Recruitment of Mad1 to metaphase kinetochores is sufficient to reactivate the mitotic checkpoint

Abstract

The mitotic checkpoint monitors kinetochore-microtubule attachment and prevents anaphase until all kinetochores are stably attached. Checkpoint regulation hinges on the dynamic localization of checkpoint proteins to kinetochores. Unattached, checkpoint-active kinetochores accumulate multiple checkpoint proteins, which are depleted from kinetochores upon stable attachment, allowing checkpoint silencing. Because multiple proteins are recruited simultaneously to unattached kinetochores, it is not known what changes at kinetochores are essential for anaphase promoting complex/cyclosome (APC/C) inhibition. Using chemically induced dimerization to manipulate protein localization with temporal control, we show that recruiting the checkpoint protein Mad1 to metaphase kinetochores is sufficient to reactivate the checkpoint without a concomitant increase in kinetochore levels of Mps1 or BubR1. Furthermore, Mad2 binding is necessary but not sufficient for Mad1 to activate the checkpoint; a conserved C-terminal motif is also required. The results of our checkpoint reactivation assay suggest that Mad1, in addition to converting Mad2 to its active conformation, scaffolds formation of a higher-order mitotic checkpoint complex at kinetochores.

Figure 1.

Endogenous FKBP depletion improves efficiency of rapamycin-mediated recruitment. (A) Diagram of a DNA cassette used to constitutively express Mis12-GFP-FKBP and miRNA, and inducibly express mCherry-FRB. The cassette is integrated between Lox acceptor sites downstream of the EF1a promoter (Khandelia et al., 2011; for details see Materials and methods). (B) Schematic representation of rapamycin-mediated recruitment of mCherry-FRB to kinetochore-localized Mis12-GFP-FKBP. (C) HeLa cells expressing Mis12-GFP-FKBP, mCherry-FRB, and either an empty miRNA backbone or miRNA against the 3′ UTR of endogenous FKBP were imaged before and ∼1 min after the addition of 500 nM rapamycin (rap). Images are representative of three independent experiments (quantified in Fig. S1 C). Bar, 5 µm.

Figure 2.

Mad1 recruitment to metaphase kinetochores activates the mitotic checkpoint. Cells expressing FRB-mCherry-Mad1, FKBP miRNA, and either Mis12-GFP-FKBP (A, C, and D) or Mis12-FKBP (B) are shown. (A) Time lapse shows FRB-mCherry-Mad1 removal from the last mCherry-positive kinetochore pair before anaphase (at 22 min). (B) Cells were treated with rapamycin (rap) or vehicle (DMSO) control for 10 min, fixed, and stained for Mad2. Control cells, which have not recruited mCherry, were also stained for Hec1. Images shown are representative of three independent replicates. (C and D) Cells were treated with either rapamycin or vehicle at metaphase (t = 0), then monitored for anaphase onset for 30 min. (C) Images show Mad1 recruitment and metaphase arrest in rapamycin-treated but not control cells. Bars, 5 µm. (D) The graph shows the cumulative percentage of cells entering anaphase over time (n ≥ 24 cells for each condition, pooled from six independent replicates).

Figure 3.

Kinetochore levels of Mps1 and BubR1 are not affected by Mad1 recruitment at metaphase. Cells expressing Mis12-FKBP, FRB-mCherry-Mad1, and FKBP miRNA were treated with rapamycin (rap) or vehicle for 10 min, then fixed. (A) Cells were stained for BubR1 together with Hec1 to label kinetochores in control cells, which have not recruited mCherry, and BubR1 levels at kinetochores were quantified. (B) Cells were stained for Mps1, together with CENP-C to label kinetochores in control cells, and Mps1 levels at kinetochores were quantified. Bars, 5 µm. Error bars represent SEM from three independent experiments.

Figure 4.

A conserved RLK motif in the C-terminal domain of Mad1 is required to activate the checkpoint. (A) Schematic representation of Mad1 constructs. The black outline indicates the Mad2-binding region and the black bar indicates an RLK motif mutated to AAA. (B–D) Cells expressing Mis12-GFP-FKBP, FKBP miRNA, and FRB-mCherry-Mad1 fragments as indicated. (B) Cells were treated with rapamycin (rap) or vehicle for 10 min, then fixed and stained for DNA and Mad2. Images are representative of three independent experiments (quantified in Fig. S3 C). (C) Cells were treated with rapamycin at metaphase (t = 0), then monitored for anaphase onset for 30 min. Vehicle controls are shown in Fig. S2. (D) The graph shows the cumulative percentage of cells entering anaphase over time (n ≥ 26 cells for each condition, pooled from at least independent replicates). Bars, 5 µm.