Discovery of miRNA-mRNA regulatory networks in glioblastoma reveals novel insights into tumor microenvironment remodeling

Abstract

Adult glioblastoma (GBM) is a highly aggressive primary brain tumor, accounting for nearly half of all malignant brain tumors, with a median survival rate of only 8 months. Treatment for GBM is largely ineffective due to the highly invasive nature and complex tumor composition of this malignancy. MicroRNAs (miRNA) are short, non-coding RNAs that regulate gene expression by binding to messenger RNAs (mRNA). While specific miRNA have been associated with GBM, their precise roles in tumor development and progression remain unclear. In this study, the analysis of miRNA expression data from 743 adult GBM cases and 59 normal brain samples identified 94 downregulated miRNA and 115 upregulated miRNA. Many of these miRNA were previously linked to GBM pathology, confirming the robustness of our approach, while we also identified novel miRNA that may act as potential regulators in GBM. By integrating miRNA predictions with gene expression data, we were able to associate downregulated miRNA with tumor microenvironment factors, including extracellular matrix remodeling and signaling pathways involved in tumor initiation, while upregulated miRNA were found to be associated with essential neuronal processes. This analysis highlights the significance of miRNA in GBM and serves as a foundation for further investigation.

Keywords: Extracellular matrix; Glioblastoma; Growth signaling; Tumor microenvironment; microRNA.

Fig. 1

Schematic overview of the miRNA analysis pipeline.

Fig. 2

PCA and volcano plots identify differences in miRNA expression between GBM and controls in nine datasets. Each panel highlights data from one dataset: GSE158284 (a), GSE109628 (b), GSE90603 (c), GSE65626 (d), GSE25631 (e), GSE91014 (TCGA) (f), GSE165937 (g), GSE214252 (h) and PDC000204 (i). Left side of each panel shows PCA clustering based on miRNA expression data, with GBM brain samples in light blue and non-cancerous brain samples in light red. On the right side, volcano plots illustrate differentially expressed miRNA. Total number of significantly dysregulated hits in GBM (q-value < 0.05, log2 fold change > 1) is noted for each dataset. Significantly upregulated miRNA are highlighted in dark red, while downregulated miRNA are in dark blue.

Fig. 3

Identification of miRNA and protein-coding genes widely dysregulated in GBM. Volcano plot illustrating dysregulated miRNA (a) and protein-coding genes (b) in GBM across all 9 and 3 datasets, respectively. Total number of significantly dysregulated hits in GBM (q-value < 0.05, log2 fold change > 1) is noted for plot. Significantly upregulated genes are highlighted in red (a) and orange (b), while downregulated genes are in blue (a) and purple (b).

Fig. 4

Pathways linked with downregulated miRNA hits showcase their involvement in ECM remodeling, growth, and immune-related signaling. (a) Network map of pathway enrichment analysis from upregulated genes targeted by downregulated miRNA. (b) Dot plot revealing targets of candidate downregulated miRNA associated with ECM remodeling pathways. (c) Dot plot revealing targets of candidate downregulated miRNA associated with growth-stimulating and immune-related signaling pathways.

Fig. 5

Pathways linked with upregulated miRNA hits highlight their involvement in core neuronal processes. (a) Network map of pathway enrichment analysis from downregulated genes targeted by upregulated miRNA. (b) Dot plot revealing targets of candidate upregulated miRNA associated with core neuronal pathways.