Osteosarcoma and miRNAs: combining in silico miRNA analysis and proteomic profiling in search of potential diagnostic panels

Abstract

Osteosarcoma (OS) is the most frequent bone tumor in pediatrics and presents two critical clinical challenges, metastasis and chemoresistance. Better diagnostic and prognostic tools for OS disease progression are in need. Here we propose the use of micro-RNAs (miRNAs) as alternative diagnostic biomarkers for OS. MiRNAs are small and stable non-coding RNAs that can be obtained from liquid biopsies of different body fluids such as plasma, that in the last years have been proposed as diagnostic and prognostic biomarkers. The aim of this work was to assess an OS miRNAs database and contrast it with our own molecular and functional profiling in an OS model with metastatic behavior, in order to propose possible miRNAs as biomarker candidates. We analyzed circulating miRNAs present in the plasma of 15 healthy donors and 20 OS patients (10 with localized OS and 10 with metastatic OS) using the miRNAs dataset GSE65071. Our analysis revealed that miR-34a-5p, -200a-3p, -582-5p, -624-5p and let-7a-3p were upregulated in OS patients plasma as compared to healthy donors (fold change: 0.43; 0.78; 0.78; 0.95; 0.5 respectively; p < 0.0001), while miR-27a-3p and -221-3p were found downregulated in the plasma of OS patients as compared to healthy donors (fold change: -0.73; -2.64 respectively; p < 0.0001). There was no difference in expression between localized and metastatic OS for these miRNAs. Bioinformatics analysis of the target genes for these miRNAs revealed that they are implicated in the regulation of different cancer-related biological pathways like ECM- receptor interaction, cell cycle control and EMT, in coincidence with our proteomic approach on metastatic and non-metastatic OS cells. These results strengthen the in-silico search and constitute a proof of concept on the use of this cross-omic approach as a tool for the identification of potential miRNAs as liquid biopsy biomarkers for diseases characterized by scarce extensive population-based data.