Identifying and Targeting 3D Regulatory Nodes that Maintain Glioblastoma Programs

PROJECT SUMMARY Glioblastoma (GBM) is a highly aggressive and incurable brain tumor. The inter-patient and intra-tumoral heterogeneity of GBM, resulting from genetic alterations and epigenetic plasticity, poses a major challenge in its treatment. GBM IDH-wt tumors are composed by different proportions of transcriptionally defined cell states, which although resemble neurodevelopmental cell types, are highly plastic and interconvertible -rather than hierarchic- as we and others have shown. This suggests the presence of a core regulatory logic that enables toggling among different transcriptional states, endowing GBM tumors with increased phenotypic plasticity and fitness. Here, we aim to unravel core regulatory modules that are critical for GBM programs with a particular focus on enhancers, which together with transcription factors govern cell-state specific programs. Enhancer dysregulation by genetic variants and epigenetic mechanisms is increasingly appreciated as a key process in oncogenic transformation and drug resistance. However, dissecting and modulating enhancer function remains very challenging due to the large number of putative enhancers and the complex ways they control their target genes in the context of the three-dimensional (3D) genome. By constructing 3D enhancer-promoter interaction networks in four patient-derived glioma stem-like cells (GSCs) and normal neuronal stem cells -as controls- we have identified a subset of GSC-specific hyperconnected enhancers, which we coin "3D regulatory hubs”. 3D hubs harbor genes with robust and coordinated transcriptional levels that enrich for oncogenic pathways and are associated with worse patient outcomes. Importantly, epigenetic perturbation of a highly-recurrent enhancer hub in GSCs resulted in concordant donwregulation of multiple hub-connected genes, leading into significant shifts in the transcriptional states and altered clonogenic and proliferation capacities. Building on this foundational work, we propose that de novo 3D regulatory hubs (established by genetic or epigenetic mechanisms) lie in the core of GBM networks, where they connect and control multiple target genes, resulting in non- linear effects on the transcriptional program and oncogenic behavior. To address this hypothesis, our interdisciplinary team will combine advanced chromatin topology assays, computational modeling and network analysis with state-of-the-art epigenetic engineering and proteomics tools as well as powerful ex vivo and in vivo functional assays. Specifically, we will pursue the following aims: (i) characterizing the inter-patient and intra- tumoral heterogeneity of 3D regulatory networks and identifying conserved structures across patients and states, (ii) predicting and targeting candidate central hubs and interrogating the molecular and functional consequences and (iii) uncovering critical players of hub organization and unique vulnerabilities. The findings will provide insights into enhancer-based reprogramming of cancer fate, opening new avenues for therapeutic targeting of GBM and establishing a paradigm for identifying and prioritizing key enhancers and regulatory factors in oncogenic programs. 1