Understanding the recruitment and regulatory functions of CTCF during development

CCCTC-binding factor (CTCF) is a conserved transcription factor with thousands of binding sites in mammalian genomes that are defined by a 20 base-pair DNA consensus motif. CTCF plays many regulatory roles, having activator, repressor, and enhancer blocker activity. What determines CTCF’s functional effect at a given locus and the factors that contribute to regulating CTCF binding are poorly understood. To investigate if differences in binding site sequence play a role in CTCF transcriptional function and recruitment, we characterized changes in genome-wide CTCF binding and gene expression during differentiation of mouse embryonic stem cells (mESCs). We found that developmentally regulated CTCF binding occurs preferentially at sites with low similarity to the consensus (LowOc sites) in contrast to sites with high similarity to the consensus (HighOc sites). By measuring the affinity of CTCF for selected sites, we show that sites lost during differentiation are enriched in motifs associated with weaker CTCF binding in vitro. By comparing changes in CTCF binding with changes in gene expression during differentiation, we also show that LowOc and HighOc sites may be associated with distinct regulatory functions. Our results suggest that the regulatory control of CTCF is dependent in part on specific motifs within its binding site. CTCF is additionally implicated in the regulation of imprinted genes, which are expressed in a parent-of-origin specific manner. Grb10 is an imprinted gene which switches expression from a maternal-specific isoform to a paternal-specific isoform during neuronal commitment. CTCF binds on the paternal allele of the Grb10 imprint control region (ICR), a regulatory element critical for Grb10 imprinted expression. To assess CTCF’s role in the switch in Grb10 expression, we have adapted a differentiation system to generate motor neurons in vitro from mESCs. In differentiated motor neurons, the maternal isoform of Grb10 (Grb10-1A) is repressed, while the paternal Grb10 isoform is activated (Grb10-1B), recapitulating the switch observed in vivo. CTCF knockdown in mESCs suggest that CTCF may act as a repressor of Grb10-1A on the paternal allele. This optimized model system will enable further dissection of CTCF’s role at Grb10