Datasheet3_Construction and analysis of a survival-associated competing endogenous RNA network in breast cancer.pdf

BackgroundRecently, increasing studies have shown that non-coding RNAs are closely associated with the progression and metastasis of cancer by participating in competing endogenous RNA (ceRNA) networks. However, the role of survival-associated ceRNAs in breast cancer (BC) remains unknown.
MethodsThe Gene Expression Omnibus database and The Cancer Genome Atlas BRCA_dataset were used to identify differentially expressed RNAs. Furthermore, circRNA-miRNA interactions were predicted based on CircInteractome, while miRNA-mRNA interactions were predicted based on TargetScan, miRDB, and miRTarBase. The survival-associated ceRNA networks were constructed based on the predicted circRNA-miRNA and miRNA-mRNA pairs. Finally, the mechanism of miRNA-mRNA pairs was determined. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of survival-related mRNAs were performed using the hypergeometric distribution formula in R software.The prognosis of hub genes was confirmed using gene set enrichment analysis.
ResultsBased on the DE-circRNAs of the top 10 initial candidates, 162 DE-miRNAsand 34 DE-miRNAs associated with significant overall survival were obtained. The miRNA target genes were then identified using online tools and verified using the Cancer Genome Atlas (TCGA) database. Overall, 46 survival-associated DE-mRNAs were obtained. The results of GO and KEGG pathway enrichment analyses implied that up-regulated survival-related DE-mRNAs were mostly enriched in the “regulation of cell cycle” and “chromatin” pathways, while down-regulated survival-related DE-mRNAs were mostly enriched in “negative regulation of neurotrophin TRK receptor signaling” and “interleukin-6 receptor complex” pathways. Finally, the survival-associated circRNA-miRNA-mRNA ceRNA network was constructed using 34 miRNAs, 46 mRNAs, and 10 circRNAs. Based on the PPI network, two ceRNA axes were identified. These ceRNA axescould be considered biomarkers for BC.GSEA results revealed that the hub genes were correlated with “VANTVEER_BREAST_CANCER_POOR_PROGNOSIS”, and the hub genes were verified using BC patients' tissues.
ConclusionsIn this study, we constructed a circRNA-mediated ceRNA network related to BC. This network provides new insight into discovering potential biomarkers for diagnosing and treating BC.

背景 近年来，越来越多的研究表明，非编码RNA（non-coding RNAs）通过参与竞争性内源RNA（competing endogenous RNA, ceRNA）网络，与癌症的进展及转移密切相关。然而，与生存相关的ceRNA在乳腺癌（breast cancer, BC）中的作用仍未明确。
方法 本研究采用基因表达综合数据库（Gene Expression Omnibus, GEO）与癌症基因组图谱乳腺癌数据集（The Cancer Genome Atlas BRCA_dataset）筛选差异表达RNA。进一步基于CircInteractome数据库预测环状RNA-微小RNA（circRNA-miRNA）互作关系，基于TargetScan、miRDB及miRTarBase数据库预测微小RNA-信使RNA（miRNA-mRNA）互作关系。基于上述预测得到的circRNA-miRNA及miRNA-mRNA对构建与生存相关的ceRNA网络。最终明确miRNA-mRNA对的调控机制。采用R软件中的超几何分布公式，对生存相关信使RNA进行基因本体（Gene Ontology, GO）与京都基因与基因组百科全书（Kyoto Encyclopedia of Genes and Genomes, KEGG）富集分析。利用基因集富集分析（gene set enrichment analysis, GSEA）验证核心基因的预后价值。
结果 基于排名前10的初始候选差异表达环状RNA（DE-circRNAs），共筛选得到162个差异表达微小RNA（DE-miRNAs）及34个与总生存显著相关的差异表达微小RNA。随后通过在线工具预测微小RNA靶基因，并通过癌症基因组图谱（The Cancer Genome Atlas, TCGA）数据库进行验证。最终共获得46个与生存相关的差异表达信使RNA（DE-mRNAs）。GO与KEGG通路富集分析结果显示，上调的生存相关差异表达信使RNA主要富集于"细胞周期调控"与"染色质"通路，而下调的生存相关差异表达信使RNA则主要富集于"神经营养因子TRK受体信号负调控"与"白细胞介素-6受体复合物"通路。最终，基于34个微小RNA、46个信使RNA及10个环状RNA，构建了与生存相关的circRNA-miRNA-mRNA ceRNA网络。通过蛋白质相互作用（protein-protein interaction, PPI）网络筛选得到2条ceRNA轴，可作为乳腺癌的潜在生物标志物。GSEA分析结果显示，核心基因与"VANTVEER_BREAST_CANCER_POOR_PROGNOSIS"基因集显著相关，并通过乳腺癌患者组织样本验证了核心基因。
结论 本研究构建了一种介导乳腺癌发生发展的环状RNA相关ceRNA网络，该网络为发掘乳腺癌诊断与治疗的潜在生物标志物提供了新的研究视角。