Splitting the nucleus: what's wrong with the tripartite ring model?

The segregation of sister DNA molecules at mitosis involves their traction to opposite poles by microtubules attached to kinetochores. By creating tension required to stabilize kinetochore microtubules, sister chromatid cohesion has a key role in ensuring that sister kinetochores attach to microtubules with opposing polarity, a process known as biorientation. Cohesion is mediated by a cohesin complex whose Smc1, Smc3, and kleisin subunits form a tripartite ring thought to hold sister DNAs together by entrapping them (the ring model). Sister chromatid disjunction at the onset of anaphase is triggered by a thiol protease called separase whose activation, only when all chromosomes have bioriented, opens the cohesin ring by cleaving its kleisin subunit. Separase is inhibited by the binding of an inhibitory chaperone called securin whose destruction at the hands of a ubiquitin protein ligase called the anaphase-promoting complex/cyclosome (APC/C) is essential for kleisin cleavage and sister chromatid disjunction. We describe microinjection experiments showing that cohesin cleavage and Cdk1 down-regulation are sufficient to drive formation of daughter nuclei in cells arrested in metaphase due to inactivation of the APC/C and describe chemical cross-linking experiments consistent with the ring model. How sister DNAs enter the cohesin ring and are retained inside for long periods of time after the completion of DNA replication remains poorly understood.