Condensin II is a structural-maintenance-of-chromosomes complex involved in organizing DNA into mitotic chromosomes. Its activity is tightly controlled: the protein MCPH1 acts as an interphase inhibitor, whereas M18BP1 is a mitotic activator. Phosphorylation favors the association of M18BP1 with condensin II, but how that interaction changes the molecular architecture of the complex was not fully understood. The study focused on human condensin II and used a phosphorylated M18BP1 fragment spanning residues 873–1132, designated M18BP1P, to investigate DNA binding, compaction and the underlying conformational changes.

Direct imaging of individual condensin II molecules attached to a 48.5-kilobase lambda-DNA tether showed that M18BP1P increased both the frequency and duration of DNA-binding events. With ATP present, condensin II alone moved along DNA in a manner consistent with one-dimensional diffusion and transient engagement. Its mean residence time on the substrate increased from 5.3 ± 0.5 seconds without M18BP1P to 9.4 ± 0.6 seconds with the phosphorylated activator, while its diffusion coefficient fell almost tenfold, from 0.11 ± 0.01 to 0.01 ± 0.003 square micrometres per second. A truncated M18BP1 mutant lacking the phosphorylation sites required for condensin II binding did not produce these effects.

Fluorescently labelled condensin II and M18BP1P also colocalized on DNA, whereas M18BP1P alone did not bind detectably. Stable association therefore required the two proteins together. The complexes formed at both 0.2 and 1 piconewton of DNA tension, suggesting that the enhanced loading was not restricted to one force regime. ADP-BeF3, which locks the condensin ATPase domains in a prehydrolysis state, supported stable, non-diffusive DNA binding, indicating that ATP hydrolysis was not required for the association itself. In the absence of nucleotides, little or no binding was detected.

Compaction assays demonstrated that the activated complexes could also reorganize DNA. When a DNA tether was relaxed and then stretched between optical traps, condensin II–M18BP1P complexes generated a distinctive sawtooth force–distance profile, while condensin II alone produced a curve similar to that of DNA without protein. Fluorescence signals that initially occupied separate positions along the DNA converged into brighter clusters, indicating that neighbouring proteins assembled into structures capable of resisting substantial mechanical force. Human condensin I compacted DNA and formed similar clusters without requiring an auxiliary activator, highlighting a functional difference between the two human condensin complexes.

The experiments nevertheless indicated two separable modes of compaction. One depended on ATP hydrolysis and was associated with smaller sawtooth features that may reflect rupture of DNA loops. A second was independent of ATP hydrolysis and appeared to operate when stable condensin II–M18BP1P complexes were close enough to associate, producing persistent compaction events. At the lowest condensin concentration tested, individual complexes were separated by several kilobases, yet appreciable compaction remained detectable. Compaction increased as more complexes bound the DNA, and the ATPase-deficient mutant compacted DNA at higher protein concentrations even though it was ineffective at lower concentrations.

Cryo-electron microscopy explained how this change in activity occurs. A structure of DNA-bound condensin II, reconstructed at an average resolution of 6.8 ångströms, showed the characteristic clamped architecture: the SMC2 and SMC4 ATPase heads and the CAP-D3 and CAP-H2 subunits surround the DNA. The coiled-coil arms and hinge region remained flexible and were not resolved. Although M18BP1P was present during sample preparation, neither it nor the CAP-G2 subunit could be identified clearly in the final DNA-bound map, consistent with their greater flexibility or displacement from the core.

A separate structure of condensin II without DNA or M18BP1P, refined to an average resolution of 4.6 ångströms, captured an autoinhibited state. In this arrangement, CAP-G2 simultaneously contacted the SMC2 ATPase head and the SMC4 coiled-coil region, bridging the core in a way that was not compatible with DNA entrapment. The CAP-H2 N-terminal domain and the CAP-D3 C-terminal tail helped stabilize this closed configuration. A higher-resolution structure of the condensin II–M18BP1P complex then showed that binding of the activator displaced CAP-G2 and shifted the two HEAT-repeat-containing subunits, CAP-G2 and CAP-D3, as a rigid body by about 60 ångström.

Microscale thermophoresis supported the structural interpretation. The SMC2/4 dimer bound CAP-G2 with a dissociation constant of 1.3 ± 0.6 micromolar, but no binding was detected when M18BP1P was present at 2 micromolar. The unphosphorylatable M18BP1 mutant did not inhibit the interaction. These results support a model in which M18BP1P occupies a site close to the inhibitory SMC2–CAP-G2 interface, destabilizing the contact and allowing CAP-G2 to move away from its autoinhibitory position. The authors propose that M18BP1P binding and phosphorylation of the CAP-D3 tail provide partly redundant routes for relieving inhibition, although the contribution of phosphorylation at the specific CAP-D3 residue T1415 remains unresolved because it was not detected in the condensin II preparations used.

The study offers a detailed model for the multilayered activation of condensin II, in which phosphorylation-dependent M18BP1 binding primes the complex for DNA engagement and ATP-associated movement can then support loop-related compaction and chromosome organization. It also suggests that condensin II may gain mechanical strength not only through the activity of individual complexes but through cooperative assembly of neighbouring proteins into stable clusters. That possibility is consistent with models in which loop extrusion begins chromosome folding before non-extruding condensins assemble into reinforcing structures. However, the short DNA substrate used for cryo-electron microscopy may bypass parts of the physiological loading reaction, and the optical-tweezer experiments used isolated lambda-DNA rather than native mitotic chromosomes. Further work will therefore be needed to determine whether the observed assemblies form in cells and whether preventing them changes chromosome architecture or stability.

Subject of Research: Molecular mechanisms that activate human condensin II for DNA compaction in vitro

Article Title: Mechanism of M18BP1-mediated activation of condensin II for mitotic DNA compaction

Article References: Borsellini, A., Chamera, S., Cecatiello, V., Andrecka, J., & Vannini, A. (2026). Mechanism of M18BP1-mediated activation of condensin II for mitotic DNA compaction. Nature Structural & Molecular Biology. https://doi.org/10.1038/s41594-026-01902-8

Image Credits: AI Generated

DOI: 10.1038/s41594-026-01902-8

Keywords: condensin II, M18BP1, DNA compaction, mitotic chromosomes, single-molecule imaging, cryo-electron microscopy, optical tweezers, chromosome organization, protein autoinhibition, cell division, Mechanism, M18BP1-mediated