Integrative epigenomic analysis of high and low Glioma-CpG island Methylator Phenotype (G-CIMP): characterization and methods development

Abstract

Gliomas, including glioblastoma which is the most lethal intracranial tumor, exhibit a very complex development that can lead to treatment resistance and recurrences. This heterogeneity can interfere in the diagnosis and, therefore, in the therapeutic approach. Currently, gliomas are classified based on histology and then graded on a scale of I to IV, depending on the degree of malignancy. However, the histological features are generally inferred by pathologists' observation of a biological sample collected from the tumor, which can cause diagnostic discrepancies. This scenario has been changing for the past years with constant updates on the molecular profiling of gliomas and the advance of technologies that allow the enforcement of these results in the clinic. The most recent study about the genomic and epigenomic landscape of gliomas showed a distinct subgroup of IDH mutant samples with poor survival. This subtype (GCIMP-low) is characterized by a loss of DNA methylation, based on array data (Illumina HumanMethylation27 and HumanMethylation450), when compared to other IDH mutant samples also without codeletion of chromosomes 1p and 19q. (GCIMP-high). We propose to analyze the changes across the entire epigenome between GCIMP-low and GCIMP-high samples in order to identify candidate deregulated regions, thereby understanding progression among IDH mutant samples. The glioma samples will be acquired and processed using tools developed by our group, such as TCGAbiolinks (an R/Bioconductor package). Complementary epigenomic data including ChIP-seq and known regulatory elements from external databases (TCGA, ENCODE and Roadmap) in normal brain and non-brain and in tumor samples will be obtained by the development of a new tool called biOMICs+ which will allow data integration from different sources. The regulatory networks that we propose to identify in this project will be annotated, based on DNA methylation and gene expression profiles of glioma samples, and characterized to expand the knowledge of molecular basis of tumors, specially gliomas, with the aim of understanding mechanisms related to initiation and progression of brain cancer. (AU)