STEM-26. ALTERED LIPID METABOLISM MARKS GLIOBLASTOMA STEM AND NON-STEM CELLS IN SEPARATE TUMOR NICHES
Abstract Clinical glioblastoma is marked by a strikingly heterogeneous mix of cell types, cellular metabolisms, and cellular microenvironments spread in different spatial locations throughout a tumor. We have created 3-dimensional organoid models that partially mimic the transition zone between nutrient-rich cellular tumor regions and nutrient-poor psuedopallisading and perinecrotic tumor zones. We found a dramatic disparity in lipid droplet presence between these regions with high lipid accumulation in the hypoxic organoid cores of a wide spectrum of patient derived specimens. This is accompanied by regionally restricted upregulation of HILPDA gene expression in the cores of our models, in clinical GBM specimens but not lower grade brain tumors, and localized specifically to pseudopallisading regions of patient tumors. We further show that lipid droplet accumulation overall marks perinecrotic and pseudopallisading regions in clinical GBM, indicating broadly altered lipid metabolism between these distinct cell populations. High lipid droplet accumulation is largely restricted to the non-stem cell populations of GBM organoids and sorted xenograft tumors whereas the stem cells are lipid-poor, suggesting lipid levels may not be simply a product of the microenvironment but also may be a reflection of cell state. We performed global lipidomic analysis on prospectively sorted stem and non-stem cells of multiple patient-derived models and found that GBM stem cells have comparatively decreased levels of neutral lipids, indicating a significant metabolic shift compared to non-stem cells from the same patient. Conversely, GBM stem cells have significantly increased levels of rare specific lipid species, and also display altered phospholipid synthesis and species specific alterations in phospholipid classes. Our findings suggest avenues for therapy by targeting the altered lipid metabolic pathways of these disparate tumor cell populations.