MicroRNA-Mediated Dynamic Bidirectional Shift between the Subclasses of Glioblastoma Stem-like Cells

Abstract

Large-scale transcriptomic profiling of glioblastoma (GBM) into subtypes has provided remarkable insight into the pathobiology and heterogeneous nature of this disease. The mechanisms of speciation and inter-subtype transitions of these molecular subtypes require better characterization to facilitate the development of subtype-specific targeting strategies. The deregulation of microRNA expression among GBM subtypes and their subtype-specific targeting mechanisms are poorly understood. To reveal the underlying basis of microRNA-driven complex subpopulation dynamics within the heterogeneous intra-tumoral ecosystem, we characterized the expression of the subtype-enriched microRNA-128 (miR-128) in transcriptionally and phenotypically diverse subpopulations of patient-derived glioblastoma stem-like cells. Because microRNAs are capable of re-arranging the molecular landscape in a cell-type-specific manner, we argue that alterations in miR-128 levels are a potent mechanism of bidirectional transitions between GBM subpopulations, resulting in intermediate hybrid stages and emphasizing highly intricate intra-tumoral networking.

Keywords: Polycomb repressive complex; chromatin; exosomes; glioblastoma; heterogeneity; microRNA; non-coding RNA; stem cells; subtype transition.

Figure 1. MiR-128 expression drives GSC subtype shift

A. MiR-128 is diversely expressed in GSC subclasses. QPCR analysis of mature miR-128 in GSCs (n=8). Data shown as mean ± SD (** p <0.001). B. Putative miR-128 targets are diversely expressed in GSCs subclasses. 5666 putative miR-128 target genes were assessed based on the value of expression. (p <0.05, fold >2 cutoff is indicated) (left). The expression of miR-128 putative target genes in GSCs stratifies P from M GSCs (n=3). Targets significantly deregulated in M GSCs (n = 569) and P GSCs (n = 741) were queried with a gene signature retrieved from TCGA and identified by clustering with subtype prediction (classical – C, blue; mesenchymal – M, red; proneural – P, magenta; neural – N, green) (right). C. Engineered GSCs with overexpression and knock-down of miR-128. QPCR analysis of mature miR-128 in GSCs (n=3). Data shown as ± SD, (**p<0.001) (left). Deregulation of miR-128 in GSCs shifts subtype-specific gene signature. QPCR analysis of subtype-specific 10-gene signatures in M GSCs (middle) and P GSCs (right). Data are shown as a heat map with cell and gene clustering. NC – negative control (empty pCDH), 128 – overexpression of miR-128 (pCDH-miR-128), aNC – negative control (empty MIR-ZIP), a128 – knockdown of miR-128 (MIR-ZIP-anti-miR-128). D. Deregulation of miR-128 shifts GSC transcriptome clustering. Gene microarray analysis of 3017 genes significantly (p <0.05, fold >2) deregulated in GSCs (n=3) in a miR-128-dependent fashion. Data are shown as a heat map with cell clustering (left) and as a PCA analysis (middle). The expression of genes deregulated by miR-128 in GSCs results in a transitional signature resembling the classical subtype. Targets significantly deregulated in M and P GSC (n = 3017) were queried with a gene signature retrieved from TCGA and identified by clustering with subtype prediction (n= 485, overall error rate >0.05) (classical – C, blue; mesenchymal – M, red; proneural – P, magenta; neural – N, green) (right).

Figure 2. PRC complex proteins and their effectors are miR-128-dependent controllers of GSC subtype shift

A. MiR-128 deregulation affects expression of PRCs, their chromatin marks, and PRC-dependent genes in GSCs. Western blot analysis of selected factors in control and miR-128 de-regulated GSCs (n=3). B. Expression of PRC-dependent genes is shifted by miR-128. Microarray platform analysis of a 30 PRC-dependent gene signature in control and miR-128 de-regulated GSCs. Data are shown as a heat map with cell and gene clustering (n=3). C. Expression of PRC-dependent genes is diversely shifted by Bmi1/Suz12 knockdown and miR-128. Microarray platform analysis of a 30 PRC-dependent gene signature in control, miR-128-overexpressing and Bmi1/Suz12 siRNA-transfected M GSCs. Data are shown as a heat map with cell and gene clustering (n=3). D. Bmi1/Suz12 knockdown deregulates the GSC transcriptome in a subtype-dependent manner. Gene microarray analysis of genes significantly (p <0.05, fold >2) deregulated in GSCs in a Bmi1/Suz12-dependent fashion (n=3). Data are shown as a heat map with cell and gene clustering (left) and as a PCA analysis (middle). The expression of genes deregulated by Bmi1/Suz12 knockdown in M GSCs results in a transitional signature resembling the classical subtype. Genes significantly deregulated in M and P GSC (n=937) were queried with a gene signature retrieved from TCGA and identified by clustering with subtype prediction (n=380, overall error rate >0.05) (classical – C, mesenchymal – M, proneural – P, neural – N) (right).

Figure 3. MiR-128 mitigates the aggressive phenotype of M GSCs both in vivo and in vitro

A. MiR-128 overexpression reduces M GSC tumor burden. MRI imaging (left) (scale bar: 2.5 mm); and tumor volume quantification (right) (n=3). Data as mean ± SD. **p<0.01. B. MiR-128 overexpression in M GSC tumors is associated with prolonged survival. Median survival: M GSC control = 11 days, M GSC miR-128 = 16 days. N = 8 mice per group. C. MiR-128 targets PRCs’ components in vivo. Western blot analysis of M GSC cells (day 0) and in M GSC tumors (day 10) (left). Western blot quantification (middle, n = 3) and qPCR analysis of miR-128 expression in tumors (day 10, right, n = 3). Data as mean ± SD, *p< 0.05, **p<0.01. D. MiR-128 overexpression shifts subtype-specific gene signature in vivo. QPCR analysis of a 10-gene subtype-specific signature in M GSC tumors (left). MiR-128 overexpression shifts the expression of PRC-dependent genes in vivo. Microarray platform analysis of a 30 PRC-gene signature in M GSC tumors (n=3). Data as a heat map with cell/gene clustering (right).

Figure 4. Targets of miR-128 and miR-128-dependent PRC associated genes predict patient outcome in GBM subtypes

A. MiR-128 knockdown in P GSC tumors is associated with shorter survival. Micrographs of tumors (left) and survival (right) are shown. N = 4 or 5 mice per group; average survival: P GSC control: 168 days, P GSC amiR-128: 143 days, scale bar: 1 mm (upper), and 200 µm (lower). B. Expression of miR-128 affects the survival of heterogeneous GSC-originated tumors. Micrographs of tumors (left) and survival (right) are shown. N = 4 mice per group; average survival: M GSC miR-128/P GSC control: 19 days, M GSC miR-128/P GSC amiR-128: 9 days, scale bar: 500 µm (upper), and 50 µm (lower). C. MiR-128-dependent transcriptome separates patients’ survival prediction. Genes deregulated in GSCs queried with TCGA gene signature were identified by clustering with class (top bar) and subtype (bottom bar) (left), miR-128 deregulated target genes (middle; value of expression, p <0.05, fold >2, n = 43) and class survival prediction (right). D. MiR-128 deregulation in M GSCs and P GSCs shifts the transcriptome towards a transitional hybrid classical-like subtype.