The chromosomal condensin complex is best known for its role in promoting mitotic chromosome condensation. In addition, it is required for sister chromatid resolution during their segregation in anaphase (1–4
). The two activities of condensin are separable. Inactivation of budding yeast condensin results in 1.5-fold longer chromosome arms in mitosis, which should prevent only the longest chromosomes from fully segregating, but the resolution of most chromosomes is impaired (3
). Furthermore, Aurora B kinase is required for condensin-dependent compaction, but not resolution, of the budding yeast ribosomal DNA (rDNA) locus (7
). A role of condensin in both chromosome compaction and resolution is conserved in higher eukaryotes (4
). Similarly, the condensin-related SMC complex in prokaryotes is required for both compaction and segregation of the bacterial nucleoid, implying that chromosome resolution is a fundamental aspect of condensin function (14
The chromosome segregation failure in condensin mutant yeast cells is reminiscent of the phenotype observed after inactivation of topoisomerase II (topo II), an enzyme that removes catenanes that persist between sister chromatids following DNA replication (16–18
). It has therefore been suggested that condensin facilitates DNA decatenation by topoisomerase II. Stimulation of topo II by condensin was reported in Drosophila
, but could not be seen in Xenopus
or yeast (4
). Alternative models in which condensin resolves transcription-dependent chromosome links have therefore been put forward (20
). Recent studies in prokaryotes have reported a direct protein interaction between the Escherichia coli
condensin subunit MukB and the E. coli
decatenating topoisomerase topo IV (22
). However, the impact of this interaction on the decatenation activity of topo IV remained weak and an alternative role of the MukB–topo IV interaction in chromosome folding has been suggested (24
). Whether and how chromosome folding might indirectly facilitate sister chromatid decatenation remains an interesting open question.
A reason why the interplay between condensin and topo II in chromosome resolution remains poorly understood is the difficulty of visualizing sister chromatid catenation. Intertwinings between linear chromosomes are not maintained after DNA isolation and cannot therefore be visualized by conventional gel electrophoresis. Catenanes between circular chromosomes, however, can be visualized. This has early on been used to characterize the cell-cycle dependence of catenane formation in budding yeast (25
) with an aim of investigating whether catenanes contribute to sister chromatid cohesion. This latter question has recently been addressed in more detail and a role for cohesin, but not condensin, in maintaining catenanes after their synthesis in S-phase has been documented (26
). These studies have not addressed how catenanes that persist between sister chromatids following DNA replication are eventually resolved. This was the topic of another recent study that correlated condensin and microtubule-dependent positive supercoiling of plasmid DNA in vivo
with its facilitated decatenation by topo II in vitro
). A possible role of condensin in DNA decatenation in vivo
was not addressed in this study due to the difficulty of visualizing plasmid catenanes.
Despite the ubiquitous occurrence of DNA catenation as the consequence of DNA replication, and the importance of its resolution, chromosome decatenation during mitosis has not yet been directly observed in vivo. Here, we follow the catenation status of circular minichromosomes of three sizes during cell-cycle progression in budding yeast. We find that the majority of catenanes produced during DNA replication are rapidly resolved, but that complete catenane removal requires condensin, a dependency that becomes more pronounced as chromosome size increases. These results provide evidence that condensin promotes sister chromatid decatenation to facilitate eukaryotic chromosome segregation.