A systematic investigation of the role of simple repeats in transcription factor binding in vivo.

Repeated genomic sequences, including simple repeats have been long regarded as genetic noise, but have recently emerged as potential regulators of gene expression in eukaryotic organisms. While recent evidence suggests that short tandem repeats (STRs), transposable elements, and other classes of repeative DNA influence transcription factor (TF) binding to DNA, the precise physical mechanisms underlying this interaction remain poorly understood.1,2 In the case of STRs, theoretical models propose that sequence symmetries contribute to nonspecific TF binding, with homopolymer sequences exerting the strongest effects.3 Similarly, these repetitive elements may modulate transcription by altering histone protein affinity for DNA, consequently affecting nucleosome occupancy.4 Our research was motivated by Horton et al.'s in vitro investigations demonstrating the stronger MAX transcription factor binding to DNA motifs flanked by short tandem repeats compared to random flanking sequences. Their findings suggested that simple repeats’ function potentially provide additional binding sites or alter local DNA structure to improve TF localization.5 Importantly, the results of Horton et al. suggest that the contribution of short tandem repeats to TF binding is greater than what would be expected simply from their contribution to the affinity of the sequence, as measured through positional weight matrices. Having gained these insights, a critical gap is still persisting in our knowledge regarding whether STRs alter TF occupancies in living cells. Indeed, as noted above, Horton et al’s results were obtained in vitro, while the mechanisms of TF binding in vivo are made more complex by the dynamics and cell type dependence of chromatin accessibility. This thesis addresses this gap by systematically investigating the in vivo role of STRs played in TF binding. We use a large database of ChIP-sequencing data and a computational approach based on multiple regression to systematically address the question of how repeated genomic sequences, and short tandem repeats in particular, affect TF binding, either by providing direct binding sites, or by affecting chromatin accessibility.