RT Journal Article
SR Electronic
T1 Chromatin Blueprint of Glioblastoma Stem Cells Reveals Common Drug Candidates for Distinct Subtypes
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 370726
DO 10.1101/370726
A1 Paul Guilhamon
A1 Michelle M Kushida
A1 Graham MacLeod
A1 Seyed AM Tonekaboni
A1 Florence MG Cavalli
A1 Fiona J Coutinho
A1 Nishani Rajakulendran
A1 Xinghui Che
A1 Naghmeh Rastegar
A1 Mona Meyer
A1 Xiaoyang Lan
A1 Nuno M Nunes
A1 Uri Tabori
A1 Michael D Taylor
A1 Benjamin Haibe-Kains
A1 Stephane Angers
A1 Peter B Dirks
A1 Mathieu Lupien
YR 2018
UL http://biorxiv.org/content/early/2018/07/17/370726.1.abstract
AB Glioblastoma (GBM) is an aggressive form of brain cancer with a median survival of 12.6 months1 and for which the standard treatment of surgery, radiotherapy and temozolomide, provides only an additional 2.5 months in the small subset of responsive patients2. Despite extensive characterization and stratification of the bulk primary tumours, no targeted therapies have been successfully developed1,3.GBM tumours are rooted in glioblastoma stem cells (GSCs) that have self-renewal and tumour-initiating capacities4. GSCs also drive disease progression in vivo5,6. Although the mutational landscape of GSCs is well established7,8, and their epigenetic profile, based on DNA methylation and histone modifications, has been described for a few samples7–9, the variability in chromatin accessibility and resulting functional heterogeneity across GSCs has not been previously investigated. Indeed, GSCs derived from different patient tumours were shown to share numerous common features in their chromatin accessibility landscape10. However, phenotypic variability, such as differentiation capacity, was also reported between GSCs that exhibit specific differences in their chromatin accessibility11, warranting a comprehensive assessment of heterogeneity across GSCs.Here we reveal three novel and distinct GSC subtypes based on the integrative analysis of chromatin accessibility, DNA methylation and gene expression on a cohort of 27 patient tumour-derived GSCs. Each GSC subtype is regulated by a specific set of transcription factors, uniquely essential for growth in the respective subtypes. Through a single-cell clonal analysis, we show that a GBM tumour can harbour more than one GSC subtype. In addition, we not only identify subtype-biased growth inhibitors through our drug response screening assay but also show that all GSC subtypes commonly express the serotonin receptor 5-HT2 and are sensitive to the dopamine/serotonin receptor ligand perphenazine. Overall, our results suggest that patients could benefit from serotonin receptor inhibitors or a combination therapy to address the GSC heterogeneity using drugs targeting the different GSC subtypes populating each GBM tumour.