|Ph.D Student||Pollak Yaroslav|
|Subject||Studying Looping Based Transcriptional Regulation Using|
Synthetic Biology and Statistical Mechanics
|Department||Department of Nanoscience and Nanotechnology||Supervisor||Professor Roee Amit|
|Full Thesis text|
Enhancer-regulated genes are abundant in all organisms. While it is generally agreed that DNA looping plays a central role in enhancer regulatory output, most enhancer structural features, such as the importance of having several binding sites for a given TF and their particular arrangements, remain poorly understood. Consequently, computationally predicting an enhancer regulatory output function has proven to be difficult. To address this problem, we devised a self-avoiding wormlike chain model for enhancers. At its basis, the model relies on numerically generating statistical ensembles of thick chains decorated with adherent protrusions, characterized by a particular size, position along the chain, and relative orientation. We employed a sequential importance sampling algorithm to ensure that these ensembles faithfully sample the underlying configurational distribution. With these ensembles, we studied the ability of the polymer chain to loop with and without an adherent protrusion, consequently simulating the effect of DNA-bound TFs on regulatory output. In the first paper, we studied the looping probability of bare DNA molecules. We confirmed the scaling-theory prediction of the looping probability power-law constant of -1.92 in the entropic regime. We also showed that in the elastic regime the end-to-end chain distribution is highly anisotropic. This anisotropy, combined with the excluded volume constraints, increase in the J-factor of the self-avoiding worm-like chain by half an order of magnitude relative to its non-self-avoiding counterpart, partially explaining the anomalous results of recent cyclization experiments, in which short dsDNA molecules exhibited an increased propensity to loop. In the second paper, we studied the effect of DNA-bound TFs on regulatory output of promoter-proximal enhancers. Based on the numerical results we constructed libraries of synthetic bacterial enhancers to test the predictions of the numerical model, and characterized them in in vivo experiments in bacteria. Our combined findings suggest that protein excluded-volume can account for both up-regulating and down-regulating effects in bacterial enhancers. We found that the nature and magnitude of the regulatory effect are influenced by the size of the TF, the number of bound TFs and their relative arrangement within the enhancer. We also showed that the magnitude of the effect is maximal for TFs bound at the center of the looping segment. The nature and magnitude of the effect are highly sensitive to the location of the TF binding site and exhibit an oscillating pattern whose period matches the dsDNA helical repeat. Additionally bound TFs can augment or diminish the effect, depending on their relative orientation to the other TFs. In the third paper, we focused on regulation for promoter-distal enhancers, more typical of eukaryotes. In this case, the elastic regulatory effects of the second paper are negligible. Using our model, we showed that the excluded volumes of DNA and a TF bound within one Kuhn length either upstream or downstream of one of the loop termini can block the “line-of-sight” of the other terminus, generating an eclipse-like effect, reducing the looping probability. Unlike the previously reported elastic effects, this eclipse-like effect is independent of looping-segment length for sufficiently long looping segments.