Characterizing the Dynamics of Genome Evolution in Tumorigenesis

Cancer is a genetic disease. Transformation to malignancy occurs primarily by genome aberrations, which provide the means for acquiring essential cancer cellular programs. Although most tumors exhibit a common set of cancer capabilities, the patterns of genome aberrations can vary extensively (reviewed by Hanahan & Weinberg, 2000). Genome aberrations harbored by a tumor largely depend on the tissue type. However, many studies show that even histopathologically similar tumors can harbor heterogeneous genome aberration patterns. The reasons why histopathologically similar tumors undergo alternative genome evolutions that lead to heterogeneous genome aberration patterns (i.e. causes) or how such genomic heterogeneity translate to cancer phenotype (i.e. consequences) are not well understood. I have undertaken an integrative functional genomics approach to understanding causes and consequences of alternative genome evolutions. In this approach, (1) an experimental tumor system is used to generate histopathologically similar yet genetically diverse tumors, (2) experimental genomics assays are used to interrogate these tumors for molecular alterations on a genome-wide level, and (3) computational methods are used to translate genome-wide information into biologically meaningful knowledge. Investigating causes of alternative genome evolutions led to a finding that the initial genotype at transformation related protein 53 locus was important for subsequent genome evolution, providing evidence for a model that differences in initial premalignant conditions modulate alternative genome evolutions. Through an investigation of the relationship between genome and transcriptome changes, I identified a link between a deleted portion of mouse chromosome 6, and a conserved genome-wide transcriptional signature with prognostic value in multiple human cancer cohorts. This suggests that genome alterations influence overall cancer outcome via global transcriptional changes. Taken together, these findings demonstrate that a functional genomics approach can provide valuable insight into the dynamic nature of the genome and the role of genome evolution during tumorigenesis