Release of Mps1 from kinetochores is crucial for timely anaphase onset

Abstract

Mps1 kinase activity is required for proper chromosome segregation during mitosis through its involvements in microtubule-chromosome attachment error correction and the mitotic checkpoint. Mps1 dynamically exchanges on unattached kinetochores but is largely removed from kinetochores in metaphase. Here we show that Mps1 promotes its own turnover at kinetochores and that removal of Mps1 upon chromosome biorientation is a prerequisite for mitotic checkpoint silencing. Inhibition of Mps1 activity increases its half-time of recovery at unattached kinetochores and causes accumulation of Mps1 protein at these sites. Strikingly, preventing dissociation of active Mps1 from kinetochores delays anaphase onset despite normal chromosome attachment and alignment, and high interkinetochore tension. This delay is marked by continued recruitment of Mad1 and Mad2 to bioriented chromosomes and is attenuated by Mad2 depletion, indicating chronic engagement of the mitotic checkpoint in metaphase. We propose that release of Mps1 from kinetochores is essential for mitotic checkpoint silencing and a fast metaphase-to-anaphase transition.

Figure 1.

Mps1 accumulates on kinetochores when inhibited. (A) Immunolocalization of LAP-Mps1 in U2OS cells cotransfected with Mps1 shRNA and LAP-Mps1-WT or LAP-Mps1-KD and treated with nocodazole and MG132. Immunoblot shows expression of LAP-Mps1-WT and -KD in whole-cell lysates. (B) Immunolocalization of Mps1 and Mad2 in HeLa cells treated as indicated. Graph represents quantitation of fluorescence intensities (±SEM, 5 cells per condition, 22 kinetochores/cell). (C and D) Quantitation of fluorescence intensities at kinetochores of U2OS cells transfected and treated as indicated (±SEM, 8 cells per condition, 22 kinetochores/cell). (E) UTRM-LAP-Mps1^M602G cells were treated as indicated, and 0.81-µm² areas around single kinetochores (green squares in top panel) or in the cytoplasm (white squares) were bleached at t = 1 s. Graphs show average fluorescence intensities, shaded areas indicate SDs, and percentages indicate average recovery between 10 and 12 s. (F) Mps1 localization on bioriented and mono-oriented kinetochores in a HeLa cell treated as indicated.

Figure 2.

Tethering Mps1 to kinetochores extends metaphase. (A) Immunolocalization of LAP-Mis12-Mps1(-WT) in U2OS during indicated phases of cell cycle. (B) Percentage of mitotic (Mpm2 positive) U2OS cells that were transfected and treated as indicated, as determined by flow cytometry. Immunoblot shows levels of LAP-Mps1 and LAP-Mis12-Mps1 in total cell lysate. (C and D) Time spent in prometaphase and metaphase of transiently transfected U2OS cells with indicated plasmids and treated as indicated. Each horizontal bar represents a single cell.

Figure 3.

Sustained Mps1 activity at kinetochores does not affect microtubule attachment or chromosome biorientation. (A) Time spent in prometaphase and metaphase of U2OS cells transfected with indicated plasmids and treated as indicated. Each horizontal bar represents a single cell. (B) Stills of Video 1 showing a LAP-Mis12-Mps1-WT–expressing U2OS cell after Mps1-IN-1 addition during metaphase. Graph shows distribution of time after Mps1-IN-1 addition for cells to initiate anaphase. Average time is 12.4 min ± 3.0 min (SD), n = 45 cells. (C) Distribution of interkinetochore distances in U2OS cells transfected and treated as indicated. For LAP-Mis12-Mps1–expressing cells, only the cells showing clear Mad2 staining on kinetochores were analyzed. (D) Cold-stable microtubules in U2OS cells expressing the indicated constructs. Examples are shown for metaphases that show no, low/medium, or high Mad2 at kinetochores. Averages of absolute total spindle intensities per cell are indicated (±SD).

Figure 4.

Persistent presence of Mps1 maintains Mad1 and Mad2 on attached, bioriented kinetochores. (A and B) Mad2 localization at kinetochores of U2OS cells transiently transfected with LAP-Mps1-WT or LAP-Mis12-Mps1-WT, treated as indicated. Graph shows distribution of cells with high, low/medium, or no Mad2 signal on kinetochores. (C) Close-up of Mad2 localization at individual kinetochore pairs of U2OS cells treated and transfected as in A. (D) Mad2 intensity related to interkinetochore distance of kinetochore pairs in cells transfected and treated as in A.

Figure 5.

A model for mitotic checkpoint regulation by Mps1. Prophase: Mps1 is recruited in an Aurora B–dependent manner to unattached and/or tensionless kinetochores, where its activity is stimulated (red), possibly through dimerization. At kinetochores, Mps1 promotes error correction by enhancing Aurora B activity, ensures kinetochore binding of Mad1, inhibits checkpoint silencing mechanisms, and replenishes an interphasic pool of cytoplasmic Mps1 that stabilizes APC/C^Cdc20 inhibitory complexes. Prometaphase: After establishing Mad1 localization, Mps1 promotes kinetochore-dependent catalysis of APC/C^Cdc20 inhibitory complexes via conformational activation of Mad2, and contributes to its own removal from kinetochores. Dotted arrow indicates possibility that Mps1 contributes to maintenance of Mad1 at unattached kinetochores. The unknown details of what pools of Mps1 are dimers is represented by the question mark. Metaphase: The fast turnover of Mps1 at kinetochores allows its removal from kinetochores after stable, bioriented attachment, causing checkpoint silencing and ultimately APC/C^Cdc20 activity toward Cyclin B and Securin. Mis12-Mps1–induced prolonged metaphase: When Mps1 is not removed from kinetochores after biorientation, checkpoint silencing cannot occur. Question mark indicates the likely contribution of a cycling, and thus also cytoplasmic, pool of Mis12-Mps1 (∼60%) with the uncertainty of whether this fusion protein functions as a dimer.