ARTICLE doi:10.1038/nature10896 Structure of the mitotic checkpoint complex William C. H. Chao 1 *, Kiran Kulkarni 1 *, Ziguo Zhang 1 , Eric H. Kong 1 & David Barford 1 In mitosis, the spindle assembly checkpoint (SAC) ensures genome stability by delaying chromosome segregation until all sister chromatids have achieved bipolar attachment to the mitotic spindle. The SAC is imposed by the mitotic checkpoint complex (MCC), whose assembly is catalysed by unattached chromosomes and which binds and inhibits the anaphase-promoting complex/cyclosome (APC/C), the E3 ubiquitin ligase that initiates chromosome segregation. Here, using the crystal structure of Schizosaccharomyces pombe MCC (a complex of mitotic spindle assembly checkpoint proteins Mad2, Mad3 and APC/C co-activator protein Cdc20), we reveal the molecular basis of MCC-mediated APC/C inhibition and the regulation of MCC assembly. The MCC inhibits the APC/C by obstructing degron recognition sites on Cdc20 (the substrate recruitment subunit of the APC/C) and displacing Cdc20 to disrupt formation of a bipartite D-box receptor with the APC/C subunit Apc10. Mad2, in the closed conformation (C-Mad2), stabilizes the complex by optimally positioning the Mad3 KEN-box degron to bind Cdc20. Mad3 and p31 comet (also known as MAD2L1-binding protein) compete for the same C-Mad2 interface, which explains how p31 comet disrupts MCC assembly to antagonize the SAC. This study shows how APC/C inhibition is coupled to degron recognition by co-activators. The fidelity of chromosome separation in mitosis is governed by an evolutionarily conserved cell-cycle checkpoint mechanism called the SAC 1,2 . The SAC arrests the mitotically dividing cell to allow complete chromosome attachment to the bipolar mitotic spindle. The essence of the SAC is to block the onset of anaphase by inhibiting APC/C- mediated ubiquitin-dependent degradation of securin and mitotic cyclin. Components of the SAC that are responsible for detecting unattached kinetochores and for propagating signals to the APC/C have been identified 3,4 , but the molecular basis underlying these pro- cesses is only partially understood 1,2 . Mad2 and Mad3 (BubR1 in metazoans) mediate APC/C inhibition through their association with its co-activator subunit Cdc20 (refs 5–12). Mad2, Mad3 and Cdc20 (together with mitotic checkpoint protein Bub3) form the MCC that directly binds the APC/C to inhibit substrate recognition 13,14 . Mad2 and Mad3 cooperate to antagonize Cdc20-dependent activation of the APC/C 12 , with Mad3–Cdc20 interactions requiring the pre-assembly of a Cdc20–Mad2 complex 12,15–18 . Thus, SAC signalling occurs through the generation of the Cdc20–Mad2 complex, a process initiated by Mad1, which is the Mad2 receptor at unattached kinetochores. Central to the association of Mad2 with Cdc20 is the inter-conversion of Mad2 between the open (O-Mad2) and closed (C-Mad2) structural states 1,19 . These states of Mad2 differ in the topology of a carboxy- terminal b-sheet that repositions in C-Mad2 to enable binding to its protein ligands, Mad1 or Cdc20 (refs 20–22). In the template model for SAC activation, Mad1 interactswith C-Mad2, generating the C-Mad2– Mad1 complex that subsequently recruits O-Mad2 through the C-Mad2 dimerization interface. By inducing the conformational trans- ition of O-Mad2 to C-Mad2, the Mad1-bound C-Mad2 subunit cata- lyses the binding of Mad2 to Cdc20 (refs 17, 23). APC/C activity and its substrate recruitment are dependent on its co- activators (Cdc20 and Cdh1) 24 , which recognize APC/C substrates through two destruction motifs (degrons); the D box 25 and the KEN (Lys-Glu-Asn) box 26 . Mad3 contains a KEN box that is essential for MCC assembly 15,27,28 , suggesting that Mad3 may act as a pseudosubstrate to block substrate recognition by APC/C Cdc20 . However, other studies showing that the promotion of ubiquitin-mediated degradation of Cdc20 by the SAC is dependent on the KEN box of Mad3 (refs 18, 27, 29) indicate that there is a more complex mechanism controlling APC/C Cdc20 activity. The mechanisms underlying APC/C activation after SAC silencing are also poorly understood. In metazoans, p31 comet antagonizes the SAC 30 by functioning as a structural mimic of Mad2 that binds at the Mad2 dimerization interface to inhibit the conforma- tional activation of O-Mad2 (ref. 31). UbcH10, assisted by p31 comet , catalyses Cdc20 ubiquitination, which leads to the disassembly of the MCC 32 . To understand the molecular mechanisms underlying the mitotic checkpoint complex, we determined the crystal structure of the fission yeast MCC. The structure shows how Mad2 and Mad3 cooperatively inhibit Cdc20, and indicates how p31 comet would antagonize MCC assembly. The structure of Cdc20 in the context of the MCC offers the first opportunity to visualize degron recognition by co-activators. The interaction between Mad2 and Mad3 positions the Mad3 KEN box towards the KEN-box receptor of the Cdc20 WD40 domain. Additionally, an unexpected D-box mimic of the Mad3 C terminus reveals the D-box-binding site on Cdc20, thus demonstrating the structural basis of D-box recognition by co-activators. Overall structure of the MCC We generated the fission yeast MCC by co-expressing Cdc20, Mad2 and Mad3 in insect cells. The complex comprises Cdc20 with all functional domains (C box, Mad2-binding motif, WD40 domain and Ile-Arg tail (Supplementary Fig. 1)), Mad2 locked in its closed conformation that facilitates binding to Cdc20 (ref. 31), and Mad3 truncated after the tetratricopeptide repeat (TPR) domain 33 and thus lacking its C-terminal KEN box. Bub3 was omitted from the complex because previous studies indicated that Bub3 was not an integral part of MCC in fission yeast 28 and was not required for MCC-mediated inhibition of human APC/C 34 . 1 Division of Structural Biology, Institute of Cancer Research, Chester Beatty Laboratories, 237 Fulham Road, London, SW3 6JB, UK. *These authors contributed equally to this work. 208 | NATURE | VOL 484 | 12 APRIL 2012 Macmillan Publishers Limited. All rights reserved ©2012