Differential metabolic profiles mediated SLC1A2 confer resistance in glioblastoma therapies : stratification goals and implementation

• Main Author: Koh, Lynnette Wei Hsien
• Other Authors: Tan Nguan Soon
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Science::Biological sciences::Molecular biology
• Online Access: https://hdl.handle.net/10356/138683
• Institution: Nanyang Technological University

Glioblastoma (GBM) is the most lethal malignant brain tumor, where its epigenetic, metabolic and molecular heterogeneity contribute largely to therapeutic resistance and relapse. Better elucidation of key oncogenic nodes as well as metabolic regulators driving the proliferation and drug resistance of glioma-propagating cells (GPCs) will be imperative for development of more efficacious therapeutic strategies. We previously implicated an elevated superoxide: hydrogen peroxide (O2- : H2O2) ratio as a tumor cell pro-survival signal and established a reactive oxygen species (ROS) gene signature from our GPCs, with clinical validation in major patient glioma databases. Enrichment of our ROS signature with global epigenetic regulator profiles identified EZH2, the catalytic subunit of Polycomb repressive complex 2 (PRC2) that is frequently dysregulated in GBM tumors. EZH2 therefore represents a promising target for ROS-stratified patients. Whilst current EZH2 inhibitors (EZH2i) compete with S-adenosyl methionine (SAM) binding, disrupting its catalytic SET function, there have been no clinical trials listed exploring the application of EZH2i in adult stratified GBM patients. We now demonstrate that patient stratification identifies individuals most likely to receive treatment benefit from EZH2 inhibition. Interestingly, there is growing evidence for methyltransferase-independent functions of EZH2, which could account for the inability of current EZH2i to fully suppress its oncogenic potential. We provide insights into the mechanisms of EZH2 in constitutively activating NF-κB signaling in ROS(+) GPCs, independent of its methyltransferase activity but forming a ternary complex with other NF-κB subunits. In contrast, EZH2 negatively regulates NF-κB target gene expression in ROS(-) GPCs by directing repressive H3K27me3 mark on their promoters, thereby abrogating ROS-mediated chemoresistance. These distinct EZH2 mechanisms in both responder and non-responder patient cohorts thus warrant combinational therapeutic strategies, and/or novel EZH2i that function independently of the SET domain. Collectively, our study reveals the complexity of therapeutic decisions aimed at eradicating the self-renewing, tumorigenic phenotype.