Decoding dynamic miRNA:ceRNA interactions unveils therapeutic insights and targets across predominant cancer landscapes
Selcen Ari Yuka,Alper Yilmaz
DOI: https://doi.org/10.1186/s13040-024-00362-4
2024-04-19
BioData Mining
Abstract:Competing endogenous RNAs play key roles in cellular molecular mechanisms through cross-talk in post-transcriptional interactions. Studies on ceRNA cross-talk, which is particularly dependent on the abundance of free transcripts, generally involve large- and small-scale studies involving the integration of transcriptomic data from tissues and correlation analyses. This abundance-dependent nature of ceRNA interactions suggests that tissue- and condition-specific ceRNA dynamics may fluctuate. However, there are no comprehensive studies investigating the ceRNA interactions in normal tissue, ceRNAs that are lost and/or appear in cancerous tissues or their interactions. In this study, we comprehensively analyzed the tumor-specific ceRNA fluctuations observed in the three highest-incidence cancers, LUAD, PRAD, and BRCA, compared to healthy lung, prostate, and breast tissues, respectively. Our observations pertaining to tumor-specific competing endogenous RNA (ceRNA) interactions revealed that, in the cases of lung adenocarcinoma (LUAD), prostate adenocarcinoma (PRAD), and breast invasive carcinoma (BRCA), 3,204, 1,233, and 406 ceRNAs, respectively, engage in post-transcriptional intercommunication within tumor tissues, in contrast to their absence in corresponding healthy samples. We also found that 90 ceRNAs are shared by the three cancer types and that these ceRNAs participate in ceRNA interactions in tumor tissues compared to those in normal tissues. Among the 90 ceRNAs that directly interact with miRNAs, we uncovered a core network of 165 miRNAs and 63 ceRNAs that should be considered in RNA-targeted and RNA-mediated approaches in future studies and could be used in these three aggressive cancer types. More specifically, in this core interaction network, ceRNAs such as GALNT7, KLF9, and DAB2 and miRNAs like miR-106a/b-5p, miR-20a-5p, and miR-519d-3p may have potential as common targets in the three critical cancers. In contrast to conventional methods that construct ceRNA networks using differentially expressed genes compared to normal tissues, our proposed approach identifies ceRNA players by considering their context within the ceRNA:miRNA interactions. Our results have the potential to reveal distinct and common ceRNA interactions in cancer types and to pinpoint critical RNAs, thereby paving the way for RNA-based strategies in the battle against cancer.
mathematical & computational biology