Analysis And Construction Of CeRNA Network Based On TCGA Transcriptome Data To Identify Potential Prognostic Indicators Of Triple-negative Breast Cancer

Objective and BackgroundBreast cancer is a heterogeneous disease with different histological abnormalities,cytogenetic abnormalities,and differences in response and prognosis to treatments.It accounts for 16% of all female cancers.Triple-negative breast cancer(TNBC)refers to a particular subtype of breast malignant tumor that does not express the genes for estrogen receptor(ER),progesterone receptor(PR),or human epidermal growth factor receptor 2(Her2/neu).Generally,this subtype has stronger invasion,higher recurrence and metastasis rates,shorter survival time than those of non-TNBC.In addition,existing targeting and endocrine therapy strategies are only applicable to the patients with HR or HER2-positive breast cancer.For triple negative breast cancer,conventional chemotherapy remains the primary treatment.Therefore,a deeper understanding of the molecular mechanisms of triple negative breast cancer will help develop new strategies for cancer treatment.Tumor biomarkers generally refer to characteristic indices for onset and progression,which can be objectively measured and evaluated to determine the tumor stage.Examining a disease-specific biomarker can help clinical identification,early diagnosis and prevention,as well as monitoring during treatment.In recent years,long non-coding RNA(lncRNA),microRNA(miRNA)and message RNA(mRNA)have been reported to play crucial roles in many biological processes,thus having become key biomarkers for the diagnosis and treatment of tumors.LncRNA can compete with the target mRNA of miRNA to reduce free miRNA content,and regulate such mRNA.Until now,the interaction mechanism for the regulatory network of lncRNA-miRNA-mRNA in TNBC remains unclear.MethodsWe collected TNBC RNA expression profile data and relevant clinical features from The Cancer Genome Atlas(TCGA).A cluster analysis was explored to show different mRNA,miRNA,lncRNA expression patterns.Gene ontology(GO),Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway enrichment analyses were performed to analyze the functions of the intersecting mRNAs.miRanda and Targetscan bioinformatics algorithms were used to predict potential relationships among RNAs.Subsequently,we constructed a competitive endogenous RNA(ceRNA)network of TNBC by targeting interrelations with significantly aberrant expression data between miRNA and mRNA or lncRNA.Survival analysis used to determine the RNA expression levels and survival times.Finally,real-time quantitative PCR further determined the expression level of the core gene in the triple negative breast cancer cell line.Results(1)Screening of DEGs.Using |logFC|>3,FDR<0.001 as a screening criterion for differential expression of mRNAs,lncRNAs,and miRNAs.There were 1238(including 818 up,420 down)and 220(including 170 up and 50 down)differentially expressed mRNAs and lncRNAs between early(StageⅠ&Ⅱ)and paracancer groups respectively,and 937(including 735 up,202 down)and 110(including 92 up and 18 down)differentially expressed mRNAs and lncRNAs between advanced(Stage Ⅲ&Ⅳ)and paracancer groups respectively.In addition,there were 51(including 39 up and 12 down)differentially expressed miRNAs between early(StageⅠ&Ⅱ)and control groups,and 42(including 34 up,8 down)ones between advanced(Stage Ⅲ&Ⅳ)and control groups.Compared with the paracancer group,a total of 686 common differentially expressed mRNAs were found in early and advanced groups,of which 519 were up-regulated and 167 were down-regulated.There were 50 differentially expressed lncRNAs,of which 38 were up-regulated and 12 were down-regulated.Additionally,there were 26 differentially expressed miRNAs,of which 19 were up-regulated and 7 were down-regulated.(2)GO and Pathway Enrichment Analyses of DEGs.We employed the DAVID database to conduct GO and pathway analyses for the above differentially expressed mRNAs.The up-regulated genes mainly involved the signaling pathways of cell cycle,p53 and ECM-receptor interaction,and the down-regulated genes involved the cAMP and AMPK signaling pathways.(3)Construction of ceRNA regulatory network.In this study,the correlation coefficient between RNA and miRNA is r<-0.3 and p<0.05,as a screening criterion that may form a regulatory relationship of ceRNA.Subsequently,the relationship between DE-mRNA and DE-miRNA conforming to the above criteria was compared in the TragetScan database,and the relationship between DE-lncRNA and DE-miRNA was compared in the miRanda database.Based on the analyses above,262 miRNA-mRNA pairs and 14 lncRNA-miRNA pairs were obtained.Subsequently,we combined the files of these pairs and visualized them using Cytoscape,yielding a ceRNA topology network.(4)Pathway and Go analyses of differentially expressed mRNAs in ceRNA network.The 20 most enriched function sets are filtered through the Metascape database.GO analysis showed that regulation of growth,regulation of secretion and mitotic cell cycle phase transition were significantly enriched biological processes,and the signaling pathways of PPAR,protein digestion and absorption,and regulation of lipolysis in adipocytes were involved.(5)Prognostic Analysis of ceRNA Network Nodes.We subjected 14 lncRNAs,11 miRNAs and 262 mRNAs in the network to K-M survival(Survminer Software package)analysis respectively.Finally,we selected five molecules that were significantly correlated with prognosis,TERT,TRIML2,PHBP4,hsa-mir-1-3p and hsa-mir-133a-3p.(6)RT-qPCR showed that the expression levels of TRIML2 gene in TNBC cell lines were significantly higher than that in normal mammary cell line.ConclusionTaken together,our findings represent new knowledge for a better understanding the ceRNA network in TNBC biology and pave the way to improved diagnosis and prognosis of TNBC.