Integrated genomic & epigenomic analyses of glioblastoma multiforme: Methods development and application

Promoter DNA hypermethylation is generally associated with transcriptional gene silencing. Studies show that silencing a critical tumor suppressor genes may contribute to tumorigenesis. While the guidelines that govern methylation patterns at promoter CpG islands during the pathogenesis of individual cancers are still unclear, it is widely known that certain genes carry a higher frequency of DNA methylation in select tumors whereas other genes are methylated across most types of tumors.;Using open source tools, a pipeline and a set of analytical tools were created to integrate multiple high-throughput datasets (genomic and epigenomic) in order to identify and understand the genes that are epigentically modulated and biologically silenced in Glioblastoma multiforme, a highly aggressive form of brain tumor. The ultimate goal for these integrative analyses is to bring these identified marks to the point where they can be developed as targets for diagnostic methods and treatment strategies.;The following thesis can be divided into three main sections. The first section presents an overall understanding of Epigenetics and its role in human cancer, specifically Glioblastoma multiforme (Chapter 1). The second section focuses primarily on the development and utilization of specific open source tools and analytical scripts in order to assist in analzying and integrating multiple high-throughput data derived for multiple samples (Chapter 2). Finally, the third and largest section of this thesis deals specifically with harnessing the tools created in Chapter 2 in order to fully describe and understand the molecular and clinical features associated with Glioblastoma multiforme (Chapter 3--5). The following summarizes the overall findings associated with Chapters 3--5.;We profiled promoter DNA methylation alterations associated with The Cancer Genome Atlas (TCGA) project in 272 glioblastoma tumors. Unsupervised analyses of these data revealed a novel molecular subgroup of samples (24/272, 8.8%) with highly concordant gene promoter methylation including a large number of tumor specific hyper-methylated loci (3,098 loci), indicating the existence of a glioma-CpG Island Methlylator Phenotype (G-CIMP). We characterized these G-CIMP-positive samples by integrating available TCGA data consisting of clinical features, DNA sequence alterations (somatic mutation and copy number variations), and transcriptome expression.;These G-CIMP-positive patients are younger at the time of diagnosis (median age=36) and display significantly improved outcome (median survival=150 weeks) compared to G-CIMP-negative patients. G-CIMP-positive tumors are predominantly of the proneural expression subtype (21/24,87.5%), and tightly completely associated with IDH1 somatic mutations.;Next, we analyzed copy number variation differences in proneural G-CIMP positive tumors and identified 2,875 genes that had significant copy number changes and reduction of chromosome 7 gain and 10 loss. Interestingly, we observed significant chromosome gains in 10p15.3-p11.21 and 8q23.1-q24 in proneural G-CIMP-positive compared to proneural G-CIMP-negative. We also identified significant differences in both DNA methylation (1,550 genes) and gene expression (1,575 genes) and the integration of these two experimental results identified a total of 300 genes with significant DNA hypermethylation and gene expression changes in proneural G-CIMP-positive tumors.;Gene ontology analyses showed G-CIMP-specific down-regulation of genes associated with the mesenchyme subtype, tumor invasion and the extracellular matrix as the most significant terms. Genes with roles in transcriptional silencing and chromatin structure modifications showed increased gene expression in proneural G-CIMP-positive tumors. Meta-analysis identified significant overlap with down-regulated genes in low-intermediate grade glioma compared to Glioblastoma Multiforme in a variety of previously published datasets. In addition, de novo DNA motif analysis identified a distinct motif (5'-CCCCTGGGG-3') significantly associated with G-CIMP specific locus. This motif revealed a strong homology with a well known palindromic motif which was found to be highly associated with EBF1, a known epigenetic modulator which promotes demethylation and chromatin remodeling in B-lymphocyte.;In summary, we profiled promoter DNA methylation alterations in GBM tumors, and identified and characterized a unique subtype of human glioma tumors that are highly associated with several different and distinct clinical and molecular features.