Emerging Evidence of Chromosome Folding by Loop Extrusion

Geoffrey Fudenberg*, Nezar Abdennur*, Maxim Imakaev, Anton Goloborodko, Leonid Mirny


Cold Spring Harb Symp Quant Biol: Chromosome Segregation & Structure. 2017; 82:45-55. Epub 2018 May 4. doi:10.1101/sqb.2017.82.034710

[bioRxiv doi:10.1101/264648]

Chromosome organization poses a remarkable physical problem with many biological consequences: how can molecular interactions between proteins at the nanometer scale organize micron-long chromatinized DNA molecules, insulating or facilitating interactions between specific genomic elements? The mechanism of active loop extrusion holds great promise for explaining interphase and mitotic chromosome folding, yet remains difficult to assay directly. We discuss predictions from our polymer models of loop extrusion with barrier elements, and review recent experimental studies that provide strong support for loop extrusion, focusing on perturbations to CTCF and cohesin assayed via Hi-C in interphase. Finally, we discuss a likely molecular mechanism of loop extrusion by SMC complexes.

HiGlass Displays

Experimental phenotypes are consistent with predictions from loop extrusion simulations

Top row: unperturbed experimental Hi-C maps, replotted from indicated studies.

Bottom Row: Hi-C maps for indicated perturbations.

Left column: ​Schwarzer et al. used tissue-specific CRE-inducible gene deletion in mouse liver cells to deplete
Nipbl (Schwarzer et al. 2017).

Middle column: ​ Nora et al. used an auxin-inducible degron (AID) system to deplete CTCF in mESCs (Nora et al. 2017).

Right column: ​Haarhuis et al. deleted Wapl in the Hap1 haploid human cell line, via CRISPR (Haarhuis et al. 2017).

Experimental phenotypes are consistent with predictions from loop extrusion simulations. Top row: unperturbed experimental Hi-C maps, replotted from indicated studies

Bottom Row: Hi-C maps for indicated perturbations.

Left column: ​Rao et al. used AID to deplete Scc1/Rad21 in HCT116 cells. Note that these cells are not synchronized (Rao et al. 2017).

Middle and Right columns: ​ Wutz et al. used AID to deplete Scc1/Rad21 (bottom center) and CTCF (bottom right) in synchronized HeLa cells. They also used RNAi to deplete Wapl (top right) and Pds5A/B (not shown) (Wutz et al. 2017).

TAD Anatomy

Interphase Hi-C data from mouse ES and neural progenitor cells (Bonev et al. 2017) illustrating contact patterns that collectively describe mammalian TADs: including contact enrichment/insulation, “flames” (a.k.a. lines, stripes or tracks), peaks and peak grids. Note that there exist similar features associated with genome compartmentalization also visible as squares along the diagonal of a Hi-C map, but differ in that they further extend into a checkered grid in cis and in trans. While these latter patterns overlap with and are very often conflated with TADs in the literature, they have been shown to arise from the action of a completely independent and competing biological mechanism from that which produces the features highlighted here.