LncRNA MIR155HG functions as a ceRNA for inhibition of lung adenocarcinoma growth and prediction of prognosis

Introduction: Long non-coding RNAs (lncRNAs) functioning as competing endogenous RNAs (ceRNAs) play critical roles in tumour progression. However, prognosis-related ceRNA networks in lung adenocarcinoma (LUAD) have not been well characterised. Material and methods: LUAD datasets were downloaded from the TCGA database, and the patients were divided into metastasis and non-metastasis groups. The differential expression of lncRNAs (DELs), miRNAs (DEMs), and mRNAs (DEGs) was analysed using the Limma package. Next, interactions between miRNA, lncRNA, and mRNA were predicted by miRcode, miRTar-Base, miRDB, and TargetScan. The ceRNA network was constructed based on these interactions using Cytoscape software. DEG enrichment analysis was performed by GO and KEGG. After the prognosis analysis, we further screened molecules and constructed the prognosis-related ceRNA network. Moreover, the interactions between lncRNA, miRNA, and mRNA were validated by biological experiments. Results: 854 DELs, 150 DEMs, and 2211 DEGs between metastasis and non-metastasis LUAD patients were identified. Functional enrichment analysis suggested that DEGs were closely related to key biological processes involved in LUAD progression. The prognosis-related ceRNA network included 1 miRNA, 2 lncRNAs, and 4 mRNAs. In this network, MIR155HG and ADAMTS9-AS2 can function as ceRNAs of miR-212 to regulate EPM2AIP1, LAX1, PRICKLE2, and CD226. Moreover, our study confirmed that MIR155HG inhibited the proliferation, migration, and invasion of LUAD cells by sponging miR-212-3p to regulate CD226. Conclusions: This ceRNA network contributes to understanding the pathogenesis of LUAD. Furthermore, the molecules in the network are valuable predictive factors for LUAD prognosis as well as potential therapeutic biomarkers.