Chromatin organization is inextricably linked to its dynamics. The loop extrusion factor (LEF) model provides a framework for how topologically associating domains (TADs) arise: cohesin or condensin extrude DNA loops, until they encounter boundary elements, namely CTCF. However, a characteristic subdiffusive behavior of MSD is observed in fission yeast on the seconds timescale and experiments show that cohesin or condensin largely constrains chromatin mobility. Such finding is inconsistent with prior LEF model.
I develop a new LEF model in which LEF loading depends on the underlying architecture of transcriptional units and chromatin remodeling is identified as the essential ATP-dependent activity that drives chromosome motion. I demonstrate that this model predicts TADs, including TAD boundaries lacking CTCF binding sites, comparably well to prior CTCF-dependent LEF models for the mouse genome, and also successfully predicts TADs in the CTCF-lacking model, S. pombe. Lastly, I perform polymer dynamic simulations and show that the DNA-looping by cohesin and condensin largely constrain chromatin mobility, which is in agreement with experiments.