Deconvolution of Genetic Heterogeneity in Glioblastoma
Date
2024-04-02
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Glioblastoma, an aggressive brain tumor, exhibits substantial genetic heterogeneity that contributes to functional variations and poor outcomes. To comprehensively understand this heterogeneity, I reconstructed phylogenies using multiple samples per patient. Phylogenies were predominantly branching and revealed that higher genetic diversity was positively associated with survival in glioblastoma and that standard therapy increased diversity. Surprisingly, recently diagnosed patients fared better, potentially linking advances in disease management to evolutionary adaptations. In a novel approach, I assessed the validity of phylogenies using long-range mutation phasing with linked reads, which supported most clones and the first branching point, thereby instilling confidence in findings derived from these phylogenies. Phylogenetic reconstruction incorporated a diverse array of sample types that reflected distinct selective pressures. Longitudinal samples capture changes in clonal composition in response to selective pressure of treatment. In contrast, phylogenies generated from multiregional and/or cell line, or xenograft models identified adaptations to microenvironmental differences or culture conditions. Through subsampling, I showed that two samples were the minimum required to capture most mutations, clones, and the initial branching point. Using the phylogenies, I implicated mutational signatures as drivers of clonal divergence events during the development of treatment resistance, including SBS11, previously associated with temozolomide resistance, and DNA mismatch repair (MMR) deficiency signatures SBS6 and SBS15. MMR signatures operated largely independently of each other during clonal evolution, affected different pathways, and had distinct associations with survival. SBS23, a signature of unknown etiology, was also significantly associated with recurrent disease and SBS11 and SBS6-relevant processes. Finally, to explore the relationship between genetic and functional diversity, I integrated expression from RNA and proteome data with genome-based phylogenies using a multi-omics approach. The resulting biological communities demonstrated normal cell infiltration and tumor cell states that broadly recapitulated glioblastoma subtypes and were associated with phylogenetic branching, glioblastoma subtypes, and survival length. Finally, I observed multiple communities annotated with diverse cell types and states linked to the mesenchymal subtype, suggesting diversity within this subtype. In summary, I developed a novel workflow that encompasses phylogenetic reconstruction of glioblastoma evolutionary trajectories and links cell state phenotypes through multi-omic integration.
Description
Keywords
phylogenetics, genetic heterogeneity, clonal evolution
Citation
Gillmor, A. (2024). Deconvolution of genetic heterogeneity in glioblastoma (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.