The Smc5/6 complex is a DNA loop-extruding motor

**Background:** Structural maintenance of chromosomes (SMC) protein complexes—cohesin, condensin, and Smc5/6—are essential for chromosome organization. While cohesin and condensin are known to organize chromosomes by extruding DNA loops, the molecular function of Smc5/6 remained largely unknown. Smc5/6 has been implicated in DNA repair by homologous recombination, chromosome segregation, and replication fork stability, but whether it performs loop extrusion was unclear.

**Methods:** The authors purified wild-type (WT) octameric Smc5/6 from Saccharomyces cerevisiae, along with hexameric complexes lacking Nse5/6, pentameric complexes lacking Nse2 and Nse5/6, and ATPase-deficient mutants (KE for ATP binding, EQ for ATP hydrolysis). They used a single-molecule assay with 48.5-kbp λ-DNA molecules tethered at both ends to a glass surface, stained with Sytox Orange, and imaged by total internal reflection microscopy. Loop extrusion was visualized under constant buffer flow or in its absence. Smc5/6 was fluorescently labeled via SNAP-tags on Nse4, Nse2, or Nse5 subunits (labeling efficiency 68±10% and 70±10%). Photobleaching step counting, mass photometry, and Langmuir-Hill analysis were used to determine the oligomeric state required for loop extrusion. ATPase activity was measured using radiolabeled ATP.

**Key Results:** Smc5/6 formed DNA loops in an ATP hydrolysis-dependent manner: 78% of DNA molecules formed loops (n_tot=233 for 2 nM Smc5/6). No looping was observed without ATP, with AMP-PNP, or with KE/EQ mutants. Loop extrusion occurred at a rate of 1.1±0.5 kbp/s (n_tot=102), similar to human cohesin (0.5–1.0 kbp/s) and yeast condensin (0.6 kbp/s). The average loop size was ~16 kbp (n_tot=100). Loop extrusion was force-sensitive, with a stalling force of 0.5±0.1 pN. The majority of events (89%) were two-sided, with DNA length decreasing on both sides of the loop. Photobleaching analysis of labeled Smc5/6 showed that loop-extruding complexes were dimers: 82% of loop events showed a Smc5/6 signal, with 43% exhibiting two-step bleaching and 36% one-step bleaching. Probability calculations indicated that the observed ratios most closely correlated with 100% dimer fraction. The Hill coefficient from Langmuir-Hill fitting was n_H=1.84, confirming cooperative dimerization. In contrast, non-looping Smc5/6 complexes predominantly translocated unidirectionally along DNA (80% of non-looping events, n_nonlooping=45) at an average velocity of 1.5±0.2 kbp/s (n_tot=32), and 92% of translocating units were monomers. Hexameric complexes lacking Nse5/6 showed an ~15-fold increase in looping probability compared to WT octamer, but similar extrusion rates (1.3±0.6 vs 1.1±0.5 kbp/s) and dwell times (270±190 vs 317±229 s). Pentameric complexes lacking Nse2 showed no looping (n_tot=300). Mass photometry revealed that hexamers formed over tenfold more dimers (25% of total counts) than octamers (2%), and adding purified Nse5/6 to hexamers reduced dimerization to 3%. High-salt washing (1 M NaCl) disrupted all extruded loops and removed all loop-extruding complexes, but a minor fraction (7%) of non-looping octamers remained bound, indicating topological entrapment occurs at low frequency and requires Nse5/6.

**Clinical Implications:** This study establishes that Smc5/6 is a DNA loop-extruding motor, revealing a conserved mechanism among all three eukaryotic SMC complexes. The finding that Smc5/6 functions as a dimer for loop extrusion but as a monomer for translocation, and that Nse5/6 negatively regulates loop initiation, provides molecular insight into how Smc5/6 contributes to genome integrity. These results open new avenues for understanding how defects in Smc5/6-mediated loop extrusion may underlie genomic instability and disease, though direct clinical applications remain distant as this is fundamental molecular biology.