Curled microtubules in the presence of GMPCPP and XMAP215.

Looks can be deceiving. According to two articles in this issue, proteins that look like microtubule stabilizing proteins at times do just the opposite, revealing activities that can both build and destroy microtubules.

Originally described as a Xenopus microtubule stabilizing protein, XMAP215 is a defining member of a large family of microtubule-associated proteins. Depletion of XMAP215 or its homologues leads to decreased spindle microtubule length in several systems, including fly, yeast, and worm. On page 349, however, Shirasu-Hiza et al. find that XMAP215 also promotes depolymerization of microtubules stabilized with a nonhydrolyzable GTP analogue (GMPCPP). This destabilizing activity, like its stabilizing activity, is specific to microtubule plus ends. The new work recalls a 10-year-old report demonstrating that XMAP215 has both activities in vitro.

Sirasu-Hiza et al. used EM analysis to reveal a structure that supports a peeling-like mechanism of XMAP215, similar to that of KinI kinesin. Previously, the plus ends of microtubules stabilized by GMPCPP have been thought to resemble a “GTP cap,” a structure postulated to exist at the ends of growing microtubules. Here, the authors suggest instead that GMPCPP-stabilized structures may mimic a “paused state”—a hypothetical third state in microtubule dynamics, intermediate between the growing and shrinking states. They propose that XMAP215 destabilizes this paused state and increases either polymerization or depolymerization rates depending on cellular conditions, thus explaining its dual activities.

On page 359, van Breugel et al. find another XMAP215 family member with destabilizing activity—the budding yeast homologue Stu2p. In vitro, Stu2p depolymerized microtubules by binding directly to plus ends, probably hindering tubulin dimer addition and thus increasing catastrophe rates. In contrast to the short spindle microtubules seen previously in stu2p mutants, cytoplasmic microtubules of stu2p interphase cells are longer than those in the wild type. Thus, for both yeast and frog proteins, cellular context, such as cell cycle status or protein localization, may determine their effects on microtubules. It remains to be seen whether destabilizing activity has been overlooked in other family members. ▪