EPEN-17.IN VITRO MODELLING OF PEDIATRIC SUPRATENTORIAL EPENDYMOMAS USING CEREBRAL ORGANOIDS

Abstract

Abstract BACKGROUND Ependymomas (EPN) are glial tumors of the central nervous system occurring in children and adults. ZFTA-fusion positive and YAP1-fusion positive EPN are the main supratentorial (ST) subgroups in children. While most ST-EPN-YAP1 patients survive, only 50% of ZFTA positive patients survive longer than 5 years. Improving survival of ST-EPN patients requires a better understanding of tumor biology. Human embryonal stem cell-derived (hESC) brain organoids provide a novel opportunity to model ST-EPN and study the impact of oncogenic fusions on tumor development within a healthy brain environment. METHODS hESC-derived cerebral organoids were genetically modified with YAP1 or ZFTA fusion genes and histologically and molecularly analyzed using antibody stainings, bulk and single cell RNA sequencing. RESULTS scRNA-seq analyses showed that our cerebral organoids mimic embryonal brain development and that radial glia, the presumed cell-of-origin of ST-EPN are abundant in 11-day old organoids. At this timepoint, electroporation of oncogenic ZFTA-RELA or YAP1-MAMLD1/YAP1-FAM118B fusions led to ectopic tumor outgrowth. Histological and molecular analyses showed that ZFTA and YAP1 EPN tumor organoids displayed different phenotypes and fusion-specific gene signatures, closely resembling human ZFTA and YAP1 EPN patient samples. ScRNA-seq data of organoid tumors showed a skewed differentiation compared to normal development with ZFTA tumors being more neuronal and YAP1 tumors having a more extracellular matrix-like phenotype. Analysis of the healthy compartment showed that YAP1 tumors influence the differentiation of the healthy cells and in both subtypes a new cell cluster was identified with high expression of tumor associated markers. Intercellular interactions revealed potential targetable tumor(specific) interactions such as YBX1-NOTCH1 in ZFTA or GNAI2-CAV1 in YAP1 tumor organoid models, as well as CD99-CD81 in both. CONCLUSIONS These models contribute to a better molecular and biological understanding of ependymomas and can be used to identify targeted therapies, especially those targeting the tumor microenvironment.