Stratification of Adult Glioblastoma with Signaling and Single Cell Biology

Abstract

Genomic and transcriptomic profiling have revealed extensive cellular diversity in adult glioblastoma, a rapidly fatal brain tumor. These approaches, however, have neither successfully stratified patient outcome nor determined specific, individualized therapeutic strategies. This dissertation presents a modular analysis pipeline that integrates 1) a novel cell preparation procedure of surgically resected adult glioblastomas, 2) high-dimensional single-cell proteomics quantification, and 3) automated risk-guided population discovery. This approach revealed biologically distinct glioblastoma cells that were linked to patient outcomes, independent of known clinical predictors. The newly discovered prognostic glioblastoma cells were defined by abnormal developmental and signaling programs, which can be immediately evaluated as targets for therapy. A connection between clinically distinct glioblastoma cells and the tumor microenvironment was also revealed and should be mechanistically explored for alternative therapeutic avenues. This single-cell analysis workflow can be immediately integrated into routine clinical practice and can be broadly applied in other intractable malignancies. The application of large-scale single-cell snapshot proteomics for the interrogation of cellular mechanisms underlying tumor pathobiology and for high-throughput drug discovery is also explored.