P01.01.A RIBOSOME COLLISION IS A TARGETABLE VULNERABILITY IN GLIOMA

Abstract BACKGROUND Glioblastoma (GB) is the most common malignant adult brain tumor. Modern chemotherapies and immunotherapies have failed to improve the outcome of GB. Thus, there is a need for developing novel approaches to target GB. Programmable cell death (PCD) programs have been long hailed as potential targets for cancer. Ferroptosis, an iron dependent cell death characterized by lipid peroxidation, has been identified as a potential target for various cancers’ therapy, including GB. However, how GB responds to ferroptosis remains unknown, hindering effective targeting. MATERIAL AND METHODS Genome-wide negative CRISPR screening was conducted in U87MG glioma cell line using 2 different classes of ferroptosis inducers. CRISPR/Cas9 mediated KO of the identified target (EDF1) was conducted to validate the downstream effect. RNA-seq and Ribo-seq were used to study the translational impact of EDF1 on glioma cells. Downstream validation of dysregulated pathways was conducted using gene knockdown and various methods. RESULTS Genome-wide CRISPR screening revealed that the top gene in 2 independent screens was EDF1, a sensor of ribosome collisions. Induction of Ferroptosis in glioma cells activated the ribotoxic stress pathway (RSR) and the integrated stress response pathway (ISR) and upregulated EDF1 translationally. Ribo-seq analysis revealed significant ribosome stalling at distinct codons and activation of a G/C-ending codon biased translation characteristic of stress response. EDF1 knockout (KO) led to reduced proliferation and stemness of glioma cells. EDF1 KO rendered glioma cells extremely sensitive to ferroptosis and interfered with the activation of the RSR in response to ferroptosis inducers. RNA-seq and Ribo-seq analysis revealed dysregulated transcription and translation in response to EDF1 KO with and without stress induction. In normal growth conditions, we observed severe ribosome stalling at hydrophobic amino acids and mismatch between transcription and translation. During stress, EDF1 KO cells were unable to properly mount the stress-responsive codon biased translational program and showed severe ribosome stalling at sublethal ferroptosis inducers concentrations. Sequencing analysis revealed an upregulation of mitochondria-related pathways at the translational level. Flux analysis revealed hyperactive mitochondria that were structurally different from normal glioma mitochondria. Further analysis of the sequencing data revealed the dysregulated expression of various genes linked to glioma oncogenesis and ferroptosis resistance. Of note, Androgen receptor (AR) showed a near complete disappearance of its protein levels. AR is known to promote oncogenesis and ferroptosis resistance CONCLUSION our results indicate that targeting the ribosome quality control machinery that operates during oxidative stress is a viable target for GB therapy.