Gene regulation in three dimensions

Abstract

Gene regulation has traditionally mainly been viewed as a 1D and possibly 2D process. In the 1D view the genome is seen as a linear string of nucleotides, where one or more transcription factors (TFs) bind to transcription factor binding sites (TFBSs), and thereby regulate the expression of genes that are nearby in the linear genomic sequence. The 2D process can be described with DNA looping where proteins bind to distal binding sites and bring them in close proximity of the transcription start site (TSS). However, in reality such interactions take place in 3D space meaning multiple interactions, possibly also between separate chromosomes. The binding of TFs to genomic DNA is experimentally studied using mainly the ChIP-Seq protocol. In some cases, the motif for transcription factor binding is found only in a subset of peaks. In this project we looked at the phenomenon of motifless (ML) binding in Vitamin D Receptor (VDR) ChIP-Seq experiments where interestingly as much as 50% of the identified peaks seem to be ML. These motifless binding sites may be caused by 3D interactions or DNA looping where the DNA strand folds back and interacts with itself or distally with another chromosome. Here we use statistics and computer software to measure localization of ML sites in 3D as well as 1D space. This is completed using previously generated genomic annotation data from ENCODE and other datasets to measure differences between ML peaks and regions containing known TF binding sites.