Nucleosome repositioning in glioblastoma

Glioblastoma multiforme (GBM) is the most malignant type of brain cancer in adults, with highly aggressive behaviour and low patient survival. The glial cells that are responsible for forming GBM show abnormalities in gene expression that can be caused by genetic or epigenetic causes. In particular, the positioning of nucleosomes along the genome determines the accessibility of DNA to regulatory proteins and therefore modulates gene expression. In this project I investigate nucleosome positioning in relation to other features, such as transcription factor (TF) binding, ALU repeats and histone modification changes. I show that the relationship between TF binding and nucleosome positioning in brain tissues, as well as cell-free DNA (cfDNA) from peripheral blood, is a highly promising system to investigate the deregulation of chromatin in cancer cells and diagnose GBM patients. The computational analysis performed here is based on experimental MNase-seq data in paired normal and tumour brain tissues from GBM patients as well as cfDNA from GBM patients and healthy individuals. I have determined GBM-specific changes in nucleosome occupancy profiles around binding sites of 20 glioblastoma-related TFs. Major changes of nucleosome positioning were found around binding sites of CTCF, RBPJ, MYC, KLF9 and JMJD6, which I propose for the role of new liquid biopsy markers. Moreover, I studied differences in nucleosome occupancy at different classes of genomic features such as ALU repeats, which showed unexpected resemblance to transcription start sites. Finally, I compiled a dataset of genomic regions that underwent GBM-specific nucleosome repositioning and showed that Principal Component Analysis (PCA) based on nucleosome occupancy values in these regions could be used for patient diagnosis. The results of my work have implications for future development of liquid biopsy assays for patient stratification based on nucleosomal DNA.