Structural maintenance of chromosomes (SMC) proteins like cohesin, condensin, and Smc5/6 are key to chromosome structure and dynamics. Using single-molecule visualization, we discovered that these SMCs constitute an entirely new class of molecular motors that extrude loops into DNA. I will present some of the latest biophysics results from our lab. Strikingly, SMCs can easily bypass DNA-binding proteins on DNA, even ones that exceed the size of these large ring-shaped SMCs—demonstrating that loop extrusion occurs nontopologically. While virtually all DNA-binding proteins are accommodated into SMC-extruded loops, long linear protein arrays on DNA can efficiently halt loop extrusion—as we discovered for dense arrays of Rap1, the DNA-binding telomeric protein of yeast, with implications for telomere biology. Due to biochemical interactions, the DNA-binding protein CTCF can also stop SMC traversal, which makes it define topologically associating domains along chromosomes. Surprisingly, we found the CTCF poses a barrier to loop extrusion that is dependent on DNA tension. Contrary to some reports in the literature, we find that all eukaryotic SMCs extrude DNA asymmetrically, while the direction of loop extrusion can switch over time. For cohesin, direction switches strongly correlate with the turnover of the subunit NIPBL—indicating that SMCs contain a direction switch subunit. Using magnetic tweezers, we recently demonstrated that all three SMC complexes induce the same large negative twist (i.e., a linking number change of −0.6) at each loop-extrusion step into the extruded loop, independent of step size. ATP binding was found to be the twist-inducing event. Most recently, we also explored what happens when human cohesin encounters condensin I or II, and I present the latest results at the conference. Overall, these single-molecule findings reveal mechanistic principles of how SMCs regulate the genome architecture.
Cees Dekker (Sun,) studied this question.