EPCO-63. LIPID METABOLIC HETEROGENEITY ACROSS GLIOMA CELLULAR STATES REVEALS ENVIRONMENTAL DEPENDENCIES

Abstract Gliomas are lethal malignancies composed of heterogeneous and dynamic subpopulations of cellular states that resemble both normal neurodevelopmental cell types and adaptive responses to stress within the tumor microenvironment (TME). While cellular plasticity enables gliomas to survive environmental pressures, the mechanisms driving glioma state dynamics remain unclear. To explore how glioma state heterogeneity functionally and metabolically interacts with the brain TME, we conducted a multi-omic analysis across 392 glioma specimens, including patient tumors, orthotopic xenografts and gliomasphere cultures using bulk RNA/whole-exome sequencing and lipidomic profiling. Single-cell transcriptome sequencing of a diverse panel of glioma tumors (n = 21) integrated with bulk tumor transcriptome profiling revealed a spectrum of cellular identities associated with distinct lipidomic profiles. Comparison of matched patient and derived models showed that the non-native environments, such as gliomasphere culture, restricted glioma state diversity and enriched for states characterized by elevated de novo fatty acid synthesis. Across environmental contexts, glioma state plasticity was coupled to lipid metabolic plasticity, enabling tumors to concordantly remodel their state composition and lipid profiles in response to environmental constraints. Of note, certain tumors exhibited limited state and metabolic plasticity – specifically those enriched for states within oligodendroglial and neuronal lineages. These gliomas demonstrated reduced de novo lipid synthesis capacity, and a concordant increased reliance on exogenous lipid scavenging within the brain microenvironment for survival. Together, these results link glioma heterogeneity and plasticity to lipid metabolic reprogramming, and highlight how distinct cellular state profiles result in environment-dependent metabolic liabilities.