Engineering CTCF DNA-Binding Specificity to Alter Gene Expression and Genome Topology in Human Cells

Abstract

The 3D organization of the eukaryotic genome is an integral part of cell homeostasis and differentiation. Genome organization is a multi-tiered system regulated in large part by a host of transcription factors and chromatin interacting proteins, such as CCCTC-binding factor (CTCF). CTCF is a ubiquitous DNA-binding protein involved in higher-order topological organization of the genome via establishment of topologically associated domains (TADs). CTCF is recruited to the genome by 40 bp CTCF binding sites (CBSs) that contain a highly conserved 15bp motif. CBSs are frequently mutated in cancer and developmental diseases leading to loss of CTCF binding and subsequent gene misregulation, but studying the mechanistic consequences of CTCF function is complicated by its broad functionality within eukaryotic cells. In addition, CTCF has been implicated in gene regulation through the formation of promoter-enhancer loops, but it is not clear if cohesin or RNA is the cofactor in this process. To address this limitation, I sought to define and alter the DNA-binding determinants of CTCF so as to facilitate structure/function studies of this important regulator. Using a bacterial-two-hybrid (B2H) reporter system, I first identified the nucleotides in the CBS that are essential for CTCF binding. I used this knowledge to generate a series of variant CBSs (vCBS) that are no longer bound efficiently by wild-type CTCF. Leveraging the B2H as a selection system, I evolved CTCF variants with altered binding specificities for these vCBSs. Utilizing the CTCF-regulated proto-oncogene MYC as an endogenous human gene reporter, I demonstrated that these engineered CTCF variants could reproduce the normal biological role of CTCF in cellula and could be used to define the functional consequences of mutating CTCF domains on expression of this gene, providing evidence that RNA, not cohesin, is the facilitating cofactor of establishing the CTCF-mediated promoter-enhancer loop. I have developed a system to study the mechanistic requirements for CTCF-mediated gene expression without the confounding pleiotropic effects, allowing for site specific analysis of CTCF mediated gene expression. This work could be applied to creating a toolbox of variant transcription factors with novel DNA-recognition profiles for application in epigenetic engineering.