Identification and Characterization of Cis-Regulatory Elements in the Human Genome

Abstract

Maintaining precise spatiotemporal control of gene expression patterns is essential for the proper functioning of cells, tissues, and organisms. In eukaryotic cells, this control is established through the use of multiple systems, which regulate expression at the levels of chromatin, DNA, RNA and proteins. At the DNA level, regulation is controlled by cis-regulatory elements (CREs): modular non-coding sequences with varying functions mediated through the binding of transcription factors. Variation within these non-coding sequences is increasingly understood to contribute to human phenotype and disease, however mapping and characterizing CREs is challenging, as non-coding sequences comprise 98.8% of the human genome. For this reason, in addition to their flexibility and scalability, episomal reporter assays have been, and continue to be, the primary tool used to test for CRE function in non-coding sequences. As our appreciation of the complexity and interconnectedness of cis-regulatory systems increases, so does the complexity of the assays designed to interrogate CRE function. However, with increasing complexity comes the potential for confounding effects within assay systems. In Chapter 1 of this thesis, I review the roles of different CRE classes in cellular regulation, the types of assays used to characterize CREs, and address the ways in which the transcription factor, CRE, and chromatin layers of regulation interconnect. I discuss how models are needed that account for the interactions of elements within and across regulatory systems, and for the role of silencers in these systems, in a genomic context. In Chapter 2, I discuss several considerations for plasmid-based reporter assay design in light of this increasing diversity and complexity. I provide supporting data for the impact of each component, discuss how it can impact interpretation, provide models for improving design, and demonstrate the utility of plasmid-based systems for modeling CRE mechanisms. Due to the potential functionality of each component in a complex plasmid system, tools that support full, rather than partial plasmid sequence validation are needed. I address this in Chapter 3, where I present OnRamp, a combined protocol and analysis toolset that leverages the long-read nature of the nanopore sequencing platform to facilitate rapid, affordable, and accessible multiplexed full-plasmid sequencing. Finally, in Chapter 4, I use a modified regulatory assay panel to characterize enhancer, silencer, and enhancer blocker activity across a single regulatory topologically associating domain (TAD) of the human genome, containing the genes PRDM1 (crucial to B-, NK- and T-lymphocyte differentiation) and ATG5 (an essential autophagy-related gene). Using assay data and previously generated high-throughput datasets, I generate a model of regulation in this region which incorporates chromatin, CRE, and transcription-factor level systems to account for the differential regulation of the two genes both within the TAD and across two cancer cell lines - K562 (myelogenous leukemia) and HepG2 (hepatocellular carcinoma). Together, this work contributes to the improved design and fidelity of plasmid-based reporter assays for the study of cis-regulatory elements and generates a functional model for regulatory dynamics in a previously relatively uncharacterized region of the human genome containing genes important for basic cellular and immune function.