Preview

PDF, English (Doctoral Dissertation) - main document Download (7MB) | Terms of use

Abstract

BACKGROUND Cancer is a direct consequence of genomic aberrations, such as somatic copy number alterations that frequently occur in the cancer genome affecting not only oncogenic genes, but also multiple passenger and potential co-driver genes. An intrinsic feature resulting from such a disruption of the genome is deregulation of the tumor metabolic landscape, as a result of which, multiple metabolic genes have been identified as oncogenes, tumor suppressor genes or targets of oncogenic signaling.

RESULTS Here we elucidate that linear proximity of metabolic and cancer-causing genes in the genome can lead to metabolic remodeling through copy number co-alterations. We observed that cancer-metabolic gene pairs are unexpectedly often proximally positioned in the chromosomes and share loci with altered copy number, thus being either co-deleted or co-amplified across all cancers analyzed (19 cancer types from The Cancer Genome Atlas). We have developed an analysis pipeline - Identification of Metabolic Cancer Genes (iMetCG), to infer the functional impact on oncogenic metabolism from such co-alteration events and delineate genes truly driving cancer metabolism from those that are neutral. Using this approach, we have identified novel and well known metabolic genes that target crucial pathways relevant for tumors. Moreover, using these identified metabolic genes we were able to classify tumors based on its tissue and developmental origins. We further observed that these putative metabolic cancer genes (identified across cancers) had higher network connectivity, were indicators for patient survival, had significant overlap with known cancer metabolic genes and shared similar features with known cancer genes in terms of their isoform diversity, evolutionary rate and selection pressure.

CONCLUSIONS This thesis provides novel insights into the functional mechanism of metabolic regulation and rewiring of the metabolic landscape in cancer cells. Our pan-cancer, genomic data driven approach revealed a hitherto unknown generic mechanism for large scale metabolic reprogramming in cancer cells based on linear gene proximities and identified 119 new metabolic cancer genes likely to be involved in remodeling tumor cell metabolism. Furthermore, our newly identified metabolic cancer genes will serve as a vital resource to the experimental community engaged in tumor metabolism and genomics research to further expand the scope of this field.