Résumé
Mitotic checkpoints delay cell cycle progression in response to alterations in the mitotic apparatus, thus ensuring correct chromosome segregation. While improper spindle orientation activates the Bub2/Bfa1-dependent checkpoint in budding yeast, delaying exit from mitosis, lack of bipolar kinetochore-microtubule attachment activates a signal transduction cascade that prevents both anaphase onset and exit from mitosis by inhibiting the Cdc20/APC (Anaphase Promoting Complex)-mediated proteolysis of securin and inactivation of mitotic cyclin-dependent kinases (CDKs), respectively. Proteolysis of the securin Pdsl is necessary to liberate the separase Esp1, which then triggers sister chromatid separation, whereas inactivation of mitotic CDKs is a prerequisite for exit from mitosis and for starting a new round of DNA replication in the next cell cycle. In budding yeast, this latter checkpoint response involves the proteins Mad1, 2, 3, Bub1 and Bub3, whose vertebrate counterparts localize to unattached kinetochores. Mutations that alter other kinetochore proteins result in mitotic checkpoint activation, while the ndc10-1 mutation not only impairs kinetochore function, but also disrupts the checkpoint response, indicating a role for Ndc10 in this process. Here we present evidence that Ndc10 is not part of the Bub2/Bfa1-dependent pathway, and its role in the checkpoint response might also be different from that of the other Mad and Bub proteins. Indeed, Ndc10, unlike other mitotic checkpoint proteins, is not required for the mitotic block induced by overexpression of the Mpsl protein kinase, which is implicated in mitotic checkpoint control. Furthermore, the delay in mitotic exit caused by non-degradable Pds1, which does not require Mad and Bub proteins, depends on Ndc10 function. We propose that a pathway involving Ndc10 might monitor defects in the mitotic apparatus independently of the Mad and Bub proteins. Since the Espl separase is required for exit from mitosis in both ndc10-1 and nocodazole-treated mad2delta cells, the two signal transduction cascades might ultimately converge on the inactivation of Esp1.