Centromere errors get chewed out

Microtubule (MT) ends that invade into the inner centromere may get chewed up, say Ryoma Ohi, Timothy Mitchison (Harvard Medical School, Boston, MA), and colleagues. The process should help prevent MTs emanating from a single pole from attaching to both sister kinetochores. Figure ICIS (red, left) helps destroy invading microtubules (red, right). Ohi proposes that the chewing is performed by the KinI kinesin MCAK. He found a new MT-binding protein, ICIS, that binds to MCAK, relies on it for localization at the inner centromere, and further activates MCAK's MT-depolymerizing activity in vitro. Although ICIS depletion causes widespread MT polymerization, probably because of co-depletion of MCAK, it almost certainly has a more specific function at the inner centromere. Just away from the inner centromere—specifically near the kinetochore—a high density of MTs is desirable if the cell is to ensure kinetochore capture. “But it's also dangerous,” says Ohi. Slightly errant MTs can skirt around the closest kinetochore, cross the inner centromere, and attach to the more distant kinetochore. Ohi predicts that ICIS-stimulated MCAK intercepts these MTs before they can reach their incorrect target. He now proposes to test whether interference with the MCAK/ICIS system causes attachment errors, as his model predicts. ▪ Reference: Ohi, R., et al. 2003. Dev. Cell. 5:309–321. [PubMed]