Back

Identification of Injury-Response State Determinants in Glioblastoma Stem Cells

Molaei, F.; MacLeod, G.; Haider, S.; Tiffany, A.; Oteng, F. M.; Berman, J. M.; Skirzynska, A.; Shoichet, M. S.; Schramek, D.; Dirks, P. B.; Angers, S.

2025-04-12 cancer biology
10.1101/2025.04.07.647626 bioRxiv
Show abstract

Glioblastoma (GBM) is a common and highly lethal type of primary brain tumor in adults. Therapeutic failure is partly attributed to a fraction of Glioblastoma Stem Cells (GSCs) that show high levels of heterogeneity and plasticity. GSCs exist in a transcriptional gradient between two states: Developmental (Dev) and Injury Response (IR) in which IR-GSCs exhibit more invasive behaviors. While previous studies have identified fitness genes in GSCs, the genes required to establish and maintain the Dev and IR states remain poorly defined. To identify the regulators of the IR GSC state, we performed a phenotypic genome-wide CRISPR-Cas9 knockout (KO) screen in patient-derived GSCs based on cell surface expression of the IR marker CD44. Notably, we found that perturbations of the histone acetyltransferase EP300 in IR GSCs led to decreased CD44 cell surface expression, significant downregulation of gene expression signatures associated with the IR state, and to decreased self-renewal and invasion. Furthermore, genetic targeting of Ep300 in a mouse GBM model delayed tumor initiation and/or progression. Collectively, our results establish EP300 as a regulator of the IR state in GSCs and provide a mechanistic basis for its therapeutic targeting in GBM. SignificanceA genome-wide phenotypic CRISPR-Cas9 screen in a patient-derived Glioblastoma Stem Cell line identified the genes required to maintain the Injury-Response cellular state, with a focus on the histone acetyl transferase gene EP300. This study suggests how therapeutic targeting of cellular state could reduce the aggressiveness of GBM tumors.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.