Chromosome biorientation—the attachment of a replicated chromosome to both poles of a spindle—is not a simple matter of pointing sister kinetochores in opposite directions and hoping for the best. Attachment errors need to be corrected. Tomoyuki Tanaka (University of Dundee, UK) and colleagues suggest that many sister kinetochores in budding yeast initially attach to a single pole, but then the Ipl1p kinase triggers detachment and reorientation of one sister kinetochore in each pair.
The majority of ipl1 kinetochores segregate into the yeast bud along with the older spindle pole body. The problem does not seem to be in the resolution of replicated centromere DNA into two separate kinetochores, at least not exclusively, as the same preferential association with the old spindle pole was seen when chromosome replication was prevented in ipl1 mutants.
This association with the old spindle pole presumably arises because only the old spindle pole is around and available to make attachments early, when kinetochores are duplicated. The new (and later-arriving) spindle pole got a better shot at attaching to ipl1 kinetochores when DNA replication was delayed. A quick dose of microtubule depolymerizing drugs also helped matters, presumably by substituting for Ipl1p's proposed function in destabilizing kinetochore–microtubule interactions.
This theory is consistent with earlier in vitro evidence for ATP- and Ipl1p-dependent destabilization of kinetochore–microtubule interactions. A more detailed model for Ipl1p action comes from the localization of Aurora B (the mammalian 3Ipl1p) and its counteracting phosphatase PP1 by Jason Swedlow (University of Dundee, UK). He saw differences based on chromosome attachment status. When both kinetochores are attached to a single pole, Ipl1p may promote instability by phosphorylating the nearby kinetochore proteins. But when attachment to opposite poles is achieved, the poles pull the two kinetochores apart, and away from the centrally located Ipl1p. This allows PP1 to take over and stabilize the attachment. The tension is lost when chromosomes separate at anaphase onset, so Ipl1p must be dispersed from the kinetochore before it triggers destabilization once again. ▪
Unreplicated DNA (top) is distributed equally unless Ipl1p is nonfunctional (bottom).