Stem-17. implicating astrocyte progenitor signatures from the developing human subventricular zone in glioblastoma progression and recurrence

Abstract Although astrocyte-like progenitors in the adult subventricular zone (SVZ) have been put forward as the glioblastoma (GBM) cell of origin, the mechanisms that trigger astrogenesis in the developing SVZ are poorly understood. We previously reconstructed astrocyte lineages using computational tools and identified PDGFRβ+ astrogenesis-committed radial glia (RG-A) and EGFR+ gliogenic intermediate progenitor cells (gIPCs). We have, however, yet to identify the transcription factors (TFs) that drive the differentiation of these progenitors in the pre- and postnatal SVZ. As malignant cells are known to co-opt neurodevelopmental pathways, the exploitation of TF driver vulnerabilities in malignant cells could bring novel therapeutic strategies to hinder GBM progression or recurrence. Here, we integrated single nucleus RNA- (snRNA-seq) and ATAC-sequencing (snATAC-seq) datasets from 1) surgically resected GBMs (n = 5), 2) the subventricular zone and neocortex from non-pathological adult cortices (n = 3) and 3) the corresponding germinal matrix and cortical plate of non-pathological cortices ranging from 17 to 41 gestational weeks (n = 17) and 21 to 301 postnatal days in age (n = 9). To prioritize candidate drivers of astrogenesis computationally, we correlated motif enrichment with gene expression over pseudotime in gIPC and RG-A lineages. Subsequently, we projected transcriptional signatures from non-pathological neocortices onto GBMs. Interestingly, we observed high proportions of tumor cells with preferentially enriched prenatal gIPC-like signature in EGFR amplified GBMs and with preferentially enriched postnatal RG-A-like signature in GBMs with PFGFRα or FGFR3 amplification. As tumor recurrence is associated with a mesenchymal-like switch, we were intrigued to find that RG-A-like cells exhibited a mesenchymal-like phenotype in a recurrent GBM. Currently, we are developing an in vitro assay to manipulate the expression of astrogenic drivers and force the differentiation of patient-derived, primary GBM cells into astrocytes; a strategy that we will later apply in patient-derived xenograft models.