Prognostic relevance of correlated co-expression of coding and noncoding RNAs in cervical cancers

Abstract

AbstractHuman papillomavirus (HPV) drives cervical cancer (CaCx) pathogenesis and viral oncoproteins jeopardize global gene expression in such cancers. We aimed to identify differentially expressed coding (DEcGs) and long noncoding (sense intronic and Natural Antisense Transcripts) RNA genes (DElncGs) in HPV16-positive CaCx patients (N=44) compared to HPV-negative normal individuals (N=34). Thereby, employing strand-specific RNA-seq, we determined the relationships between DEcGs and DElncGs and their clinical implications. Gene set enrichment and protein-protein interaction analyses of DEcGs revealed enrichment of processes crucial for abortive virus life cycle and cancer progression. The DEcGs formed 16 gene clusters, portraying cancer-related functions. We recorded significantly correlated co-expression of 79 DElncGs with DEcGs at proximal genomic loci, and 24 such pairs portrayed significantly altered correlation coefficients among patients, compared to normal individuals. Of these, 6 DEcGs belonged to 5 gene clusters, one of which was survival-associated. Out of the 24 correlated DEcG: DElncG pairs, 3 pairs were identified, where expression of both members was significantly associated with patient overall survival. Besides being prognostically relevant, disruption of the significant correlative relationships of such gene pairs in CaCx bears immense potential for patient-targeted therapy.Graphical Abstract:A schematic representation of the analyses undertaken to draw insights on the biological relevance of joint analysis of co-expressed and significantly correlated DElncGs and DEcGs considering the ncNATs and sense intronic DElncGs, in HPV16-positive CaCx patients compared with HPV-negative normal individuals.