Glioblastoma multiforme (GBM) is almost universally fatal with median survival of 12-18 months from the time of initial diagnosis and new therapeutic approaches are needed. Epigenetics and genetics co-operate at all stages of carcinogenesis, including in brain tumors. Epigenetic mechanisms revolve around the nucleosome, a structure comprised of a core of histone proteins and DNA, which is wrapped around the histone core. A variety of epigenetic mechanisms work together to generate chromatin states that facilitate or inhibit gene activation. These mechanisms include DNA methylation (at CpG sites), histone variants and modifications and nucleosome positioning. Genome wide studies of chromatin states have focused on individual aspects of epigenetic regulation and as a result an integrated map of epigenetic regulatory mechanisms, which together determine transcriptional state, has not been completed. Using a novel technique that I have been developing (GNOMe-seq), I will generate genome wide maps of DNA methylation and nucleosome positioning within the same DNA strand of normal human neural progenitor cells and GBM cells. This will identify aberrantly silenced genes, which require DNA demethylation in addition to chromatin remodeling for reactivation. This data will be combined with localization profiles of the H2A variant, H2A.Z. In addition to being present at important genomic regulatory regions like enhancers and insulators, H2A.Z is localized to active and poised gene promoters. Combining DNA methylation, nucleosome occupancy and H2A.Z localization data will identify which genomic loci are active, poised and silenced, thereby providing critical information detailing aberrant chromatin structures present in GBM which has important implications for epigenetic therapy. Armed with this information, in the independent phase, I will treat GBM cells with the DNA demethylating agent 5-Aza-CdR and measure the ability of demethylation to restore normal gene expression patterns and resolve aberrant chromatin structures. In addition to measuring DNA methylation, nucleosome occupancy and H2A.Z localization, I will assess tumorigenic potential by measuring neurosphere formation. During the mentored phase of the award, I will generate the genome-wide maps of DNA methylation, nucleosome occupancy and H2A.Z localization, which will characterize the epigenetic landscape of normal human neural progenitors and GBM cells. I will also learn new bioinformatics techniques necessary for analyzing genome wide epigenomic data and participate in career development opportunities that will help in obtaining and succeeding as an independent position at an academic research institution.

Public Health Relevance

GBM tumors are difficult to treat, and little is known about the epigenetic landscape of these cells. Understanding the epigenetic alterations in these cells and how they respond to DNA demethylation treatment will have important implications for future therapies. Here I will investigate several levels of epigenetic regulation and how they are altered in brain tumors.

National Institute of Health (NIH)
National Cancer Institute (NCI)
Career Transition Award (K99)
Project #
Application #
Study Section
Subcommittee G - Education (NCI)
Program Officer
Schmidt, Michael K
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Southern California
Schools of Medicine
Los Angeles
United States
Zip Code
Kelly, Theresa K; Liu, Yaping; Lay, Fides D et al. (2012) Genome-wide mapping of nucleosome positioning and DNA methylation within individual DNA molecules. Genome Res 22:2497-506