Dynamics of CTCF- and cohesin-mediated chromatin looping revealed by live-cell imaging

Abstract

Animal genomes are folded into loops and topologically associating domains (TADs) by CTCF and loop-extruding cohesins, but the live dynamics of loop formation and stability remain unknown. Here, we directly visualized chromatin looping at the Fbn2 TAD in mouse embryonic stem cells using super-resolution live-cell imaging and quantified looping dynamics by Bayesian inference. Unexpectedly, the Fbn2 loop was both rare and dynamic, with a looped fraction of approximately 3 to 6.5% and a median loop lifetime of approximately 10 to 30 minutes. Our results establish that the Fbn2 TAD is highly dynamic, and about 92% of the time, cohesin-extruded loops exist within the TAD without bridging both CTCF boundaries. This suggests that single CTCF boundaries, rather than the fully CTCF-CTCF looped state, may be the primary regulators of functional interactions.

Fig. 1.. Endogenous labeling and tracking of the Fbn2 loop with super-resolution live cell imaging.

(A) Fluorescent labeling of Fbn2 loop anchors does not perturb the Fbn2 TAD. Micro-C contact map comparing the parental untagged (C59, top left) and tagged (C36, bottom right) cell lines. Red triangles: CTCF motifs with orientation. C36 ChIP-seq shows CTCF (GSM3508478) and cohesin (SMC1A; GSM3508477) binding as compared to Input (GSM3508475). Fbn2 is not expressed (RNA-seq GSE123636); annotation: GRCm38). Genome coordinates: mm10. (B) Overview of tagging and readout using 3D distance. (C) Overview of the genome-edited cell lines (left) and a representative maximum intensity projection (MIP) of a cell nucleus showing two pairs of “dots” (right). (D) Representative 3D trajectory over time of a dot pair. MIPs of the 3D voxels centered on the mScarlet dot (top) and 3D distances between dots (bottom) are shown. (E) 3D distance probability density functions of dot pairs (n=32,171; n=46,163; n=13,566 distance measurements for C27, C36, C65 respectively) (F) Localization error corrected 2-point mean squared displacement (MSD) plots (n=358; n=491; n=147 trajectories in C27; C36; C65 respectively).

Fig. 2.. Degradation of CTCF, cohesin, and WAPL reveal their role in loop extrusion and looping-mediated chromosome compaction.

(A) Micro-C data for the AID-tagged clones for RAD21 (left), CTCF (middle), and WAPL (right), showing control data (no IAA treatment; top half) and protein degradation data (3 hours post IAA; bottom half) with schematics illustrating the expected effect. (B) Representative trajectories with (colored lines) or without IAA treatment (gray lines) for each AID-tagged clone. (C) 3D distance probability density functions of dot pairs (n=45,379; n=10,469; n=18,153 distance measurements for ΔRAD21 (2 hr), ΔCTCF (2 hr), ΔWAPL (4 hr) depletion conditions respectively, and n=17,605; n=11,631; n=21,001 for the same clones without treatment). (D) Localization error corrected 2-Point MSD plots for the AID-tagged clones (left) (n=537; n=137; n=215 trajectories in ΔRAD21 (2 hr), ΔCTCF (2 hr), ΔWAPL (4 hr) depletion conditions respectively, and n=183; n=151; n=257 without treatment (gray lines)). The effective tether length is obtained by computing the ratio of the steady-state variance of each clone to the value in the RAD21-depletion condition (note that 2<R²> is also the asymptotic value of the MSD; see also Supplementary Material). (E) Representative 3D polymer conformation from simulations mimicking the ΔRAD21 (95% cohesin depletion) (left) and ΔCTCF (100% CTCF depletion) (right) depletion conditions. Red: unextruded segment; Blue: extruded segment. (F) Matching simulations to the data to obtain loop extrusion parameters (3 fit parameters; see Supplementary Materials). The extrusion rate is for two-sided extrusion.

Fig. 3.. Bayesian inference of looping dynamics (BILD) reveals rare and dynamic CTCF loops.

(A) Example trajectory from polymer simulations with loop extrusion. Extrusion shortens the effective tether (red: unextruded length, ground truth from simulations) between the CTCF sites. A ground truth loop is formed when the tether is minimal and cohesin is stalled at both CTCF sites (black bar). BILD captures these accurately (purple bar). (B) Schematic overview of BILD. Building on the analytical solution to the Rouse model, we employ hierarchical Bayesian model to determine the optimal looping profile for single trajectories. (C) Illustrative examples of inferred profiles on real trajectory data. (D) Kaplan-Meier survival curves rescaled by the inferred looped fraction. Gray lines are maximum likelihood fits of a single exponential to the data, accounting for censoring. (E) Fraction of time the Fbn2-locus spends in the fully looped conformation. Error bars are bootstrapped 95% confidence intervals. (F) Median loop lifetimes from the Kaplan-Meier survival curves (squares) or exponential fits (crosses). Confidence intervals are determined from the confidence intervals on the Kaplan-Meier curve and the likelihood function of the exponential fit, respectively. Where the upper confidence limit on the survival curve did not cross below 50% an arrowhead indicates a semi-infinite confidence interval.

Fig. 4.. Comprehensive picture of the Fbn2 TAD.

(A) Comparison of Micro-C data for the C36 line to in silico Micro-C of our best-fit simulation, map (left) and contact probability scaling (right). (B) BILD applied to the same simulation (green), comparing to C36 (WT) experimental data (blue). (C) Number of cohesins forming the looped state in simulations (n = 18,789). (D) “Anatomy” of the Fbn2 TAD. Quantitative description of the Fbn2 TAD using both real data (blue) and our best-fit simulation (green). Cohesin processivity and density and CTCF stalling probability and lifetime boost are simulation parameters. Fraction of time in different conformations was extracted from simulation ground truth, using effective tether lengths of 1.1 kb and 505 kb as cutoffs to define “fully looped” and “fully unlooped”, respectively. Fraction of TAD unextruded was calculated using the mean tether length over the full simulation.