Microtubule plus ends (green) become concentrated in the spindle midzone at anaphase B (bottom).

It takes a polar catastrophe—oodles of them, actually—to get through anaphase, suggest Cheerambathur et al. on page 995. Their modeling reveals that concentrating microtubule catastrophes near the spindle poles might drive the second half of anaphase.

Anaphase in fly cells is a two-step process: first the chromosomes migrate to the poles (anaphase A), then the spindle lengthens, pushing the DNA further apart (anaphase B). To understand the microtubule changes that lead to anaphase B, the authors combined modeling studies with in vivo observations of microtubule dynamics.

The experimental studies revealed that microtubule plus ends became concentrated in the central spindle at the onset of anaphase B. The changes were driven by cell cycle–regulated signaling pathways, not by any intrinsic properties of microtubules.

Determining which spindle proteins are targeted by these pathways is tricky, however. Most candidates have a host of different mitotic functions, making specific effects difficult to pick out. The group thus turned to computational modeling.

Only one of their envisioned scenarios accounted for both the plus-end redistribution and the rapid microtubule turnover dynamics in the spindle. In this model, microtubule plus ends near the poles were threefold more likely switch from growth to depolymerization—an event known as a catastrophe—than were plus ends at the midzone.

The authors imagine that losses at the poles concentrate plus ends in the midzone and thus make it more likely that a microtubule from one pole will capture a partner from the other pole. This bias should expand the microtubule overlap at the midzone, where the pushing forces that elongate the spindle are generated. The remaining short microtubules near the pole, meanwhile, probably keep hold of the chromosomes.

To test their model experimentally, the group will have to screen for factors that affect this catastrophe gradient. It is not yet clear how the gradient is generated. Perhaps it reflects a gradient of a depolymerase activity, such as kinesin-13, which is known to hover near the poles.