Analisi integrativa dell'espressione di circRNA in tessuto canceroso e tessuto adiacente del colon di pazienti con cancro del colon-retto

The widespread use of RNA sequencing (RNA-Seq) technologies supported the characterization of a peculiar class of transcripts synthesized by an alternative splicing event defined as back-splicing. This process involves the covalent bond between splicing donor and splicing acceptor resulting in RNAs having a circular structure, namely circular RNAs (circRNAs). Despite circRNAs were thought to be an erroneous splicing machinery bioproduct, an increasing number of functionally relevant circRNAs is reported. A dysregulated activity of circRNA in the modulation of gene expression in disease conditions, especially in cancer progression, was extensively reported. However, the early mechanisms driving the circRNA alterations in these contexts are less understood. Colorectal cancer (CRC) is one of the deadliest cancer diseases, especially because tumors are diagnosed at a late disease stage. A deep analysis of circRNAs profiles in a large cohort of CRC tissues can help to characterize the molecular driver of the circRNA dysregulation and to discover a circRNA-based signature that might be used for the early disease diagnosis. In this Master Thesis project, I performed an integrative analysis of the circRNAs expression profiles in 87 CRC and 27 adenoma tissues matched with healthy adjacent colonic mucosa tissue from a cohort of Italian subjects. The circRNA prediction tool CIRI2 was used for circRNA prediction and quantification of their expression starting from total RNA-Seq data. Subsequently, the circRNA molecular features and expression pattern were explored with respect to patient clinical parameters, to their host gene levels, and to RNA binding proteins (RBP) expression. By paired differential expression analysis, 47 dysregulated circRNAs were observed, among which 43 were down-regulated in the CRC tissue. This down-regulation trend was observed at genome-wide level and it was confirmed among different disease stages and tumor molecular subtypes. In addition, the circRNA down-regulation was observed in advanced adenomas and was more pronounced in early than late CRC stages. The analysis of the circRNA host genes showed a significant relation between host gene expression and circRNA dysregulation with the down-regulated circRNAs associated with the highest correlations. The integration with the patient clinical data showed differential expression in respect to specific covariates, particularly age, tumor localization and grade. Analysis of RBPs involved in circRNA biogenesis showed a significant association between the number of detected circRNAs and the RBM20 and QK1 levels, two genes significantly down-regulated in CRC tissues. These data support that multiple molecular alterations may drive a widespread circRNA down-regulation in CRC that might represent a novel early molecular perturbation occurring at the first steps of colorectal carcinogenesis.