Label-free tissue proteomics can classify oral squamous cell carcinoma from healthy tissue in a stage-specific manner

Abstract

Objectives: No prognostic or predictive biomarkers for oral squamous cell carcinoma (OSCC) exist. We aimed to discover novel proteins, altered in OSCC, to be further investigated as potential biomarkers, and to improve understanding about pathways involved in OSCC. Materials and Methods: Proteomic signatures of seven paired healthy and OSCC tissue samples were identified using ultra-definition quantitative mass spectrometry, then analysed and compared using Anova, principal component analysis, hierarchical clustering and OPLS-DA modelling. A selection of significant proteins that were also altered in the serum from a previous study (PMID: 28632724) were validated immunohistochemically on an independent cohort (n = 66) to confirm immunopositivity and location within tumour tissue. Ingenuity Pathways Analysis was employed to identify altered pathways. Results: Of 829 proteins quantified, 257 were significant and 72 were able to classify healthy vs OSCC using OPLS-DA modelling. We identified 19 proteins not previously known to be upregulated in OSCC, including prosaposin and alpha-taxilin. KIAA1217 and NDRG1 were upregulated in stage IVa compared with stage I tumours. Altered pathways included calcium signalling, cellular movement, haematological system development and function, and immune cell trafficking, and involved NF-kappa B and MAPK networks. Conclusions: We found a set of proteins reliably separating OSCC tumour from healthy tissue, and multiple proteins differing between stage I and stage IVa OSCC. These potential biomarkers can be studied and validated in larger cohorts.