The epigenome is the dynamic interface between our changing environment and the static genome, and understanding it is a goal of immense importance to human health. The Costello laboratory is interested in the interaction between genomic and epigenomic mechanisms in gene expression in normal brain and brain cancer. DNA methylation is one of the most important epigenetic mechanisms through its influence on gene expression and chromosome stability.
Genetic and epigenetic mechanisms contribute to the development of human brain tumors, yet the typical analysis of tumors is focused on only one or the other mechanism. This approach has led to a biased, primarily genetic view of human tumorigenesis. Epigenetic alterations such as aberrant DNA methylation are sufficient to induce tumor formation, and can modify the incidence and tumor type in genetic models of cancer. These initial studies raise important questions which our research addresses: understanding the degree to which genetic and epigenetic pathways cooperate in human brain tumorigenesis, the identity of the specific cooperating genes and how they interact functionally to determine the differing biological and clinical course of tumors. To facilitate these studies we have also developed new methods for whole epigenome analysis using next generation sequencing.
Current Research Projects
Next-generation sequencing of the GBM stem cell epigenome
A significant subset of CpG islands, including the MGMT promoter, exhibit intra-tumoral heterogeneity of aberrant methylation. We are addressing the hypothesis that the intra-tumoral heterogeneity in methylation may in part reflect differences between the CD133+ cancer stem cell subpopulation and the CD133- cell subpopulations within each tumor, and could underlie the tumorigenic potential of the CD133+ subpopulation. As a first step to addressing this hypothesis, we will determine if intra-tumoral heterogeneity of methylation corresponds in part to CD133+ and CD133- GBM cell subpopulations. We have devised a new method for identification of every significantly methylated CpG island in a DNA sample, using immunoprecipitation of methylated DNA (MeDIP) followed by next-generation sequencing (MeDIP-seq).
We are applying MeDIP-seq to unsorted primary GBMs and their CD133 FACS-sorted subpopulations. The results of these proposed experiments will substantially increase our understanding of the relative contribution of epigenetic mechanisms to GBM growth, particularly in the tumorigenic/cancer stem cell compartment, identify new methylation markers to improve clinical application of the MGMT methylation assay, and stimulate further mechanistic studies to address functional consequences of disrupting aberrant methylation on the tumorigenic potential of cancer stem cells. More
Integrated Epigenome Maps of Human Embryonic and Adult Cells
We are working cooperatively with other Mapping Centers and the Data Coordination Center (EDACC) funded by this Roadmap mechanism to comprehensively map epigenomes of select human cells with significant relevance to complex human disease. Our group, consisting of scientists at UCSF, UC Davis, UCSC and the British Columbia Genome Sciences Centre has the broad expertise that this project requires. We are focusing on cells relevant to human health and complex disease including cells from the blood, brain, breast and human embryonic stem cells (aim 1). We will incorporate high quality, homogeneous cells from males and females, and two predominant racial groups, and biological replicates of each cell type.
Production of comprehensive maps will include 6 histone modifications selected for their opposing roles in regulating active and inactive chromatin (aim 2), DNA methylation (aim 3) and miRNA and gene expression (aim 4). This epigenetic data, along with genetic and expression data will be integrated using advanced informatics (aim 5) to address fundamental roles of epigenetics in differentiation, maintenance of cell-type identity and gene expression.
Our cell and data production pipeline will incorporate verification and data validation with independent methods, and will operate under a model motivated by increased data production and decrease cost. We envision that our group in conjunction with the other REMC teams, the EDACC, ENCODE, future EHHD (Epigenetics of Human Health and Disease) centers and the NIH Roadmap program will develop methods, tools and reference epigenome maps for the research community that will make the promise of epigenetics in understand and treating human complex disease a reality. Our reference epigenomes will enable new disciplines including human population epigenetics, comparative epigenomics, neuroepigenetics, and therapeutic epigenetics for tissue regeneration and reversal of disease. More
Reversal of HOXA9 Oncogene Activation by Inhibition of PI3K: Epigenetic Mechanism and Prognostic Significance to Human Glioblastomas
HOX genes encode transcriptional regulators with roles in embryonic development and tumorigenesis. In a subset of glioblastomas (GBMs), HOX gene clusters are aberrantly activated within confined chromosomal domains. Transcriptional activation of the HOXA cluster is reversible by a phosphoinositide 3-kinase (PI3K) inhibitor, but not an mTOR inhibitor, through an epigenetic mechanism involving histone H3K27 trimethylation. Aberrant expression of HOXA9 is independently predictive of shorter overall and progression-free survival in two independent sets of GBM patients, and improved survival prediction by MGMT promoter methylation. Functional studies of HOXA9, a potent oncogene in leukemias, show its capacity to decrease apoptosis, increase cellular proliferation and increase GBM cells resistance to therapy.
Role for intragenic DNA methylation in gene expression
Although DNA methylation is commonly found in gene bodies, its biological significance is unclear. Using the Autism gene SHANK3 as a model, we show that intragenic methylation is tissue- and cell type-specific, but also differs significantly within a single cell type from distinct brain regions. The intragenic methylation is also gene region–specific and evolutionarily conserved. Because DNA methylation is known to influence the activity of 5’-promoter sequences, we searched for genetic and epigenetic evidence of promoter activity embedded within SHANK3. Using this integrated and cross-species approach, we identified two intragenic regions that have promoter activity that is blocked by methylation in vitro, that are differentially methylated in vivo, and which drive transcription of novel and likely protein-coding SHANK3 transcripts. The constitutively unmethylated 5’-CpG island promoter and the differentially methylated intragenic promoters of SHANK3 are coordinately regulated, but by distinct epigenetic mechanisms. In 15/22 additional genes (68.2%) having tissue-specific intragenic methylation, the differentially methylated sites coincided precisely with multiple features of promoters. Altogether, these results support an evolutionarily conserved, cell context-specific role for intragenic DNA methylation in regulating the activity of alternate promoters. We suggest that intragenic DNA methylation has a greater role in gene expression than methylation of 5’-CpG island promoters. More