Targeting cellular diversity and interactivity in paediatric diffuse glioma

Background Paediatric-type diffuse high-grade glioma is a biologically and clinically diverse group of brain tumours, different to those that arise in adults; for the vast majority of children with these lesions, outcomes remain dismal.

We have demonstrated the extensive inter- and intra-tumoral heterogeneity which defines subgroups of the disease, highlighting the unique developmental origins and dysregulated epigenetic processes underlying tumorigenesis.

We have identified and isolated distinct tumour cell subpopulations on the basis of genotype, lineage and identity, and established the presence of co-operative positive interactions in driving tumour phenotypes.

Understanding the mechanisms driving these cellular interactions and identifying ways to target subpopulation-specific dependencies will be critical in exploiting these observations to develop novel treatments.

Aims We will refine our understanding of the mechanisms driving tumour cell subpopulations through single cell epigenomics and map these cell types in the context of the tumour microenvironment through spatial profiling.

We will explore subclonal fitness interactions in vitro and use evolutionary principles to identify optimal treatment scheduling based upon complex mixture modelling.

We will develop genetic and pharmacological degrader strategies to target individual, epigenetically-driven cellular vulnerabilities.

Methods Chromatin accessibility and conformation will be applied at the single cell level using scATAC-seq and Hi-C/Micro-C, linked to gene (scRNA-seq) and protein (CyTOF) expression.

Tissue sections will be explored using a combination of spatial genomics (RNA-ISH), transcriptomics (Visium), and proteomics (mass cytometry) on human tissue specimens, mouse models, and ex vivo human brain slices.

Novel barcoding techniques allowing for the retrieval of cellular subpopulations with any fitness profile will be adapted for high-content screening and molecular profiling to read spatially resolved information in situ.

These will also be applied to complex co-culture systems of subpopulations (including non-tumour cells) and combined with mathematical modelling of cellular interactions.

Tuneable protein degrader-based approaches (d-TAG) will target neurodevelopmental transcription factor dependencies of these cells.

Validated targets will be prioritised for pharmacological degradation using proof-of-principle ImiD screens and proteome-wide interrogation of protein:protein interactions and druggable binding partners.

How the results of this research will be used The proposed studies exploit newly accessible single cell techniques to resolve the distinct chromatin landscapes of these subpopulations, track cellular interactions temporally and in response to therapeutic challenge, and visualise them spatially.

We will use these insights to assess the biological consequences and therapeutic opportunities of targeting these unique lineage-specific dependencies to pave the way for future drug discovery initiatives.