Bioinformatics analysis of miRNA expression profile between primary and recurrent glioblastoma.

OBJECTIVE: Glioblastoma (GBM) is the most malignant brain tumor with rapid relapse. The goal of this study is to identify microRNAs (miRNAs) involved in recurrent GBM. MATERIALS AND METHODS: miRNA transcription profile data (GSE32466) were downloaded, including 12 primary GBM samples and 12 recurrent GBM samples. Then, limma package was utilized to identify differentially expressed miRNAs (DEMs) with the criteria of false discovery rate < 0.05 and vertical bar log2 fold change vertical bar >= 1. Thereafter, miTarget and TargetScan databases were used to predict the potential target genes of DEMs. Regulatory co-expression network was constructed based on co-expressed genes and potential miRNA-gene pairs, and then, pathway analysis was conducted. Furthermore, database miRWalk was used to screen out known GBM-associated miRNAs from the identified DEMs. RESULTS: A total of 71 DEMs were identified between primary and recurrent GBM samples, and 2684 potential target genes were found. Besides, regulatory co-expression network was constructed, including 12 DEMs and 81 potential target genes. These genes significantly enriched in ECM-receptor interaction, ribosome, and focal adhesion pathways, and DEMs like hsa-miR-320a, hsa-miR-139-5p, has-miR-128, hsa-miR-140-5p, and hsa-miR-146b-5p had high degree. Notably, 7 DEMs in network were known GBM-associated miRNAs recorded in database miRWalk. CONCLUSIONS: DEMs like hsa-miR-320a, hsa-miR-139-5p, has-miR-128, hsa-miR-146b-5p, hsa-let-7c, hsa-miR-128, and hsa-let-7a might participate in recurrent GBM. These results would pave ways for further study of recurrent GBM mechanism, and for the prevention and treatment of recurrent GBM. However, more experimental verifications are required to prove these predictions.