Therapeutic compound discovery and validation of drugs targeting a recurrent glioblastoma (GBM) phenotype using LINCS compounds via QUADrATiC analyses


Session type:

Shahnaz Al Rashid1,Paul O'Reilly1,Philip Dunne1,Matthew Alderdice1,Thomas Flannery1,Kevin Prise1,Shu-Dong Zhang1,Darragh McArt1
1Queen's University Belfast



Glioblastoma (GBM) is the most aggressive of malignant primary brain tumours. Despite multi-modal therapy, including surgical resection, radiotherapy and temozolomide chemotherapy, GBM continues to be uniformly lethal. Resistance and ultimately recurrence is thought to originate from a sub-population of cells residing in the initial tumour, identified as glioma-initiating or glioma-stem cells (GSC). These GSC have been shown to be intrinsically more invasive, with enhanced repair and metabolic pathways, and enriched in recurrent tumours. This study utilized a novel connectivity mapping software developed in-house, QUADrATiC (O’Reilly, et. al., BMC Bioinformatics, 2016), to identify candidate FDA-approved compounds within the LINCS (Library of Integrated Network-based Cellular Signatures) database to target therapy-resistant GSC.


A differentially expressed gene list was generated by contrasting gene expression profiles from publicly available NCBI GEO microarray datasets of clinically resected primary and recurrent glioblastoma tissues. QUADrATiC was used to map connections between this gene list and drug responses from the LINCS database. Among the top compounds identified, rosiglitazone (rosi), a ligand of the PPARg nuclear receptor, was validated in vitro, as a single agent and in combination with, radiation and/or temozolomide. Clonogenic and western blot assays were used to determine non-stem and GSC cell kill and cell death pathways, respectively.


Rosiglitazone significantly reduced survival of both non-stem glioma cells and GSC populations, alone and in combination with radiation and temozolomide (p < 0.05). Furthermore, rosi sensitized GSC significantly (p < 0.05) more than non-stem glioma cells. Rosi induced the p53 pathway which may mediate GSC cell death.


This work validates the efficacy of rosi at targeting GSC and supports further exploration of its use in combination with current treatments. This work also highlights the capability of the QUADrATiC package at identifying optimal agents for use in a clinical setting, specifically targeting the GSC sub-populations within these heterogeneous tumours.