Figure 1.

Polymer model of loop extrusion with barrier elements recapitulates features of interphase chromosome folding (see also Supplemental Movie 1). (A) Illustrations of the four key parameters governing the dynamics of interphase loop extrusion: LEF velocity, LEF lifetime, LEF separation, and barrier strength. Characterizing how changes to these parameters affect Hi-C maps in silico allows us to make experimental predictions for perturbations. (B) To compare our models with Hi-C experiments, we generate ensembles of conformations for each set of parameters, and then compute average contact maps. To compare with imaging experiments, we can calculate other observables (e.g., pairwise distance between loci). (C) Interphase Hi-C data from mouse neural progenitor cells (Bonev et al. 2017), plotted with HiGlass (Kerpedjiev et al. 2017), annotated with features that can emerge via loop extrusion in blue (i–iv). Arc diagrams depict how stochastic configurations of LEF-mediated loops in distinct nuclei can lead to the population-averaged features. Chromatin loops directly held by LEFs are depicted with yellow arcs, whereas dashed gray arcs depict “transitive loops” from sets of adjacent LEFs. (i) Insulation, observed as squares along the diagonal of Hi-C maps (i.e., TADs), arises when extrusion barriers halt LEF translocation. LEFs then facilitate additional contacts within TADs, but not between TADs. (ii) Flames (or tracks), observed as straight lines often emerging from the borders of TADs, arise when LEFs become halted on one side at a barrier while continuing to extrude from the other side (referred to as “lines” in Fudenberg et al. 2016). (iii) Peaks of enriched contact frequency often appear at the corners of TADs and also often coincide with intersection points of flames. These peaks emerge as a result of LEFs being halted on both sides by extrusion barriers. (iv) Peak grids can emerge either when internal boundaries are skipped or via transitive sets of LEF-mediated loops.