EPCO-06. ESTABLISHING THE EPIGENETIC LANDSCAPE OF THE UNEXPLORED EZHIP-EXPRESSING DIFFUSE MIDLINE GLIOMA SUBTYPE

Abstract INTRODUCTION Diffuse midline glioma (DMG) is a terminal paediatric brain cancer which has no current cure, leading to an urgent demand to re-evaluate therapeutic strategies. DMG is an ‘epigenetically driven’ cancer, as most patients harbor the epigenetic mutation H3K27M. In 2020, a novel DMG subtype was discovered for cases lacking H3K27M, which instead aberrantly expressed the Enhancer of zeste inhibitory protein (EZHIP). Interestingly, EZHIP is a ‘mimic’ of H3K27M due to sequence similarity. This project aims to study the novel EZHIP-DMG subtype by establishing the epigenetic profile of these tumours for comparison with other DMG subtypes, and explore the EZHIP-specific epigenetic effects to predict the best therapeutic strategy. METHODS To compare epigenetic profiles of DMG subtypes, patient-derived DMG cell lines expressing EZHIP or H3K27M, H3-wildtype (without EZHIP) and normal controls were used. Epigenomic profiling was performed in cell lines for active and repressive chromatin marks by ChIP-sequencing (H3K27ac, H3K4me3, H3K4me1, H2A.Zac, H3K36me3, H3K27me3 and H3K9me3) and for chromatin accessibility by ATAC-sequencing. Integration of transcriptomic signatures was done using RNA-sequencing. RESULTS Overall, EZHIP-DMG cell lines are epigenetically more similar to H3.3K27M-DMG than H3-wildtype DMG cell lines. This was confirmed by principal component analysis for chromatin accessibility, histone modifications and gene expression. However, specific epigenetic patterns of EZHIP-DMG were also identified. Particularly, for H3K27ac and H2A.Zac histone modifications, which putatively mark active enhancers. Further analysis of the top variant H3K27ac-enhancers by hierarchical clustering revealed unique enhancers in EZHIP-DMG, reasoning the differences observed between subtypes. CONCLUSION This study is the first to confirm that EZHIP has epigenetic similarities with H3K27M in DMG beyond H3K27me3 loss, but also, EZHIP-DMG may have unique signatures at enhancers that could influence therapeutic response. Further work into identifying EZHIP-DMG specific targets in addition to targeting common epigenetic vulnerabilities for H3K27M/EZHIP in DMG will be a promising therapeutic strategy.