Diffuse gliomas are the most common primary malignant brain tumors in adults. Existing therapies provide only modest benefits, so there is a pressing need for new treatment strategies. Genetic studies of gliomas have identified a number of recurrent mutations, inspiring many important efforts to develop targeted treatments. Nevertheless, because it is uncertain when or if these efforts will succeed, it is critical to expand the therapeu- tic search space for gliomas. Like all cancers, mutations that cause gliomas transform cellular identity by alter- ing gene expression. However, the transcriptional phenotypes that distinguish malignant glioma cells and cells of the glioma microenvironment from cells in the normal adult human brain remain poorly understood. This pro- ject will use novel analytical and experimental strategies to precisely define the transcriptional phenotypes that most reliably distinguish malignant and non-malignant cell classes in glioma from cells in the normal adult hu- man brain. Our central hypothesis is that integrative gene coexpression analysis of intact tissue samples can reveal the core transcriptional identities of distinct cell classes in gliomas, thereby highlighting the impact of glioma genotypes on gene expression.
In Aim 1, we will perform meta-analysis and integrative deconvolution of transcriptomes from >4K intact tissue samples from astrocytomas, oligodendrogliomas, and glioblastomas to distill transcriptional profiles of distinct cell classes. We will compare cell class-specific transcriptional signa- tures between gliomas and normal brains and validate predicted differences histologically.
In Aim 2, we will determine the most consistent molecular phenotypes of IDH1 R132H+ malignant cells in lower-grade gliomas by analyzing covariation of mutant allele frequencies and molecular features over serial sections of frozen tu- mor specimens. We will validate predictions histologically and through comparisons with normal human brain.
In Aim 3, we will perform multiscale and multiomic analysis of subclonal diversity in spatially mapped subre- gions of lower-grade gliomas. We will validate subclonal diversity and transcriptional phenotypes through tar- geted single-nucleus RNA-seq. Collectively, these experiments will provide fundamental insights into molecular mechanisms that promote gliomagenesis and substantially expand the therapeutic search space for adult ma- lignant gliomas.
Adult malignant gliomas are deadly brain tumors with limited treatment options. This project proposes a new strategy for identifying the genes that these tumors use to grow and survive in the human brain. This information is important for developing new ways to attack these tumors.
