Prognostic microRNAs in high-grade glioma reveal a link to oligodendrocyte precursor differentiation

Abstract

MicroRNA expression can be exploited to define tumor prognosis and stratification for precision medicine. It remains unclear whether prognostic microRNA signatures are exclusively tumor grade and/or molecular subtype-specific, or whether common signatures of aggressive clinical behavior can be identified. Here, we defined microRNAs that are associated with good and poor prognosis in grade III and IV gliomas using data from The Cancer Genome Atlas. Pathway analysis of microRNA targets that are differentially expressed in good and poor prognosis glioma identified a link to oligodendrocyte development. Notably, a microRNA expression profile that is characteristic of a specific oligodendrocyte precursor cell type (OP1) correlates with microRNA expression from 597 of these tumors and is consistently associated with poor patient outcome in grade III and IV gliomas. Our study reveals grade-independent and subtype-independent prognostic molecular signatures in high-grade glioma and provides a framework for investigating the mechanisms of brain tumor aggressiveness.

Keywords: astrocytoma; glioblastoma; glioma; microRNA; oligodendrocyte; prognosis.

Figure 1. A computational analysis identifies common prognostic molecular signatures in high-grade astrocytoma

(A) Differentially expressed microRNAs were identified separately in GIIIA and GBM and data were merged to create a common high-grade microRNA profile associated with prognosis. Targets of significant microRNAs were predicted and pathway analysis suggests that gene expression pathways associated with OP cells may predict patient outcome. Fold change data for each microRNA differentially expressed between each cell in the OP differentiation pathway was correlated with microRNA fold change data calculated between poor and good prognosis groups (and IDH1 mutation/IDHwt tumors) in GIIIA and GBM. (B) MicroRNA expression profiles for all 597 TCGA malignant glioma (GIIIA and GBM) were correlated with the expression values of each cell type in the OP differentiation pathway [27].

Figure 2

(A-B) Fold changes of the differentially expressed microRNA expression between the good and poor prognosis groups in GIIIA and GBM. (C) Plot of the microRNAs differentially expressed between good and poor prognosis groups when data from GBM and GIIIA are combined. 63 microRNAs (in red) are significantly altered between good and poor prognosis groups (p<0.05) and have a z-value of at least 2/−2. (D) The targets of the 63 microRNAs associated with patient outcome were predicted and pathway analysis revealed a significant enrichment of genes involved in several OP-related pathways.

Figure 3. Correlation coefficients comparing the fold change of microRNA expression between each stage in the OP pathway and the GIIIA and GBM good and poor prognosis groups

The top 6 rows relate to data from Letzen et al. and the bottom rows refer to data from Goff et al. and Risueño et al. [27,40,41]. The highest negative correlation is the transition from OP1 to OP2 and the highest positive correlation is the transition from GP to OP1.

Figure 4

(A) Density plots of the Spearman's correlation coefficients for each cell type with all GIIIA and GBM tumors in the TCGA. All cell types in the OP differentiation pathway show positive correlation of microRNA expression with each tumor. The oligodendrocyte lineage and OP1 cells show the most significant positive correlations with the tumors. (B) Heatmap of correlation of each GIIIA/GBM tumor with each OP cell type. (C) Hazard ratios from Cox regression analysis of the correlation patterns of each cell type shows that OP1 microRNA expression correlation is the most predictive in terms of prognosis. MicroRNA profiles of all cell types were significantly associated with survival (p<0.05); however, there is a peak in statistical power when OP1 cells are used as the predictor. (D) Kaplan Meier plot of the OP1 correlation coefficients for grade III and IV gliomas. Groups are separated above and below the median correlation of microRNA expression between OP1 and tumor.