Capture Hi-C to identify regulatory variants and target genes influencing breast cancer risk

Abstract

Genome wide association studies have discovered approximately 200 breast cancer risk single nucleotide polymorphisms, most of which map to non-protein-coding regions. To understand the mechanisms influencing disease risk, identification of the genes, non-coding RNAs and casual variants mediating these associations is required. One of the methods that allows functional characterisation of cancer risk loci is Capture Hi-C (CHi-C). CHi-C provides a high-throughput, high-resolution approach for studying physical interactions between long-range regulatory elements and their targets and has previously been used to identify putative target genes and to prioritise credible variants at a subset of risk loci. To date, however, CHi-C data have only been generated in breast cancer and immortalised 'normal' breast epithelial cell lines. Additionally, most studies have used HindIII digested libraries, which result in an average resolution of 10 kb. the aims of this project were to: 1 - Generate region CHi-C libraries in breast epithelial and fibroblast cell lines using three different protocols to identify and optimise the most suitable method for library generation in primary cells; 2 - Generate higher resolution region CHi-C data in two types of primary breast cells (luminal epithelial cells and fibroblasts) to identify regulatory variants and target genes influencing breast cancer risk; 3 - Compare cell line data to the primary cell data to evaluate the usefulness of cell lines as model systems; 4 - Generate cell line promoter CHi-C data to validate region CHi-C findings and to identify 'indirect' interactions; 5 - Using data generated in primary breast fibroblasts, determine whether a subset of breast cancer loci may act via the stroma to influence the risk.