Slow checkpoint activation kinetics as a safety device in anaphase.

Chromosome attachment to the mitotic spindle in early mitosis is guarded by an Aurora B kinase-dependent error correction mechanism [1, 2] and by the spindle assembly checkpoint (SAC), which delays cell-cycle progression in response to errors in chromosome attachment [3, 4]. The abrupt loss of sister chromatid cohesion at anaphase creates a type of chromosome attachment that in early mitosis would be recognized as erroneous, would elicit Aurora B-dependent destabilization of kinetochore-microtubule attachment, and would activate the checkpoint [5, 6]. However, in anaphase, none of these responses occurs, which is vital to ensure progression through anaphase and faithful chromosome segregation. The difference has been attributed to the drop in CDK1/cyclin B activity that accompanies anaphase and causes Aurora B translocation away from centromeres [7-12] and to the inactivation of the checkpoint by the time of anaphase [10, 11, 13, 14]. Here, we show that checkpoint inactivation may not be crucial because checkpoint activation by anaphase chromosomes is too slow to take effect on the timescale during which anaphase is executed. In addition, we observe that checkpoint activation can still occur for a considerable time after the anaphase-promoting complex/cyclosome (APC/C) becomes active, raising the question whether the checkpoint is indeed completely inactivated by the time of anaphase under physiologic conditions.