A Confirmatory Study To Probe The Reproducibility Of Cancer Cell Lines As Lab Tools To Identify Molecular Subtype-Specific Therapies And Therapeutic Targets Against Glioblastoma Stem-Like Cells

Objective: Evidence that some high-impact biomedical results cannot be repeated has stimulated interest in development of standardization and normalization tools. Cancer cell lines are the workhorse of preclinical cancer research and their ability as test-matrix to prove reproducible results has been challenged. Stratifying tumors according to their DNA-methylation and transcriptome has revolutionized neuropathology and raises hopes to develop personalized management options for brain cancer patients. Here, we probe the reproducibility of glioblastoma stem-like cell lines (GSCs) to function as a study tool for the identification of tumor-subtype-specific therapeutic targets and therapies. Methods: A collection of phenotypically diverse GSCs was assembled. In vitro aging was enforced by controlled passaging of n=30 passages and comparative analysis of aging markers before and at the end of the experiment. High-resolution (at least n=5 different concentrations of the substances), medium-throughput (231 clinical-relevant drugs) pharmacology was performed using automation-assistance. Therapy response is presented by quantification of IC50 and AUC in terms of inhibition of cellular growth. Quantification of gene expression was performed using whole transcriptome RNA sequencing and global DNA-methylation was characterized using EPIC BeadChip. In silico prediction of drug response was performed by correlating the transcriptome with gene expression/drug-response data retrieved from publically-available sources. Results: Classification of cell models into transcriptional tumor subtypes was feasible but models show limited difference in DNA-methylation. Reproducibly of drug-response was limited with significant alterations observed in every 8th of the tested compounds. Our comparative results provide a large and technical most accurate alignment of in silico drug screen of GSCs to its physical acquired counterpart. Conclusions: GSCs are a practical study tool to perform throughput pharmacology testing in molecular subtype-specific manner. Unequivocal association of cell models into established molecular subtypes as identified in clinical specimens is partly feasible. Meanwhile our project is ongoing, we hypothesize the generation of a multi OMICS/chemo responsome compendium is an valuable approach to increase the robustness of cell line-guided studies aiming to discover therapeutic targets and therapeutics. Citation Format: Ann-Christin Nickel, Daniel Picard, Gonzalo Torga, Hans-Jakob Steiger, Daniel Hanggi, Maria Stella Carro, Marc Remke, Ulf Dietrich Kahlert. A confirmatory study to probe the reproducibility of cancer cell lines as lab tools to identify molecular subtype-specific therapies and therapeutic targets against glioblastoma stem-like cells [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1929.