further analysed by treating cells with 10ug/mL and 20ug/mL (−)-epigallocatechin gallate (EGCG), a major polyphenol in green tea. EGCG has been previously reported to block EGFR expression in wide variety of various cancer cell types. Proteins were extracted and western blotting performed to determine the effect on EGFR expression. Results: Biomarker overexpression was observed in 72 (56%) cases for p53, 23 (18%) for p16, 45 (35%) for Cyclin D1, 72 (56%) for EGFR and 3 (2%) for Survivin. Multiple logistic regression analysis revealed that tongue tumours (p=0.012) and late stage cancers (p=0.031) were more likely to have cyclin D1 overexpression. Younger patients significantly more often had cyclin D1 overexpression (p=0.008) and non-smoking non-drinking (NSND) patients had more p16 expression (p=0.043). In contrast, smokers were more likely to have EGFR overexpression (p=0.033). Concurrent overexpression of p53 and cyclin D1 was observed (p = 0.030). HPV DNA was found in only 3 OSCC cases, which were also p16 IHC positive. Two cases were genotyped as HPV-16 and one as HPV-33. Only one of the HPV-16 cases was confirmed as harbouring transcriptionally active virus. Oral malignant keratinocyte cell lines were used as a model to compare muted CDKN2A and non-muted CDKN2A oral cancers and assess phenotypic differences. The two cell lines significantly differed in proliferation and migration rates. Explorer antibody array results confirmed significant differences in the expression of various proteins between two cell lines. However, proteins in the p16 network did not differ as confirmed by western blotting. p16 expression was not found in normal growth conditions for the non-mutated cell line. H2O2 did not induce p16 expression. Cell proliferation of both cell lines was significantly reduced at 48 hours when treated with 20ug/mL EGCG. However, after 72 hours of treatment the effect of EGCG on cell proliferation ceased. Treatment of both cell lines with 10ug/mL and 20ug/mL of EGCG resulted in significant reduction in cell migration. EGCG did not block EGFR expression in these cell lines. Conclusion: Smoking, site and stage of OSCC can influence biomarker expression, with p16 overexpression specifically observed in NSNDs, indicating fundamental differences in the mechanisms of oral carcinogenesis among different patient cohorts. However, HPV was not found to be significantly associated with OSCC and therefore p16 cannot be used as a surrogate marker for HPV in OSCC cases. A non-viral mechanism is most likely responsible for p16 overexpression in OSCC. Oral squamous cell carcinoma cell lines H357 and H400 cannot be used as a model to represent the patient non-HPV OSCC samples that are p16 positive, and are not suitable as a model for p16 protein dysregulation in OSCC. EGCG transiently inhibits both cell proliferation and migration of oral cavity cancer cells. This effect is independent of EGFR