This project elucidated the molecular mechanisms by which SMC complexes—condensin, cohesin, and Smc5/6—structure genomes through DNA loop extrusion. Using high-resolution magnetic tweezers, single-molecule fluorescence imaging, and complementary in vivo analyses, we defined key kinetic, structural, and regulatory principles underlying this process.
We resolved discrete loop extrusion steps by individual SMC complexes, revealing median step sizes of 20–40 nm (corresponding to ~200 bp) that depend on DNA tension. ATP binding, rather than hydrolysis, was identified as the step-generating event, consistent with a scrunching mechanism in which ATP-dependent engagement between hinge and globular domains drives loop growth. Molecular dynamics simulations demonstrated that the observed large step sizes arise from DNA flexibility at low forces.
Condensin and cohesin were found to translocate past nucleosomes, RNA polymerase, dCas9, and even 200 nm DNA-bound nanoparticles—well beyond the SMC ring dimensions. This ability persisted even for single-chain cohesin incapable of ring opening, providing strong evidence for a nontopological loop extrusion mechanism that excludes DNA passage through the SMC ring. We analyzed how extended DNA-bound linear protein arrays, such as those at telomeres, influence SMC activity. Combined in vitro and live-cell experiments showed that continuous high-density arrays completely stall condensin-driven loop extrusion, in contrast to single bound proteins. Stalling efficiency scales with array density, length, and stiffness, establishing quantitative parameters for extrusion-blocking architectures.
We further demonstrated that SMC complexes couple extrusion to DNA topology: each step introduces ~0.6 negative supercoils. For cohesin, negative supercoiling required ATPase-head engagement, DNA clamping, and hinge-domain DNA binding, indicating that DNA is twisted as it is constrained between these elements. Cohesin mutants defective in supercoiling formed shorter DNA loops in vivo, and depletion of topoisomerase I produced a similar, though weaker, phenotype, showing that torsional relaxation facilitates processive loop extrusion.
Regulation by CTCF was dissected using single-molecule assays that directly visualized CTCF–cohesin interactions. We found that CTCF blocks cohesin-mediated loop extrusion in a DNA-tension-dependent manner, revealing it to be an active modulator rather than a passive barrier. Dissection of the CTCF N-terminal region identified two regulatory motifs: KTYQR, which fully impedes extrusion, and YDF, which converts cohesin into a unidirectional extruder by stabilizing STAG1–DNA contacts. These motifs act through distinct but synergistic mechanisms to control the positioning and stability of cohesin loops.
Together, these studies provide a unified mechanistic framework for SMC-driven DNA loop extrusion, integrating step size, topology, and regulatory modulation by chromatin-bound factors. The findings redefine current models of chromosome organization and establish essential constraints for future structural and theoretical studies of SMC dynamics.
In a complementary approach, we established and optimized protocols for a genome-in-a-box experimental platform designed to reconstitute and study chromosome-scale DNA organization under controlled conditions. Here, genome-length DNA molecules are confined within microfabricated chambers together with SMC complexes and other DNA-processing proteins, enabling direct visualization of the structure and dynamics of megabasepair-scale DNA in vitro. Using this system, we examined chromosome condensation driven by yeast condensin acting on long DNA substrates.
To achieve this, we developed microfluidic circuitry that allows the precise manipulation, confinement, and imaging of intact genomes. We introduced a sequential microfluidic extraction and purification method that isolates bacterial nucleoids within individual microchambers, eliminating any need for manual transfer or pipetting of fragile, megabase-size DNA molecules and thereby preventing their fragmentation. Using this device, we successfully extracted and analyzed single Bacillus subtilis chromosomes. Upon on-chip cell lysis, the bacterial genome expanded while native DNA-binding proteins were efficiently removed; subsequently, exogenous proteins could be reintroduced by diffusion, allowing controlled reconstitution experiments.
This integrated microfluidic platform provides a robust route to interrogate genome-scale DNA–protein interactions and to study SMC-driven chromosome organization under defined biochemical and physical conditions.
