The impact of changes in chromatin architecture on cancer phenotypes and tumor progression

New York University School Of Medicine, New York NY

Investigators

Linked publications, trials & patents

Abstract

OVERALL SECTION: SUMMARY Chromatin is organized into boundary delimited territories and sub-territories, within which there is a high level of coordination of transcriptional states. The larger domains, defined as compartments A and B, are respectively comprised of active and inactive chromatin, which can be further subdivided into highly self- interacting topologically associated domains (TADs) separated from each other by ‘insulating boundaries’. TADs form via ‘loop-extrusion’, in which cohesin rings extrude DNA until they encounter two convergently oriented CTCF binding sites, which form the base of a loop. Insulating boundaries limit inter-TAD interactions, to ensure that enhancers predominantly contact promoters within the same domain. Cohesin-mediated loops bring distal cis regulatory elements (CREs) into physical contact with the promoters of genes they control. CRE contacts commonly form between a single CRE-gene pair. However, we and others have found that gene regulation can also occur in ‘CRE hubs’, regions that encompass multiple CREs and their target genes. CRE hubs form interaction networks of highly connected regulatory elements (~100s kb wide). In contrast to individual CRE-gene pairs, CRE hubs are likely to have stronger functional consequences on cell fitness, cell state and pathological outcome. Because CRE hubs are rare compared to non-hub CREs, they are less likely to be deregulated by stochastic mutations, hence they confer robustness. At the same time, they represent weak points of the regulatory network since genes within CRE hubs can be deregulated by a single ‘targeted’ mutation. Importantly, genomic structural variants (SVs) can potentially generate novel CRE hubs with oncogenic functions. In addition, mutant CTCF and transcription factors or chromatin modifiers can also alter the composition of CRE hubs. Using novel long-read sequencing based 3D architectural methods, members of this P01 identified chromunities, which connect distinct CRE hubs via bridging enhancers thereby generating large interconnected regulatory modules. Chromunities are also susceptible to gain and loss of function lesions as postulated for CRE hubs. Based on this rationale, we hypothesize that CRE hubs and chromunities provide prime candidates for identifying critical genomic regions that contribute to different developmental and cancer specific cell states. We will explore these models, and their relevance to developmental states and hematological malignancies, in the following Aims: 1. Determine whether and how mutations of CTCF and linker histone 1 disrupt or distort CRE hubs and chromunities, leading to the formation of oncogenic transcriptional programs. 2. Determine whether complex genomic structural variants form novel aberrant CRE hubs or chromunities that drive cancer phenotypes. 3. Generate CRE hub maps encompassing T-ALL, B-ALL, T cell lymphoma and B cell lymphoma to identify potential vulnerabilities specific to, or shared among each cancer type. 4. Use CRISPR tools and mouse genomics to “dissect” CRE hubs that encompass target genes with therapeutic potential. 1