Comprehensive analysis of exosomal circRNA-miRNA-mRNA network in breast cancer.

Abstract

Exosomal circRNAs played critical roles in tumor development and progression and might be novel biomarkers in the diagnosis and treatment of various cancers. However, the biological functions and clinical implications of exosomal circRNAs in breast cancer are unclear. Expression profiles of exosomal circRNAs in exoRBase 2.0 database were used to identify differentially expressed exosomal circRNAs in breast cancer. The LASSO and SVM-RFE algorithm followed by multivariate logistic regression analysis were performed to construct the diagnostic model. The target genes of circRNAs were selected by combing differential expression analysis and CSCD, TargetScan and ENCORI database. Univariate and multivariate survival analysis were conducted to construct a survival-associated exosomal circRNA-miRNA-mRNA network. GSVA and CIBERSORT algorithm were used to evaluate the cancer hallmarks and immune cells in breast cancer and spearman correlation analysis was used to investigate their correlations with the circRNA-miRNA-mRNA network. A total of 347 upregulated and 3 downregulated exosomal circRNAs were identified in breast cancer patients. The diagnostic model basing on 14 exosomal circRNAs showed a high area under the curve (AUC) value in both the training (AUC = 0.98) and validation dataset (AUC = 0.94). A total of 70 miRNAs and 1147 mRNAs were selected as the downstream targets of circRNAs and were revealed to participate into tumor-associated pathways including PI3K-AKT, MAPK, RAS and RAP1 pathways, calcium signaling pathways and transcriptional misregulations. The constructed survival-associated exosomal circRNA-miRNA-mRNA network contained 9 exosomal circRNAs, 12 miRNAs and 10 mRNAs and showed complicated correlations and interactions within networks. Cancer hallmark pathways including TGF-β, KRAS and MYC signaling pathways, tumor angiogenesis, epithelial-mesenchymal transition, DNA repair and G2M checkpoint and immune cells including CD4+ and CD8+ T cells, dendritic cells, mast cells, macrophage cells, memory B cells and natural killer cells were closely correlated with the circRNA-miRNA-mRNA network. Our study was the first to systematically analyze the exosomal circRNAs in breast cancer. We established an exosomal circRNA diagnostic model and constructed a survival-associated exosomal circRNA-miRNA-mRNA network. Our results revealed the complicated functions and potential mechanisms of exosomal circRNA-miRNA-mRNA network in breast cancer which need to be further validated in the future.