Abstract
BACKGROUND: Long noncoding RNAs (lncRNAs) can act as competitive endogenous RNAs (ceRNAs) to compete with mRNAs for binding miroRNAs (miRNAs). The dysregulated triplets, composed by mRNAs, lncRNAs, and miRNAs, contributed to the development and progression of diseases, such as cancer. However, the roles played by triplet biomarkers are not fully understand in glioblastoma multiforme (GBM) patient survival. OBJECTIVES: Here, we constructed a differential triplet interaction network (TriNet) between GBM and normal tissues and identified GBM survival related triplets. METHODS: Four significantly dysregulated modules, enriched differentially expressed molecules, were identified by integrating affinity propagation method and hypergeometric method. Furthermore, knockdown of TP73-AS1 was implemented by siRNA and the expression of RFX1 was examined in U87 cells by qRT-PCR. The apoptosis of U87 cells was investigated using MTT assay and Acridine orange/Ethidium bromide (AO/EB) assay. RESULTS: We randomly split GBM samples into training and testing sets, and found that these four modules can robustly and significantly distinguish low- and high-survival patients in both two sets. By manually curated literatures for triplets mediated by core interactions, we found that members involved tumor invasion, proliferation, and migration. The dysregulated triplets may cause the poor survival of GBM patients. We finally experimentally verified that knockdown of TP73-AS1, an lncRNA of one triplet, could not only reduce the expression of RFX1, an mRNA of this triplet, but also induce apoptosis in U87 cells. CONCLUSIONS: These results can provide further insights to understand the functions of triplet biomarkers that associated with GBM prognosis.