Linking kinetochore-microtubule binding to the spindle checkpoint

Abstract

The spindle checkpoint blocks cell-cycle progression until chromosomes are properly attached to the mitotic spindle. Popular models propose that checkpoint proteins associate with kinetochores to produce a "wait anaphase" signal that inhibits anaphase. Recent data suggest that a two-state switch results from using the same kinetochore proteins to bind microtubules and checkpoint proteins. At least eight protein kinases are implicated in spindle checkpoint signaling, arguing that a traditional signal transduction cascade is integral to spindle checkpoint signaling.

Figure 1. A current model of spindle checkpoint signaling highlights the importance of MCC formation as the step catalyzed by the kinetochore

Green and purple are proteins required for microtubule binding. Checkpoint proteins (red and yellow shapes) are recruited to kinetochores that are not binding microtubules (central image). In the absence of microtubules the checkpoint proteins act to catalyze the assembly of the MCC checkpoint complex (bottom image), which diffuses from the kinetochore to inhibit the anaphase promoting complex (APC). Mad2 has both open and closed states and although it is a globular protein it is drawn as a ring to emphasize transitions between these two states. Mad2^closed on kinetochores binds Mad2^open, which in turn binds Cdc20, Bub1 and Bub3 to form MCC. Microtubule attachment inhibits the signal by two mechanisms. The checkpoint proteins Mad1/Mad2 and RZZ are shown being “stripped” by dynein, which carries them away from the kinetochore by walking towards the minus end of the microtubule (top image). CENP-E activates BubR1 kinase activity unless it binds microtubules, which is also important for silencing the checkpoint.

Figure 2. An updated model for spindle checkpoint signaling

A) Dual role of KMN in the kinetochore as both microtubule anchor and a scaffold to generate spindle checkpoint signals. Overall color scheme is the same as Figure 1. KMN (purple and green) is located in the outer plate of the kinetochore where both binds microtubules (arrow left) and spindle checkpoint protein (arrow right). KMN contains at least two microtubule-binding interfaces one in the KNL-1 subunit and another in the Hec1 subunit of the Ndc80 complex and approximately eight KMNs generate a binding pocket (not shown). In the absence of microtubules KMN has both direct and indirect interactions with checkpoint proteins. KNL-1 binds Bub3/Bub1 and Bub3/BubR1. The Ndc80 subunit can bind a coiled-coil region of Mad1 in a two-hybrid assay. Finally through the Zwint protein, the Mis12 complex binds RZZ, which can strip Mad1 from kinetochores. The kinetochore activates the checkpoint by acting as a scaffold to recruit and activate Bub1 and BubR1 kinases as well as other kinases recently implicated in checkpoint signaling (stars). B) A schematic Map of the Mad1 protein highlighting kinase interactions (stars) and a potential signal transduction network initiated after Mad1 recruitment to the kinetochore.

Figure 3. A possible role for Dynein in silencing the checkpoint signals by removing checkpoint proteins associated with KMN complexes that are not binding microtubules

Since there is more KMN than is necessary to bind microtubules there may be two independent steps in checkpoint silencing. First, microtubule binding sterically prevents checkpoint protein binding to potential microtubule binding sites. Second, dynein “strips” checkpoint proteins from surrounding KMN complexes that are not associated with microtubules.