The spindle checkpoint is an intracellular signaling network during mitosis that senses and responds to kinetochores not under bi-orientation and delays anaphase (Figure 1). In the past two decades, our research, along with contributions from many others, has established that unattached kinetochores recruit and activate checkpoint proteins. The activated checkpoint proteins collaborate to inhibit Cdc20, the mitotic activator of APC/C, thereby delaying anaphase. A critical APC/C inhibitor is the mitotic checkpoint complex (MCC) comprising BubR1-Bub3-Mad2-Cdc20. Microtubule binding and tension generation at kinetochores displace and inactivate checkpoint proteins, and silence the checkpoint. Checkpoint signaling is thus controlled by the spatiotemporally regulated antagonism between checkpoint proteins and microtubules at kinetochores.
A highlight of our work is the demonstration that Mad2 is a prion-like, multi-state protein. The conformational transition of Mad2 is critical for MCC assembly and checkpoint signaling, and is regulated by positive (Mad1) and negative (p31comet) regulators and by phosphorylation. More recently, we have established the KMN network of kinetochore proteins (consisting of Knl1-Zwint, the Ndc80 complex, and the Mis12 complex) as a major signaling platform of the spindle checkpoint in human cells. We have characterized the mechanisms by which KMN is anchored to centromeric chromatin during mitosis. Furthermore, we have shown that competition between the checkpoint kinase Mps1 and microtubules for binding to the Ndc80 complex constitutes a direct mechanism for detection of unattached kinetochores (Figure 2). Finally, we have shown that a heterodimeric kinase complex of Bub1 and Plk1 phosphorylates Cdc20 and inhibits APC/C in a mechanism that is parallel, but not redundant, to MCC formation. Both mechanisms are required to sustain mitotic arrest in response to spindle defects.
Our ongoing research focuses on the in vitro reconstitution of KMN-dependent spindle checkpoint signaling and microtubule-mediated inactivation of this process in the test tube. These experiments will define the minimal checkpoint sensor, and establish the activation and inactivation mechanisms of the checkpoint. Another ongoing effort is to explore the physiological functions of spindle checkpoint components using mouse genetics. Finally, we are interested in studying the plasticity of mitotic programs and consequences of chromosome missegregation in different cell types using the human embryonic stem cells and induced pluripotent stem cells as models.