Resolution fails (right) when cohesin stays put.

Hirano/CSHL

Achromosomal glue needs to get out of the way so that sister chromatids can resolve into two distinct entities, say Ana Losada, Michiko Hirano, and Tatsuya Hirano (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY).Cohesin sticks sister chromatids together from the moment of their replication. During interphase this may aid in homology-based DNA repair. But when it is time to compact chromosomes in mitosis, all that glue could confuse matters. Up to 95% of the cohesin disperses, with the rest remaining in place until chromosome pairs finally split during anaphase.

Hirano and his colleagues found that two kinases—polo-like kinase (Plx1) and aurora B—were necessary for the initial dispersal. Plx1 probably operates via its direct phosphorylation of cohesin, whereas aurora B seems to hit another target such as histone H3.

Cohesin was no longer lost from chromosomes when frog extracts were depleted of both kinases. But, to the team's surprise, this had no effect on either the loading of condensin or its ability to condense the chromosomes. Based on temporal correlation, some had predicted that cohesin had to leave to make way for the incoming condensin.

Instead, cohesin's persistence resulted in a failure of sister chromatid resolution. Pairs of sister chromatids formed single rod-shaped structures. In this circumstance, says Hirano, “the condensin is not smart enough to recognize that there are two double helices.”

But in normal cells, with cohesin out of the way, Hirano predicts that the condensin operates on only a single DNA molecule. Condensin's compacting activity should reel in a single DNA molecule, while allowing topoisomerase to untangle the strands that have not yet been packed away. ▪

Reference:

Losada, A., et al.
2002
.
Genes Dev.
16
:
3004
–3016.