Abstract Three-dimensional genome organization has emerged as a critical component for the proper regulation of gene expression. Recent years have seen a rapid expansion of the understanding of many of the basic features that define how genomes are organized in space inside of cells, including the identification of features such as A/B compartments, Topologically Associated Domains, and chromatin loops. Furthermore, there is evidence that mutations that alter 3D genome organization can contribute to human disease. This is most evident for a class of mutations known as structural variants, which includes translocations, inversions, tandem duplications, and deletions. When these mutations disrupt sequence features that are critical for 3D genome structure, such as the boundaries between Topologically Associating Domains, this can lead to enhancer-promoter rewiring, changes in gene expression, and phenotypic consequences. Such effects have been observed both in the context of germline structural variants that contribute to syndromic disorders of development as well as somatic structural variants that can lead to cancer. While it has become clear that structural variants can alter 3D genome organization and gene expression, more recent studies that comprehensively examined structural variants and gene expression indicate their relationship is considerably more complex. Specifically, in only a minority of instances do structural variants lead to changes in expression of neighboring genes. Therefore, why structural variants can have dramatic consequences on 3D genome structure and gene expression in some contexts but not others is currently unclear. This proposal will investigate the relationship between structural variants, 3D genome organization, and gene expression in cancer genomes with the goal of understanding where and when structural variants will actually lead to changes in gene expression that may contribute to oncogenesis. Specific aim 1 will test whether only specific sets genes are sensitive to structural variant induced changes in enhancer-promoter communication by examining changes in 3D genome structure and gene expression in haplotype resolved human tumor samples. Specific aim 2 will use CRISPR/Cas9 genome engineering to evaluate the effects of structural variant partner regions on induction of oncogene expression. Specific aim 3 will assess the role of intra-tumor heterogeneity on the effects of structural variants on 3D genome structure by using novel multi-omic methods for profiling DNA methylation and 3D genome structure simultaneously within single cells derived from patient tumor samples. Successful completion of these aims will result in a deeper understanding of the relationship between structural variation, 3D genome organization, and gene regulation in the context of cancer genomes. In the long term, this will facilitate the use of information derived from structural variants and 3D genome structure on determining patient prognosis and on identifying novel therapeutic targets in cancer.

Narrative Cancer can be caused by a certain class of DNA mutations, called “structural variants”, that break apart DNA, then re-join it incorrectly. This can cause DNA to fold improperly and can contribute to genes being turned on inappropriately; however, when and where this leads to altered gene activation is not clear. This proposal will investigate when, where, and how structural variants cause DNA misfolding and inappropriate gene activation in cancer cells.