Dissecting spatiotemporal heterogeneity of glioblastoma evolution under therapy

PROJECT SUMMARY: Glioblastoma (GBM) is an uncurable form of primary brain tumor with extremely poor prognosis. Despite multi- modal therapy including surgery, irradiation and chemotherapy, all patients experience tumor progression. No standard of care is established in recurrent or progressive GBM. The identification of a neuronal cellular state of

GBM as a more differentiated state enriched at recurrence and periphery of the tumor provides new insights into how neuronal activity regulates tumor invasion. Deconvolution of normal cell types from single-cell RNAseq and bulk tumors revealed that neuronal state of GBM is associated with high infiltration of non-malignant cells. The

intricate synaptic communications between neurons and brain tumor cells are crucial for glioma progression and resistance to standard therapies, which is supported by ample data from the rapidly emerging field of “cancer neuroscience”. However, the molecular mechanisms driving enhanced neuronal activity at recurrence remains

to be understood. Dissecting the spatiotemporal dynamics of glioma eco-system during evolution will be necessary to identify therapeutic vulnerability of recurrent GBM. Here, proteogenomics and single-nuclei RNAseq profiling of matched primary and recurrent GBM IDH wild-type suggest that the evolutionary transition

from a more proliferative-progenitor towards the neuronal state in GBM is regulated by both genetic and post- genetic molecular events and potential functional interactions between malignant and non-malignant cells. In Aim 1, the development of a multiomics-based network diffusion approach will enlighten subnetworks of

proteins/phospho-proteins significantly affected by upstream genetic events driving activation of neuronal programs during progression. Generation of in silico knock-out networks screen and integrative analysis of proteomics and pharmacological data of cancer cell lines will prioritize lethal and essential proteins in the

subnetworks to identify potential therapeutic vulnerabilities in neuronal-recurrent GBM. In Aim 2, single-nuclei and spatial transcriptomics profiling of matched primary and recurrent GBM IDH wild-type will identify the functional connections between neurodevelopmental tumor cellular states and cell types in tumor

microenvironment. The development of a spatial informed cell–cell communications algorithm and the reconstruction of intercellular signaling networks will infer the key functional interactions between neurodevelopmental tumor cellular states and cell types in tumor microenvironment along with the potential effect

of these interactions on downstream regulatory molecular pathways. These studies lay the foundation for my future research program and will advance the understanding of the molecular mechanisms mediating glioma connectomes and driving glioma invasion. These studies will advance the neuroscience field through discovery

of targetable pathways and proteins providing therapeutic opportunities for recurrent GBM.