MSACL 2020 US Abstract

Full-scan Metabolomic Profiling of Glioblastoma Stem Cells Discovers Metabolic Pathway Variation Associated with Clinical Outcomes

Amy A. Caudy(1), Olga Zaslaver (2), Michelle Kushida (3), Polina Sutyrina (2) Artee Luchman (4), Sam Weiss (4), Peter Dirks (3), Adam P. Rosebrock (1)
(1) Stony Brook School of Medicine, Stony Brook, USA (2)University of Toronto, Toronto, Canada. (3) The Hospital for Sick Children, Toronto, Canada, (4) University of Calgary, Alberta, Canada

Warning: Undefined variable $headshot in /var/www/html/view_abstract/view_abstract_in_program.php on line 704

Amy Caudy (Presenter) Stony Brook University

Presenter Bio: Dr. Caudy is a Research Associate at Stony Brook University, and has led the metabolomics analysis in the Stand Up To Cancer Canada Brain Tumor Stem Cell Dream Team. She uses full scan metabolomics to discover biomarkers and enzyme function.

Relevant Financial Disclosures (within past 24 months) Honorarium/Expenses Agilent Grant/Research Support Agilent Stock/Bonds Maple Flavored Solutions Discussing Products? NO, presenter will NOT present educational content related to products/services of the ineligible company(ies) listed above. Conflict mitigation NOT REQUIRED.

Introduction: Glioblastoma multiforme (GBM) is a deadly cancer with few treatment options. Rare stem cells underlie glioblastoma tumor drug resistance and inevitable recurrence. Our team created methods to reproducibly culture these cells, and we have now profiled tumor stem cell lines from more than 40 pediatric and adult patients using genomics, transcriptomics, chromatin accessibility, and metabolomics methods. We are using these data to propose new drug targets by determining metabolic pathways that are active in tumors as compared to normal, nontumorigenic neural stem cells.

Methods: This project generated more than 1000 intracellular and extracellular brain tumor stem cell metabolite extracts across a period of years. Rather than delaying analysis until complete sample accrual, we developed reproducible and robust methods for metabolomics analysis that allow sample- and batch-wise correction. We analyzed each sample using an negative-mode ion-paired method that retains polar metabolites and an acidic reverse phase method that resolves isomeric compounds in the TCA cycle. In order to quantify relative metabolite levels across samples, we incorporate a stable isotope labeled biological reference material into each sample. We employ multiple sample- and batch-wise methods of normalization to allow robust comparison.

Results: Patient-derived tumor stem cell lines vary in their levels of the lysine metabolite alpha-aminoadipic acid (a-AA). Variation in a-AA has been linked to glioblastoma patient survival by other studies. We find that changes in the levels of other lysine catabolic intermediates are consistent with a-AA variation is driven by altered lysine use. Using RNA-seq data, we predict the genes that underlie these alterations and propose that this variation in lysine catabolism is the consequence of differences in mitochondrial function.
We also observe marked differences in the extracellular levels of the epigenetic metabolite methylthioadenosine. These differences are explained by the expression level of the methylthioadenosine phosphorylase MTAP. Our metabolomic observations challenged the predictions of MTAP status made by whole-genome sequencing, demonstrating a role for metabolomics in accurate tumor phenotyping.
Glutamate consumption is an important driver of cell growth. Residual glutamate in spent culture media varies, showing differences in glutamate consumption. Correlation with RNA-seq data shows that expression of the GLUL glutamine synthase is the major driver of this variation. We also observe that the degree of injury response signal is correlated with several metabolites in central carbon metabolism.

Conclusions & Discussion: Our study demonstrates the potential for our full-scan metabolomics approach to discover clinically relevant metabolites. Integration of metabolomics analysis with genomics and gene expression data proposes potential routes for altering the growth and survival of these tumor stem cells.